GUEST EDITORIAL

Business science and evidence-based

decision making
Donald J. Rinchuse
Irwin, Pa

I
read with interest 2 excellent articles in the evidence-based decision making, dental and resident
December issue of the American Journal of Ortho- education, and foundational principles.
dontics and Dentofacial Orthopedics; that is, a
Guest Editorial by Dr Ernst M. Taeger1 (“An argument THE CONTEXTUAL ENVIRONMENT
for healthy skepticism in orthodontics”) and a Letter to
There is an alarming concern that orthodontists are
the Editor by Dr David W. Chambers2 (“Shift in
losing claim to their specialty. The landscape of ortho-
continuing education model”). Both authors argued
dontics has certainly changed and continues to evolve.
for increased attention to evidence-based science in
There is much competition for orthodontic patients
continuing education and residency training programs.
that is driven by the participation of corporate practices,
Despite the evidence-based science, the authors believe
do-it-yourself orthodontics, entrepreneurial orthodon-
there is a focus on anecdotes and weak clinical studies
tists with multiple offices, nonorthodontists doing or-
used to promote the marketing and selling of orthodon-
thodontics and in particular using clear aligners, an
tic products.
influx of new orthodontic graduates, use of 3-
Both are condemnatory of the emphasis on business
dimensional printing to manufacture appliances, mar-
science. Chambers2 writes, “Clinical studies based on bits
keting and direct mailing of appliances to customers,
of evidence, often out of context, are now being added
and so on. Much has been written about these encroach-
to continuing education programs. there is more to be
ments on the so-called traditional orthodontic practice.
made from selling products than understanding ortho-
Parenthetically, if one looks at the above-cited
dontic outcomes.” Taeger1 adds, “Has business science
competitive environment of orthodontics, with few ex-
clouded all of our ability to be perceptive evidence-
ceptions, it is orthodontist competing against orthodon-
based clinicians?”
tist. Even in corporate orthodontic practices, are not
Furthermore, Taeger1 believes that healthy skepti-
“real” orthodontists providing the treatments? “We
cism of the “fake news” involves clinicians and ortho-
have met the enemy and he is us.”3 This point has to
dontic residents developing and exercising the skill of
be considered when setting up the American Association
discernment, which relies on the ability to think critically.
of Orthodontists marketing advocacy programs that
Taeger1 said that discernment is “. the quality of being
inform the public of the benefits of receiving orthodon-
able to grasp and comprehend what is often obscure,
tic care from a certified orthodontist.
even the ability to decide between fact and error.” He
These contextual forces, as well as the significant
adds, “. lack of discernment will lead to a multitude
indebtedness of recent orthodontic graduates, have led
of sins, the consequences of which will, ultimately,
orthodontists to consider the business and economic
come to haunt us.”
side of orthodontics over the professional and service
In addition, Taeger1 claims that some orthodontists
side. Whether orthodontists work on their own, for
“.have long forgotten, or have never been taught,
others as independent contractors, as associates, or in
the time-tested fundamental principles of orthodontics,
a corporate practice, all have a heightened desire to
of which the most frequently ignored are the limitations
make money. This could lead to over treatments and
of the denture.”
attention to products, treatments, and procedures that
Let us now consider Drs Taeger1 and Chambers’2
bring in the most money, albeit at the expense of consid-
points further, as they relate to the contextual environ-
erations for science and evidence.
ment of orthodontics, sociability in orthodontics,
Am J Orthod Dentofacial Orthop 2021;160:159-62 SOCIABILITY
0889-5406/$36.00
Ó 2021 by the American Association of Orthodontists. All rights reserved. There is the human need for fellowship and camara-
https://doi.org/10.1016/j.ajodo.2021.03.011 derie, which is often displayed by belonging to a group.
159
160
This is true for orthodontists as well. Among other
things, belonging to a group gives a sense of worth. In
some instances, however, group loyalty can suppress
the thinking and reasoning of the individual. By this I
mean, there are certain tenets and principles that groups
espouse, and the followers accept. Whether it be a study
group or more formal venues such as professional orga-
nizations, associations, and societies, all those in the
group pledged allegiance to that group’s precepts. Those
who most uphold the doctrines of the group are the ones
who are most respected. Indeed, an orthodontist’s prac-
tice and treatments clearly reflect the ideas of the group,
and this is irrespective of clinical science and evidence.
EVIDENCE-BASED DECISION MAKING
Is there a gullibility pandemic going on in orthodon-
tics? If we rightfully demanded science and evidence
from our infectious disease experts during the coronavi-
rus disease 2019 pandemic, why will we not demand the
very same standards for orthodontic clinical decisions?
The nebulous, intangible notion of evidence-based sci-
ence is pitted against the visible, hyped-up, sensational-
ized convictions heralded by the orthodontic vendors,
key opinion leaders (KOLs), and their journal magazines.
These intuitively appealing notions and pseudoscience
are spun into a web of deceit by the charlatans of ortho-
dontics. Victims are typically mesmerized by an item of
fact intertwined with a sense of reasonability. Once
captivated, apocryphal notions are then espoused about
a product, regimen, theory, or teaching. The spurious
claims marketed by the pretenders for “the biggest and
the best” are without evidentiary support and focus their
assertions on “cherry picked” case reports and anecdotal
evidence. This has fostered a following devoted to fabri-
cated falsehoods. Let me be clear, not all companies and
KOL’s are guilty of advancing unproven claims to sell
products, and I would like to believe that this is the
exception rather than the rule.
Furthermore, it would be foolish to think that all or-
thodontists and graduate orthodontic residents will
follow the evidence-based decision-making teachings,
even when well delivered (ie, “I know what I like and I
like what I know!”). As Dr Johnston facetiously said,
“When everything ‘works’ and nobody dies, evidence-
based practice is for many an unnecessary elaboration
that serves only to interfere with the orderly flow of
commerce.”4
Similarly, it has been professed that the success of or-
thodontic treatment is not proof of the correctness of
treatment. That is, most orthodontic treatments are suc-
cessful to the extent that the outcomes of treatment are
an improvement over the pretreatment conditions.
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Guest editorial
However, was the treatment “correct” from the perspec-
tive of efficiency, effectiveness, and risk-reward? Related
to risk-reward, did the treatment cause any harm (ie,
decay, decalcification, periodontal disease, root resorp-
tion, worsening of the face, dehiscence, fenestration,
etc). On the reward side, did the treatment provide a
benefit in regard to stability, function, dental esthetics,
facial balance, health-related quality of life, and so
forth?
It should be pointed out that not all orthodontic clin-
ical decisions can be “evidence-based.” For some gray
areas of orthodontic clinical practice, there is little to
no evidence in which to base judgments. Some examples
of these muddled areas are early treatment, direct vs in-
direct bonding, 2-step retraction vs en masse retraction,
what constitutes good orthodontics, open bite treat-
ments, and so on. In these incidences, orthodontic clini-
cians have the freedom to choose a course of treatment
on the basis of low-level evidence, experience, or logic.
Animus and vitriol exchanges should not ensue
among orthodontic brethren over who has the best rem-
edy for clinical decisions in a murky area of orthodontic
practice. It is discomforting when practitioners express a
sense of superiority in presuming that their way is the
preeminent and only way in these uncertain areas of or-
thodontic practice (ie, “whoever does not practice the
way I do, is not as good an orthodontist as I am”). I close
this section with a quote from St Augustine related to
differing religious views, but his words are applicable
to our discussion: “In essentials unity, in non-
essentials liberty, and in all things love.”5
EDUCATION
The views promulgated in the repositories of Drs
Taeger1 and Chambers2 are certainly not new. The cry
for science and evidence-based decision making dates
back to at least the 1920s. In his 1926 report on the con-
dition of dental education in the United States and Can-
ada, William Gies6 petitioned for university-based
education over the existing forms consisting of precep-
torships and proprietary schools. He called for
university-based dental schools to foster the principles
of science and evidence.
It was hoped that the introduction of problem-based
learning (PBL) and case-based learning in the mid to
late 1980s would help to promote critical thinking in
the predoctoral dental school curricula.7,8 Critical
thinking skills developed in the predoctoral dental
school curricula would then be meliorated in the post-
doctoral dental school curricula. In theory, this was a
good initiative, but based on the way it was approached,
it fell short of expectations. To explain, dental schools’
Guest editorial
plan for the delivery of PBL was to merely add several
courses to the existing curricula that had literature re-
views and group-type interactions and titled them PBL.
To truly develop critical thinking skills in dental stu-
dents, evidence-based learning needed to be expansive
and more than adding a few PBL sessions. Dental educa-
tors needed to possess the pedagogic skills to provide
higher-level cognitive learning in every aspect of their
teaching, such as learning objectives, teaching sessions,
and evaluations (examinations), according to Bloom’s
Taxonomy of Education Objectives.9 It is not good
enough to teach for knowledge and understanding.
Dental educators must develop critical thinking skills
in students so they can apply knowledge as well as
analyzing, synthesizing, and evaluating information.9
Knowledge that cannot be applied is useless. Einstein
has been quoted as saying, “Education is that which re-
mains after one has forgotten what one has learned in
school.” It is the concepts that remain and an attitude
that allows receptivity to new ideas.
Obviously, orthodontic graduate programs have to
teach some aspect of evidence-based science and deci-
sion making as required by accreditation standards
(American Dental Association Commission on Dental
Accreditation). However, do educators teach enough of
this, and is this aspect of an orthodontist’s clinical deci-
sion making evaluated and rewarded (eg, ABO Board
Certification)? In the teaching of research and
evidence-based decision making, there needs to be a
teaching of epistemology (how knowledge is acquired),
as supported by eclectic, heuristic, and PBL. These steps
may better prepare graduates to possess the skills of
discernment and critical thinking as advocated by Dr
Taeger.1
Parenthetically, discernment can be considered by
some as a gift that some have, and others do not have.
In this regard, educators may only be able to impact
some learner’s ability to discern and to a limited extent.
In addition and contrast, it would be foolish to think that
educators can teach gullibility “out of” any student.
Common sense is not so common.
There will always be those in a society who believe
alternative facts, but this should not be the rule as it ap-
plies to a specialty. Aristotle has said, “It is the mark of an
educated mind to be able to entertain a thought without
accepting it.” As mentioned earlier, orthodontic resi-
dents, along with existing practitioners, need the knowl-
edge and skill to identify spurious claims of charlatans,
product manufacturers, those KOL’s with conflicts of in-
terest, product and treatment-oriented study groups,
op-eds, and hyped-up propaganda-type magazines.
American Journal of Orthodontics and Dentofacial Orthopedics
161
There needs to be a focus on how knowledge is acquired
which pits science against the unscientific ways knowl-
edge can be acquired, such as tenacity, rationalism, au-
thority, intuition, and empiricism.10,11 Of course, the
scientific method uses the above listed unscientific
ways in various manners in its protocol(s) but not as
the determining factor.
Dr Alex Ditmarov12 listed 10 “tongue and cheek”
ways to build wealth in orthodontic practice at the
expense of truth and professionalism. His list includes:
advertising as a nonextraction orthodontist, using a su-
perior or magical appliance, becoming a discipline of
some pseudoscience school, becoming a friend with
airway, and so on. An example of a tactic used by the
charlatan is the Galileo ploy (ie, people of Galileo’s
day mocked and laughed at him, as well as Columbus
and Copernicus). We wrote an answer to this thinking
in one of our articles, “They also laugh at clowns in
the circus.”13 Be reminded, the burden of proof for
claims of miracle products, treatments, and cures are
in the hands of those making the claims, not the critics!
Perhaps the American Association of Orthodontists,
among other considerations, could gather a team of
prominent orthodontic clinical scientists to develop
an evidence-based program that all graduate ortho-
dontic programs could use. This program should
discuss ways to recognize deceptive claims of ortho-
dontic hucksters.
FOUNDATIONAL PRINCIPLES
Dr Taeger1 suggests that orthodontists get back to
the basics and look to the foundational principles of
clinical orthodontics, the least of which is the strict limits
of the denture. As an orthodontic resident in the mid-
1970s under the tutorage of the late esteemed ortho-
dontic professor, Dr Viken Sassouni (Chair and Program
Director at the University of Pittsburgh), I was taught the
need to extract teeth to not violate the arch perimeters as
nature established for each patient; the immutability of
the arch width of the mandibular canines. We residents
needed to have study models present at the chairside
for each patient. We then had to form each patient’s
archwires to correspond to the patient’s original arch-
forms. This was done out of a straight wire, not starting
with a generic preformed archwire. The only resilient
wire we had back then for initial leveling and aligning
was Twist-a-Flex wire (not nickel-titanium wires)
made from bundled small steel strains woven into a sin-
gle wire. Even these wires had to be individually shaped.
All this in an effort to preserve each patient’s pretreat-
ment archforms.
August 2021  Vol 160  Issue 2
162
When examining a new patient with crowded teeth,
an often-asked question is, “Can I squeeze all the teeth
in with a little IPR (interproximal reduction) and some
expansion?” The answer to this question was, most
often, yes. But the more important question is, should
I? To wit, will I cause more harm than benefits as it re-
lates to stability, facial and dental esthetics, and dental
and periodontal health? After 45 years in orthodontic
practice, I can honestly say I have lamented the extrac-
tion of teeth for only a few patients. In contrast, I have
regretted not removing teeth for many, many patients.
Orthodontic educators and students alike should reflect
on the teachings of our eminent nonexpansion ortho-
dontic pioneers such as Drs Alan Brodie, Charlie Tweed,
Hayes Nance, Alton Moore, and Dick Riedel, to mention
a few.
THE FUTURE
Although the outlook for orthodontics and especially
private practice may appear bleak, there is the anticipa-
tion that we have seen our darkest moments. It has been
said, “The night is darkest just before the dawn”
(Batman-The Dark Knight, 2008). As the onset of a
new day comes forth, we have the opportunity to deter-
mine if the day will bring sun or storm. We look toward a
rainbow of hope that will guide our future. However, we
are at an inflection point of choice. On the one hand, we
can let the competing, contextual, and economic forces
of the marketplace of orthodontics play out, unhindered,
and accept the consequences as they may be. Or, on the
other hand, we can take a preemptive stand and discover
ways and strategies to tip the odds in our favor. When
the specialty is ready, orthodontic heroes will appear.
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Guest editorial
Let those in our specialty with the profiles in courage
necessary to change the direction of our specialty step
forth, and let the journey begin!
REFERENCES
1. Taeger EM. An argument for healthy skepticism in orthodontics.
Am J Orthod Dentofacial Orthop 2020;158:775-6.
2. Chambers DW. Shifts in the continuing education model. Am J Or-
thod Dentofacial Orthop 2020;158:777.
3. Kelly W. “We have met the enemy and he is us”. Available at:
https://improvegovernment.org/wp-content/uploads/2016/07/Po
sition-Paper-AIG-April-2016.pdf. Accessed April 21, 2021.
4. Johnston LE. 21st century orthodontics: when everything works,
can anything matter? Semin Orthod 2019;25:307-8.
5. Saint Augustine of Hippo. The Confessions of Saint Augustine. Salt
Lake City: Project Gutenberg; 2002.
6. Gies WJ. Dental Education in the United States and Canada. New
York: Carnegie Foundation; 1926.
7. Rinchuse DJ, Zullo T. The cognitive level demands of a dental
school’s predoctoral, didactic examinations. J Dent Educ 1986;
50:167-71.
8. Rinchuse DJ, Zullo T, Rinchuse DJ. The cognitive level demands of
the National Board Dental Examination. J Dent Educ 1987;51:
543-5.
9. Bloom BS, editor. Taxonomy of Educational Objectives. New York:
Longman; 1956.
10. Helmstadter GC. Research Concepts in Human Behavior: Educa-
tion, Psychology, Sociology. New York: Appleton-Century-Crofts;
1970.
11. Rinchuse DJ, Rinchuse DJ, Kandasamy S. Evidence-based versus
experience-based views on occlusion and TMD. Am J Orthod Den-
tofacial Orthop 2005;127:249-54.
12. Ditmarov A. How to win orthodontic patients and cash in on igno-
rance? Available at: orthodonticgrammar.com. Accessed April 21,
2021.
13. Rinchuse DJ, Sweitzer EM, Rinchuse DJ, Rinchuse DL. Understand-
ing science and evidence-based decision making in orthodontics.
Am J Orthod Dentofacial Orthop 2005;127:618-24.
READERS' FORUM

Letters to the editor

Hypoxia-inducible factor-1a may be the
first host response in orthodontic tooth
movement

A n article published in the May 2021 issue of the

American Journal of Orthodontics and Dentofacial
Orthopedics advocated how orthodontic loading leads
to activation of autophagy during orthodontic tooth
movement (OTM); however, little is known about how
it is regulated.1 The article included original research
findings and was appreciated greatly, but it could be
supplemented with some important information on the
role of the first host response taking place during
OTM, that is, hypoxia, as identified by some of the au-
thors in a recent review article.2 Similarly, others have
identified hypoxia-inducible factor-1a (HIF-1a) in
OTM.3
In the early 1990s, King et al4 published their
work from the University of Florida on OTM in rats.
Some of their histologic sections were donated to
the University of Rochester Eastman Dental Center Or-
thodontic Program, where Dr Brian McLellan, a
second-year orthodontic resident at the time, analyzed
them for HIF-1a expression using the method of
immunohistochemistry.
Figure 1 depicts the induction of HIF-1a in the spec-
imens during OTM as early as 12 hours after the applica-
tion of orthodontic force in the alveolar bone cells Fig 1. Induction of HIF-1a in the specimens during OTM
located proximal to the root apex (black staining). as early as 12 hours after the application of orthodontic
Therefore, HIF-1a is rapidly activated by orthodontic force in the alveolar bone cells located proximal to the
forces through a posttranslational mechanism (hypox- root apex (black staining).
ia-driven removal of its cognate inhibitor) and does
not require de novo synthesis like other factors involved
Stephanos Kyrkanides
in OTM. Furthermore, hypoxia-induced autophagy ap- Rochester, NY
pears to be mediated by HIF-1a in bone.5 It appears
on the basis of available literature that HIF-1a might Am J Orthod Dentofacial Orthop 2021;160:163–164
be an initiator of OTM, and autophagy, one of several 0889-5406/$36.00
Ó 2021 by the American Association of Orthodontists. All rights reserved.
processes involved in OTM induced by HIF-1a. https://doi.org/10.1016/j.ajodo.2021.04.011
Figure 2 summarizes the molecular cascade of events
that take place after the application of orthodontic
forces. Notably, the role of cyclooxygenase-2 is high- REFERENCES
lighted, which is the target of pain inhibitors often 1. Li Y, Jacox LA, Coats S, Kwon J, Xue P, Tang N, et al. Roles of auto-
used by patients, including ibuprofen, naproxen, and phagy in orthodontic tooth movement. Am J Orthod Dentofac Or-
other over-the-counter nonsteroidal antiinflammatory thop 2021;159:582-93.
drugs. In contrast, when the orthodontic pressure results 2. Li Y, Jacox LA, Little SH, Ko CC. Orthodontic tooth movement: the
biology and clinical implications. Kaohsiung J Med Sci 2018;34:
in anoxia (complete lack of oxygen), then cell death will 207-14.
ensue, and the teeth will only move through undermin- 3. Chen D, Wu L, Liu L, Gong Q, Zheng J, Peng C, et al. Comparison of
ing resorption.6 HIF1A-AS1 and HIF1A-AS2 in regulating HIF-1a and the

163
164
Fig 2. The molecular cascade of events after the application of orthodontic forces.
osteogenic differentiation of PDLCs under hypoxia. Int J Mol Med
2017;40:1529-36.
4. King GJ, Keeling SD, Wronski TJ. Histomorphometric study of alveolar
bone turnover in orthodontic tooth movement. Bone 1991;12:401-9.
5. Yellowley CE, Genetos DC. Hypoxia signaling in the skeleton: impli-
cations for bone health. Curr Osteoporos Rep 2019;17:26-35.
6. Roberts-Harry D, Sandy J. Orthodontics. Part 11: Orthodontic tooth
movement. Br Dent J 2004;196:391-4: quiz 426.
Author's response
T hank you for your letter. I agree that this is an inter-
esting subject, which can be crucial for the biology of
orthodontic tooth movement (OTM).
It has been well known for 100 years that aseptic
inflammation regulates the biological process of bone
resorption during OTM. There must exist a mechanism
that controls inflammation and the resorptive process
to prevent the pathogenesis induced by orthodontic
loading. However, factors that subside the orthodontic
inflammation to uphold homeostasis of the periodontal
tissues are rarely studied. In our article, we suggested
that autophagy may serve as such a regulatory role dur-
ing OTM. In addition, the administration of an auto-
phagy activator (eg, rapamycin) decreases tooth
movement and osteoclast recruitment in rats, suggest-
ing that autophagy could downregulate the inflamma-
tory response during OTM.1 Furthermore, Jacox et al2
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Readers' forum
demonstrated that autophagy is primarily induced in
macrophage-lineage cells (the precursor of osteoclast)
by orthodontic loading in a force-dependent manner
and plays a role during OTM.
In the early 1990s, King et al3 showed hypoxia-
inducible factor-1a (HIF-1a) was induced as early as
12 hours after the application of compressive force dur-
ing OTM. Other studies have shown that HIF-1a could
bind to hypoxia response elements of .100 genes (eg,
VEGF and BNIP3) and regulates their production.4
More evidence supports crosstalk between HIF-1a and
autophagy.5-7 Thus, the letter hypothesizes that HIF-1a
might be the initiator of OTM and autophagy. The
findings from Li et al1 showed an increase in BECN1 (a
macroautophagy gene) expression 3 days after loading
indirectly supports this hypothesis. Although this can be
a reasonable hypothesis, it would require rigorous
investigations to unveil their upstream and downstream
relationships.
Recently, the Ko laboratory (Li’s colleague) reported
that another autophagy gene (eg, optineurin [OPTN])
binds Nrf2 to regulate intracellular reactive oxygen spe-
cies and the survival of bone cells.8 Because a major role
of HIF-1a is to prevent excess mitochondrial reactive ox-
ygen species production under hypoxic conditions, it is
possible both HIF-1a and optineurin may express paral-
lel without a regulatory relationship. It is suggested that
164
Fig 2. The molecular cascade of events after the application of orthodontic forces.
osteogenic differentiation of PDLCs under hypoxia. Int J Mol Med
2017;40:1529-36.
4. King GJ, Keeling SD, Wronski TJ. Histomorphometric study of alveolar
bone turnover in orthodontic tooth movement. Bone 1991;12:401-9.
5. Yellowley CE, Genetos DC. Hypoxia signaling in the skeleton: impli-
cations for bone health. Curr Osteoporos Rep 2019;17:26-35.
6. Roberts-Harry D, Sandy J. Orthodontics. Part 11: Orthodontic tooth
movement. Br Dent J 2004;196:391-4: quiz 426.
Author's response
T hank you for your letter. I agree that this is an inter-
esting subject, which can be crucial for the biology of
orthodontic tooth movement (OTM).
It has been well known for 100 years that aseptic
inflammation regulates the biological process of bone
resorption during OTM. There must exist a mechanism
that controls inflammation and the resorptive process
to prevent the pathogenesis induced by orthodontic
loading. However, factors that subside the orthodontic
inflammation to uphold homeostasis of the periodontal
tissues are rarely studied. In our article, we suggested
that autophagy may serve as such a regulatory role dur-
ing OTM. In addition, the administration of an auto-
phagy activator (eg, rapamycin) decreases tooth
movement and osteoclast recruitment in rats, suggest-
ing that autophagy could downregulate the inflamma-
tory response during OTM.1 Furthermore, Jacox et al2
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Readers' forum
demonstrated that autophagy is primarily induced in
macrophage-lineage cells (the precursor of osteoclast)
by orthodontic loading in a force-dependent manner
and plays a role during OTM.
In the early 1990s, King et al3 showed hypoxia-
inducible factor-1a (HIF-1a) was induced as early as
12 hours after the application of compressive force dur-
ing OTM. Other studies have shown that HIF-1a could
bind to hypoxia response elements of .100 genes (eg,
VEGF and BNIP3) and regulates their production.4
More evidence supports crosstalk between HIF-1a and
autophagy.5-7 Thus, the letter hypothesizes that HIF-1a
might be the initiator of OTM and autophagy. The
findings from Li et al1 showed an increase in BECN1 (a
macroautophagy gene) expression 3 days after loading
indirectly supports this hypothesis. Although this can be
a reasonable hypothesis, it would require rigorous
investigations to unveil their upstream and downstream
relationships.
Recently, the Ko laboratory (Li’s colleague) reported
that another autophagy gene (eg, optineurin [OPTN])
binds Nrf2 to regulate intracellular reactive oxygen spe-
cies and the survival of bone cells.8 Because a major role
of HIF-1a is to prevent excess mitochondrial reactive ox-
ygen species production under hypoxic conditions, it is
possible both HIF-1a and optineurin may express paral-
lel without a regulatory relationship. It is suggested that
Readers' forum
the hypothesis should depict specific autophagy forms
such as macroautophagy, microautophagy, or mitoph-
agy, etc.
Ching-Chang Ko
Columbus, Ohio
Am J Orthod Dentofacial Orthop 2021;160:164–165
0889-5406/$36.00
Ó 2021 by the American Association of Orthodontists. All rights reserved.
https://doi.org/10.1016/j.ajodo.2021.04.012
REFERENCES
1. Li Y, Jacox LA, Coats S, Kwon J, Xue P, Tang N, et al. Roles of auto-
phagy in orthodontic tooth movement. Am J Orthod Dentofacial
Orthop 2021;159:582-93.
2. Jacox L, Tang N, Li Y, Bocklage C, Coats S, Graves C, et al. Orthodon-
tic loading activates cell-specific autophagy in a force-dependent
manner. Am J Orthod Dentofac Orthop 2021: In press.
3. King GJ, Keeling SD, Wronski TJ. Histomorphometric study of alve-
olar bone turnover in orthodontic tooth movement. Bone 1991;12:
401-9.
4. Bellot G, Garcia-Medina R, Gounon P, Chiche J, Roux D,
Pouyss
egur J, et al. Hypoxia-induced autophagy is mediated
through hypoxia-inducible factor induction of BNIP3 and BNIP3L
via their BH3 domains. Mol Cell Biol 2009;29:2570-81.
5. Qu€aschling T, Friedrich D, Deepe GS Jr, Rupp J. Crosstalk Between
autophagy and hypoxia-inducible factor-1a in antifungal immu-
nity. Cells 2020;9:2150.
6. Chen D, Wu L, Liu L, Gong Q, Zheng J, Peng C, et al. Comparison of
HIF1A-AS1 and HIF1A-AS2 in regulating HIF-1a and the osteo-
genic differentiation of PDLCs under hypoxia. Int J Mol Med
2017;40:1529-36.
7. Zhang H, Bosch-Marce M, Shimoda LA, Tan YS, Baek JH, Wesley JB,
et al. Mitochondrial autophagy is an HIF-1-dependent adaptive
metabolic response to hypoxia. J Biol Chem 2008;283:10892-903.
8. Xue P, Hu X, Chang E, Wang L, Chen M, Wu T-H, et al. Deficiency of
optineurin enhances osteoclast differentiation by attenuating NRF-
2 mediated antioxidant response. Exp Mol Med 2021;53:667-80.
Correcting skeletal open bite with clear
aligners and miniscrews
I read with a great deal of interest the case report by
Pinho and Santos in the February issue of the Amer-
ican Journal of Orthodontics and Dentofacial Orthope-
dics.1 It was most impressive how they were able to close
such a severe skeletal open bite without surgery! As a
long-time user of Invisalign, I, too, have often found it
a better way to close open bite malocclusions than tradi-
tional orthodontic treatment, but I was never able to
achieve such a profound result as this case report
demonstrated.
As the orthodontic specialty care director at the Mis-
souri School of Dentistry and Oral Health, one of my re-
sponsibilities is to educate the dental students on how to
evaluate orthodontic treatment. To this end, I teach
American Journal of Orthodontics and Dentofacial Orthopedics
165
them how to do cephalometric tracings, and I was look-
ing forward to using this case as an example of the cor-
rect way to close a skeletal open bite, that is, posterior
intrusion rather than anterior extrusion.
Evaluating the information presented by the authors,
I, unfortunately, have to take issue with their findings as
to the mechanism of the open bite correction. In their
conclusion, they state: “The occlusal plan control with
mandibular—molar intrusion allowed a counterclock-
wise rotation the mandible to permit bite closure.” I do
not think that the superimposition cephalometric radio-
graphs shown in Figure 14 justify this conclusion. I
believe that the mandibular border outline is, within
the margin of tracing error, almost perfectly superim-
posed between the before and after cephalometric
tracings. This would indicate that there was no counter-
clockwise rotation of the mandible. Instead, looking at
the before and after the position of the incisors, it is
evident that there was a large degree of extrusion (espe-
cially the mandibular incisors).
Given the cost and trauma of orthognathic surgery,
the treatment performed was certainly appropriate and
successful. Nonetheless, it is important to acknowledge
the actual mechanism of the correction and not to
delude ourselves that temporary anchorage devices are
capable of producing the same counterclockwise open
bite correction that is typically achieved with orthog-
nathic surgery.
Robert Waxler
St. Louis, Mo
Am J Orthod Dentofacial Orthop 2021;160:165
0889-5406/$36.00
Ó 2021 by the American Association of Orthodontists. All rights reserved.
https://doi.org/10.1016/j.ajodo.2021.04.010
REFERENCE
1. Pinho T, Santos M. Skeletal open bite treated with clear aligners and
miniscrews. Am J Orthod Dentofacial Orthop 2021;159:224-33.
Author's response
W e appreciate our colleague's comments and agree
that aligners are a powerful tool for open bite
correction. However, in patients with skeletal open bite
with high magnitude, as in the case report we presented,
we are sure to program the concomitant use of
auxiliaries, namely miniscrews.
We reinforce what has already been explained in the
article, in which the purpose was not to convey the
message that everything is solved with camouflage using
miniscrews. However, in this specific complex case, as
August 2021  Vol 160  Issue 2
Readers' forum
the hypothesis should depict specific autophagy forms
such as macroautophagy, microautophagy, or mitoph-
agy, etc.
Ching-Chang Ko
Columbus, Ohio
Am J Orthod Dentofacial Orthop 2021;160:165
0889-5406/$36.00
Ó 2021.
https://doi.org/10.1016/j.ajodo.2021.04.012
REFERENCES
1. Li Y, Jacox LA, Coats S, Kwon J, Xue P, Tang N, et al. Roles of auto-
phagy in orthodontic tooth movement. Am J Orthod Dentofacial
Orthop 2021;159:582-93.
2. Jacox L, Tang N, Li Y, Bocklage C, Coats S, Graves C, et al. Orthodon-
tic loading activates cell-specific autophagy in a force-dependent
manner. Am J Orthod Dentofac Orthop 2021: In press.
3. King GJ, Keeling SD, Wronski TJ. Histomorphometric study of alve-
olar bone turnover in orthodontic tooth movement. Bone 1991;12:
401-9.
4. Bellot G, Garcia-Medina R, Gounon P, Chiche J, Roux D,
Pouyss
egur J, et al. Hypoxia-induced autophagy is mediated
through hypoxia-inducible factor induction of BNIP3 and BNIP3L
via their BH3 domains. Mol Cell Biol 2009;29:2570-81.
5. Qu€aschling T, Friedrich D, Deepe GS Jr, Rupp J. Crosstalk Between
autophagy and hypoxia-inducible factor-1a in antifungal immu-
nity. Cells 2020;9:2150.
6. Chen D, Wu L, Liu L, Gong Q, Zheng J, Peng C, et al. Comparison of
HIF1A-AS1 and HIF1A-AS2 in regulating HIF-1a and the osteo-
genic differentiation of PDLCs under hypoxia. Int J Mol Med
2017;40:1529-36.
7. Zhang H, Bosch-Marce M, Shimoda LA, Tan YS, Baek JH, Wesley JB,
et al. Mitochondrial autophagy is an HIF-1-dependent adaptive
metabolic response to hypoxia. J Biol Chem 2008;283:10892-903.
8. Xue P, Hu X, Chang E, Wang L, Chen M, Wu T-H, et al. Deficiency of
optineurin enhances osteoclast differentiation by attenuating NRF-
2 mediated antioxidant response. Exp Mol Med 2021;53:667-80.
Correcting skeletal open bite with clear
aligners and miniscrews
I read with a great deal of interest the case report by
Pinho and Santos in the February issue of the Amer-
ican Journal of Orthodontics and Dentofacial Orthope-
dics.1 It was most impressive how they were able to close
such a severe skeletal open bite without surgery! As a
long-time user of Invisalign, I, too, have often found it
a better way to close open bite malocclusions than tradi-
tional orthodontic treatment, but I was never able to
achieve such a profound result as this case report
demonstrated.
As the orthodontic specialty care director at the Mis-
souri School of Dentistry and Oral Health, one of my re-
sponsibilities is to educate the dental students on how to
evaluate orthodontic treatment. To this end, I teach
American Journal of Orthodontics and Dentofacial Orthopedics
165
them how to do cephalometric tracings, and I was look-
ing forward to using this case as an example of the cor-
rect way to close a skeletal open bite, that is, posterior
intrusion rather than anterior extrusion.
Evaluating the information presented by the authors,
I, unfortunately, have to take issue with their findings as
to the mechanism of the open bite correction. In their
conclusion, they state: “The occlusal plan control with
mandibular—molar intrusion allowed a counterclock-
wise rotation the mandible to permit bite closure.” I do
not think that the superimposition cephalometric radio-
graphs shown in Figure 14 justify this conclusion. I
believe that the mandibular border outline is, within
the margin of tracing error, almost perfectly superim-
posed between the before and after cephalometric
tracings. This would indicate that there was no counter-
clockwise rotation of the mandible. Instead, looking at
the before and after the position of the incisors, it is
evident that there was a large degree of extrusion (espe-
cially the mandibular incisors).
Given the cost and trauma of orthognathic surgery,
the treatment performed was certainly appropriate and
successful. Nonetheless, it is important to acknowledge
the actual mechanism of the correction and not to
delude ourselves that temporary anchorage devices are
capable of producing the same counterclockwise open
bite correction that is typically achieved with orthog-
nathic surgery.
Robert Waxler
St. Louis, Mo
Am J Orthod Dentofacial Orthop 2021;160:165
0889-5406/$36.00
Ó 2021 by the American Association of Orthodontists. All rights reserved.
https://doi.org/10.1016/j.ajodo.2021.04.010
REFERENCE
1. Pinho T, Santos M. Skeletal open bite treated with clear aligners and
miniscrews. Am J Orthod Dentofacial Orthop 2021;159:224-33.
Author's response
W e appreciate our colleague's comments and agree
that aligners are a powerful tool for open bite
correction. However, in patients with skeletal open bite
with high magnitude, as in the case report we presented,
we are sure to program the concomitant use of
auxiliaries, namely miniscrews.
We reinforce what has already been explained in the
article, in which the purpose was not to convey the
message that everything is solved with camouflage using
miniscrews. However, in this specific complex case, as
August 2021  Vol 160  Issue 2
Readers' forum
the hypothesis should depict specific autophagy forms
such as macroautophagy, microautophagy, or mitoph-
agy, etc.
Ching-Chang Ko
Columbus, Ohio
Am J Orthod Dentofacial Orthop 2021;160:165
0889-5406/$36.00
Ó 2021.
https://doi.org/10.1016/j.ajodo.2021.04.012
REFERENCES
1. Li Y, Jacox LA, Coats S, Kwon J, Xue P, Tang N, et al. Roles of auto-
phagy in orthodontic tooth movement. Am J Orthod Dentofacial
Orthop 2021;159:582-93.
2. Jacox L, Tang N, Li Y, Bocklage C, Coats S, Graves C, et al. Orthodon-
tic loading activates cell-specific autophagy in a force-dependent
manner. Am J Orthod Dentofac Orthop 2021: In press.
3. King GJ, Keeling SD, Wronski TJ. Histomorphometric study of alve-
olar bone turnover in orthodontic tooth movement. Bone 1991;12:
401-9.
4. Bellot G, Garcia-Medina R, Gounon P, Chiche J, Roux D,
Pouyss
egur J, et al. Hypoxia-induced autophagy is mediated
through hypoxia-inducible factor induction of BNIP3 and BNIP3L
via their BH3 domains. Mol Cell Biol 2009;29:2570-81.
5. Qu€aschling T, Friedrich D, Deepe GS Jr, Rupp J. Crosstalk Between
autophagy and hypoxia-inducible factor-1a in antifungal immu-
nity. Cells 2020;9:2150.
6. Chen D, Wu L, Liu L, Gong Q, Zheng J, Peng C, et al. Comparison of
HIF1A-AS1 and HIF1A-AS2 in regulating HIF-1a and the osteo-
genic differentiation of PDLCs under hypoxia. Int J Mol Med
2017;40:1529-36.
7. Zhang H, Bosch-Marce M, Shimoda LA, Tan YS, Baek JH, Wesley JB,
et al. Mitochondrial autophagy is an HIF-1-dependent adaptive
metabolic response to hypoxia. J Biol Chem 2008;283:10892-903.
8. Xue P, Hu X, Chang E, Wang L, Chen M, Wu T-H, et al. Deficiency of
optineurin enhances osteoclast differentiation by attenuating NRF-
2 mediated antioxidant response. Exp Mol Med 2021;53:667-80.
Correcting skeletal open bite with clear
aligners and miniscrews
I read with a great deal of interest the case report by
Pinho and Santos in the February issue of the Amer-
ican Journal of Orthodontics and Dentofacial Orthope-
dics.1 It was most impressive how they were able to close
such a severe skeletal open bite without surgery! As a
long-time user of Invisalign, I, too, have often found it
a better way to close open bite malocclusions than tradi-
tional orthodontic treatment, but I was never able to
achieve such a profound result as this case report
demonstrated.
As the orthodontic specialty care director at the Mis-
souri School of Dentistry and Oral Health, one of my re-
sponsibilities is to educate the dental students on how to
evaluate orthodontic treatment. To this end, I teach
American Journal of Orthodontics and Dentofacial Orthopedics
165
them how to do cephalometric tracings, and I was look-
ing forward to using this case as an example of the cor-
rect way to close a skeletal open bite, that is, posterior
intrusion rather than anterior extrusion.
Evaluating the information presented by the authors,
I, unfortunately, have to take issue with their findings as
to the mechanism of the open bite correction. In their
conclusion, they state: “The occlusal plan control with
mandibular—molar intrusion allowed a counterclock-
wise rotation the mandible to permit bite closure.” I do
not think that the superimposition cephalometric radio-
graphs shown in Figure 14 justify this conclusion. I
believe that the mandibular border outline is, within
the margin of tracing error, almost perfectly superim-
posed between the before and after cephalometric
tracings. This would indicate that there was no counter-
clockwise rotation of the mandible. Instead, looking at
the before and after the position of the incisors, it is
evident that there was a large degree of extrusion (espe-
cially the mandibular incisors).
Given the cost and trauma of orthognathic surgery,
the treatment performed was certainly appropriate and
successful. Nonetheless, it is important to acknowledge
the actual mechanism of the correction and not to
delude ourselves that temporary anchorage devices are
capable of producing the same counterclockwise open
bite correction that is typically achieved with orthog-
nathic surgery.
Robert Waxler
St. Louis, Mo
Am J Orthod Dentofacial Orthop 2021;160:165
0889-5406/$36.00
Ó 2021 by the American Association of Orthodontists. All rights reserved.
https://doi.org/10.1016/j.ajodo.2021.04.010
REFERENCE
1. Pinho T, Santos M. Skeletal open bite treated with clear aligners and
miniscrews. Am J Orthod Dentofacial Orthop 2021;159:224-33.
Author's response
W e appreciate our colleague's comments and agree
that aligners are a powerful tool for open bite
correction. However, in patients with skeletal open bite
with high magnitude, as in the case report we presented,
we are sure to program the concomitant use of
auxiliaries, namely miniscrews.
We reinforce what has already been explained in the
article, in which the purpose was not to convey the
message that everything is solved with camouflage using
miniscrews. However, in this specific complex case, as
August 2021  Vol 160  Issue 2
166
€rk general superimposition; B, Bjo
Fig. A, Bjo €rk maxillary
€rk mandibular superimposition.
superimposition; C, Bjo
the superimposition and clinical results revealed, there
was a high posterior vertical control, which contributed
to the success achieved.
Having in mind the comments on the cephalometric
superimposition, we considered it pertinent to present a
general superimposition (Fig A) with reference planes, as
well as to present the maxillary (Fig B) and mandibular
(Fig C) superimpositions, which are undoubtedly an
asset for the answer to the presented doubts. Regarding
the total superimposition (Fig A), a counterclockwise
rotation of the mandible is observed, with a decrease
in the FMA. Despite the omission in the article, we can
also prove it with a counterclockwise rotation of the
mandible through the measurements of the occlusal
plane that went from the initial 6.8 to 5.5 . Admittedly,
we should have included the value of this measure, as it
would be an important supporting factor.
Considering the new opportunity to explain the
obtained results, and because it is omitted in the article,
in the maxillary (Fig B) and mandibular (Fig C)
superimposition areas, we can prove that at the
mandibular level, a slight first molar intrusion was
made, which in this patient with the magnitude of
hyperdivergence presented was undoubtedly an asset
for vertical control. As our colleague mentions, this is a
patient with an increased complexity degree for most of
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Readers' forum
the traditional orthodontic treatments; however, with
the present technical system (aligners and miniscrews),
the success of orthodontic treatment was increased.
In the superimposition of the maxilla (with the palate
superimposition) (Fig B), a significant intrusion of the
maxillary first molars was observed, perhaps being the
factor that most contributed to a counterclockwise
rotation of the mandible.
To further reinforce our conclusion about the
counterclockwise rotation of the mandible, in
Merrifield's Z angle, an increase of the angle from 67
to 70 can be observed. It should also be noted that the
increase in the IMPA (which went from 87.7 to 93 ) would
aggravate the Z angle but was not observed, further
reinforcing the occurrence of counterclockwise rotation.
In the general superimposition (Fig A), considering the
counterclockwise rotation of the mandible, it gives an
illusion of excessive extrusion of the mandibular incisors,
which when evaluated in the mandibular super-
imposition, is not confirmed, being observed a slight
extrusion, but not as relevant as referred by our colleague.
In the maxilla superimposition, it was verified relative
extrusion, not pure extrusion, as mentioned in the
manuscript.
In conclusion, the success of this treatment was due
to the posterior vertical control. Considering the initial
open bite magnitude, a minimal extrusion of the incisors
reinforces the occurrence of counterclockwise rotation
of the mandible as a major contribution to close the
open bite. In addition to the cephalometric super-
impositions that are undoubtedly a scientifically proven
asset, we have a clinical improvement at the dental and
facial level, namely in the lip competence that reinforces
what was explained in this answer.
Teresa Pinho
Paredes, Portugal
Am J Orthod Dentofacial Orthop 2021;160:165–166
0889-5406/$36.00
Ó 2021.
https://doi.org/10.1016/j.ajodo.2021.04.007
Skeletal open bite treated with clear
aligners and miniscrews
A case report published in February 2021 by Teresa
Pinho and Mariana Santos describes the treatment
of a 16-year-old female patient with a skeletal open
bite and temporomandibular dysfunction (Pinho T,
Santos M. Skeletal open bite treated with clearaligners
and miniscrews Am J Orthod Dentofacial Orthop
2021;159:224-33).
166
€rk general superimposition; B, Bjo
Fig. A, Bjo €rk maxillary
€rk mandibular superimposition.
superimposition; C, Bjo
the superimposition and clinical results revealed, there
was a high posterior vertical control, which contributed
to the success achieved.
Having in mind the comments on the cephalometric
superimposition, we considered it pertinent to present a
general superimposition (Fig A) with reference planes, as
well as to present the maxillary (Fig B) and mandibular
(Fig C) superimpositions, which are undoubtedly an
asset for the answer to the presented doubts. Regarding
the total superimposition (Fig A), a counterclockwise
rotation of the mandible is observed, with a decrease
in the FMA. Despite the omission in the article, we can
also prove it with a counterclockwise rotation of the
mandible through the measurements of the occlusal
plane that went from the initial 6.8 to 5.5 . Admittedly,
we should have included the value of this measure, as it
would be an important supporting factor.
Considering the new opportunity to explain the
obtained results, and because it is omitted in the article,
in the maxillary (Fig B) and mandibular (Fig C)
superimposition areas, we can prove that at the
mandibular level, a slight first molar intrusion was
made, which in this patient with the magnitude of
hyperdivergence presented was undoubtedly an asset
for vertical control. As our colleague mentions, this is a
patient with an increased complexity degree for most of
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Readers' forum
the traditional orthodontic treatments; however, with
the present technical system (aligners and miniscrews),
the success of orthodontic treatment was increased.
In the superimposition of the maxilla (with the palate
superimposition) (Fig B), a significant intrusion of the
maxillary first molars was observed, perhaps being the
factor that most contributed to a counterclockwise
rotation of the mandible.
To further reinforce our conclusion about the
counterclockwise rotation of the mandible, in
Merrifield's Z angle, an increase of the angle from 67
to 70 can be observed. It should also be noted that the
increase in the IMPA (which went from 87.7 to 93 ) would
aggravate the Z angle but was not observed, further
reinforcing the occurrence of counterclockwise rotation.
In the general superimposition (Fig A), considering the
counterclockwise rotation of the mandible, it gives an
illusion of excessive extrusion of the mandibular incisors,
which when evaluated in the mandibular super-
imposition, is not confirmed, being observed a slight
extrusion, but not as relevant as referred by our colleague.
In the maxilla superimposition, it was verified relative
extrusion, not pure extrusion, as mentioned in the
manuscript.
In conclusion, the success of this treatment was due
to the posterior vertical control. Considering the initial
open bite magnitude, a minimal extrusion of the incisors
reinforces the occurrence of counterclockwise rotation
of the mandible as a major contribution to close the
open bite. In addition to the cephalometric super-
impositions that are undoubtedly a scientifically proven
asset, we have a clinical improvement at the dental and
facial level, namely in the lip competence that reinforces
what was explained in this answer.
Teresa Pinho
Paredes, Portugal
Am J Orthod Dentofacial Orthop 2021;160:166
0889-5406/$36.00
Ó 2021.
https://doi.org/10.1016/j.ajodo.2021.04.007
Skeletal open bite treated with clear
aligners and miniscrews
A case report published in February 2021 by Teresa
Pinho and Mariana Santos describes the treatment
of a 16-year-old female patient with a skeletal open
bite and temporomandibular dysfunction (Pinho T,
Santos M. Skeletal open bite treated with clearaligners
and miniscrews Am J Orthod Dentofacial Orthop
2021;159:224-33).
Readers' forum
The case report was very informative, innovative, and
of great knowledge, but we have some questions.
The authors mentioned that they observed Class III
molar tendency bilaterally (p 224), but Figure 1 clearly
depicts a Class I molar relationship. Furthermore,
Figure 2 shows an end-on molar relationship bilaterally.
Kindly clarify the molar relationship of the patient.
In the diagnosis and etiology section, the authors
mention that the patient had labial incompetence
because of increased overjet. However, labial compe-
tence is seen in extraoral pretreatment photographs
(Fig 1).
Finally, the authors stated that mandibular posterior
intrusion was done with the use of a mandibular buccal
shelf implant, which led to counterclockwise rotation of
the mandible. The Table shows a decrease in FMA, from
38 to 32.1 , but the superimposition shows no evident
change. Can the authors comment on the posterior
intrusion in aligner therapy because of the bite-block
effect, which is also a confounding factor?
Yash Agarwal
Manish Goyal
Mukesh Kumar
Amandeep Kaur
Jharkhand and Uttar Pradesh, India
Am J Orthod Dentofacial Orthop 2021;160:166–167
0889-5406/$36.00
Ó 2021 by the American Association of Orthodontists. All rights reserved.
https://doi.org/10.1016/j.ajodo.2021.05.005
Author’s response
T hank you for your letter. We addressed some of these
issues in our author’s response to Dr Waxler and will
try to answer your questions here.
We referred to a bilateral Class III tendency on
maximum intercuspation with the forced condyles in
an anomalous position, not a true Class III relationship.
As Figure 1 showed, the mesiobuccal cusp of the maxil-
lary molar is slightly behind the buccal groove of the
mandibular molar. However, the dental relationship
became a bilateral Class II in centric condyle, which cor-
responds to a nonforced reference condyle position. We
cannot forget that this patient had a temporomandib-
ular disorder, with a high, different position between
maximum intercuspation and centric relation.
Although the patient's lips are closed in Figure 1, the
pretreatment photograph, this position is clearly forced;
you can see the evident roughness of the chin (the so-
called orange peel menton), typical of labial incompe-
tence because of forced lip contact.
American Journal of Orthodontics and Dentofacial Orthopedics
167
Regarding posterior intrusion in aligner therapy, clin-
ically, if we compare Figures 7 and 8, which show the
aligners inserted before and after the posterior intrusion,
we can observe that the molars have intruded with an
evident step with adjacent premolars in which a pro-
gramed intrusion was also contemplated, because of
the initial occlusal contacts on premolars. This fact is
corroborated by the photographs in Figure 9, after use
of the number 25 aligner, 7 months after the start of
treatment. Until aligner number 25, no true extrusion
was programed on the canines and incisors, the similar
step between these teeth on the maxillary and mandib-
ular arch, that can be confirmed comparing Figures 7
and 8. We addressed this question in a previous author's
response (to the letter from Dr Waxler) with new general,
maxillary and mandibular superpositions. Having in
mind the comments on the cephalometric superimposi-
tion, we considered it pertinent to present a general su-
perimposition with reference plans, as well as to present
the maxillary and mandibular superimpositions, which
are undoubtedly an asset for the answer to the presented
doubts. Regarding the total superimposition, a counter-
clockwise rotation of the mandible is observed, with a
decrease in the FMA angle. Despite the omission in the
article, we can also prove it with counterclockwise rota-
tion of the mandible through the measurements of the
occlusal plane that went from the initial 6.8 to 5.5 .
We should have reported the value of this measure in
the article, as it was an important supporting factor.
The new maxillary and mandibular superimposition
show that at the mandibular level, a slight first molar
intrusion was made, which in this patient, with the
magnitude of hyperdivergence presented, was undoubt-
edly an asset for vertical control. This was a patient with
an increased degree of complexity for most of the tradi-
tional orthodontic treatments; however, with the present
technical system (aligners and miniscrews), the success
of orthodontic treatment was increased.
The superimposition of the maxilla shows a signifi-
cant intrusion of the maxillary first molars, perhaps
being the factor that most contributed to a counter-
clockwise rotation of the mandible.
To further reinforce our conclusion about the coun-
terclockwise rotation of the mandible, in Merrifield's z
angle, an increase of the angle from 67 to 70 can be
observed. It should also be noted that the increase in
the IMPA (which went from 87.7 to 93 ) would aggra-
vate the z angle but was not observed, further reinforc-
ing the occurrence of counterclockwise rotation. The
general superimposition, considering the counterclock-
wise rotation of the mandible, gives an illusion of
August 2021  Vol 160  Issue 2
Readers' forum
The case report was very informative, innovative, and
of great knowledge, but we have some questions.
The authors mentioned that they observed Class III
molar tendency bilaterally (p 224), but Figure 1 clearly
depicts a Class I molar relationship. Furthermore,
Figure 2 shows an end-on molar relationship bilaterally.
Kindly clarify the molar relationship of the patient.
In the diagnosis and etiology section, the authors
mention that the patient had labial incompetence
because of increased overjet. However, labial compe-
tence is seen in extraoral pretreatment photographs
(Fig 1).
Finally, the authors stated that mandibular posterior
intrusion was done with the use of a mandibular buccal
shelf implant, which led to counterclockwise rotation of
the mandible. The Table shows a decrease in FMA, from
38 to 32.1 , but the superimposition shows no evident
change. Can the authors comment on the posterior
intrusion in aligner therapy because of the bite-block
effect, which is also a confounding factor?
Yash Agarwal
Manish Goyal
Mukesh Kumar
Amandeep Kaur
Jharkhand and Uttar Pradesh, India
Am J Orthod Dentofacial Orthop 2021;160:167
0889-5406/$36.00
Ó 2021.
https://doi.org/10.1016/j.ajodo.2021.05.005
Author’s response
T hank you for your letter. We addressed some of these
issues in our author’s response to Dr Waxler and will
try to answer your questions here.
We referred to a bilateral Class III tendency on
maximum intercuspation with the forced condyles in
an anomalous position, not a true Class III relationship.
As Figure 1 showed, the mesiobuccal cusp of the maxil-
lary molar is slightly behind the buccal groove of the
mandibular molar. However, the dental relationship
became a bilateral Class II in centric condyle, which cor-
responds to a nonforced reference condyle position. We
cannot forget that this patient had a temporomandib-
ular disorder, with a high, different position between
maximum intercuspation and centric relation.
Although the patient's lips are closed in Figure 1, the
pretreatment photograph, this position is clearly forced;
you can see the evident roughness of the chin (the so-
called orange peel menton), typical of labial incompe-
tence because of forced lip contact.
American Journal of Orthodontics and Dentofacial Orthopedics
167
Regarding posterior intrusion in aligner therapy, clin-
ically, if we compare Figures 7 and 8, which show the
aligners inserted before and after the posterior intrusion,
we can observe that the molars have intruded with an
evident step with adjacent premolars in which a pro-
gramed intrusion was also contemplated, because of
the initial occlusal contacts on premolars. This fact is
corroborated by the photographs in Figure 9, after use
of the number 25 aligner, 7 months after the start of
treatment. Until aligner number 25, no true extrusion
was programed on the canines and incisors, the similar
step between these teeth on the maxillary and mandib-
ular arch, that can be confirmed comparing Figures 7
and 8. We addressed this question in a previous author's
response (to the letter from Dr Waxler) with new general,
maxillary and mandibular superpositions. Having in
mind the comments on the cephalometric superimposi-
tion, we considered it pertinent to present a general su-
perimposition with reference plans, as well as to present
the maxillary and mandibular superimpositions, which
are undoubtedly an asset for the answer to the presented
doubts. Regarding the total superimposition, a counter-
clockwise rotation of the mandible is observed, with a
decrease in the FMA angle. Despite the omission in the
article, we can also prove it with counterclockwise rota-
tion of the mandible through the measurements of the
occlusal plane that went from the initial 6.8 to 5.5 .
We should have reported the value of this measure in
the article, as it was an important supporting factor.
The new maxillary and mandibular superimposition
show that at the mandibular level, a slight first molar
intrusion was made, which in this patient, with the
magnitude of hyperdivergence presented, was undoubt-
edly an asset for vertical control. This was a patient with
an increased degree of complexity for most of the tradi-
tional orthodontic treatments; however, with the present
technical system (aligners and miniscrews), the success
of orthodontic treatment was increased.
The superimposition of the maxilla shows a signifi-
cant intrusion of the maxillary first molars, perhaps
being the factor that most contributed to a counter-
clockwise rotation of the mandible.
To further reinforce our conclusion about the coun-
terclockwise rotation of the mandible, in Merrifield's z
angle, an increase of the angle from 67 to 70 can be
observed. It should also be noted that the increase in
the IMPA (which went from 87.7 to 93 ) would aggra-
vate the z angle but was not observed, further reinforc-
ing the occurrence of counterclockwise rotation. The
general superimposition, considering the counterclock-
wise rotation of the mandible, gives an illusion of
August 2021  Vol 160  Issue 2
168
excessive extrusion of the mandibular incisors, which
when evaluated in the mandibular superimposition, is
not confirmed, being observed a slight extrusion, but
not as relevant as referred by the colleague. The maxil-
lary superimposition verified relative extrusion, not
pure extrusion, as mentioned in the manuscript. In
conclusion, the success of this treatment was due to
the posterior vertical control. Considering the initial
open bite magnitude, a minimal extrusion of the inci-
sors, reinforces the occurrence of counterclockwise rota-
tion of the mandible as a major contribution to close the
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Readers' forum
open bite. In addition to the cephalometric superimpo-
sitions that are undoubtedly a scientifically proven asset,
we have a clinical improvement at the dental and facial
level, namely in the lip competence that reinforces what
was explained in this answer.
Teresa Pinho
Paredes, Portugal
Am J Orthod Dentofacial Orthop 2021;160:167–168
0889-5406/$36.00
Ó 2021.
https://doi.org/10.1016/j.ajodo.2021.05.006
ETHICS IN ORTHODONTICS
The foundation
Peter M. Greco
Philadelphia, Pa
M
rs Bell is typical of the adult patients that find
their way to your practice. Many are wary and
critical when they arrive, and a few even
display a bit of an edge when they meet you. Your recep-
tionist warns you that Mrs Bell is no different. From her
first phone call, she needs a bit of coddling to satisfy her
requirements for an appointment time. She swiftly ques-
tions your receptionist to determine if you “run on time.”
She chooses to leave many questions in her medical his-
tory unanswered. And in reply to the question that asks if
she was dissatisfied with any previous treatment, she an-
swers affirmatively without an explanation.
She appears at your office promptly and is seated
immediately but is less than cordial with your staff as
she settles into the operatory. As you introduce yourself,
you sense a slight air of belligerence as she asks if you
received and reviewed her x-rays from her dentist. You
immediately ask yourself if she might be a patient with
whom you'd prefer not to get involved. But you know
the high-quality dental group who referred her for pre-
prosthetic orthodontic preparation and figure that if
you had any doubts after seeing her, you could seek their
input about the potential of satisfying her orthodontic
needs.
After a short chat about her orthodontic concerns
and your subsequent explanation of the objectives of
the visit, the examination goes smoothly. You thor-
oughly outline your role in her rehabilitation and
patiently answer a myriad of the well-grounded ques-
tions she had compiled. You request she contemplate
your treatment recommendations and call should she
have any future questions. As she rises from the chair,
she smiles broadly and tells you that your examination
and explanation were the most thorough and compre-
hensive that she has ever had.
Trust is the foundation of our specialty. Simply think
of the icons that created the binding unity that led to the
prestigious legacy of this specialty's membership. Akin to
our trust of those who accepted and mentored us, so too,
have we sustained that commitment to generate trust as
we communicate with our patients, residents, and
Am J Orthod Dentofacial Orthop 2021;160:169
0889-5406/$36.00
Ó 2021 by the American Association of Orthodontists. All rights reserved.
https://doi.org/10.1016/j.ajodo.2021.05.001
colleagues. Because actions truly speak louder than
words, our follow-through in any of our promises is
essential to establishing trust in our reputation, both
individually and collectively. Our reverence for trust
compels us to become part of a consistent mission—
whether as clinicians, academicians, or residents—by
creating the common purpose and objective to do
good. Trust is a 2-way street with 1 lane based on trans-
parency and the other on competency.1
The establishment of trust in an orthodontist differs
from that of trust in friends. Think of the time it took to
develop a cherished friendship in which mutual trust
abounds. Contrast this to a potential patient who might
make the commitment to time and expense within mi-
nutes of meeting us. Then compound that leap with
the inequity of knowledge between our patient and us,
and we find a patient in a unique position of vulnera-
bility.2
We cannot and should not sell orthodontic therapy.
What we provide is vastly elective. But we can establish
sufficient confidence and trust to generate the motiva-
tion for patients to eagerly enlist in our care. Despite
all the tricks we read from the practice gurus and trade
magazines, patients and parents—especially the discrim-
inating ones—will proceed with treatment on the basis of
trust that they cultivate from our understanding and
appreciation of their history, needs, and desires.
Just 1 day after your initial examination, Mrs Bell
calls back and wants to know how soon she can begin
treatment with you. She declares, “I'm totally on board!”
and pledges to follow any directives you prescribe for
her.
It's been said we vote with our feet. No doubt trust
had a role in Mrs Bell's ballot.
REFERENCES
1. Jaffe D. The essential importance of trust: how to build it or restore
it. Available at: forbes.com/sites/dennisjaffe/2018/12/05/the-
essential-importance-of-trust-how-to-build-it-or-restore-it/?sh5
e42cfc464fe5. Accessed April 23, 2021.
2. Rule J, Veatch RM. Chapter 7. Fidelity: obligations of trust and
confidentiality. In: Bywaters L, Harmon L, editors. Ethical Questions
In Dentistry. 2nd ed. Chicago: Quintessence Publishing; 2004. p.
125.
169
SYSTEMATIC REVIEW

Machine learning and orthodontics,

current trends and the future
opportunities: A scoping review
Hossein Mohammad-Rahimi,a Mohadeseh Nadimi,b Mohammad Hossein Rohban,a Erfan Shamsoddin,c
Victor Y. Lee,d and Saeed Reza Motamediane
Tehran, Iran, and New York, NY

Introduction: In recent years, artificial intelligence (AI) has been applied in various ways in medicine and
dentistry. Advancements in AI technology show promising results in the practice of orthodontics. This scoping
review aimed to investigate the effectiveness of AI-based models employed in orthodontic landmark
detection, diagnosis, and treatment planning. Methods: A precise search of electronic databases was conduct-
ed, including PubMed, Google Scholar, Scopus, and Embase (English publications from January 2010 to July
2020). Quality Assessment and Diagnostic Accuracy Tool 2 (QUADAS-2) was used to assess the quality of the
articles included in this review. Results: After applying inclusion and exclusion criteria, 49 articles were included
in the final review. AI technology has achieved state-of-the-art results in various orthodontic applications,
including automated landmark detection on lateral cephalograms and photography images, cervical vertebra
maturation degree determination, skeletal classification, orthodontic tooth extraction decisions, predicting the
need for orthodontic treatment or orthognathic surgery, and facial attractiveness. Most of the AI models used
in these applications are based on artificial neural networks. Conclusions: AI can help orthodontists save
time and provide accuracy comparable to the trained dentists in diagnostic assessments and prognostic predic-
tions. These systems aim to boost performance and enhance the quality of care in orthodontics. However, based
on current studies, the most promising application was cephalometry landmark detection, skeletal classification,
and decision making on tooth extractions. (Am J Orthod Dentofacial Orthop 2021;160:170-92)

A
rtificial intelligence (AI) refers to a system's abil- already visible in our daily lives, from web searches to
ity to mimic human-like intelligence or be content filtering on social media and consumer products
defined as making effective and right decisions like smartphones, cameras, and cars.2-4
according to a gold standard.1 The increasing availability One of the main subcategories of AI is machine
of data, computing power, and improvements in ana- learning. Machine learning, which needs training data,
lytics methods allow AI to be integrated with many as- is a technique to provide predictions of new data and
pects of modern society. Its ubiquitous impacts are conditions on the basis of the previously learned data's
statistical pattern. This technique allows the computer
a
model to improve over time through experience without
Computer Engineering Department, Sharif University of Technology, Tehran,
Iran.
classic explicit programming.5 Deep learning is a subset
b
Department of Medical Physics and Biomedical Engineering, Tehran University of machine learning, which requires a model to be fed a
of Medical Sciences, Tehran, Iran.
c
large amount of data to learn features about the data
National Institute for Medical Research Development, Tehran, Iran.
d
Private practice, New York, NY.
with abstractions from multiple processing layers, with
e
Department of Orthodontics, School of Dentistry, & Dentofacial Deformities the privilege that does not require much engineering ef-
Research Center, Research Institute of Dental Sciences, Shahid Beheshti Univer- forts to preprocess the data. Deep learning methods have
sity of Medical Sciences, Tehran, Iran.
All authors have completed and submitted the ICMJE Form for Disclosure of Po-
been notably used in visual object recognition and ob-
tential Conflicts of Interest, and none were reported. ject detection.4 Being in the age of big data, world gov-
Address correspondence to: Saeed Reza Motamedian, Department of Orthodon- ernments and companies employed AI and its subfields
tics, School of Dentistry, Shahid Beheshti University of Medical Sciences, Danesh-
joo Blvd, Evin, Shahid Chamran Highway, Tehran 1983963113, Iran; e-mail,
as one of the leading strategies to deal with complexity
drmotamedian@gmail.com. in real-world decision making on the basis of large
Submitted, December 2020; revised, January 2021; accepted, February 2021. amounts of data.6
0889-5406/$36.00

2021 by the American Association of Orthodontists. All rights reserved.
Medicine is one of the fields that has been enhanced
https://doi.org/10.1016/j.ajodo.2021.02.013 by AI. Specifically, AI's ability to process a large amount
170
Mohammad-Rahimi et al
of data reduced the likelihood of neglecting valuable in-
formation. It is a potent and reliable tool to help physi-
cians by reducing diagnosis time.7,8 Promising
applications of AI diagnostic models have been reported
in radiology, dermatology, and oncology studies. These
examples include automatic detection of pulmonary
nodules, prostate cancer, coronary artery calcification,
differentiating skin lesions, lung nodules into benign
or malignant, and assessing bone age.2,9
AI also has applications in the field of dentistry.
Studies suggest that AI can become a powerful
decision-making tool within dentistry to promote clin-
ical care.10 Diagnostic imaging is the most notable use
case for the use of AI in dentistry. Currently, applications
and research in AI dental radiology focus on the diag-
nosis of osteoporosis,11 classification/segmentation of
maxillofacial cysts and tumors,12,13 description of peri-
apical disease,14 cephalometric landmarks detection,15
etc.
AI has shown to be an effective solution for the diag-
nosis and evaluation of orthodontic problems. Ortho-
dontic treatments for malocclusion can be categorized
as either extraction or nonextraction treatments.16 This
decision is traditionally made on the basis of clinical
experience gained over time; therefore, it is hard for
new practitioners to make these decisions.17 Investiga-
tions suggest that deep learning methods could help
to resolve this problem.18 Deciding if a patient requires
orthognathic surgery can also be challenging for practi-
tioners. AI has shown potential in this field to help clini-
cians determine whether surgical intervention is
necessary.19,20 Cephalometric analysis is extensively
used in orthodontics to diagnose facial growth anoma-
lies.21 The manual localization of cephalometric land-
marks on x-ray images is a time-consuming approach
with a high error rate.22 Recent studies demonstrate
outstanding achievements in landmark detection by AI
methods, especially using deep learning models.15,23
Furthermore, researchers can predict esthetics following
orthognathic surgery using AI, which can be a good cri-
terion for operating (Fig 1). Because orthodontics may
be unfamiliar with the terms used in the present study,
the authors provide a table with terms and definitions
(Supplementary Table I).
The studies mentioned above have shown promising
results regarding the effectiveness of AI to automate de-
cision making in orthodontics. Further investigations
must review these methods and evaluate the effective-
ness in achieving the mentioned goals in real-world or-
thodontic scenarios. This study aims to review how
different AI models perform in various orthodontic ap-
plications: orthodontic diagnosis, treatment planning,
and prognosis.
American Journal of Orthodontics and Dentofacial Orthopedics
171
MATERIAL AND METHODS
Protocol
The present systematic review follows Preferred Re-
porting Items for Systematic Reviews and Meta-
Analyses (PRISMA) extension for scoping reviews.24 A
filled version of the PRISMA form is provided in
Supplementary Table II. This study's research question
was, What are the applications of machine learning
techniques and their performances in the field of ortho-
dontics? The study looked for publications that evalu-
ated the performance of any machine learning or deep
learning approaches in the following domains: (1) anal-
ysis of orthodontics data, (2) prediction of outcomes in
the orthodontics treatments, (3) orthodontic diagnosis,
and (4) orthodontics treatment planning.
Eligibility criteria
The following inclusion criteria were used in the se-
lection of the articles:
1. Studies that used machine learning or deep learning
algorithms, by definition, a set of algorithms that
automatically detect patterns in data improved
through experience.
2. Studies that compared the measurement of model
outcomes with ground truth or gold standards.
Exclusion criteria were as follows:
1. Studies that used any machine learning or deep
learning approaches for problems not directly
related to orthodontics (eg, sleep apnea).
2. Studies that did not provide a clear explanation of
the machine learning or deep learning model that
was used to solve their problem.
3. Review studies were excluded.
Information sources and search
An electronic search was conducted on PubMed,
Google Scholar, Scopus, and Embase to find the relevant
literature. The search was limited to English publications
from January 2010 to June 2020. Various combinations
of the following keywords were used in the search pro-
cedure: machine learning, deep learning, neural
network, artificial intelligence, cephalo*, orthodon*.
The search results are presented in Table I. Endnote X9
(Clarivate, Philadelphia, Pa) was used as a reference
manager to manage the search results. Using this tool,
duplicate studies were removed. Furthermore, references
that were cited within the retrieved papers were reviewed
for finding any missing studies. Books, book sections,
generics, and thesis were first excluded.
August 2021
Vol 160
Issue 2
172 Mohammad-Rahimi et al

Fig 1. Graphical abstract of machine/deep learning application in the orthodontics. Machine/deep

learning approaches can be used for analyzing various types of orthodontics records to provide an ac-
curate and accelerated diagnosis. Furthermore, in conjunction with the patient records themselves,
these outputs can be used to provide an efficient treatment plan.

Selection of sources of evidence dataset specifications, data preprocessing procedure,

To identify eligible journal papers and conference
proceedings, 2 investigators (H.M.-R. and M.N.)
screened the title and abstracts on the basis of inclusion
and exclusion criteria independently. Then, the full texts
of potentially eligible publications were retrieved for
further assessments. Considering the inclusion and
exclusion criteria, 2 investigators identified the eligible
publications in this stage independently. Any disagree-
ments were resolved through consensus.
Data charting process
The data charting process was conducted by 2 inves-
tigators independently. Following the completion of the
charting process, any disagreements were discussed and
resolved through consensus.
Data items
The following data were extracted for the corre-
sponding groups of studies: the studies' objective,
the best-applied machine learning or deep learning
model architecture, model measurements, and model
performance (on the basis of the best model).
Critical appraisal of individual sources of evidence
For assessing the quality and risk-of-bias (RoB) of
the included studies, the QUADAS-2 was used.25 Us-
ing this tool, RoBs were evaluated in 4 domains: pa-
tient selection, index tests, reference standard, and
flow and timing. Using QUADAS-2, the authors rated
concerns regarding the included studies’ applicability
in 3 different domains: patient selection, index test,
reference standard. There were 3 options in each
domain: high, low, and unclear RoB. Because refer-
ence standard was the most influencing factor in the
relevant studies, the authors considered the reference
standard as the primary domain. If it was high or un-
clear, the RoB of the whole study was set as high or
unclear, respectively. If it was low, the RoB of the
study was determined on the basis of other domains.

August 2021
Vol 160
Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Mohammad-Rahimi et al 173

Table I. Results of primary electronic search on the selected databases

Database Keywords No. results
Pubmed ((“machine learning” OR “deep learning” OR “neural network” OR “artificial intelligence”) AND 102
(“cephalo*” OR “orthodon*”)) AND ((“2010/01/01”[Date - Publication] : “2020/06/30”
[Date - Publication]))
Google Scholar ((“machine learning” OR “deep learning” OR “neural network” OR “artificial intelligence”) AND 315
(“cephalo*” OR “orthodon*”))
Scopus ('machine learning':ti,ab,kw OR 'deep learning':ti,ab,kw OR 'neural network':ti,ab,kw OR 'artificial 210
intelligence':ti,ab,kw) AND ('cephalo*':ti,ab,kw OR 'orthodon*':ti,ab,kw) AND [2010-2020]/py
Embase ((TITLE-ABS-KEY (“machine learning”) OR TITLE-ABS-KEY (“deep learning”) OR TITLE-ABS-KEY 62
(“neural network”) OR TITLE-ABS-KEY (“artificial intelligence”)) AND (TITLE-ABS-KEY (“cephalo*”) OR
TITLE-ABS-KEY (“orthodon*”))) AND PUBYEAR . 2009
All searches conducted on July 4, 2020.

If a study had at least 2 high domains or 1 high and
1 unclear, the RoB of the whole study was set as
high RoB. Otherwise, if a study had at least 2 unclear
domains, the RoB of the entire study was set as un-
clear. Two investigators completed the evaluation
independently. Any disagreements were resolved
through consensus.
Synthesis of results
The included studies were divided into 4 categories
on the basis of their objective and the application of ma-
chine learning: (1) landmark detection in the lateral
cephalometry, (2) diagnosis and problem analysis in or-
thodontics, (3) treatment planning and prognosis, and
(4) other studies.
Because there were numerous landmark detection
publications in lateral cephalometry, it was considered
a separate category. The primary outcomes in all publi-
cations were measurable or had predictive outcomes for
evaluating the machine learning model. Using a wide
range of specific diagnostic tools in each publication,
conducting meta-analysis was impossible.
RESULTS
Selection of sources of evidence
A total of 689 studies were retrieved in our initial
search in the following databases: PubMed (n 5 102),
Google Scholar (n 5 315), Scopus (n 5 210), and Em-
base (n 5 62). After removing the duplicates and title/
abstract screening, 61 studies were selected for full-
text eligibility assessments. Finally, considering the in-
clusion and exclusion criteria, 49 studies remained.
Moreover, on the basis of the included studies’ refer-
ences, 3 new publications were added to the selected
studies. The PRISMA flow diagram is presented in
Figure 2.
Characteristics of sources of evidence
The publication year of various types of machine
learning studies is presented in Figure 3. As it can be
seen, considering that we included studies before July
2020, there is notable growth in the publication of this
study from the year 2019.
Critical appraisal within sources of evidence
Results of the RoB assessment of included studies
are presented in Supplementary Table III. A total of
77.55% of studies (38/49) were identified as low RoB
studies, whereas 14.29% and 8.16% of included studies
were identified as studies with high and unclear RoB,
respectively.
Results of individual sources of evidence and
synthesis of results
Landmark detection in lateral cephalometry. Twenty-
one publications were included regarding using machine
learning algorithms to detect orthodontic landmarks
(Table II). Except for Kunz et al,26 the primary measur-
able outcomes for evaluating the included studies were
at least 1 of the following: successful detection rate in
the range of 2-mm (2-mm SDR), mean radial error,
and classification accuracy (on the basis of classifica-
tion accuracy classified per clinical parameter). These
measurements are defined as follows27:
#fj :k Ld ðjÞ  Lr ðjÞ k \2g
2  mm SDR 5 3 100%
#U
where Ld and Lr are the location of the predicted and
referenced landmark, respectively. #U is the number of
detections made and j ˛ U. Furthermore, we defined
these as follows:
PN
Ri
MRE 5 i 5 1
N

American Journal of Orthodontics and Dentofacial Orthopedics August 2021
Vol 160
Issue 2
174 Mohammad-Rahimi et al

Fig 2. PRISMA flowchart.

pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
where the radial error (R) is Dx 2 1Dy2 , and Dx and Dy
are the absolute distances in the x-direction and y-direc-
tion between the predicted and referenced landmarks,
respectively. N is the number of landmarks. Finally, we
defined these as follows:
TP1TN
Classification Accuracy 5
#All Samples
where TP and TN are the number of true positive and
true negative samples, respectively.
Because most of the publications used International
Symposium on Biomedical Imaging (ISBI) 2014 and
2015 grand challenges in dental x-ray image
dataset,23,28-42 their results are comparable. The 2015
grand challenge in dental x-ray image dataset has 2
different test sets (Test 1 and Test 2 data) for the
performance measurement. A few studies used data
augmentation techniques to relieve small sample size
issues.26,32,43 More recent studies employed deep
learning models and artificial neural networks (ANNs)
to detect landmarks.15,23,26,28-35,37,43-45 Other studies
used variations of ensemble methods and random
forest.36,38-42
Regarding 2-mm SDR, Oh et al,28 Song et al,23 Gil-
mour et al,29 and Zhong et al30 showed the best perfor-
mance on the 2015 grand challenge in dental x-ray

August 2021
Vol 160
Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Mohammad-Rahimi et al
Fig 3. Characteristics of included studies by year of
publication.
image dataset by reaching more than 86% 2-mm SDR.
Gilmour et al29 reported the lowest mean radial error
(1.01 6 0.85 mm) on the 2015 grand challenge in dental
x-ray image dataset through a foveated pyramid atten-
tion approach by fine-tuning a pretrained artificial
network. Lindner et al38 showed 83.4% classification ac-
curacy on cephalometry analysis. Kunz et al26 showed
that, with the exception of SN-MeGo, there was no sig-
nificant difference in cephalometry analysis following
landmark detection.
Diagnosis and problem analysis in orthodontics. A to-
tal of 14 studies were included in analyzing orthodontic
data to diagnose, problem classification, and orthodon-
tic assessments (Table III). Major measurable outcomes
were area under curve (AUC) score, defined as the area
under the receiver operating characteristic curve,46
and classification accuracy. Similar to cephalometric
landmark detection, most studies used deep learning
techniques and ANN.22,47-57 No studies used publicly
available datasets.
When analyzing radiographs, some studies used the
cephalometry radiographs (used the cephalograms as
the model inputs),22,47,52 whereas others used measure-
ments of cephalograms analysis.49,53,57,58 Only 1 study
used 3-dimensional (3D) cone-beam computed tomog-
raphy images.48 Machine learning approaches were
applied to x-ray images for the prediction of cervical
vertebra maturation degree (accuracy between 72%
and 100%),47,49,52,53 maxillary structure variation
(average dice ratio of 0.800),48 maxillofacial skeletal
classifications (accuracy between 74.51% and
American Journal of Orthodontics and Dentofacial Orthopedics
175
96.40%),22,58 and detecting craniofacial patterns (with
the accuracy of 93%).57
Moreover, 4 studies used facial images or their mea-
surements as their input features.50,51,55,59 In these
included studies, machine learning techniques were
used for analyzing photographs to predict facial land-
mark detection (78.19% 1-mm SDR),50 facial attractive-
ness (accuracy between 58% and 83%),51,59 and facial
distortions (with the accuracy of 64.8%).55 Other than
the mentioned studies, genetic risk assessment of non-
syndromic orofacial cleft (AUC between 0.64 and 0.90)
was done using genetic data,54 and prediction of arch-
form was conducted using diagnostic cast data (with
an accuracy of 76.32%).56
Treatment planning and prognosis. A total of 12
studies were included regarding AI-based orthodontic
treatment and prognosis (Table IV). Notable measurable
outcomes were classification accuracy and AUC.
Most studies used random forest algorithms60-63 and
ANN.17-19,64 No studies used publicly available datasets.
In the reviewed studies, machine learning and deep
learning techniques were used for decision making in or-
thodontic treatment planning, which included extract-
ing teeth or not (accuracy between 80% and
96%),17-19,60,64 assessing treatment procedure, and ter-
minating orthodontic treatment (accuracy between
94.2% and 98.7%),61,65 using orthognathic surgery or
orthodontic treatment (accuracy between 91.9% and
96%),19,20,66 predicting blood loss before orthognathic
surgery (highly correlation between the predicted and
the actual blood loss values with P 5 0.001),62 designing
treatment plan with free-form certificates (F1 score of
58.5),67 and detecting essential features that contribute
most to the success of preorthodontic treatments for
cleft lip and palate patients (with the accuracy of
88.89%).63
Most studies used cephalogram analysis measure-
ments.17-20,60,61,63,66 Just 1 study used an end-to-end
approach used the cephalogram as the input to a convo-
lutional neural network (CNN).20 Moreover, one study
used facial images or their measurements as their
input.65 One study used natural language processing
to extract meaningful data and treatment plans from
dental certificates.67
Other related studies. Two studies were included in
this section (Table V). Zhang et al68 used deep learning
to convert 2D lateral cephalometry images to 3D images,
with a mean contour deviation of 0.12 mm. Further-
more, Pei et al69 used a deep model to sketch the
anatomic structure on cephalograms with average con-
tours error of 33.3 mm.
August 2021
Vol 160
Issue 2
August 2021
Vol 160
Issue 2

176
Table II. Summary of the findings regarding the landmark detection in lateral cephalometry
Performance measurements

No. Data Best model Classification

Author (y) Dataset source Dataset size landmarks preprocessing architecture 2-mm SDR MRE accuracy
Oh et al (2020)28 ISBI 2015 grand 400 cephalograms 19 Data augmentation A deep anatomic 86.20% (test 1) 11.77 6 10.13 pixels -
challenge in dental (consists of context feature 75.89% (test 2) (test 1)
x-ray image geometry and learning framework 14.55 6 8.22 pixels
intensity that enforces the (test 2)
transforms) CNN
Song (A) ISBI 2015 grand (A) 400 cephalograms 19 Extracted patches for Transfer learning 86.4% (dataset A, 1.077 mm (dataset A, -
et al (2020)23 challenge in dental (B) 100 images (only each landmark (400 using pretrained test 1) test 1)
x-ray image for test) patches for each ResNet-50 74.0% (dataset A, 1.542 mm (dataset A,
(B) A dataset provided landmark on each test 2) test 2)
by Shandong image, totally 62.0% (dataset B) 2.1 mm (dataset B)
University 60,000)
Kunz A private orthodontic 1792 cephalograms 18 Data augmentation Customized CNN Pearson product–moment correlation coefficients: r . 0.864 with
et al (2020)26 dental practice (postaugmentation) (rotation, tilting, P \ 0.001.
parallel shifting, Absolute mean differences: .0.37 for angular parameters, .0.20
mirroring, noise mm for metric parameters and .0.25% for the proportional
American Journal of Orthodontics and Dentofacial Orthopedics

adding as well as parameter. There were no significant differences between

changes in the gold standard and the predictions. (except for SN-MeGo)
brightness and
contrast)
Kim (A) Obtained from 2 (A) 2075 cephalograms 23 Cropped to a width- 2-stage DNN using a 82.92% (dataset A) 1.37 mm (dataset A) 91.17% (dataset A)
et al (2020)44 medical institutions, (B) 400 cephalograms height ratio of 1.0; stacked hourglass 84.53% (dataset B) 1.16 mm (dataset B) 83.13% (dataset B)
(B) ISBI 2015 grand region of interest network
challenge in dental images of each
x-ray image landmark with the
original resolution
were extracted
Hwang Not mentioned 1311 cephalograms 80 - YOLOv3 algorithm - 1.46 6 2.97 mm -
et al (2020)15 with custom (compared with
modifications human
1.50 6 1.48 mm)

Mohammad-Rahimi et al
American Journal of Orthodontics and Dentofacial Orthopedics

Mohammad-Rahimi et al
Table II. Continued

Performance measurements

No. Data Best model Classification

Author (y) Dataset source Dataset size landmarks preprocessing architecture 2-mm SDR MRE accuracy
Gilmour and ISBI 2015 grand 400 cephalograms 19 Extracted patches for Transfer learning 88.32% (test 1) 1.01 6 0.85 mm -
Ray (2020)29 challenge in dental each landmark using pretrained 77.05% (test 2) (test 1),
x-ray image ResNet-50 with 1.33 6 0.74 mm
foveated pyramid (test 2)
attention algorithm
Zhong ISBI 2015 grand 300 cephalograms 19 Cropped to An attention-guided 86.74% 1.14 6 1.03 mm -
et al (2019)30 challenge in dental (Test 2 was 1935 3 1935 pixels. deep regression
x-ray image excluded) Scaled image by model through 2
0.15 times stage U-net (using
the expansive
exploration)
Song ISBI 2015 grand 400 cephalograms 19 Extracted patches for Transfer learning 85.0% (test 1), 1.147 mm (test 1), -
et al (2019)31 challenge in dental each landmark (200 using pretrained 81.8% (test 2) 1.223 mm (test 2)
x-ray image patches for each ResNet-50
landmark)
Qian ISBI 2015 grand 400 cephalograms 19 Data augmentation An improved faster R- 82.5% (test 1), - -
et al (2019)32 challenge in dental using multiscaling CNN (using 72.4% (test 2)
x-ray image strategy (narrow multitask loss)
image size from 0.3
to 0.9), produced
more than 1000
cephalograms
Park Obtained from Seoul 1311 cephalograms 80 - YOLOv3 and single 80.4% - -
et al (2019)45 National University shot multibox
Dental Hospital detector algorithm
with custom
modifications
August 2021
Vol 160
Issue 2

Nishimoto Scraping 219 cephalograms 10 Data augmentation by Customized CNN - 17.02 6 11.13 pixels -
et al cephalograms on rotating, deviating,
(2019)43 the internet and changing
contrast produced
more than 7803
cephalograms
Goutham ISBI 2015 grand 400 cephalograms 7 A total of 6750 Modified u-net 65.13% - -
et al challenge in dental segmentation map
(2019)33 x-ray image for each landmark
point, adaptive
histogram
equalization

177
August 2021
Vol 160
Issue 2

178
Table II. Continued

Performance measurements

No. Data Best model Classification

Author (y) Dataset source Dataset size landmarks preprocessing architecture 2-mm SDR MRE accuracy
Dai et al ISBI 2014 grand 300 cephalograms 19 Cropping with Adversarial encoder- Between 35% and 2.5-7.5 mm for each -
(2019)34 challenge in dental template matching decoder networks 40% for each landmark
x-ray image landmark
Chen et al (A) ISBI 2015 grand (A) 400 cephalograms, 19 - End-to-end deep 86.21% (test 1) and 1.25 mm (test 1) and
(2019)35 challenge in dental (B) 1857 cephalograms learning with novel 73.89% (test 2), 1.47 mm (test 2).
x-ray image attentive feature Best result on other Best result on other
(B) 5 new datasets pyramid usion data sets: 94.73% datasets: 0.88 mm
collected (A-E) for module using
the purpose of the Inception
study
Wang et al (A) ISBI 2015 grand (A) 300 cephalograms (A) 19 SIFT-based patch Multiresolution (A) 73.37% (A) 1.69 6 1.43 mm
(2018)36 challenge in dental (Test 2 was (B) 45 feature extraction decision tree (B) 72.08% (B) 1.71 6 1.39 mm
x-ray image excluded), regression voting
(B) Dataset provided (B) 165 cephalograms
by Peking University
American Journal of Orthodontics and Dentofacial Orthopedics

School and Hospital

of Stomatology
(private)
Arık, S., ISBI 2014, 2015 grand 400 cephalograms 19 Down sampled by 3 Customized CNN 75.58% (2014), - 75.92%
et al. challenge in dental (by taking the 75.37% (2015-test (2015-test 1)
(2017)37 x-ray image average of each 1), and 67.68% and 76.75%
3 3 3 patch) for (2015-test 2) (2015-test 2)
dimensionality
reduction
Lindner, C. ISBI 2015 grand 400 cephalograms 19 - RF regression-voting 84.7% 1.2 mm 78.4% (over all
et al challenge in dental using constrained classes),
(2016)38 x-ray image local model 83.4% (over all
framework subjects)
Lindner and ISBI 2015 grand 400 cephalograms 19 - RF regression-voting 74.95% 1.67 6 1.65 mm 77%
Cootes challenge in dental using constrained
(2014)40

Mohammad-Rahimi et al
x-ray image local model
framework
Chu et al ISBI 2014 grand 200 cephalograms 19 - Combination of RF 77.79% (4mm-SDR) - -
(2014)42 challenge in dental regression-based
x-ray image landmark detection
with sparse shape
composition model
based landmark
correction
Mohammad-Rahimi et al 179

DISCUSSION

MRE, mean radial error; CNN, convolutional neural network; DNN, deep neural network; R-CNN, region based convolutional neural network; SIFT, scale-invariant feature transform; RF, random
Clinical decision-making necessitates the practitioner
Classification
accuracy
- to apply broad knowledge to analyze each specific pa-

-
tient for developing the diagnosis and prognosis to
determine treatment strategies.70 AI application in
health care has improved the accuracy in the diagnosis
Performance measurements

and treatment planning.18,19 The present review aimed

to review papers that use machine learning or deep
1.83 6 1.81 mm

learning approaches for enhancing diagnosis and

MRE

About 2 mm
decision-making procedures in orthodontics.
Although appreciating AI's novelty in the orthodon-
tics discipline is of utmost importance, the rigorous clas-
sification of AI platforms is imperative to improve
clinical outcomes further. The authors divided AI appli-
65.26 6 18.26 %
2-mm SDR

cations in orthodontics into 3 distinct classes: cepha-

lometry landmark detection, diagnosis, and treatment
planning.
77.58%

Cephalometry landmark detection

simple pixel-based

Cephalometric analyses are widely considered as a

randomized trees

multiresolution
combined with
architecture

Random decision
Best model

critical diagnostic tool in determining the best therapeu-

forest-based

tic options.71 Landmark detection in lateral cephalom-

Ensembles of

likelihoods
different resolutions extremely

features

etry is an initial phase to define the primary

maxillofacial classifications (skeletal classifications and
intermaxillary relations, soft tissue classifications,
etc).26,72 Measurements performed on the basis of the
Downsizing into 6
preprocessing

landmarks provide supportive information for the pro-

vider to determine the optimal treatment plan. The
Data

more information clinicians have, the better they can

treat the patient.28
The most important challenge in conventional ceph-
alometric tracing by dentists is a relatively time-
landmarks

consuming process.28 Moreover, high intrapersonal

No.

and interpersonal variations of landmark tracing are

19

19

other problems that can lead to errors in orthodontic

problem diagnosis and, consequently, treatment plan-
200 cephalograms

200 cephalograms
Dataset size

ning and decision making.73 To solve these challenges,

applying AI could ultimately increase the efficiency of
assessments for clinicians.28
In the relevant studies, deep learning models23,28-32
generally showed better performance than classical
machine learning techniques.36,39-41 However, in 2016,
challenge in dental

challenge in dental
Dataset source

Challenge grand

Lindner et al38 reached a comparable accuracy of

Mirzaalian, H., ISBI 2014 Grand
ISBI 2014 grand

x-ray image

x-ray image

84.7% 2-mm SDR using a classical machine learning al-

gorithm (random forest regression-voting using con-
Table II. Continued

strained local model framework). Other studies also

revealed that it seems random forest based among clas-
sical machine learning approaches showed the best re-
et al (2015)39

et al (2014)41

sults regarding landmark detection.38,40-42 One of the

Author (y)
Vandaele

most promising approaches was to extract numerous

forest.

patches for each landmark and apply a deep learning

model.23,28,29,31 Gilmour et al29 and Song et al23 reached

American Journal of Orthodontics and Dentofacial Orthopedics August 2021
Vol 160
Issue 2
August 2021
Vol 160
Issue 2

180
Table III. Summary of the findings regarding the diagnosis and problem analysis in orthodontics
Performance
measurements

Data structure Best model

Author (y) Object Data source and size Data preprocessing architecture Accuracy AUC score Others
Amasya Cervical vertebra University 647 Landmarking on ANN 86.93% - Sagittal
et al (2020)49 maturation degree archive cephalograms cervical vertebra, mean
the measurement sensitivity:
data of landmarks 93.55%
on vertebra were Vertical
given to the model mean
(totally 56 sensitivity:
features) 94.59%
Yu et al End-to-end skeletal Obtained from 5890 Under-sampling, A modified DenseNet Sagittal mean Sagittal AUC: Sagittal
(2020)22 classification Ewha Womans cephalograms data augmented with pretrained accuracy: 0.965 to 0.991 mean
(Sagittal and University by rotation, weights for the 95.70% Vertical AUC: specificity:
Vertical) Medical Center shifting, ImageNet Vertical mean 0.967 to 0.995 96.77%
histogram accuracy: Vertical
equalization, and 96.40% mean
American Journal of Orthodontics and Dentofacial Orthopedics

adaptive specificity:
histogram 97.29%
equalization Sagittal
mean
sensitivity:
93.55%
Vertical
mean
sensitivity:
94.59%
Makaremi End-to-end cervical Not 650 Sobel filter Customized CNN 85% - -
et al vertebra mentioned cephalograms Resize: 488 3 488,
(2020)47 maturation degree 244 3 244,
64 3 64
Chen et al Assess maxillary Collected from 36 cone-beam Not mentioned A machine learning - - Dice ratio
(2020)48 structure variation Peking computed algorithm using of maxilla:

Mohammad-Rahimi et al
in unilateral University tomography learning-based 0.800 6
canine impaction Hospital multisource 0.029
integration
framework for
segmentation
American Journal of Orthodontics and Dentofacial Orthopedics

Mohammad-Rahimi et al
Table III. Continued

Performance
measurements

Data structure Best model

Author (y) Object Data source and size Data preprocessing architecture Accuracy AUC score Others
Makaremi End-to-end cervical Not 1870 Sobel filter Customized CNN Reached 1 - -
et al vertebra mentioned cephalograms Entropy filter 00% accuracy
(2019)52 maturation degree Resize: 512 3 512, with entropy
256 3 256, filter
128 3 128
K€
ok Cervical vertebra Not 300 The measurement k-NN, Decision tTee, Decision tree ANN for -
et al maturation degree mentioned cephalograms data of 20 ANN, SVM for CSV1 (97.1%) CSV1
(2019)53 landmarks on and CSV2 (90.5%); (0.99),
vertebra were SVM for CVS3 CSV2
given to the model (73.2%) and (0.96),
CVS4 (58.5%); CSV4
kNN CVS5 (60.9%) (0.90)
for CVS 6 (78.7%) and
CSV6
(0.96);
SVM for
CSV3
(0.96),
CSV5
(0.92)
Rao et al Identifying and Obtained from Twenty-two Cropped to YOLO trained on - - 21.81%
(2019)50 analyzing facial volunteers by 2-dimensional 512 3 512 large-scale Celeb- landmarks
landmarks on 2- different image facial images Faces Attributes in the error
dimensional facial sensors (with 418 facial Dataset (for face range of
images landmark points detection). Active 0-1 mm,
and 220 landmark shape model 34.09%
August 2021
Vol 160
Issue 2

measures) algorithm trained landmarks

on 1000 celeb- in the error
face dataset range of
(finding the facial 2-3 mm,
landmarks)

181
August 2021
Vol 160
Issue 2

182
Table III. Continued

Performance
measurements

Data structure Best model

Author (y) Object Data source and size Data preprocessing architecture Accuracy AUC score Others
Patcas Evaluate facial Not mentioned Frontal and profile All face images were A computational - - AI evaluation
et al attractiveness of images of 20 brought to an algorithm of cleft patients:
(2019)51 treated cleft treated left-sided equal size of consisting of 4.75 6 1.27
patients and cleft patients and 256 3 256 pixels Matthias face Laypeople:
controls by AI 10 controls with a centered detector and 4.24 6 0.81,
face and a 40% pretrained VGG- Orthodontists:
background 16 trained to 4.82 6 0.94,
margin extract facial oral Surgeons:
features 4.74 6 0.83
associated with (all Ps $ 0.19)
attractiveness, to Facial
provide a attractiveness
prediction of facial of controls
attractiveness was rated
American Journal of Orthodontics and Dentofacial Orthopedics

significantly
higher by
humans than AI
(all Ps # 0.02)
Zhang Genetic Risk Infants attending the 382 patients and 43 risk and protective LR, SVM - Han -
et al Assessment of Smile Train in 709 control alleles were Chinese
(2018)54,68 Infants with Yantai, Shandong collected from best model:
Nonsyndromic (Han population) GWAS catalog LR with
Orofacial Cleft and Kashi, database used for 0.90
Xinjiang (Uyghur model input AUCUyghur
population) Chinese
best
model: SVM
with 0.64 AUC
Murata Automated Obtained from 352 frontal Data augmented by A hybrid model using 64.8 6 7.7 % - -
et al diagnostic of Department of facial images. inverting (which CNN and RNN

Mohammad-Rahimi et al
(2017)55 facial distortions Orthodontics at The labels were were labeled in the with an attention
Osaka University three-grade opposite direction) mechanism using
Dental Hospital distortions of (totally 704 fine-tuned VGG-
the mouth, jaw, images) 19
and entire face
American Journal of Orthodontics and Dentofacial Orthopedics

Mohammad-Rahimi et al
Table III. Continued

Performance
measurements

Data structure Best model

Author (y) Object Data source and size Data preprocessing architecture Accuracy AUC score Others
no-Sandoval An automatic to
Ni~ Obtained from 229 cephalograms Cephalometric SVM 74.51% - Geometric
et al (2016)58 classify skeletal faculty of (labeled with measurements mean: 0.73
patterns (sagittal) Odontology, cephalometric were given to the
National landmarks and model
University of analysis)
Colombia
Yu et al Facial attractiveness Obtained from 6 108 patients with The of landmarks SVR 71.8% - -
(2014)59 from orthodontic universities in frontal, lateral, configuration with
photographs China and frontal the corresponding
smiling three-point scale
photographs grade, was used as
before and after the input data
orthodontic
treatment (labeled
from most
attractive to least
attractive)
Budiman Qualitative archform Obtained from 190 dental casts of Input parameters ANN 76.32% - -
(2013)56 diagnostic (the Postgraduate maxillae were intercanine
shape of archform) Clinic Faculty of width, canine
Dentistry depth, intermolar
University of width and molar
Indonesia and 3 depth
other
orthodontists
Diaz et al Detect craniofacial Collected from 100 cephalograms 30 cephalometric K-means algorithm; 93% - -
August 2021
Vol 160
Issue 2

(2013)57 morphology Chilean patients features were J48 decision tree

patterns extracted
CNN, convolutional neural network; kNN, k-nearest neighbors; SVM, support vector machine; GWAS, genome-wide association study; LR, Linear Regression; SVR, Support Vector Regressor.

183
August 2021
Vol 160
Issue 2

184
Table IV. Summary of the findings regarding the treatment planning and prognosis
Performance measurements

Data structure Data Best model AUC

Author (y) Object Data source and size preprocessing architecture Accuracy score Others
Suhail Constructing an Obtained from a 287 patients Totally 19 RF ensemble 94.4% - -
et al decision making private center (medical charts, diagnostic learning
(2020)60 system for teeth cephalometric features were
extraction x-rays, extracted
panoramic
radiographs,
facial
photographs,
and intraoral
photograph)
Dharma Predicting Obtained from the 310 dental Categorical RF 98.71% 0.88 Sensitivity:
et al cessation of Division of malocclusion variables were 92.50%
(2020)61 orthodontic Orthodontics, treatments converted to a Specificity:
treatments University records Likert-scale 99.63%
Dental Hospital, varaibles
American Journal of Orthodontics and Dentofacial Orthopedics

Peradeniya, Sri
Lanka
Riri et al Evaluating the Obtained from a 1207 Facial and skin Tree-based 94.28% Sensitivity:
(2020)65 progression of private center extraoral, color features classification 95.35%
treatment intraoral, were extracted Specificity:
and mold 99.60%
image of Cohen
98 patients Kappa:
93.44
F1-measure:
92.6
Lee Differential Obtained from a 333 Totally 50 Modified- 91.9% 0.96 Sensitivity:
et al diagnosis of the Korea University cephalograms cephalometric Alexnet 85.2%
(2020)20 orthognathic Ansan Hospital landmarks were Specificity:
surgery detected and 97.3%
used

Mohammad-Rahimi et al
Stehrer Predict blood loss Obtained from 950 - RF - - Highly
et al before department of patients significant
(2019)62 orthognathic Cranio correlation
surgery Maxillofacial between the
Surgery of the predicted and
Kepler the actual
University Clinic blood loss
Linz, Austria values
(P \ 0.001,
mean
difference
was 7.4 6
172.3 mL)
American Journal of Orthodontics and Dentofacial Orthopedics

Mohammad-Rahimi et al
Table IV. Continued

Performance measurements

Data structure Data Best model AUC

Author (y) Object Data source and size preprocessing architecture Accuracy score Others
Li et al Determination of Obtained from 302 cephalometric Totally, 24 ANN Extraction- Extraction- Extraction-
(2019)64 extraction- Department of and features were nonextraction: nonextraction: nonextraction:
nonextraction Orthodontics, demographic extracted 94.0% 0.982 Sensitivity:
approach, West China data Extraction 94.6%
extraction Hospital of patterns: Specificity:
patterns, and Stomatology in 83.3% 93.8%
anchorage Chengdu, China Anchorage
patterns in the patterns:
treatment 92.8%
planning
Choi Determination and Obtained from 316 Totally, 18 ANN Surgery/ - -
et al surgery/ Department of cephalograms features were nonsurgery
(2019)19 nonsurgery Orthodontics, extracted decision:
decision, Seoul National 96%
surgeryype and University Surgery-type
extraction Dental Hospital and extraction
decision decision: 91%
Kajiwara Designing Not mentioned 990 Feature vectors SVM - - F1 score: 0.585
et al treatment plan certificates were Spearman's rho:
(2019)67 on the basis of written by constructed 0.584
free-form dentists using Bag-of- (treatment
documents Words, prioritization
written by Universal task)
dentists Sentence
Encoder and
one-of-kind
methods
August 2021
Vol 160
Issue 2

Thanathorn Evaluate the need Obtained from a 1000 15 orthodontic BN 96% 0.91 Sensitivity: 95%
wong, for orthodontic private center patients treatment Specificity: 100%
(2018)66 treatment in variables were Kappa
patients with extracted as value 5 0.894
permanent input features
dentition
Omar Explore features in Obtained from 18 Totally, 7 features RF 88.89% - -
et al success of Queen Elizabeth patients were extracted
(2018)63 preorthodontic II Hospital, Kota
treatments for Kinabalu Sabah
cleft lip and
palate patients

185
186 Mohammad-Rahimi et al

88.32% and 86.4% 2-mm SDR on the ISBI 2015 grand

challenge in dental x-ray image dataset using extracted
Others
patches with a pretrained transfer learning model.
-

-
In general, the application of deep learning in land-
mark detection can be divided into 3 different main ap-
proaches, including the regression-based method,74
Performance measurements

region proposal classification,32,37 and heatmap regres-

sion.75 In the regression-based method, the CNN-based
frameworks learn to directly regress x- and y-coordi-
score
AUC

- nates of landmarks.76 In the second approach, the

models aim to predict the region proposal-wise features
on anatomic landmarks. Then, image patches will be
classified into landmark categories and used to localize
nonextraction:

landmark coordinates.32 Using this approach, Arık

patterns: 84
Accuracy

et al37 classified pixel-wise features using CNN to predict

Extraction-

Extraction

the location of the landmarks. In the heatmap regres-

93%

sion, the deep learning models learn to predict the heat-

80%

maps representation of landmarks (eg, heatmap on the

basis of the distance between pixels and landmarks). In
this example, x- and y-coordinates of landmarks can
architecture
Best model

be localized through the local or global minima in the

heat maps.76 The advantages of this approach are
ANN

ANN

computational efficiency, simplicity, and end-to-end

learning.28
preprocessing

features were

features were

Diagnosis
Data

extracted

extracted

An essential step in treatment planning in orthodon-

Totally, 18

Totally, 23

tics is analyzing patients’ data and proper diagnosis.

These AI-based technologies have been a powerful tool
in disease prediction and diagnosis, ultimately aiding
physicians in providing the ideal treatment plan.10 These
Data structure

cephalograms

cephalograms

AI-based systems can be used to assist the orthodontist

and size

RF, random forest; BN, Bayesian network; SVM, support vector machine.

in providing standard treatment to the patient and maxi-

mizing the chances of achieving the set goals. The
156

200

specialist can use AI technology to diagnose better and

improve clinical decision making.77
One of the important guides for growing patients to
Dental Hospital
in Seoul, South
Seoul National
Data source

determine the proper time to start treatment is detecting

Not mentioned
Obtained from

University

a patient's skeletal maturation and bone age. The cervi-

Korea

cal vertebral maturation (CVM) degree, which can be

performed on lateral cephalograms, is the common
approach to detecting patients' skeletal age and matura-
tion.78 The major problem in detecting CVM is that it has
approach in the
patterns in the
Determination of

Determination of
nonextraction

nonextraction
approach and

been reported that there are less than 62% intraobserver

extraction-

extraction-
extraction
Object

treatment

treatment

agreement and 50% interobserver agreement among or-

planning

planning
Table IV. Continued

thodontists.79 Machine learning or deep learning algo-

rithms have been used for detecting CVM degrees on
cephalograms.47,49,52,53 Amasya et al49 and K€ ok et al53
marked a specific set of landmarks cervical vertebrae
(2016)18

(2010)17
Author (y)

and extracted features from landmarks' relations to

et al
Kim
and

each other. Then, they fed their models with extracted

Jung

Xie

features. The disadvantage of their approach was that

August 2021
Vol 160
Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Mohammad-Rahimi et al 187

they required additional hand-crafted features, which is

errors: 0.768 mm
Average landmarks
errors: 33.3 mm
deviation of the

when given the

The mean surface time-consuming, and it heavily relied on the quality of

Average contours
from 0.78 mm
ACB decrease

to 0.12 mm
initial manual landmark localization. Using CNN and
Results

constraint
temporal
by preprocessing images by entropy filter, Makaremi
et al52 reported 100% accuracy on cropped cephalo-
grams on the region of interest. Although their results
seem to be interestingly favorable, they did not mention

The contours errors

how images were cropped. For future studies, the au-
the mean contour
Performance

thors suggest developing an AI-based tool that can

deviation of
measures

errors (mm)
landmarks
(mm) and
detect CVM degrees using lateral cephalograms through
the ACB

end-to-end approaches, which can be performed using

the whole lateral cephalograms as their models' input
or developing a 2-stage algorithm in which the cropping
procedure is standardized using another model.
connections between

Boltzmann machine
convolutional layers
VGG-face model, with
long-jump residual
Model architecture

Detection of the craniofacial and skeletal patterns us-

ing machine learning approaches was another domain in
A bimodal deep
intermediate

AI-based orthodontic diagnosis.22,48,57,58 Correct classi-

fication of the skeletal pattern is a crucial step in ortho-
dontic diagnosis, and it is essential for successful
decision making and treatment planning.80,81 Similar
to CVM studies, Ni~ no-Sandoval et al58 suggested a 2-
size of 1 3 1 3 1.
Data preprocessing

190 with a voxel

step approach through landmark localization, which is

and rescaled to

CBCT images are

200 3 200 3

prone to errors and biases in the final diagnosis. This

were cropped

resolution of
rescaled to a
200 3 200;
Cephalograms

bias is because the definitive diagnosis is based on

manually identified landmarks fed into the machine
learning model. However, regarding skeletal classifica-
-

tion on the basis of lateral cephalograms, Yu et al22 pro-

posed an end-to-end CNN-based deep learning model.
724 cephalograms
30 cephalograms
Data structure

Using a pretrained transfer learning model, they reached

2D, 2-dimensional; ACB, anterior cranial base; CBCT, cone-beam computed tomography.
and size

95.70% and 96.40% accuracy in sagittal and vertical di-

mensions, respectively. Similarly, AI-based end-to-end
medical diagnostic systems on the basis of x-ray images
or photographs using CNN structures showed their
Table V. Summary of the findings regarding other studies

clinician-level classification performance regarding the

obtained from
a longitudinal

Not mentioned
Data source

diagnosis of pneumonia82 and melanoma.83 Despite

Cephalograms

population

their high accuracy reported by Yu et al,22 a possible

dataset

source of bias in their model inhibits widespread gener-

alization of it because it was trained on lateral cephalo-
grams taken from 1 device with similar properties.
A temporal consistent 2D-3D

Therefore, this model should be evaluated on images

Anatomic structure sketcher

from other resources before its widespread application

craniofacial structures
registration technique
enabling 3D growth

in different clinical settings.

for cephalograms
measurements of
Object

Other than x-ray images, photography im-

ages50,51,55,59 and diagnostic casts56 can be considered
great resources to be analyzed by machine learning for
orthodontic diagnosis. Similar to landmark detection
on cephalograms studies, Rao et al50 developed a model
to localize landmarks on photography images, which can
Pei et al (2013)69

be used for analyzing orthodontic problems. Using pho-

(2018)54,68

tographs, it is possible to assess the degree of malocclu-

Zhang et al
Author (y)

sion and jaw abnormality, leading to masticatory

dysfunction, apnea syndromes, and orthodontic-
related problems.55 To achieve these goals, Murata

American Journal of Orthodontics and Dentofacial Orthopedics August 2021
Vol 160
Issue 2
188
et al55 study was the only study that proposed deep end-
to-end learning to diagnose various types of facial dis-
tortions and reached 62.9% accuracy. However, their
research was a step toward photography-based diag-
nostic systems; their low-accuracy and training proced-
ure only on one specific ethnic group should be resolved
in the future for the clinical application. Furthermore,
photography images can be used for the assessment of
facial attractiveness.51,59 Facial appearance and attrac-
tiveness are crucial in patients' quality of life and psy-
chosocial well-being, so they should be considered by
orthodontists.84 Because laypeople and clinicians' per-
ceptions of attractiveness can have biases because ofva-
rious reasons,51 developing an unbiased tool to evaluate
patients’ attractiveness can be beneficial.
Only Budiman56 used orthodontic diagnostics cast in
his study. He fed measurements on diagnostic casts to an
ANN to predict the shape of the arch. Successful long-
term stability of orthodontic treatments will be acquired
when the archform would be maintained during the
treatment procedure.85 Therefore, the correct diagnosis
of archform is a crucial step in treatment planning.
Future studies can use 3D scanned models of the diag-
nostic cast for analyzing with AI tools with the advance-
ment in technologies. One of the future applications of
using AI in analyzing diagnostic casts is developing a
model to select the best archwire on the basis of the pa-
tients’ archform to enhance treatment outcomes.86
Treatment planning and decision making
Treatment planning is one of the most substantial
steps in the orthodontic procedure. AI helps this process
by providing information learned from experienced
experts. Prediction of treatment outcomes can aid
the clinician in better decision making. The predetermi-
nation of esthetic and clinical outcomes could help
both the surgeon and the patient make the optimal
decision.62,63
Tooth crowding and protrusions can be considered
the main reasons for tooth extraction in orthodon-
tics.17,87 Although there is no right answer for tooth
extraction, experienced specialist decisions can be
learned and become an additional reference for less
experienced practitioners.18 Studies extracted orthodon-
tic features on the basis of a mixture of cephalometric
data, photographs from patients, and clinical examina-
tion.17-19,60,64 Most studies applied ANN for this pur-
pose. Xie et al17 proposed an ANN-based detection of
teeth extraction. Two features, anterior teeth uncovered
by incompetent lips and IMPA (L1-MP) were the most
contributive features in their study. The limitation of
their research was its low generalizability in the test
August 2021
Vol 160
Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Mohammad-Rahimi et al
set, which needs further investigation. In addition to
answering the tooth extraction question, Jung and
Kim18 determined the extraction position. They divided
their data into validation and test set to improve gener-
alization problems in the Xie et al17 study and prevent
overfitting. The main limitation of their study was
confinement to nonsurgical treatment patients. Howev-
er, Choi et al19 just investigated extraction needed in or-
thognathic surgeries and achieved a 91% success rate in
diagnosing surgery-type and extraction decision. Li
et al64 achieved 94% extraction accuracy in their
extraction-nonextraction study, which is 14% higher
than Xie et al.17 The partial derivatives method was im-
plemented for feature importance detection, which was
more stable than Xie et al.17 Suhail et al,60 unlike other
study’ methodologies, used the random forest algo-
rithm. This study's limitations included excluding
mandibular incisor extraction, second premolar extrac-
tions, and confining outcomes to nonsurgical orthodon-
tic procedures, which the latter was also observed in
other studies.17,18,64 It is necessary to do a comprehen-
sive study that incorporates all exceptions for real appli-
cations in clinics. Probably by collecting a high amount
of orthodontic patients, we can reach a promising result.
Another important characteristic of AI in treatment
is its role in evaluating the treatment process63,65,66
and its cessation.61 Riri et al65 investigated machine
learning algorithms to classify orthodontic images
to evaluate each session's progression using a tree-
based classifier and facial features and skin color
detection. Dharma61 predicted discontinuation of
orthodontic treatments of dental patients. On the basis
of this study, the cessation of the treatment was highly
influenced by active treatment duration. Thanathorn-
wong66 developed an orthodontic expert system on
the basis of a Bayesian network to help general practi-
tioners assess the need for orthodontic treatment. The
employed Bayesian network consisted of relevant vari-
ables on the basis of patient intraoral and extraoral
orthodontic-related data. Despite the limited number
of training data, there was a remarkable similarity be-
tween the orthodontists' judgments and the proposed
expert system's results. The system should be further
improved by adding more data and dental conditions
for orthodontic treatment assessment for use in large
community hospitals with a shortage of orthodontic
specialists.
The other significant role of AI in orthodontic treat-
ment is its diverse orthognathic surgery applica-
tions.19,20,62 Lee et al20 examined the need for
orthognathic surgery or orthodontic treatment using
the deep learning method, with the advantage of elimi-
nating errors caused by landmark measurement values in
Mohammad-Rahimi et al
traditional neural network studies.19 In the aforemen-
tioned study, the important regions resulted from AI
showed considerable differentiation between the maxilla
and mandible for deciding which treatment be used. The
remarkable limitation of deep learning studies is data
consistency problems and image distortion caused by
dental prostheses and the need for a large amount of
data. Stehrer et al62 predicted blood loss value before or-
thognathic surgery, allowing for an optimal treatment
plan before the surgery. It was the first study in this
domain, and more studies from other clinics and surgical
procedures need to be done to confirm the results of this
study.
Future of machine or deep learning applications in
orthodontics and its limitations
Orthodontics is a specialty that will continue to
evolve, especially with advancements in AI.88 However,
the study results suggested that we are at the inception
of orthodontics leveraging AI diagnosis and decision
making. One of the foremost opportunities for AI-
based systems regarding orthodontics, which has not
been assessed in the literature, is its usages in precision
medicine. Given the advances in orthodontics and preci-
sion medicine, researchers are trying to propose custom-
ized treatment approaches on the basis of each patient's
characteristics to enhance treatment outcomes. It has
been reported that moving toward precision orthodon-
tics could be the next paradigm shift in orthodontic
treatments.89 In contrast, using the capabilities and po-
tentials of machine learning and deep learning, data-
derived precision medicine approaches can be facilitated
and enhanced.90
Despite the high performance of AI-based ap-
proaches in medical diagnostics publication, these
methods are not used widely in clinical settings.91
Regarding the AI-assisted approach in orthodontics,
there are still many shortcomings. For instance, not
considering oral diseases and prioritizing dental treat-
ment plans are evident ones.77,92 Another critical limita-
tion of the currently proposed machine or deep learning
models is that they cannot determine the etiology of or-
thodontic problems in each individual.77 These issues
could only be solved by an experienced professional clin-
ical who could extensively analyze the treatment options
while considering interdisciplinary problems (eg, restor-
ative needs). In contrast, because of the deep learning
models’ black-box nature, they do not provide any
explanation or knowledge for their outcomes. However,
recent medical image analysis studies tried to propose
approaches for explainability and interpretability of AI
model outcomes.91
American Journal of Orthodontics and Dentofacial Orthopedics
189
CONCLUSIONS
AI can be considered a high-potential assistant to
perform diagnostic assessments and prognostic predic-
tions or determine treatment plans (especially for more
complicated patients). However, the current scoping re-
view results implied that we are far from widespread
clinical application of end-to-end machine learning-
based tools. With this in mind, the most promising appli-
cations of these approaches were landmark detection on
lateral cephalograms, skeletal classification, and deci-
sion making on tooth extraction.
AUTHOR CREDIT STATEMENT
Hossein Mohammad-Rahimi contributed to concep-
tualization, investigation, data curation, and original
draft preparation; Mohadeseh Nadimi: contributed to
the investigation, data curation, and original draft prep-
aration; Mohammad Hossein Rohban contributed to
conceptualization, methodology, supervision; Erfan
Shamsoddin contributed to data curation and original
draft preparation; Victor Y. Lee contributed to original
draft preparation, review, and editing; and Saeed Reza
Motamadian contributed to the conceptualization,
methodology, and supervision.
SUPPLEMENTARY DATA
Supplementary data associated with this article can
be found, in the online version, at https://doi.org/
10.1016/j.ajodo.2021.02.013.
REFERENCES
1. Hamet P, Tremblay J. Artificial intelligence in medicine. Meta-
bolism 2017;69S:S36-40.
2. Bouletreau P, Makaremi M, Ibrahim B, Louvrier A, Sigaux N. Arti-
ficial intelligence: applications in orthognathic surgery. J Stomatol
Oral Maxillofac Surg 2019;120:347-54.
3. Jiang F, Jiang Y, Zhi H, Dong Y, Li H, Ma S, et al. Artificial intelli-
gence in healthcare: past, present and future. Stroke Vasc Neurol
2017;2:230-43.
4. LeCun Y, Bengio Y, Hinton G. Deep learning. Nature 2015;521:
436-44.
5. Jordan MI, Mitchell TM. Machine learning: trends, perspectives,
and prospects. Science 2015;349:255-60.
6. Thrall JH, Li X, Li Q, Cruz C, Do S, Dreyer K, et al. Artificial intelli-
gence and machine learning in radiology: opportunities, challenges,
pitfalls, and criteria for success. J Am Coll Radiol 2018;15:504-8.
7. Joda T, Bornstein MM, Jung RE, Ferrari M, Waltimo T,
Zitzmann NU. Recent trends and future direction of dental
research in the digital era. Int J Environ Res Public Health 2020;
17:1987.
8. Mohammad-Rahimi H, Nadimi M, Ghalyanchi-Langeroudi A,
Taheri M, Ghafouri-Fard S. Application of machine learning in
diagnosis of COVID-19 through X-ray and CT images: a scoping
review. Front Cardiovasc Med 2021;8:638011.
August 2021
Vol 160
Issue 2
190
9. Hung K, Montalvao C, Tanaka R, Kawai T, Bornstein MM. The use
and performance of artificial intelligence applications in dental
and maxillofacial radiology: a systematic review. Dentomaxillofac
Radiol 2020;49:20190107.
10. Brickley MR, Shepherd JP, Armstrong RA. Neural networks: a new
technique for development of decision support systems in
dentistry. J Dent 1998;26:305-9.
11. Lee KS, Jung SK, Ryu JJ, Shin SW, Choi J. Evaluation of transfer
learning with deep convolutional neural networks for screening
osteoporosis in dental panoramic radiographs. J Clin Med 2020;
9:392.
12. Imran A, Jayanthi P, Tanveer S, Gobu SC. Classification of odonto-
genic cysts and tumors—Antecedents. J Oral Maxillofac Pathol
2016;20:269-71.
13. Abdolali F, Zoroofi RA, Otake Y, Sato Y. Automatic segmentation
of maxillofacial cysts in cone beam CT images. Comput Biol Med
2016;72:108-19.
14. Setzer FC, Shi KJ, Zhang Z, Yan H, Yoon H, Mupparapu M, et al.
Artificial intelligence for the computer-aided detection of periap-
ical lesions in cone-beam computed tomographic images. J Endod
2020;46:987-93.
15. Hwang HW, Park JH, Moon JH, Yu Y, Kim H, Her SB, et al. Auto-
mated identification of cephalometric landmarks: Part 2-Might it
be better than human? Angle Orthod 2020;90:69-76.
16. Saelens NA, De Smit AA. Therapeutic changes in extraction versus
non-extraction orthodontic treatment. Eur J Orthod 1998;20:
225-36.
17. Xie X, Wang L, Wang A. Artificial neural network modeling for
deciding if extractions are necessary prior to orthodontic treat-
ment. Angle Orthod 2010;80:262-6.
18. Jung SK, Kim TW. New approach for the diagnosis of extractions
with neural network machine learning. Am J Orthod Dentofacial
Orthop 2016;149:127-33.
19. Choi HI, Jung SK, Baek SH, Lim WH, Ahn SJ, Yang IH, et al. Arti-
ficial intelligent model with neural network machine learning for
the diagnosis of orthognathic surgery. J Craniofac Surg 2019;30:
1986-9.
20. Lee K-S, Ryu J-J, Jang HS, Lee D-Y, Jung S-K. Deep convolutional
neural networks based analysis of cephalometric radiographs for
differential diagnosis of orthognathic surgery indications. Appl
Sci 2020;10:2124.
21. Heidb€ uchel KL, Kuijpers-Jagtman AM, Freihofer HP. Facial growth
in patients with bilateral cleft lip and palate: a cephalometric
study. Cleft Palate Craniofac J 1994;31:210-6.
22. Yu HJ, Cho SR, Kim MJ, Kim WH, Kim JW, Choi J. Automated skel-
etal classification with lateral cephalometry based on artificial in-
telligence. J Dent Res 2020;99:249-56.
23. Song Y, Qiao X, Iwamoto Y, Chen Y-w. Automatic cephalometric
landmark detection on X-ray images using a deep-learning
method. Appl Sci 2020;10:2547.
24. Tricco AC, Lillie E, Zarin W, O’Brien KK, Colquhoun H,
Levac D, et al. PRISMA extension for scoping reviews
(PRISMA-ScR): checklist and explanation. Ann Intern Med
2018;169:467-73.
25. Whiting PF, Rutjes AW, Westwood ME, Mallett S, Deeks JJ,
Reitsma JB, et al. QUADAS-2: a revised tool for the quality assess-
ment of diagnostic accuracy studies. Ann Intern Med 2011;155:
529-36.
26. Kunz F, Stellzig-Eisenhauer A, Zeman F, Boldt J. Artificial intelli-
gence in orthodontics: evaluation of a fully automated cephalo-
metric analysis using a customized convolutional neural
network. J Orofac Orthop 2020;81:52-68.
August 2021
Vol 160
Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Mohammad-Rahimi et al
27. Wang CW, Huang CT, Lee JH, Li CH, Chang SW, Siao MJ, et al. A
benchmark for comparison of dental radiography analysis algo-
rithms. Med Image Anal 2016;31:63-76.
28. Oh K, Oh I-S, Van nhat Le T, LEE D-W. Deep anatomical context
feature learning for cephalometric landmark detection. IEEE J Bio-
med Health Inform 2021;25:806-17.
29. Gilmour L, Ray N. Locating cephalometric x-ray landmarks with
foveated pyramid attention. Proceedings of the Third Conference
on Medical Imaging with Deep Learning; 2020 Jul 6-9; Montreal,
Canada. New York: PMLR; 2020.
30. Zhong Z, Li J, Zhang Z, Jiao Z, Gao X. An attention-guided deep
regression model for landmark detection in cephalograms. In:
Shen D, et al., editors. Medical Image Computing and Computer
Assisted Intervention – MICCAI 2019. MICCAI 2019: International
Conference on Medical Image Computing and Computer-Assisted
Intervention; 2019 Oct 13-17; Shenzhen, China. Cham,
Switzerland: Springer; 2019. p. 540-8.
31. Song Y, Qiao X, Iwmoto Y, Chen YW. Semi-automatic cephalo-
metric landmark detection on x-ray images using deep learning
method. In: Liu Y, Wang L, Zhao L, Yu Z, editors. Advances in Nat-
ural Computation, Fuzzy Systems and Knowledge Discovery.
ICNC-FSKD 2019: The International Conference on Natural
Computation, Fuzzy Systems and Knowledge Discovery; 2019
Jul 20-22; Kunming, China. Cham, Switzerland: Springer; 2019.
p. 585-92.
32. Qian J, Cheng M, Tao Y, Lin J, Lin H. CephaNet: an improved faster
R-CNN for cephalometric landmark detection. In: 2019 IEEE 16th
International Symposium on Biomedical Imaging. ISBI 2019: 16th
International Symposium on Biomedical Imaging; 2019 Apr 8-11;
Venice, Italy. New York: IEEE; 2019. p. 868-71.
33. Goutham E, Vasamsetti S, Kishore P, Sardana H. Automatic local-
ization of landmarks in cephalometric images via modified U-Net.
In: 2019 10th International Conference on Computing, Communi-
cation and Networking Technologies. ICCCNT 2019: 10th Interna-
tional Conference on Computing, Communication and
Networking Technologies (ICCCNT); 2019 Jul 6-8; Kanpur, India.
New York: IEEE; 2019. p. 1-6.
34. Dai X, Zhao H, Liu T, Cao D, Xie L. Locating anatomical landmarks
on 2D lateral cephalograms through adversarial encoder-decoder
networks. IEEE Access 2019;7:132738-47.
35. Chen R, Ma Y, Chen N, Lee D, Wang W. Cephalometric landmark
detection by attentive feature pyramid fusion and regression-
voting. In: Shen D, et al., editors. Medical Image Computing and
Computer Assisted Intervention – MICCAI 2019. MICCAI 2019:
International Conference on Medical Image Computing and
Computer-Assisted Intervention; 2019 Oct 13-17; Shenzhen,
China. Cham, Switzerland: Springer; 2019. p. 873-81.
36. Wang S, Li H, Li J, Zhang Y, Zou B. Automatic analysis of lateral
cephalograms based on multiresolution decision tree regression
voting. J Healthc Eng 2018;2018:1797502.
€ Ibragimov B, Xing L. Fully automated quantitative ceph-
37. Arık SO,
alometry using convolutional neural networks. J Med Imaging
(Bellingham) 2017;4:014501.
38. Lindner C, Wang CW, Huang CT, Li CH, Chang SW, Cootes TF. Fully
automatic system for accurate localisation and analysis of cepha-
lometric landmarks in lateral cephalograms. Sci Rep 2016;6:
33581.
39. Vandaele R, Maree R, Jodogne S, Geurts P. Automatic Cephalo-
metric X-Ray Landmark Detection Challenge 2014: a machine
learning tree-based approach. Proceedings of the 14th Interna-
tional Congress for Stereology and Image Analysis; 2015 Jul 7-
10; Liege, Belgium. Liege, Belgium: University of Liege; 2015.
Mohammad-Rahimi et al
40. Lindner C, Cootes TF. Fully automatic cephalometric evaluation
using random forest regression-voting. Proceedings of the IEEE
International Symposium on Biomedical Imaging; 2015 Apr 16-
19; Brooklyn Bridge, NY. New York: IEEE; 2015.
41. Mirzaalian H, Hamarneh G. Automatic globally-optimal pictorial
structures with random decision forest based likelihoods for ceph-
alometric x-ray landmark detection. Proceedings of the IEEE Inter-
national Symposium on Biomedical Imaging. New York: IEEE;
2014. p. 15-27.
42. Chu C, Chen C, Nolte LP, Zheng G. Fully automatic cephalometric
x-ray landmark detection using random forest regression and
sparse shape composition. Proceedings of the IEEE International
Symposium on Biomedical Imaging; 2014 Apr 29-May 2; Beijing,
China. New York: IEEE; 2014. p. 30-8.
43. Nishimoto S, Sotsuka Y, Kawai K, Ishise H, Kakibuchi M. Personal
computer-based cephalometric landmark detection With deep
learning, using cephalograms on the Internet. J Craniofac Surg
2019;30:91-5.
44. Kim H, Shim E, Park J, Kim YJ, Lee U, Kim Y. Web-based fully auto-
mated cephalometric analysis by deep learning. Comput Methods
Programs Biomed 2020;194:105513.
45. Park JH, Hwang HW, Moon JH, Yu Y, Kim H, Her SB, et al. Auto-
mated identification of cephalometric landmarks: part 1—Compar-
isons between the latest deep-learning methods YOLOV3 and SSD.
Angle Orthod 2019;89:903-9.
46. Lobo JM, Jimenez-Valverde A, Real R. AUC: a misleading measure
of the performance of predictive distribution models. Glob Ecol
Biogeography 2008;17:145-51.
47. Makaremi M, Lacaule C, Mohammad-Djafari A. Determination of
the cervical vertebra maturation degree from lateral radiography.
Proceedings 2020;33:30.
48. Chen S, Wang L, Li G, Wu TH, Diachina S, Tejera B, et al. Machine
learning in orthodontics: introducing a 3D auto-segmentation and
auto-landmark finder of CBCT images to assess maxillary constric-
tion in unilateral impacted canine patients. Angle Orthod 2020;90:
77-84.
49. Amasya H, Yildirim D, Aydogan T, Kemaloglu N, Orhan K. Cervical
vertebral maturation assessment on lateral cephalometric radio-
graphs using artificial intelligence: comparison of machine
learning classifier models. Dentomaxillofac Radiol 2020;49:
20190441.
50. Rao GKL, Srinivasa AC, Iskandar YHP, Mokhtar N. Identification
and analysis of photometric points on 2D facial images: a machine
learning approach in orthodontics. Health Technol 2019;9:
715-24.
51. Patcas R, Timofte R, Volokitin A, Agustsson E, Eliades T,
Eichenberger M, et al. Facial attractiveness of cleft patients: a
direct comparison between artificial-intelligence-based scoring
and conventional rater groups. Eur J Orthod 2019;41:428-33.
52. Makaremi M, Lacaule C, Mohammad-Djafari A. Deep learning and
artificial intelligence for the determination of the cervical vertebra
maturation degree from lateral radiography. Entropy 2019;21.
53. K€ok H, Acilar AM, _Izgi MS. Usage and comparison of artificial in-
telligence algorithms for determination of growth and develop-
ment by cervical vertebrae stages in orthodontics. Prog Orthod
2019;20:41.
54. Zhang SJ, Meng P, Zhang J, Jia P, Lin J, Wang X, et al. Machine
learning models for genetic risk assessment of infants with non-
syndromic orofacial cleft. Genomics Proteomics Bioinformatics
2018;16:354-64.
55. Murata S, Lee C, Tanikawa C, Date S. Towards a fully automated
diagnostic system for orthodontic treatment in dentistry. In:
2017 IEEE 13th International Conference on e-Science. e-Science
American Journal of Orthodontics and Dentofacial Orthopedics
191
2017: The 13th IEEE International Conference on e-Science; 2017
Oct 24-27; Auckland, New Zealand. New York: IEEE; 2017. p. 1-8.
56. Budiman JA. Neural network in developing software for Indentify-
ing arch form. Int J Artif Intell Appl 2013;4:1-11.
57. Araya-Dıaz P, Ruz GA, Palomino HM. Discovering craniofacial pat-
terns using multivariate cephalometric data for treatment decision
making in orthodontics. Int J Morphol 2013;31:1109-15.
58. Ni~no-Sandoval TC, Guevara Perez SV, Gonzalez FA, Jaque RA,
Infante-Contreras C. An automatic method for skeletal patterns
classification using craniomaxillary variables on a Colombian pop-
ulation. Forensic Sci Int 2016;261:159.e1-6.
59. Yu X, Liu B, Pei Y, Xu T. Evaluation of facial attractiveness for pa-
tients with malocclusion: a machine-learning technique employ-
ing Procrustes. Angle Orthod 2014;84:410-6.
60. Suhail Y, Upadhyay M, Chhibber A, Kshitiz. Machine learning for
the diagnosis of orthodontic extractions: a computational analysis
using ensemble learning. Bioengineering 2020;7:55.
61. Dharmasena R, Nawarathna L, Nawarathna R. Predicting cessation
of orthodontic treatments using a classification-based approach.
Biom Biostat Int J 2020;61-7.
62. Stehrer R, Hingsammer L, Staudigl C, Hunger S, Malek M, Jacob M,
et al. Machine learning based prediction of perioperative blood loss
in orthognathic surgery. J Craniomaxillofac Surg 2019;47:
1676-81.
63. Omar ZA, Chin SN, Sentian A, Hamzah N, Yassin F. Exploring
contributing features of pre-graft orthodontic treatment of cleft
lip and palate patients using random forests. Trans Sci Technol
2018;5:5-11.
64. Li P, Kong D, Tang T, Su D, Yang P, Wang H, et al. Orthodontic
Treatment Planning based on Artificial Neural Networks [sci rep:
2037]. Sci Rep 2019;9:2037.
65. Riri H, Ed-Dhahraouy M, Elmoutaouakkil A, Beni-Hssane A,
Bourzgui F. Extracted features based multi-class classification of
orthodontic images. Int J Electr Comput Eng 2020;10:3558.
66. Thanathornwong B. Bayesian-based decision support system for
assessing the needs for orthodontic treatment. Healthc Inform
Res 2018;24:22-8.
67. Kajiwara T, Tanikawa C, Shimizu Y, Chu C, Yamashiro T,
Nagahara H. Using natural language processing to develop an
automated orthodontic diagnostic system. arXiv:1905.13601 [Pre-
print]. 2019. Available at: https://arxiv.org/abs/1905.13601.
Accessed June 2, 2021.
68. Zhang Y, Pei Y, Qin H, Guo Y, Ma G, Xu T, et al. Temporal consistent
2D-3D registration of lateral cephalograms and cone-beam
computed tomography images. In: Shi Y, Suk HI, Liu M, editors. Ma-
chine Learning in Medical Imaging. MLMI 2018: 9th International
Workshop on Machine Learning in Medical Imaging; 2018 Sep
16; Granada, Spain. Cham, Switzerland: Springer; 2018. p. 371-9.
69. Pei Y, Liu B, Zha H, Han B, Xu T. Anatomical structure sketcher for
cephalograms by bimodal deep learning. Proceedings of the British
Machine Vision Conference; 2013 Sep 9-13; Bristol, UK. BMVC;
2013.
70. Dur~ao AR, Alqerban A, Ferreira AP, Jacobs R. Influence of lateral
cephalometric radiography in orthodontic diagnosis and treatment
planning. Angle Orthod 2015;85:206-10.
71. McNamara JA Jr. A method of cephalometric evaluation. Am J Or-
thod 1984;86:449-69.
72. Hwang JJ, Lee JH, Han SS, Kim YH, Jeong HG, Choi YJ, et al. Strut
analysis for osteoporosis detection model using dental panoramic
radiography. Dentomaxillofac Radiol 2017;46:20170006.
73. Kamoen A, Dermaut L, Verbeeck R. The clinical significance of er-
ror measurement in the interpretation of treatment results. Eur J
Orthod 2001;23:569-78.
August 2021
Vol 160
Issue 2
192
74. Lee H, Park M, Kim J. Cephalometric landmark detection in dental
x-ray images using convolutional neural networks, Proceedings of
the 2017 SPIE Medical Imaging; 2017. Orlando, FL.
75. Davison AK, Lindner C, Perry DC, Luo W, Cootes TF. Landmark lo-
calisation in radiographs using weighted heatmap displacement
voting. In: Vrtovec T, Yao J, Zheng G, Pozo J, editors. Computa-
tional Methods and Clinical Applications in Musculoskeletal Imag-
ing. MSKI 2018: International Workshop on Computational
Methods and Clinical Applications in Musculoskeletal Imaging;
2018 Sep 16; Granada, Spain. Cham, Switzerland: Springer;
2019. p. 73-85.
76. Noothout JMH, De Vos BD, Wolterink JM, Postma EM,
Smeets PAM, Takx RAP, et al. Deep learning-based regression
and classification for automatic landmark localization in medical
images. IEEE Trans Med Imaging 2020;39:4011-22.
77. Faber J, Faber C, Faber P. Artificial intelligence in orthodontics.
APOS Trends Orthod 2019;9:201-5.
78. Baccetti T, Franchi L, McNamara JA Jr. The cervical vertebral
maturation (CVM) method for the assessment of optimal treat-
ment timing in dentofacial orthopedics. Semin Orthod 2005;
11:119-29.
79. Gabriel DB, Southard KA, Qian F, Marshall SD, Franciscus RG,
Southard TE. Cervical vertebrae maturation method: poor repro-
ducibility. Am J Orthod Dentofacial Orthop 2009;136:478:e1-7;
discussion 478.
80. da Fontoura CS, Miller SF, Wehby GL, Amendt BA, Holton NE,
Southard TE, et al. Candidate gene analyses of skeletal variation
in malocclusion. J Dent Res 2015;94:913-20.
81. Doff MH, Hoekema A, Pruim GJ, Huddleston Slater JJ, Stegenga B.
Long-term oral-appliance therapy in obstructive sleep apnea: a ceph-
alometric study of craniofacial changes. J Dent 2010;38:1010-8.
82. Rajpurkar P, Irvin J, Zhu K, Yang B, Mehta H, Duan T, et al. CheX-
Net: radiologist-level pneumonia detection on chest x-rays with
August 2021
Vol 160
Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Mohammad-Rahimi et al
deep learning. arXiv:1711.05225 [Preprint]. 2017. Available at:
https://arxiv.org/abs/1711.05225. Accessed June 2, 2021.
83. Esteva A, Kuprel B, Novoa RA, Ko J, Swetter SM, Blau HM, et al.
Dermatologist-level classification of skin cancer with deep neural
networks. Nature 2017;542:115-8.
84. Sinko K, Jagsch R, Prechtl V, Watzinger F, Hollmann K,
Baumann A. Evaluation of esthetic, functional, and quality-of-
life outcome in adult cleft lip and palate patients. Cleft Palate
Craniofac J 2005;42:355-61.
85. Bishara SE, Jakobsen JR, Treder J, Nowak A. Arch width changes
from 6 weeks to 45 years of age. Am J Orthod Dentofacial Orthop
1997;111:401-9.
86. Haddadpour S, Motamedian SR, Behnaz M, Asefi S, Bagheban AA,
Abdi AH, et al. Agreement of the clinician’s choice of archwire se-
lection on conventional and virtual models. Angle Orthod 2019;
89:597-604.
87. Araujo TMd, Caldas LD. Tooth extractions in Orthodontics: first or
second premolars? Dent Press J Orthod 2019;24:88-98.
88. Masse J. Will the orthodontic profession disappear. J Dent Sleep
Med 2019;49:6-7.
89. Jheon AH, Oberoi S, Solem RC, Kapila S. Moving towards precision
orthodontics: an evolving paradigm shift in the planning and de-
livery of customized orthodontic therapy. Orthod Craniofac Res
2017;20(Suppl 1):106-13.
90. Filipp FV. Opportunities for artificial intelligence in advancing pre-
cision medicine. Curr Genet Med Rep 2019;7:208-13.
91. Singh A, Sengupta S, Lakshminarayanan V. Explainable deep
learning models in medical image analysis. J Imaging 2020;
6:52.
92. Cheng E, Chen J, Yang J, Deng H, Wu Y, Megalooikonomou V,
et al. Automatic Dent-landmark detection in 3-D CBCT dental
volumes. Annu Int Conf IEEE Eng Med Biol Soc 2011;2011:
6204-7.
Mohammad-Rahimi et al
Supplementary Table I. Definitions of unfamiliar terms used in the manuscript for dentists and orthodontists
Phrase
Artificial intelligence (AI)
Machine learning
Deep learning
Artificial neural network
Transfer learning
Feature Extraction
Data preprocessing
Bayesian network
Overfitting
Training set
Validation set
Test set
Pretrained model/network
End-to-end model
Data augmentation
Ensemble methods
Random forest
Attention-Based Model
American Journal of Orthodontics and Dentofacial Orthopedics
192.e1
Definition
A branch of science that machines try to imitate human intelligence when doing things.1
Part of AI, which its goal is to build computers that can learn through experience.2
A subset of machine learning enables computers to learn from experience without any feature engineering.3
A network with multiple layers of neurons is inspired by the human nervous system.4
A method that improves the model's performance with the help of the knowledge transferred from different but
related source domains.5
The process of extracting be informative and nonredundant features from the data facilitates subsequent learning.6
A step in machine learning that features get transformed or encoded to be easily interpreted by the algorithm.7
A network that uses Bayesian inference for computing probabilities.8
A condition that happens when the model performs well just in the training process, not in the generalization step.9
A dataset in which the model initially is fitted on that.10
A dataset is used for hyparameter tuning by providing an unbiased evaluation of the model.10
A dataset is used for the evaluation of a final model by providing unbiased evaluation.10
A previously trained model on a large dataset, similar to the problem we want to solve.11
Training a complex learning system by running a single model.12
Techniques used to increase the data samples by adding slightly transformed copies of already existing data.13
Techniques that combine multiple models result to improve performance.14
An ensemble learning method consists of multiple decision trees for classification and regression.15
Attention-based models belong to a class of models commonly called sequence-to-sequence models. As the name
suggests, the aim of these models is to produce an output sequence given an input sequence; that is, in general,
of different length.16
August 2021
Vol 160
Issue 2
192.e2 Mohammad-Rahimi et al

Supplementary Table II. Preferred reporting items for systematic reviews and meta-analyses extension for scoping
reviews (PRISMA-ScR) checklist
Reported
Section Item PRISMA-ScR checklist item on page no.
Title
Title 1 Identify the report as a scoping review. 1
Abstract
Structured summary 2 Provide a structured summary that includes (as applicable): background, 1
objectives, eligibility criteria, sources of evidence, charting methods,
results, and conclusions that relate to the review questions and objectives.
Introduction
Rationale 3 Describe the rationale for the review in the context of what is already known. 2, 3
Explain why the review questions/objectives lend themselves to a scoping
review approach.
Objectives 4 Provide an explicit statement of the questions and objectives being addressed 3
with reference to their key elements (eg, population or participants,
concepts, and context) or other relevant key elements used to
conceptualize the review questions and/or objectives.
Methods
Protocol and 5 Indicate whether a review protocol exists; state if and where it can be 3, 4
registration accessed (eg, a Web address); and if available, provide registration
information, including the registration number.
Eligibility criteria 6 Specify characteristics of the sources of evidence used as eligibility criteria 4
(eg, years considered, language, and publication status), and provide a
rationale.
Information sources* 7 Describe all information sources in the search (eg, databases with dates of 4
coverage and contact with authors to identify additional sources), as well
as the date the most recent search was executed.
Search 8 Present the full electronic search strategy for at least 1 database, including 4
any limits used, such that it could be repeated.
Selection of sources of 9 State the process for selecting sources of evidence (ie, screening and 4, 5
evidencey eligibility) included in the scoping review.
Data charting processz 10 Describe the methods of charting data from the included sources of evidence 5
(eg, calibrated forms or forms that have been tested by the team before
their use, and whether data charting was done independently or in
duplicate) and any processes for obtaining and confirming data from
investigators.
Data items 11 List and define all variables for which data were sought and any assumptions 5
and simplifications made.
Critical appraisal of 12 If done, provide a rationale for conducting a critical appraisal of included 5
individual sources of sources of evidence; describe the methods used and how this information
evidence§ was used in any data synthesis (if appropriate).
Synthesis of results 13 Describe the methods of handling and summarizing the data that were 6
charted.
Results
Selection of sources of 14 Give numbers of sources of evidence screened, assessed for eligibility, and 6
evidence included in the review, with reasons for exclusions at each stage, ideally
using a flow diagram.
Characteristics of 15 For each source of evidence, present characteristics for which data were 6
sources of evidence charted and provide the citations.
Critical appraisal within 16 If done, present data on critical appraisal of included sources of evidence (see 7
sources of evidence item 12).
Results of individual 17 For each included source of evidence, present the relevant data that were 7-9
sources of evidence charted that relate to the review questions and objectives.
Synthesis of results 18 Summarize and/or present the charting results as they relate to the review 7-9
questions and objectives.

August 2021
Vol 160
Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Mohammad-Rahimi et al 192.e3

Supplementary Table II. Continued

Reported
Section Item PRISMA-ScR checklist item on page no.
Discussion
Summary of evidence 19 Summarize the main results (including an overview of concepts, themes, and 10-15
types of evidence available), link to the review questions and objectives,
and consider the relevance to key groups.
Limitations 20 Discuss the limitations of the scoping review process. 15, 16
Conclusions 21 Provide a general interpretation of the results with respect to the review 16
questions and objectives, as well as potential implications and/or next
steps.
Funding
Funding 22 Describe sources of funding for the included sources of evidence, as well as None
sources of funding for the scoping review. Describe the role of the funders
of the scoping review.

JBI, Joanna Briggs Institute.

*When sources of evidence (see second footnote) are compiled from, such as bibliographic databases, social media platforms, and Web sites; yA more
inclusive/heterogeneous term used to account for the different types of evidence or data sources (eg, quantitative and/or qualitative research, expert
opinion, and policy documents) that may be eligible in a scoping review as opposed to only studies. This is not to be confused with information
sources (see first footnote); zThe frameworks by Arksey and O'Malley (6) and Levac et al (7) and the JBI guidance (4, 5) refer to the process of data
extraction in a scoping review as data charting; §The process of systematically examining research evidence to assess its validity, results, and rele-
vance before using it to inform a decision. This term is used for items 12 and 19 instead of risk-of-bias (which is more applicable to systematic
reviews of interventions) to include and acknowledge the various sources of evidence that may be used in a scoping review (eg, quantitative
and/or qualitative research, expert opinion, and policy document).
From: Tricco AC, Lillie E, Zarin W, O'Brien KK, Colquhoun H, Levac D, et al. PRISMA extension for scoping reviews (PRISMAScR): checklist and
explanation. Ann Intern Med 2018;169:467-73.

American Journal of Orthodontics and Dentofacial Orthopedics August 2021
Vol 160
Issue 2
192.e4
REFERENCES
1. Suchman LA, Trigg RH. Artificial intelligence as craftwork. Cam-
bridge: Cambridge University Press; 1993.
2. Jordan MI, Mitchell TM. Machine learning: trends, perspectives,
and prospects. Science 2015;349:255-60.
3. Goodfellow I, Bengio Y, Courville A, Bengio Y. Deep Learning.
Cambridge: MIT press; 2016.
4. Dongare A, Kharde R, Kachare AD. Introduction to artificial neural
network. Int J Eng Innov Technol 2012;2:189-94.
5. Zhuang F, Qi Z, Duan K, Xi D, Zhu Y, Zhu H, et al. A comprehensive
survey on transfer learning. Proc IEEE 2021;109:43-76.
6. Sarangi S, Sahidullah M, Saha G. Optimization of data-driven fil-
terbank for automatic speaker verification. Digit Signal Process
2020;104.
7. Kotsiantis S, Kanellopoulos D, Pintelas P. Data preprocessing for
supervised leaning. Int J Comput Sci 2006;1:111-7.
8. Friedman N, Geiger D, Goldszmidt M. Bayesian network classifiers.
Mach Learn 1997;29:131-63.
9. Tetko IV, Livingstone DJ, Luik AI. Neural network studies. 1. Com-
parison of overfitting and overtraining. J Chem Inf Comput Sci
1995;35:826-33.
August 2021
Vol 160
Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Mohammad-Rahimi et al
10. Goodfellow I, Bengio Y, Courville A. Deep Learning. Cambridge:
MIT Press; 2016.
11. Marcelino P. Transfer learning from pre-trained models - Towards
Data Science [Internet]. Towards Data Science. 2018. Available at:
https://towardsdatascience.com/transfer-learning-from-pre-trained
-models-f2393f124751. Accessed April 30, 2021.
12. Bojarski M, Del Testa D, Dworakowski D, Firner B, Flepp B, Goyal P,
et al. End to end learning for self-driving cars. arXiv:1604.07316
[Preprint]. 2016. Available at: https://arxiv.org/abs/1604.07316.
Accessed June 2, 2021.
13. Van Dyk DA, Meng X-L. The art of data augmentation. J Comp
Graph Stat 2001;10:1-50.
14. Pham NT, Foo E, Suriadi S, Jeffrey H, Lahza HFM. Improving perfor-
mance of intrusion detection system using ensemble methods and
feature selection. Proceedings of the Australasian Computer Science
Week Multiconference; 2018 Jan 29-Feb 2; Brisbane, Australia.
15. Ho TK. Random decision forests. Proceedings of 3rd International
Conference on Document Analysis and Recognition; 1995 Aug 14-
16; Montreal, Canada. New York: IEEE; 1995.
16. Chorowski JK, Bahdanau D, Serdyuk D, Cho K, Bengio Y. Atten-
tion-based models for speech recognition. Adv Neural Inf Process
Syst 2015;28:577-85.
ORIGINAL ARTICLE

Biomonitoring of children and

adolescents using orthodontic appliances
made of acrylic resins through
micronucleus testing of exfoliated buccal
and palatal mucosa cells
Joao Pedro Pedrosa Cruz,a Nilton Cesar Nogueira dos Santos,a Matheus Melo Pithon,a,b and
Eneida de Morais Marc
ılio Cerqueirac
Jequi
e, Bahia, and Rio de Janeiro, Brazil

Introduction: The present study aimed to investigate the occurrence of micronuclei and the degenerative nu-
clear alterations indicative of apoptosis and necrosis in the exfoliated buccal and palatal mucosa cells of children
and adolescents using orthodontic appliances made of self-curing acrylic resins. Methods: The micronuclei and
nuclear alterations were evaluated in a minimum of 2000 cells collected from the cheeks and palates of 30 pa-
tients of both sexes, aged between 6 and 12 years. The cell evaluations were performed before appliance instal-
lation and 15 to 21 days after installation. A conditional test was used. In all analyses, the significance level was
5%. Results: No differences in the occurrence of micronuclei or the nuclear degenerative changes indicative of
apoptosis and necrosis were observed in the cheek cells between the 2 time points (P . 0.05). However, eval-
uation of the palate cells revealed a significant increase in the occurrence of micronuclei and the nuclear alter-
ations indicative of apoptosis (P \ 0.01) but not of karyolysis (P . 0.05). Conclusions: Direct contact of
orthodontic appliances made of acrylic resins with the oral mucosa increases the incidence of chromosomal
damage and degenerative nuclear alterations. (Am J Orthod Dentofacial Orthop 2021;-:---)

A
crylic resins have been widely used in dental
practice since the mid-20th century,1 having
been used in the making of removable ortho-
dontic appliances since the orthodontics field was intro-
duced. Orthodontic acrylic resins are organic compounds
that are classified as polymers that consist predomi-
nantly of methyl methacrylate monomers.2
a
Department of Health, State University of Southwestern Bahia, Jequi
e, Bahia,
Brazil.
b
Department of Pediatric Dentistry and Orthodontics, School of Dentistry,
Federal University of Rio de Janeiro, Rio de Janeiro, Brazil.
c
Department of Biological Science, State University of Feira de Santana, Feira de
Santana, Bahia, Brazil.
All authors have completed and submitted the ICMJE Form for Disclosure of
Potential Conflicts of Interest, and none were reported.
Address correspondence to: Joao Pedro Pedrosa Cruz, Department of Health,
Southwestern University of Bahia, Jos
e Moreira Sobrinho Ave, Jequiezinho,
44076-140 Jequi
e, Bahia, Brazil; e-mail, jppcruz@uesb.edu.br.
Submitted, May 2019; revised and accepted, March 2020.
0889-5406/$36.00
Ó 2021 by the American Association of Orthodontists. All rights reserved.
https://doi.org/10.1016/j.ajodo.2020.03.037
Despite the various methods used in the polymeriza-
tion of acrylic resins, the monomer to polymer conver-
sion is never complete.3,4 This results in varying
amounts of methyl methacrylate monomers being
released into the oral cavity during orthodontic appli-
ance use.4,5
In addition to altering the final physical properties of
the resin, these residual monomers can lead to local and
systemic tissue reactions when coming into contact with
saliva and oral soft tissues.4,6 These reactions can man-
ifest as local chemical irritation, hypersensitivity, and/or
inflammation of the mucosa.4 The potential genotoxic
effects of methyl methacrylate are not yet entirely un-
derstood. Consequently, a study in this area is required
because genetic damage at an early age can lead to
the development of health problems throughout life7,8;
biomonitoring is one approach.
Among the methods used for biomonitoring children
and adolescents, the micronucleus test, applied to exfo-
liated buccal mucosa cells, has been a frequent choice
because of the comfort and speed associated with
1
2 Cruz et al
collecting the sample material.9,10 Micronuclei are struc-
tures that result from whole chromosomes or chromo-
somal fragments that, because they do not bind to
spindle fibers, are not included in the nuclei of daughter
cells, remaining instead in the cytoplasm of interphase
cells.11
In this context, the present study aimed to investigate
the occurrence of chromosomal (ie, micronuclei) damage
and the degenerative nuclear changes that indicate
apoptosis and necrosis in exfoliated cells of the buccal
and palatal mucosa of children and adolescents using
orthodontic appliances made of acrylic resin. In addition,
the hypothesis that removable orthodontic appliances
made of acrylic resin do not cause chromosomal damage
or degenerative nuclear changes was investigated.
MATERIAL AND METHODS
Before the start of the study, the project was submit-
ted to the research ethics committee (Feira de Santana
State University), from which it obtained a favorable
opinion. Participation was voluntary, and the partici-
pants were given the freedom to remove themselves
from the study at any time. All participants signed a con-
sent form, which was also signed by their legal guard-
ians. The study was carried out from January 2018 to
April 2019.
Thirty patients, representing men and women and
aged 6-12 years and with dental indications for the
use of removable orthodontic appliances made of self-
curing acrylic resin, participated in the study. The sample
was randomly selected from a cohort of 353 patients
from a school clinic. Definition of the sample size took
previous studies that used the micronucleus test in exfo-
liated oral mucosa cells into consideration.12 In addition,
a minimum of 20 patients has been recommended for
studies in which chromosomal alterations are used as
risk biomarkers in both exposed and control groups.13
The exclusion criteria adopted were (1) use of any
medication, (2) infection (viral or bacterial), (3) oral le-
sions, (4) use of oral antiseptics in the last month, (5)
exposure to x-rays (radiography, tomography or any
radiological procedure in the last 3 months) or to any
other genotoxicity in the same period, and (6) chemo-
therapy or radiation therapy at any point.
After obtaining the superior models, removable or-
thodontic appliances were made. The apparatus con-
sisted of an acrylic plate with posterior and anterior
Hawley retainer clips (Fig 1).
After making the staples and isolating the plaster
model, the acrylic (Jet; Classico, Sao Paulo, Sao Paulo,
Brazil), previously manipulated in a Paladon pot (Nova
OGP, Bragança Paulista, Sao Paulo, Brazil), was adjusted
- 2021
Vol -
Issue - American Journal of Orthodontics and Dentofacial Orthopedics
Fig 1. Model of orthodontic appliance used.
from 2.5 parts polymer to 1.0-part monomer. After
noting that the resin surface was dull, the appliances
were immediately treated with an Ortho Class appliance
(Classico), in which a hydraulic pressure of 25-30 pounds
was applied. This was followed by finishing and polish-
ing using drills and abrasive tips.
The participants were instructed to use the orthodon-
tic appliance continuously (ie, 24 hours per day) and re-
move it only during meals and clean the teeth. Each
participant was given a form on which to note the pe-
riods when the appliance was not in their mouth.
The material for cytologic analysis was collected by
gently scraping the buccal mucosa of the cheek and
the mucosa of the palate with a cytobrush (CooperSurgi-
cal, Trumbull, Conn) rinsing the mouth with filtered wa-
ter.14 These samples were taken immediately before the
installation of the orthodontic appliance (Moment I) and
then 15-21 days after (Moment II).
The collected material was smeared onto a clean glass
slide on which were 2 drops of saline (0.9% NaCl). The
slides holding the samples were dried at room tempera-
ture. After drying, they were fixed in a methanol/acetic
acid solution (Reagen, Rio de Janeiro, Rio de Janeiro,
Brazil) at a 3:1 ratio. After 24 hours, hydrolysis was per-
formed in a hydrochloric acid (5 N HCl) solution for 20 mi-
nutes, followed by washing in distilled water 3 times.
After again drying the slides at room temperature,
the samples were stained with Schiff reagent (US
Merck, Kenilworth, NJ) for 90 minutes and then rinsed
and dried. Finally, 1% Fast Green counterstaining was
performed (Sigma-Aldrich), after which the slides were
washed with absolute alcohol, and coverslips were af-
fixed using Entelan (US Merck, Kenilworth, NJ).
The cytologic analysis was performed in a blind test
by a single, trained evaluator using binocular optical
microscopy at 203, 403, and 1003 magnification.15
Micronuclei were identified as being rounded structures,
distinctly separate from the nucleus, with well-defined
borders and measuring about one-third to one-fifth of
Cruz et al 3

Fig 2. Photomicrographs of exfoliated cells with (A) micronuclei, (B) karyorrhexis, (C) condensed
chromatin, (D) pyknosis, and (E) karyolysis. Magnification 10003.

the size of the nucleus. They possessed a chromatin
structure, were colored similarly to that and were view-
able in the same plane.
At least 2000 cells from each region were analyzed
per individual. Only cells with intact cytoplasm were
counted. For the palate cells, because of keratinization,
there were cells that presented only the stained cyto-
plasm, and so 2 slides had to be analyzed from each in-
dividual to allow the minimum number of cells to be
counted.
In addition to the micronuclei, degenerative nuclear
alterations, indicative of apoptosis (pycnosis—reduced
nuclear structure; karyorrhexis—fragmentation of the
nucleus into small round or oval bodies within the intact
cytoplasm; condensed chromatin–more intensely
stained in relation to the nucleus; and necrosis (or kar-
yolysis)—the dissolution of the nucleus, visually marked
by its absence) were also recorded (Fig 2, A-E).
In addition to these changes, nuclear projections
(broken eggs and nuclear buds) were also recorded.
The broken eggs had rounded structures, color, and
chromatin distribution similar to that of the nucleus
and were connected to the nucleus by a thin chromatin
filament (Fig 3, A). Nuclear buds were observed as pro-
jections resulting from the strangulation of small areas
of the nuclear surface, causing a rounded protuberance
to emerge (Fig 3, B).15
Statistical analysis
The results of the cytologic analysis were evaluated
using the conditional test for the comparison of pro-
portions in rare-event situations—an alternative test
to the chi-square significance test—in line with Fisch-
er's exact test. This test is considered to be suitable
for cytogenetic evaluation when the detection of a
given cytologic alteration requires the computation of

Fig 3. Photomicrographs of exfoliated cells with (A) a nuclear bud and (B) broken eggs. Magnification
10003.

American Journal of Orthodontics and Dentofacial Orthopedics - 2021
Vol -
Issue -

4 Cruz et al

Table I. Distribution of means and standard devia- Table III. Occurrence of broken eggs and nuclear buds
tions of ages according to sex in cheek mucosa at the 2 moments
Sex n % Age 6 standard deviation Moment I Moment II
Male 18 60 10.04 6 1.33 Nuclear
alterations Obs Exp Obs Exp c2; df 5 1; P . 0.05
Female 12 40 8.61 6 2.05
Total 30 100 9.47 6 1.77 Broken egg 44 50.43 57 50.57 1.6379
Nuclear bud 25 30.46 36 30.54 1.9536
Obs, observed; Exp, expected; df, degrees of freedom.
a large number of cells.14,16 In all analyses, the signif-
icance level was 5%.
Micronuclei were significantly more abundant in the
cells collected at Moment II than at Moment I
RESULTS
(c2 5 4.0291; df 5 1; P \ 0.05). These data are pre-
The sample consisted of 30 patients (18 men, 12 sented in Table IV.
women) aged 6-12 years (mean, 9.47 6 1.77 years). Similar results were found in the evaluation of the
Table I shows the distribution of the participants by occurrence of apoptosis, inferred from the sum of asso-
sex and average age. The average daily extent of appli- ciated nuclear alterations (karyorrhexis, condensed chro-
ance use was 19.45 6 2.089 hours. All of the partici- matin, and pycnose) (see Table IV).
pants who were present at the start of the study stayed Kariolysys, nuclear buds, and broken eggs did not
until the study was complete. differ in abundance between the moments. These data
The analysis of differences in the occurrence of mi- are presented in Table V.
cronuclei in cells collected from the mucosa of the
cheeks between Moment I (before installation) and DISCUSSION
Moment II (15-21 days after installation), using the con-
The genotoxic effects of materials used in dentistry
ditional test for the comparison of proportions in situa-
have been investigated in several studies.8,9,16-32
tions of rare events, did not show any significant
Recognizing that materials used in dental practice can
differences (c2 5 0.9782; degree of freedom (df) 5 1;
cause cellular alterations that can translate into DNA
P . 0.05). These data are presented in Table II.
damage makes the study of their genotoxic effects
Evaluation of the occurrence of apoptosis in cells
highly relevant and encourages the biomonitoring of
collected at Moments I and II, considering the sum of
exposed populations as an important preventative
the karyorrhexis, condensed chromatin, and pycnose,
measure. The prevention of DNA damage is especially
did not show any significant differences (c2 5 3.5426;
relevant in younger populations because there is more
df 5 1; P . 0.05). These data are presented in Table II.
time for mutations to occur during their lives.10,19
The occurrence of necrosis inferred using the sum of
The micronucleus testing of exfoliated buccal mu-
associated nuclear degenerative alterations (karyolysis,
cosa cells has been systematically employed in the
karyorrhexis, condensed chromatin, and pyknosis) did
genetic biomonitoring of populations exposed to
not differ between the 2 moments (c2 5 3.3971;
genotoxicity.20-22 This method is particularly useful for
df 5 1; P . 0.05) (see Table II).
the identification of chromosomal damage in children
Similar results were observed when comparing the
because the collection method is noninvasive and its
occurrence of broken eggs and nuclear buds for the 2
execution is simple and easy.9,10,23
moments (Table III).

Table II. Occurrence of micronuclei, apoptosis, and necrosis in cheek mucosa between the 2 moments
Total Micronuclei Micronuclei Apoptosis* Apoptosis* Necrosisy Necrosisy
Moment n cells (obs) (exp) c2 (obs) (exp) c2 (obs) (exp) c2
I 30 61709 28 31.96 0.9782 1415 1465.99 3.5426 1612 1665.22 c2 5 3.3971
II 30 61878 36 32.04 df 5 1 1521 1470.01 df 5 1 1723 1669.78 df 5 1
Total 60 123587 64 64.00 P . 0.05 2936 2936.00 P . 0.05 3335 3335.00 P . 0.05

obs, observed; exp, expected; df, degrees of freedom.

*Skaryorrhexis, condensed chromatin, and pyknosis.
y
Skaryorrhexis, condensed chromatin, pyknosis, and karyolysis.

- 2021
Vol -
Issue - American Journal of Orthodontics and Dentofacial Orthopedics

Cruz et al
Table IV. Occurrence of micronuclei and apoptosis from palate mucosa at the 2 moments
Moment n Total cells Micronuclei (obs) Micronuclei (exp)
I 30 61315 17 23.95
II 30 61554 31 24.05
Total 60 122869 48 48.00
Obs, observed; Exp, expected; df, degrees of freedom.
*Skaryorrhexis, condensed chromatin, and pyknosis.
The mechanisms that lead to the formation of micro-
nuclei occur in the cells of the basal layers of the epithelial
tissue when, to guarantee tissue renewal, they undergo
mitotic division. In the development of the maturation
process of the epithelium, the cells of the basal layer
differentiate and migrate to more superficial layers, in
which they are exfoliated. This process lasts from 10 to
21 days—an interval suitable for sampling when an eval-
uation is necessary before and after a given exposure to
genotoxins.9,24 The period between the sampling adop-
ted in this study was therefore favorable for detection
of the genotoxic damage that might have been induced
by the acrylic resins in orthodontic appliances.
The results of studies in which the genotoxic poten-
tial of acrylic resins, or methyl methacrylate monomers
and their main constituents, have been evaluated are
conflicting.25,26 Although some international agencies
have classified methyl methacrylate as noncarcino-
genic,27 in a longitudinal study investigating cancer
mortality among dental surgeons exposed to this sub-
stance, an increase in deaths from this disease was
observed when compared with general surgeons.27 It
can be concluded that health professionals exposed to
methyl methacrylate may have an increased risk of pre-
mature death from malignant tumors.
In addition, there have been few studies in which the
genotoxicity of these resins has been evaluated through
the analysis of micronuclei in exfoliated buccal mucosa
cells.28,29 Topajiche et al29 described a genotoxic effect
(inferred from a higher incidence of these structures) in
the exfoliated buccal mucosa cells of dentists exposed
to the monomers of acrylic resins when compared with
a control group; however, the authors could not
Table V. Occurrence of karyolysis, broken eggs, and
nuclear buds in palate mucosa at the 2 moments
Nuclear
Moment I Moment II
alterations Obs Exp Obs Exp c2; df 5 1; P . 0.05
Karyolysis 185 203.60 223 204.40 3.3929
Broken eggs 43 50.40 58 50.60 2.1698
Nuclear buds 23 27.45 32 27.55 1.4379
Obs, observed; Exp, expected; df, degrees of freedom.
American Journal of Orthodontics and Dentofacial Orthopedics
5
c2 Apoptosis* (obs) Apoptosis* (exp) c2
4.0291 1047 1126.30 11.1463
df 5 1 1210 1130.70 df 5 1
P \ 0.05 2257 2257.00 P \ 0.01
determine whether the effect was exclusively associated
with the resins because the exposed group had also had
contact with other materials. In contrast, Azhar et al28
found no significant differences in the occurrence of mi-
cronuclei in exfoliated buccal mucosa cells when
comparing exposed professionals to a group of students
and dentists who had not been exposed to these resins.
This is the first study in which exfoliated and kerati-
nized mucosa cells have been directly exposed to ortho-
dontic appliances made of acrylic resin, although the
genotoxic potential of several agents and conditions
have already been evaluated in these cells in many
studies.7,30,31
The greater abundance of micronuclei in the postex-
posure palate cells, as opposed to the cheek cells, suggests
that the methyl methacrylate monomers released into the
oral cavity are not effective in inducing chromosomal
damage but that direct contact with the orthodontic
appliance is sufficient to induce such damage.32,33
The genotoxicity of acrylic resins was also deter-
mined on the basis of the higher abundance of degener-
ative nuclear alterations, indicative of apoptosis, in the
palate mucosa cells in the postexposure period.
Apoptosis is one of the mechanisms of natural tissue
renewal under physiological control but is indicative of
damage when it occurs excessively.34,35
Increased apoptosis rates have also been found in
studies evaluating genotoxicity induced by the contact
of buccal mucosa with fixed orthodontic appliances.36,37
Faccioni et al36 noted significant positive correlations
between metal concentration (ie, cobalt) and the num-
ber of comets and apoptotic cells, and between nickel
concentrations and the number of comet cells.
Cytotoxicity, as assessed in the present study, was not
observed, and further studies are needed to elucidate the
actual cytotoxic potential of the self-curing acrylic resins
used in dentistry. Rose et al38 classified these resins as
slightly cytotoxic on the basis of an investigation of the
effects of methyl methacrylate on fibroblast cultures, us-
ing Mosmann's proliferation-inhibition test; however, af-
ter soaking the appliances in water for 3 days, the same
resins were classified as noncytotoxic or at around the
limit for such a classification. Jorge et al1 reported that
the results of cytotoxicity tests on the acrylic resins used
- 2021
Vol -
Issue -
6 Cruz et al
in dentistry had their limitations in terms of applicability
under clinical conditions; however, the direct-contact tis-
sue reactions already reported in the literature indicate the
need for care in both handling and considering the tech-
nical issues associated with, the use of these materials.
Certain studies have described an association be-
tween increased nuclear projections (broken eggs and
nuclear buds) and genotoxicity in exfoliated buccal mu-
cosa cells,39,40 whereas other studies have found no such
association.41,42 The results obtained from the present
study support the findings of the latter investiga-
tions,41,43 indicating that these structures are associated
with the natural process of cell differentiation.
CONCLUSIONS
This study showed that direct contact of orthodontic
appliances made of acrylic resins with the buccal mucosa
induces an increase in the occurrence of chromosomal
damage and degenerative nuclear alterations. Thus,
the hypothesis that removable orthodontic appliances
made of acrylic resin do not cause chromosomal damage
or degenerative nuclear alterations was rejected. Howev-
er, the results obtained suggest that further studies are
needed to evaluate the real genotoxic potential of ortho-
dontic appliances made of acrylic resins.
AUTHOR CREDIT STATEMENT
Joao Pedro Pedrosa Cruz contributed to conceptual-
ization; Nilton Cesar Nogueira dos Santos contributed to
formal analysis and investigation; Matheus Melo Pithon
contributed to methodology and project administration;
Eneida de Morais Marc
ılio Cerqueira contributed to
funding acquisition, supervision, validation, visualiza-
tion, original draft preparation, and review and editing
manuscript.
ACKNOWLEDGMENTS
We thank Professor Ma
ıza Alves Lopes and Professor
Jos
e Roberto Cardoso Meireles for the support by the
Toxicological Genetic Laboratory of the State University
of Feira de Santana. We also thank Professor L
ıvia Maria
Andrade de Freitas for the support with the patients and
Professor Raildo da Silva Coqueiro and Professor Lean-
dro Barros for the suggestions in the statistics.
REFERENCES
1. Jorge JH, Giampaolo ET, Machado AL, Vergani CE. Cytotoxicity of
denture base acrylic resins: a literature review. J Prosthet Dent
2003;90:190-3.
2. Nik TH, Shahroudi AS, Eraghihzadeh Z, Aghajani F. Comparison of
residual monomer loss from cold-cure orthodontic acrylic resins
processed by different polymerization techniques. J Orthod
2014;41:30-7.
- 2021
Vol -
Issue - American Journal of Orthodontics and Dentofacial Orthopedics
3. Alves PV, Lima Filho RM, Telles E, Bolognese A. Surface roughness
of acrylic resins after different curing and polishing techniques.
Angle Orthod 2007;77:528-31.
4. Gonçalves TS, de Menezes LM, Silva LE. Residual monomer of au-
topolymerized acrylic resin according to different manipulation
and polishing methods. An in situ evaluation. Angle Orthod
2008;78:722-7.
€ urk F, Ates B, Malkoc MA, Kelestemur U.
5. Iça RB, Ozt€ € Level of resid-
ual monomer released from orthodontic acrylic materials. Angle
Orthod 2014;84:862-7.
6. Baker S, Brooks SC, Walker DM. The release of residual monomeric
methyl methacrylate from acrylic appliances in the human mouth:
an assay for monomer in saliva. J Dent Res 1988;67:1295-9.
7. Cerqueira EM, Meireles JR, Lopes MA, Junqueira VC, Gomes-
Filho IS, Trindade S, et al. Genotoxic effects of X-rays on kerati-
nized mucosa cells during panoramic dental radiography. Dento
Maxillo Facial Radiol 2008;37:398-403.
8. Cunha AS, Castillo WO, Takahashi CS, K€ uchler EC, Segato RAB, da
Silva LAB, et al. Genotoxic and cytotoxic effects of Haas appliance
in exfoliated buccal mucosa cells during orthodontic treatment.
Angle Orthod 2018;88:590-5.
9. Angelieri F, Carlin V, Martins RA, Ribeiro DA. Biomonitoring of
mutagenicity and cytotoxicity in patients undergoing fixed ortho-
dontic therapy. Am J Orthod Dentofacial Orthop 2011;139:
e399-404.
10. Ceretti E, Feretti D, Viola GC, Zerbini I, Limina RM, Zani C, et al.
DNA damage in buccal mucosa cells of pre-school children
exposed to high levels of urban air pollutants. PLoS One 2014;9:
e96524.
11. Motgi AA, Chavan MS, Diwan NN, Chowdhery A, Channe PP,
Shete MV. Assessment of cytogenic damage in the form of micro-
nuclei in oral epithelial cells in patients using smokeless and
smoked form of tobacco and non-tobacco users and its relevance
for oral cancer. J Cancer Res Ther 2014;10:165-70.
12. Burgaz S, Demircigil GC, Yilmazer M, Ertaş N, Kemaloglu Y,
Burgaz Y. Assessment of cytogenetic damage in lymphocytes
and in exfoliated nasal cells of dental laboratory technicians
exposed to chromium, cobalt, and nickel. Mutat Res 2002;521:
47-56.
13. Au WW, Cajas-Salazar N, Salama S. Factors contributing to dis-
crepancies in population monitoring studies. Mutat Res 1998;
400:467-78.
14. Santos NC, Ramos ME, Ramos AF, Cerqueira AB, Cerqueira EM.
Evaluation of the genotoxicity and cytotoxicity of filling pastes
used for pulp therapy on deciduous teeth using the micronucleus
test on bone marrow from mice (Mus musculus). Mutagenesis
2016;31:589-95.
15. Thomas P, Fenech M. Buccal micronucleus cytome assay. Methods
Mol Biol 2011;682:235-48.
16. Brandao Pde T, Gomes-Filho IS, Cruz SS, Jde P-SS, Trindade SC,
Souza Lda C, et al. Can periodontal infection induce genotoxic ef-
fects? Acta Odontol Scand 2015;73:219-25.
17. Westphalen GH, Menezes LM, Pr
a D, Garcia GG, Schmitt VM,
Henriques JA, et al. In vivo determination of genotoxicity induced
by metals from orthodontic appliances using micronucleus and
comet assays. Genet Mol Res 2008;7:1259-66.
€ urk F, Karatas OH, Yalcin M. Evaluation of the
18. Toy E, Yuksel S, Ozt€
genotoxicity and cytotoxicity in the buccal epithelial cells of pa-
tients undergoing orthodontic treatment with three light-cured
bonding composites by using micronucleus testing. Korean J Or-
thod 2014;44:128-35.
19. Idolo A, Grassi T, Bagordo F, Panico A, De Giorgi M, Serio F, et al.
Micronuclei in exfoliated buccal cells of children living in a cluster
Cruz et al
area of Salento (Southern Italy) with a high incidence of lung can-
cer: the IMP.AIR study. Int J Environ Res Public Health 2018;15:
1659.
20. Bolognesi C, Roggieri P, Ropolo M, Thomas P, Hor M, Fenech M,
et al. Buccal micronucleus cytome assay: results of an intra- and
inter-laboratory scoring comparison. Mutagenesis 2015;30:
545-55.
21. Dutta S, Bahadur M. Cytogenetic analysis of micronuclei and cell
death parameters in epithelial cells of pesticide exposed tea garden
workers. Toxicol Mech Methods 2016;26:627-34.
22. Tadin A, Gavic L, Govic T, Galic N, Zorica Vladislavic N, Zeljezic D.
In vivo evaluation of fluoride and sodium lauryl sulphate in tooth-
paste on buccal epithelial cells toxicity. Acta Odontol Scand 2019;
77:386-93.
23. Aykanat B, Demircigil GC, Buyan N, Baskin E, Gulleroglu K,
Fidan K, et al. Micronuclei and other nuclear anomalies in buccal
epithelial cells of children with chronic kidney disease. Arh Hig
Rada Toksikol 2016;67:317-25.
24. Thomas P, Holland N, Bolognesi C, Kirsch-Volders M, Bonassi S,
Zeiger E, et al. Buccal micronucleus cytome assay. Nat Protoc
2009;4:825-37.
25. Schweikl H, Schmalz G, Spruss T. The induction of micronuclei
in vitro by unpolymerized resin monomers. J Dent Res 2001;80:
1615-20.
26. Yang HW, Chou LS, Chou MY, Chang YC. Assessment of genetic
damage by methyl methacrylate employing in vitro mammalian
test system. Biomaterials 2003;24:2909-14.
27. Albertini RJ. The lower alkyl methacrylates: genotoxic profile of
non-carcinogenic compounds. Regul Toxicol Pharmacol 2017;
84:77-93.
28. Azhar DA, Syed S, Luqman M, Ali AA. Evaluation of methyl meth-
acrylate monomer cytotoxicity in dental lab technicians using
buccal micronucleus cytome assay. Dent Mater J 2013;32:519-21.
29. Topajiche SS, Smitha BR, Murthy SJ, Patil MS, Karthik Kumar R,
Solanki MP. Genotoxic effect of methyl methacrylate and
Nickelcobalt-chromium in dental lab technicians: A micronuclei
and cytomorphometric study of buccal mucosal cells. Toxicol Int
2015;22:46.
30. Sheikh S, Pallagatti S, Grewal H, Kalucha A, Kaur H. Genotoxicity
of digital panoramic radiography on oral epithelial tissues. Quin-
tessence Int 2012;43:719-25.
31. Yang P, Hao S, Gong X, Li G. Cytogenetic biomonitoring in individ-
uals exposed to cone beam CT: comparison among exfoliated
buccal mucosa cells, cells of tongue and epithelial gingival cells.
Dento Maxillo Facial Radiol 2017;46:20160413.
American Journal of Orthodontics and Dentofacial Orthopedics
7
32. Natarajan M, Padmanabhan S, Chitharanjan A, Narasimhan M.
Evaluation of the genotoxic effects of fixed appliances on oral
mucosal cells and the relationship to nickel and chromium concen-
trations: an in-vivo study. Am J Orthod Dentofacial Orthop 2011;
140:383-8.
33. Heravi F, Abbaszadegan MR, Merati M, Hasanzadeh N, Dadkhah E,
Ahrari F. DNA damage in oral mucosa cells of patients with fixed
orthodontic appliances. J Dent (Tehran) 2013;10:494-500.
34. Joshi MS, Verma Y, Gautam AK, Parmar G, Lakkad BC, Kumar S.
Cytogenetic alterations in buccal mucosa cells of chewers of areca
nut and tobacco. Arch Oral Biol 2011;56:63-7.
35. Galluzzi L, Vitale I, Aaronson SA, Abrams JM, Adam D, Agostinis P,
et al. Molecular mechanisms of cell death: recommendations of the
Nomenclature Committee on Cell Death 2018. Cell Death Differ
2018;25:486-541.
36. Faccioni F, Franceschetti P, Cerpelloni M, Fracasso ME. In vivo
study on metal release from fixed orthodontic appliances and
DNA damage in oral mucosa cells. Am J Orthod Dentofacial Orthop
2003;124:687-93: discussion 93-4.
37. Kapadia JM, Agarwal AR, Mishra S, Joneja P, Yusuf AS,
Choudhary DS. Cytotoxic and genotoxic effect on the buccal mu-
cosa Cells of patients undergoing fixed orthodontic treatment. J
Contemp Dent Pract 2018;19:1358-62.
38. Rose EC, Bumann J, Jonas IE, Kappert HF. Contribution to the bio-
logical assessment of orthodontic acrylic materials. Measurement
of their residual monomer output and cytotoxicity. J Orofac Or-
thop 2000;61:246-57.
39. Quadras DD, Nayak USK, Kumari NS, Priyadarshini HR, Gowda S,
Fernandes B. In vivo study on the release of nickel, chromium,
and zinc in saliva and serum from patients treated with fixed or-
thodontic appliances. Dent Res J (Isfahan) 2019;16:209-15.
40. Çelik A, Yildirim S, Ekinci SY, Taşdelen B. Bio-monitoring for the
genotoxic assessment in road construction workers as determined
by the buccal micronucleus cytome assay. Ecotoxicol Environ Saf
2013;92:265-70.
41. Torres-Bugar
ın O, De Anda-Casillas A, Ram
ırez-Mu~ noz MP,
S
anchez-Corona J, Cant
u JM, Z
un~iga G. Determination of diesel
genotoxicity in firebreathers by micronuclei and nuclear abnor-
malities in buccal mucosa. Mutat Res 1998;413:277-81.
42. Bohrer PL, Filho MS, Paiva RL, da Silva IL, Rados PV. Assessment of
micronucleus frequency in normal oral mucosa of patients
exposed to carcinogens. Acta Cytol 2005;49:265-72.
43. Khan AW, Nersesyan A, Knasm€ uller S, Moshammer H, Kundi M.
Nuclear anomalies in exfoliated buccal cells in Pakistani cotton
weavers. Mutagenesis 2015;30:613-9.
- 2021
Vol -
Issue -
ORIGINAL ARTICLE

Bias in a blink: Shedding light on implicit

attitudes toward patients with a cleft lip
Rany M. Bous,a Anthony Lyamichev,b Ashleigh Kmentt,b and Manish Valiathanc
Cleveland, Ohio

Introduction: Previous studies have shown that patients with cleft lip and/or palate may be stigmatized in soci-
ety. The objective of this study was to use an implicit association test to evaluate the subconscious biases of
non–health care providers and orthodontists against patients with a repaired cleft lip (CL). Methods: Respon-
dents participated in an implicit association test. Pictures of patients with CL and controls were shown to partic-
ipants, along with terms representing positive and negative attributes. Participants were prompted to match
pictures to the attributes. The software algorithm detected whether the participants were more likely to associate
CL with positive or negative terms than controls. Demographic information was collected to measure the asso-
ciation between some sociodemographic factors and implicit biases. Results: Of 130 valid participants, 52 were
orthodontists and 78 were non–health care providers. The entire sample displayed a significant implicit bias
against CL (P \0.001). Overall, orthodontists tended to exhibit slightly higher levels of implicit biases against
CL than non–health care providers, but the difference was not significant when controlling for
sociodemographic factors (P 5 0.34). Females showed significantly lower implicit biases against CL than
males (P 5 0.046). Spearman correlations showed that older people and those who reported a more
conservative political affiliation tended to show slightly higher levels of implicit biases against CL (P \0.007).
Conclusions: Orthodontists and non–health care providers showed moderate but significant levels of implicit
biases against patients with clefts. Males, older age groups, and patients with a more conservative political
affiliation tended to exhibit slightly higher levels of biases than females, younger people, and those with a
more liberal political affiliation. (Am J Orthod Dentofacial Orthop 2021;160:200-8)

C
left lip and/or palate (CL/P) is the most common
craniofacial condition, with an incidence rate of
about 1 in 700 births in the United States.1
Although the majority of the affected patients undergo
multiple surgical procedures to close the CL and opti-
mize the esthetics, most of them still endure a visible
facial difference throughout their lives.
The association between appearance and social stigma
is established in social sciences research.2 Seeman and
Goffman3 defined stigma as “The phenomenon whereby
an patient with an attribute which is deeply discredited by
a
Craniofacial and Special Care Orthodontics, Department of Orthodontics, School
of Dental Medicine, Case Western Reserve University, Cleveland, Ohio.
b
School of Dental Medicine, Case Western Reserve University, Cleveland, Ohio.
c
Mt Sinai-Dr Edward Reiter Fellowship Program, Department of Orthodontics,
School of Dental Medicine, Case Western Reserve University, Cleveland, Ohio.
All authors have completed and submitted the ICMJE Form for Disclosure of
Potential Conflicts of Interest, and none were reported.
Address correspondence to: Rany M. Bous, Craniofacial and Special Care Ortho-
dontics, Department of Orthodontics, School of Dental Medicine, Case Western
Reserve University, 9601 Chester Ave, Cleveland, Ohio 44106; e-mail,
rmb184@case.edu.
Submitted, September 2019; revised, March 2020; accepted, April 2020.
0889-5406/$36.00
Ó 2021 by the American Association of Orthodontists. All rights reserved.
https://doi.org/10.1016/j.ajodo.2020.04.023
his/her society is rejected as a result of the attribute.” Pre-
vious studies have shown that patients with CL/P may be
stigmatized in society. Eye-tracking studies have found
that the lip and nose regions among patients with CL
draw more attention than unaffected controls.4 Multiple
studies have shown this cohort to face various forms of
stigmas, being considered outcasts,5 referred to with an-
imal terms,6 and being rated significantly less attractive
and less friendly than their nonaffected peers.7 Patients
with CL/P are less likely to be involved in conversations,8
and they reported encountering daily social challenges,
including stares, comments, and questions about their
facial differences.9,10 Similarly, affected patients are
more likely to be bullied or teased because of their facial
differences.11 These stigmas may extend to the level of
discrimination and may negatively affect their employ-
ment prospects.12 Negative social interactions among
this population were found to be associated with lower
self-esteem, decreased satisfaction with appearance,
and an elevated risk of depression.13
Although many studies have explored explicit biases
against patients with CL, as mentioned above, our un-
derstanding of the implicit (subconscious) biases toward
this cohort remain limited. Understanding implicit biases

200
Bous et al
is an important component in understanding overall
stereotyping behaviors, as a person may have explicit
equalitarian views and yet display subconscious preju-
dices that may affect how one behaves toward a sub-
group of patients.14
The implicit association test (IAT) was developed by
Greenwald et al15 in 1998 to provide a tool that re-
searchers may use to explore hidden biases. Because its
development, the IAT has become a widely used instru-
ment for assessing implicit biases. IATs measure the
speed of response of participants to evaluate the relative
strength with which they associate specific target groups
(stimuli) with positive or negative attributes. For
example, when the IAT is used to measure racial biases,
people typically respond more quickly if positive attri-
butes share the same response key with white racial stim-
uli and negative attributes share the same key with black
racial stimuli than vice versa.15,16
More recently, researchers have shown interest in
evaluating implicit biases among health care pro-
viders.14,17 Multiple studies have reported the presence
of implicit biases among health care providers, but the
evidence is conflicting regarding whether the provider’s
implicit bias impacts the treatment decisions.14,18,19
However, many studies have concluded that physician’s
implicit biases may be associated with poorer interper-
sonal interactions, anticipating lower compliance from
the patient, and a less positive patient's perception of
the encounters with health care providers.17-19
The current study aimed to use an IAT to evaluate if
non–health care providers and orthodontists may hold
subconscious biases against patients with a surgically re-
paired cleft lip (CL) and to assess the association between
these biases and specific sociodemographic factors.
MATERIAL AND METHODS
The protocol of this study was approved by the insti-
tutional review board (# STUDY20181087). Recruited
participants were stratified into 2 groups: orthodontists
and non–health care providers. For this study, everyone
who was not an orthodontist or orthodontic resident
was considered a non–health care provider. Exclusion
criteria: patients with a vision disorder that would inhibit
them from seeing details, patients not fluent in the En-
glish language, participants aged under 21 years, and
patients with a CL and/or palate. We excluded patients
with CL/P as those patients only constitute a small per-
centage of the population, and to avoid introducing a
confounding factor, because previous studies have
shown patients from a specific minority group to favor
their in-group.20 Furthermore, dental students, general
dentists, and physicians were excluded from the study
American Journal of Orthodontics and Dentofacial Orthopedics
201
because they can neither be classified as orthodontists
or non–health care providers.
Enrollment was voluntary, and participants were re-
cruited from an opportunity sample. Recruitment was
conducted over 8 weeks (from April 2019 to June
2019). Non–health care providers were recruited in per-
son, by e-mail, and through the daily University news-
letter. Orthodontists were recruited in person at the
university’s orthodontic department, and directors of
all orthodontic residency programs in the United States
were contacted via e-mail and asked to share the link
to the study with their residents and colleagues. More-
over, orthodontists in a 32-km radius surrounding the
institution were identified from the American Associa-
tion of Orthodontist’s orthodontist locator Web site,
which yielded 38 records, and were subsequently con-
tacted by phone and/or e-mail and asked to participate.
All participants were provided a link to the Web site
where the test was hosted, and they were asked to take
it in a private, quiet place. All data collected were anon-
ymous. A consent form was provided on the Web site
before starting the test.
Demographic information such as sex, highest level
of education, age, race, political affiliation, education
level, and whether the participant was an orthodontist
or orthodontic resident were collected. Age was cate-
gorized into 5 categories as follows: 21-25, 26-35,
36-45, 46-55, and .55 years. Political affiliation was
collected on a scale from 1 to 7 (1 5 strongly liberal;
7 5 strongly conservative). Furthermore, the sociode-
mographic survey inquired about the highest level of
education and whether the participant was a general
dentist, dental student, dental resident, orthodontist,
orthodontic resident, physician, medical resident, or
none of the above.
The IAT test was designed by Inquisit 5 Web Version
(Millisecond, Seattle, Wash).16,21 All the Data were
collected on the Web, hosted at the Inquisit Millisecond
servers.
A total of 16 nonsmiling frontal photographs of con-
senting patients from the university’s orthodontic
department were used: 8 patients with a surgically re-
paired CL (5 males, 3 females) and 8 controls (CON),
matched for age, gender, and ethnicity. CL photographs
were labeled as Target A, whereas CON photographs
were labeled as Target B for the purposes of data anal-
ysis. Eight positive adjectives (attribute A) and 8 negative
adjectives (attribute B) were used to assess whether par-
ticipants were more likely to associate CL with positive or
negative attributes and vice versa. The positive adjectives
were good, confident, achiever, successful, healthy,
friendly, intelligent, and sociable, whereas the negative
August 2021
Vol 160
Issue 2
202 Bous et al

adjectives were bad, failure, misfit, terrible, miserable,

Table I. Sequence and description of blocks used in
unsociable, isolated, and undesirable. The choice of
the current IAT
the adjectives was based on efforts to include a simple
diverse set of attributes that encompass perceived socia- Item assigned to Item assigned to
bility, health, success, and intelligence. No. of left key (E) right key (I)
Block* trials Function response response
In the current study, we used the 7 Block IAT, as
1 20 Practice CL No CL
described by Greenwald et al,22 and designed it using 2 20 Practice Positive attributes Negative attributes
the Picture IAT hosted by Inquisit (Millisecond). The IAT 3 20 Test Positive and CL Negative and no CL
design included 7 blocks in total. Blocks 1, 2, and 5 4 40 Test Positive and CL Negative and no CL
were used as training blocks where data were not 5 20 Practice No CL CL
6 40 Test Positive and no CL Negative and CL
collected. Training blocks are used for calibration and to
7 40 Test Positive and no CL Negative and CL
get the participant acquainted with the test. Blocks 3,4,
6, and 7 are the actual test blocks for which data were *For all the blocks, the attributes and targets alternate on the screen,
collected. When the test begins, the participants will be and the attributes and targets were randomly shown to participants.
asked to press either E on the left side of the keyboard
or I on the right side of the keyboard as quickly as possible whether the differences would be influenced by the
after a random target or attribute is displayed in the middle group’s sociodemographic factors.
of the screen. Table I describes the 7 blocks used in the cur- To assess if certain sociodemographic factors among
rent study. Ultimately, if the attribute is strongly associ- the entire sample were associated with the levels of im-
ated with the target, the time taken to respond to the plicit bias exhibited, the following tests were conducted:
correct answer (E or I) will be faster, and if it is not asso- Mann-Whitney U test to compare the mean D-scores of
ciated with the target, the expected time to respond would males and females and Spearman’s correlation coeffi-
be slower. The time it takes to respond is recorded in mil- cient to assess the association between D-scores and
liseconds, and the software algorithm generates a final D- age group, and D-scores and political affiliation. Spear-
score, which ranges from 2 to 2. In the current study, a man’s correlation coefficient was used to measure the
negative D-score would indicate a preference for CON over correlations because age and political affiliation did
CL (higher association of CL to negative attributes and not show normal distributions (P \0.01), and they
CON to positive attributes), whereas a positive D-score were recorded on an ordinal scale.
would indicate the opposite. More specifically, 0.1 to
0.34 is considered slight bias, 0.35 to 0.63 is moder- RESULTS
ate bias, and less than 0.64 is strong bias.23 A total of 364 participants logged into our Web link.
However, only 144 completed the test. There were 130
Statistical analysis valid responses once participants who are general den-
Statistical analysis was conducted using SPSS statis- tists (n 5 10) and physicians (n 5 4) were excluded.
tical analysis software (version 25; IBM, Armonk, NY). Of the 130 valid participants, 52 were orthodontists
First, we ran the Shapiro-Wilk test to determine the and 78 were non–health care providers. The entire sam-
normality of the distribution of the D-scores, age, and ple included 69 females, 60 males, and 1 preferred not to
political affiliation. Because the D-scores were not nor- answer. Demographics of the participants are displayed
mally distributed (P 5 0.039), we used the following in Table II.
nonparametric tests: a 1-sample Wilcoxon sign rank Overall, responses were faster on the IAT when CL
test was used to compare the D-scores of the entire sam- was paired with negative attributes and CON was paired
ple to a hypothesized control, in which a hypothesized with positive attributes, compared with the reversed
value of zero (neutral responses or no biases) was used pairings (mean D-score, 0.39). Approximately 80% of
as a control. Mann-Whitney U test was performed for participants showed this pattern.
the intergroup comparison to compare the mean D- One-sample Wilcoxon signed rank test showed that
scores of orthodontists and non–health care providers. the entire sample (n 5 130) displayed a moderate,
Because the demographics of the orthodontist and yet statistically significant, implicit bias against patients
non–health care provider groups were different, we sub- with a CL (mean D-score 5 0.39 6 0.47; median
sequently ran an analysis of covariance (ANCOVA) to D-score, 0.44, P \0.001) (Table III).
compare the scores of orthodontists and non–health The ANCOVA comparing the D-scores between
care providers again while controlling for the age, orthodontists (adjusted mean D-score, 0.43)
gender, and political affiliation, in an attempt to identify and Non–health care providers (adjusted mean

August 2021
Vol 160
Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Bous et al 203

Table II. Demographics of participants Table III. One-sample Wilcoxon signed rank test
comparing the IAT scores of the entire sample against
Demographics
a hypothesized control set at 0
Gender
Males (n 5 60) 95% CI
Females (n 5 69)
Prefer not to answer (n 5 1) Group Mean (SD) Median Lower Upper P value
Age group Entire sample* 0.39 (0.47) 0.44 0.47 0.31 \0.0001y
21-26 (n 5 35) H0 0 0
26-35 (n 5 52)
Note. H0 set at 0; where 0 5 no bias toward any group.
36-45 (n 5 12)
CI, confidence interval; SD, standard deviation; H0, hypothesized
46-55 (n 5 10)
control.
.55 (n 5 21)
*One-sample Wilcoxon signed rank test.
Highest level of education y
Statistically significant at P \0.05, according to Wilcoxon signed
High school graduate (n 5 2)
rank test.
College student (n 5 13)
Associate’s degree (n 5 4)
Bachelor’s degree (n 5 22)
Master’s degree (n 5 30) Table IV. One-way analysis of covariance comparing
Other advanced degree (n 5 9)
Graduate school (n 5 14)
the IAT scores of orthodontists vs non–health care
DMD (n 5 32) providers while controlling for age, gender, and polit-
PhD (n 5 4) ical affiliation as covariates
Race
White (n 5 91) 95% CI
Adjusted
African American (n 5 11)
Group Mean (SD) mean F Lower Upper P
Asian (n 5 11)
Orthodontists 0.51 (0.43) 0.43 0.749 0.57 0.29 0.384*
Native American (n 5 1)
Laypeople 0.31 (0.48) 0.35 0.46 0.24
Other (n 5 16)
Political affiliation CI, confidence interval; SD, standard deviation.
Strongly liberal (n 5 12) *Not statistically significant at P \0.05, after controlling for age,
Moderately liberal (n 5 39) gender, and political affiliation.
Slightly liberal (n 5 23)
Middle (n 5 34)
Slightly conservative (n 5 14)
Moderately conservative (n 5 6) Table V. Mann-Whitney U test comparing the IAT
Strongly conservative (n 5 2) scores of orthodontists vs laypeople and comparing
Orthodontist status the scores of males vs females
Laypeople (n 5 78)
Orthodontist or orthodontic resident (n 5 52) 95% CI
Adjusted
Group Mean (SD) mean Lower Upper P value
D-score 5 0.35) showed no statistically significant dif- Orthodontistsy 0.51 (0.43) 5.1 0.38 0.64 0.02*
Laypeople 0.31 (0.48) 0.35
ferences between the 2 groups (P 5 0.38) when control- Malesy 0.47 (0.46) 0.44 0.32 0.002 0.046*
ling for age, gender, and political affiliation (Table IV). Females 0.31 (0.45) 0.4
However, Mann-Whitney U test showed that, in the
CI, confidence interval; SD, standard deviation.
absence of controlling for the confounding sociodemo- *Statistically Significant at P \0.05, according to Mann-Whitney
graphic factors, orthodontists exhibited significantly U test.
y
higher levels of implicit biases against CL than non– Mann-Whitney U test.
health care providers (mean D-scores 5 0.51 6 0.43
and 0.31 6 0.48, respectively; P 5 0.02) (Table V). that older people and those that reported a more conser-
Females showed significantly lower levels of implicit vative political affiliation tended to show slightly higher
biases against CL than males (mean D-score, levels of implicit biases against CL.
0.31 6 0.45 and 0.41 6 0.46, respectively;
P 5 0.046) (Table V). Spearman correlations showed a DISCUSSION
weak but statistically significant negative correlation be- The primary aim of the current study was to assess
tween D-scores and age (r 5 0.24, P 5 0.007) whether patients with CL are perceived differently in so-
(Table VI; Fig 1) and D-scores and political affiliation ciety. Our findings were in accordance with our hypoth-
(r 5 0.23, P 5 0.007) (Table VI; Fig 2), indicating esis, as the entire sample showed moderate levels of

American Journal of Orthodontics and Dentofacial Orthopedics August 2021
Vol 160
Issue 2
204
Table VI. Spearman correlations between D-scores
and age group and D-scores and political affiliation
Demographic factor Correlation to D-score (r) P value
Political affiliation 0.234 0.007*
Age group 0.235 0.007*
*Statistically significant at a P \0.05.
subconscious associations of CL with negative attributes
and CON with positive attributes.
Studies have shown that IAT may be a beneficial tool
in studying implicit biases. The IAT has shown satisfac-
tory internal consistency, ranging from 0.7 to 0.9,24 and
moderate test-retest reliability over several studies,
which tended to be considerably higher than other mea-
sures of implicit associations.24,25 One strength of the
IAT is that it often reveals associations that participants
may not be aware of or would prefer not to reveal, sug-
gesting that it is more resistant to deliberative alter-
ations, or Hawthorne effects, compared with explicit
measures.24 In addition, IATs were shown to have mod-
erate levels of predictive validity for discriminatory
behavior and satisfactory convergent and discriminant
validity in relation to explicit self-measures, concluding
that the 2 measures are related but distinct attitude
constructs.24
Stone and Wright26 used an IAT to compare the per-
ceptions of the general public to people with facial dif-
ferences or those in a wheelchair and found that people
with a facial difference were perceived more negatively
than those in a wheelchair. Similarly, Grandfield et al27
reported a statistically significant implicit preference
for those with clear skin compared with those with a
skin disorder. In contrast to our findings, Roberts
Fig 1. Correlation plot between age group and D-score.
August 2021
Vol 160
Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Bous et al
et al28 found no significant differences between the
IAT scores when comparing those with various facial dif-
ferences (such as CL, burn scars, etc) to matched con-
trols. However, their study was conducted in Australia,
and the difference between their findings and the find-
ings of the current study may be attributed to regional
differences or the fact that they included various forms
of disfigurement, whereas we only included CL.28 In a
recent study, Hartung et al29 found their participants
to display implicit negative biases against disfigured
faces, although they did not show an explicit bias.
They established neural evidence for their findings
through functional magnetic resonance imaging scans,
which found disfigured faces to evoke lower neural re-
sponses in the anterior cingulate and medial prefrontal
cortex, which may reflect inhibition of empathy.29
Former studies have found that pictures of people
with a facial disfigurement may elicit an emotional
response of disgust or revulsion.30-32 Researchers have
proposed that facial disfigurement may be interpreted
as a possible sign of disease and that the emotion of
disgust may be an inherent response to avoid a patient
who may be contagious, contributing to prejudicial
behaviors.33-35 Therefore, it is possible that some of
the negative associations toward people with a cleft
may be attributed to an emotional feeling of disgust
rather than pure bias.
In our study, orthodontists showed slightly higher
levels of implicit bias against patients with CL than
non–health care providers, although the scores of the
2 groups did not statistically significantly differ when
controlling for sociodemographic confounding factors.
The slightly elevated level of bias among orthodontists
may be partially attributed to the fact that orthodontists
are hyperaware of facial features, as they are trained to
Bous et al
Fig 2. Correlation plot between political affiliation and D-score.
detect facial differences more easily so that their eyes
may focus on the area surrounding the mouth more.
The other more significant reason for the difference, as
revealed by the ANCOVA, was the unequal distribution
of the demographic features of the non–health care pro-
viders and orthodontists, as our orthodontist group
included a higher percentage of more conservative polit-
ical affiliations, older age groups, and slightly higher
percentage of males, all factors that were associated
with higher levels of bias. Thus, it may be possible that
orthodontists may hold higher levels of bias than non–
health care providers, but mostly because of their socio-
demographic factors and not merely because of their
specialty.
Notably, orthodontic treatment is a key component
in the multidisciplinary approach to treating patients
with clefts, as patients with a CL/P often undergo exten-
sive orthodontic treatment to address their condition,
and they often interact with orthodontists over many
years.36 Thus, the perception of the patient and the qual-
ity of their orthodontic visit interactions may be of
utmost importance in maintaining the patient’s compli-
ance and achieving satisfaction with treatment. This be-
comes even more important when we consider the
processes with which clinicians make decisions. Accord-
ing to cognitive psychology studies, humans have 2
distinct decision-making mechanisms.37,38 The analytic
system is primarily responsible for making slow, delib-
erate, controlled, rational conscious decisions, whereas
the heuristic system is responsible for automatic, fast,
and intuitive decisions.37,38 The latter system may be
American Journal of Orthodontics and Dentofacial Orthopedics
205
affected with implicit biases, which may reflect on the
clinician’s decision making and attitude, especially in
fast-paced environments with numerous demands that
require fast, automatic decisions, such as an orthodontic
practice.37,38 Implicit bias may cause health care
providers to unintentionally make assumptions about
their patients on the basis of stereotypes, such as having
lower compliance expectations, assuming an exaggera-
tion of symptoms, accepting different treatment
standards, or may lead to poorer patient-doctor
interactions.17-19 Although implicit biases may conflict
with the health care provider’s explicit beliefs, they
could potentially impact treatment decisions, patient
interactions, and overall health outcomes and thus
warrant further investigations.
Our results showed that females were slightly less
likely to exhibit implicit biases, which is consistent with
previous reports in the literature.39,40 Other studies that
used the IAT in health care found that female physicians
tend to show lower racial biases than their male counter-
parts.41 Researchers have proposed a few theories that
may explain this gender effect. The theory of gendered
prejudice postulates that men display higher levels of
negative prejudices because of gender roles, superior
physical build, or societal positions of power.40 Moreover,
women tend to exhibit higher levels of empathy than
males, making them more accepting of other people’s dif-
ferences.42,43 A third possible explanation is social domi-
nance theory, which suggests that men have a greater
desire to dominate others than women do, which may
be reflected in their social group attitudes.44,45 In
August 2021
Vol 160
Issue 2
206
contrast, Pruett et al46 found no gender differences when
evaluating implicit biases toward disabilities, implying
that gender differences in prejudice may be multifactorial
and not consistent across all the scales.
In the current study, participants from an older age
group tended to show slightly higher levels of implicit
bias toward CL. Similarly, other studies found that older
people tend to show increased levels of various, but
mostly racial, explicit, and implicit prejudices, although
the results were inconsistent.44,47 Although some re-
searchers argue that this effect may be attributed to
generational differences, others have suggested that
the increased levels of prejudice among older partici-
pants may not be necessary because of holding stronger
preferences, but because of a decreased capacity for in-
hibiting their automatic responses, as a result of age-
related declines in inhibition capacity.44,47 In contrast,
other studies found no age-related differences in racial
implicit or explicit biases.48
Exploring the relationship between political affilia-
tion and implicit biases, we found that participants
with a more conservative political affiliation tended to
show slightly higher levels of bias toward CL than partic-
ipants who reported a more liberal political affiliation.
Similar results were reported by a meta-analysis, which
found conservative subjects to show stronger implicit
and explicit social group preferences than liberal
subjects.44 It is worth noting that the results of this
meta-analysis showed that liberals held the same social
preferences as conservative subjects, just more weakly.44
Previous functional magnetic resonance imaging
research has shown conservatives to display a height-
ened emotional disgust response to repulsive pictures
compared with liberal subjects.49 This may provide 1
possible explanation for the increased levels of implicit
biases shown by the more conservative participants in
the current study, as previous studies have shown pic-
tures of people with facial disfigurement to elicit an
emotional response of disgust.30-32,50 Another possible
explanation is the social dominance orientation theory,
which refers to one’s degree of desire to dominate
among social groups.45 Pratto et al45 have reported
that sociopolitical conservatism was positively correlated
with one’s social dominance orientation, which was in
turn correlated with prejudicial behavior against other
social groups, and thus may be 1 possible explanation
to our findings.45 However, the findings from our study
should be interpreted with caution, keeping in mind that
the correlations between D-scores and age and political
affiliation were weak, although statistically significant.
Although many studies have explored the explicit
attitudes of people toward patients with clefts, we are
unaware of previous studies that evaluated implicit
August 2021
Vol 160
Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Bous et al
attitudes toward this population. We used a well-
known instrument to explore implicit biases which may
provide additional insights into how patients with CL
are perceived in society by revealing associations that
participants may not be aware of or would prefer not
to reveal for political correctness.25 IATs were found to
be more effective than explicit measures when assessing
socially sensitive topics, such as the topic under study
here.24
Moreover, we attempted to include the views of or-
thodontists to try and understand how clinicians may
perceive this population. Orthodontists are involved in
the care of patients with a cleft from infancy to adult-
hood, and thus we considered them to be an essential
group to study. However, it is important to point out
that not all orthodontists are routinely involved in treat-
ing patients with clefts, which makes it possible that
studying the implicit attitudes of craniofacial orthodon-
tists may yield different results. Future studies may eval-
uate how these implicit biases may correlate to the
orthodontists’ treatment decision making and patient
interactions.
A limitation of our study was an unequal distribution
of the demographic features of the sample and that the
sample was not robust enough to yield a high number of
participants within each subgroup. In the same vein, the
demographics of non–health care providers and ortho-
dontists were different, as our orthodontist group
included a higher percentage of more conservative polit-
ical affiliations, older age groups, and a slightly higher
percentage of males. In addition, we were unable to
study the effects of the participants’ ethnicity because
the sample was predominantly white (70%). Finally,
because of limitations in the recruitment process, the
majority of the participants were college-educated and
thus may not reflect the general views of the entire pop-
ulation. Thus, it is important to note that the findings of
this study may be considered exploratory. Future studies
with larger, more diverse samples may be needed to
determine the differences among subgroups with more
confidence and overcome the limitations above.
One may argue that the IAT in a research experiment
is no more than a manual dexterity or hand-eye coordi-
nation test.51 Gatewood et al14 refuted this claim,
emphasizing the importance of proper explanation and
framing of implicit bias and IAT testing before partici-
pants take it. On the whole, results of IAT testing tend
to show an inherent preference for the more socially
valued group over the lesser. Although the IAT construct
is highly defendable, it is crucial to note its limitations.
Researchers have warned against the use of an IAT as a
definitive personal diagnosis mechanism. However,
research points to the fact that the IAT is useful in so
Bous et al
far as a recognition tool to expose possible implicit ten-
dencies and to start the conversation for further discus-
sion, debriefing, and conscious modification, for implicit
bias can be unlearned.52
Findings from the current study support an argument
for raising awareness among the general population and
clinicians regarding the implicit biases that one may hold
against patients with cleft. It is essential that orthodon-
tists who treat patients with CL are introspective of their
own biases to avoid having these biases reflect on their
treatment decisions and daily interactions with this
potentially vulnerable group of patients. Although
some authors have argued that because patients may
be unaware of their implicit biases, they should not be
blamed or held responsible for them.53 Holroyd54 pro-
posed that our responsibility may lay in being attentive
to our actions and introspective of our biases to try
and prevent them from manifesting as discriminatory
actions or attitudes. As clinicians, it is crucial that we
care for our patients regardless of their differences.
Because our automatic implicit attitudes may affect
our perceptions and actions, it may be advantageous
to incorporate specific bias training and awareness ses-
sions in clinicians’ curriculums.38 Concern about the po-
tential negative implications of implicit bias influencing
clinical decisions may be partially mitigated with the
hope of debiasing coaching. Gatewood et al14 piloted
an education program among nursing students
involving a preparatory lesson on the nature of implicit
bias, an assessment of each patient’s own implicit bias
via IAT, and finally, a self-reflection discussion activity.
This experimental activity resulted in overall positive
feedback about increasing awareness of bias from the
students involved. It furthermore affirmed that solely
completing an IAT test is insufficient to confront implicit
bias with the intention of positive change but may have
to be supplemented with prior education on the nature
of bias as well as debriefing discussions on how to pro-
cess and improve on an often difficult topic to reveal and
for which to profess ownership.14
Being mindful of one’s biases is an essential step to
minimizing the incidence of any discriminatory behav-
iors against this cohort of patients, and the potential
adverse effects on their quality of life.54
CONCLUSIONS
Orthodontists and non–health care providers tend to
show moderate but significant levels of implicit biases
against patients with CL. Males, patients from an older
age group, and patients with a more conservative polit-
ical affiliation tend to exhibit slightly higher levels of
biases than females, younger people, and those with a
American Journal of Orthodontics and Dentofacial Orthopedics
207
more liberal political affiliation. Future studies
with more diverse sample sizes are needed for a better
understanding of the nature of bias toward patients
with a cleft and how it may affect orthodontist-patient
interactions.
AUTHOR CREDIT STATEMENT
Rany Bous contributed to conceptualization, meth-
odology, validation, formal analysis, investigation, re-
sources, writing, and project administration; Anthony
Lyamichev contributed to software, data curation, and
investigation; Ashleight Kmentt contributed to data cu-
ration, visualization, and writing; and Manish Valiathan
contributed to supervision, writing and review, and
funding acquisition.
REFERENCES
1. Tanaka SA, Mahabir RC, Jupiter DC, Menezes JM. Updating the
epidemiology of cleft lip with or without cleft palate. Plast Re-
constr Surg 2012;129:511e-8e.
2. Langlois JH, Kalakanis L, Rubenstein AJ, Larson A, Hallam M,
Smoot M. Maxims or myths of beauty? A meta-analytic and theo-
retical review. Psychol Bull 2000;126:390-423.
3. Seeman M, Goffman E. Stigma: notes on the management of
spoiled identity. Am Sociol Rev 1964;29:77.
4. Morzycki A, Wong A, Hong P, Bezuhly M. Assessing attentional
bias in secondary cleft lip deformities: an eye-tracking study. Cleft
Palate Craniofac J 2019;56:257-64.
5. Kesande T, Muwazi LM, Bataringaya A, Rwenyonyi CM. Preva-
lence, pattern and perceptions of cleft lip and cleft palate among
children born in two hospitals in Kisoro District, Uganda. BMC Oral
Health 2014;14:104.
6. Pausch NC, Herzberg PY, Wirtz C, Hemprich A, Dhanuthai K,
Hierl T, et al. German animal terms for oral cleft deformity: a Leip-
zig survey. J Craniomaxillofac Surg 2012;40:e236-42.
7. Krueckeberg SM, Kapp-Simon KA, Ribordy SC. Social skills of pre-
schoolers with and without craniofacial anomalies. Cleft Palate
Craniofac J 1993;30:475-81.
8. Kapp-Simon KA, Mcguire DE. Observed social interaction patterns
in adolescents with and without craniofacial conditions. Cleft Pal-
ate Craniofac J 1997;34:380-4.
9. Rumsey N. Body image & congenital conditions with visible differ-
ences. In: Cash TF, Pruzinsky T, editors. Body Image: A Handbook
of Theory, Research, and Clinical Practice. New York: Guilford
Press; 2002. p. 226-33.
10. Turner SR, Thomas PW, Dowell T, Rumsey N, Sandy JR. Psycholog-
ical outcomes amongst cleft patients and their families. Br J Plast
Surg 1997;50:1-9.
11. Feragen KB, Stock NM, Sharratt ND, Kvalem IL. Self-perceptions of
romantic appeal in adolescents with a cleft lip and/or palate. Body
Image 2016;18:143-52.
12. Madera JM, Hebl MR. Discrimination against facially stigmatized
applicants in interviews: an eye-tracking and face-to-face investi-
gation. J Appl Psychol 2012;97:317-30.
13. Feragen KB, Stock NM. A longitudinal study of 340 young people
with or without a visible difference: the impact of teasing on self-
perceptions of appearance and depressive symptoms. Body Image
2016;16:133-42.
August 2021
Vol 160
Issue 2
208
14. Gatewood E, Broholm C, Herman J, Yingling C. Making the invis-
ible visible: implementing an implicit bias activity in nursing edu-
cation. J Prof Nurs 2019;35:447-51.
15. Greenwald AG, Mcghee DE, Schwartz JLK. Measuring individual
differences in implicit cognition: the implicit association test. J
Pers Soc Psychol 1998;74:1464-80.
16. Dasgupta N, Mcghee DE, Greenwald AG, Banaji MR.
Automatic preference for White Americans: eliminating
the familiarity explanation. J Exp Soc Psychol 2000;36:
316-28.
17. FitzGerald C, Hurst S. Implicit bias in healthcare professionals: a
systematic review. BMC Med Ethics 2017;18:19.
18. Dehon E, Weiss N, Jones J, Faulconer W, Hinton E, Sterling S. A
systematic review of the impact of physician implicit racial bias
on clinical decision making. Acad Emerg Med 2017;24:895-904.
19. Hall WJ, Chapman MV, Lee KM, Merino YM, Thomas TW,
Payne BK, et al. Implicit racial/ethnic bias among health care pro-
fessionals and its influence on health care outcomes: A systematic
review. Am J Public Health 2015;105:e60-76.
20. Rudman LA, Feinberg J, Fairchild K. Minority members’ implicit at-
titudes: automatic ingroup bias as a function of group status. Soc
Cogn 2002;20:294-320.
21. De Clercq A, Crombez G, Buysse A, Roeyers H. A simple and sensi-
tive method to measure timing accuracy. Behav Res Methods Ins-
trum Comput 2003;35:109-15.
22. Greenwald AG, Nosek BA, Banaji MR. Understanding and using the
implicit association test: I. An improved scoring algorithm. J Pers
Soc Psychol 2003;85:197-216.
23. Greenwald AG, Poehlman TA, Uhlmann EL, Banaji MR. Under-
standing and using the implicit association test: III. Meta-
analysis of predictive validity. J Pers Soc Psychol 2009;97:17-41.
24. Nosek BA, Greenwald AG, Banaji MR. The implicit association test
at age 7: a methodological and conceptual review. In: Bargh JA,
editor. Social Psychology and the Unconscious: The Automaticity
of Higher Mental Processes. London, United Kingdom: Psychology
Press; 2007. p. 265-92.
25. Egloff B, Schmukle SC. Predictive validity of an Implicit Association
Test for assessing anxiety. J Pers Soc Psychol 2002;83:1441-55.
26. Stone A, Wright T. Evaluations of people depicted with facial
disfigurement compared to those with mobility impairment. Basic
Appl Soc Psych 2012;34:212-25.
27. Grandfield TA, Thompson AR, Turpin G. An attitudinal study of re-
sponses to a range of dermatological conditions using the implicit
association test. J Health Psychol 2005;10:821-9.
28. Roberts RM, Neate GM, Gierasch A. Implicit attitudes towards peo-
ple with visible difference: findings from an Implicit Association
Test. Psychol Health Med 2017;22:352-8.
29. Hartung F, Jamrozik A, Rosen ME, Aguirre G, Sarwer DB,
Chatterjee A. Behavioural and neural responses to facial disfigure-
ment. Sci Rep 2019;9:8021.
30. Stone A, Potton A. Emotional responses to disfigured faces and
Disgust Sensitivity: an eye-tracking study. J Health Psychol
2019;24:1191-200.
31. Stone A, Potton A. Emotional responses to disfigured faces: the in-
fluences of perceived anonymity, empathy, and disgust sensitivity.
Basic Appl Soc Psych 2014;36:520-32.
32. Shanmugarajah K, Gaind S, Clarke A, Butler PE. The role of disgust
emotions in the observer response to facial disfigurement. Body
Image 2012;9:455-61.
33. Ackerman JM, Becker DV, Mortensen CR, Sasaki T, Neuberg SL,
Kenrick DT. A pox on the mind: disjunction of attention and mem-
ory in the processing of physical disfigurement. J Exp Soc Psychol
2009;45:478-85.
August 2021
Vol 160
Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Bous et al
34. Schaller M, Neuberg SL. Chapter one - danger, disease, and the na-
ture of prejudice(s). Adv Exp Soc Psychol 2012;46:1-54.
35. Ryan S, Oaten M, Stevenson RJ, Case TI. Facial disfigurement is
treated like an infectious disease. Evol Hum Behav 2012;33:639-46.
36. Santiago PE, Grayson BH. Role of the craniofacial orthodontist on
the craniofacial and cleft lip and palate team. Semin Orthod 2009;
15:225-43.
37. Hicks EP, Kluemper GT. Heuristic reasoning and cognitive biases:
are they hindrances to judgments and decision making in ortho-
dontics? Am J Orthod Dentofacial Orthop 2011;139:297-304.
38. Schnierle J, Christian-Brathwaite N, Louisias M. Implicit bias: what
every pediatrician should know about the effect of bias on health
and future directions. Curr Probl Pediatr Adolesc Health Care
2019;49:34-44.
39. Assari S. Interaction between race and gender on implicit racial
bias against blacks. Int J Epidemiol Res 2018;5:43-9.
40. McDonald MM, Navarrete CD, van Vugt M. Evolution and the psy-
chology of intergroup conflict: the male warrior hypothesis. Philos
Trans R Soc B Biol Sci 2012;367:670-9.
41. Sabin J, Nosek BA, Greenwald A, Rivara FP. Physicians’ implicit
and explicit attitudes about race by MD race, ethnicity, and gender.
J Health Care Poor Underserved 2009;20:896-913.
42. Schulte-R€ uther M, Markowitsch HJ, Shah NJ, Fink GR, Piefke M.
Gender differences in brain networks supporting empathy. Neuro-
image 2008;42:393-403.
43. Hojat M, Gonnella JS, Nasca TJ, Mangione S, Vergare M, Magee M.
Physician empathy: definition, components, measurement, and
relationship to gender and specialty. Am J Psychiatry 2002;159:
1563-9.
44. Nosek BA, Smyth FL, Hansen JJ, Devos T, Lindner NM,
Ranganath KA, et al. Pervasiveness and correlates of implicit atti-
tudes and stereotypes. Eur Rev Soc Psychol 2007;18:36-88.
45. Pratto F, Sidanius J, Stallworth LM, Malle BF. Social dominance
orientation: a personality variable predicting social and political
attitudes. J Pers Soc Psychol 1994;67:741-63.
46. Pruett SR, Chan F. The development and psychometric validation
of the disability attitude implicit association test. Rehabil Psychol
2006;51:202-13.
47. Von Hippel W, Silver LA, Lynch ME. Stereotyping against your will:
the role of inhibitory ability in stereotyping and prejudice among
the elderly. Pers Soc Psychol Bull 2000;26:523-32.
48. Hutchings VL. Change or more of the same? Evaluating racial at-
titudes in the Obama era. Public Opin Q 2009;73:917-42.
49. Ahn WY, Kishida KT, Gu X, Lohrenz T, Harvey A, Alford JR, et al.
Nonpolitical images evoke neural predictors of political ideology.
Curr Biol 2014;24:2693-9.
50. Curtis V, de Barra M, Aunger R. Disgust as an adaptive system for
disease avoidance behaviour. Philos Trans R Soc Lond B Biol Sci
2011;366:389-401.
51. Gonzalez CM, Kim MY, Marantz PR. Implicit bias and its relation to
health disparities: a teaching program and survey of medical stu-
dents. Teach Learn Med 2014;26:64-71.
52. Staats C. 2014 state of the science: implicit bias review. Available
at: https://kirwaninstitute.osu.edu/research/2014-state-science-
implicit-bias-review#::text5With%20the%20results%20in%20
thethe%20Science%3A%20Implicit%20Bias%20Review. Accessed
September 2019.
53. Saul J. Implicit bias, stereotype threat, and women in philosophy.
In: Hutchison K, Jenkins F, editors. Women in Philosophy: What
Needs to Change? Oxford, United Kingdom: Oxford University
Press; 2013. p. 39-60.
54. Holroyd J. Implicit bias, awareness and imperfect cognitions.
Conscious Cogn 2015;33:511-23.
ORIGINAL ARTICLE

Evaluation of antimicrobial potential and

surface morphology in thin films of
titanium nitride and calcium phosphate
on orthodontic brackets

bia Pereira,c

nior,b Monalessa Fa
Licia Pacheco Teixeira,a Leonardo Cabral Gontijo,b Adonias Ribeiro Franco Ju
Ricardo Pinto Schuenck,c and Juliana Malacarne-Zanond
Vit
oria, Esp
ırito Santo, Brazil

Introduction: The goal of this research was to experimentally evaluate the surface morphology and adhesion
capacity of Streptococcus mutans (U159) on brackets with thin films of titanium nitride (TN) and of titanium nitride
doped with calcium phosphate (TNCP). Methods: Twenty-four metallic brackets were equally allocated to 3
groups (n 5 8), according to the type of covering (no covering, TNCP, and TN). The coatings were deposited
by cathodic cage (TNCP and TN groups) and were evaluated by scanning electron microscopy and energy
dispersive x-ray spectrometry. The biofilm formation of S. mutans on the surface of brackets was determined
by crystal violet assay and subsequent optical density quantification. Results: There was homogeneity on
the surface morphology of the tie wing area in all groups, whereas the TNCP group has presented particles in
the slot. After 24 hours, a biofilm of S. mutans was formed in all the observed groups. The optical density ob-
tained in all 3 groups was similar (no covering, 0.347 6 0.042; TNCP, 0.238 6 0.055; TN, 0.226 6 0.057),
with no statistically relevant difference (P 5 0.06). Conclusions: The thin film of TNCP has altered the surface
of the bracket's slot, whereas the coatings of TN and TNCP have not altered the superficial morphology of the tie
wings. The presence of coatings have not influenced the formation of the S. mutans biofilm on the surface of
metallic brackets. (Am J Orthod Dentofacial Orthop 2021;160:209-14)

W
hite spot lesions affect the majority of ortho-
dontic patients. As brackets work as retention
sites, areas of demineralized enamel may
occur when binomial biofilm-sugar is present.1-3 The
antimicrobial activity of some substances has been
widely studied for the purpose of reducing biofilm
formation and, therefore, the development of white
spot lesions and caries. For orthodontic patients,
a
Graduate Program in Clinical Dentistry, Federal University of Esp
ırito Santo,
Vit
oria, Esp
ırito Santo, Brazil.
b
Graduate Program in Metallurgical and Materials Engineering, Federal Institute
of Esp
ırito Santo, Vit
oria, Esp
ırito Santo, Brazil.
c
Department of Pathology, Graduate Program in Infectious Diseases, Federal
University of Esp
ırito Santo, Vit
oria, Esp
ırito Santo, Brazil.
d
Department of Prosthodontics, Graduate Program in Clinical Dentistry, Federal
University of Esp
ırito Santo, Vit
oria, Esp
ırito Santo, Brazil.
All authors have completed and submitted the ICMJE Form for Disclosure of
Potential Conflicts of Interest, and none were reported.
Address correspondence to: Licia Pacheco Teixeira, Graduate Program in Clinical
Dentistry, Federal University of Esp
ırito Santo, Praça Philogomiro Lannes 200/
305, Vit
oria, Esp
ırito Santo 29060-740, Brazil; e-mail, liciapacheco@gmail.com.
Submitted, July 2019; revised and accepted, April 2020.
0889-5406/$36.00
Ó 2021.
https://doi.org/10.1016/j.ajodo.2020.04.024
preventive actions are limited to oral hygiene
instructions, fluoridation, and application of varnishes
and fluorides.4-6
Among all orthodontic materials, brackets play a sig-
nificant role in biofilm retention and, as a consequence,
in enamel demineralization, for they remain attached to
the teeth during orthodontic treatment. In addition,
their complex design results in difficult cleansing.7-9
The incorporation of antimicrobial agents in the
bracket is an option to prevent caries regardless of
patient compliance.
Titanium nitride (TN) coating has been used in
dentistry since the 1980s, but it was only by the year
2000 that its antimicrobial potential began to be stud-
ied, especially in implants.10 Several authors have re-
ported smaller bacteria attachment to surfaces covered
with TN in comparison with control groups.11-14
Amorphous calcium phosphate is a biomaterial used
for dental regeneration. Its addition to orthodontic
bonding materials means an alternative for secondary
prevention of caries, especially in noncooperative pa-
tients who have a high cariogenic potential. In addition,
its known positive effect on the remineralization of areas
209
210
peripheral to brackets,15,16 an inhibitory action on mi-
croorganisms’ growth was recently observed in amor-
phous calcium phosphate-coated titanium.17
The bactericidal action reported in the literature and
the excellent biocompatibility of calcium phosphate and
TN led us to select these experimental thin films for this
research. Thus, the objective of this study was an in vitro
evaluation of the surface morphology and the adhesion
capacity of Streptococcus mutans on brackets with thin
films of titanium nitride doped with calcium phosphate
(TNCP) and of TN on its own.
MATERIAL AND METHODS
Twenty-four stainless steel brackets for mandibular
incisors (Edgewise, 0.022-in slot, area size 10.72 mm2;
Morelli, Sorocaba, Brazil) were equally and randomly
divided into 3 groups according to coating types.
Brackets belonging to the no covering (NC) group did
not receive any coating. In the TNCP and TN groups
brackets were coated with TNCP and TN alone, respec-
tively, through cathodic cage deposition, an innovative
technique based on multiple hollow cathode effects.
This technique presents some advantages compared
with the conventional methods, including the simplicity
of the apparatus involved and formation of homogenous
thin films even in surfaces of complex geometry, such as
orthodontic brackets.18
The depositions were made in a plasma chamber with
a perforated cathodic cage made of commercially pure
titanium (Ti) grade II (100 mm 3 64 mm 3 1 mm), in
a plasma atmosphere of 20% nitrogen and 80%
hydrogen, for 330 minutes with constant pressure and
temperature (106.66 Pa and 240 6 5 C). The brackets
were positioned within the cage, on an insulating sur-
face. The cathodic cage acted as a cathode and plasma
was not formed on the surface of the samples.
Two brackets from each group were used for sam-
ple characterization, whereas the remaining ones were
used in the microbiological evaluation. Surface
morphology of the bracket slot and tie wing areas of
each group were examined using a Carl Zeiss EVO 40
series scanning electron microscope (SEM), at 1,000
X magnification (Carl Zeiss, Jena, Germany). Energy
dispersive x-ray spectroscopy (EDS) analysis was per-
formed, allowing the identification of chemical ele-
ments in the brackets through the interaction
between the primary electron beam and the samples.19
Streptococcus mutans U159 was inoculated in a
brain heart infusion (BHI) medium (Becton Dickinson,
Franklin Lakes, New Jersey) and incubated for 24 hours
at 100 rpm, 37 C and 5% CO2 partial pressure. From
this inoculum, standardized suspensions were prepared
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Teixeira et al
at 108 Colony Forming Unit (CFU)/mL (optical density
[OD]620 5 1.0). OD is a common method for estimating
the concentration of bacteria in a liquid medium, being
related to the number of actual bacterial cells through a
standard curve relating these 2 parameters. CFUs were
verified by plating serial dilutions of each inoculum
onto a BHI medium (Becton Dickinson).
The remaining brackets of each group (NC, TNCP,
and TN) were sterilized at 121 C, 1 atm for 30 minutes
(Stericlan 12; Sandersmedical, Santa Rita, Brazil) and
evaluated after 24 hours of incubation. Each experiment
was conducted in independent triplicates, in 6 indepen-
dent repetitions with a total of 18 tests. Blanks consisted
of brackets put under the same experimental conditions,
but with no microorganisms.
On 96-well microtiter plates, the brackets were sub-
merged in aliquots of 100 mL of S. mutans suspension
in the standardized BHI (or only BHI for blank groups)
for 24 hours at 37 C and 5% carbon dioxide partial pres-
sure. The growth medium did not contain any added
sugar or sucrose.
After the incubation period, the brackets were gently
washed with distilled water and adhered cells were fixed
with 99% methanol (Isofar, Rio de Janeiro, Brazil) for
15 minutes. The brackets were dried and immersed in
a 0.1% crystal violet solution (Isofar, Rio de Janeiro,
Brazil) for 20 minutes. The excess of crystal violet was
removed and each well washed 5 times with distilled wa-
ter. Then, the crystal violet impregnated in the biofilm
was dissolved with the addition of 100 mL of acetic
acid 33% (Isofar) and the OD of the cells measured at
590 nm (BioPhotometer Plus, Eppendorf, Hamburg,
Germany), to quantify the total biomass of the adhered
biofilm. Absorption values from the control group were
subtracted from those obtained from the experimental
groups to eliminate medium interference.20
Statistical analysis
Intraclass correlation coefficient was applied to test
the repeatability of assessments and normality was as-
sessed through the Shapiro-Wilk test. A comparative
analysis of S. mutans biofilm formation and the type
of coating among the groups was analyzed by 1-way
analysis of variance. The level of significance adopted
was a 5 5%. The software Bioestat statistical (version
5.3; Mamirau
a, Bel
em, Brazil) was used for statistical
analysis.
RESULTS
Deposition of thin films resulted in darkened brackets
in the TNCP and TN groups, when compared with the NC
group (Fig 1).
Teixeira et al
Fig 1. Visual aspect of the brackets in groups NC, TNCP,
and TN after the deposition of the correspondent thin films
via cathodic cage (when applied).
Figure 2 depicts scanning electron micrographs of
superficial aspects of the brackets in NC, TNCP, and
TN groups, respectively. A homogeneous surface
morphology could be observed on the bracket tie wings
of all groups. In contrast, Figure 3 shows the presence of
particles of different sizes on the bracket's slot area with
a thin film of TNCP, what does not occur in NC and TN
groups.
Analysis through EDS confirmed the presence of Ti,
calcium, and phosphorus on the surface of brackets
from the TNCP group and Ti on those from the TN
group. Shadows of the elements iron, chrome, manga-
nese, nickel, and silicon are present in all evaluations,
as these are chemical constituents of the brackets
(Table I).
By evaluating the control group we observed that the
samples had not been contaminated throughout the
study. The coefficient of interclass correlation was
considered satisfactory (intraclass correlation coeffi-
cient, 0.667, with intraclass correlation 95%,
P 5 0.085) and the Shapiro-Wilk test revealed that the
data were normally distributed.
Our results suggested no influence of the coating-
type on S. mutans biofilm formation in the different
groups after 24 hours of incubation (P 5 0.06), as is
shown in Table II.
Fig 2. Homogeneous surface morphology on bracket tie wings in groups NC, TNCP, and TN respec-
tively (scanning electron microscopy with an augmentation of 10003).
American Journal of Orthodontics and Dentofacial Orthopedics
211
DISCUSSION
The association between acidogenic bacteria and the
presence of sugar is an essential event in the demineral-
ization of tooth enamel. The presence of orthodontic
braces increases the number of sites for biofilm reten-
tion, which makes appropriate cleaning more difficult
and may favor microbial attachment. Once attachment
has occurred, cellular proliferation may lead to the
developing of a pathogenic biofilm, the main cause for
white spot lesions. This paper aimed to evaluate the
influence of thin film deposition on the surface
morphology and the microbial attachment to metallic
brackets. Reducing S. mutans adherence to brackets
could prevent white spot lesions in orthodontic patients.
Cathodic cage deposition altered the coloring of
metallic brackets, a change inherent to physical deposi-
tion techniques, also noted by other authors.9,21 Because
the esthetic perception of orthodontic braces is related
to the amount of exposed metal,22,23 the darkening of
brackets is an important observation. Cao et al24 used
a sol-gel deposition process in esthetic brackets as a
way to avoid the expected darkening by cathodic pulver-
ization. Coating did inhibit colonies from forming,
although 5 layers of film were necessary. Coatings
made of many layers tend to display lower attachment
to the surface. Despite the esthetic loss, the deposition
by cathodic cage allowed the formation of a single
thin film layer.
The uniformity of bracket surfaces was evaluated by
scanning electron microscopy in 2 different regions (tie
wing and slot area). There was an equivalence on the su-
perficial aspect of the 3 groups in the tie wings area, with
calcium phosphate particles having been observed in the
slot of the TNCP group.
Although bracket tie wings contain the orthodontic
wire through metallic or elastomeric ligatures, the slot
area is of fundamental importance for the tooth's
August 2021  Vol 160  Issue 2
212 Teixeira et al

Fig 3. Scanning electron microscopy of the slot in group TNCP, with an augmentation of 10003.
Arrows show calcium phosphate particles detected. EHT, electron high tension; WD, working distance.

orthodontic movement. Irregularities in the slot area to the qualitative results obtained with SEM in the
may influence the frictional coefficient of the region, im- TNCP group should be the focus of new studies.
pairing the mechanism of sliding.25 Another hypothesis for the presence of particles at
It is important to point out that SEM generates 2- the slot area in the TNCP group may be a lower degree
dimensional images, and the evaluation of the roughness of crystallinity. When analyzing Ti discs coated with cal-
of both slots and brackets should be done quantitatively cium phosphate, Kulkarni Aranya et al17 observed the
in a 3-dimensional manner.25 The quantitative evalua- distinctive morphology of individual microcrystals
tion of roughness in the slot area as well as its relation because the calcifying solutions had different composi-
tions. The antibacterial activity found in all coated discs
was inversely proportional to the degree of crystallinity,
Table I. Composition and concentration of chemical measured through x-ray diffraction analysis. Their re-
elements in the brackets of the different groups sults indicate that a higher crystallinity degree tends to
according to EDS analysis lead to a smaller dissolution rate and, consequently, a
Atomic percentage (%)
smaller effect on the bacterial population.
Calcium phosphate needs to be solubilized to be able
Chemical elements NC TNCP TN to perform its remineralizing effect on the tooth
Fe 59.7 54.8 43.7 enamel.26 However, the degree of solubility necessary
O 0.5 5.6 12.8
to allow a coating with calcium phosphate to act in tooth
Cr 16.1 15.4 12.2
Ni 7.0 6.9 6.2 remineralization is uncertain. Moreover, it is important
C 7.8 10.5 8.9 not to compromise the superficial integrity of the film.
Cu 1.8 1.8 1.6 The long-term effects of coatings made with dental bio-
Mn 1.5 1.4 1.3 materials are still questionable. In addition, the mainte-
Si 1.5 1.1 6.7
nance of mechanical properties in the long-term is yet
S 0.2 0.2 0.2
Ti – 0.9 2.3 uncertain, and gradual liberation may result not only
Ca – 0.2 – in material loss, but, more importantly, such loss could
P – 0.2 – lead to a consequent increase in roughness and bacteria
Ag 3.9 1.0 4.1 attachment.27

August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Teixeira et al
Table II. OD obtained for each group in the presence
of S. mutans after 24 hours incubation
Groups Mean Standard deviation
NC 0.347* 0.047
TNCP 0.238* 0.055
TN 0.226* 0.057
*Statistically significant differences, by 1-way analysis of variance
(P 5 0.06).
The presence of sulfur and silver in all groups was re-
vealed by the EDS analysis. Sulfur is used to ease the
machinability of steel, whereas the presence of silver cor-
responds to the process of bracket welding. The Ti pre-
sent in the TNCP group comes from the Ti cage used
for the deposition of the coating.
We observed lower cellular attachment to the brackets
covered in experimental thin films, with no damage to the
uniformity of the brackets’ surfaces having been detected.
In the present study, we chose to evaluate calcium phos-
phate and TN as experimental thin films because of
their well-documented bactericidal action and excellent
biocompatibility. However, we found no evidence of anti-
microbial characteristics in the experimental thin films
evaluated here, as the total stained biomass in the bio-
films was similar. There were consistently no differences
as to the adherence of S. mutans between groups.
These results are similar to those of Brusca et al28 and
Papaioannou et al,29 and opposite to the results reported
by Rammohan et al,30 who compared the patterns of
attachment and growing of S. mutans on the surface
of metallic, ceramics, and plastic brackets.
Papaioannou et al29 also evaluated the brackets in
the presence of an acquired pellicle, unveiling the impor-
tant role played by the salivary pellicle in reducing differ-
ences on the enamel surface. The pellicle may have either
a positive (owing to the presence of specific receptors) or
a negative (when the surface free energy of adjacent ma-
terials is reduced) influence in the bacterial attach-
ment.31,32 During the initial colonization, the effect of
the acellular layer of adsorbed salivary proteins is posi-
tive because the receptors in the acquired pellicle
interact with the surface of pioneer microorganisms.33
When evaluating brackets not coated in salivary pellicle,
only the properties of the material surfaces can influence
bacterial attachment.31 Therefore, other factors that in-
fluence the formation of biofilm, such as roughness and
surface free energy, should be investigated.
Some points may have influenced the present results
and should be thus reported: though widely used, the
crystal violet assay is not specific and does not allow
evaluation of cellular viability. In addition, the biofilms
American Journal of Orthodontics and Dentofacial Orthopedics
213
evaluated in this study are monospecific and do not
represent a faithful mirror of the oral microbiota, though
they do permit an analysis of the capacity of 2 different
thin films in interfering with the properties of attach-
ment and formation of an S. mutans biofilm on the
brackets. Although they do work as an attempt to imitate
what occurs in a buccal cavity, in vitro studies cannot
completely predict the behavior of materials because of
laboratory limitations and the inability to reproduce all
the factors that can be found intraorally.
Among the different methods for coating metals,
deposition by cathodic cage was the chosen one for
the present study. It consists in a deposition method
developed in 2007 and its industrial scale usage is still
being implemented. Even if the technique has the
advantage of forming a more uniform, thin pellicle—
including on surfaces with a complex geometry such
as orthodontic brackets—further studies are required.
CONCLUSIONS
TNCP and TN thin films did not interfere with the
attachment of S. mutans to the surface of brackets.
Moreover, the presence of a thin film of TNCP was found
to alter slot surface, whereas neither TN nor TNCP
coverings were capable of modifying the surface
morphology of metallic brackets’ tie wings. It appears
that such coatings do not reduce the growth of S. mu-
tans, being, therefore, unable to prevent the bacterial ef-
fect on demineralization as anticipated with the coatings
in the context of this study.
AUTHOR CREDIT STATEMENT
Licia Pacheco Teixeira: Conceptualization, Method-
ology, Writing - original draft. Leonardo Cabral
Gontijo: Resources, Writing - review & editing. Adonias
Ribeiro Franco J
unior: Validation, Resources.
Monalessa F
abia Pereira: Methodology, Validation,
Writing - review & editing. Ricardo Pinto Schuenck:
Conceptualization, Writing - review & editing. Juliana
Malacarne-Zanon: Supervision.
REFERENCES
1. Julien KC, Buschang PH, Campbell PM. Prevalence of white spot
lesion formation during orthodontic treatment. Angle Orthod
2013;83:641-7.
2. Guzm
an-Armstrong S, Chalmers J, Warren JJ. Ask us. White
spot lesions: prevention and treatment. Am J Orthod Dentofacial
Orthop 2010;138:690-6.
3. Cury JA, De Oliveira BH, Dos Santos AP, Tenuta LM. Are fluoride
releasing dental materials clinically effective on caries control?
Dent Mater 2016;32:323-33.
4. Kirschneck C, Christl JJ, Reicheneder C, Proff P. Efficacy of fluoride
varnish for preventing white spot lesions and gingivitis during
August 2021  Vol 160  Issue 2
214
orthodontic treatment with fixed appliances-a prospective ran-
domized controlled trial. Clin Oral Investig 2016;20:2371-8.
5. Fern
andez-Ferrer L, Vicente-Ru
ız M, Garc
ıa-Sanz V, Montiel-
Company JM, Paredes-Gallardo V, Almerich-Silla JM, et al. Enamel
remineralization therapies for treating postorthodontic white-spot
lesions: a systematic review. J Am Dent Assoc 2018;149:
778-86.e2.
6. H€ochli D, Hersberger-Zurfluh M, Papageorgiou SN, Eliades T. In-
terventions for orthodontically induced white spot lesions: a sys-
tematic review and meta-analysis. Eur J Orthod 2017;39:122-33.
7. Ahn SJ, Lee SJ, Kook JK, Lim BS. Experimental antimicrobial or-
thodontic adhesives using nanofillers and silver nanoparticles.
Dent Mater 2009;25:206-13.
8. do Ros
ario Junior AF, Knop LA, Baboni FB, Rymovicz AU,
Tanaka OM, Rosa EA. Differential adhesion of Streptococcus
mutans to metallic brackets induced by saliva from caries-free
and caries-active individuals. J Investig Clin Dent 2011;2:
197-200.
9. Shah AG, Shetty PC, Ramachandra CS, Bhat NS, Laxmikanth SM.
In vitro assessment of photocatalytic titanium oxide surface modi-
fied stainless steel orthodontic brackets for antiadherent and
antibacterial properties against Lactobacillus acidophilus. Angle
Orthod 2011;81:1028-35.
10. Jabbari YSA, Fehrman J, Barnes AC, Zapf AM, Zinelis S,
Berzins DW. Titanium nitride and nitrogen ion implanted coated
dental materials. Coatings 2012;2:160-78.
11. Gr€ossner-Schreiber B, Griepentrog M, Haustein I, M€ uller WD,
Lange KP, Briedigkeit H, et al. Plaque formation on surface modi-
fied dental implants. An in vitro study. Clin Oral Implants Res
2001;12:543-51.
12. Scarano A, Piattelli M, Vrespa G, Caputi S, Piattelli A. Bacterial
adhesion on titanium nitride-coated and uncoated implants: an
in vivo human study. J Oral Implantol 2003;29:80-5.
13. Annunziata M, Oliva A, Basile MA, Giordano M, Mazzola N,
Rizzo A, et al. The effects of titanium nitride-coating on the topo-
graphic and biological features of TPS implant surfaces. J Dent
2011;39:720-8.
14. Jeyachandran YL, Narayandass SK. The effect of thickness of tita-
nium nitride coatings on bacterial adhesion. Trends Biomater Artif
Organs 2010;24:90-3.
15. Skrtic D, Antonucci JM, Eanes ED. Amorphous calcium
phosphate-based bioactive polymeric composites for mineralized
tissue regeneration. J Res Natl Inst Stand Technol 2003;108:
167-82.
16. Paschos E, Geiger FJ, Malyk Y, Rudzki I, Wichelhaus A, Ilie N. Ef-
ficacy of four preventive measures against enamel demineraliza-
tion at the bracket periphery-comparison of microhardness and
confocal laser microscopy analysis. Clin Oral Investig 2016;20:
1355-66.
17. Kulkarni Aranya A, Pushalkar S, Zhao M, Legeros RZ, Zhang Y,
Saxena D. Antibacterial and bioactive coatings on titanium
implant surfaces. J Biomed Mater Res A 2017;105:2218-27.
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Teixeira et al
18. Alves C Jr, de Ara
ujo FO, Ribeiro KJB, da Costa JAP, Sousa RRM, de
Sousa RS. Use of cathodic cage in plasma nitriding. Surf Coatings
Technol 2006;201:2450-4.
19. Zinelis S, Annousaki O, Makou M, Eliades T. Metallurgical charac-
terization of orthodontic brackets produced by metal injection
molding (MIM). Angle Orthod 2005;75:1024-31.
20. Mattos-Graner RO, Napimoga MH, Fukushima K, Duncan MJ,
Smith DJ. Comparative analysis of Gtf isozyme production
and diversity in isolates of Streptococcus mutans with
different biofilm growth phenotypes. J Clin Microbiol 2004;
42:4586-92.
21. Cao B, Wang Y, Li N, Liu B, Zhang Y. Preparation of an orthodontic
bracket coated with an nitrogen-doped TiO(2-x)N(y) thin film and
examination of its antimicrobial performance. Dent Mater J 2013;
32:311-6.
22. Feu D, Catharino F, Duplat CB, Capelli Junior J. Esthetic perception
and economic value of orthodontic appliances by lay Brazilian
adults. Dental Press J Orthod 2012;17:102-14.
23. Rosvall MD, Fields HW, Ziuchkovski J, Rosenstiel SF,
Johnston WM. Attractiveness, acceptability, and value of ortho-
dontic appliances. Am J Orthod Dentofacial Orthop 2009;135:
276: e1-12; discussion 276-7.
24. Cao S, Wang Y, Cao L, Wang Y, Lin B, Lan W, et al. Preparation and
antimicrobial assay of ceramic brackets coated with TiO2 thin films.
Korean J Orthod 2016;46:146-54.
25. Assad-Loss TF, Neves RML, Mucha JN. Composiç~ao qu
ımica e
aspecto superficial do slot de br
aquetes met
alicos. Rev Dent Press
Ortodon Ortop Facial 2008;13:85-96: Portuguese.
26. Uysal T, Amasyali M, Koyuturk AE, Ozcan S. Effects of different
topical agents on enamel demineralization around orthodontic
brackets: an in vivo and in vitro study. Aust Dent J 2010;55:268-74.
27. Øilo M, Bakken V. Biofilm and dental biomaterials. Materials 2015;
8:2887-900.
28. Brusca MI, Chara O, Sterin-Borda L, Rosa AC. Influence of different
orthodontic brackets on adherence of microorganisms in vitro.
Angle Orthod 2007;77:331-6.
29. Papaioannou W, Gizani S, Nassika M, Kontou E, Nakou M. Adhe-
sion of Streptococcus mutans to different types of brackets. Angle
Orthod 2007;77:1090-5.
30. Rammohan SN, Juvvadi SR, Gandikota CS, Challa P, Manne R,
Mathur A. Adherence of Streptococcus mutans and Candida albi-
cans to different bracket materials. J Pharm Bioallied Sci 2012;
4(Suppl 2):S212-6.
31. Ahn SJ, Lim BS, Yang HC, Chang YI. Quantitative analysis of the
adhesion of cariogenic streptococci to orthodontic metal brackets.
Angle Orthod 2005;75:666-71.
32. Ahn SJ, Ahn SJ, Wen ZT, Brady LJ, Burne RA. Characteristics of
biofilm formation by Streptococcus mutans in the presence of
saliva. Infect Immun 2008;76:4259-68.
33. Li F, Weir MD, Fouad AF, Xu HH. Effect of salivary pellicle on antibac-
terial activity of novel antibacterial dental adhesives using a dental
plaque microcosm biofilm model. Dent Mater 2014;30:182-91.
ORIGINAL ARTICLE

A clinically friendly viscoelastic finite

element analysis model of the mandible
with Herbst appliance
Zahra Heidari Zadi,a Amir J. Bidhendi,b Ali Shariati,c and Eung-Kwon Paed
Montreal, Qu
ebec, Canada, Dumfries, and Montclair, Va, and Baltimore, Md

Introduction: As a powerful numerical approximation tool, finite element analysis (FEA) has been widely used to
predict stress and strain distributions in facial bones generated by orthodontic appliances. Previous FEA models
were constructed on the basis of a linear elastic phase of the bone response (eg, elastic bone strains to loading).
However, what is more useful for clinical understanding would be predicting long-term strains and displacements
of bone-segments responding to loading, yet tissue responses are (1) not promptly observable and (2) hard to
predict in nature. Methods: Viscoelastic property of the mandibular bone was incorporated into FEA models to
visualize long-term, time-dependent stress and strain patterns in the mandible after being exposed to orthopedic
stress. A mandible under loading by a Herbst appliance was modeled, and outcomes of the constructed elastic
and viscoelastic models were compared. Results: Patterns and magnitudes of the displacement throughout the
mandible predicted by the viscoelastic model were exhibited in accordance with previous clinical outcomes of
Herbst appliance therapy. The elastic models exhibited similar displacement patterns; however, the magnitude
of the displacements in the models was invariably small (approximately 1 per 100) compared with those outputs
of corresponding viscoelastic models. The corresponding maximum stress level in our viscoelastic mandible
subjected to the Herbst appliance with the same loading was considerably low and relaxed in various regions
when compared with the elastic model. Conclusions: We suggest that a viscoelastic model of the mandible
mimics our general prediction of orthopedic treatment outcomes better than those by elastic models. (Am J
Orthod Dentofacial Orthop 2020;-:---)

T
o measure the clinical effectiveness of an ortho-
dontic appliance, stress exerted by the orthodontic
appliance to the bone needs to be analyzed
because the loading applied to the bone through the
corresponding strain in the soft tissue matrix is respon-
sible for bone remodeling over time.1,2 Throughout the
years, many approaches, such as brittle lacquer, photo-
elasticity, and holography,3 have been used to study
the effects of orthodontic force on bones. Finite element
analysis (FEA) simulates complex biologic structures and
a expanders5 to Class II correctors,6,7 facemasks,8-10 and
Private practice, Dumfries, Va.
b
Department of Plant Science, McGill University, Montreal, Qu
ebec, Canada.
c
Parsa Engineering LLC, Montclair, Va.
d
Department of Orthodontics and Pediatric Dentistry, School of Dentistry, Uni-
versity of Maryland, Baltimore, Md.
All authors have completed and submitted the ICMJE Form for Disclosure of Po-
tential Conflicts of Interest, and none were reported.
Address correspondence to: Eung-Kwon Pae, Department of Orthodontics and
Pediatric Dentistry, School of Dentistry, University of Maryland, 650 W Baltimore
St, Baltimore, MD 20101; e-mail, eungkpae@gmail.com.
Submitted, November 2019; revised and accepted, April 2020.
0889-5406/$36.00
Ó 2020.
https://doi.org/10.1016/j.ajodo.2020.04.017
their biomechanical behaviors under different condi-
tions, and various forces were used in orthodontics for
many decades. In 1984, Williams et al4 first used FEA
as a tool to study the center of rotation of maxillary in-
cisors in relation to elastic properties of the periodontal
ligament. However, owing to a lack of reports on mate-
rial properties and oversimplified geometries, most
studies using FEA were remote from clinical applica-
tions. Using FEA, many researchers attempted to show
stress and strain distributions on the maxilla and
mandible generated by orthodontics appliances such as
temporary anchorages devices.11
As studies using FEA gain popularity in the ortho-
dontic field, we note that the validity of their research re-
lies on the soundness of input data. Therefore, defining
proper material properties, accuracy in geometry, appli-
cable forces, and boundary conditions, as well as types of
analysis depending on the nature of the problem, are
crucial for the soundness of a model. Digital Imaging
and Communications in Medicine files converted from
3-dimensional (3D) cone-beam computed tomography

1
2 Zadi et al

(CBCT) images can conveniently be exported to an FEA

Table. Material properties of the mandible
software package, which enables researchers to build in-
dividual models to test. Material properties of cortical bone
Modulus of
Despite all major advancements in the field, most elasticity (E) Poisson ratio (y) Retardation period (t)
previous studies examining the clinical effects of ortho- 13,700 MPa 0.3 50 min
dontic appliances employed a set of linear elastic mate-
rial properties to simulate behaviors of viscoelastic bone
tissue showing nonlinear behaviors.12,13 With elastic For viscoelastic models, the Prony series parameters
models, it is impossible to calculate displacements of were chosen for the study because the material behaves
the bone over a long period of treatment time, which close to the Maxwell model.17 The retardation period
is crucial for studying the end results of orthodontic ap- was assumed to be 50 minutes,18 and the treatment
pliances. In addition, the propagation of stress and its re- period was assumed to be 4300 hours, which approxi-
sultants in the bone during the treatment process cannot mates 6 months. Although the Kelvin-Voigt model is
be captured in an elastic model because, in general, an usually used in the literature to capture the viscoelastic
elastic model can only express instantaneous behaviors behavior of the cortical bone, it appears reasonable for
of the bone. In contrast, a viscoelastic model factors a clinical orthodontic treatments to assume that the
time-effect into account.13 Thus, a viscoelastic model cortical bone shows significant plastic behavior as
express changes over time. In this study, we aimed to well.19 This behavior in the cortical bone is often
compare structural behaviors of a viscoelastic model of observed and suggested by researchers in biomechanical
the mandible with Herbst appliance in action compared engineering.20
with those of a linear elastic model of the mandible. After defining the material properties, the boundary
Herbst appliances were chosen because there are ample conditions were imposed as the translational lock in all
clinical data which would help examine and understand the global directions for elements on the condylar heads
the results of this study.14,15 of the mandible, as shown in Figure 1. To study the
deformation of the mandible, degrees of freedom (ie,
MATERIAL AND METHODS movement of the node in 1 or more directions x, y,
and z) must be restricted to avoid rigid body motion.
Briefly, the captured geometry of a mandible from Such constraints are termed as boundary conditions. In
CBCT in Digital Imaging and Communications in Medicine addition, a static force of 40 N in the vertical and 60 N
image was converted to a stereolithography (STL) file. in the horizontal direction was imposed through masti-
Then, the file was transformed into a 3D computer- catory muscles to the first molar regions on the mandible
aided design model that can be interpreted by Finite
element method (FEM) software (Abaqus; Dassault Sys-
temes Simulia Corp, Providence, RI) for analysis. The ge-
ometry was then discretized (meshed), and material
properties (such as elastic or viscoelastic) were assigned,
and then finally, the applied forces from the appliance as
well as the boundary conditions were specified.
A full volume CBCT image on a boy aged 10 years
with skeletal Class II was used. The CBCT was taken
with CareStream CS9300 (Carestream Health, Rochester,
New York, NY) at the following settings: 90 kVp; 5 mA;
exposure time of 8 seconds; resolution of 180 micro-
meter. The Ma 4. Invivo software (Anatomage, San
Jose, Calif) was used to derive the file in STL format.
The STL files were then transferred to Abaqus, which is
a FEA software package with pre- and postprocessing
capabilities. Using the 3D image obtained from the
CBCT, the geometry was imported and meshed using
other modules of Abaqus. Material properties16—
namely, Young modulus (or modulus of elasticity) and Fig 1. Boundary conditions applied to the areas in or-
Poisson ratio—were assigned in accordance with the ange color on the condylar heads. Arrows in yellow indi-
values in the Table. cate the directions of force applied.

- 2020  Vol -  Issue - American Journal of Orthodontics and Dentofacial Orthopedics

Zadi et al 3

Fig 2. Elastic modeling displacement. The symphysis

area exhibits a displacement of approximately 0.04 mm.

to simulate the loading from the Herbst appliance. These

forces were adopted from the average bite-force, and the
forces from the masticatory muscles reported previously
in the literature.16,21
A pair of models using 2 separate material properties
were constructed and analyzed. For 1 model, the elastic
properties of the cortical bone were incorporated for the
immediate elastic response. To account for the time
parameter, the identical model except for viscoelastic ma-
terial properties was constructed and analyzed. One of the
purposes of this phase was to see the progression of how
stresses and strains emerge and propagate through the
course of treatment during Herbst appliance treatment.
In addition, the magnitude of displacement, as well as
the change in stress levels due to creep, were of interest
in this study. In the postprocessing phase (which is the
last phase of an FEA study), we observed the results of
applied forces in the form of displacements and stress dis-
tributions. These results were visualized by contour maps
in colored magnitudes of the outputs. One should note
that the value of the displacement obtained from our
elastic model could be very small because of the high
elastic modulus of the cortical bone that we adopted. In
addition, the elastic response would reflect a transitory
spontaneous behavior of the mandible because creep or
stress relaxation were not a part of the interpretation.
RESULTS
In both models, we exhibit principal stresses, von
Mises stresses, and the magnitude of each displacement.
von Mises stress was calculated to predict yielding of
the bone.22,23 The total number of nodes employed for
Fig 3. Elastic model: A, shows von Mises stress contour;
B, shows Maximum principal stress contour. Colormaps
are in megapascals (MPa).
the model was 26,260, along with the total number of
elements of 52,235.
Figure 2 shows the magnitude of displacements
when the elastic material properties were incorporated.
Various colors in different areas represent the range of
their corresponding displacements; red indicates an
instant and maximum displacement, and blue indicates
minimum displacements. The FEM analysis revealed that
the maximum displacement resulted from the Herbst
appliance in the elastic model was 0.04 mm at the chin
in a forward and downward direction.
Figures 3, A and B show the von Mises model and the
maximum principal stresses, respectively. The colors
represent different ranges of stress values in various re-
gions (ie, red for the maximum and blue for the mini-
mum values). FEM analysis exhibits the areas of stress
in the mandible immediately after force applications.

American Journal of Orthodontics and Dentofacial Orthopedics - 2020  Vol -  Issue -

4 Zadi et al
The areas of lighter green color indicating the highest
stresses in our elastic model were approximately
1.8-3.6 MPa for von Mises and 2.2-4.3 MPa for
maximum principal stresses. The stress patterns were
more concentrated at the buccal and lingual ramus areas
and around the first molars.
Figure 4 shows the results of stresses and displace-
ments when a visco-analysis with viscoelastic material
properties were used. In this model, the maximum
magnitude of displacement (in red color) at the chin
and the alveolar bone area for incisors was approxi-
mately 3.1 mm. (See Video, available at www.ajodo.
org) for dynamic visualization of the achieved theoretical
displacement over 6 months.
In Figure 5, each panel represents the stresses distrib-
uted in the viscoelastic model immediately after the
loading begins (A and C) and at the end of the assumed
treatment period (B and D). One can observe that the
areas under high stress (green areas) were reduced in
size and more localized over time. Note that areas covered
in green were reduced in B and D compared with A and C.
Although the amount of force should remain constant
during treatment, there appears stress relaxation in the
areas (as stress receded from the origin of loading). At
the beginning and end of treatment, stressed areas accu-
mulated in the condylar neck and the alveolar bone
around posterior teeth, as shown in Figure 5.
DISCUSSION
The treatment effects expressed in the mandible by
functional appliances have long been analyzed clinically
using cephalometric images based on statistics. Being a
Fig 4. Viscoelastic modeling displacement. The symphy-
sis area exhibits a displacement of approximately
3.067 mm. See the Video (available at www.ajodo.org)
for animation.
- 2020  Vol -  Issue - American Journal of Orthodontics and Dentofacial Orthopedics
clinical study, the longitudinal cephalometric analysis
takes time and provides study results only after comple-
tion of the course of treatment. The accuracy of the
cephalometric approach relies on the number of subjects
participating in the study. Furthermore, the cephalo-
metric analysis only measures a mixture of displacement
and growth at the end of the treatment, which offers a
prediction of the treatment effects.
In contrast, FEA is an objective tool that can attest to
the mechanical effect (independent of growth effects) of
the functional appliances associated with the shape and
size of anatomic structures immediately. The majority of
the FEM studies to date have modeled the mandible or
maxilla as an elastic material.5-7,24-27 However, as
briefed earlier, it may only make sense to model the
bone as a viscoelastic material.20,28 The fundamental
drawbacks of modeling bone as an elastic material are
as follows: (1) elastic modeling only provides instanta-
neous stress and displacement magnitudes at the time
of applying forces to the model. Thus, the actual
behavior of the bone over time cannot be examined;
and (2) results of the elastic model does not simulate
clinical outcomes because values indicating instanta-
neous displacement are invariably very small as shown
in this study.
In this article, we offered more clinically acceptable
models with viscoelastic elements showing more clini-
cally relevant mechanical properties of the mandible.
To test the practicality of our method, we chose to simu-
late the effects of a Herbst appliance on the mandible.
The rationale behind this decision was the abundance
of Herbst-based clinical studies for the Herbst appliance
had been widely accepted as a Class II functional thera-
peutic tool. This study compared the elastic and visco-
elastic FEA results and validated the magnitude of
displacement approximates their corresponding clinical
values expected from Herbst appliance therapy. The re-
sults of this comparison substantiate why the field
should begin incorporating the viscoelastic properties
of the bone for FEA models.
Our Herbst appliance in the models exhibited a
downward and forward displacement of the mandible
as the condyles immobilized in the condylar sockets.
This outcome, serendipitously, follows the results
demonstrated in the majority of previous publications
that studied Herbst appliance clinically. For instance,
Pancherz et al29,30 found that the chin was displaced
anteriorly and inferiorly by 1.9-3.1 mm. In our visco-
elastic model, we assumed the treatment with Herbst
was 6 months, and the amount of force exerted by the
appliance was constant. As a result, our model achieved
3.1 mm of displacement anteriorly and inferiorly at the
chin point at the end of treatment. In contrast, the value
Zadi et al 5

Fig 5. Viscoelastic models for von Mises stress and Maximum principal stress. Each panel indicates:
A, von Mises stresses at the beginning of treatment; B, von Mises stresses at the end of treatment;
C, maximum principal stresses at the beginning of treatment; D, maximum principal stresses at the
end of treatment. Colormaps are in megapascals (MPa).

of displacement that our elastic model achieved was
0.044 mm.
CONCLUSIONS
The objective of this study was to introduce a visco-
elastic FEA analysis of the mandible and to examine if
the viscoelastic model may yield more clinically compat-
ible outcomes. Results from our viscoelastic model
demonstrated that Herbst appliance results in a down-
ward and forward chin displacement if the patient uses
the appliance for 6 months (see Video for animation,
available at www.ajodo.org). Although this modeling
effect does not represent a real result of growth
modification, this FEA model using viscoelastic
elements visualizes an average clinical outcome. A
time-independent elastic model cannot demonstrate
this displacement effect. This study validates that
viscoelastic models of the bone are superior and more
clinically relevant than elastic models for FEM analysis
in our field.
A viscoelastic model should provide a better mathe-
matical interpretation representing an outcome of the
orthopedic effects of orthodontic appliances. This claim
may be somewhat preposterous because we did not pro-
vide any statistical evidence for the magnitude of
displacement. The downward and forward displacement
of the chin point of 3.1 mm appears to be empirical, but
ought to inevitably be hypothetical because this magni-
tude is based on 1 mandible and several given boundary
conditions. Nonetheless, this report opens a door for
further studies in a de novo direction. Examining the
reliability of an FEA model constructed with viscoelastic
elements maybe 1 example for such future studies, and
testing whether the initial shape of a mandible could

American Journal of Orthodontics and Dentofacial Orthopedics - 2020  Vol -  Issue -

6 Zadi et al
affect expected results of orthopedic appliances could be
another.
SUPPLEMENTARY DATA
Supplementary data associated with this article can
be found, in the online version, at https://doi.org/
10.1016/j.ajodo.2020.04.017.
REFERENCES
1. Sun Z, Lee E, Herring SW. Cranial sutures and bones: growth and
fusion in relation to masticatory strain. Anat Rec A Discov Mol Cell
Evol Biol 2004;276:150-61.
2. Gr€ oning F, Liu J, Fagan MJ, O'Higgins P. Why do humans have
chins? Testing the mechanical significance of modern human sym-
physeal morphology with finite element analysis. Am J Phys
Anthropol 2011;144:593-606.
3. Trivedi S. Finite element analysis: a boon to dentistry. J Oral Biol
Craniofac Res 2014;4:200-3.
4. Williams KR, Edmundson JT. Orthodontic tooth movement ana-
lysed by the finite element method. Biomaterials 1984;5:347-51.
5. Işeri H, Tekkaya AE, Oztan O, Bilgiç S. Biomechanical effects of
rapid maxillary expansion on the craniofacial skeleton, studied
by the finite element method. Eur J Orthod 1998;20:347-56.
6. Chaudhry A, Sidhu MS, Chaudhary G, Grover S, Chaudhry N,
Kaushik A. Evaluation of stress changes in the mandible with a
fixed functional appliance: a finite element study. Am J Orthod
Dentofacial Orthop 2015;147:226-34.
7. Maruo IT, Maruo H, Saga AY, de Oliveira DD, Argenta MA,
Tanaka OM. Tridimensional finite element analysis of teeth
movement induced by different headgear forces. Prog Orthod
2016;17:18.
8. Gautam P, Valiathan A, Adhikari R. Skeletal response to maxillary
protraction with and without maxillary expansion: a finite element
study. Am J Orthod Dentofacial Orthop 2009;135:723-8.
9. Gautam P, Valiathan A, Adhikari R. Stress and displacement pat-
terns in the craniofacial skeleton with rapid maxillary expansion:
a finite element method study. Am J Orthod Dentofacial Orthop
2007;132:5.e1-11.
10. Park JH, Bayome M, Zahrowski JJ, Kook YA. Displacement and
stress distribution by different bone-borne palatal expanders
with facemask: a 3-dimensional finite element analysis. Am J Or-
thod Dentofacial Orthop 2017;151:105-17.
11. Sarmah A, Mathur AK, Gupta V, Pai VS, Nandini S. Finite element
analysis of dental implant as orthodontic anchorage. J Contemp
Dent Pract 2011;12:259-64.
12. Zhao S, Arnold M, Ma S, Abel RL, Cobb JP, Hansen U, et al. Stan-
dardizing compression testing for measuring the stiffness of
human bone. Bone Joint Res 2018;7:524-38.
13. Ojanen X, Tanska P, Malo MKH, Isaksson H, V€a€an€anen SP,
Koistinen AP, et al. Tissue viscoelasticity is related to tissue compo-
sition but may not fully predict the apparent-level viscoelasticity in
human trabecular bone - an experimental and finite element
study. J Biomech 2017;65:96-105.
- 2020  Vol -  Issue - American Journal of Orthodontics and Dentofacial Orthopedics
14. Pancherz H. The mechanism of Class II correction in Herbst appli-
ance treatment. A cephalometric investigation. Am J Orthod 1982;
82:104-13.
15. Pancherz H. The Herbst appliance–its biologic effects and clinical
use. Am J Orthod 1985;87:1-20.
16. Srirekha A, Bashetty K. Infinite to finite: an overview of finite
element analysis. Indian J Dent Res 2010;21:425-32.
17. Park SW, Schapery RA. Methods of interconversion between linear
viscoelastic material functions. Part I—a numerical method based
on Prony series. Int J Solids Struct 1999;36:1653-75.
18. Wu Z, Ovaert TC, Niebur GL. Viscoelastic properties of human
cortical bone tissue depend on gender and elastic modulus. J
Orthop Res 2012;30:693-9.
19. Kim DG, Huja SS, Lee HR, Tee BC, Hueni S. Relationships of viscos-
ity with contact hardness and modulus of bone matrix measured
by nanoindentation. J Biomech Eng 2010;132:024502.
20. Johnson TPM, Socrate S, Boyce MC. A viscoelastic, viscoplastic
model of cortical bone valid at low and high strain rates. Acta
Biomater 2010;6:4073-80.
21. Kikuchi M, Korioth TWP, Hannam AG. The association among
occlusal contacts, clenching effort, and bite force distribution in
man. J Dent Res 1997;76:1316-25.
22. Yang C, Wang C, Deng F, Fan Y. Biomechanical effects of cortico-
tomy approaches on dentoalveolar structures during canine retrac-
tion: a 3-dimensional finite element analysis. Am J Orthod
Dentofacial Orthop 2015;148:457-65.
23. Tamura N, Takaki T, Takano N, Shibahara T. Three-dimensional
finite element analysis of bone fixation in bilateral sagittal split
ramus osteotomy using individual models. Bull Tokyo Dent Coll
2018;59:67-78.
24. Tominaga JY, Tanaka M, Koga Y, Gonzales C, Kobayashi M,
Yoshida N. Optimal loading conditions for controlled movement
of anterior teeth in sliding mechanics. Angle Orthod 2009;79:
1102-7.
25. Ozaki H, Tominaga JY, Hamanaka R, Sumi M, Chiang PC,
Tanaka M, et al. Biomechanical aspects of segmented arch me-
chanics combined with power arm for controlled anterior tooth
movement: a three-dimensional finite element study. J Dent Bio-
mech 2015;6:1758736014566337.
26. Schneider J, Geiger M, Sander FG. Numerical experiments on long-
time orthodontic tooth movement. Am J Orthod Dentofacial Or-
thop 2002;121:257-65.
27. V
asquez M, Calao E, Becerra F, Ossa J, Enr
ıquez C, Fresneda E.
Initial stress differences between sliding and sectional mechanics
with an endosseous implant as anchorage: a 3-dimensional finite
element analysis. Angle Orthod 2001;71:247-56.
28. Roychowdhury A, Pal S, Saha S. Stress analysis of an artificial
temporal mandibular joint. Crit Rev Biomed Eng 2000;28:
411-20.
29. Pancherz H, Malmgren O, H€agg U, Omblus J, Hansen K. Class II
correction in Herbst and Bass therapy. Eur J Orthod 1989;11:
17-30.
30. Pancherz H, Bjerklin K, Hashemi K. Late adult skeletofacial
growth after adolescent Herbst therapy: a 32-year longitudinal
follow-up study. Am J Orthod Dentofacial Orthop 2015;147:
19-28.
Zadi et al
APPENDIX
MATHEMATICAL IDEALIZATIONS AND
SIMPLIFICATIONS
Viscoelasticity is the property of materials showing both
viscous and elastic behavior when undergoing deforma-
tion. Viscous materials strain continuously with time
when stress is applied to them, whereas elastic materials
show strain when they are stretched, and as soon as the
stress is released, they return to their original form.1
Stress relaxation and creep phenomena are 2 important
properties of viscoelastic materials.2
To model the behavior of materials, mathematical ideal-
izations and simplifications are essential. For instance, in
the elastic materials, the elasticity can be demonstrated
as shown in the Supplementary Figure in which E is
the modulus of elasticity, and s is the stress. This model
shows that the rate of displacement (strain) is the same
as the force, but as soon as the force is removed, the
spring recoils to its original shape. Viscosity or plasticity
can be shown as in Supplementary Figure in which s is
the stress that is applied to the dashpot, and h is the vis-
cosity. This model shows that as the forces are applied,
the dashpot opens (get displaced), and once the forces
are removed, it will stay in that position. It is good to
note that the speed of displacement is controlled by
the amount of force and h.
Different mathematical idealizations models such as
Maxwell, Kelvin-Voigt, and standard linear solid model
can be used to predict and simulate the viscoelastic ma-
terial response under different loading conditions. As
shown in Supplementary Figure, the behavior of visco-
elastic materials is idealized using combinations of
springs and dampers. The elastic part is characterized us-
ing a spring and the viscosity part by a damper (dashpot)
with the properties and behaviors that are already ex-
plained. This combination can be in series and in paral-
lel. Each component in the series has equal forces,
whereas each component of a parallel system has a
similar displacement. Two of the most popular combina-
tions that have been used to describe the behavior of
American Journal of Orthodontics and Dentofacial Orthopedics
6.e1
bone are the Voigt and Maxwell models, which are
shown in Supplementary Figures, C and D, respectively.
The Maxwell model, which is shown below in
Supplementary Figure C is a damper and a spring in se-
ries. This model shows that the applied force on the
spring and dashpot is equal; however, when the force
is removed, the spring recoils, but the dashpot does
not. Under initial displacement (strain), this model al-
lows for gradual stress relaxation while under constant
loading, gradual displacement occurs—a phenomenon
that is also known as creep.2 The other model that is
very well known and used to capture the behavior of bio-
logical materials such as bone is Kelvin-Voigt model in
which a damper and a spring are acting in parallel. In
this model, the force is more in the dashpot initially until
it is open fully, then it is maximum in the spring. It is
good to note that generalized models such as the stan-
dard linear solid model (see Supplementary Fig E) and
generalized Maxwell models are used to capture mate-
rials viscoelastic behavior as well.
In our study, we hypothesized to use the Prony parame-
ters according to the Maxwell model to allow both stress
relaxation and creep behavior. This choice was made on
the basis of the intuition that in the clinical setting, we
can observe relatively large displacements even by
applying small forces during a long period. The Maxwell
model was chosen for modeling the viscoelastic behavior
of the mandibular bone. In addition, recent studies have
shown plastic properties for bone in line with our current
assumption.3
REFERENCES
1. Johnson TP, Socrate S, Boyce MC. A viscoelastic, viscoplastic model
of cortical bone valid at low and high strain rates. Acta Biomater
2010;6:4073-80.
2. Wikipedia. Viscoelasticity. Available at: https://en.wikipedia.org/
wiki/Viscoelasticity. Accessed July 30, 2020.
3. Zhao S, Arnold M, Ma S, Abel RL, Cobb JP, Hansen U, et al. Stan-
dardizing compression testing for measuring the stiffness of human
bone. Bone Joint Res 2018;7:524-38.
- 2020  Vol -  Issue -
6.e2 Zadi et al

Supplementary Fig. Constitutive models of linear viscoelasticity: A, spring; B, damper (dashpot);

C, Maxwell model; D, Kelvin-Voight model; E, standard solid model; F, generalized Maxwell model.

- 2020  Vol -  Issue - American Journal of Orthodontics and Dentofacial Orthopedics

ORIGINAL ARTICLE

Long-term follow-up of late maxillary

orthopedic advancement with the
Liou-Alternate rapid maxillary expansion-
constriction technique in patients with
skeletal Class III malocclusion
Maria Costanza Meazzini,a Camilla Torre,b Alessandro Cappello,b Roberto Tintinelli,b Elena De Ponti,c
and Fabio Mazzolenid
Monza and Milan, Italy

Introduction: The objective of this study was to evaluate short and long-term results of the application of the
alternate rapid maxillary expansion/constriction (Alt-RAMEC) technique in patients with skeletal Class III
malocclusion. Methods: Forty-two white patients were consecutively treated with the Alt-RAMEC technique.
The average age of the patients was 12.7 6 1.6 years (range, 9.4-15.9 years) before protraction. The
average age at long-term follow-up was 19.4 6 2.8 years (range, 17.2-26.9 years). A sample of nontreated
patients with Class III malocclusion from the archives of orthognatic surgery in our hospital was used as a
control group. The initial records were matched for sex, the severity of Class III malocclusion, and age
(mean, 12.1 6 1.4 years; range, 9.7-14.1 years) with the old records available in the archive. The control
sample had records presurgery (mean, 19.8 6 2.2 years; range, 16.6-21.6 years). Results: The sagittal
advancement of A-point, after the application of the technique, was 5.43 6 2.71 mm. Some mandibular dentoal-
veolar adaptation was noted. The position of the maxilla was stable in the long term. In contrast, the control group
showed limited growth at the maxillary level during the long-term follow-up period. Conclusions: Our results
showed that the Alt-RAMEC technique, performed at the correct time, with a double-hinged expander,
followed by Class III spring or elastic traction, 24 h/d, allows for satisfactory maxillary protraction, with stable
long-term results. The comparison with a sample of matched nontreated patients with Class III malocclusion
allowed to suggest the positive effect of the treatment on the maxillary position vs the natural evolution of the
Class III skeletal discrepancy. (Am J Orthod Dentofacial Orthop 2021;160:221-30)

T
he incidence of maxillary hypoplasia in the white
population is reported to be about 5%.1 Devel-
a
oping midfacial retrusion in children has been
University of Milan, Smile House, Regional Center for CLP, Department of
Maxillo-Facial Surgery, Santi Paolo e Carlo Hospital, Milan and Department of
conventionally treated with protraction facemasks at
Maxillo-Facial Surgery, University of Milano-Bicocca, San Gerardo Hospital, an early age.2 Rapid maxillary expansion (RME) loosens
Monza, Italy.
b
the articulations of the maxillary complex from the rest
Private practice, Milan, Italy.
c
Medical Physics Department, University of Milano-Bicocca, San Gerardo Hospi-
of the skull, whereby rendering more effective maxillary
tal, Monza, Italy. protraction.3 A metric average anterior movement of
d
Maxillo-Facial Surgery, University of Milano-Bicocca, Department of Maxillo-- more than 3 mm at point A was reported in a meta-
Facial Surgery, San Gerardo Hospital, Monza, Italy.
All authors have completed and submitted the ICMJE Form for Disclosure of Po-
analysis on maxillary skeletal base protraction with face-
tential Conflicts of Interest, and none were reported. mask and RME in patients with Class III malocclusion.4
Address correspondence to: Maria Costanza Meazzini, University of Milan, Smile Although Woon and Thiruvenkatachari,5 in a recent sys-
House, Regional Center for CLP, Department of Maxillo-Facial Surgery, Santi
Paolo e Carlo Hospital, Via di Rudin
ı 8, Milan 20121, Italy; e-mail,
tematic review, noted the lack of evidence of any lasting
cmeazzini@yahoo.it. long-term effect of early maxillary protraction. The
Submitted, November 2019; revised and accepted, April 2020. alternate RME and constriction (Alt-RAMEC) technique6
0889-5406/$36.00
Ó 2021 by the American Association of Orthodontists. All rights reserved.
is a protocol that allows disarticulating the circummax-
https://doi.org/10.1016/j.ajodo.2020.04.027 illary sutures in patients who are close to the end of
221
222
craniofacial growth; the technique uses a 2-hinged rapid
palatal expander, which is unique in its biomechanics,
expanding and rotating each half of the maxilla out-
ward. Timing of treatment seems fundamental for the
success of the technique in the long term. The treatment
is started when the vertebral stage of maturation is be-
tween V2 and V3 (second and third stage of vertebral
maturation).7 Liou6 has shown a significant advance-
ment of A-point in cleft patients (5.8 mm) in 6 months,
and those results remained stable without significant
maxillary relapse after 5 years. Similar results (5.7 mm)
were reported by Meazzini et al8 in patients with unilat-
eral cleft lip and palate.
The objective of this study was to assess the long-
term validity of this technique in patients with noncleft
Class III malocclusion. This study followed the principles
of the declaration of Helsinki.
MATERIAL AND METHODS
A modification of the Liou-Alt-RAMEC and maxillary
protraction technique has been applied by the authors in
117 patients: of these patients, 49 had noncleft, nonsyn-
dromic, Class III malocclusion.
Inclusion criteria for this study were (1) white pa-
tients; (2) patients with noncleft, nonsyndromic Class
III malocclusion consecutively treated with the Alt-
RAMEC and maxillary protraction technique; (3) verte-
bral stage of maturation V2 V3 at the beginning of
treatment (usually corresponding to late deciduous or
permanent dentition); (4) skeletal Class III malocclusion
with no functional shift; (5) no mandibular asymmetries
were included; and (6) full cooperation of the patients
during treatment.
Seven of the consecutively treated patients had to be
excluded because of a severe lack of cooperation. There-
fore, the actual total study sample was 42. The average
age of the patients in the sample was 12.7 6 1.6 (range,
9.4-15.9) years before (T0) and 14.0 6 1.1 (range, 10.4-
15.4) years after maxillary protraction (T1). Of these pa-
tients, 21 had follow-up records longer than 6 years. The
average age at long-term (Tlt) was 19.4 6 2.8 (range,
17.2-26.9) years.
A group of patients with Class III malocclusion,
extrapolated from our archives of orthognatic surgery,
were searched. Only 17 out of the 120 patients with Class
III malocclusion had early records, which could be
matched for sex, cephalometric severity of Class III
malocclusion, and average age (12.1 6 1.4 years; range,
9.7-14.1 years) at T0 and used as a control sample. The
sample had preorthognathic surgical records (Tlt) at an
average age of (19.8 6 2.2 years, range 16.6-21.6 years).
The double-hinged maxillary expander (DHME)
consisted of a jackscrew in the center and 2 hinges of
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Meazzini et al
rotation posteriorly6 (Fig 1, A). In the mandibular
arch, a double lingual arch with anterior hooks was sol-
dered on molar and premolar bands8 (Fig 1, B). The
treatment protocol, as suggested by Liou,6 consisted
in 7 cycles with 7 days of expansion and 7 days of
constriction, 1 mm per day, alternatively. After 7 weeks
of alternate expansion-constriction, mild mobility of
the whole maxilla was felt clinically, and a mild discom-
fort was reported by the patient, especially at the para-
nasal area. In 30% of the patients, there was a need to
go up to 9 or 11 cycles to achieve mildly perceivable
maxillary mobility.
We modified the original protocol by adding tempo-
rary skeletal anchorage devices (TADs) provided by 2
maxillary and 2 mandibular titanium miniscrews
(Cortical Anchorage Miniscrews; Forestadent, Pforz-
heim, Germany) (Fig 1, C). Two TADs were positioned
in the maxilla between the roots of the first molars
and the second premolars, in the mandible between
the roots of the canines and the lateral incisors. TADs
were used indirectly, with ligature wires to the dental ap-
pliances. In 10 of the patients who had significant lower
anterior crowding and missing maxillary teeth (incisors,
premolars, or in 1 patient's first molars), some dental
movement was desirable to close the space and relieve
mandibular crowding, and therefore no TADs were used.
After the completion of the expansion/constriction
cycles, the technique included 5-8 months of active
maxillary protraction.
The maxillary protraction was delivered by a pair of
noncompliant tooth-borne, intraoral maxillary protrac-
tion springs (Fig 1, D). The springs produced 300 g of
force per side. Given the relatively frequent breakage
of the b-titanium springs, all patients continued pro-
traction with intraoral elastics (300 g), to be used
24 hours a day, also during mealtimes (Fig 1, C).
For each patient, a lateral cephalometric radiograph
was obtained at T0 and T1 and the long term (Tlt).
Cone-beam computed tomography systems scans were
not offered by our national health system until recently
and were, therefore, not available for most patients.
Lateral cephalometric tracings were digitized with
software (version 1.5.1; Delta-Dent, Milan, Italy)
and superimposed on the anterior cranial base, ori-
enting on Sella-Nasion (SN) line. For linear measure-
ments, a constructed true horizontal line at 7 to SN
was used (Fig 2, A). The method was previously
described.9,10 Superimpositions of the tracings at T0
and Tlt in 1 treated and 1 nontreated patient are de-
picted in Figure 2, B and C. Lateral x-rays and clin-
ical photographs pretreatment, posttreatment, and at
long-term follow-up of 2 patients are shown in
Figures 3 and 4.
Meazzini et al 223

Fig 1. A, Double-hinged maxillary expander. B, Mandibular double lingual arch. C, TADs positioned in
the maxillary and mandibular arch for indirect anchorage. Note the heavy intraoral elastics, substituting
protraction springs. D, Protraction springs.

Fig 2. A, Cephalometric superimposition T0 T1. The dotted line shows the T0 tracing (pretreatment),
the continuous line the T1 tracing (at the completion of treatment). B, Cephalometric superimposition
from T0 to Tlt of a patient-matched at T0 for sex and severity of maxillary hypoplasia who has not been
treated with Alt-RAMEC. C, Cephalometric superimposition from T0 to Tlt of a patient who has been
treated with Alt-RAMEC. The dotted line shows the T0 tracing; the continuous line shows the cephalo-
metric tracing at Tlt. S, sellion 5 midpoint of the fossa hypophysealis; N, nasion 5 anterior point at the
frontonasal suture; A, subspinal 5 deepest anterior point in the maxilla; Oclp, posterior occlusal
point 5 distal cusp of the maxillary first molar; UI, maxillary incisor point 5 incisal edge of the maxillary
central incisor; LI, mandibular incisor point 5 incisal edge of the mandibular central incisor; Go,
gonion 5 midpoint of the angle of the mandible; B, supramental 5 deepest anterior point of the
mandible; Gn, gnathion 5 most anterior inferior point of the contour of the symphysis.

Statistical analysis if there were any skeletal differences or any age differ-
After the Shapiro-Wilk normality test, descriptive sta- ence between the treated sample and the nontreated
tistics of data at T0, T1, and Tlt were calculated for the control (for correct age and skeletal matching). A chi-
treated sample. A Student t test was carried out to check square test was used to assess differences in gender

American Journal of Orthodontics and Dentofacial Orthopedics August 2021  Vol 160  Issue 2
224 Meazzini et al

Fig 3. A, Profile of a female patient aged 12.5 years at pretreatment. B, Lateral x-ray pretreatment. C,
Lateral occlusal photograph pretreatment. D, Profile at the end of maxillary protraction. E, Lateral x-ray
postprotraction with occlusal bite blocks for vertical control, which were removed before orthodontic
alignment. F, Lateral occlusal photograph postprotraction with posterior occlusal bite blocks for vertical
control. G, Profile 6.7 years posttreatment. H, Lateral x-ray 6.7 years posttreatment. I, Lateral occlusal
photograph 6.7 years posttreatment when patient aged of 19.8 years.

August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Meazzini et al 225

Fig 4. A, Profile of a female patient aged 12.9 years at pretreatment. B, Lateral x-ray pretreatment. C,
Lateral occlusal photograph pretreatment. D, Profile of the patient at the end of maxillary protraction.
Note the support of the upper lip. E, Lateral x-ray posttreatment. In this patient, TADs were used in
the mandibular arch. F, Lateral occlusal photograph posttreatment. G, Profile at smile 7 years posttreat-
ment. H, Lateral x-ray 7 years posttreatment. I, Lateral occlusal photograph 7 years posttreatment
when patient was aged 20.4 years.

American Journal of Orthodontics and Dentofacial Orthopedics August 2021  Vol 160  Issue 2
226
distribution between the sample and control group (for
correct gender matching).
An analysis of variance (ANOVA) for repeated mea-
surements (within-subjects ANOVA for correlated sam-
ples) at 3-time points (T0, T1, and Tlt) was performed
for all the cephalometric measurements (as the same var-
iable was being analyzed at different time points) to
detect any overall changes in mean scores over 3-time
points in the treated group.
To compare the differences in the linear and angular
measurements at T0 and Tlt between treated and non-
treated patients, an independent samples t test was car-
ried out.
Given a large number of multiple comparisons, a
Benjamini-Hochberg correction procedure was applied.
The Raw P values, significant using the Benjamini-
Hochberg procedure with the false discovery rate, were
used.11
A power analysis was run for each test, given the
sample size reached in the collection of the data, setting
a (type I error) at 0.05 and a large population size effect.
The power (1 b) of each test was over 0.89 with G*Po-
wer (Heinrich Heine University, Duesseldorf, Germany.
A Cronbach a intraclass correlation coefficient was
used to assess cephalometric intraexaminer reliability.
Point detection and measurements were performed
twice by the same operator (C.T.) at 6-month intervals
on 10 randomly selected patients, which is suitable to
assess test-retest reliability. Statistical analysis was car-
ried out with Stata software (version 10; StataCorp, Col-
lege Station, Tex).
RESULTS
The intraclass correlation coefficient used to assess
the consistency of the single rater was 0.892, thus indi-
cating good intrarater reliability. After the Benjamini-
Hochberg correction, the P value was set at 0.029.
Matching of the treated and nontreated samples was
adequate (Table I).
At T1, the maxilla was advanced at A-point on
average 5.43 6 2.71 mm (P \ 0.000). SNA increased
by 4.6 6 2.8 (P \ 0.000). Wits occlusal indicator was
increased 7.5 mm (P \ 0.000). The mandibular posterior
rotation was not significant, but Pg and B-point had a
significant vertical lowering. Mandibular incisors were
retroclined at T1, though not significantly. The anterior
height of the mandible (symphyseal height) was signifi-
cantly increased after treatment. Nasion was advanced
1.7 6 1.5 mm (P \ 0.029). The ANOVA run showed
that the variance between samples was justified by factor
time (T0, T1, and Tlt). All angular and linear measure-
ments and the results of the ANOVA for repeated mea-
sures are listed in Table II. Three patients developed
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Meazzini et al
reciprocal temporomandibular joint clicks, which were
addressed with physical therapy.
The maxilla remained stable, without significant
relapse (point A horizontal from T1 to Tlt,
0.57 6 2.63 mm), in patients at long-term posttreat-
ment (average age, 19.4 6 2.8; range, 17.2-26.9 years).
The mandible grew, after the completion of treatment
(T1 Tlt), at B-point on average of 2.1 6 2.17 mm
and Pog of 3.1 6 2.19 mm. Statistical significance of
each angular and linear change between the end of
treatment and the long-term follow-up is depicted in
Table II.
None of the female patients in the long term needed
further treatment, whereas 3 male (14%) patients had an
edge-to-edge dental relationship, which could warrant
either surgery or dental compensation.
Clinically none of the patients presented peri-
odontal problems or loss of vitality of teeth. None of
the patients had a permanent temporomandibular joint
disfunction with pain in the long term, though 2 pa-
tients still presented a reciprocal click during mandib-
ular movement.
The control group was formed by patients with Class
III malocclusion matched for cephalometric severity of
Class III malocclusion and matched for sex and age,
not treated with any late orthopedics. There were no sig-
nificant differences in age and gender distribution at T0
and Tlt between the sample and control groups. At T0,
there were no significant cephalometric differences
than the treated group, confirming the adequacy of
the cephalometric matching. There was no significant
difference in terms of the postretention period between
the 2 samples (Table I). The comparison of the horizontal
and vertical changes from T0 to Tlt between the treated
and the control group (Table II) showed that, although in
the control group there was an average sagittal growth
of A-point of 1.9 6 3.75 mm after adolescence (from
12.7 to 19.4 years), the position of point A in the treated
patients was still close to the advancement measured at
T1 (T0 T1, 5.4 6 2.7 mm; T0 Tlt, 4.9 6 2.9 mm)
with a comparable long-term follow-up period (from
12.1 to 19.8 years). Mandibular horizontal growth was
also significantly different at the level of Pogonion. Hor-
izontal growth of Pg from T0 to Tlt was 6.4 6 3.9 mm in
nontreated and 4.1 6 4.3 mm in treated patients. Hor-
izontal growth of B-point from T0 to Tlt was
5.0 6 4.4 mm in nontreated, whereas it was
2.2 6 4.5 mm in treated patients. Vertical displacement
of Pg and Go from T0 to Tlt between nontreated and
treated patients was not significantly different. Compar-
ison of all other metric variables from T0 to Tlt between
treated and nontreated patients are listed in Table II and
depicted in Figure 2, B and C.
Meazzini et al 227

Table I. Comparisons between treated and control sample: Adequacy of age, sex, postretention period, and skeletal
matching. Descriptive statistics (average age, gender, postretention period, and angular measurements) and compar-
ison of dentoskeletal variables at T0 and Tlt of the treated sample and the nontreated control sample during the same
period of growth
Diff treated Diff treated
Treated sample Nontreated control vs control Treated sample Nontreated control vs control
Values (T0) (T0) (T0) (Tlt) (Tlt) (Tlt)
Age, y 12.7 6 1.6 12.1 6 1.4 0.6 19.4 6 2.8 19.8 6 2.2 0.4
(9.4-15.9) (9.7-14.1) (17.2-26.9) (16.6-21.6)
Gender 73% F; 27% M 69% F; 31% M 4 73% F; 27% M 69% F; 31% M 4
Postretention period, y 6.7 6 3.3 7.7 6 4.0 1.0
SNA,  78.01 6 4.76 79.06 6 4.12 1.05 82.26 6 4.15 79.41 6 4.43 2.80*
SNB,  80.51 6 5.03 80.99 6 3.82 0.48 80.88 6 4.82 83.86 6 5.18 3.02
ANB,  2.50 6 2.79 1.93 6 3.30 0.57 1.38 6 2.46 3.89 6 4.29 5.27**
PNS-ANS 3 GoGn,  26.64 6 5.59 27.63 6 4.94 0.99 27.10 6 5.49 26.97 6 4.34 0.02
Wits appraisal, mm 7.33 6 3.41 7.23 6 2.90 0.10 2.31 6 2.67 8.49 6 6.60 6.18**
UI 3 PNS-ANS,  115.09 6 8.69 118.07 6 9.02 2.98 113.30 6 6.63 118.60 6 9.37 5.23**
UI 3 Li.  130.47 6 10.81 119.88 6 25.17 10.59 132.55 6 9.58 128.04 6 9.37 4.80
SN 3 GoGn,  35.80 6 11.85 34.21 6 4.91 1.59 33.96 6 6.61 32.58 6 6.02 1.05
H Symph, mm 35.54 6 3.03 36.01 6 4.03 0.47 39.74 6 3.96 36.42 6 5.31 3.32
Note. Values are mean 6 standard deviation. P values of the Student t test for continuous variables and chi-square test for gender.
Treated sample (T0), tracings pretreatment; Treated sample (Tlt), long-term evaluation; Nontreated control (T0), tracings made at an average age
of 12.1 years (matched to the treated sample, 12.7 years); Nontreated control (Tlt), tracing at the end of growth of the control sample; F, female
patients; M, male patents.
*P \ 0.01; **P \ 0.029 (Benjamini-Hochberg correction).

Table II. Treatment results and long-term follow-up of Alt-RAMEC: angular and horizontal and vertical linear mea-
surements. Average angular and linear cephalometric measurements at different time points: changes between pre-
treatment and immediately postprotraction (T1 T0), changes between posttreatment and long-term evaluation
(Tlt T1)
Values T0 Difference (T1 T0) T1 Difference (Tlt T1) Tlt
SNA,  78.01 6 4.76a 4.63 6 2.82 82.64 6 3.48b 0.38 6 1.89 82.26 6 4.15b
SNB,  80.51 6 5.03a 0.32 6 2.49 80.19 6 4.22a 0.69 6 1.15 80.88 6 4.82b
ANB,  2.50 6 2.79a 5.05 6 2.19 2.55 6 2.24b 1.35 6 1.34 1.38 6 2.46c
PNS-ANS 3 GoGn,  26.64 6 5.59a 1.34 6 3.55 27.9 3 6 5.23a 0.90 6 2.61 27.10 6 5.49a
Wits appraisal, mm 7.33 6 3.41a 7.36 6 4.32 0.03 6 3.22b 2.60 6 3.15 2.31 6 2.67c
UI 3 PNS-ANS,  115.09 6 8.69a 0.50 6 5.30 115.08 6 6.32a 1.42 6 4.31 113.30 6 6.63a
LI 3 GoGn,  88.60 6 5.95a 3.02 6 7.36 85.58 6 8.8a 1.30 6 7.40 86.88 6 7.16a
UI 3 LI,  130.47 6 10.81a 2.37 6 8.79 131.41 6 10.10a 1.91 6 7.26 132.55 6 9.58a
SN 3 GoGn,  35.80 6 11.85a 0.30 6 2.90 36.10 6 5.76a 2.14 6 2.39 33.96 6 6.61a
H symph, mm 35.54 6 3.03a 3.67 6 2.56 39.16 6 3.79b 0.81 6 2.72 39.74 6 3.96b
A Vert, mm 44.83 6 4.41a 2.97 6 2.46 47.80 6 4.78b 1.58 6 3.00 49.38 6 4.06b
A Horiz, mm 59.64 6 6.13a 5.43 6 2.71 65.07 6 5.58b 0.57 6 2.63 64.50 6 6.09b
B Vert, mm 83.38 6 7.66a 3.33 6 3.41 86.72 6 7.77b 1.21 6 3.76 87.93 6 6.59b
B Horiz, mm 60.56 6 10.16a 0.48 6 3.48 61.14 6 8.79a 2.19 6 2.17 63.33 6 6.59c
N Vert, mm 7.85 6 0.46a 0.13 6 0.28 7.98 6 0.52b 0.04 6 0.21 7.94 6 0.50b
N Horiz, mm 63.95 6 3.84a 1.71 6 1.54 65.64 6 4.14b 0.45 6 1.75 66.09 6 4.18b
Pog Vert, mm 93.02 6 8.33a 4.78 6 4.41 97.80 6 8.94b 1.47 6 5.22 99.27 6 7.98b
Pog Horiz, mm 61.28 6 10.10a 0.97 6 2.61 62.25 6 9.75b 3.13 6 2.19 65.38 6 10.70c
ANS Vert, mm 39.73 6 3.81a 2.37 6 2.14 41.10 6 3.96b 2.18 6 2.29 43.38 6 3.41b
ANS Horiz, mm 64.83 6 5.52a 5.00 6 5.60 69.83 6 7.46b 0.98 6 5.39 68.85 6 5.90b
Note. Values are mean 6 standard deviation. Analysis of variance for repeated measures: different letters on the same line indicate a statistically
significant difference (P \ 0.029 [Benjamini-Hochberg correction]).
T0, Cephalometric x-rays before treatment; T1, after maxillary protraction; Tlt, long-term follow-up; SNA, SNB, ANB, SN 3 GoGn, SN 3 PNS-
ANS, angular changes relative to the cranial base (SN) of A-point, B-point, mandibular plane (GoGn) and nasal plane (PNS-ANS); PNS-ANS 3 GoGn,
angular changes relative to the nasal plane of the mandibular plane; UI 3 PNS-ANS, LI 3 GoGn, angular changes of maxillary incisors and mandib-
ular incisors position relative to mandibular and nasal plane; Wits appraisal, changes in Wits index of A B points positions relative to occlusal
plane; H Symph, changes in the height of the symphysis measured from mandibular incisal edge to Gn point (A, B, N, Pog, and ANS); Vert and Horiz,
vertical and horizontal movement of A-point, B-point, nasion, Pogonion, anterior nasal spine, relative to the reference line.

American Journal of Orthodontics and Dentofacial Orthopedics August 2021  Vol 160  Issue 2
228
DISCUSSION
The results of this retrospective clinical study support
the hypothesis that a repetitive weekly protocol of Alt-
RAMEC, using a 2-hinged expander with skeletal
anchorage and intraoral traction, allows obtaining sig-
nificant advancement of the maxilla in patients with
Class III malocclusion and that the results remain stable
in the long term.
To demonstrate the effectiveness of a technique, a
randomized trial is mandatory; nevertheless, we have
made an attempt to understand better the significance
of the results of this retrospective longitudinal study
by matching for sex and age the treated patients with
a sample of patients with nontreated Class III malocclu-
sion who initially had the same severity of skeletal
malocclusion. Furthermore, we are reporting the long-
term results of this technique in over half of the patients,
who have now reached the completion of growth
(19.4 6 2.8 years).
The total amount of maxillary advancement was over
5 mm in 6-12 months; the long-term evaluation re-
vealed that the maxillary advancement was stable.
Other authors have reported treatment with a modi-
fication of the Alt-RAMEC technique, but contrary to the
original protocol, using a traditional RME.12,13 Kaya
et al12 reported an average maxillary advancement of
2 mm, adding miniplates and a facemask, whereas Can-
turk and Celikoglu13 reported over 3 mm maxillary
advancement applying night time facemask and daytime
Class III elastics. These studies do not show any long-
term follow-up data. Al-Mozany,14 using a traditional
RPE, added palatal TADs and Class III elastics, obtaining
3.2 mm advancement.
Isci et al15 showed a larger advancement (4.1 mm),
quite close to the result of this study, using an RPE
and facemask 16 h/day, although, again, no long-term
data were reported. There are reports in the literature
on different applications of the Alt-RAMEC but applied
at a much younger age and with no long-term follow-
up, therefore, not comparable with this study. The rea-
sons that might have allowed a greater advancement
in the present study (point A, 5.4 mm in patients with
Class III malocclusion), in Liou6 (point A, 5.8 mm in cleft
patients), and Meazzini et al10 (point A 5.7 mm in cleft
patients)8 may be more than one.
First, the 2-hinged expander has a specific geometry
that rotates each half of the maxilla outward.6 According
to some authors, this allows a better loosening of all cir-
cummaxillary sutures.16 Therefore, studies that apply a
straight RME instead of a DHME might obtain less total
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Meazzini et al
advancement, as the lateral circummaxillary sutures (zy-
gomaticomaxillary, temporzygomatic, and pterigomax-
illary sutures) mainly bend but do not open sufficiently
to allow the same maxillary protraction.12-15 This
seems to be an important clinical difference, and
although it has already been shown in animals,16 it
certainly needs further investigation in patients.
In this study, we are unable to assess any effect of the
DHME itself, before protraction, in terms of the sagittal
position of the maxilla, given the fact that no radio-
graphic records were taken at the completion of the
expansion-constriction cycles, but only at the end of
the protraction phase.
In addition to the maxilla, the nasal bones in most
patients were also very mildly displaced and protracted
anteriorly (Table II). This explains the discomfort re-
ported by many patients over the nasal area during the
Alt-RAMEC, again suggesting that the nasomaxillary
complex is being disarticulated. Histologically, Wang
et al16 demonstrated that what occurs at the circummax-
illary sutures is not a simple deposition of osteoid, as
found in RME, but a process of sutural stretching and
protraction osteogenesis.
Second, springs or intraoral elastics are used 24 hours
per day, even during eating time. This is a fundamental
advantage over a facemask, which is usually only worn
during nighttime.
Third, the skeletal anchorage allows distributing the
forces of traction directly to the bones, with less dentoal-
veolar compensation. Mandibular incisors were only
mildly retroclined, and maxillary incisors were not pro-
clined, confirming the sagittal efficacy of TADs in con-
trolling dental compensations.
We believe that the most important aspect which jus-
tifies the success of this technique might be the timing.7
The expansion/constriction cycle is started when sutures
are still open, though the growth peak is almost reached.
The protraction mechanics are held till the process of
craniofacial growth is significantly decelerated. The later
timing of treatment seems to be the key to stability,
given the fact that in patients treated in the early mixed
dentition, there is no evidence of a long-term stability5
and that in a recent systematic review and meta-
analysis, Lin et al17 concluded that posttreatment
changes in patients treated early, reflect substantial
relapse.
The availability of a control group consisting of pa-
tients with nontreated Class III malocclusion, which pre-
sented a similar degree of Class III skeletal discrepancy at
the same age in which the study sample was treated, was
Meazzini et al
of considerable interest (Tables I and II). In nonpro-
tracted patients, the maxilla, as shown by other au-
thors,7,18 has a smaller horizontal growth in the
second decade of life, whereas the mandible grows and
autorotates19 (Fig 2, A). In the treated sample, mandib-
ular growth occurs but is mildly reduced compared with
the controls. In contrast, in the maxilla, the average for-
ward movement of 5.4 mm obtained with the Alt-
RAMEC does not grow further, but neither relapses
significantly after more than 6 years of follow-up. This
differential allows for the stability of the long-term
result (Fig 2, B and C).
As mentioned, the average maxillary advancement
obtained in our study was 5.4 6 2.7 mm. In a recent
study on LeFort I osteotomy, Almukhtar et al20 reported
an average advancement of 5.9 6 1.7 mm in patients
treated with orthognatic surgery at the completion of
growth. This seems to suggest that some of the patients
treated with Alt-RAMEC will not avoid surgery. The age
range at the long-term follow-up of this sample was
17.2-26.9 years, and the sample was relatively small.
Although none of the female patients showed any
relapse, 3 of the male patients presented with a mild
Class III occlusion at 20-21 years of age, which might
warrant either the surgical correction or dental compen-
sation. Therefore, although long-term results are com-
forting, particularly in male patients, we are still
unable to predict the proportion of patients who will
not require orthognatic surgery.
However, a significant advantage of this technique is
the psychological aspect: the appearance of the patient
improves during adolescence, which is a remarkably
difficult period in terms of awareness and psychosocial
adjustment.21
Clearly, this technique has several drawbacks: first, a
high collaboration is needed. Seven out of 49 patients
(14%) abandoned treatment very early, finding it impos-
sible to follow the opening/closing cycles. We have not
been able to retrospectively assess the actual number
of patients who refused to undergo the treatment once
properly informed, neither the number of patients who
decided they would rather wait for surgery at the
completion of growth, even for financial reasons (sur-
gery is free in our country, whereas orthodontics is
only partially paid by the national health care system).
Second, even when TADs were used, some dentoal-
veolar modifications were observed in the mandibular
arch, with a mild retroclination and extrusion of the
mandibular incisors, although this value normalizes in
the long term. Third, no significant vertical lowering of
ANS and point A was obtained even when the maxilla
was not only sagittally, but also vertically deficient
(Table II).
American Journal of Orthodontics and Dentofacial Orthopedics
229
CONCLUSIONS
Our results seem to point out that the Liou-Alt-
RAMEC technique, performed:
1. With the right timing, close to growth peak.
2. With the correct double-hinged expander, which al-
lows true mobilization of the maxilla and not a
traditional RPE.
3. Followed by Class III springs or elastic traction,
applied 24 h/d, and not only nightly traction allows
for a satisfactory maxillary protraction, with a stable
long-term result. Although the study is retrospective
and may not give the evidence of a prospective ran-
domized trial, the comparison with a sample of pa-
tients with nontreated Class III malocclusion allows
suggesting the positive effect of the treatment on
the maxillary position vs the natural evolution of
adolescents in Class III malocclusion of comparable
age who were not treated.
Finally, as the long-term age of the patients in this
study ranged from 17.2 to 26.9 years; furthermore,
follow-up is needed to assess the proportion of patients
who will avoid final orthognathic surgery.
AUTHOR CREDIT STATEMENT
Maria Costanza Meazzini contributed to conceptual-
ization, methodology, supervision, original draft prepa-
ration, and draft reviewing and editing; Camilla Torre,
Roberto Tintinelli, Alessandro Capello, and Fabio Maz-
zoleni contributed to study investigation; Elena De Ponti
contributed to data curation and validation.
ACKNOWLEDGMENTS
The authors thank Dr Andrea Montanari and Chico
Onlus for the generous donations to Smile House
through Operation Smile Italy.
REFERENCES
1. Perillo L, Masucci C, Ferro F, Apicella D, Baccetti T. Prevalence of
orthodontic treatment need in southern Italian schoolchildren. Eur
J Orthod 2010;32:49-53.
2. Kim JH, Viana MA, Graber TM, Omerza FF, BeGole EA. The effec-
tiveness of protraction face mask therapy: a meta-analysis. Am J
Orthod Dentofacial Orthop 1999;115:675-85.
3. Ngan P, H€agg U, Yiu C, Merwin D, Wei SH. Soft tissue and dentos-
keletal profile changes associated with maxillary expansion and
protraction headgear treatment. Am J Orthod Dentofacial Orthop
1996;109:38-49.
4. Foersch M, Jacobs C, Wriedt S, Hechtner M, Wehrbein H. Effective-
ness of maxillary protraction using facemask with or without
maxillary expansion: a systematic review and meta-analysis. Clin
Oral Investig 2015;19:1181-92.
5. Woon SC, Thiruvenkatachari B. Early orthodontic treatment
for Class III malocclusion: a systematic review and meta-
August 2021  Vol 160  Issue 2
230
analysis. Am J Orthod Dentofacial Orthop 2017;151:
28-52.
6. Liou EJ, Tsai WC. A new protocol for maxillary protraction in cleft
patients: repetitive weekly protocol of alternate rapid maxillary ex-
pansions and constrictions. Cleft Palate Craniofac J 2005;42:
121-7.
7. Baccetti T, Franchi L, McNamara JA Jr. An improved version of the
cervical vertebral maturation (CVM) method for the assessment of
mandibular growth. Angle Orthod 2002;72:316-23.
8. Meazzini MC, Zappia LB, Tortora C, Autelitano L, Tintinelli R.
Short- and long-term effects of late maxillary advancement with
the Liou-Alt-RAMEC protocol in unilateral cleft lip and palate.
Cleft Palate Craniofac J 2019;56:159-67.
9. Meazzini MC, Allevia F, Mazzoleni F, Ferrari L, Pagnoni M,
Iannetti G, et al. Long-term follow-up of syndromic craniosynos-
tosis after Le Fort III halo distraction: a cephalometric and CT eval-
uation. J Plast Reconstr Aesthet Surg 2012;65:464-72.
10. Meazzini MC, Basile V, Mazzoleni F, Bozzetti A, Brusati R. Long-
term follow-up of large maxillary advancements with distraction
osteogenesis in growing and non-growing cleft lip and palate pa-
tients. J Plast Reconstr Aesthet Surg 2015;68:79-86.
11. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a
practical and powerful approach to multiple testing. J R Stat Soc
B Methodol 1995;57:289-300.
12. Kaya D, Kocadereli I, Kan B, Tasar F. Effects of facemask treatment
anchored with miniplates after alternate rapid maxillary expan-
sions and constrictions; a pilot study. Angle Orthod 2011;81:
639-46.
13. Canturk BH, Celikoglu M. Comparison of the effects of face mask
treatment started simultaneously and after the completion of the
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Meazzini et al
alternate rapid maxillary expansion and constriction procedure.
Angle Orthod 2015;85:284-91.
14. Al-Mozany S. Treatment of Class III Malocclusions Using Tempo-
rary Anchorage Devices (TADs), the Alt-RAMEC Protocol and In-
termaxillary Class III Elastics in the Growing Patient. A
Prospective Clinical Study. Sydney, Australia: University of Syd-
ney; 2011.
15. Isci D, Turk T, Elekdag-Turk S. Activation-deactivation rapid
palatal expansion and reverse headgear in Class III cases. Eur J Or-
thod 2010;32:706-15.
16. Wang YC, Chang PM, Liou EJ. Opening of circumaxillary sutures by
alternate rapid maxillary expansions and constrictions. Angle Or-
thod 2009;79:230-4.
17. Lin Y, Guo R, Hou L, Fu Z, Li W. Stability of maxillary protrac-
tion therapy in children with Class III malocclusion: a system-
atic review and meta-analysis. Clin Oral Investig 2018;22:
2639-52.
18. Cortella S, Shofer FS, Ghafari J. Transverse development of the
jaws: norms for the posteroanterior cephalometric analysis. Am J
Orthod Dentofacial Orthop 1997;112:519-22.
19. Bj€ork A. Facial growth rotation–reflections on definition and
cause. Proc Finn Dent Soc 1991;87:51-8.
20. Almukhtar A, Ayoub A, Khambay B, McDonald J, Ju X. State-of-
the-art three-dimensional analysis of soft tissue changes following
Le Fort I maxillary advancement. Br J Oral Maxillofac Surg 2016;
54:812-7.
21. Takatsuji H, Kobayashi T, Kojima T, Hasebe D, Izumi N, Saito I,
et al. Effects of orthognathic surgery on psychological status of
patients with jaw deformities. Int J Oral Maxillofac Surg 2015;
44:1125-30.
ORIGINAL ARTICLE

Diagnostic accuracy of lateral

cephalograms and cone-beam computed
tomography for the assessment of sella
turcica bridging
Ashley Marie Acevedo,a Manuel Lagravere-Vich,b and Thikriat Al-Jewairc
Bridgeport, CT, Edmonton, Alberta, Canada, and Buffalo, NY

Introduction: The purpose of this research was to assess the diagnostic accuracy of sella turcica bridging on
lateral cephalograms when compared with true sella turcica bridging determined via cone-beam computed
tomography (CBCT). Methods: A cross-sectional study was conducted using CBCT images from which
lateral cephalograms were generated. The study included 185 subjects (118 females and 67 males; age
range, 10-30 years; mean age, 16.63 6 4.20 years). Sella turcica landmarks and related measurements were
calculated for both diagnostic modalities and analyzed by 1 examiner. Subjects were classified into 1 of 3
outcome groups: no bridging, partial bridging, and complete bridging. Diagnostic accuracy was evaluated
using sensitivity, specificity, positive and negative predictive values, and receiver operator characteristic
curves. Results: Ten patients were diagnosed as complete bridging on CBCT, whereas 31 patients were diag-
nosed as complete bridging on lateral cephalogram. Although the lateral cephalogram detected all subjects with
complete bridging, it incorrectly classified 12% of subjects. The percent agreement between both diagnostic
methods was 55.68%, with a kappa statistic of 0.22 on the right sella turcica and 0.20 on the left sella turcica,
indicating fair but statistically significant agreement. The overall accuracy of lateral cephalograms as a diag-
nostic modality in discriminating between no bridging and partial bridging was good as determined with the
area under the curve values of 0.86 and 0.85 for right and left sides, respectively. Conclusions: Although lateral
cephalograms overestimate patients with complete bridging compared to CBCTs, they are a suitable screening
modality for accurately suggesting complete sella turcica bridging and differentiating between patients with no
bridging and partial bridging. (Am J Orthod Dentofacial Orthop 2021;160:231-9)

T
he development of sella turcica and its morpho-
logic variants has been extensively studied in the
literature.1,2 The sella turcica is a saddle-shaped
depression on the intracranial surface of the sphenoid
bone. There are 5 different morphologic variations of
sella turcica as reported by Axelsson et al1: oblique ante-
rior wall, sella turcica bridging, double contour of the
a bridging,3 ossification of the dura mater between the
Private practice, Bridgeport, CT.
b
Division of Orthodontics, University of Alberta, Edmonton, Alberta, Canada.
c
Department of Orthodontics, School of Dental Medicine, State University of New
York at Buffalo, Buffalo, NY.
All authors have completed and submitted the ICMJE Form for Disclosure of Po-
tential Conflicts of Interest, and none were reported.
The study was approved by the University at Buffalo Health Sciences Institutional
Review Board (STUDY00001251).
Address correspondence to: Thikriat Al-Jewair, Department of Orthodontics,
School of Dental Medicine, State University of New York at Buffalo, 140 Squire
Hall, 3435 Main St, Buffalo, NY 14214; e-mail, thikriat@buffalo.edu.
Submitted, November 2019; revised, March 2020; accepted, April 2020.
0889-5406/$36.00
Ó 2021 by the American Association of Orthodontists. All rights reserved.
https://doi.org/10.1016/j.ajodo.2020.04.025
floor, irregularity in the posterior dorsum sella, and py-
ramidal shape of dorsum sella.
The sella turcica bridge is a complete ossification of
the interclinoid ligaments, which may occur either
unilaterally or bilaterally.2 Although the exact cause of
bridging is not known, multiple theories have been pro-
posed, including an abnormality in the embryologic
development of the sphenoid bone resulting in
anterior and posterior clinoid processes,4 or because of
focal infections of the pituitary gland.5
This normal variant of sella turcica occurs with a re-
ported incidence of 1.1%-13% of the general popula-
tion, with an increased prevalence in those with severe
craniofacial disproportions.1,6-11 In addition, several
studies have reported an increased frequency of
bridging in adolescents with dental anomalies, such as
canine impaction (odds ratio, 3.93; 95% confidence
interval [CI], 1.43-10.7)3 and tooth transposition (odds

231
232
ratio, 2.9; 95% CI, 1.1-7.9).12 As calcification in the sella
turcica region can appear in early childhood, recognizing
sella turcica bridging on diagnostic imaging may allow
for early diagnosis and timely intervention of palatal
canine impactions, potentially reducing the complexity
of treatment.3 As well, it can serve as a screening method
for rare genetic diseases such as Axenfeld-Rieger syn-
drome with PITX2 mutation.13 This is an autosomal
dominant disorder that is caused by the abnormal migra-
tion of neural crest cells.14 Although the phenotype in
this syndrome is heterogenous, midface hypoplasia,
dental hypodontia, and sella turcica anomalies are
commonly present.13,15
Because of the 2-dimensional (2D) view provided in
conventional lateral cephalograms, it is difficult to
ascertain a clear distinction between a sella turcica
bridge when there is complete bony fusion and the
appearance of fusion between the anterior and posterior
clinoid processes because of radiographic superimposi-
tion. In previous autopsy studies, the occurrence of a
true sella turcica bridge varied between 2% and 6%.9
However, sella turcica bridging diagnosed on lateral
cephalograms is reported at a higher frequency because
of the difficulty distinguishing between a true sella tur-
cica bridge and a pseudobridge. As a result, previous
studies have questioned the reliability of using 2D
radiographs for accurate diagnosis of sella turcica
bridging.9,16
Technological advances in 3-dimensional (3D) imag-
ing using cone-beam computed tomography (CBCT)
offer noteworthy advantages in the quality and quantity
of anatomic data. 3D CBCT has the advantage of over-
coming challenges of superimposition and differential
magnification of bilateral structures.17 Furthermore,
3D CBCT scans acquired from a patient can be used to
generate 2D lateral cephalograms, thus minimizing
further cost and radiation exposure to the patient.18,19
The objective of this cross-sectional study was to assess
the diagnostic accuracy of sella turcica bridging on con-
ventional lateral cephalograms when compared with
true sella turcica bridging determined via CBCT (consid-
ered the reference standard).
MATERIAL AND METHODS
This cross-sectional study was approved by the State
University of New York at Buffalo Health Sciences Insti-
tutional Review Board (no. STUDY00001251). The sam-
ple included CBCT images of patients ranging from 10 to
30 years of age assessed between January 2005 and
August 2015 at 1 oral and maxillofacial surgery practice.
The inclusion criteria were full volume CBCT images,
clear representation of sella turcica, and any type of
malocclusion. Records of subjects with a significant
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Acevedo, Lagravere-Vich, and Al-Jewair
pathology in the maxillofacial region, clear craniofacial
syndromic anomalies, and CBCT images with motion ar-
tifacts were all excluded.
CBCT images were acquired using an i-CAT Cone
Beam 3D dental imaging system (version 3.1.62; Imag-
ing Sciences International, Hatfield, Pa) at these set-
tings: 3-7 mA, 120 kV, exposure time of 40 seconds,
voxel size of 0.4 mm, and a focal spot of 0.5 mm and
a scanning area of 16 3 13 cm. The majority of scans
were taken at 0.4 mm voxel size for 40 seconds with a
greyscale range of 14 bits, except for 3 scans taken at
0.3 mm voxel size for 40 seconds.
The CBCT images were obtained by an experienced
technician. All patients were seated in an upright posi-
tion with their heads oriented with the occlusal plane
parallel to the floor. Head and chin support were used
to stabilize the head position, and all teeth were out of
occlusion by having the patient bite on a cotton roll.
The images were exported in digital imaging and com-
munications in medicine format then imported into Dol-
phin 3D Imaging System (version 11.7.05.66 Premium;
Dolphin Imaging and Management Solutions, Chats-
worth, Calif) for analysis.
The original CBCT volumes were standardized by
setting the orientation of the axial (x), midsagittal (y),
and coronal (z) planes. The axial plane (x) was set to
the Frankfort Horizontal plane. The midsagittal plane
(y) was defined by nasion (N), anterior nasal spine
(ANS), and basion (Ba) landmarks.20 Then, the multipla-
nar views were configured as 0.4 mm slices to match the
thickness of the original scan.
Synthesized lateral cephalograms were generated
from each standardized CBCT volume in the perspective
projection, using mechanical porion as the projection
center, to simulate the geometry of the conventional
lateral cephalograms. Because the original CBCT vol-
umes were oriented to Frankfort Horizontal, synthesized
lateral cephalograms were also oriented to Frankfort
Horizontal. The Dolphin 1 filter was then used to
improve image sharpness.
An onscreen 0.5 mm marker was used to identify each
landmark in the multiplanar views and 3D rendered im-
age (Table I and Fig 1). A total of 6 landmarks were iden-
tified, and 4 measurements were calculated. Coordinate
data for each landmark was used to calculate the
Euclidean distance between the 2 paired points. These
landmarks were not measured directly on the CBCT
reconstruction.
To ensure the investigator was blinded, random case
identification numbers were generated and reordered for
landmark identification on lateral cephalograms, which
occurred 2 weeks after landmarks were identified on
CBCT images (Table I). Distances between the landmarks
Acevedo, Lagravere-Vich, and Al-Jewair 233

Table I. Sella turcica landmarks and measurements on CBCT and lateral cephalograms
Landmark Definition Measurement Definition
CBCT
TS Tuberculum sellae; midpoint on the anterior TS to DS Length of sella turcica
boundary of sella identified on the
midsagittal plane
DS Dorsum sellae; midpoint on the posterior Right ACP-PCP Interclinoid distance on right side
boundary of the sella turcica on the
midsagittal plane
R ACP Apex of the anterior clinoid process on the Left ACP-PCP Interclinoid distance on left side
right side
L ACP Apex of the anterior clinoid process on the left Complete sella turcica bridging Distance between ACP-PCP equals zero
side
R PCP Apex of the posterior clinoid process on the
right side
L PCP Apex of the posterior clinoid process on the
right side
ACP Apex of the anterior clinoid process
PCP Apex of the posterior clinoid process
Lateral cephalograms
TS Anterior boundary of sella turcica TS to DS Length of sella turcica
DS Posterior boundary of sella turcica ACP-PCP Interclinoid distance
ACP Apex of the anterior clinoid process
PCP Apex of the posterior clinoid process
R, right; L, left.

were calculated using the 2 Pt Line measurement tool in
the Dolphin 3D Imaging software. (Fig 2).
Quantification of sella turcica bridging was
performed using the method of Sundareswaran and
Nipun,21 which uses the ratio of interclinoid
distance (ACP-PCP) to length (TS-DS) to classify the
extent of interclinoid calcification into 3 categories:
no bridging (ratio, $33%), partial bridging
(ratio, .0%-33%), and complete bridging (ratio,
0%). For the CBCT images, sella turcica bridging was
quantified for the right and left sides of the sella tur-
cica separately (Fig 3). For the lateral cephalograms,
the measurements were conducted without differenti-
ation between sella turcica sides. The data were
collected, analyzed, and measured by 1 investigator
(A.M.A) who was trained in sella turcica landmark
identification on CBCT images by an experienced
investigator.

Fig 1. Identification of landmarks on CBCT: Multiplanar slices.

American Journal of Orthodontics and Dentofacial Orthopedics August 2021  Vol 160  Issue 2
234
Fig 2. Identification of landmarks on lateral cephalogram:
A, distance calculated between ACP and PCP; B, dis-
tance calculated between TS and DS.
The interclass correlation coefficient (ICC) was calcu-
lated to assess the intraexaminer reliability. Two weeks
after from initial examination, 10 randomly selected ra-
diographs were remeasured on both diagnostic modal-
ities. The same 10 radiographs were also compared
against an experienced orthodontist to assess interexa-
miner reliability. The method error was determined using
Dahlberg's formula.
Statistical analysis
Data were analyzed using SPSS Software for Win-
dows (version 23.0; IBM, Armonk, NY). Subjects were
classified into 1 of the 3 outcome groups by calculating
the ratio between ACP-PCP/TS-DS and using a cutoff of
33%. Diagnostic accuracy of lateral cephalograms was
evaluated using sensitivity, specificity, positive predic-
tive value (PPV), negative predictive value (NPV), and
receiver operator characteristic curves for outcome
groups. Agreement was assessed using the kappa static.
RESULTS
A total of 218 records were assessed, and 32 were
excluded for lack of visualization of sella turcica and
one because of the possibility of pathology. The study
included 185 subjects (118 female and 67 male) with a
mean age of 16.63 6 4.20 years. There was an approx-
imate difference of 9 months between mean ages for
Fig 3. Depiction of different sella turcica bridging types on CBCT: A, complete bridging; B, partial
bridging; and C, no bridging.
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Acevedo, Lagravere-Vich, and Al-Jewair
males and females (females, 16.36 6 4.10 years; male,
17.10 6 4.59 years).
There was a high agreement for the repeated mea-
surements between the 2-time points on CBCT (ICC:
length, 0.972; interclinoid distance R, 0.990; interclinoid
distance L, 0.989) and lateral cephalogram (ICC: length,
0.979; interclinoid distance, 0.992). In addition, there
was a high agreement between both examiners with all
coefficients near the maximum of one (ICC: length,
0.961; interclinoid distance R, 0.996; interclinoid dis-
tance L, 0.998). The average error is depicted in Table II.
The average interclinoid distance on lateral cephalo-
gram was 4.01 mm and the sella turcica length (TS-DS)
was 10.58 mm (Table III). With respect to the measure-
ments taken on CBCT, the average interclinoid distances
were 6.14 mm and 6.10 mm for the right and left sides,
respectively. The average sella turcica length (TS-DS)
was 10.13 mm.
Using lateral cephalograms, the prevalence of sella
turcica bridging was 54.6% (n 5 101), of which 16.8%
(n 5 31) was complete and 37.8% (n 5 70) was partial
bridging (Table IV). While using CBCT, the prevalence of
sella turcica bridging on the right was 13.0% (n 5 24),
with 5.4% (n 5 10) complete and 7.6% (n 5 14) partial
bridging. Similar prevalence was observed on the left
side for complete (5.4%, n 5 10) and partial bridging
(n 5 6.5%, n 5 12).
As seen in Table V, 10 patients were diagnosed as
complete bridging on CBCT, whereas 31 patients were
diagnosed as complete bridging on lateral cephalogram.
In addition, 14 patients were classified as partial
bridging on CBCT, whereas 70 patients were classified
as partial bridging on lateral cephalogram, with only 9
patients in agreement. In measuring the diagnostic ac-
curacy of the lateral cephalogram, sensitivities of
52.2%, 64.3%, and 100.0% for the no bridging, partial
bridging, and complete bridging groups were calculated,
respectively.
Although the lateral cephalogram detected all sub-
jects with complete bridging, it incorrectly classified
12.0% of subjects, demonstrated by the specificity of
Acevedo, Lagravere-Vich, and Al-Jewair 235

Although the percent agreement between both

Table II. Mean error of measurements according to
methods for the right was 55.68%, the kappa statistic
Dahlberg's formula
was 0.22 on the right and 0.20 on the left sides, indi-
Variable Method error cating fair but significant agreement (P \ 0.001). A
Distance TS-DS 0.26 truth table was generated listing the 9 different possibil-
Interclinoid distance R 0.24 ities of sella turcica bridging on CBCT to better assess the
Interclinoid distance L 0.20
agreement among outcome groups (Table VI).
Interclinoid distance (ACP-PCP) 0.20
Sella turcica length (TS-DS) 0.23
ACP-PCP/TS-DS 0.02 DISCUSSION
R, right; L, left. Sella turcica, teeth, and parts of the face share a com-
mon origin, that is, neural crest cells. Therefore, abnor-
malities in the anterior wall of the sella turcica have been
Table III. Mean sella turcica measurements on lateral linked to abnormalities in other structures, particularly
cephalograms and CBCT* in the frontonasal field.11 One important link is the rela-
tion between sella turcica bridging and dental anoma-
Mean, Standard
Measurement N mm deviation Minimum Maximum lies, including canine impaction. Because sella turcica
Lateral cephalograms can be routinely visualized on lateral cephalograms,
TS-DS 185 10.58 1.69 6.90 15.00 this study aimed to establish the relative diagnostic
ACP-PCP 154 4.01 2.05 0.40 11.30 capability of using lateral cephalograms to predict dental
ACP-PCP/TS-DS 154 0.37 0.16 0.04 0.83 anomalies in patients with sella turcica bridging.
CBCT
This cross-sectional study assessed the diagnostic ac-
TS-DS 185 10.13 1.57 5.77 16.28
R ACP-PCP 175 6.14 2.13 1.75 14.23 curacy of sella turcica bridging using conventional
L ACP-PCP 175 6.10 1.87 1.30 11.78 lateral cephalograms compared with true sella turcica
R ACP-PCP/TS-DS 175 0.61 0.19 0.19 1.20 bridging on CBCT imaging. The frequency of complete
L ACP-PCP/TS-DS 175 0.60 0.18 0.14 1.11 sella turcica bridging, when diagnosed with CBCT, was
R, right; L, left. 5.4% (10 complete bridging out of 185 subjects) and
*Complete bridging omitted to display minimum and maximum 16.8% (31 complete bridging out of 185 subjects)
measurements not including 0.
when diagnosed with synthesized lateral cephalogram.
The results from the CBCT compare favorably to the
wide range of reported values in the literature; however,
Table IV. Frequencies of sella turcica bridging on
the results from the lateral cephalograms were greater
lateral cephalograms and CBCT
than previous reports.7,22,23
CBCT Several 2D radiographic studies have investigated the
incidence of complete sella turcica bridging. Jewett24 re-
Lateral
Sella turcica bridging cephalograms Right Left ported an incidence of 13% in a study of 100 normal
No bridging 84 (45.4) 161 (87.0) 163 (88.1) subjects, M€ uller25 reported an incidence of 3.85%, and
Partial bridging 70 (37.8) 14 (7.6) 12 (6.5) Cederberg et al7 reported an incidence of 8% in a study
Complete bridging 31 (16.8) 10 (5.4) 10 (5.4) of 255 subjects. Several autopsy studies have also re-
Total 185 (100.0) 185 (100.0) 185 (100.0)
ported similar results. In an autopsy study of 250 skulls
Values are n (%). by Archana et al,22 incidences of sella turcica bridging
were classified on the basis of the right or left intercli-
88.0% for this category. Furthermore, out of those sub- noid ligaments. The total reported incidence of sella tur-
jects who were found to have complete bridging, only cica bridging was 4%, with 0.8% occurring on the right,
32.3% actually had complete bridging as suggested by 2.4% on the left, and 0.8% bilaterally. Busch26 reported
the PPV. When considering the PPV of the no bridging an incidence of 1.54% for complete bridging, Bergland
category, of those subjects found to have no bridging, et al6 reported an incidence of 6%, and Platzer reported
100.0% actually had no bridging. The lateral cephalo- an incidence of 5.9%. In a recent study evaluating 3D
gram demonstrated good accuracy in differentiating be- imaging using CBCT, Ortiz reported an incidence of
tween no bridging and partial bridging on both sides 9.3% for complete bridging.23 The increased frequency
with an area under the curve of 0.86 (95% CI, 0.78- of complete bridging on lateral cephalogram can be
0.93; P \ 0.001) and 0.85 (95% CI, 0.77-0.93; partly explained by the 2D view provided, making it diffi-
P \ 0.001), respectively (Figs 4 and 5). cult to discriminate between a true sella turcica bridge

American Journal of Orthodontics and Dentofacial Orthopedics August 2021  Vol 160  Issue 2
236 Acevedo, Lagravere-Vich, and Al-Jewair

Table V. Comparison of sella turcica bridging counts for right and left sides using CBCT as gold standard
Lateral cephalograms

Sella turcica No bridging (n) Partial bridging (n) Complete bridging (n) Total Sensitivity (%) Specificity (%) PPV (%) NPV (%)
CBCT
Right No bridging 84 61 16 161 52.2 100.0 100.0 23.8
Partial bridging 0 9 5 14 64.3 64.3 12.9 95.7
Complete bridging 0 0 10 10 100.0 88.0 32.3 100.0
Total 84 70 31 185
Left No bridging 84 63 16 163 51.5 100.0 100.0 21.8
Partial bridging 0 7 5 12 58.3 63.6 10.0 95.7
Complete bridging 0 0 10 10 100.0 88.0 32.3 100.0
Total 84 70 31 185

and the appearance of fusion between the clinoid pro-
cesses because of radiographic superimposition.
The overall incidence of partial bridging in this study
was 37.8% and 7.05% when diagnosed through lateral
cephalograms and CBCT, respectively. M€ uller25 and Je-
24
wett found an incidence of 3.2% and 13% for partial
bridging, respectively. Cederberg et al7 reported a higher
incidence of 68.8% partial bridging in a radiographic
study of 255 subjects. In his study, partial bridging
was classified into 2 categories on the basis of whether
the interclinoid ligament was more or less than half
calcified. No explanations were given as to how the no
bridging subjects were classified. In the recent 3D study
by Ortiz,23 an incidence of 9.3% for complete bridging
and 40.6% for partial bridging were reported. In this
study, sella turcica bridging was classified on the basis
of an objective classification system, whereas autopsy
studies by Lang,27 Archana et al,22 Kolagi et al,28 and
Skrzat et al2 used a visual classification system, which
can introduce some subjectivity and bias. In this study,
the interclinoid distance was 2 mm smaller when
measured using lateral cephalograms than on CBCT.
This difference was not surprising and could be due to
the superimposition of structures commonly seen on
2D lateral cephalometric images, which may have
impacted the linear measurements.

Fig 4. Receiver operator characteristic (ROC) curve of Fig 5. Receiver operator characteristic (ROC) curve of
no bridging and partial bridging classifications for right no bridging and partial bridging classifications for left sella
sella turcica bridging. turcica bridging.

August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Acevedo, Lagravere-Vich, and Al-Jewair
Table VI. Comparison of sella turcica bridging on lateral cephalograms and CBCT using multiple combinations of
sella bridging
CBCT No bridging
No bridge R–no bridge L 84
No bridge R–partial bridge L
No bridge R–complete bridge L
Partial bridge R–no bridge L
Partial bridge R–partial bridge L
Partial bridge R–complete bridge L
Complete bridge R–no bridge L
Complete bridge R–partial bridge L
Complete bridge R–complete bridge L
Total 84
R, right; L, left.
One of the most useful tools of currently available
CBCT software for orthodontic purposes is the ability
to reconstruct 2D conventional images, including pano-
ramic and lateral cephalometric images. In the process of
creating a lateral cephalogram, the patient's head is ori-
ented in a virtual space similar to the orientation used in
a cephalostat, thus eliminating the head orientation lim-
itation of conventional lateral cephalograms.29 In addi-
tion, many types of software allow the user to choose an
orthogonal or perspective projection-type when creating
CBCT synthetic cephalograms, the latter matching the
magnification and distortion of conventional cephalo-
grams with high accuracy.30,31
Several studies have reported that craniofacial mea-
surements from CBCT reconstructed lateral cephalo-
grams are similar to those obtained from conventional
lateral cephalograms. Although some significant differ-
ences were found between measurements, mainly linear
mandibular and midsagittal measurements, the differ-
ences were not clinically significant.30,32-34 In a study
by Moshiri et al,33 in which distances between anatomic
landmarks on skulls were compared with measurements
on CBCT synthesized lateral cephalograms, synthetic
cephalograms proved to be more accurate for linear
measurements in the sagittal plane. In a similar study,
van Vlijmen et al34 reported good intraobserver reli-
ability for all measurements and concluded that mea-
surements on CBCT synthesized lateral cephalograms
are comparable to conventional cephalograms. Because
of these conclusions, we found it suitable to use CBCT
synthesized lateral cephalograms to assess the diag-
nostic accuracy of sella turcica bridging when compared
with true sella turcica bridging determined via CBCT.
The classification of sella turcica bridging was not al-
ways in agreement between the 2 diagnostic methods.
Lateral cephalogram overestimated patients with partial
American Journal of Orthodontics and Dentofacial Orthopedics
237
Lateral cephalograms
Partial bridging Complete bridging Total
58 10 152
3 2 5
4 4
5 2 7
4 2 6
1 1
4 4
1 1
5 5
70 31 185
and complete bridging and underestimated patients
with no bridging for both sides. However, it correctly
identified all 10 patients with complete bridging on
the right but incorrectly classified 21 patients as com-
plete bridging because of superimposition of the bilat-
eral interclinoid calcifications when the following
scenarios presented: complete bridging on the left and
partial bridging on the right, complete bridging on the
left and no bridging on the right, partial bridging on
both sides, partial bridging on the right and no bridging
on the left, and no bridging on the right and partial
bridging on the left. The superimposition of bilateral
structures leading to the identification of a pseudo-
bridge is a result of the inherent systematic error present
in lateral cephalograms because of the 2D representation
of a 3D object, resulting in errors of projection and
magnification.35,36 These errors must be taken into ac-
count in the diagnosis of sella turcica bridging when
the radiographic appearance of fusion may be due to su-
perimposition of the interclinoid ligaments and not
necessarily because of real bony fusion.9
Overall, the synthesized lateral cephalogram gener-
ated good accuracy in differentiating between no
bridging and partial bridging and excellent accuracy in
identifying complete bridging. Although the patients
with partial and complete bridging were overestimated
and patients with no bridging were underestimated,
there was fair and statistically significant agreement be-
tween the CBCT and synthesized lateral cephalogram.
As a result, this study concluded that lateral cephalo-
grams are an acceptable diagnostic modality in accu-
rately detecting complete sella turcica bridging and
differentiating between no bridging and partial
bridging. Although patients with complete bridging
were overestimated with more false positives, using
this modality, paying closer attention to patients when
August 2021  Vol 160  Issue 2
238
the possibility of sella turcica bridging is present may
improve early diagnosis of canine impactions3 and allow
for proper interceptive management, therefore, reducing
the need for surgical exposure and comprehensive care.
Because sella turcica bridging is a developmental abnor-
mality that occurs at the prenatal stage9 and its dimen-
sions do not change significantly over time,37 early
diagnosis may occur during the first orthodontic
screening at age 7 years. In addition, the significant as-
sociation between sella turcica bridging and skeletal
Class III growth pattern suggests that early screening
for bridging could predict craniofacial growth.38
Because there is no standardized and objective clas-
sification of sella turcica bridging in the literature,
several authors have established their own quantitative
and qualitative classification systems. This study utilized
an objective formula established by Sundareswaran and
Nipun.21 Although this classification has the advantage
of measuring the extent of interclinoid calcification,
there was no evidence provided regarding the rationale
behind the cutoff point of 33%. In addition, several
other types of bridging involving the middle clinoid pro-
cesses have been reported and not accounted for using
this classification. When using lateral cephalograms to
diagnose sella turcica bridging, it is possible that a sella
turcica bridge involving the middle clinoid processes
may be inaccurately identified as an ossification between
the anterior and posterior clinoid processes.
CONCLUSIONS
The lateral cephalogram correctly identified all 10
patients with complete bridging but incorrectly classified
21 patients as complete because of the superimposition
of the bilateral interclinoid calcifications. Although all
subjects with complete bridging were correctly classified,
12.0% of subjects were incorrectly classified.
The overall accuracy of the lateral cephalogram as a
diagnostic modality in discriminating between no
bridging and partial bridging was good as determined
by receiver operator characteristic analysis with the
area under the curve values of 0.86 and 0.85 for right
and left sides, respectively. Although lateral cephalo-
grams overestimate patients with complete bridging
compared with CBCTs, they are a suitable screening mo-
dality to accurately suggest complete sella turcica
bridging and differentiating between patients of no
bridging and partial bridging.
AUTHOR CREDIT STATEMENT
Ashley Acevedo contributed to conceptualization,
data curation, methodology, investigation, formal anal-
ysis, original draft preparation, and draft review and
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Acevedo, Lagravere-Vich, and Al-Jewair
editing; Manuel Lagravere-Vich contributed to draft re-
view and editing; and Thikriat Al-Jewair contributed to
data curation, methodology, formal analysis, supervi-
sion, original draft preparation, and draft review and
editing.
ACKNOWLEDGMENTS
The authors thank Robert Dunford for his help with
the statistical analysis, Pamela Ortiz for her support
during the data collection stage, and Vandana Kumar
and Carlos Flores-Mir for their comments on the
manuscript.
REFERENCES
1. Axelsson S, Storhaug K, Kjaer I. Post-natal size and morphology of
the sella turcica. Longitudinal cephalometric standards for Norwe-
gians between 6 and 21 years of age. Eur J Orthod 2004;26:
597-604.
2. Skrzat J, Mroz I, Marchewka J. Bridges of the sella turcica - anat-
omy and topography. Folia Med Cracov 2012;52:97-101.
3. Ali B, Shaikh A, Fida M. Association between sella turcica bridging
and palatal canine impaction. Am J Orthod Dentofacial Orthop
2014;146:437-41.
4. Skrzat J, Szewczyk R, Walocha J. The ossified interclinoid liga-
ment. Folia Morphol (Warsz) 2006;65:242-5.
5. Carstens M. Die Selladiagnostik. Fortschr Geb Rontgenstr Nu-
klearmed 1949;71:257-72.
6. Bergland RM, Ray BS, Torack RM. Anatomical variations in the pi-
tuitary gland and adjacent structures in 225 human autopsy cases.
J Neurosurg 1968;28:93-9.
7. Cederberg RA, Benson BW, Nunn M, English JD. Calcification of
the interclinoid and petroclinoid ligaments of sella turcica: a radio-
graphic study of the prevalence. Orthod Craniofac Res 2003;6:
227-32.
8. Alkofide EA. The shape and size of the sella turcica in skeletal Class I,
Class II, and Class III Saudi subjects. Eur J Orthod 2007;29:457-63.
9. Becktor JP, Einersen S, Kjaer I. A sella turcica bridge in subjects
with severe craniofacial deviations. Eur J Orthod 2000;22:69-74.
10. Jones RM, Faqir A, Millett DT, Moos KF, McHugh S. Bridging and
dimensions of sella turcica in subjects treated by surgical-
orthodontic means or orthodontics only. Angle Orthod 2005;75:
714-8.
11. Kjaer I, Becktor KB, Lisson J, Gormsen C, Russell BG. Face, palate,
and craniofacial morphology in patients with a solitary median
maxillary central incisor. Eur J Orthod 2001;23:63-73.
12. Leonardi R, Farella M, Cobourne MT. An association between sella
turcica bridging and dental transposition. Eur J Orthod 2011;33:
461-5.
13. Meyer-Marcotty P, Weisschuh N, Dressler P, Hartmann J, Stellzig-
Eisenhauer A. Morphology of the sella turcica in Axenfeld-Rieger
syndrome with PITX2 mutation. J Oral Pathol Med 2008;37:
504-10.
14. Bahn CF, Falls HF, Varley GA, Meyer RF, Edelhauser HF,
Bourne WM. Classification of corneal endothelial disorders based
on neural crest origin. Ophthalmology 1984;91:558-63.
15. Childers NK, Wright JT. Dental and craniofacial anomalies of
Axenfeld-Rieger syndrome. J Oral Pathol 1986;15:534-9.
16. Leonardi R, Barbato E, Vichi M, Caltabiano M. A sella turcica
bridge in subjects with dental anomalies. Eur J Orthod 2006;
28:580-5.
Acevedo, Lagravere-Vich, and Al-Jewair
17. Kapila S, Conley RS, Harrell WE Jr. The current status of cone beam
computed tomography imaging in orthodontics. Dentomaxillofac
Radiol 2011;40:24-34.
18. Schulze D, Heiland M, Thurmann H, Adam G. Radiation exposure
during midfacial imaging using 4- and 16-slice computed tomog-
raphy, cone beam computed tomography systems and conven-
tional radiography. Dentomaxillofac Radiol 2004;33:83-6.
19. Wen J, Liu S, Ye X, Xie X, Li J, Li H, et al. Comparative study of
cephalometric measurements using 3 imaging modalities. J Am
Dent Assoc 2017;148:913-21.
20. Ludlow JB, Gubler M, Cevidanes L, Mol A. Precision of cephalo-
metric landmark identification: cone-beam computed tomography
vs conventional cephalometric views. Am J Orthod Dentofacial Or-
thop 2009;136:312: e1-10; discussion 312-3.
21. Sundareswaran S, Nipun CA. Bridging the gap: sella turcica in uni-
lateral cleft lip and palate patients. Cleft Palate Craniofac J 2015;
52:597-604.
22. Archana R, Anita R, Jyoti C, Punita M, Rakesh D. Incidence of
osseous interclinoid bars in Indian population. Surg Radiol Anat
2010;32:383-7.
23. Ortiz P. The association between sella turcica bridging and maxil-
lary palatal canine impaction: a CBCT study [thesis]. Buffalo: Uni-
versity at Buffalo, State University of New York; 2016.
24. Jewett CH. Teleroentgenography of the sella turcica with observa-
tions on one hundred normal cases. Am J Roentgenol 1920;7:
352-5.
25. M€ uller F. Die bedeutung der sellabr€ ucke f€
ur das auge. Klin Monbl
Augenheilkd Augenarztl Fortbild 1952;120:298-302.
26. Busch W. Morphology of sella turcica and its relation to the pitu-
itary gland. Virchows Arch Pathol Anat Physiol Klin Med 1951;
320:437-58.
27. Lang J. Structure and postnatal organization of heretofore unin-
vestigated and infrequent ossifications of the sella turcica region.
Acta Anat (Basel) 1977;99:121-39.
28. Kolagi S, Herur PG, Patil G, Rairam GB. Complete sella turcica
bridges prevalence and dimensions. J Anat Soc India 2011;60:22-5.
American Journal of Orthodontics and Dentofacial Orthopedics
239
29. Paula LK, Solon-de-Mello Pde A, Mattos CT, Ruellas AC,
Sant’Anna EF. Influence of magnification and superimposition
of structures on cephalometric diagnosis. Dental Press J Orthod
2015;20:29-34.
30. Kumar V, Ludlow JB, Mol A, Cevidanes L. Comparison of conven-
tional and cone beam CT synthesized cephalograms. Dentomaxil-
lofac Radiol 2007;36:263-9.
31. Lamichane M, Anderson NK, Rigali PH, Seldin EB, Will LA. Accu-
racy of reconstructed images from cone-beam computed tomog-
raphy scans. Am J Orthod Dentofacial Orthop 2009;136:156:
e1-6; discussion 156-7.
32. Cattaneo PM, Bloch CB, Calmar D, Hjortshøj M, Melsen B. Com-
parison between conventional and cone-beam computed
tomography-generated cephalograms. Am J Orthod Dentofacial
Orthop 2008;134:798-802.
33. Moshiri M, Scarfe WC, Hilgers ML, Scheetz JP, Silveira AM,
Farman AG. Accuracy of linear measurements from imaging plate
and lateral cephalometric images derived from cone-beam
computed tomography. Am J Orthod Dentofacial Orthop 2007;
132:550-60.
34. van Vlijmen OJ, Berge SJ, Swennen GR, Bronkhorst EM,
Katsaros C, Kuijpers-Jagtman AM. Comparison of cephalometric
radiographs obtained from cone-beam computed tomography
scans and conventional radiographs. J Oral Maxillofac Surg
2009;67:92-7.
35. Baumrind S, Frantz RC. The reliability of head film measurements.
1. Landmark identification. Am J Orthod 1971;60:111-27.
36. Baumrind S, Frantz RC. The reliability of head film measurements.
2. Conventional angular and linear measures. Am J Orthod 1971;
60:505-17.
37. Scribante A, Sfondrini MF, Cassani M, Fraticelli D, Beccari S,
Gandini P. Sella turcica bridging and dental anomalies: is there
an association? Int J Paediatr Dent 2017;27:568-73.
38. Meyer-Marcotty P, Reuther T, Stellzig-Eisenhauer A. Bridging of
the sella turcica in skeletal Class III subjects. Eur J Orthod 2010;
32:148-53.
August 2021  Vol 160  Issue 2
ORIGINAL ARTICLE

Simulation of orthodontic tooth

movement during activation of an
innovative design of closing loop using
the finite element method
Tuan Nguyen Anh,a Ryo Hamanaka,a Sachio Jinnai,a Hiroya Komaki,a Satoshi Yamaoka,a Jun-ya Tominaga,a
Yoshiyuki Koga,b and Noriaki Yoshidaa
Nagasaki, Japan

Introduction: Although many attempts have been made to study the mechanical behavior of closing loops, most
have been limited to analyses of the magnitude of forces and moments acting on the end of the closing loop. The
objectives of this study were to simulate orthodontic tooth movement during the activation of a newly designed
closing loop combined with a gable bend and to investigate the optimal loop activation condition to achieve the
desired tooth movement. Methods: We constructed a 3-dimensional model of maxillary dentition reproducing
the state wherein a looped archwire combined with a gable bend was engaged in brackets and tubes.
Orthodontic tooth movements were simulated for both anterior and posterior teeth while varying the degree of
gable bend using the finite element method. Results: The incorporation of a 5 gable bend into the newly de-
signed closing loop produced lingual crown tipping for the central incisor and bodily movement for the first molar.
The incorporation of 10 and 15 gable bends produced bodily movement and root movement, respectively, for
the central incisor and distal tipping for the first molar. Conclusions: Torque control of the anterior teeth and
anchorage control of the posterior teeth can be carried out effectively and simply by reducing by half the thick-
ness of a teardrop loop with a height of 10 mm and a 0.019 3 0.025-in cross-section, to a distance of 3 mm from
its apex, and by incorporating various degrees of gable bend into the loop corresponding to the treatment plan.
(Am J Orthod Dentofacial Orthop 2021;160:240-9)

S
ince the establishment of the Tweed technique,
closing loops have been widely used for space
closure in the treatment of premolar extraction
patients.1,2 Contrary to sliding mechanics, wherein fric-
tion generated during space closure can reduce the rate
of tooth movement,3,4 loop mechanics is a frictionless
treatment system. Therefore, this technique has the po-
tential to produce predetermined moment-to-force (M/
F) ratios to achieve controlled movement of the anterior
teeth accurately.2,5-7 In contrast, the loop mechanics
a
Department of Orthodontics and Dentofacial Orthopedics, Nagasaki University
Graduate School of Biomedical Sciences, Nagasaki, Japan.
b
Department of Orthodontics, Nagasaki University Hospital, Nagasaki, Japan.
All authors have completed and submitted the ICMJE Form for Disclosure of
Potential Conflicts of Interest, and none were reported.
Address correspondence to: Noriaki Yoshida, Department of Orthodontics and
Dentofacial Orthopedics, Nagasaki University Graduate School of Biomedical Sci-
ences, 1-7-1 Sakamoto, Nagasaki 852-8588, Japan; e-mail, nori@nagasaki-u.
ac.jp.
Submitted, November 2019; revised, March 2020; accepted, April 2020.
0889-5406/$36.00
Ó 2021.
https://doi.org/10.1016/j.ajodo.2020.04.026
system has several disadvantages. That is, the M/F
ratio produced by conventionally designed loops is
too low to achieve bodily or root movement of the
anterior teeth.5 Another shortcoming is that closing
loops made of stainless steel wire could produce an
excessively heavy force, especially with the incorporation
of gable bends.2,5,8,9 Furthermore, the use of an archwire
with a larger cross-section could generate a greater
retraction force in the case of the 0.022-in bracket
slot system.
To generate a low force magnitude and increase the
M/F ratio for achieving better control of the anterior
teeth during space closure, many loop designs with
complicated shapes have been developed by extending
both the horizontal length and the vertical height.10-20
Sumi et al21 developed a simple design of a 10-mm-
high teardrop loop with a wire cross-section reduced in
thickness by 50% for a distance of 3 mm from the
apex. This newly designed closing loop produces a
gentle force and a higher M/F ratio of 9.3 without add-
ing a gable bend. However, the M/F ratios obtained from

240
Anh et al
Fig 1. A, Three-dimensional finite element model of maxillary dentition, into which a looped archwire is
engaged into brackets and tubes. B, A 10-mm-high teardrop loop, with the wire cross-section reduced
in thickness by 50% for a distance of 3 mm from the apex.
these numerical studies do not apply directly to the
incisor but rather to the canine bracket adjacent to the
loop. However, because of the difficulty of conducting
numerical analysis for the mechanical behavior of a
looped archwire, the tooth movement generated by acti-
vating this new closing loop design has not yet been
investigated. The purposes of the present study were
to simulate orthodontic tooth movement of not only
the anterior but also the posterior teeth after the appli-
cation of the newly designed loop with a partially
reduced wire cross-section in the 0.022-in bracket slot
system and to investigate the effect of the degree of ga-
ble bend on the tooth movement pattern using the finite
element method (FEM).
MATERIAL AND METHODS
Three-dimensional (3D) images of a maxillary denti-
tion, taken with a multi-image cone-beam computed
tomography scanner (3DX; J. Morita, Kyoto, Japan),
were exported as digital imaging and communication
in medicine file to 3D image processing and editing
software (Mimics, version 10.02; Materialise, Leuven,
Belgium). After reconstructing the digital imaging and
communication in medicine data to 3D surface data,
they were imported to FEM preprocessing and postpro-
cessing software (Patran 2017; MSC Software, Los
Angeles, Calif). After remeshing, the teeth were aligned
to the ideal positions of the normal dentition, assuming
that the case model was diagnosed as maxillary
protrusion, the first premolar extractions were indicated
(Fig 1, A).
Each tooth model was constructed using thin shell
elements with a thickness of 3 mm to develop a more
simplified model for reducing the analysis time. A Young
modulus of 18,600 MPa and a Poisson ratio of 0.3 were
assigned in accordance with previous studies.22,23 The
American Journal of Orthodontics and Dentofacial Orthopedics
241
periodontal ligament (PDL) was constructed using
wedge elements with a uniform thickness of
0.2 mm,24-26 Young's modulus of 0.05 MPa, and a
Poisson ratio of 0.3.26-28
The fixed orthodontic appliance was composed of
0.022 3 0.028-in slot brackets and a 0.019 3 0.025-
inch stainless steel archwire. These were constructed us-
ing solid modeler software (Solidworks 2016; Dassault
Syst
emes, Velizy-Villacoublay, France) and imported to
the preprocessing software. An archwire and brackets
were modeled using 8-node hexahedral elements with
Young's modulus of 200 GPa and a Poisson ratio of
0.3.27,28 All materials were set to be isotropic. Each
bracket was attached to the corresponding tooth with
rigid body elements. A symmetric model of the maxillary
dentition was developed to reduce the number of ele-
ments and minimize the analysis time. The front end
of the archwire, on the midsagittal plane, was con-
strained in the transverse (x) direction for symmetric
analysis.
Long-term orthodontic tooth movements were simu-
lated using a bone remodeling algorithm that has been
described in detail elsewhere.29 Simply, in the first
step, an orthodontic force is applied to the tooth while
displacements are constrained at the outer surface of
the PDL. After an initial tooth displacement is produced,
each node forming the outer surface of the PDL is dis-
placed so that the PDL is restored to its original config-
uration and a thickness of 0.2 mm in the next step. In the
following step, a second orthodontic force is applied to
the tooth while displacements are retrained at the outer
nodes of the PDL, and again, the geometry of the PDL is
updated. Then, 2 steps—the initial displacement and
bone remodeling—are iterated to simulate orthodontic
tooth movement on the assumption that initial tooth
displacement could be a predictor of long-term tooth
August 2021  Vol 160  Issue 2
242
Fig 2. A, Actuator element connecting the distal end of
the second molar tube and a point distal to the tube on
the archwire. B, Loop activation by lengthening the length
of the actuator element.
movement.30 All analyses were performed using a FEM
package (Marc 2016; MSC Software).
The basic design of the closing loop examined in this
study was a teardrop that was 10 mm in height and
2 mm in width, bent from a 0.019 3 0.025-in stainless
steel archwire. The thickness of the cross-section of the
loop was reduced by 50% for a distance of 3 mm from
the apex (Fig 1, B).
The tooth movement during space closure was simu-
lated by activating the loop by 1 mm using an actuator
element connecting the distal end of the second molar
tube and a point distal to the tube on the archwire
(Fig 2). The actuator element is a truss element with var-
iable length. Lengthening of the actuator element
causes distal movement of the posterior portion of the
archwire, which enables activation of the loop in the
same manner as a tie-back in clinical practice.
We tested 5 different mechanical conditions of loop
activation. First, we investigated the effect of a reduction
in the cross-section at the loop apex by comparing the
tooth movements when the thickness of the cross-
section of the loop apex was reduced by 0% and 50%.
Next, we compared the effect of different degrees of ga-
ble bend (5 , 10 , and 15 ) incorporated into the loop on
the tooth movement pattern.
The loading condition, wherein the loop with gable
bend is activated, was simulated in the following steps.
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Anh et al
First, we constructed the model representing the state
of neutral activation, namely, the state before activation
of the loop, instead of engaging the bent archwire with
the closing loop combined with a gable bend into the
brackets and tubes. To construct the model in which
the closing archwire combined with a gable bend is
passively inserted into the brackets and tubes on the
posterior teeth (Fig 3, A), 2 coordinate systems were es-
tablished on the basis of the loop configuration (Fig 3,
B). One is the coordinate system whose origin is located
at the corner of the loop base of the anterior portion of
the archwire, and the other is the coordinate system
whose origin is located at the corner of the loop base
of the posterior portion of the archwire (CS2). The orig-
inal model was divided into 3 segments. The anterior
segment consisted of the anterior portion of the archwire
mesial to the loop, including the anterior teeth models
and corresponding PDLs, the middle segment was the
closing loop, and the posterior segment included the
posterior portion of the archwire distal to the loop,
including the posterior teeth models and corresponding
PDLs.
The loop and posterior segments of the model were
rotated in 2 steps (Fig 3, C-E). First, the posterior
segment was rotated around the z-axis at the origin of
CS2 by one-half of the angulation of the gable bend
(Fig 3, C). That is, each node forming the archwire distal
to CS2 was rotated. At the same time, the posterior teeth
and corresponding brackets were also rotated while
maintaining their relative position to the archwire.
Next, the loop and posterior segments were rotated
around the z-axis at the origin of the coordinate system
whose origin is located at the corner of the loop base of
the anterior portion of the archwire by one-half of the
angulation of the gable bend (Fig 3, D). The model
before activating the loop with a gable bend was con-
structed in this manner (Fig 3, A and E).
Next, the mechanical conditions of loop activation
were reproduced as follows in the case that a gable
bend is incorporated into the loop. Every node of the
loop and posterior segments of the model was displaced
to its original position (Fig 4, A and B). Because the
archwire was engaged into the brackets, it also caused
the displacement of the archwire, including the loop
combined with a gable bend. Then, the loop was acti-
vated by 1 mm using the actuator element (Fig 4, C),
and space was closed after the movements of the ante-
rior and posterior teeth (Fig 4, D). At this time, displace-
ments were constrained at the outer surface of the PDL,
and consequently, long-term tooth movement was
simulated on the basis of a bone remodeling algorithm.7
Tooth movements of the central incisor and first molar
Anh et al 243

Fig 3. Procedure for constructing the FEM model in which the closing archwire combined with a gable
bend is passively inserted into the brackets and tubes on the posterior teeth: A, two coordinate systems
of the anterior and posterior parts of the loop model. Origins are indicated by red points; B, FEM model
in which the bent closing archwire with a gable bend is passively engaged; C, the posterior segment
was rotated around the z-axis at the origin CS2 by one-half degree angles of gable bend; D, the
loop and posterior segments were rotated around the z-axis at the origin of CS1 by one-half degree
angle of gable bend; E, after construction of the model into which the looped archwire with a gable
bend was engaged into the brackets and tubes. CS1, the coordinate system whose origin is located
at the corner of the loop base of the anterior portion of the archwire.

were represented as a combination of translation and
rotational displacement at the center of resistance (CR)
in the sagittal plane.
RESULTS
Figure 5 shows the relationship between the degree
of rotation and translational displacement at the CR of
the central incisor in the sagittal plane when comparing
the loops with and without reduction. The horizontal
axis shows the translational displacement of the CR in
the sagittal plane, the vertical axis indicates the degree
of labiolingual tipping, and the negative signs indicate
lingual crown tipping. The degree of lingual crown
tipping was increased almost linearly with increases in
the amount of translational displacement in both pa-
tients. The degree of lingual crown tipping was 1.56
on the loop activation of 1 mm for the normal loop
without reduction. In contrast, the degree of lingual
crown tipping for the loop in which the thickness of
the cross-section was reduced by 50% for a distance
of 3 mm from the apex was 0.78 , which was half that
of the normal loop.
Figure 6 shows the degree of rotation as a function of
translational displacement at the CR of the central
incisor in the sagittal plane when varying the degree of
gable bend (from 0 , 5 , 10 , and 15 ) incorporated
into the loop in which the thickness of the cross-
section was reduced by 50% for a distance of 3 mm
from the apex. The vertical axis indicates the degree of
labiolingual tipping. Positive signs indicate lingual root
tipping, and negative signs lingual crown tipping.
When gable bends of 0 , 5 , 10 , and 15 were incorpo-
rated into the loop, the degree of rotation of the central
incisor was 0.78 , 0.50 , 0.04 , and 1.14 , respec-
tively. The degree of lingual crown tipping with a gable
bend of 5 was smaller than that with a gable bend of 0 .
When a gable bend of 10 was placed, the central incisor
showed lingual root tipping, although the value was
quite small. With a gable bend of 15 , a higher degree
of lingual root tipping was observed.
Figure 7 shows the relationship between the degree
of rotation and translational displacement at the CR of
the first molar in the sagittal plane when comparing
the loops with and without reduction. The vertical axis

American Journal of Orthodontics and Dentofacial Orthopedics August 2021  Vol 160  Issue 2
244 Anh et al

Fig 4. Procedure for activating the closing loop combined with a gable bend: A, the state wherein the
looped archwire combined with a gable bend is passively engaged in the brackets and tubes on the
posterior teeth. A curved arrow indicates the direction of rotational displacement of the posterior
segment; B, after displacement of the loop and posterior segments to their original positions; C, acti-
vation of the loop using an actuator element; D, after deactivation of the loop.

indicates the degree of mesiodistal tipping. Positive
signs indicate mesial tipping, and negative signs distal
tipping. The degree of mesial tipping of the first molar
was 0.84 on the loop activation of 1 mm for the normal
loop without reduction. When the cross-section of the
loop was partially reduced, the degree of mesial tipping
was 0.35 , which was less than half that of the normal
loop.
Figure 8 shows the degree of rotation as a function of
the translational displacement at the CR of the first
molar in the sagittal plane when varying the degree of
gable bend (from 0 , 5 , 10 , and 15 ) incorporated

August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Anh et al
Fig 5. Degree of rotation of the central incisor as a func-
tion of the translational displacement of the CR in the
sagittal plane when comparing the loops with and without
reduction. Negative signs indicate lingual crown tipping.
Fig 6. Degree of rotation of the central incisor as a func-
tion of the translational displacement of the CR in the
sagittal plane when the newly designed loop was acti-
vated combined with gable bends of 0 , 5 , 10 , and
15 . Positive signs indicate lingual root tipping, and nega-
tive signs lingual crown tipping.
into the loop in which the thickness of the cross-section
was reduced by 50% for a distance of 3 mm from the
apex. When gable bends of 0 , 5 , 10 , and 15 were
incorporated into the loop, the degree of rotation of
the first molar was 0.35 , 0.02 , 0.20 and 0.52 ,
respectively. The first molar showed distal tipping
when 5 , 10 , and 15 of gable bends were incorporated,
and the higher the degree of gable bend, the more
distally the first molar tipped. By contrast, mesial tipping
of 0.35 was observed when no gable bend was incorpo-
rated. With a gable bend of 5 , the degree of distal
tipping was 0.02, which was almost zero. That is, bodily
movement of the first molar was likely to occur by
American Journal of Orthodontics and Dentofacial Orthopedics
245
Fig 7. Degree of rotation of the first molar as a function of
the translational displacement of the CR in the sagittal
plane when comparing the loops with and without reduc-
tion. Positive signs indicate mesial tipping.
Fig 8. Degree of rotation of the first molar as a function of
the translational displacement of the CR in the sagittal
plane when the newly designed loop was activated com-
bined with gable bends of 0 , 5 , 10 , and 15 . Positive
signs indicate mesial tipping, and negative signs distal
tipping.
adding a gable bend of 5 to the newly designed closing
loop.
DISCUSSION
The loop mechanics system has been widely used to
close the extraction space in orthodontic treatment.
Although many attempts have been made to study the
mechanical behavior of the closing loop,15,20,31,32
because of several difficulties, it has not been clarified
what condition of loop activation or design allows for
achieving controlled movement of the anterior and pos-
terior teeth. First, it is extremely difficult to prescribe
displacement boundary conditions in FEM analysis,
August 2021  Vol 160  Issue 2
246
especially when a gable bend is incorporated into the
loop. That is, on insertion of the bent archwire into the
brackets, each node on the archwire corresponding to
the mesial and distal ends of the brackets should be
accurately displaced to the bracket position. Because
the procedure for prescribing boundary conditions on
the interface between the bracket and the archwire be-
comes enormously complicated, such analyses would
not necessarily converge to a solution.
Furthermore, most previous studies11,14,15,33-36 have
been limited to conducting analyses to determine the
magnitudes of force and moment acting on the end of
the closing loop. In other words, the value of the M/F
ratio produced at the bracket adjacent to the loop, not
on the central incisor or first molar, was obtained from
the numeric analyses. In addition, it remains difficult
to predict actual tooth movements after the
application of certain values of force and moment
because the force system changes in the course of loop
deactivation.
The above-mentioned unsolved problems could be
successfully overcome in the present study. We con-
structed a model reproducing the state in which the
looped archwire with a gable bend is passively engaged
into the brackets without exerting force and moment on
each tooth by rotating each node forming the posterior
segment corresponding to the applied degree of gable
bend. After the insertion of the looped archwire with a
gable bend, the posterior segment was displaced into
its original position along with the posterior portion of
the archwire, which facilitated the procedure for pre-
scribing boundary conditions on the interface between
the bracket and the archwire. In addition, the movement
patterns of the central incisor and first molar during the
course of loop deactivation could be simulated using the
FEM that could predict long-term orthodontic tooth
movement on the basis of a bone remodeling
algorithm.29
However, the PDL was constructed as a linear
isotropic material to develop a more simplified model,
which substantially reduces the analysis time in this
study. Because the previous study37 suggested that the
nonlinear behavior of the PDL should be taken into
consideration for precise simulation of tooth movement,
nonlinear finite element analysis will be necessary for a
future study of ours. In addition, further clinical tests
and investigations by combining the Taguchi method
and FEM to analyze more design factors at the same
time would help to precisely determine the ideal ortho-
dontic design for tooth movement.
Although the frictionless technique of loop me-
chanics has the potential to produce a preprogrammed
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Anh et al
force system for achieving the desired type of tooth
movement by incorporating gable bends into closing
loops,8 the M/F ratio generated by conventional loops
is too low to achieve controlled movement of the ante-
rior teeth.2,6,8,9,11,13,19,20 To increase the M/F ratio and
thereby achieve better control of the anterior teeth dur-
ing space closure, many loop designs with complicated
shapes have been developed by extending the horizontal
length and the vertical height.10-20 Sumi et al21 devel-
oped a simple design of a 10-mm-high teardrop loop,
with a 0.019 3 0.025-in cross-section reduced in thick-
ness by 50% for a distance of 3 mm from the apex, and
therefore applicable in the 0.022-in bracket slot system.
This newly designed closing loop produces a gentle
force, even with a 0.019 3 0.025-in wire and a high
M/F ratio (9.3) without adding a gable bend. According
to previous studies,2,7-9,38,39 bodily movement of the
anterior teeth is achieved by applying an M/F ratio of
10; therefore, this loop design is more likely to produce
bodily movement. However, the tooth movement gener-
ated by activating this new design of the closing loop has
not yet been analyzed in a practical manner. The present
study aimed to investigate how not only the anterior
teeth but also the posterior teeth will move when em-
ploying the loop with a partially reduced wire
cross-section.
FEM analyses showed that a cross-sectional reduc-
tion in the apical portion of the loop would reduce the
degree of lingual crown tipping of the central incisor
by 50%, thereby preventing uncontrolled tipping (Fig
5). This finding was supported by a previous study21 re-
porting that the M/F ratio increased from 5.8 to 9.3 as
the reduction rate of the wire cross-section increased
from 0% to 50%. Although the previous study suggested
that the central incisor was expected to show a move-
ment pattern close to bodily movement, the present
study clearly demonstrated that lingual crown tipping
was generated. This result may be partly because of
the play between the brackets and archwire, which
causes the loss of anterior torque control. In addition,
the M/F ratio of 9.3 acting on the loop, which was placed
distal to the canine, could not be effectively transmitted
to the central incisor because the canine might absorb a
greater part of the antitipping moment generated by
loop activation. As a result, it is considered that bodily
movement is not produced through the application of
an M/F ratio of 9.3 acting on the loop end.
The degree of gable bend incorporated into the loop
made a difference in central incisor tipping patterns (Fig
6). When a gable bend of 5 was incorporated into the
loop, lingual crown tipping was observed after the
completion of loop deactivation. Increasing the degree
Anh et al
of gable bend from 5 to 10 decreased the degree of
lingual crown tipping from 0.50 to 0.04 , which
was close to zero. In other words, bodily movement is
likely to occur with a gable bend of 10 . In contrast,
the placement of a gable bend of 15 produced lingual
root tipping, which indicates that the antitipping
moment generated at the bracket on the central incisor,
which resists the tendency of lingual crown tipping, is
increased with increases in the degree of gable bend.
Thus, the results obtained from the present study sug-
gest that a reduction in the cross-section at the loop
apex would substantially increase the M/F ratio and de-
creases the traction force, which enables us to use loop
mechanics combined with gable bend even in the
0.022-in bracket slot system. Our newly developed clos-
ing loop has the potential to achieve the desired anterior
tooth movement, such as lingual crown tipping, bodily
movement, or root movement, by combining it with
the placement of different degrees of gable bend de-
pending on the patient's clinical condition.
In regard to molar movement, mesial tipping was
observed when activating the normal loop without
cross-sectional reduction (Fig 7). The degree of mesial
tipping was decreased when the wire cross-section of
the loop was partially reduced. Increasing the degree
of gable bend from 0 to 5 for the newly designed
loop changed the direction of molar tipping from mesial
to distal, although the tipping degree was almost zero. In
other words, with a gable bend of 5 , the first molar
moved almost bodily in the mesial direction (Fig 8),
and with gable bends of 10 and 15 , the degree of distal
tipping of the first molar was increased to 0.20 and
0.52 , respectively, which indicates that incorporating
a gable bend larger than 10 into the loop produced
distal tipping of the first molar. This finding suggests
that placing a gable bend into the loop with cross-
sectional reduction could effectively reinforce the
anchorage by preventing mesial tipping of the posterior
teeth.
Because mesial tipping of the posterior teeth, as well
as lingual crown tipping of the anterior teeth, could
cause a vertical bowing effect, thereby generating deep
overbite, further closure of the residual extraction space
would be prevented. The use of this newly designed clos-
ing loop could further improve treatment efficiency and
shorten treatment time by avoiding the side effects
generated during space closure.
The incorporation of a 5 gable bend into our newly
designed closing loop can produce lingual crown tipping
of the central incisor for the correction of Angle Class II,
Division 1 malocclusion. The placement of a 10 gable
bend into the loop can achieve bodily movement of
the central incisor when its inclination is within the
American Journal of Orthodontics and Dentofacial Orthopedics
247
normal range. The incorporation of a 15 gable bend
into the loop can achieve root movement of the central
incisor for the correction of Angle Class II, Division 2
malocclusion. At the same time, the placement of the
gable bend can minimize anchorage loss by moving
the first molar bodily or tipping it distally.
Clinically, a teardrop loop in the height of 10 mm is
applicable for the maxillary dentition in most patients.
However, for the patients with shallow mucobuccal
fold or the mandibular dentition, a 10 mm-high loop
may cause impingement to the mucosa. In such patients,
the loop height should be decreased from 10 to 8 mm.
Burstone and Koenig2 reported that the vertical height
of the loop is the dominant factor influencing the M/F
ratio, and the shorter the loop, the lower the M/F ratio
is produced. Thus, the reduction of the loop height de-
creases the M/F ratio, which could cause loss of control
of anterior tooth movement during space closure. It is,
therefore, recommended to place a certain degree of
twist in the anterior portion of the archwire for applying
additional torque.
Further confirmation of these findings, which were
obtained using FEM analyses, in a clinical study would
be a great help in establishing an optimal treatment
plan and achieving speedy, effective, and accurate or-
thodontic tooth movement.
CONCLUSIONS
The finite element model developed in the present
study enabled the prediction of orthodontic tooth move-
ment when loop mechanics combined with a gable bend
was employed for space closure. The analysis showed
that not only torque control of the anterior teeth but
also anchorage control of the posterior teeth could be
carried out effectively by simply reducing the thickness
of a teardrop loop with a height of 10 mm and a
0.019 3 0.025-in cross-section by 50% for a distance
of 3 mm from its apex, and by incorporating various de-
grees of gable bend into the loop corresponding to the
case. The additional advantages of this new closing
loop are that it is easily fabricated at the chairside by
grinding with a turbine handpiece and applicable in
the 0.022-in bracket slot system.
ACKNOWLEDGMENTS
Funding: This research did not receive any specific
grant from funding agencies in the public, commercial,
or not-for-profit sectors.
CREDIT AUTHORSHIP CONTRIBUTION STATEMENT
Tuan Nguyen Anh: Investigation, Software, Writing -
review & editing, Conceptualization, Methodology. Ryo
August 2021  Vol 160  Issue 2
248
Hamanaka: Methodology. Sachio Jinnai: Validation.
Hiroya Komaki: Conceptualization, Methodology, Vali-
dation. Satoshi Yamaoka: Validation. Jun-ya Tominaga:
Resources, Writing - review & editing. Yoshiyuki Koga:
Resources, Writing - review & editing. Noriaki Yoshida:
Writing - review & editing, Supervision, Project
administration.
REFERENCES
1. Tweed CH. The application of the principles of the edge-wise arch
in the treatment of malocclusions: I. Angle Orthod 1941;11:5-11.
2. Burstone CJ, Koenig HA. Optimizing anterior and canine retrac-
tion. Am J Orthod 1976;70:1-19.
3. Nanda R, Ghosh J. Biomechanical considerations in sliding me-
chanics. In: Nanda R, editor. Biomechanics in Clinical Orthodon-
tics. Philadelphia: Saunders; 1997. p. 188-217.
4. Tominaga JY, Ozaki H, Chiang PC, Sumi M, Tanaka M, Koga Y,
et al. Effect of bracket slot and archwire dimensions on anterior
tooth movement during space closure in sliding mechanics: a 3-
dimensional finite element study. Am J Orthod Dentofacial Orthop
2014;146:166-74.
5. Proffit WR. Closure of extraction space. In: Proffit WR, Fields HW,
Sarver DM, editors. Contemporary Orthodontics. 4th ed. St Louis:
Mosby Elsevier; 2007. p. 592-601.
6. Nanda R, Ghosh J. Biomechanical basis of extraction space closure.
In: Nanda R, editor. Biomechanics in Clinical Orthodontics. Phila-
delphia: Saunders; 1997. p. 156-87.
7. Nanda R, Ghosh J. Principles of biomechanics. In: Nanda R, editor.
Biomechanics in Clinical Orthodontics. Philadelphia: Saunders;
1997. p. 6-8.
8. Braun S, Garcia JL. The Gable bend revisited. Am J Orthod Dento-
facial Orthop 2002;122:523-7.
9. Yoshida N, Jost-Brinkmann PG, Koga Y, Kobayashi K, Obiya H,
Peng CL. Moment/Kraft-Verh€altnisse von Kontraktionsb€ogen
w€ahrend der Deaktivierung. Kieferorthop€adie 2004;18:175-82:
German.
10. Dinçer M, G€ulsen A, T€
urk T. The retraction of upper incisors with
the PG retraction system. Eur J Orthod 2000;22:33-41.
11. Coimbra MER, Penedo ND, de Gouv^ea JP, Elias CN, de Souza
Araujo MT, Coelho PG. Mechanical testing and finite element
analysis of orthodontic teardrop loop. Am J Orthod Dentofacial
Orthop 2008;133:188: e9-13.
12. Burstone CJ. The segmented arch approach to space closure. Am J
Orthod 1982;82:361-78.
13. Techalertpaisarn P, Versluis A. How do mechanical responses at
closing loop ends vary when loop position changes? A systematic
analysis of vertical, T-, and L-loops. Oral Sci Int 2013;10:
58-64.
14. Techalertpaisarn P, Versluis A. Mechanical properties of Opus clos-
ing loops, L-loops, and T-loops investigated with finite element
analysis. Am J Orthod Dentofacial Orthop 2013;143:675-83.
15. Kuhlberg AJ, Burstone CJ. T-loop position and anchorage control.
Am J Orthod Dentofacial Orthop 1997;112:12-8.
16. Rose D, Quick A, Swain M, Herbison P. Moment-to-force charac-
teristics of preactivated nickel-titanium and titanium-
molybdenum alloy symmetrical T-loops. Am J Orthod Dentofacial
Orthop 2009;135:757-63.
17. Martins RP, Buschang PH, Martins LP, Gandini LG Jr. Optimizing
the design of preactivated titanium T-loop springs with Loop soft-
ware. Am J Orthod Dentofacial Orthop 2008;134:161-6.
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Anh et al
18. Xia Z, Chen J, Jiangc F, Li S, Viecilli RF, Liu SY. Load system of
segmental T-loops for canine retraction. Am J Orthod Dentofacial
Orthop 2013;144:548-56.
19. Siatkowski RE. Continuous arch wire closing loop design, optimi-
zation, and verification. Part I. Am J Orthod Dentofacial Orthop
1997;112:393-402.
20. Siatkowski RE. Continuous arch wire closing loop design, optimi-
zation, and verification. Part II. Am J Orthod Dentofacial Orthop
1997;112:487-95.
21. Sumi M, Koga Y, Tominaga JY, Hamanaka R, Ozaki H, Chiang PC,
et al. Innovative design of closing loops producing an optimal
force system applicable in the 0.022-in bracket slot system. Am J
Orthod Dentofacial Orthop 2016;150:968-78.
22. Romeed SA, Malik R, Dunne SM. Stress analysis of occlusal forces
in canine teeth and their role in the development of non-carious
cervical lesions: abfraction. Int J Dent 2012;2012:234845.
23. Cai Y, Yang X, He B, Yao J. Finite element method analysis of the
periodontal ligament in mandibular canine movement with trans-
parent tooth correction treatment. BMC Oral Health 2015;15:106.
24. Coolidge ED. The thickness of the human periodontal membrane.
J Am Dent Assoc Dent Cosmos 1937;24:1260-70.
25. Vollmer D, Bourauel C, Maier K, J€ager A. Determination of the
centre of resistance in an upper human canine and idealized tooth
model. Eur J Orthod 1999;21:633-48.
26. Reimann S, Keilig L, J€ager A, Bourauel C. Biomechanical finite-
element investigation of the position of the centre of resistance
of the upper incisors. Eur J Orthod 2007;29:219-24.
27. Ozaki H, Tominaga JY, Hamanaka R, Sumi M, Chiang PC,
Tanaka M, et al. Biomechanical aspects of segmented arch me-
chanics combined with power arm for controlled anterior tooth
movement: a three-dimensional finite element study. J Dent Bio-
mech 2015;6:1758736014566337.
28. Tominaga JY, Tanaka M, Koga Y, Gonzales C, Kobayashi M,
Yoshida N. Optimal loading conditions for controlled movement
of anterior teeth in sliding mechanics. Angle Orthod 2009;79:
1102-7.
29. Hamanaka R, Yamaoka S, Anh TN, Tominaga JY, Koga Y,
Yoshida N. Numeric simulation model for long-term orthodontic
tooth movement with contact boundary conditions using the finite
element method. Am J Orthod Dentofacial Orthop 2017;152:
601-12.
30. Soenen PL, Dermaut LR, Verbeeck RM. Initial tooth displacement
in vivo as a predictor of long-term displacement. Eur J Orthod
1999;21:405-11.
31. Ferreira MA, Assumpç~ao R, Luersen MA, Borges PC. Mechanical
behaviour of a prototype orthodontic retraction spring: a
numerical-experimental study. Eur J Orthod 2013;35:414-20.
32. Martins RP, Buschang PH, Gandini LG Jr. Group A T-loop for dif-
ferential moment mechanics: an implant study. Am J Orthod Den-
tofacial Orthop 2009;135:182-9.
33. Vanderby R Jr, Burstone CJ, Solonche DJ, Ratches JA. Experimen-
tally determined force systems from vertically activated orthodon-
tic loops. Angle Orthod 1977;47:272-9.
34. Almeida L, Ribeiro A, Parsekian Martins R, Viecilli R, Parsekian
Martins L. Nickel titanium T-loop wire dimensions for en masse
retraction. Angle Orthod 2016;86:810-7.
35. Hoenigl KD, Freudenthaler J, Marcotte MR, Bantleon HP. The
centered T-loop–a new way of preactivation. Am J Orthod Dento-
facial Orthop 1995;108:149-53.
36. Rodrigues FRM, Borges PC, Luersen MA, Ferreira MdA. Three-
dimensional analysis of an orthodontic delta spring. Rev Bras
Eng Biomed 2014;30:248-56.
Anh et al
37. Bourauel C, Freudenreich D, Vollmer D, Kobe D, Drescher D,
J€ager A. Simulation of orthodontic tooth movements. A compari-
son of numerical models. J Orofac Orthop 1999;60:136-51.
38. Chiang PC, Koga Y, Tominaga J, Ozaki H, Hamanaka R, Sumi M,
et al. Effect of gable bend incorporated into loop mechanics on
American Journal of Orthodontics and Dentofacial Orthopedics
249
anterior tooth movement: comparative study between en masse
retraction and two-step retraction. Orthod Waves 2015;74:55-61.
39. Burstone C, van Steenbergen E, Hanley K. Modern edgewise me-
chanics and the segmented arch technique. Ormco: Farmington;
1995.
August 2021  Vol 160  Issue 2
ORIGINAL ARTICLE

Efficacy of Invisalign attachments:

A retrospective study
Theresa Karras,a Maharaj Singh,b Emelia Karkazis,b Dawei Liu,c Ghada Nimeri,c and Bhoomika Ahujac
Milwaukee, Wis, and Chicago, Ill

Introduction: The purpose of this study was to compare the efficacy of Invisalign's (Align Technology, Santa
Clara, Calif) optimized and conventional attachments on rotational and extrusive tooth movements.
Methods: Initial, predicted, and achieved digital dental models from 100 orthodontic patients were exported
from Invisalign's ClinCheck software as stereolithography files and subsequently imported into the Slicer CMF
program (version 4.7.0; http://www.slicer.org) for superimpositions on posterior teeth with no planned
movement. Rotational and extrusive measurements for both optimized and conventional attachments were
made on 382 teeth from the superimposition of the initial and predicted models (predicted movement) and
from the superimposed initial and achieved models (achieved movement). Predicted and achieved
movements were compared along with movements of teeth with optimized and conventional attachments.
Results: Differences between accuracies of tooth movements using optimized vs conventional attachments
for both rotation and extrusion were neither statistically nor clinically significant. Mean predicted values were
larger than mean achieved values for all attachment types and movements (P \ 0.0001). For extrusion, the
mean difference between predicted and achieved movements was clinically significant (0.40 mm and 0.62 mm
for optimized and conventional attachments, respectively). Overall, the mean accuracy was 57.2%. Mean
accuracy was 63.2% for rotation and 47.6% for extrusion. Interproximal reduction or spacing did not
significantly affect accuracy. Conclusions: Conventional attachment types may be just as effective as
Invisalign's proprietary optimized attachments for rotations of canines and premolars and extrusion of incisors
and canines. Clinicians should consider overcorrecting tooth movements, especially anterior tooth extrusion.
(Am J Orthod Dentofacial Orthop 2021;160:250-8)

A
lthough fixed orthodontic appliances are still
widely used today, the advent of removable clear
aligners has undoubtedly revolutionized the
field of orthodontics in recent years. In 1997, Align
Technology (Santa Clara, Calif) developed Invisalign,
which is arguably the most used and recognizable clear
aligner system today. Initially, each Invisalign aligner
was programmed to move a tooth 0.25 to 0.33 mm
over 14 days.1,2 In 2016, Invisalign changed its protocol
a
Formerly, Department of Orthodontics, Marquette University School of
Dentistry, Milwaukee, Wis; currently, Private practice, Chicago, Ill.
b
Marquette University School of Dentistry, Milwaukee, Wis.
c
Department of Orthodontics, Marquette University School of Dentistry,
Milwaukee, Wis.
All authors have completed and submitted the ICMJE Form for Disclosure of
Potential Conflicts of Interest, and none were reported.
The study protocol was approved by the Institutional Review Board of Marquette
University (no. HR-1811026964).
Address correspondence to: Bhoomika Ahuja, Department of Orthodontics,
Marquette University School of Dentistry, 1801 W Wisconsin Ave, Rm 201,
Milwaukee, WI 53233; e-mail, bhoomika83@gmail.com.
Submitted, November 2019; revised and accepted, April 2020.
0889-5406/$36.00
Ó 2021.
https://doi.org/10.1016/j.ajodo.2020.04.028
from two-week wear to weekly aligner switches,
decreasing treatment time by up to 50%.3 Each aligner
is to be worn for 20-22 hours a day to be effective.1,4
Several studies have evaluated the accuracy of the
Invisalign system by superimposing predicted and
achieved virtual models over unmoved posterior
teeth using 3-dimensional (3D) superimposition
software.2,5-10 Although it is possible that the teeth
superimposed on may move during treatment, more
stable landmarks (ie, palatal rugae) are not available
on Invisalign's predicted models because they only
illustrate teeth and attached gingiva. In addition, most
of these studies were conducted before the release of
Align Technology's SmartTrack (LD30; Align
Technology) material developed in 2013 and before
weekly aligner switches were recommended in 2016.
A recent systematic review concluded that Invisalign
could predictably level, tip, and derotate anterior teeth,
but not canines and premolars. The authors found that
limitations of Invisalign also include posterior arch
expansion through bodily tooth movement, closure of
extraction spaces, improvement of occlusal contacts,
extrusion of maxillary incisors, and correction of large

250
Karras et al
anteroposterior and vertical discrepancies.11 To increase
effectiveness, composite attachments are bonded to
teeth so that the aligner can be more retentive and to
facilitate tooth movement.12
The first Invisalign attachments were conventional
attachments that were either ellipsoid or rectangular in
shape. The ellipsoid shape is considered the least
effective attachment today because of its small size
and lack of a defined active surface.12 Conventional
rectangular attachment dimensions, prominence, degree
of beveling, and position on the tooth may be changed
according to clinician preference in the ClinCheck Pro
software (Align Technology) and are still widely used
today. Optimized attachments, a type of SmartForce
feature introduced in 2009, are engineered and patented
by Align Technology to create precise biomechanical
forces on teeth, thus increasing the predictability of
tooth movement.12 They vary by shape and are
automatically placed by the ClinCheck software when a
certain amount and type of planned tooth movement
is detected. Optimized rotation attachments are
automatically placed onto canines or premolars when a
rotation of $5
is detected. Maximum rotational
velocity is 2
per stage. Optimized extrusion attachments
are applied on to incisors or canines when $0.5 mm
extrusion is detected by the software. Maximum linear
velocity is 0.25 mm per stage.13
Unlike optimized attachments, conventional
attachments are not unique to Invisalign and are used
by other companies offering clear aligners or software
to create in-office aligners using 3D printers. Although
the precision of orthodontic tooth movements with
Invisalign has been studied, the effectiveness of the
different attachment types, among other aligner
variables, has not been considered. This research aimed
to compare the efficacy of optimized and conventional
attachment types on rotations of canines and premolars
and extrusion of anterior teeth—two movements
reported to be the most difficult to achieve predictably
with Invisalign.11 Results can help guide dentists in their
choice of attachment types or in considering any
overcorrection of movements when treatment planning
with Invisalign or another clear aligner software.
MATERIAL AND METHODS
This retrospective study consisted of 382 teeth from
digital dental models of 100 orthodontic patients aged
11-63 years (32 males and 68 females with a mean
age of 28 years 2 months). The sample teeth were derived
from 97 maxillary arches and 60 mandibular arches.
Some patients were used more than once because they
had a refinement scan available with qualifying teeth
for a total of 120 subjects. All patients were
American Journal of Orthodontics and Dentofacial Orthopedics
251
treated with Invisalign (Align Technology) by 1 of 2
orthodontists in private practice outside of Milwaukee,
Wis and Chicago, Ill between October 2016 and August
2018. Both orthodontists had been providing Invisalign
for at least 5 years before when the patients were started.
A power analysis indicated that a sample size of at least
64 teeth per group would be needed to have a power of
95% with a significance level (a) of 0.05. The number of
attachment types were: 163 optimized rotation (43%),
72 conventional rotation (19%), 81 optimized extrusion
(21%), and 66 conventional extrusion (17%). Aligners
were changed once a week according to the
manufacturer's and clinician's recommendations at the
time. The average number of aligners per series was
20, corresponding to an average treatment time of
5 months. Spacing was present or interproximal
reduction (IPR) performed on either side of 61 out of
the 382 teeth studied (16%). The study protocol was
approved by the Institutional Review Board of
Marquette University.
The main inclusion criteria were as follows:
(1) presence of optimized or conventional rotation or
extrusion attachments in the planned ClinCheck;
(2) completion of the initial series of aligners, resulting
in either a refinement or final scan; (3) no planned
movement of at least one posterior tooth per side of
the dental arch; (4) good compliance reported with
aligner wear; (5) full permanent dentition; and
(6) treatment beginning in 2016 or later. The exclusion
criteria were: (1) patients in the primary or mixed
dentition; (2) new dental restorations or extractions
during treatment; (3) the use of any auxiliaries, such as
elastics or vibrational devices; and (4) patients with
any orofacial syndromes or malformations.
To detect which teeth had conventional attachments
placed primarily for rotation or extrusion, the previous
unaccepted ClinChecks were reviewed to confirm that
an optimized rotation or extrusion attachment was
removed and replaced by a conventional one. Removal
and replacement of an optimized attachment would
indicate that conventional attachments were placed on
teeth with planned rotations of $5
or planned
extrusion of $0.5 mm, which are the thresholds for
optimized attachments to be placed. Predicted rotation
was divided into mild (\45
), moderate (45
-55
), or
advanced (.55
), whereas predicted extrusion was also
divided into mild (\2.5 mm), moderate (2.5-3.5 mm),
or advanced (.3.5 mm), according to Align
Technology's classifications.14
Initial, predicted, and achieved digital dental models
were exported from the ClinCheck software as
stereolithography files. The initial and final models
from the original ClinCheck were labeled as “initial”
August 2021  Vol 160  Issue 2
252
and “predicted,” respectively. The models from the
midtreatment refinement scan or the models from the
final scan at the end of treatment (whichever came first)
were labeled as “achieved.” The stereolithography files
were then imported into the 3D Slicer CMF
program (version 4.7.0; http://www.slicer.org) for
superimpositions and measurements. Fiducial markers
were placed on the central pits of posterior teeth planned
to have no movement, and a region of interest was
selected to include the entire occlusal surface, at a
minimum, to superimpose on. Gingival margins were
not included as superimposition landmarks because
the virtual gingiva in treatment simulations may be
inaccurate and misleading. Initial and predicted models
were superimposed to measure predicted tooth
movements, whereas initial and achieved models were
superimposed to measure achieved movements (Fig 1).6
Measurements were made on the teeth as follows:
(1) for rotations of canines and premolars, two
landmarks were manually placed on each tooth, the
points were automatically connected to form a straight
line, and the angle (yaw) between the two lines from
each model was calculated by the software in degrees
(
) (Fig 2). The landmarks used were usually buccal
and lingual cusp tips on premolars or a cusp tip and
cingulum on canines. If the cusp tips or cingula were
ill-defined or the points not reproducible, the most
mesial and distal points of each tooth were used; and
(2) for extrusion of incisors and canines, one point was
chosen near the center of the incisal edge or cusp tip
of each tooth, and the vertical distance between the
Fig 1. 3D model superimposition using 3D Slicer CMF: A, regions of interest on unmoved
second molars of initial (top) and predicted (bottom) maxillary arches; B, arches after they were
superimposed.
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Karras et al
two points on each model was calculated in millimeters
(mm) (Fig 3).
To account for any error in model superimposition
because of inadvertent vertical movement of
posterior teeth superimposed on, all achieved extrusive
measurements were adjusted by comparing them to a
control tooth. The control teeth were typically
directly adjacent to those being measured so that they
were roughly in the same anteroposterior position
along the dental arch. Control teeth were measured
to confirm no predicted vertical movement
(0 6 0.05 mm). If the movement was achieved even
though no movement was predicted, it was assumed
this was because of either intrusion or eruption of the
teeth superimposed on. The achieved value from a
control tooth was subtracted from the achieved value
of the adjacent tooth of interest to calculate the true
extrusion of the latter.
Statistical analysis
To calibrate the principal investigator to a uniform
measuring method, all of the measurements were
performed only after initially completing several
measurements as a practice exercise. The same examiner
repeated 40 of the rotational measurements and 40 of
the extrusive measurements by random within a
3-week interval to assess intraexaminer reliability. The
intraclass correlation coefficient was excellent, with a
score of 0.970 (95% confidence interval [CI],
0.944-0.984) for overall mean difference values. For
rotation, Cronbach's alpha was 0.965 (95% CI,
Karras et al
Fig 2. Rotational measurements.
Fig 3. Extrusive measurements.
0.914-0.986). For extrusion, intrarater reliability had a
value of 0.907 (95% CI, 0.780-0.962).
Any tooth measured to have a negative achieved
value for a vertical movement, indicating intrusion,
was changed to 0 mm because no extrusion was
achieved. This was done to avoid large negative
percentages when calculating accuracy (%
accuracy 5 100  [(|predicted  achieved|)/(|
predicted|) 3 100]). In this equation, the absolute value
of the difference between predicted and achieved
movements was taken to ensure that percent accuracy
never exceeded 100% for the teeth that achieved
movements beyond what was predicted. To account
for this same situation, the absolute value was also
taken when calculating the discrepancy between
predicted and achieved measurements in degrees and
millimeters to avoid yielding negative values that would
affect the mean without accounting for directionality.
To reduce the number of variables, similar types of
teeth were grouped, including contralateral teeth,
maxillary first and second premolars, mandibular first
and second premolars, and mandibular central and
lateral incisors. Independent t tests (two-tailed) were
used to compare mean predicted and achieved
movements between optimized and conventional
attachments. Paired t tests (two-tailed) were used to
compare mean predicted and mean achieved movements
American Journal of Orthodontics and Dentofacial Orthopedics
253
within groups. A one-way analysis of variance was used
to compare the mean accuracies of movements among
tooth types. Data analysis was performed using
Statistical Analysis Software (version 9.4; SAS, Cary,
NC) at a significance level of P \ 0.05.
RESULTS
Descriptive statistics for both rotation and extrusion
with optimized and conventional attachment types are
presented in Tables I–IV. When comparing the efficacy
of optimized and conventional attachments, the mean
differences in raw values were higher for conventional
attachments, and mean percent accuracies were higher
for optimized attachments, but this did not reach
statistical significance for both rotation and extrusion
(P .0.05) (Table V).
For all tooth movements and attachment types, mean
predicted values were significantly larger than mean
achieved values (P \ 0.0001) (Table VI). Table VII shows
mean accuracies by tooth type for both rotation and
extrusion. The mean accuracy for all movements studied
was 57.2%. The mean accuracy for rotation was 63.2%,
whereas, for extrusion, it was 47.6%. The most severe
planned movements for both rotation (74.0
) and
extrusion (4.21 mm) had an accuracy of 64%. The least
accurate tooth movement was mandibular canine
extrusion with a conventional attachment (16.1%). The
most accurate tooth movement was extrusion of the
maxillary central incisor with a conventional attachment
(73.9%), followed closely by rotation of the maxillary
premolar with an optimized attachment (72.8%).
Sixteen teeth were shown to intrude an average of
0.16 mm, so these achieved movements were changed
to 0 mm for extrusion, yielding a 0% accuracy.
A one-way analysis of variance showed that there was
no statistically significant difference between tooth
types for rotation and extrusion when considering raw
data measurements. When mean percent accuracies
were compared, there was no significant difference
between tooth types for rotation, but there was for
extrusion. A Bonferroni post-hoc test concluded that
the mean accuracies for maxillary canine extrusion
(41.7%) and mandibular canine extrusion (27.1%)
were significantly lower than that of the maxillary
central incisor (66.3%) at P \ 0.05.
When comparing teeth that had spacing or IPR to
those without, the mean accuracy for both conditions
was 57%. An independent t test showed that for all
teeth, IPR or spacing only slightly improved accuracy
by 0.2% [standard deviation, 28.9; 95% CI, 7.7 to
8.2], and this did not reach statistical significance
(P .0.05).
August 2021  Vol 160  Issue 2
254 Karras et al

Table I. Descriptive statistics for optimized rotation Table III. Descriptive statistics for optimized extrusion
attachments attachments
Standard Standard
Tooth Movement n Mean deviation Tooth Movement n Mean deviation
Maxillary Predicted (
) 38 14.26 9.97 Maxillary Predicted (mm) 11 1.45 0.77
canine central
Achieved (
) 9.71 7.37 incisor
jPredicted  achievedj (
) 4.94 6.62 Achieved (mm) 1.36 1.25
Accuracy (%) 65.9 22.9 jPredicted  achievedj 0.52 0.37
Maxillary Predicted (
) 36 12.65 12.73 (mm)
premolar Accuracy (%) 58.7 24.6
Achieved (
) 9.68 8.83 Maxillary Predicted (mm) 40 1.00 0.51
jPredicted  achievedj (
) 3.54 5.86 lateral
Accuracy (%) 72.8 23.6 incisor
Mandibular Predicted (
) 35 15.49 11.04 Achieved (mm) 0.54 0.49
canine jPredicted  achievedj 0.50 0.31
Achieved (
) 12.23 9.61 (mm)
jPredicted  achievedj (
) 3.89 4.60 Accuracy (%) 44.8 29.3
Accuracy (%) 68.0 25.9 Maxillary Predicted (mm) 19 1.01 0.85
Mandibular Predicted (
) 54 14.42 8.49 canine
premolar Achieved (mm) 0.52 0.48
Achieved (
) 9.62 7.95 jPredicted  achievedj 0.50 0.53
jPredicted  achievedj (
) 5.74 5.59 (mm)
Accuracy (%) 58.6 28.3 Accuracy (%) 46.6 35.7
Mandibular Predicted (mm) 9 1.10 0.87
incisor
Achieved (mm) 0.72 0.54
Table II. Descriptive statistics for conventional rota- jPredicted  achievedj 0.38 0.42
tion attachments (mm)
Accuracy (%) 64.8 24.3
Standard Mandibular Predicted (mm) 2 0.59 0.37
Tooth Movement n Mean deviation canine
Maxillary Predicted (
) 17 11.18 7.29 Achieved (mm) 0.23 0.12
canine jPredicted  achievedj 0.37 0.49
Achieved (
) 7.11 6.61 (mm)
jPredicted  achievedj (
) 4.45 4.59 Accuracy (%) 54.5 53.8
Accuracy (%) 57.9 29.8
Maxillary Predicted (
) 10 11.94 8.79
premolar specifically optimized attachments available to compare
Achieved (
) 5.08 2.83 to conventional ones. No published study to date has
jPredicted  achievedj (
) 6.86 7.73 compared the efficacy of Invisalign's two attachment
Accuracy (%) 48.1 23.4
types for any tooth movement.
Mandibular Predicted (
) 19 15.78 8.53
canine Because of the strict inclusion and exclusion criteria,
Achieved (
) 9.50 7.71 the majority of the patients included in this study were
jPredicted  achievedj (
) 6.84 6.39 Class I malocclusions, in which the clinicians had a
Accuracy (%) 60.5 25.0 preference of using optimized or conventional
Mandibular Predicted (
) 26 14.41 8.93
attachments to help resolve rotations or extrude teeth.
premolar
Achieved (
) 7.66 4.80 Overall, the study sample was representative of the
jPredicted  achievedj (
) 6.84 8.23 general, orthodontic population, as 99% of rotations
Accuracy (%) 58.6 28.3 and 95% of extrusion measured were considered to be
mild (\45
and \2.5 mm, respectively), according to
Align Technology's classifications.
DISCUSSION A clinically discernible amount of malrotation was
The current study focused on intra-arch considered to be 15
on the basis of a previous study
measurements of two tooth movements reported to be by Kravitz et al2 that also assessed the accuracy of the
the least accurate with Invisalign—rotation of canines Invisalign system. For extrusion, a 0.2 mm discrepancy
and premolars and extrusion of incisors and canines. was chosen to be clinically significant because that is
These movements were also chosen because they have the limit of resolution of the human eye,15 and because

August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Karras et al 255

Table IV. Descriptive statistics for conventional extru- Table VI. Comparison of mean predicted and achieved
sion attachments tooth movements
Standard Mean difference
Tooth Movement n Mean deviation Attachment Standard
Maxillary Predicted (mm) 11 1.37 1.21 type n (Predicted  achieved) deviation P value*
central Optimized 163 4.01 6.22 \0.0001
incisor rotation (
)
Achieved (mm) 0.94 0.73 Conventional 72 6.01 7.14 \0.0001
jPredicted  achievedj 0.44 0.53 rotation (
)
(mm) Optimized 81 0.40 0.47 \0.0001
Accuracy (%) 73.9 18.4 extrusion
Maxillary Predicted (mm) 30 1.03 0.58 (mm)
lateral Conventional 66 0.62 0.45 \0.0001
incisor extrusion
Achieved (mm) 0.51 0.37 (mm)
jPredicted  achievedj 0.52 0.37
*Paired t test; statistical significance at P \ 0.05.
(mm)
Accuracy (%) 48.3 23.7
Maxillary Predicted (mm) 13 1.23 0.75
canine
Table VII. Mean accuracy of tooth movements
Achieved (mm) 0.45 0.43 Rotation Extrusion
jPredicted  achievedj 0.78 0.57
(mm) Mean Mean
Accuracy (%) 34.5 28.0 accuracy Standard accuracy Standard
Mandibular Predicted (mm) 7 1.46 0.67 Tooth n (%) deviation n (%) deviation
incisor Maxillary 22 66.3 22.6
Achieved (mm) 0.55 0.66 central
jPredicted  achievedj 0.92 0.29 incisor
(mm) Maxillary 70 46.3 26.9
Accuracy (%) 27.7 33.3 lateral
Mandibular Predicted (mm) 5 1.00 0.47 incisor
canine Maxillary 55 63.4 25.2 32 41.7 32.9
Achieved (mm) 0.22 0.28 canine
jPredicted  achievedj 0.77 0.30 Maxillary 46 67.4 25.5
(mm) premolar
Accuracy (%) 16.1 18.8 Mandibular 16 48.5 33.5
incisor
Mandibular 54 65.4 25.6 7 27.1 32.7
canine
Table V. Comparison of optimized and conventional Mandibular 80 59.3 28.7
attachments premolar

Absolute mean
difference
Tooth (jConventional  Standard P 1.61
or 7.3%. For extrusion, the mean difference was
movement n optimizedj) deviation value
0.14 mm or 4.3%. Though optimized attachments had
Rotation 235 1.61
6.11 0.0638
Extrusion 147 0.14 mm 0.42 0.0523
a higher mean accuracy than conventional attachments
for both movements, these differences were neither
Mean accuracy clinically nor statistically significant. This may be
difference (%)
because most of the conventional rotation attachments
Tooth (Optimized  Standard
movement n conventional) deviation P value used were rectangular and 3 mm long, which are
Rotation 235 7.3 26.4 0.0533 typically larger than optimized attachments. A larger
Extrusion 147 4.3 29.8 0.3819 attachment provides a greater surface area for the
aligner to push on, thus improving efficacy. In addition,
many clinicians choose to bevel conventional horizontal
orthodontists are trained to focus on anterior attachments gingivally, resembling the design of
microesthetics during the finishing stages of treatment. optimized extrusion attachments. This configuration
When comparing optimized and conventional provides a surface perpendicular to the force vector
attachments, the mean difference for rotation was needed for the extrusion of anterior teeth. Optimized

American Journal of Orthodontics and Dentofacial Orthopedics August 2021  Vol 160  Issue 2
256
extrusion attachments may still be more effective
because of the intentional gap left between the
attachment and aligner and between the tooth and
aligner on the incisal surface. This space is meant to
allow clearance for the tooth to extrude unimpeded.
In general, it was found that for all tooth movements,
predicted values were higher than achieved values, and
these results were statistically significant. For rotation,
the mean difference between predicted and achieved
values ranged from 4.01
to 6.01
for optimized and
conventional attachments, respectively. Although these
discrepancies were not clinically significant,
practitioners may still elect to overcorrect canine and
premolar rotations by 5 6 1
on the basis of the findings
of this study. For extrusion, the mean difference ranged
from 0.40 mm to 0.62 mm for optimized and
conventional attachments, respectively. Because these
differences are to be considered clinically significant,
clinicians should not only plan to overcorrect anterior
extrusion, but they should also be mindful about how
the prescribed tooth movement is staged. This is because
it is likely that some extrusion or intrusion observed with
clear aligners is “relative” and because of retroclination
and proclination, respectively. In fact, anterior extrusion
may be more predictable if the teeth are initially
proclined labially and then retracted through space
closure, rather than if overcorrected by 0.5 6 0.1 mm
(the discrepancy found in this study).
A study by Kravitz et al2 found that the least accurate
tooth movement with Invisalign was extrusion of
incisors (29.6%) and that, for rotation, the least accurate
tooth was the canine (35.8%). The same study also
reported that for rotations greater than 15
, accuracy
significantly fell by up to 52.5%. According to the
literature, derotation of a cylindrical tooth is difficult
because aligners tend to lose anchorage and slip off
because of a lack of undercuts and the round tooth
shape.10 It would make sense that well-designed
attachments would provide more retention and an active
surface area for forces to be applied to, as long as the
aligners fit well. Even though Kravitz et al7 found that
the presence of attachments did not significantly
improve the accuracy of canine rotation, the most
common attachment in that study was the vertical
ellipsoid, which is rarely used today.
A systematic review reported 29.1%-49.7% accuracy
for canine and premolar derotation.11 The current study
found a mean accuracy of 63.2% for these rotational
movements and 47.6% for extrusion of anterior teeth.
These findings are higher than previous studies show.
Differences may be due to several factors, including
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Karras et al
the introduction of SmartTrack material in 2013, smaller
sample sizes in previous studies, variable patient
compliance, frequency of aligner switches, and the
presence or absence of attachments. The overall mean
accuracy of tooth movements in this study was found
to be 57.2%, but it must be noted that only rotation
and extrusion were analyzed in patients without any
treatment adjuncts, such as interarch elastics. Therefore,
these results should not be generalized for all patients
treated with the Invisalign system. However, this study
does provide useful information on the accuracy of these
two difficult tooth movements and the efficacy of
attachment types used over a series of aligners.
A recent study by Charalampakis et al6 found that
intrusion was the least accurate of all linear movements
and that extrusion of incisors appeared to be accurate.
This disagrees with previous findings and may be
because of unplanned intrusion of posterior teeth
superimposed on. The intrusion of posterior teeth
between 0.25 mm to 0.5 mm is often observed with
Invisalign because of plastic thickness.1 Because each
aligner is 0.75 mm thick occlusally,16 the appliance has
a posterior bite-block effect, which would cause the
incisors to appear to extrude more than planned.6 In
addition, because only adult patients were used in the
previous study, the eruption of posterior teeth was likely
not encountered, so anterior intrusion would have
appeared to be less accurate than extrusion. To
overcome this limitation, control teeth were used in
the current study to measure true achieved extrusion.
These control teeth happened to extrude or intrude
between 0 mm and 0.5 mm, likely because of
second molar intrusion or eruption, respectively.
In this study, IPR and spacing were grouped because
both conditions would, in theory, reduce friction and
collisions between teeth during movement. Kravitz
et al7 reported that IPR improved the accuracy of canine
rotations but that this was not statistically significant.
The findings from the current study also found that
there was no significant improvement in accuracy
when IPR was used or when spacing was present, with
both the presence and absence of these conditions
reaching 57% accuracy. However, a small sample size
of teeth had IPR or spacing (16%). In addition, IPR is
most commonly performed on mandibular anterior teeth
to help resolve crowding, and anterior rotations were not
evaluated in this study.
The study sample included 37 patients that were
aged \18 years. Of these, 16 patients were aged
between 11 and 14 years who were growing and may
have had second molars erupting. Although control
Karras et al
teeth were used, both of these patient factors may have
affected digital model superimposition on posterior
teeth, as every case used for this study was superimposed
on second molars, at a minimum. Although control teeth
were thought to have moved in the vertical dimension
because of posterior intrusion or eruption, they may
also have moved if they had an attachment on them
for planned movement in a different plane of space or
simply for support of the aligner.
Other limitations of this retrospective study included
the inability to account for certain variables, such as
periodontal support, conventional attachment features,
simultaneous tooth movements, and patient compliance.
Though most of the patients used initial scans, this study
also used refinement scans of some patients that already
had prior tooth movement. Any existing mobility and
altered periodontal support may have influenced results,
along with general periodontal status and bone density
ranging from patient to patient. Furthermore, the
location, size, orientation, and beveling of conventional
attachments were not considered. Although having an
optimized rotation and extrusion attachment applied
indicates that those were the primary movements for a
specific tooth, it is possible that other minor simultaneous
movements were occurring in different planes of space,
compromising the predictability of the studied
movements. As mentioned previously, some of the
extrusion observed may have been due to retraction or
retroclination rather than pure vertical movement. To
overcome this, clinicians and researchers may consider
planning more than one movement in separate stages
for individual teeth. Pure movement may be possible
with clear aligners because the plastic can act as a
boundary for any other concurrent movements, unlike
with fixed appliances. However, this is still difficult to
achieve because most clear aligner systems, including
Invisalign, do not require the use of radiographs, and
thus, do not take into account a tooth's root
length, angulation, and center of resistance when
planning movements. Relying solely on the digital crowns
of teeth can reduce the accuracy of tooth movement,
leading to “non-tracking” and even unwanted
movements.
Future studies evaluating tooth movement should be
prospective and consider using 2-dimensional lateral
cephalograms or 3D cone-beam computed tomography
imaging to superimpose on stable landmarks rather than
posterior teeth, which may inadvertently move during
orthodontic treatment. They should only include
patients that were evaluated from the start of treatment
to the end to assess final results. In addition, other
attachment variables previously mentioned should be
considered, and additional types of tooth movement
American Journal of Orthodontics and Dentofacial Orthopedics
257
assessed to further compare the efficacy of conventional
and optimized attachments.
CONCLUSIONS
1. Differences between mean accuracies of tooth
movements using Invisalign's optimized vs
conventional attachments were neither statistically
nor clinically significant. This applies to the
derotation of canines and premolars and extrusion
of incisors and canines, specifically.
2. There was a statistically significant difference
between the amount of tooth movement that was
predicted and the amount that was achieved. For
extrusion, this was also clinically significant
(.0.2 mm), but for rotation, it was not (\15
).
3. Derotation of canines and premolars was
accomplished with a 63.2% mean accuracy, and
extrusion of incisors and canines was achieved
with a mean accuracy of 47.6%. Overall, the mean
accuracy of both these movements was 57.2%.
4. IPR or spacing did not significantly affect accuracy
for rotations of canines and premolars and for
extrusion of anterior teeth. A larger sample size of
teeth with this treatment or condition is needed
for more definitive conclusions.
With more companies offering clear aligners and with
the emergence of in-house 3D printing, one can feel
confident in knowing that conventional attachment
types may be just as effective as Invisalign's proprietary
optimized attachments for rotations of canines and
premolars and for extrusion of incisors and canines.
However, to improve the predictability of anterior
extrusion, clinicians may aim to achieve this movement
primarily through retraction and space closure with
gingivally beveled attachments and should plan for
overcorrection of up to 0.5 6 0.1 mm. Clinicians may
also consider overcorrecting rotations by 5 6 1
to
improve accuracy with the Invisalign system. Even with
planned overcorrection, patients should always be aware
of the possibility of needing refinement aligners to
achieve clinically acceptable results.
ACKNOWLEDGMENTS
The authors would like to thank the orthodontists,
who wish to remain anonymous, for providing the
patient sample for this study.
AUTHOR CREDIT STATEMENT
Theresa Karras contributed to conceptualization,
data curation, investigation, methodology, original draft
preparation, and draft review and editing; Maharaj
August 2021  Vol 160  Issue 2
258
Singh contributed to formal analysis and resources;
Emelia Karkazis contributed to data curation; Dawei
Liu and Ghada Nimeri contributed to study validation;
Bhoomika Ahuja contributed to the methodology,
project administration, supervision, and draft review
and editing.
REFERENCES
1. Phan X, Ling PH. Clinical limitations of Invisalign. J Can Dent
Assoc 2007;73:263-6.
2. Kravitz ND, Kusnoto B, BeGole E, Obrez A, Agran B. How well does
Invisalign work? A prospective clinical study evaluating the
efficacy of tooth movement with Invisalign. Am J Orthod
Dentofacial Orthop 2009;135:27-35.
3. Align Technology. Inc. Frequently asked questions. Available at:
https://www.invisalign.com/frequently-asked-questions. Accessed
June 18, 2021.
4. Align Technology. Inc. Living with Invisalign clear aligners.
Available at: www.invisalign.com/how-invisalign-works/living-with-
invisalign. Accessed June 16, 2019.
5. Buschang PH, Ross M, Shaw SG, Crosby D, Campbell PM. Predicted
and actual end-of-treatment occlusion produced with aligner
therapy. Angle Orthod 2015;85:723-7.
6. Charalampakis O, Iliadi A, Ueno H, Oliver DR, Kim KB. Accuracy of
clear aligners: a retrospective study of patients who needed
refinement. Am J Orthod Dentofacial Orthop 2018;154:47-54.
7. Kravitz ND, Kusnoto B, Agran B, Viana G. Influence of attachments
and interproximal reduction on the accuracy of canine rotation
with Invisalign. A prospective clinical study. Angle Orthod 2008;
78:682-7.
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Karras et al
8. Krieger E, Seiferth J, Saric I, Jung BA, Wehrbein H. Accuracy of
InvisalignÒ treatments in the anterior tooth region. First results.
J Orofac Orthop 2011;72:141-9.
9. Krieger E, Seiferth J, Marinello I, Jung BA, Wriedt S, Jacobs C,
et al. InvisalignÒ treatment in the anterior region: were the
predicted tooth movements achieved? J Orofac Orthop 2012;
73:365-76.
10. Simon M, Keilig L, Schwarze J, Jung BA, Bourauel C. Treatment
outcome and efficacy of an aligner technique–regarding incisor
torque, premolar derotation and molar distalization. BMC Oral
Health 2014;14:68.
11. Papadimitriou A, Mousoulea S, Gkantidis N, Kloukos D. Clinical
effectiveness of InvisalignÒ orthodontic treatment: a systematic
review. Prog Orthod 2018;19:37.
12. Paquette DE, Colville C, Wheeler T. Clear aligner treatment. In:
Graber L, Vanarsdall R, Vig K, Huang G, editors. Orthodontics:
Current Principles and Techniques. 6th ed. St Louis: Elsevier;
2017. p. 778-811.
13. Align Technology. Inc. SmartForce features and attachments.
Available at: https://s3.amazonaws.com/learn-invisalign/docs/
SmartForce%20Features%20and%20Attachments_en-gb-en.pdf.
Accessed June 16, 2019.
14. Align Technology. Inc. Invisalign tooth movement assessment
overview. Available at: https://s3.amazonaws.com/learn-inv
isalign/docs/us/ToothAssessment.pdf. Accessed June 16, 2019.
15. Bille JF, B€
uchler-Costa J, M€ uller F. Optical quality of the human eye:
the quest for perfect vision. In: Bille JF, Harner CFH, Loesel FF, editors.
Aberration-Free Refractive Surgery: New Frontiers in Vision.
Heidelberg, Germany: Springer; 2004. p. 61-82.
16. Gao L, Wichelhaus A. Forces and moments delivered by the PET-G
aligner to a maxillary central incisor for palatal tipping and
intrusion. Angle Orthod 2017;87:534-41.
ORIGINAL ARTICLE

Biomechanical analysis for total

distalization of the maxillary dentition: A
finite element study
Jun Kawamura,a Jae Hyun Park,b Yukio Kojima,c Naohiko Tamaya,d Yoon-Ah Kook,e Hee-Moon Kyung,f
and Jong-Moon Chaeg
Gifu, Nagoya, and Fukui, Japan, and Mesa, Ariz, and Seoul, Daegu, and Iksan, South Korea

Introduction: This study aimed to identify the tooth movement patterns relative to various force angulations
(FAs) when distalizing the total maxillary dentition. Methods: Long-term orthodontic movement of the maxillary
dentition was simulated by accumulating the initial displacement of teeth produced by elastic deflection of the
periodontal ligament using a finite element analysis. Distalization forces of 3 N were applied to the archwire be-
tween the maxillary canine and first premolar at 5 different FAs ( 30 , 15 , 0 , 15 , and 30 ) to the occlusal
plane. Results: Maxillary incisors and molars showed lingual and distal tipping at all FAs, respectively. At a force
angulation of 30 , almost bodily distalization of the total maxillary dentition occurred, but incisors showed consid-
erable lingual tipping because of the effect of clearance gap (0.003-in, 0.022 3 0.025-in bracket slot,
0.019 3 0.025-in archwire) and elastic deflection of the archwire. Medial displacement of the maxillary
anterior teeth occurred because of lingual tipping during distalization. The occlusal plane rotated clockwise at
all FAs because of extrusion of the maxillary incisors and intrusion of the maxillary second molars, and the
amounts decreased as FA increased. Conclusions: Tooth movement patterns during distalization of the total
maxillary dentition were recognized. With an understanding of the mechanics, a proper treatment plan can be
established. (Am J Orthod Dentofacial Orthop 2021;160:259-65)

C
lass II malocclusion correction can be accom- prudent differential diagnosis considering the face, skel-
plished with extraction or nonextraction treat- etal pattern, and dentition.1,2
ment. The decision should be made by a For nonextraction treatment, headgear3 and Class II
elastics4,5 have been traditionally used for distalization
of total maxillary dentition. However, they included un-
a
b
Private practice, Gifu, Japan. desirable side effects and were highly dependent on pa-
Postgraduate Orthodontic Program, Arizona School of Dentistry & Oral Health,
tient compliance, therefore alternative treatment
A. T. Still University, Mesa, Ariz, Graduate School of Dentistry, Kyung Hee Uni-
versity, Seoul, South Korea. modalities such as pendulum6 and distal jet7 were intro-
c
Private practice, Nagoya, Japan. duced because they required only minimal patient coop-
d
Private practice, Fukui, Japan.
e eration, but unfortunately, some reciprocal anchorage
Department of Orthodontics, Seoul St. Mary's Hospital, Catholic University of
Korea, Seoul, South Korea. loss occurred during distalization with these methods.
f
Department of Orthodontics, School of Dentistry, Kyungpook National Univer- A nonextraction approach has become more popular
sity, Daegu, South Korea.
g for Class II treatment by total distalization of the maxil-
Department of Orthodontics, School of Dentistry, Wonkwang University, Wonk-
wang Dental Research Institute, Iksan, South Korea, Postgraduate Orthodontic lary dentition with the advent of temporary skeletal
Program, Arizona School of Dentistry & Oral Health, A. T. Still University, anchorage devices. For distalization of the maxillary
Mesa, Ariz.
dentition, an indirect skeletal anchorage-supported
All authors have completed and submitted the ICMJE Form for Disclosure of Po-
tential Conflicts of Interest, and none were reported. pendulum8 and distal jet9 were introduced to prevent
This paper was supported by Wonkwang University in 2021. reciprocal anchorage loss.
Address correspondence to: Jong-Moon Chae, Department of Orthodontics,
Miniscrews and miniplates have been placed in
School of Dentistry, Wonkwang University, Daejeon Dental Hospital, 77
Doonsan-ro, Seo-Gu, Daejeon 35233, South Korea; e-mail, jongmoon@wku. various sites and are used widely as direct skeletal
ac.kr. anchorage devices for total distalization of the maxillary
Submitted, February 2020; revised, March 2020; accepted, April 2020.
dentition with various force angulations (FAs) from the
0889-5406/$36.00
Ó 2021 by the American Association of Orthodontists. All rights reserved. posterosuperior to the posteroinferior direction to the
https://doi.org/10.1016/j.ajodo.2020.04.029

259
260 Kawamura et al

occlusal plane in pursuit of desired tooth move-

ment.10-19
Many studies have been conducted to correct Class II
malocclusions with various treatment modalities that
use conventional and skeletal anchorage devices,3-19
but few studies have attempted to predict the tooth
movement patterns from a biomechanical point of
view. Therefore, this study aimed to recognize tooth
movement patterns and identify the biomechanics
relative to the FAs during total distalization of the
maxillary dentition.

MATERIAL AND METHODS

Figure 1 shows the finite element model. The same
model was used as in a previous study20 for total mesi-
alization of the maxillary dentition. Only the right side
of the dentition was modeled because of its bilateral
symmetry. Finite element models of teeth were con-
structed on the basis of a dental study model (i21D-
400C; Nissin Dental Products, Kyoto, Japan). Surfaces
of the teeth were divided into shell elements and were
considered to be rigid. The periodontal ligament (PDL)
was assumed to be a linear elastic material of 0.2 mm
thickness. Young's modulus and Poisson's ratio were
0.2 MPa and 0.4, respectively. The alveolar bone was
assumed to be a rigid body and was thereby excluded
from the finite element model.
The archwire was made of 0.019 3 0.025-in stainless
steel (SS) wire, whose Young's modulus and Poisson ra-
tio were 200 GPa and 0.3, respectively. The size of the
bracket slot was 0.022 3 0.025-in, the width was
4 mm for the molars and 3 mm for the other teeth.
The bracket was assumed to be rigid. A force of 3 N
was applied to the archwire between the canine and first
premolar brackets with various FAs in the sagittal plane.
The archwire slid along the bracket slots with a frictional
coefficient of 0.15. Surfaces of the archwire and the
bracket slot were overlaid with contact elements (Fig 1).
There was a clearance gap or play (0.003-in) between
the archwire and the bracket slot in the initial position
(Fig 2, A), and the bracket rotated 7.2 in the contact po-
sition (Fig 2, B). Figure 3 shows the location of the center
of resistance, which was determined in a previous
study.20
Long-term movements of the dentition were simu-
lated by the finite element method (FEM) in the same
way as in the previous study.20-23 The simulation
method consisted of the following 2 steps. First, the
initial displacement of each tooth produced by elastic
deflection of the PDL was calculated. Second, the
alveolar socket was moved in the same direction and
the same amount as the initial displacement.
Fig 1. Finite element model for simulating distalization of
the whole maxillary dentition.
Fig 2. Initial (A) and contact (B) positions of the archwire
put into the bracket. The archwire rotates by 7.2 in the
bracket slot.
Repeating these steps, the teeth moved with their
alveolar sockets. The number of repeated calculations
was equivalent to the elapsed time after the force
application. FEM software, ANSYS (version 19.1;
ANSYS, Inc, Canonsburg, Pa), was used for the
simulation.
Three-dimensional displacements were measured at
the central incisal edge (CIE), the mesiobuccal cusp of
the first molar (MBC6), and the distobuccal cusp of the
second molar (DBC7). The change of the occlusal plane
angle was calculated using vertical displacements of
these points.
RESULTS
Figure 4 shows the movement patterns of the maxil-
lary dentition at the number of repeated calculations
(N 5 400) for each force angulation. The initial tooth
positions are illustrated with a pale color. Color contours
on the root indicate distributions of mean stress in the
PDL. Tipping angles of the central incisor and the

August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Kawamura et al
Fig 3. The location of the center of resistance of the
whole dentition as determined using a rigid archwire. At
a force angulation of 30 , the line of action of the force
passes through the center of resistance.
second molar are described in Figure 4. Table I shows the
3-dimensional displacement of CIE, MBC6, and DBC7,
whereas Table II shows their tipping angles and occlusal
plane angles relative to each force angulation.
CIE showed distalization and extrusion at all FAs,
with amounts that decreased as FA increased. MBC6
showed distalization at all FAs and extrusion at FAs of
30 , 15 , and 0 with amounts that decreased as
FA increased, whereas MBC6 showed intrusion at FAs
of 15 and 30 with amounts that increased as FA
increased. DBC7 showed distalization at all FAs with
amounts that decreased as FA increased. DBC7 showed
intrusion at all FAs with amounts that were similar but
that were least at an FA of 30 . CIE showed medial
movement at all FAs with amounts that decreased as
FA increased. MBC6 and DBC7 showed buccal move-
ments at all FAs with amounts that increased as FA
increased except for MBC6 at FA of 30 and 15 ,
which showed lingual movements with amounts that
decreased as FA increased (Fig 4; Table I).
Incisors showed lingual tipping at all FAs with
amounts that decreased as FA increased. The first molar
showed lingual tipping at FAs of 30 and 15 with
amounts that decreased as FA increased, whereas the
first molar showed buccal tipping at FAs of 0 , 15 ,
and 30 with amounts that increased as FA increased.
The second molar showed buccal tipping at all FAs
with amounts that increased as FA increased. Both mo-
lars showed distal tipping at all FAs with amounts that
decreased as FA increased, and the amounts were greater
with the first molar than with the second molar. The
occlusal plane showed clockwise rotation patterns at
all FAs with amounts that decreased as FA increased.
(Fig 4; Table II).
Total maxillary dentition rotated clockwise with
clockwise rotation of the occlusal plane by extrusion of
American Journal of Orthodontics and Dentofacial Orthopedics
261
the anterior teeth and intrusion of the posterior teeth
at all FAs. The amount of lingual tipping of the incisors
was greater than the distal tipping of the molars.
Because of these tipping movements, compressive and
tensile stresses were produced near the cervix and
apex, respectively (Fig 4; Table II).
DISCUSSION
Using FEM, we simulated and investigated the tooth
movement patterns as reported in previous studies for
distalization and mesialization of the whole mandibular
dentition,22,23 and mesialization of the whole maxillary
dentition.20 In those reports, we discussed some of the
general limitations of a FEM study which were the
same in this one.20,22,23 This time, we examined the me-
chanics for distalization of the whole maxillary denti-
tion. As described in previous studies,20,22,23 tooth
movement patterns were determined by the rotation of
the whole dentition, clearance gap or play between the
archwire and bracket slot, and elastic deflection of the
archwire.
Rotation of the whole dentition with occlusal plane
change differed depending on the direction of force.
Whole dentition almost translated with a negligible rota-
tion of the occlusal plane at a force angulation of 30 ,
because the line of action of the force passed nearly
through the center of resistance of the whole maxillary
dentition. In addition, the whole dentition showed
more clockwise rotation as FA decreased, that is because
the line of action of the force passed farther away from
the center of resistance (Fig 4; Table II).
Clockwise rotation of the whole dentition caused
lingual tipping and extrusion of the central incisor and
distal tipping and intrusion of the second molars,
whereas their amount was greater in the central incisor
than in the second molar. In all FAs, the difference be-
tween both tipping angles was about 12 which was
caused by elastic deflection and clearance gap in the
anterior area of the archwire, whereas the posterior
area showed no elastic deflection and less clearance
gap, which caused less mesiodistal tipping of the poste-
rior teeth. The central incisor tipped lingually about 7
and 5 because of the clearance gap and elastic deflec-
tion of the archwire, respectively (Fig 2, B). If an archwire
with a low Young's modulus or smaller size is used,
tipping because of elastic deflection will increase. The
compensating curve bend incorporated into an archwire
can be effective in preventing the lingual tipping of the
incisor because of elastic deflection.
In clinical situations, the amount of tipping because
of the clearance gap is not definite because it varies de-
pending on the initial position of the archwire with
August 2021  Vol 160  Issue 2
262 Kawamura et al

Fig 4. Tooth movement patterns (N 5 400) for 5 FAs to the occlusal plane: A, 30 ; B, 15 ; C, 0 ; D,
15 ; E, 30 .

respect to the bracket slot. If the archwire is initially in
contact with the bracket slot, as shown in Figure 2, B,
tipping because of the clearance gap will not occur.
That is, the pretorque of the archwire can effectively pre-
vent tipping because of the clearance gap.
Lingual displacement of the incisal edge was mainly
because of the rotation of the whole dentition and clear-
ance gap between the archwire and bracket slot. This
movement may work as a mechanism for Class II maloc-
clusion correction when retracting the whole maxillary
dentition. However, in some patients, this mechanism
could worsen the facial profile, although the dental oc-
clusion may be improved. Therefore, a prudent differen-
tial diagnosis considering the face, skeletal pattern, and
dentition should be prepared for a proper treatment
plan.
Medial movement of CIE occurred during distaliza-
tion because of the interproximal spaces created during
modeling of the teeth. This means that if a diastema was
present, it might be closed during distalization (Tables I

August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Kawamura et al 263

Table I. Three-dimensional displacements of the maxillary teeth

Force angulation

Displacement 0.019 3 0.025-in stainless steel archwire


Measurement Point –30 –15 0 15 30
x (mm) CIE 4.84 4.87 4.70 4.29 3.72
MBC6 2.13 1.86 1.51 1.09 0.59
DBC7 1.51 1.36 1.11 0.77 0.35
y (mm) CIE 0.35 0.35 0.33 0.29 0.25
MBC6 0.51 0.21 0.06 0.25 0.42
DBC7 0.29 0.54 0.73 0.84 0.92
z (mm) CIE 3.01 2.49 1.87 1.14 0.30
MBC6 0.69 0.42 0.12 0.19 0.52
DBC7 0.84 0.90 0.89 0.80 0.65

Note. Negative values of force angulation mean that line of action of the force is in an anteroinferior direction along the occlusal plane and positive
values mean the opposite. Three-dimensional axes: x axis, posterior (1) and anterior ( ); y axis in the anterior teeth, lateral (1) and medial ( ); y
axis in the posterior teeth, buccal (1) and lingual ( ); z axis, superior (1) and inferior ( ).

Table II. Three-dimensional rotation of the maxillary teeth

Force angulation

0.019 3 0.025-in stainless steel archwire

Tooth Wire angulation –30 –15 0 15 30

Incisor LLA ( ) 19.53 18.95 17.64 15.33 12.18
First molar BLA ( ) 1.65 0.51 0.57 1.35 2.10
MDA ( ) 8.05 6.75 5.17 3.32 1.16
Second molar BLA ( ) 1.77 3.01 3.90 4.30 4.51
MDA ( ) 6.86 5.91 4.58 2.95 1.01
Occlusal plane angle ( ) 5.50 4.87 3.98 2.80 1.37
Note. Negative values of force angulation mean that line of action of the force is in an anteroinferior direction along the occlusal plane and positive
values mean the opposite. Angulation: LLA, labial (1) and lingual ( ); BLA, buccal (1) and lingual ( ); MDA, mesial (1) and distal ( ); occlusal
plane angle, clockwise (1) and counterclockwise ( ).
LLA, labiolingual angulation; BLA, buccolingual angulation; MDA, mesiodistal angulation.

and II). There was almost no buccolingual tipping of the
second molar at all FAs compared with previous studies
for total mesialization of the dentition20,22 because of
force application to the archwire between the canine
and first premolar. With total mesialization, a force
was applied directly to the second molar, and thereby
buccolingual tipping of the second molar increased,
and so its tipping movement was prevented by the use
of an archwire with a compensating bend or a transpa-
latal arch.20,22 However, with the total distalization of
the dentition, such methods are not needed to prevent
buccolingual tipping of the molar.
Total distalization of the maxillary dentition without
extraction of the maxillary premolars for Class II maloc-
clusion correction has become commonly available with
the development of temporary skeletal anchorage de-
vices. With the use of miniscrews and miniplates placed
in various sites, intermaxillary and intramaxillary elastic
forces were applied to distalize the whole dentition in a
predictable direction by considering the biomechanical
aspect.
Intermaxillary Class II elastomeric force using dental
anchorage is still used for Class II malocclusion correc-
tion by distalization of the whole maxillary dentition fol-
lowed by intentional or unintentional steepening of the
occlusal plane.24 FAs in these applications are 30 to
15 . These mechanics resulted in lingual tipping, ret-
rusion, extrusion of the maxillary incisor, and undesir-
able soft-tissue effects. Therefore, a counterclockwise
directional system using a high pull J-hook or skeletal
anchorage has been used to prevent the undesirable
side effects.25,26
Intramaxillary elastomeric forces using skeletal
anchorage devices such as miniscrews and miniplates
have been widely used for Class II malocclusion correc-
tion by distalization of the whole maxillary dentition.
FAs in these applications are 0 to 30 . With these me-
chanics, extrusion and lingual tipping of the maxillary

American Journal of Orthodontics and Dentofacial Orthopedics August 2021  Vol 160  Issue 2
264
incisors were reduced as FA increased, but serious lingual
tipping occurred at all FAs. To counteract the clockwise
rotation or lingual tipping of the maxillary anterior
dentition, a compensating curve14 and pretorque27 can
be added. But this counteracting mechanics can aggra-
vate the posterior open bite by more intrusion of the
maxillary posterior teeth, which necessitates up and
down elastics on the posterior teeth.28
In this study, we used 0.019 3 0.025-in SS wire,
300 g of force (unilateral), and 5 FAs ( 30 to 30 )
which were similar to clinical conditions. In previous
studies,10-19 0.019 3 0.025-in or 0.016 3 0.022-in SS
wires, 200 to 400 g of force, and various FAs were
used. The mechanics using one force application point
with various FAs featured in this study is limited in ob-
taining all desired tooth movement patterns, and
thereby various force application points, and variable
lever arm lengths in a buccal and palatal aspect have
been used to overcome these limitations.16,18,29 In our
study, the force application point was set on the archwire
between the canine and first premolar brackets, which is
a common clinical site for total distalization of the
maxillary dentition. The tooth movement pattern de-
pends on the force application point, so it also should
be considered to obtain the desired tooth movement
pattern.29
Interradicular screws have anatomic limitations and
are unable to distalize the maxillary dentition more
than 3 mm,10 so an extraalveolar screw17,30 or mini-
plate19 and palatal plate16,18 would be more favorable
for distalization. The amount of available posterior space
in the maxillary tuberosity area should be examined
before distalization.31 Even when posterior space is
insufficient for distalization, Class II malocclusion can
be corrected by interproximal reduction with distaliza-
tion of the whole dentition without premolar
extraction.32
CONCLUSIONS
Three-dimensional finite element simulation was
used to recognize the tooth movement patterns during
total distalization of the maxillary dentition. Followings
are the conclusions:
 At a force angulation of 30 , almost bodily distaliza-
tion of the total maxillary dentition occurred, but the
maxillary incisors showed considerable lingual
tipping because of the effect of clearance gap and
elastic deflection of the archwire.
 The occlusal plane rotated clockwise at all FAs
because of extrusion of the maxillary incisors and
intrusion of the maxillary second molars at all FAs.
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Kawamura et al
 Medial displacement of the maxillary anterior teeth
occurred during distalization.
 By understanding the mechanics during total distali-
zation of the maxillary dentition, proper treatment
planning can be established for Class II malocclusion
correction.
AUTHOR CREDIT STATEMENT
Jun Kawamura contributed to reviewing the litera-
ture and writing the article; Jae Hyun Park contributed
to critical revision of the article; Yukio Kojima contrib-
uted to data analysis and interpretation; Naohiko Tam-
aya contributed to reviewing the literature and
introduction; Yoon-Ah Kook contributed to reviewing
the literature and revision of the article; Hee-Moon
Kyung contributed to final approval of the article; and
Jong-Moon Chae contributed to supervising the overall
project and overall responsibility.
REFERENCES
1. Merrifield LL, Klontz HA, Vaden JL. Differential diagnostic analysis
system. Am J Orthod Dentofacial Orthop 1994;106:641-8.
2. Vaden JL, Williams RA, Goforth RL. Class II correction: extraction
or nonextraction? Am J Orthod Dentofacial Orthop 2018;154:
860-76.
3. Henriques FP, Janson G, Henriques JF, Pupulim DC. Effects of cer-
vical headgear appliance: a systematic review. Dental Press J Or-
thod 2015;20:76-81.
4. Aras I, Pasaoglu A. Class II subdivision treatment with the Forsus
Fatigue Resistant Device vs intermaxillary elastics. Angle Orthod
2017;87:371-6.
5. Janson G, Niederberger A, Garib DG, Caldas W. Root resorption in
Class II malocclusion treatment with Class II elastics. Am J Orthod
Dentofacial Orthop 2016;150:585-91.
6. Bussick TJ, McNamara JA Jr. Dentoalveolar and skeletal changes
associated with the pendulum appliance. Am J Orthod Dentofacial
Orthop 2000;117:333-43.
7. Bolla E, Muratore F, Carano A, Bowman SJ. Evaluation of maxillary
molar distalization with the distal jet: a comparison with other
contemporary methods. Angle Orthod 2002;72:481-94.
8. Kircelli BH, Pektaş ZO, Kircelli C. Maxillary molar distalization with
a bone-anchored pendulum appliance. Angle Orthod 2006;76:
650-9.
9. Kinzinger GS, G€ ulden N, Yildizhan F, Diedrich PR. Efficiency of a
skeletonized distal jet appliance supported by miniscrew
anchorage for noncompliance maxillary molar distalization. Am
J Orthod Dentofacial Orthop 2009;136:578-86.
10. Park HS, Kwon TG, Sung JH. Nonextraction treatment with micro-
screw implants. Angle Orthod 2004;74:539-49.
11. Park HS, Lee SK, Kwon OW. Group distal movement of teeth using
microscrew implant anchorage. Angle Orthod 2005;75:602-9.
12. Jeon JM, Yu HS, Baik HS, Lee JS. En-masse distalization with min-
iscrew anchorage in Class II nonextraction treatment. J Clin Orthod
2006;40:472-6.
13. Yamada K, Kuroda S, Deguchi T, Takano-Yamamoto T,
Yamashiro T. Distal movement of maxillary molars using miniscrew
anchorage in the buccal interradicular region. Angle Orthod 2009;
79:78-84.
Kawamura et al
14. Choi YJ, Lee JS, Cha JY, Park YC. Total distalization of the maxillary
arch in a patient with skeletal Class II malocclusion. Am J Orthod
Dentofacial Orthop 2011;139:823-33.
15. Bechtold TE, Kim JW, Choi TH, Park YC, Lee KJ. Distalization
pattern of the maxillary arch depending on the number of ortho-
dontic miniscrews. Angle Orthod 2013;83:266-73.
16. Kook YA, Bayome M, Trang VT, Kim HJ, Park JH, Kim KB, et al.
Treatment effects of a modified palatal anchorage plate for distal-
ization evaluated with cone-beam computed tomography. Am J
Orthod Dentofacial Orthop 2014;146:47-54.
17. Wu X, Liu H, Luo C, Li Y, Ding Y. Three-dimensional evaluation on
the effect of maxillary dentition distalization with miniscrews im-
planted in the infrazygomatic crest. Implant Dent 2018;27:22-7.
18. Park CO, Sa’aed NL, Bayome M, Park JH, Kook YA, Park YS, et al.
Comparison of treatment effects between the modified C-palatal
plate and cervical pull headgear for total arch distalization in
adults. Korean J Orthod 2017;47:375-83.
19. Sugawara J, Kanzaki R, Takahashi I, Nagasaka H, Nanda R. Distal
movement of maxillary molars in nongrowing patients with the
skeletal anchorage system. Am J Orthod Dentofacial Orthop
2006;129:723-33.
20. Kawamura J, Park JH, Kojima Y, Tamaya N, Kook Y-A, Kyung H-M,
et al. Biomechanical analysis for total mesialization of the maxil-
lary dentition: A finite element study. Am J Orthod Dentofacial
Orthop 2021;159:790-8.
21. Kojima Y, Fukui H. A finite element simulation of initial move-
ment, orthodontic movement, and the centre of resistance of the
maxillary teeth connected with an archwire. Eur J Orthod 2014;
36:255-61.
22. Kawamura J, Park JH, Kojima Y, Kook YA, Kyung HM, Chae JM.
Biomechanical analysis for total mesialization of the mandibular
American Journal of Orthodontics and Dentofacial Orthopedics
265
dentition: a finite element study. Orthod Craniofac Res 2019;22:
329-36.
23. Chae JM, Park JH, Kojima Y, Tai K, Kook YA, Kyung HM. Biome-
chanical analysis for total distalization of the mandibular denti-
tion: a finite element study. Am J Orthod Dentofacial Orthop
2019;155:388-97.
24. Braun S, Legan HL. Changes in occlusion related to the cant of the
occlusal plane. Am J Orthod Dentofacial Orthop 1997;111:184-8.
25. Klontz HA. Tweed-Merrifield sequential directional force treat-
ment. Semin Orthod 1996;2:254-67.
26. Chae JM. A new protocol of Tweed-Merrifield directional force
technology with microimplant anchorage. Am J Orthod Dentofa-
cial Orthop 2006;130:100-9.
27. Isaacson RJ, Lindauer SJ, Rubenstein LK. Moments with the edge-
wise appliance: incisor torque control. Am J Orthod Dentofacial
Orthop 1993;103:428-38.
28. Jung MH, Kim TW. Biomechanical considerations in treatment
with miniscrew anchorage. Part 1: the sagittal plane. J Clin Orthod
2008;42:79-83.
29. Sung EH, Kim SJ, Chun YS, Park YC, Yu HS, Lee KJ. Distalization
pattern of whole maxillary dentition according to force application
points. Korean J Orthod 2015;45:20-8.
30. Liu H, Wu X, Yang L, Ding Y. Safe zones for miniscrews in maxillary
dentition distalization assessed with cone-beam computed to-
mography. Am J Orthod Dentofacial Orthop 2017;151:500-6.
31. Baek ES, Hwang S, Kim KH, Chung CJ. Total intrusion and distal-
ization of the maxillary arch to improve smile esthetics. Korean J
Orthod 2017;47:59-73.
32. Jung MH. A comparison of second premolar extraction and mini-
implant total arch distalization with interproximal stripping. Angle
Orthod 2013;83:680-5.
August 2021  Vol 160  Issue 2
ORIGINAL ARTICLE

Influence of the hyrax expander screw

position on displacement and stress
distribution in teeth: A study with finite
elements
Let
ıcia Chaves Fernandes,a Robert Willer Farinazzo Vitral,a Pedro Yoshito Noritomi,b Gustavo Silva Maximiano,a
and Marcio Jose
da Silva Camposa
Juiz de Fora, Minas Gerais, and Campinas, S~ao Paulo, Brazil

Introduction: This study aimed to simulate the different positions of the hyrax appliance expander screw and
evaluate tooth displacement and the stress distribution standard on the periodontal ligament using the finite
element method. Methods: Part of the maxilla with anchorage teeth, periodontal ligament, midpalatal suture,
and the hyrax appliance was modeled, and finite element method models were created to simulate 6 different
screw positions. There were 2 vertical positions at distances of 20 mm and 15 mm from the occlusal plane.
Another position was anteroposterior, the center of the screw placed between and equidistant from the mesial
face of the first molar and the distal face of the first premolar, aligned to the center of the crown of the first molar,
with the anterior edge of the screw aligned to the distal face of the first molar. A 1 mm activation of the expander
screw was simulated. The displacement (total, vertical, and buccolingual) and the stress distribution on the peri-
odontal ligament of supporting teeth in each model were registered. Results: The model simulating the
expander screw in a more occlusal and anterior position presented higher displacement values and higher stress
concentration, followed by the model with the screw in a more posterior but same vertical position. With the
exception of the first premolar, the teeth presented cervical-apical displacement in the vestibular face and
apical-cervical displacement in palatal faces. This displacement is compatible with the vestibular inclination
associated with the activation of the expander screw. The first premolar presented an atypical tendency for
the mesial and lingual displacement of the vestibular surface and counterclockwise rotation. Conclusions:
The supporting teeth presented a tendency for vestibular crown displacement and lingual root displacement
associated with compression areas in the vestibular-cervical region and tensile strength in the linguoapical
region. Placing the expander screw in a more occlusal and anterior position generated more mechanical
stress transfer, resulting in greater dental displacement. (Am J Orthod Dentofacial Orthop 2021;-:---)

R
apid maxillary expansion (RME) is the recommen-
ded treatment to correct the transversal maxillary
deficiency associated with crossbite in patients
undergoing bone growth.1 RME is a procedure that
a
Department of Orthodontics, Juiz de Fora Federal University, Juiz de Fora,
Minas Gerais, Brazil;
b
Renato Archer Information Technology Center, Campinas, S~ao Paulo, Brazil
All authors have completed and submitted the ICMJE Form for Disclosure of Po-
tential Conflicts of Interest, and none were reported.
This study was financed in part by the Conselho Nacional de Desenvolvimento
Cient
ıfico e Tecnol
ogico (CNPq) and Coordenaç~ao de Aperfeiçoamento de Pes-
soal de N
ıvel Superior (CAPES), Finance Code 001.
Address correspondence to: Robert Willer Farinazzo Vitral, Department of Ortho-
dontics, Juiz de Fora Federal University, R 21 de Abril, 117/404, Juiz de Fora,
Minas Gerais 36025-090, Brazil; e-mail, robertvitral@gmail.com.
Submitted, September 2019; revised and accepted, April 2020.
0889-5406/$36.00
Ó 2021 by the American Association of Orthodontists. All rights reserved.
https://doi.org/10.1016/j.ajodo.2020.04.031
separates the midpalatal suture (MPS) and expands the
maxilla transversally through lateral force applied on
the maxillary teeth and maxillary bones using ex-
panders.2-4 During RME, the expected result is the
sideward expansion of both maxillary segments.5,6 How-
ever, even though the strength applied on the maxillary
bones is high,7 this procedure is not merely orthopedic,
at it causes undesirable dental inclination of the teeth
supporting the expander.5,6,8,9 This fact harms the sta-
bility and the prognosis, which restricts the orthopedic
results of the treatment.8,10
During the activation of the expanders, the higher the
undesirable dental movement, the shorter the bone
expansion obtained because the excessive vestibular
dental inclination causes the clinical limit of RME to
be achieved early. When the palatal cusps of the maxil-
lary teeth and the vestibular cusps of the mandibular

1
2 Fernandes et al
teeth touch, the MPS opening is reduced.5,6,11 Although
the orthopedic changes obtained with the RME may pre-
sent some degree of recurrence,12 dental movement is
the most unstable change in this procedure.10,13 Its re-
striction is related to a lower degree of recurrence after
correcting posterior crossbite.5,6,9
One of the devices used to perform RME is the hyrax
expander screw, which has a side expander transversal to
the MPS. When activated, this expander forces the
maxilla segments laterally.3,4 Laboratory production al-
lows changes in the palate along with both the height
and anteroposterior axes. Such changes may interfere
clinically in distributing orthopedic forces generated by
the expander and in dental effects during use, influ-
encing the efficiency and the stability of RME.7,11
The finite element method (FEM) is a computer
method applied to biomechanics that is used to deter-
mine stress and deformation in structures submitted to
different mechanical loads.14,15 In orthodontics, the
FEM has been used to analyze the tendency of move-
ment and tension distribution on teeth and craniofacial
bones during mechanic orthodontic simulations, such as
RME.7,11,16-19
A FEM study conducted by Fernandes et al7 described
a standard of tension and deformation distribution on
maxillary bone structures after placing the expander
screw into 6 different positions during maxillary expan-
sion. Only the impact on bone structures was analyzed.
The possibility of dental movement during the process
was ignored.
The objective of this study was to simulate the
different vertical and anteroposterior positions of the
hyrax appliance expander screw and to evaluate using
the FEM tooth displacement and stress distribution on
the periodontal ligament.
MATERIAL AND METHODS
A computer-aided design model from the Renato
Archer Information Technology Center, Campinas, S~ao
Paulo, Brazil, was employed. It included the maxilla,
the skull base (with the zygomatic, nasal, sphenoid,
and frontal bones), the central incisor, the lateral incisor,
the canine and the first premolar (U4), the second pre-
molars (U5), the first molar (U6), the second molars
(U7), the periodontal ligament, and a bone-suture unit
representing the MPS. The dental crowns contact and
transfer the force to each other. The model was created
(Rhinoceros 4.0; McNeel North America, Seattle, Wash)
from computerized tomography images (GE Lightspeed
Pro 16; GE Healthcare, Chicago, Ill) taken from an
adult, without any evident facial asymmetry, with all
the permanent teeth emerged, with the exception of
- 2021  Vol -  Issue - American Journal of Orthodontics and Dentofacial Orthopedics
the third molars, and without any dental restoration or
congenital or acquired craniofacial alterations. The use
of these images was approved by the ethics research
committee of the University of S~ao Paulo (no. 97/06).
The anatomic model of part of the maxilla, teeth,
periodontal ligament, and MPS was imported into the
software program FEMAP (version 10.1.1; Siemens
PLM Software, Plano, Tex), incorporating the single-
body hyrax appliance, composed of 1 expander screw
and 3 wire segments of 0.036-in diameter that joined
the screw to the U4 and U6 and the teeth to each other
(from U4 to U7), resulting in a geometric model with a
tetrahedral mesh (Fig 1).
The geometric model was subjected to mathematical
analysis (Ansys 17.2; Ansys, Inc, Canonsburg, Pa), using
a bone thickness of 2 mm and bar elements with elastic
properties to represent the MPS. A horizontal movement
restriction was imposed on the body of the device to
simulate soldering to the orthodontic bands connected
to the U4 and U6. The model structures were determined
with specific properties (Table I), and the simulated ma-
terials had elastic, isotropic, and uniform characteristics.
Six distinct positions of the expander screw were
simulated in FEM models.7 Three of them were antero-
posterior, and 2 of them were vertical. In all the simula-
tions, the expander screw was placed in the transversal
center of the palate, perpendicular to the MPS and par-
allel to the occlusal plane.
In anteroposterior position 1, the center of the screw
was positioned equidistant to the mesial face of U6 and
the distal face of U4. In anteroposterior position 2, the
center of the screw was aligned to the center of the U6
crown. In anterior position 3, the anterior edge of the
expander was aligned to the distal face of U6. Vertically,
the expander screw was positioned 20 mm (vertical po-
sition 1) and 15 mm (vertical position 2) from the
occlusal plane. Table II and Figure 2 show the positions
of all 6 models.
Fig 1. FEM model comprising bone, teeth, periodontal
ligament, MPS, and hyrax appliance.
Fernandes et al 3

The M4 model presented the highest maximum total

Table I. Mechanical properties attributed to the struc-
displacements among all models, with values close to
tures of the geometric model
0.4 mm in U6 and U5, 0.3 mm in U4, and 0.15 mm in
Poisson Young's U7. However, this model was the only one whose
Material coefficient modulus (MPa) expander screw presented total displacement values
Cortical bone14 0.3 13,700 lower than 0.533 mm (orange displacement). The M5
Trabecular bone14 0.3 1370
Teeth20 0.3 20,000
model presented a standard similar to M4 in teeth
Periodontal ligament21 0.49 0.69 displacement. However, its maximum values were
Hyrax expander14 0.33 200,000 reduced between 30% and 50%. The M6 model pre-
Midpalatal suture22 0.49 1 sented a maximum total displacement 3 times lower
than M4, with a standard displacement similar to M2.
A condition for the outline of the maxillary bone was However, its values were slightly higher in the distolin-
set for both stress distribution and displacement analysis gual area of U6.
to restrict vertical, anteroposterior, and transverse move- A diastema opened between U4 and U5. In all models,
ments in the model. A bar element was created for each U4 presented mesial and lingual displacement of the
node on the edge of the face in the MPS and perpendic- vestibular face. It also presented a counterclockwise rota-
ular to it,7 simulating the MPS during the RME. The acti- tion (Video 1, available at www.ajodo.org: M1, M2, and
vation of the hyrax appliance was achieved only by the M3; Video 2, available at www.ajodo.org: M4, M5, and
enforced displacement toward maxillary expansion. M6). This rotation was observed even in M3, where
In the MPS region, symmetry was required as a con- vestibular tooth displacement was insignificant.
dition, and the loading was recreated symmetrically on In M2, M4, M5, and M6, there was a tendency toward
the opposite side to obtain equivalent results for both a cervical-apical displacement of the vestibular face in
sides. For each model, a transversal displacement of U5, U6, and U7. However, U4 presented a contrary
0.5 mm in the center of the screw was simulated. displacement (apical-cervical) of the vestibular face in
Because of its symmetry, it was the same as 1 mm of acti- models: M1, M2, M3, and M6 (Fig 4, A). Through a
vation of the hyrax expander. lingual view, the teeth presented apical-cervical
displacement, with an exception in U4 again, which pre-
RESULTS sented apical displacement of the dental crown in
models M3 and M4 (Fig 4, B). Models M4 and M5 pre-
After simulating the opening of the expander screw, sented more significant vertical displacements, located
tooth displacement was analyzed as a whole and sepa- in U6 and U5, apical-cervically toward the palatal roots
rately on vertical and buccolingual axes. and cervical-apically toward the vestibular faces of those
The total tooth displacement in each model after teeth. Contrary to the tendency of other teeth, U4 pre-
opening the screw was evaluated from an axial point sented apical-cervical displacement toward the vestib-
of view, represented in Figure 3. The M1, M2, and M3 ular face in M1, M2, M3, and M6 and cervical-apically
models presented a similar maximum total displacement toward the lingual face of its crown in M3 and M6.
of approximately 0.2 mm. However, such displacements Evaluation of the isolated horizontal displacement
occurred in different areas. On the M1 model, there was (Fig 5) showed a general tendency for vestibular
a distolingual displacement in U6, U5, and the vestibular displacement of dental crowns, being inexpressive in
area of U4. In M2, the entire U6 crown presented U5 and U7 in model M3 and U7 in M1. Models M4
displacement. A few areas in U4 and U5 also presented and M5 presented the greatest horizontal displacements,
short displacement. In M3, only the vestibular and mesial with maximum values near the occlusal regions of U5
faces of U4 and the distolingual face of U6 presented and U6. However, in M5, the values were approximately
displacement. 55% lower than in M4.
In models M1, M2, M3, and M6, the 3-dimensional
representation (initial position) on the distal vestibular
Table II. FEM related to the anteroposterior and verti- region of the first premolar indicate that this tooth un-
cal positions of the expander screw derwent mesial movement (Fig 5, A), which causes an in-
crease in the size of the 3-dimensional models, as the
Vertical variation
expander screw is displaced posteriorly in all the vertical
Anteroposterior variation Vertical position 1 Vertical position 2
Anteroposterior position 1 Model 1 (M1) Model 4 (M4) position models 1 (M1, M2, and M3).
Anteroposterior position 2 Model 2 (M2) Model 5 (M5) Although the dental crown in U7 did not present
Anteroposterior position 3 Model 3 (M3) Model 6 (M6) any significant horizontal displacement in any of the

American Journal of Orthodontics and Dentofacial Orthopedics - 2021  Vol -  Issue -

4 Fernandes et al

Fig 2. Sagittal view of anteroposterior and vertical positions of the expander screw.

models simulated in this study, the apical region of the models M4 and M5. Such displacement, associated
lingual root of this tooth presented a lingual with the vestibular displacement of the crown,
displacement in all models (Fig 5, B). A similar reinforces the tendency of vestibular inclination of
movement was observed in the U6 lingual root in the teeth during RME.

Fig 3. Axial view of the total dental displacement (in millimeters) of simulated models.

- 2021  Vol -  Issue - American Journal of Orthodontics and Dentofacial Orthopedics

Fernandes et al 5

Fig 4. Vertical teeth displacement (in millimeters) through vestibular (A) and lingual (B) views. The
displacement in the cervical-apical direction was represented by cold colors (blue), and the apical-
cervical direction was represented by hot colors (red).

Aside from displacement, the distribution of tensile
strength (Fig 6) and compression (Fig 7) on areas repre-
senting the periodontal ligament on the posterior teeth
root surface was also evaluated.
The areas of higher tensile strength concentration
occurred in the vestibular-apical lingual root in U6 and
U7 (Fig 6, A) and the cervical region of the lingual faces
in U5 and U6 (Fig 6, B), confirming the tendency for the
vestibular inclination of these teeth after activation of
the expander screw. The values of maximum traction
(warm colors) were located at the top of the palatal
root in the first model (approximately 5 or more MPa)
in M4. The first premolar presented areas of high tensile
strength on the lingual face of the lingual root, espe-
cially on the mesial region (Fig 6, A and B), reinforcing
the tendency for rotation in this tooth.
The highest compression values (cold colors)
occurred in the cervical and lingual-apical region of
vestibular roots of U6 and U7, especially in M4 and
M5 (Fig 7, A), suggesting a tendency for vestibular
inclination and displacement of teeth. The first
premolar presented areas of compression on the vestib-
ular face of the vestibular root, especially on the mesial
region (Fig 7, A and B), indicating a tendency for rota-
tion in this tooth.
DISCUSSION
Although RME is considered a treatment to have
proven effective for patients with maxillary transverse
deficiency (maxillary atresia), its study and the improve-
ment of expander appliances aim to reduce undesirable
dental side effects and maximize orthopedic ef-
fects.11,23,24
During RME, the desirable movement is the lateral
dislocation of both right and maxillary left segments,
with a minimal inclination of teeth. However, even
with the great force7 applied during RME, it is not
possible to obtain an exclusively orthopedic maxillary
expansion. The vestibular dental inclination is undesir-
able,5,6,8,9 impairs stability and prognosis and limits
the orthopedic results of the treatment.8,10

American Journal of Orthodontics and Dentofacial Orthopedics - 2021  Vol -  Issue -

6 Fernandes et al

Fig 5. Horizontal dental displacement (in millimeters) from vestibular (A) and lingual (B) views. The
black mesh represents the initial position of teeth in the system before the activation of the expander
screw.

The FEM has been used to demonstrate tensions and
deformations imposed on teeth and bones during
RME.7,11,16,25,26 In a previous study7 simulating 6 posi-
tions of the expander screw with anteroposterior and
vertical variations, FEMs were used to simulate the
maxillary bone structure, the hyrax anchor teeth (first
premolar and first molar), and the MPS to analyze the
distribution of tensile strength on the maxilla during
RME. Then, to analyze the effect on the teeth of the
same positions of the expander screw, the study added
the second premolar, the second molar, and the peri-
odontal ligament of the involved teeth to the computer-
ized models in the experiment. The aim was to evaluate
dental movement caused by the same activation of
simulated expander appliances.
Generally, the simulated models exhibited less total
dental displacement as the expander screw was moved
posteriorly and along the palate. This result explains
the higher distance between the screw and the point of
application of force (the dental crowns of the first pre-
molar and the first molar), which amplifies the lever
arms and the extension and flexibility of the appliance's
dimensions. This increase in the extension of the appli-
ance was also associated with a higher rotation of the
first premolar, higher tensile strength in the screw
legs,11,19 lower effectiveness in transferring strength to
the maxilla,7 and reduction of dental movement in the
direction of the screw opening,11,19 as shown in the pre-
sent study.
Contrary to the tensile strength variations in the
expander screw,11,19 in the present study, the connect-
ing legs of the hyrax presented similar displacements
among all models, but M4 was the only one to present
displacement values under 0.533 mm in the anterior
position of the expander screw (orange displacement).
This may have occurred because the movement of the
expander screw after its activation is limited because
of the flexibility of its connecting legs, the opening

- 2021  Vol -  Issue - American Journal of Orthodontics and Dentofacial Orthopedics

Fernandes et al
Fig 6. Vestibular (A) and lingual (B) views of tensile stress distribution (in megapascal) on the peri-
odontal ligament.
of the MPS, and the dental displacement. The lower
flexibility of the connecting legs in M4 made the dental
displacement restrict the expander screw's movement
even more, even though it presented the highest values
among all models.
There was a greater vestibular displacement of the
occlusal areas of posterior teeth toward apical areas, fol-
lowed by lingual and extrusive displacement of lingual
root apices and increased compression on the
vestibular-cervical regions and increased tensile strength
in the lingual-apical regions of the periodontal ligament,
indicating a vestibular inclination of teeth. This ten-
dency to inclination seems to be proportional to the total
dental displacement that occurred in each model. Such a
tendency is not associated with any specific position of
the expander screw. In contrast, Araugio et al11 identi-
fied lower vestibular dental inclination as the expander
screw became farther from the occlusal plane. Even a
lingual inclination of teeth of around 5 mm over vertical
position 2 was observed. The use of a simplified com-
puter model that disregarded the structural properties
of maxillary bones and used elastic supports placed
over vestibular surfaces of dental roots to imitate the
American Journal of Orthodontics and Dentofacial Orthopedics
7
periodontal ligament may have generated a discrepancy
regarding the results of this study.
Generally, posterior teeth presented a tendency for
displacement in the apical-cervical direction of the
lingual face, especially on the lingual root apex of U6,
and in the cervical-apical direction of the vestibular
face. Such displacements are compatible with the ten-
dency for vestibular dental inclination expected
during RME,5,6,8,9 which is an idea confirmed by the
compression areas of the periodontal ligament in the
vestibular-cervical area of teeth. The atypical vertical
displacement presented in U4 in the models of lower
dental movement amplitude may be related to the mesial
and lingual displacement of the vestibular face, which
has also stood apart from the general tendency. Using
a more apical positioning of the expander screw than
in the present study, Araugio et al11 recorded an extru-
sive movement of the vestibular face of all posterior
teeth, during which U4 did not present any irregular re-
sults in the simulated models.
The extrusive displacement of lingual faces of poste-
rior teeth and the standard opening of the maxilla with
its vertex superiorly and posteriorly placed7 that
- 2021  Vol -  Issue -
8 Fernandes et al

Fig 7. Vestibular (A) and lingual (B) views of compressive stress distribution (in megapascal) on peri-
odontal ligament.

occurred on the 6 positions simulated by the expander
screw may be the reason for the RME effects described
previously,27,28 such as the inferior displacement of the
maxilla, counterclockwise rotation of the jaw, and bite
opening.
In a study from Fernandes el al,7 model M4 presented
a more efficient transfer of mechanical effort from the
expander screw toward the bone structures. This transfer
has a higher maximum tensile strength and occurs
nearer to the incisive foramen, and facilitates the initial
opening movement of the MPS.29 In the present study,
the same model presented a higher amplitude of dental
displacement, reinforcing the relationship between the
amount of maxillary expansion and inclination of
posterior teeth,30 which indicates that, among the
expander screw positions that were simulated, the
desirable effect of opening the MPS is necessarily
followed by a dental effect, which must be considered
during the performance of such types of orthopedic
treatment.
Although the relation between dental displacement
and maxillary expansion was established,30 model M6
presented the poorest distribution of tensions in the
maxilla7 and a total dental displacement higher than
models M1, M2, and M3. However, this displacement
mainly consisted of the counterclockwise rotation of
U4 and U6 and the lingual displacement of the dental
crown in U4, which made it hardly productive, or entirely
unproductive, to correct malocclusion related to trans-
verse maxillary deficiency.
In silico studies, like FEM, must be considered
before human application because they simulate
only the morphology and mechanical properties of
biological systems in a computer, disregarding cellular
and chemical mechanisms and individual biological
features, such as quantity, shape, and dental root
area; inclination and position of teeth; format, qual-
ity, and quantity of maxilla bones and alveolar pro-
cess; and MPS hardness. Consequently, this study
had the objective of specifically informing the me-
chanical consequences of changing the hyrax
expander screw position, highlighting the possibility
of obtaining different results in oral tissues during
RME. Therefore, we suggest further investigations
involving clinical trials and computerized methods to
analyze more design factors to help precisely deter-
mine, at the same time, the bone and teeth effects
of each expander device setting.

- 2021  Vol -  Issue - American Journal of Orthodontics and Dentofacial Orthopedics

Fernandes et al
CONCLUSIONS
The simulations performed in this study using the
FEM showed the following:
1. There was a tendency for displacement in the
cervical-apical direction of the vestibular face and
the apical-cervical direction of the palatal face of
teeth, with a vestibular displacement of the crown
and roots.
2. There were compression areas in the vestibular and
cervical region and stress in the lingual-apical re-
gion of teeth.
3. Placing the expander screw in a more occlusal and
anterior position generated mechanical stress trans-
fer that resulted in a greater dental displacement.
AUTHOR CREDIT STATEMENT
Let
ıcia Chaves Fernandes contributed to methodol-
ogy, formal analysis, investigation, and original draft
preparation; Robert Willer Farinazzo Vitral contributed
to conceptualization, formal analysis, draft review and
editing, supervision, and funding acquisition; Pedro
Yoshito Noritomi contributed to methodology, valida-
tion, investigation, and resources; Gustavo Silva Maxi-
miano contributed to visualization and original draft
preparation; and Marcio Jos
e da Silva Campos contrib-
uted to conceptualization, formal analysis, original draft
preparation, and project administration.
SUPPLEMENTARY DATA
Supplementary data associated with this article can
be found, in the online version, at https://doi.org/10.
1016/j.ajodo.2020.04.031.
REFERENCES
1. Podesser B, Williams S, Crismani AG, Bantleon HP. Evaluation of
the effects of rapid maxillary expansion in growing children using
computer tomography scanning: a pilot study. Eur J Orthod 2007;
29:37-44.
2. Biederman W. A hygienic appliance for rapid expansion. JPO J
Pract Orthod 1968;2:67-70.
3. Chaconas SJ, Caputo AA. Observation of orthopedic force distribu-
tion produced by maxillary orthodontic appliances. Am J Orthod
1982;82:492-501.
4. Garib DG, Henriques JF, Janson G, Freitas MR, Coelho RA. Rapid
maxillary expansion–tooth tissue-borne versus tooth-borne ex-
panders: a computed tomography evaluation of dentoskeletal ef-
fects. Angle Orthod 2005;75:548-57.
5. Starnbach HK. The effects of splitting the mid-palatal suture on
the surrounding structures. Am J Orthod 1964;50:923-4.
6. Starnbach H, Bayne D, Cleall J, Subtelny JD. Facioskeletal and
dental changes resulting from rapid maxillary expansion. Angle
Orthod 1966;36:152-64.
7. Fernandes LC, Farinazzo Vitral RW, Noritomi PY, Schmitberger CA,
Jos
e da Silva Campos M. Influence of the hyrax expander screw
American Journal of Orthodontics and Dentofacial Orthopedics
9
position on stress distribution in the maxilla: a study with finite el-
ements. Am J Orthod Dentofacial Orthop 2019;155:80-7.
8. Haas AJ. Palatal expansion: just the beginning of dentofacial or-
thopedics. Am J Orthod 1970;57:219-55.
9. Mew J. Relapse following maxillary expansion. A study of twenty-
five consecutive cases. Am J Orthod 1983;83:56-61.
10. Haas AJ. Rapid palatal expansion: a recommmended prerequisite
to Class III treatment. Trans Eur Orthod Soc 1973;311-8.
11. Araugio RM, Landre J Jr, Silva Dde L, Pacheco W, Pithon MM,
Oliveira DD. Influence of the expansion screw height on the dental
effects of the hyrax expander: a study with finite elements. Am J
Orthod Dentofacial Orthop 2013;143:221-7.
12. Sarn€as KV, Bj€ork A, Rune B. Long-term effect of rapid
maxillary expansion studied in one patient with the aid of metallic
implants and roentgen stereometry. Eur J Orthod 1992;14:
427-32.
13. Timms DJ. A study of basal movement with rapid maxillary expan-
sion. Am J Orthod 1980;77:500-7.
14. Serpe LCT, Las Casas EBd, Toyofuku ACMM, Gonz
alez-Torres LA.
A bilinear elastic constitutive model applied for midpalatal suture
behavior during rapid maxillary expansion. Res Biomed Eng 2015;
31:319-27.
15. Magesh V, Harikrishnan P, Kingsly Jeba Singh D. Finite element
analysis of slot wall deformation in stainless steel and titanium or-
thodontic brackets during simulated palatal root torque. Am J Or-
thod Dentofacial Orthop 2018;153:481-8.
16. Provatidis CG, Georgiopoulos B, Kotinas A, McDonald JP. Evalua-
tion of craniofacial effects during rapid maxillary expansion
through combined in vivo/in vitro and finite element studies.
Eur J Orthod 2008;30:437-48.
17. Jafari A, Shetty KS, Kumar M. Study of stress distribution and
displacement of various craniofacial structures following applica-
tion of transverse orthopedic forces–a three-dimensional FEM
study. Angle Orthod 2003;73:12-20.
18. Lee HK, Bayome M, Ahn CS, KIM SH, Kim KB, Mo SS, et al. Stress
distribution and displacement by different bone-borne palatal ex-
panders with micro-implants: a three-dimensional finite-element
analysis. Eur J Orthod 2014;36:531-40.
19. Matsuyama Y, Motoyoshi M, Tsurumachi N, Shimizu N. Effects of
palate depth, modified arm shape, and anchor screw on rapid
maxillary expansion: a finite element analysis. Eur J Orthod
2015;37:188-93.
20. Serpe LCT, Torres LAG, de Freitas Pinto RU, Toyofuku ACMM, de
Las Casas EB. Maxillary biomechanical study during rapid expan-
sion treatment with simplified model. J Med Imaging Health
Inform 2014;4:137-41.
21. Andersen KL, Pedersen EH, Melsen B. Material parameters and
stress profiles within the periodontal ligament. Am J Orthod Den-
tofacial Orthop 1991;99:427-40.
22. Tanne K, Hiraga J, Sakuda M. Effects of directions of maxillary
protraction forces on biomechanical changes in craniofacial com-
plex. Eur J Orthod 1989;11:382-91.
23. Trindade IE, Castilho RL, Sampaio-Teixeira AC, Trindade-
Suedam IK, Silva-Filho OG. Effects of orthopedic rapid maxillary
expansion on internal nasal dimensions in children with cleft lip
and palate assessed by acoustic rhinometry. J Craniofac Surg
2010;21:306-11.
24. Goldenberg DC, Goldenberg FC, Alonso N, Gebrin ES,
Amaral TS, Scanavini MA, et al. Hyrax appliance opening and
pattern of skeletal maxillary expansion after surgically assisted
rapid palatal expansion: a computed tomography evaluation.
Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2008;106:
812-9.
- 2021  Vol -  Issue -
10
25. Han UA, Kim Y, Park JU. Three-dimensional finite element analysis
of stress distribution and displacement of the maxilla following
surgically assisted rapid maxillary expansion. J Craniomaxillofac
Surg 2009;37:145-54.
26. Gautam P, Zhao L, Patel P. Biomechanical response of the maxil-
lofacial skeleton to transpalatal orthopedic force in a unilateral
palatal cleft. Angle Orthod 2011;81:503-9.
27. Haas AJ. The treatment of maxillary deficiency by opening the
midpalatal suture. Angle Orthod 1965;35:200-17.
- 2021  Vol -  Issue - American Journal of Orthodontics and Dentofacial Orthopedics
Fernandes et al
28. Heflin BM. A three-dimensional cephalometric study of the influ-
ence of expansion of the midpalatal suture on the bones of the
face. Am J Orthod 1970;57:194-5.
29. Liu S, Xu T, Zou W. Effects of rapid maxillary expansion on the
midpalatal suture: a systematic review. Eur J Orthod 2015;37:
651-5.
30. Bassarelli T, Dalstra M, Melsen B. Changes in clinical crown height
as a result of transverse expansion of the maxilla in adults. Eur J
Orthod 2005;27:121-8.
ORIGINAL ARTICLE

Attractiveness assessment by
orthodontists and laypeople judging
female profile modifications of Class II
Division 1 malocclusion
Kadriye Kalin,a Sebahat Yesim Iskender,b and Reinder Kuitertc
Utrecht and Amsterdam, The Netherlands

Introduction: The objective of this study was to evaluate the differences in preference between orthodontists
and laypeople, judging soft tissue digital alterations of a Class II Division 1 profile of a female patient with mandib-
ular retrognathia, produced by simulated camouflage and mandibular advancement therapy. Methods: The pro-
file image of a White woman with a Class II Division 1 mandibular retrognathic profile was digitally modified to
produce 7 pictures: 1 baseline, 3 stepwise increase in the nasolabial angle of 113 , 121 , and 129 , and 3 step-
wise increase in chin-neck length of 51 mm, 54 mm, and 57 mm. Forty-four orthodontists and 162 laypeople
assessed these 7 images. Results: The untreated baseline profile was found to be least attractive for both or-
thodontists and laypeople, with orthodontists scoring significantly lower than laypeople. The profiles represent-
ing mandibular advancement therapy were judged significantly better by both groups than camouflage therapy.
Orthodontists preferred straighter profiles than laypeople, giving the highest-ranking to a chin-neck length of
57 mm, whereas laypeople gave the highest rank to a chin-neck length of 54 mm. Conclusions: Orthodontists
prefer straighter profiles and gave a lower ranking to the untreated Class II Division 1 female profile compared
with laypeople. Orthodontists and laypeople favor mandibular advancement therapy over camouflage therapy.
However, both groups seem to prefer the effect of both treatment modalities over the untreated baseline Class II
Division 1 profile. (Am J Orthod Dentofacial Orthop 2021;160:276-82)

S
everal studies have shown that an attractive indi-
vidual is perceived as happier, more sociable, and
more successful.1,2 Attractive people are also
treated and judged more positively3,4; therefore, people
want to improve their facial esthetics.5-12 When it comes
to beauty, the face is the most important factor.13,14 This
is 1 of the main reasons people seek treatment by an
orthodontist or an oral surgeon to improve their facial
esthetics.
Treatment planning by orthodontists and oral sur-
geons involves a profile analysis. The most frequent
malocclusion in the Netherlands is Class II Division 1
a increasing the chin-neck length.21
Private practice, Utrecht, The Netherlands.
b
Northo Orthodontistenpraktijk, Amsterdam, The Netherlands.
c
Department of Orthodontics, Academisch Centrum Tandheelkunde Amsterdam,
Amsterdam, The Netherlands.
All authors have completed and submitted the ICMJE Form for Disclosure of Po-
tential Conflicts of Interest, and none were reported.
Address correspondence to: Kadriye Kalin, Private practice, Wibautstraat 29,
3555WE Utrecht, The Netherlands; e-mail, Kadriye.kalin@live.nl.
Submitted, October 2019; revised, March 2020; accepted, April 2020.
0889-5406/$36.00
Ó 2021 by the American Association of Orthodontists. All rights reserved.
https://doi.org/10.1016/j.ajodo.2020.04.032
skeletal malocclusion (prevalence of 65%).15 To correct
this malocclusion in adolescent patients, a wide range
of functional or orthopedic appliances are available,16-19
which is in contrast to adult patients Class II Division 1
malocclusion. For adult patients, 2 effective treatment
modalities are available. The first treatment modality is
camouflage treatment therapy, with retraction of
maxillary incisors after extraction of maxillary teeth
resulting in overjet reduction with the posterior
movement of the upper lip and an increased nasolabial
angle.20 The second treatment modality is the surgical
advancement of the mandible, often improving overjet,
occlusion, dentofacial pattern, and profile esthetics by
A previous study by Y€ uksel et al22 has shown that
laypeople preferred mandibular advancement over cam-
ouflage therapy in profile esthetics in adults with Class II
Division 1 malocclusion. Knowledge of what laypeople
and orthodontists consider as attractive and whether
laypeople and professionals would evaluate improve-
ment in facial attractiveness achieved by orthodontic
treatment, in the same way, could be important when

276
Kalin, Iskender, and Kuitert
considering treatment planning for adults with Class II
Division 1 malocclusion. The judgment of profile es-
thetics by professionals vs judgment by laypeople has
been repeatedly compared, but with conflicting out-
comes. Therefore, it would be interesting to assess if
there is a difference in preference between orthodontists
and laypeople regarding the outcome of the 2 different
treatment methods for adult patients with Class II Divi-
sion 1 malocclusion. The outcome might be helpful in
choosing between retracting maxillary teeth via extrac-
tion or headgear vs advancing the mandible with func-
tional appliances or oral surgery when planning
treatment of adults and adolescents with Class II Divi-
sion 1 malocclusion.
This study aimed to investigate and compare the
judgments of laypeople and orthodontists on the effect
of different soft tissue alterations on the profile of a
Class II Division 1 female subject with mandibular retro-
gnathia, produced by software simulated camouflage or
mandibular advancement treatment, by showing the
different alterations simultaneously. In addition, to
investigate and compare the amount of tolerable naso-
labial angle increase and the amount of chin-neck
length increase necessary for satisfactory profile
improvement during treatment of adult patients with
Class II Division 1 malocclusion.
MATERIAL AND METHODS
Seven modified profile pictures of a White woman
developed for a previous study22 were used. The woman
used in this study has an untreated Class II Division
1 malocclusion with a normal face height, normal
mandibular plane angle, nasolabial angle of 115.7 ,
chin-neck length of 51 mm, z-angle of 70 , angle of
facial convexity of 21 , and a Holdaway’s profile angle
of 18 . These measurements are sufficiently beyond
the respective limits between Class I and Class II soft tis-
sue profiles of 75 , 15 , and 10 .23,24
In the previous study,22 this original photograph (un-
treated female profile with Class II Division 1 malocclu-
sion) was modified with Photoshop software (Adobe,
San Jose, Calif) and Dolphin Imaging software (Dolphin
Imaging and Management Solutions, Chatsworth, Calif)
to emphasize the effects of the 2 treatment modalities.
The scanned profile image and lateral cephalogram
were both imported into software (version 11.5; Dolphin
Imaging and Management Solutions). Then the images
were linked to each other to standardize the profile im-
age with the magnification of the cephalogram.
To emphasize mandibular retrusion, the profile pic-
ture was altered with Photoshop software (Adobe). To
create the starting profile, profile D, the chin-neck length
of the original profile was reduced from 51 mm to
American Journal of Orthodontics and Dentofacial Orthopedics
277
45 mm, by subtracting 2 standard deviations from the
norm of 57 6 6 mm as determined by Lehman Jr,25,26
resulting in 45 mm for this variable (Fig 1, D).
To create 3 additional images simulating the treat-
ment outcome of mandibular advancement surgery,
the profile image was digitally altered using Dolphin Im-
aging software. The chin-neck length was increased by
1.0, 1.5, and 2.0 standard deviations resulting in chin-
neck lengths of 51 mm, 54 mm, and 57 mm, respectively
(Fig 1, E, F, and G).
To simulate the treatment effect of camouflage treat-
ment on the nasolabial angle, a similar procedure was
performed. The original nasolabial angle of 115.7 ,
modified to 104.9 to achieve the starting point (Fig 1,
D) for further alterations according to Sinno et al.27
Three additional profile types were created by gradually
increasing the nasolabial angle using Dolphin software.
The nasolabial angle was increased by 2.0, 4.0, and 6.0
standard deviations resulting in angles of 113 , 121 ,
and 129 (Fig 1, C, B, and A).
The original, untreated profile (D) (Fig 1, D) with
a chin-neck length of 45 mm and a nasolabial angle
of 104.9 was the starting point for all modifications.
This profile shows the most pronounced Class II
Division 1 features with the largest sagittal interlabial
step.
The 7 created profiles were inserted in a booklet,
including a questionnaire providing information on fac-
tors that could influence the judgment of the observers.
This questionnaire collected information about the age,
gender, and educational level of the observers. Addi-
tional information was obtained from the orthodontists
about their site of training and years of experience.
The laypersons were approached on the streets and in
several parks in Amsterdam in 2016 by an undergraduate
dental student. In addition, patients in waiting rooms at
dental practices in Amsterdam were approached by the
same dental student. The orthodontists were ap-
proached at an orthodontic congress in Soestduinen,
The Netherlands, in March 2016 by another undergrad-
uate dental student. Inclusion criteria for laypeople
included the following: aged .17 years, minimum edu-
cation of higher general secondary school, and primary
school finished in the Netherlands. The inclusion criteria
for the orthodontists was full membership of the Dutch
Orthodontic Society. In total, 162 laypeople, 53% of
them women and 47% men, and 44 orthodontists,
evenly distributed by gender, were willing to participate.
The power (a 5 0.05) of this study was 0.83, which is
calculated with G-Power 3.1.9.4 (Heinrich-Heine-Uni-
versit€at D€
usseldorf, D€usseldorf, Germany). This study
was previously approved by the Ethics Committee (no.
X14-0355).
August 2021  Vol 160  Issue 2
278 Kalin, Iskender, and Kuitert

Fig. A starting profile image (D) was digitally altered to simulate increases in the nasolabial angle by
2.0, 4.0, and 6.0 standard deviations (C, 113 ; B, 121 ; A, 129 ); mandibular advancement surgery,
increasing the chin-neck length by 1.0, 1.5, and 2.0 standard deviations (E, 51 mm; F, 54 mm;
G, 57 mm).

Table I. Mean ranking scores and standard deviation

Orthodontist Laypeople

Profile Mean Standard deviation Mean Standard deviation

A (6) 5.23 B, D, E, F, G 1.48 (6) 4.90 B, C, D, E, F, G 1.76
B (4) 4.55 A, D, E, F, G 0.95 (5) 4.40 A, D, E, F, G 1.57
C (5)* 5.09 D, E, F, G 0.94 (4)* 4.24 A, D, E, F, G 1.58
D (7)* 6.39 A, B, C, E, F, G 1.22 (7)* 5.65 A, B, C, E, F, G 1.85
E (3) 3.16 A, B, C, D, F, G 1.14 (3) 3.22 A, B, C, D, F 1.49
F (2)* 2.00 A, B, C, D, E 0.84 (1)* 2.70 A, B, C, D, E 1.72
G (1)* 1.59 A, B, C, D, E 1.25 (2)* 2.90 A, B, C, D 2.07

Note. Significant at P \0.01 differences between profiles assessed by orthodontists or laypeople are noted in superscript.
*Significant P \0.01 difference in the ranking of corresponding profiles between orthodontists and laypeople. The final ranking order is shown in
parentheses, with 1 the most attractive and 7 the least attractive.

The Figure shows the profiles printed alongside each
other on an A4 page. This presentation aimed to show
the participating observers a set of profiles offering clear
options by placing the most pronounced Class II Division
1 profile in the center (Fig 1, D) with a nasolabial angle
of 104.9 and chin-neck length of 45 mm. Assuming a
distinct selection between options for the observers for
their preferred profile, the maximum compensation
with camouflage treatment was placed on the far left
(Fig 1, A) with a nasolabial angle of 129 and the profile
with the maximum mandibular advancement surgery on
the far right (Fig 1, G) with a chin-neck length of 57 mm.
Participating laypeople and orthodontists were asked
to rank the profiles from 1 (most attractive) to 7 (least
attractive).28,29 A reliability test was carried out by 20
undergraduate dental students, who assessed the set
of profile images with the visual analogue scale (VAS)
score twice with a washout period of 2 weeks.
Statistical analysis
Friedman and Wilcoxon tests were used to analyze
facial attractiveness ranking. The influence of back-
ground factors on ranking was analyzed by Kruskal-
Wallis and Mann-Whitney U tests. To analyze the
mean ranking difference between laypeople and ortho-
dontists, Mann-Whitney U tests were used. The Pearson
correlation test was to assess the reliability by analyzing
the correlation between the ranking scores.
RESULTS
The estimated means of the rankings calculated for
the orthodontists and laypeople are shown in Table I.
The orthodontists and laypeople considered profile D
the least attractive. Profile G and profile F were consid-
ered the most attractive by orthodontists and laypeople,
respectively. Both groups considered the results of
mandibular advancement therapy (profiles E, F, and G)
significantly more attractive than camouflage therapy
(profiles A, B, and C). In addition, all camouflage therapy
profiles (A, B, and C) were rated significantly better than
the baseline (profile D). Orthodontists and laypeople
scored profile A with the largest nasolabial angle, signif-
icantly more attractive than the untreated photograph
(profile D) but significantly less attractive than profile
B with a smaller nasolabial angle. Laypeople also scored
profile A significantly less attractive than profile C with

August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Kalin, Iskender, and Kuitert
Table II. Mean ranking scores of profile D by orthodontists graduated in different locations and mean ranking scores
for profiles C, D, and G by orthodontists according to years of experience
Years of experience
Profile \6 (n 5 6) 6-10 (n 5 6) 11-16 (n 5 10)
C 5.0 5.2 5.4
D 5.7 7** 6.5
G 2.8 1.5 1.6
ACTA, Academic Centre of Dentistry Amsterdam; RUG, University of Groningen; RUN, University of Nijmegen.
*Profile D was judged significantly (P \0.01) less attractive by orthodontists trained outside the Netherlands; **Orthodontists with 6 to 10 years of
experience rated profile D significantly (P \0.01) less attractive.
a smaller nasolabial angle. Compared with laypeople,
orthodontists considered profiles resulting from
mandibular advancement therapy (profiles E, F, and G)
more attractive and profiles with camouflage therapy
(profiles A, B, and C) less attractive. This difference was
significant for profiles C, D, F, and G (Table I).
The geographic location of postgraduate orthodontic
training and the years of experience seem to influence
the judgment as seen for profile D, which was judged
significantly less attractive by the orthodontists with
foreign postgraduate training than orthodontists with
Dutch postgraduate training (Table II). Profile D was
also judged significantly more unattractive by the ortho-
dontists with 6-10 years of practice experience (Table II).
There was no significant correlation between previ-
ous orthodontic treatment and judgment by laypeople.
Education has a limited influence on the judgment by
laypeople; higher educated people judged profile
E significantly more attractive (Table III).
DISCUSSION
The current study used a modified female Class II Di-
vision 1 image because laypeople tend to be more sensi-
tive to female profile changes,30 with the untreated
specimen in the center and the modifications on the
left and the right side, a method also used by Kuroda
et al,28 Soh et al,31 and H€onn et al.32 The same ranking
system was also carried out by Y€ uksel et al.22 These
methods are proven to be dependable, and a valid
form of assessment of facial attractiveness28,31 and the
ranking system was a useful tool to determine the rank
order of the profile pictures.29 In the current study, pro-
files were allocated with a simple ranking system as
pointed out by Y€ uksel et al,22 to support indecisive
laypeople in making a clear choice presuming that allo-
cation of profiles is easier by ranking than by VAS score.
However, the ranking system, as used in the current
study, is limited, being an approximate evaluation
comparing the shown profiles among each other. Some
very close scored profiles with the ranking system in
American Journal of Orthodontics and Dentofacial Orthopedics
279
Site of graduation
.16 (n 5 22) ACTA (n 5 18) RUG (n 5 9) RUN (n 5 7) Other (n 5 10)
4.9
6.4 6.3 6.7 5.4 7*
1.3
Table III. Mean ranking scores of profile E by
laypeople according to education
Education level n Profile E
High school 31 3.3
Intermediate vocational education 45 3.7
Higher vocational education 86 2.9*
*Laypeople with higher vocational education level rated profile E
significantly (P \0.01) more attractive than laypeople with an inter-
mediate vocational education or high school.
the current study might have received equal scores in
the VAS, meaning that they could have been appreciated
at the same level of beauty as other profiles. With the use
of the ranking system, the chosen method for the assess-
ment of the profiles in the current study, assessors are
“obliged” to give a different score to each profile when
judging the attractiveness.
In the current study, modifications to reduce the
Class II features of the profile were limited to extension
of the mandible or increase of the nasolabial angle
because the intention of the current study was to find
the exclusive effect of these 2 treatment amenities on
the observers. Therefore, changes in lip posture that
may occur during Class II correction were omitted.
Changes in other features than nasolabial angle and
chin-neck length might influence the judgment on
attractiveness by laypeople and lead away from the
main subject of the study. In the current study, these in-
fluences were also limited by using the same profile to
create a number of profiles to be assessed by the ob-
servers.22,28,31,32 Using only 1 modified female profile
in the assessment could be considered a limitation of
the current study. There might have been different re-
sults if a male profile was also added for assessment
because of the difference in the perception of esthetics.
In addition, judgment by male assessors could attribute
different scores with female Class II profiles compared
with male Class II profiles. This is an interesting concern
August 2021  Vol 160  Issue 2
280
which should be considered in future studies with larger
groups of assessors. Unfortunately, this issue was not
addressed in the current study because of problems in
conducting reliable statistics between 22 male and 22
female orthodontists, with very small subgroups that
would make precise conclusions difficult.
The opinion of laypeople on facial attractiveness is an
important factor when developing treatment goals for
patients with Class II Division 1 malocclusion. Previous
studies are controversial on the difference in the judg-
ment of facial attractiveness by orthodontists and
laypeople. Lines et al,33 Marchiori et al,34 and Peerlings
et al35 found no difference between the judgment of
laypeople and professionals. Kiekens et al,36 Kerr and
O’Donnell37 and Lundstr€ om et al38 found that profes-
sionals were more critical than laypeople. However,
Spyropoulos and Halazonetis,39 Tedesco et al,40 Phillips
et al,41,42 and Giddon et al43 found that orthodontists
were less critical than laypeople. In the current study,
the profile with the largest mandibular advancement
was given the highest valuation from the orthodontists,
whereas laypeople preferred the profile with 3 mm less
advancement (profile F). Orthodontists likely look more
accurately at profiles as a consequence of their training
and experience. Laypeople may also prefer less straight
profiles than orthodontists.28,31 This assumption is
confirmed by Kokich Jr et al,44 who found that laypeople
accept deviations from ideal that are considered unat-
tractive by orthodontists and dentists. This is consistent
with the current study in which the difference in ranking
between attractive profiles (with mandibular advance-
ment) and profiles with increased nasolabial angle (den-
toalveolar compensation) was larger in the judgment
from orthodontists than in the assessment by laypeople
(Table I). It is likely that orthodontists show more
appraisal for, or observe better, the difference between
the effects on the profile of mandibular advancement
and of dentoalveolar compensation than laypeople.
The smaller standard deviations for the orthodontists
than laypeople (Table I) show that professionals observe
more precisely than laypeople. The variation among the
orthodontists was not extremely large; consequently,
their comparatively small number (44) is probably suffi-
cient for a reliable comparison with laypeople.
Profile C, with the least increase in nasolabial angle,
was appreciated significantly more by laypeople than
by orthodontists. Professionals were more able to
observe a small increase in nasolabial angle, although
the untreated profile (D) got the lowest appreciation
by both laypeople and professionals. The orthodontists
gave a significantly higher ranking, recognizing more
accurately a severe Class II Division 1 malocclusion and
judging accordingly. However, laypeople were still able
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Kalin, Iskender, and Kuitert
to recognize the improvement between profile D and
C, preferring, like the orthodontists, a decrease of the in-
terlabial step by an increase in nasolabial angle over the
original Class II Division 1 profile. This finding is consis-
tent with the study by Y€ uksel et al,22 who concluded that
Australian laypeople appreciated a profile with a moder-
ately increased nasolabial angle more than a severe Class
II Division 1 profile. In the study of Y€uksel et al,22 and the
current study, the modified untreated profile (D) was
considered the least attractive and the profiles with the
mandibular advancement treatment the most attractive,
followed by the profiles with the camouflage treatment.
Bishara and Jakobsen45 and de Almeida-Pedrin et al46
studied the changes in facial profile treated by camou-
flage therapy, and Ng et al47 showed the change in facial
attractiveness after mandibular advancement therapy.
These 3 studies showed improvement of facial attrac-
tiveness by both treatments. In addition, Lo and Hunt-
er48 concluded that the retraction of maxillary incisors
increases the nasolabial angle and improves the facial
attractiveness despite the increase in nasolabial angle.
According to the current study, an enlarged nasolabial
angle is preferred over a large sagittal interlabial step.
Straighter profiles, reached by either maxillary incisor
retraction or mandibular advancement surgery, are
considered more attractive than convex or concave pro-
files.32,49,50
No significant difference was found for the appreci-
ation by laypeople and orthodontists between profiles F
and G (chin-neck lengths 54 mm and 57 mm, respec-
tively) and between profiles B and C (nasolabial angles
121 and 113 , respectively). Consequently, the
outcome of the current study allows for some variation
in the preferred nasolabial angles and mandibular
advancement in Class II Division 1 correction. Orthodon-
tists may consider these findings when developing treat-
ment goals for patients with Class II Division 1
malocclusion, realizing that their perception of an ideal
facial profile can differ from that of the patients.
An interesting result of the current study is that the
relatively large difference between profiles D and E
(6 mm) is reflected in the relative approval by orthodon-
tists (3.2) and laypeople (2.5), whereas the difference in
approval between profiles E and F with only 3 mm
advancement is proportionally smaller (1.16 for the or-
thodontists and 0.5 for the laypeople). For the gradually
increased nasolabial angle with 8 for each step, an in-
crease in approval for both groups is also gradual. This
finding seems to point out that most observers had a
keen eye for proportions.
In the current study, higher educated people
considered profile E (with less mandibular advance-
ment than profiles F and G) less attractive than other
Kalin, Iskender, and Kuitert
laypeople, giving the profile almost the same ranking
score as the orthodontist. Maybe higher education
stimulates the ability to better distinguish small differ-
ences or changes.
In the current study, the judgment by orthodontists
is influenced by the site of graduate training and years
of experience. The untreated profile (D) was judged
significantly less attractive by orthodontists with
6-10 years of experience than colleagues with less or
more years of experience. This might originate from
the emphasis given to Class II Division 1 diagnosis
and treatment planning in adults in the period of their
graduate training. Profile D was also judged signifi-
cantly less attractive by orthodontists trained outside
the Netherlands. This may also be related to the grad-
uate programs in the United States, where most of
them were trained and where more straight profiles
are more likely preferred and involved in treatment
planning.
The current study was limited to the orthodontists
attending the same national orthodontic congress in
the Netherlands. Other studies also show a variation in
preference for different profiles in different countries.
In Brazil, the study of Marchiori et al34 found that ortho-
dontists and laypeople judged a retrusive mandibula less
attractive, but the nasolabial angle was not related to the
judgment of facial attractiveness. However, in the study
of Marchiori et al34 there was no emphasis on Class II Di-
vision 1 correction, and the variation in the judgment of
attractiveness among orthodontists and laypeople and
between orthodontists and laypeople was larger than
in the current study. In Iran, the study of Khosravanifard
et al51 showed that both male and female laypeople
preferred straighter profiles and higher nose tips. Ac-
cording to the study of Soh et al31 orthodontists and
oral surgeons in China ranked both a normal Chinese
profile and a bimaxillary retrusion to be esthetically
pleasant.
CONCLUSIONS
A difference in the assessment of attractiveness be-
tween orthodontists and laypeople was found when
judging female profile modifications of Class II Division
1 malocclusion with mandibular retrognathia.
1. The untreated baseline profile was judged the least
attractive by both laypeople and orthodontists.
2. Orthodontists rank a straighter profile more attrac-
tive than laypeople.
3. Orthodontists judge an untreated Class II Division 1
profile as less attractive than laypeople.
4. Orthodontists and laypeople appreciate both the re-
sults of the camouflage and mandibular advance-
American Journal of Orthodontics and Dentofacial Orthopedics
281
ment therapy more than the untreated profile,
with a preference for mandibular advancement.
5. According to both laypeople and orthodontists, the
most attractive nasolabial angle can vary between
113 and 121 .
6. According to both laypeople and orthodontists, the
most attractive chin-neck length can vary between
54 and 57 mm.
AUTHOR CREDIT STATEMENT
Kadriye Kalin contributed to conceptualization, vali-
dation, original draft preparation, visualization, and
project administration; Sebahat Yesim Iskender contrib-
uted to methodology, software, formal analysis investi-
gation, and data curation; and Reinder Kuitert
contributed to resources and manuscript review and
editing.
REFERENCES
1. Dion KK. Young children’s stereotyping of facial attractiveness.
Dev Psychol 1973;9:183-8.
2. Watkins LM, Johnston L. Screening job applicants: the impact of
physical attractiveness and application quality. Int J Select Assess
2000;8:76-84.
3. Langlois JH, Kalakanis L, Rubenstein AJ, Larson A, Hallam M,
Smoot M. Maxims or myths of beauty? A meta-analytic and theo-
retical review. Psychol Bull 2000;126:390-423.
4. Shahani-Denning C. Physical attractiveness bias in hiring: what is
beautiful is good. New York: Hofstra Horizon; 2003. p. 14-7.
5. Cox NH, van der Linden FP. Facial harmony. Am J Orthod 1971;60:
175-83.
6. Foster EJ. Profile preferences among diversified groups. Angle Or-
thod 1973;43:34-40.
7. Romani KL, Agahi F, Nanda R, Zernik JH. Evaluation of horizontal
and vertical differences in facial profiles by orthodontists and lay
people. Angle Orthod 1993;63:175-82.
8. Baldwin DC. Appearance and aesthetics in oral health. Community
Dent Oral Epidemiol 1980;8:244-56.
9. Kilpel€ainen PV, Phillips C, Tulloch JF. Anterior tooth position and
motivation for early treatment. Angle Orthod 1993;63:171-4.
10. Laufer D, Glick D, Gutman D, Sharon A. Patient motivation and
response to surgical correction of prognathism. Oral Surg Oral
Med Oral Pathol 1976;41:309-13.
11. Jacobson A. Psychological aspects of dentofacial esthetics and or-
thognathic surgery. Angle Orthod 1984;54:18-35.
12. Flanary CM, Barnwell GM Jr, Alexander JM. Patient perceptions of
orthognathic surgery. Am J Orthod 1985;88:137-45.
13. Mueser KT, Grau BW, Sussman S, Rosen AJ. You’re only as pretty as
you feel: facial expression as a determinant of physical attractive-
ness. J Pers Soc Psychol 1984;46:469-78.
14. Riggio RE, Widaman KF, Tucker JS, Salinas C. Beauty is more than
skin deep: components of attractiveness. Basic Appl Soc Psych
1991;12:423-39.
15. Prahl-Andersen B, Kowalski CJ. A mixed longitudinal, interdisci-
plinary study of the growth and development of Dutch children.
Growth 1973;37:281-95.
16. Fr€ankel R. The treatment of Class II, Division 1 malocclusion with
functional correctors. Am J Orthod 1969;55:265-75.
August 2021  Vol 160  Issue 2
282
17. Janson I. Skeletal and dentoalveolar changes in patients treated
with a bionator during prepubertal and pubertal growth. In:
McNamara JA Jr, Ribbens KA, Howe RP, editors. Clinical Alter-
ation of the Growing Face. Ann Arbor: The University of
Michigan; 1983.
18. McNamara JA Jr, Bookstein FL, Shaughnessy TG. Skeletal and
dental changes following functional regulator therapy on Class II
patients. Am J Orthod 1985;88:91-110.
19. Johnston LE Jr. A comparative analysis of Class II treatments: a
retrospective/prospective alternative. Clin Orthod Res 1998;1:
142-6.
20. Talass MF, Talass L, Baker RC. Soft-tissue profile changes resulting
from retraction of maxillary incisors. Am J Orthod Dentofacial Or-
thop 1987;91:385-94.
21. Worms FW, Isaacson RJ, Speidel TM. Surgical orthodontic treat-
ment planning: profile analysis and mandibular surgery. Angle Or-
thod 1976;46:1-25.
22. Y€uksel AG, Iskender SY, Kuitert R, Papadopoulou AK, Dalci K,
Darendeliler MA, et al. Differences in attractiveness comparing fe-
male profile modifications of Class II Division 1 malocclusion. Am J
Orthod Dentofacial Orthop 2017;152:471-6.
23. Farkas LG. Anthropometry of the Head and Face in Medicine. New
York: Elsevier/North Holland Inc; 1981.
24. Athanasiou AE. Orthodontic Cephalometry. Londen, United
Kingdom: Mosby-Wolfe; 1995.
25. Lehman JA Jr. Soft-tissue manifestations of aesthetic defects of
the jaws: diagnosis and treatment. Clin Plast Surg 1987;14:
767-83.
26. Hayes RJ, Sarver DM, Jacobson A. The quantification of soft tissue
cervicomental changes after mandibular advancement surgery. Am
J Orthod Dentofacial Orthop 1994;105:383-91.
27. Sinno HH, Markarian MK, Ibrahim AMS, Lin SJ. The ideal
nasolabial angle in rhinoplasty: a preference analysis of the
general population. Plast Reconstr Surg 2014;134:201-
10.
28. Kuroda S, Sugahara T, Takabatake S, Taketa H, Ando R, Takano-
Yamamoto T. Influence of anteroposterior mandibular positions
on facial attractiveness in Japanese adults. Am J Orthod Dentofac
Orthop 2009;135:73-8.
29. Chong HT, Thea KW, Descallar J, Chen Y, Dalci O, Wong R, et al.
Comparison of white and Chinese perception of esthetic Chinese
lip position. Angle Orthod 2014;84:246-53.
30. Burcal RG, Laskin DM, Sperry TP. Recognition of profile change af-
ter simulated orthognathic surgery. J Oral Maxillofac Surg 1987;
45:666-70.
31. Soh J, Chew MT, Wong HB. Professional assessment of facial pro-
file attractiveness. Am J Orthod Dentofacial Orthop 2005;128:
201-5.
32. H€onn M, Dietz K, Eiselt ML, G€oz G. Attractiveness of facial profiles
as rated by individuals with different levels of education. J Orofac
Orthop 2008;69:20-30.
33. Lines PA, Lines RR, Lines CA. Profilemetrics and facial esthetics.
Am J Orthod 1978;73:648-57.
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Kalin, Iskender, and Kuitert
34. Marchiori GE, Sodre LO, da Cunha TCR, Torres FC, Rosario HD,
Paranhos LR. Pleasantness of facial profile and its correlation
with soft tissue cephalometric parameters: perception of ortho-
dontists and lay people. Eur J Dent 2015;9:352-5.
35. Peerlings RH, Kuijpers-Jagtman AM, Hoeksma JB. A photo-
graphic scale to measure facial aesthetics. Eur J Orthod
1995;17:101-9.
36. Kiekens RMA, Maltha JC, van ’t Hof MA, Kuijpers-Jagtman AM. A
measuring system for facial aesthetics in Caucasian adolescents:
reproducibility and validity. Eur J Orthod 2005;27:579-84.
37. Kerr WJ, O’Donnell JM. Panel perception of facial attractiveness.
Br J Orthod 1990;17:299-304.
38. Lundstr€om A, Woodside DG, Popovich F. Panel assessments of
facial profile related to mandibular growth direction. Eur J Orthod
1987;9:271-8.
39. Spyropoulos MN, Halazonetis DJ. Significance of the soft tissue
profile on facial esthetics. Am J Orthod Dentofacial Orthop
2001;119:464-71.
40. Tedesco LA, Albino JE, Cunat JJ, Green LJ, Lewis EA, Slakter MJ. A
dental-facial attractiveness scale. Part I. Reliability and validity. Am
J Orthod 1983;83:38-43.
41. Phillips C, Trentini CJ, Douvartzidis N. The effect of treatment on
facial attractiveness. J Oral Maxillofac Surg 1992;50:590-4.
42. Phillips C, Tulloch C, Dann C. Rating of facial attractiveness. Com-
munity Dent Oral Epidemiol 1992;20:214-20.
43. Giddon DB, Sconzo R, Kinchen JA, Evans CA. Quantitative com-
parison of computerized discrete and animated profile preferences.
Angle Orthod 1996;66:441-8.
44. Kokich VO Jr, Kiyak HA, Shapiro PA. Comparing the perception of
dentists and lay people to altered dental esthetics. J Esthet Dent
1999;11:311-24.
45. Bishara SE, Jakobsen JR. Profile changes in patients treated with
and without extractions: assessments by lay people. Am J Orthod
Dentofacial Orthop 1997;112:639-44.
46. de Almeida-Pedrin RR, Guimar~aes LBM, de Almeida MR, de
Almeida RR, Ferreira FPC. Assessment of facial profile changes in
patients treated with maxillary premolar extractions. Dental Press
J Orthod 2012;17:131-7.
47. Ng D, De Silva RK, Smit R, De Silva H, Farella M. Facial attractive-
ness of skeletal Class II patients before and after mandibular
advancement surgery as perceived by people with different back-
grounds. Eur J Orthod 2013;35:515-20.
48. Lo FD, Hunter WS. Changes in nasolabial angle related to maxillary
incisor retraction. Am J Orthod 1982;82:384-91.
49. H€onn M, Dietz K, Godt A, G€ oz G. Perceived relative attractiveness
of facial profiles with varying degrees of skeletal anomalies. J Or-
ofac Orthop 2005;66:187-96.
50. Hambleton RS. The soft-tissue covering of the skeletal face as
related to orthodontic problems. Am J Orthod 1964;50:
405-20.
51. Khosravanifard B, Rakhshan V, Raeesi E. Factors influencing
attractiveness of soft tissue profile. Oral Surg Oral Med Oral Pathol
Oral Radiol 2013;115:29-37.
ORIGINAL ARTICLE

The influence of incisor inclination and

anterior vertical facial height on facial
attractiveness in an Asian woman
Umair Shoukat Ali,a Rashna Hoshang Sukhia,a Mubassar Fida,a Adeel Tahir Kamal,a and Ausjah Abbasb
Karachi, Pakistan

Introduction: Orthodontic treatment can help improve facial attractiveness through the modification of factors
affecting the soft tissue profile. The objectives of this study were to determine the impact of different maxillary
incisal inclinations and lower anterior facial heights (at rest and with a smile) on the facial attractiveness of an
Asian woman, as perceived by different panels of raters using visual analog scale (VAS) scores. Methods: A
cross-sectional study was performed with 66 raters equally divided into 3 panels that consisted of general
dentists, orthodontic residents, and laypersons. Raters assessed modified photographs of a subject with
various incisor inclinations and lower anterior facial height/total anterior facial height (LAFH/TAFH) on lateral
profile view. Modifications were made using Photoshop software (Adobe, San Jose, Calif). Subjective
evaluations of facial attractiveness were performed by raters using VAS. Kruskal-Wallis test was used to
compare VAS scores among raters. Mann-Whitney U test was applied to compare VAS scores between
groups. Results: Significant differences in VAS scores were found among raters for
10 (P 5 0.004) and

15 (P 5 0.021) incisal inclinations. Significant differences were found in VAS scores for
8% LAFH/TAFH
(P 5 0.044) and 4% LAFH/TAFH with smile (P 5 0.002). Conclusions: Professionals preferred normal incisal
inclinations to be the most attractive. General dentists found reduced facial height to be unattractive. Orthodontic
residents and laypersons considered increased LAFH/TAFH to be most unattractive. Smile had a negative
impact on VAS scores at extreme anterior facial height modifications. (Am J Orthod Dentofacial Orthop
2021;160:283-91)

O
ne of the important reasons patients seek ortho-
dontic care is to improve their facial attractive-
ness. Technological advancements such as
computer simulation software that predicts treatment
outcomes help orthodontists develop customized treat-
ment plans that best serve orthodontic patients.1 The re-
sults of treatment are considered to be successful when
they satisfy the patients and the societal perceptions of
beauty.2 Perception has been defined as the process by
which patterns of environmental stimuli are organized
and interpreted; it can be influenced by a variety of
a
Section of Dentistry (Orthodontics), Department of Surgery, The Aga Khan
University and Hospital, Karachi, Pakistan.
b
Karachi Medical and Dental College, Karachi, Pakistan.
All authors have completed and submitted the ICMJE Form for Disclosure of
Potential Conflicts of Interest, and none were reported.
Address correspondence to: Rashna Hoshang Sukhia, Section of Dentistry
(Orthodontics), Department of Surgery, The Aga Khan University and Hospital,
PO Box 3500, Stadium Road, Karachi 74800, Pakistan; e-mail, rashna_aga@
yahoo.com.
Submitted, September 2019; revised, March 2020; accepted, April 2020.
0889-5406/$36.00
Ó 2021 by the American Association of Orthodontists. All rights reserved.
https://doi.org/10.1016/j.ajodo.2020.04.030
physical, physiological, and social factors.3 These per-
ceptions are culturally or even religiously determined
and are transmitted to the masses via different formats,
such as the media, which has established esthetic stan-
dards and has had a direct effect on the social and per-
sonal preferences related to facial esthetics and smile
pleasantness.4,5 However, professional opinions some-
times do not coincide with the perceptions and expecta-
tions of laypersons.6
Orthodontic treatment targets the dentition and the
maxillomandibular relationships to create a considerable
impact on facial esthetics. Facial harmony must be
achieved in both the frontal and lateral aspects of the
face.7 Orthodontic treatment that relies on cephalomet-
rics to determine esthetic goals might not be able to pro-
duce the desired results.8 For example, the display of the
maxillary dentition during rest and smiling is considered
youthful and esthetically pleasing. Because the soft tis-
sue esthetics of the face (lips, nose, and chin) depend on
the underlying skeletal and dental support, an unfavor-
able change in the incisor position can lead to poor soft
tissue outcome.9 This reflects the importance of incisor
inclination on facial esthetics and should be a critical
283
284
component of treatment planning.10,11 The vertical
dimension of the face is equally important for the
orthodontist; controlling the vertical dimension is a
predictor of treatment success.12 In growing patients,
this vertical growth tendency can be managed via
different extraoral or intraoral devices such as high
pull headgears and high bite blocks.1 Surgical
intervention is required for correction of severe vertical
deformity in adults.13
Ideal smile characteristics appreciated from the fron-
tal view include an adequate incisal display and minimal
buccal corridor display.14 However, incisor inclinations
and facial heights can be best appreciated from the
lateral profile view.1 The objectives of this study were
to determine the impact of different maxillary incisal in-
clinations and lower anterior facial heights (at rest and
with a smile) on the facial attractiveness of an adult
Asian female as perceived by different panels of raters
using visual analog scale (VAS) scores. The null hypoth-
esis was that there is no impact of various incisal inclina-
tions and lower anterior facial heights (at rest and with
smile) on facial attractiveness as perceived by the
different panels of raters.
MATERIAL AND METHODS
A cross-sectional study was performed after obtain-
ing ethical approval from the institutional ethical review
board (no. 2019-1619-4175). The sample size was calcu-
lated by Open-Epi software using the findings of Ghaleb
et al,15 who reported the mean VAS scores of 10 by or-
thodontists and laypersons to be 62.23 6 13.62 and
49.99 6 14.84, respectively. The power of the study
was set at 80% with a confidence interval of 95%. It
was calculated that at least 22 subjects were required in
each group. Because we had 3 different panels of raters,
the total sample size was 66. Three different panels of
raters equally divided into general dentists (GD), ortho-
dontic residents (OR), and laypersons (LP) evaluated a se-
ries of images that showed changes in incisal inclinations
and vertical facial heights. Female subject aged 18 years,
who signed the informed consent, was included in this
study on the basis of the following inclusion criteria: a
clinical examination showed an orthognatic profile
with a pleasant smile in both frontal and profile views,
lateral cephalometric examination showed a Class I skel-
etal base (ANB, 2 6 2), maxillary incisor to sella-nasion
angle (UI/SN, 102 6 5) was well-positioned within the
cephalometric standard range, lower anterior facial
height/total anterior facial height (LAFH/TAFH) was
55% according to Eastman cephalometrics, and the sub-
ject had a Class I dental relationships with ideal overjet
and overbite. The exclusion criteria were any
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Shoukat Ali et al
craniofacial/dental anomaly or syndrome or a history of
trauma or surgery involving facial structures.
The clinician helped the patient assume the natural
head position (esthetic position) by a method recom-
mended by Bass16 so that the face was not tilted up or
down. This is an easy, reliable, and replicable way to
obtain profile pictures in clinical practice.15 Two profile
pictures were taken 1.5 meters away from the subject us-
ing a Nikon camera D3500 (24.2 million megapixels; Ni-
kon, Tokyo, Japan) with the subjects’ head fixed in the
natural head position. One photograph was in a normal
resting posture with relaxed lips, and the other one was
with a social smile that exposed the distal aspect of the
canine. Photographs were modified using an image ed-
iting software (Adobe Photoshop CS [version 8.0; Adobe,
San Jose, Calif] and Adobe Illustrator CS5 [version
15.0.1; Adobe]) according to the guidelines described
by Ghaleb et al15 and Devanna17 for incisal inclinations
alterations and angular measurements. To simulate the
changes in incisor inclinations, the crowns of the central
and lateral incisors were separately sliced using the edit-
ing software. Each tooth was considered as an individual
object with the center of rotation at the incisal edge. The
central incisor was superimposed from the tracing of the
lateral cephalograms, and the center of rotation was
placed at the incisal edge of the tooth. To maintain
the symmetry, the center of rotation of the lateral incisor
was set at the midpoint of the mesiodistal width. To
maintain the vertical positions of the maxillary incisors,
horizontal lines were drawn as tangents to the incisal
edges of the teeth, and vertical tangents were drawn
medial to the maxillary canines as the distal limit for
sagittal repositioning of the lateral incisor. Seven final
images were obtained for the female subjects (1 unal-
tered, 3 lingual, and 3 labial). This resulted in 4 cate-
gories according to the degree of proclination or
retroclination (ie, normal, mild, moderate, and severe
proclination and retroclination) (Fig 1). Artistic touches
were given when necessary to maintain the natural
appearance of the face.
The esthetic horizontal line (Hr) was taken as a stable
reference that is not altered by orthognathic or orthope-
dic treatment. The Sn-Pg0 line joins the subnasale point
(the deepest point at the junction where the profile of
the nose joins the upper lip) and the facial pogonion
(the most prominent point on soft tissue chin). This rep-
resents the lower facial third (ie, nearest reference part of
the face to the incisors) (Fig 2).
The following steps were carried out to obtain
angular measurements: (1) drawing of the line Sn-Pg0
and Hr passing through the mid-third; (2) determining
the most anterior point on the labial surface of the
maxillary incisor (obtained by the intersection of this
Shoukat Ali et al 285

Fig 1. Normal incisor inclination and 6 modifications in labial and lingual directions, with changes in
increments of 5 .

surface with the vertical tangent to the most labial Table I. Angular measurements of incisor inclination
aspect of the maxillary incisor); and (3) a tangent (Tg) of the face, for various inclinations
passing through this point resulted in 2 angular mea-
surements for each inclination: Tg/Hr is the angle Photographs incisal
inclinations,  Angle Tg/Hr,  Angle Tg/Sn-Pg0 , 
formed between incisor inclination and esthetic hori-

15 76.10
9.23
zontal, and Tg/Sn-Pg0 is the angle formed between
10 78.40
7.80
incisor inclination and lower facial third. A positive value
5 80.20
6.45
is given to the angle when the tangent is forward and a 0 85.40
4.50
negative value when the tangent is backward. 5 88.60
2.20
The angular measurements for each of the modified 10 99.70 13.50
15 113.00 26.41
incisal inclinations are given in Table I.
The anterior vertical dimension of the facial profile of Tg/Hr, the angle between incisor inclination and aesthetic horizon-
the female subject was altered by stretching and tal; Tg/Sn-Pg0 , the angle between incisor inclination and lower facial
third (positive value when the tangent is forward and negative value
depressing the image at the soft tissue points subnasale
when the tangent is backward).
and sublabiale. The LAFH/TAFH ratio of 55% was

American Journal of Orthodontics and Dentofacial Orthopedics August 2021  Vol 160  Issue 2
286
Fig 2. Angular measurements for the incisor inclination
modifications.
Fig 3. Normal LAFH/TAFH ratio at rest, and 4 modifications with 4% incremental increases (59% and
63%) and decreases (51% and 47%).
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Shoukat Ali et al
increased and decreased by 4%.12 The modified LAFH/
TAFH heights were 2 standard deviation away from the
Eastman cephalometrics standard range (55% 6 2%).
These alterations produced fewer pictures with appre-
ciable differences in facial heights for the raters to score
and judge accurately. The soft tissue contours above the
columella and below the soft tissue pogonion were not
altered and were identical for all images.18 The altered
LAFH/TAFH generated 5 lateral profile images of the fe-
male subject (47%, 51%, 55%, 59%, and 63%) at rest
and 5 lateral profile pictures on smiling, (Figs 3 and 4)
whereas the anteroposterior proportions were kept con-
stant (Class I).
The altered images were shown to the different panels
of raters on a Powerpoint presentation. These raters pro-
vided esthetic scores for the profile on the basis of facial
attractiveness using the VAS. This scale was created on
a 100 mm interrupted line anchored at 0 on the left
(very unattractive) and 10 on the right (very attractive).
Shoukat Ali et al
Fig 4. Normal LAFH/TAFH ratio with smile and 4 modifications with 4% incremental increases (59%
and 63%) and decreases (51% and 47%).
The following parameters were evaluated by the raters: (1)
impact of different incisal inclinations on facial attractive-
ness, (2) impact of different vertical anterior facial heights
without smile, and (3) impact of different vertical anterior
facial heights with smile.
Intraexaminer reliability was assessed by incorpo-
rating a duplicate image of one of the altered profiles
for each of the original test images. The raters were un-
aware of the duplicate image and were asked to score it
as an additional image.12 VAS scores were compared by
applying the intraclass correlation coefficient between
both the original and duplicate pictures, which showed
a good agreement (0.83).
Statistical analysis
Data analysis was performed using SPSS (version
19.0; SPSS, Chicago, Ill). Descriptive statistics (ie, mean
and standard deviation) were calculated for the age of
American Journal of Orthodontics and Dentofacial Orthopedics
287
raters, and median and interquartile range were calcu-
lated for the VAS scores from each category of incisor
inclinations and LAFH/TAFH altered groups. The
normality of data was determined by applying the
Shapiro-Wilk test, which yielded a nonnormal distribu-
tion. The Kruskal-Wallis test was used to determine sta-
tistically significant differences among panels of raters
for various incisor inclinations and LAFH/TAFHs. To
compare the VAS scores between groups of raters, the
Mann-Whitney U test was applied. The level of signifi-
cance was kept at P # 0.05.
RESULTS
The mean ages of the raters were 26.96 6 2.08 years
for OR, 30.59 6 ;6.39 years for GD, and 27.50 6 6.82
years for LP, whereas the gender distributions for our
raters were 8 males and 14 females (OR), 10 males and
12 females (GD), and 14 males and 8 females (LP).
August 2021  Vol 160  Issue 2
288 Shoukat Ali et al

Gender dimorphism within the raters is shown in LAFH/TAFH at rest, normal LAFH/TAFH was more
Table II. Female raters preferred normal inclinations attractive, and 8% was perceived to be least attractive
and found 15 to be displeasing. They found
4% (Table IV). When smiling, normal LAFH/TAFH was
LAFH/TAFH to be attractive and
8% to be the most deemed to be more attractive, whereas 8% was found
unattractive at rest. With smile, they found normal to be least attractive (Table V).
LAFH/TAFH to be the most attractive profile and 8% LP gave the highest scores to normal incisal inclina-
to be the most unattractive. tion and the lowest to 15 . (Table III) Normal LAFH/
Male raters were in agreement with female raters on TAFH was considered as the most attractive profile at
the most attractive (normal inclination) and unattractive rest (Table IV) and with smile. However, they rated the
(15 ) incisal inclinations. Similarly, a consensus was lowest to 8% LAFH/TAFH at rest and with smile
reached with respect to the profile attractiveness of (Table V).
normal LAFH/TAFH (55%) with smile. When rating the Comparison of the median VAS score among the 3
images with altered LAFH/TAFH at rest, the most attrac- panels of raters showed significant differences in
10
tive profile was
4%, and the most unattractive profile (P 5 0.004) and
15 (P 5 0.021) (Table VI). For
was 8%. The differences in the esthetic scores between
10 incisal inclinations, GD scored the lowest. For
males and females were statistically nonsignificant.
15 , OR scored the lowest. A comparison of VAS scores
GD gave the highest scores to normal incisal inclina- among raters showed a nonsignificant difference in
tion and the lowest to 15 (Table III). With respect to LAFH/TAFH without smile. Significant differences
LAFH/TAFH on the photograph at rest, GD preferred were found among raters VAS scores for
8% LAFH/

4% and gave the lowest scores to 8%. (Table IV) How- TAFH (P 5 0.044) and 4% LAFH/TAFH with smile
ever, for the lateral profile photograph with smile, the (P 5 0.003). For
8% LAFH/TAFH, the highest scores
GD preferred the normal LAFH/TAFH and gave the were given by LP. Similarly, LP scored highest for 4%
lowest scores to
8% (Table V). with smile (Table VI).
OR gave the highest scores to normal incisal inclina- Comparisons between each panel of rater revealed
tion and the lowest to
15 (Table III). With respect to significant differences for
10 and
15 incisal incli-
nations. Similarly, significant differences were found
for the
8% and 4% LAFH/TAFH ratio with smile be-
Table II. Gender dimorphism within raters tween raters (Table VII).
Male raters Female raters P
Variables (n 5 32) (n 5 32) value
DISCUSSION
Photographs incisor
inclinations,  The present study was conducted to determine the

15 5.00 (4.00-6.00) 4.50 (3.00-6.00) 0.970 impact of different maxillary incisal inclinations and

10 6.00 (4.25-7.00) 6.00 (4.00-7.00) 0.671

5 6.00 (5.00-7.00) 6.00 (4.00-7.25) 0.763
lower anterior facial heights (at rest and with smile) on
Normal 7.00 (5.25-8.75) 7.50 (5.00-8.00) 0.671 facial attractiveness of an adult Asian female as
5 6.00 (5.00-8.00) 6.50 (5.00-8.00) 0.629 perceived by different panels of raters using VAS scores.
10 6.00 (5.00-8.00) 6.00 (4.00-8.00) 0.965 In a recent study, only orthodontists and LP were asked
15 4.00 (2.00-5.00) 4.00 (2.00-5.00) 0.668 to evaluate subjects on the basis of facial attractive-
Photographs
(LAFH/TAFH
ness.19 The inclusion of GD as raters is crucial as they
at rest) are the first to assess patients and refer them to special-

8% (47%) 6.00 (4.00-7.00) 5.00 (3.00-6.25) 0.385 ists.20 The knowledge of their preferences will help

4% (51%) 6.00 (5.00-7.75) 6.00 (5.00-7.00) 0.935 generate a better zone of acceptability. In the past, crop-
Normal (55%) 6.00 (3.25-7.00) 6.00 (4.75-7.00) 0.767 ped images were used to assess facial esthetics, which
4% (59%) 5.00 (4.00-7.00) 5.00 (4.00-6.00) 0.805
8% (63%) 5.00 (3.00-6.75) 5.00 (3.75-6.25) 0.940
produced inconclusive or misleading results. Complete
Photographs profile photographs along with close-up dental views
(LAFH/TAFH were used to evaluate the facial attractiveness to over-
with smile) come these limitations.8

8% (47%) 4.00 (2.25-5.75) 4.00 (2.00-6.00) 0.954 In our study, all panels of raters provided the highest

4% (51%) 6.00 (3.25-7.00) 5.00 (3.00-6.00) 0.310
Normal (55%) 8.00 (7.00-8.75) 7.00 (6.00-8.00) 0.346
scores to normal incisal inclinations. These results are in
4% (59%) 6.00 (7.00-3.25) 5.00 (3.75-7.00) 0.903 concordance with Devanna,17 who also found normal
8% (63%) 3.00 (2.00-4.00) 4.00 (1.00-5.00) 0.770 inclinations to be the most attractive among the panel-
Note. Values are median (interquartile range); n 5 66. Independent ists. However, Ghaleb et al15 found that 5 are the most
t test: P #0.05; P \0.001. preferred incisal inclinations among orthodontists and

August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Shoukat Ali et al 289

Table III. Facial attractiveness scores with modified incisor inclinations by panels of raters
Photographs incisor inclinations,  GD (n 5 22) OR (n 5 22) LP (n 5 22)

15 4.00 (3.00-5.00) 3.50 (2.75-5.00) 5.50 (4.00-9.00)

10 5.00 (3.75-6.00) 6.00 (4.00-7.00) 7.00 (5.75-8.00)

5 5.5 (4.00-6.00) 7.00 (5.00-8.00) 6.00 (5.00-8.00)
0 6.50 (4.75-8.00) 8.00 (7.00-9.00) 7.00 (5.75-8.25)
5 6.00 (4.00-8.00) 6.50 (5.00-8.00) 6.50 (6.50-8.25)
10 6.00 (4.75-9.00) 6.00 (5.00-8.00) 6.00 (4.00-7.25)
15 2.5 (1.05-4.00) 4.50 (2.75-5.00) 4.00 (1.75-5.25)

Note. Values are median (interquartile range); n 5 66. Kruskal-Wallis test: P #0.05; P \0.001.

Table IV. Facial attractiveness scores of modified anterior facial heights at rest by panels of raters
Photographs (LAFH/TAFH at rest) GD (n 5 22) OR (n 5 22) LP (n 5 22)

8 (47%) 5.00 (3.75-6.00) 5.50 (4.00-7.25) 6.00 (3.75-7.00)

4 (51%) 6.00 (4.50-7.00) 6.00 (5.00-7.25) 6.00 (5.00-8.00)
Normal (55%) 5.00 (3.00-7.00) 6.00 (5.00-7.00) 6.50 (4.00-8.00)
4 (59%) 5.00 (3.75-6.25) 6.00 (3.75-6.25) 5.00 (4.00-7.00)
8 (63%) 5.00 (3.00-6.25) 5.50 (3.75-6.25) 6.00 (4.00-7.00)

Note. Values are median (interquartile range); n 5 66. Kruskal-Wallis test: P #0.05; P \0.001.

Table V. Facial attractiveness scores of modified anterior facial heights with smile by panels of raters
Photographs (LAFH/TAFH with smile) GD (n 5 22) OR (n 5 22) LP (n 5 22)

8% (47%) 2.50 (0.75-4.50) 4.00 (3.00-5.25) 4.50 (3.00-6.00)

4% (51%) 5.00 (3.75-6.00) 5.00 (3.00-6.25) 6.00 (4.75-7.00)
Normal (55%) 7.50 (4.75-8.00) 7.50 (5.75-8.00) 8.00 (7.00-9.00)
4% (59%) 6.00 (5.00-7.00) 6.00 (3.75-7.00) 7.00 (5.00-8.00)
8% (63%) 3.00 (1.00-6.00) 3.00 (2.00-4.00) 4.00 (2.00-6.00)

Note. Values are median (interquartile range); n 5 66. Kruskal-Wallis test: P #0.05; P \0.001.

GD, implicating that having slightly protrusive maxillary
incisors are considered to be more esthetic among pro-
fessionals.7 GP and LP gave the lowest score to 15 ,
whereas OR were displeased with
15 . The possibility
of OR being more lenient to 15 could be due to the
fact that OR favor greater inclinations of anterior teeth,
which provide better lip support and portray a youthful
appearance. In addition, they are aware of the conse-
quences of overretraction of maxillary incisors during
orthodontic treatment as this would result in a poor
soft tissue profile and aged appearance. Significant dif-
ferences were found between
15 and
10 among
the raters, with professionals providing the lowest scores
followed by the LP, further strengthening the fact that
more protrusive dental inclinations are believed to be
more esthetic. However, because no effects of retro-
clined incisors were shown on soft tissues in modified
pictures, LP may have failed to understand the impact
of retrusive incisors on the soft tissue profile. Ghaleb
et al15 has proposed the total facial concept and found
that an angle of 93 to the Hr and 7 to the lower facial
third to be the most esthetic incisal inclination. We
found that 85.4 to the Hr and
4.5 to the lower facial
third was the most preferred incisor inclination among
our raters.
A LAFH/TAFH of 55% was considered to be most
desirable among the various panels of raters. In a previ-
ous study, short facial heights were considered to be
more acceptable than longer facial heights on lateral
facial silhouettes.12 In a recent study, two thirds of a
sample of long-faced subjects were considered to be
pleasant by orthodontists and laypersons.19 In our sam-
ple of raters, disagreements were found among raters on
the most attractive profile without smile; however, it was
statistically insignificant. The most unattractive profile
without smile was 8% LAFH/TAFH (long face), which
is in agreement with Johnston et al.12 For evaluation
of the lateral profile photographs with smile, all raters
agreed on the normal LAFH/TAFH to be the most attrac-
tive. GD disagreed on the most unattractive profile on

American Journal of Orthodontics and Dentofacial Orthopedics August 2021  Vol 160  Issue 2
290
Table VI. Comparison of facial attractive scores among 3 groups of raters
Photographs GD (n 5 22)

10 6.00 (4.00-7.00)

15 3.50 (2.75-5.00)

8% LAFH/TAFH with smile 4.00 (3.00-5.25)
4 smile LAFH/TAFH with smile 6.00 (3.75-7.00)
Note. Values are median (interquartile range); n 5 66.
*Kruskal-Wallis test: P #0.05.
Table VII. Significant results from comparison be-
tween panel of raters for different variables
P value
Photographs GD vs OR LP vs OR
4 smile LAFH/TAFH 0.017* 0.001*
with smile

15 NS 0.49*

10 NS NS

8% LAFH/TAFH NS NS
with smile
Note. Values are median (interquartile range); n 5 66.
NS, Nonsignificant.
*Mann-Whitney U test: P #0.05.
smile with the other group of raters, but no significant
differences were found. OR and LP had an agreement
on the most attractive and unattractive profiles with
and without smile. These results are in disagreement to
different studies which concluded that LP perception
could be altered by various features such as hair, nose,
and chin.8,21 Significant differences were found between
raters when assessing 4% LAFH/TAFH with smile. OR
were more critical and provided much lower scores as
the values deviated more from the ideal LAFH/TAFH.
However, LP and GD found 4% LAFH/TAFH with smile
to be more acceptable. A significant disagreement was
found between GD and LP with the assessment of

8% LAFH/TAFH with smile in which LP found this to
be less unattractive. This portrays the variability of
acceptance among different groups.
Smile had a positive impact on facial attractiveness
when normal (55%), 4%, and
4% LAFH/TAFH were
scored. However, we found smile had a negative impact
on facial attractiveness with 8% and
8% LAFH/TAFH.
Considerably lower scores were provided by all panels of
raters to these extreme modifications, which indicated
that the profile was unattractive despite the smile.
The process of smiling involves various facial muscles
and causes elevation of the upper two thirds of the lower
face, including the upper lip.1 These dynamic changes
may have increased the prominence of the lower one
third of the face and aggravated the vertical proportion
of the face.
August 2021  Vol 160  Issue 2
Shoukat Ali et al
OR (n 5 22) LP (n 5 22) P value
5.00 (3.75-6.00) 7.00 (5.75-8.00) 0.004*
4.00 (3.00-5.00) 5.50 (4.00-9.00) 0.021*
2.50 (0.75-4.50) 4.50 (3.00-6.00) 0.044*
6.00 (5.00,7.00) 7.00 (5.00-8.00) 0.002*
A limitation of our study was that no comparison for
facial attractiveness between genders was performed.
Similarly, the impact of smile on anterior vertical heights
should also be assessed from the frontal view. Extreme
GD vs LP
modifications may have had unnatural distortion of the
NS face, which is a limitation of the Photoshop software.
Facial features such as skin color, tone, and hairstyle might
0.010* have had an impact on the facial attractiveness scores. This
0.001* study was performed using pictures of a subject with Class
0.020*
I skeletal and dental relationships and a harmonious soft
tissue profile. Patients with skeletal Class II or III
malocclusions with dentoalveolar compensations may
have an altered upper lip projection and nasolabial angle.
Therefore, the applicability of the results to all classes of
malocclusion may be deemed inappropriate. No
previous work has been done on the impact of smile on
anterior vertical dimensions to our knowledge. We
recommend the assessment of these features in frontal
view by employing methods to prevent unnatural
distortion. Similarly, the impact of various maxillary incisal
inclination and LAFH/TAFH should be assessed in other
skeletal patterns.
CONCLUSIONS
This study was conducted to determine the impact of
different maxillary incisal inclinations and lower anterior
facial heights (at rest and with smile) on facial attractive-
ness of an adult Asian female as perceived by different
panels of raters using VAS scores, drawing the following
conclusions:
1. All groups of raters preferred normal incisal inclina-
tion to be the most attractive inclination
2. LP and OR found normal LAFH/TAFH to be the most
attractive and long facial height to be unpleasant
with and without smile
3. Gender of raters had no major influence on the
facial attractiveness scores
4. Raters provided lowest VAS scores to smiling
photographs on extreme modifications (8%
and
8%) of LAFH/TAFH ratio as compared with
at rest.
American Journal of Orthodontics and Dentofacial Orthopedics
Shoukat Ali et al
AUTHOR CREDIT STATEMENT
Umair Shoukat Ali contributed conceptualization,
methodology, software, original draft preparation,
investigation, and resources; Rashna Hoshang Sukhia
and Mubassar Fida contributed to manuscript review
and editing, supervision, and validation; Adeel Tahir Ka-
mal contributed to formal analysis, visualization, manu-
script review and editing; and Ausjah Abbas contributed
to data curation.
REFERENCES
1. Proffit WR, Fields HW Jr, Sarver DM. Contemporary Orthodontics.
5th ed. St Louis: Mosby; 2013.
2. Jheon AH, Oberoi S, Solem RC, Kapila S. Moving towards precision
orthodontics: an evolving paradigm shift in the planning and de-
livery of customized orthodontic therapy. Orthod Craniofac Res
2017;20(Suppl 1):106-13.
3. Karimi-Afshar M, Torabi M, Safarian F, Dehghan MA, Karimi-
Afshar M. Effect of orthodontic treatments on quality of life in ad-
olescents. J Oral Health Oral Epidemiol 2018;7:198-204.
4. Henrich J, Boyd R. The evolution of conformist transmission and
the emergence of between-group differences. Evol Hum Behav
1998;19:215-41.
5. Cao L, Zhang K, Bai D, Jing Y, Tian Y, Guo Y. Effect of maxillary
incisor labiolingual inclination and anteroposterior position on
smiling profile esthetics. Angle Orthod 2011;81:121-9.
6. Flores-Mir C, Silva E, Barriga MI, Lagravere MO, Major PW. Lay
person’s perception of smile aesthetics in dental and facial views.
J Orthod 2004;31:204-9: discussion 201.
7. Nimbalkar S, Oh YY, Mok RY, Tioh JY, Yew KJ, Patil PG. Smile
attractiveness related to buccal corridor space in 3 different facial
types: a perception of 3 ethnic groups of Malaysians. J Prosthet
Dent 2018;120:252-6.
8. Turley PK. Evolution of esthetic considerations in orthodontics.
Am J Orthod Dentofacial Orthop 2015;148:374-9.
American Journal of Orthodontics and Dentofacial Orthopedics
291
9. Sundareswaran S, Ramakrishnan R. The Facial Aesthetic index: an
additional tool for assessing treatment need. J Orthod Sci 2016;5:
57-63.
10. Khanum A, Prashantha GS, Mathew S, Naidu M, Kumar A. Extrac-
tion vs non extraction controversy: a review. J Dent Orofac Res
2018;14:41-8.
11. Resnick CM, Daniels KM, Vlahos M. Does Andrews facial analysis
predict esthetic sagittal maxillary position? Oral Surg Oral Med
Oral Pathol Oral Radiol 2018;125:376-81.
12. Johnston DJ, Hunt O, Johnston CD, Burden DJ, Stevenson M,
Hepper P. The influence of lower face vertical proportion on facial
attractiveness. Eur J Orthod 2005;27:349-54.
13. Sahu M, Chain S, Koul R, Goel P. A new diagnostic tool for facial
typing. J Indian Orthod Soc 2017;51:294-9.
14. Li H, Cao T, Zhou H, Hou Y. Lip position analysis of young women
with different skeletal patterns during posed smiling using 3-
dimensional stereophotogrammetry. Am J Orthod Dentofacial Or-
thop 2019;155:64-70.
15. Ghaleb N, Bouserhal J, Bassil-Nassif N. Aesthetic evaluation of
profile incisor inclination. Eur J Orthod 2011;33:228-35.
16. Bass NM. Measurement of the profile angle and the aesthetic anal-
ysis of the facial profile. J Orthod 2003;30:3-9.
17. Devanna R. Turning subjective into objective: profile smile percep-
tion of I2 (incisor inclination) and its impact on treatment plan-
ning. J Orthod Res 2013;1:27-32.
18. Abu Arqoub SH, Al-Khateeb SN. Perception of facial profile attrac-
tiveness of different antero-posterior and vertical proportions. Eur
J Orthod 2011;33:103-11.
19. da Silva Goulart M, Filho LC, Claudia de Castro Ferreira Conti A,
Almeida Pedrin RR, de Miranda Ladewig V, Cardoso MA.
Evaluation of facial esthetics in long-faced white Brazilian middle
school students. Am J Orthod Dentofacial Orthop 2019;155:
812-8.
20. Hall JF, Sohn W, McNamara JA. Why do dentists refer to specific
orthodontists? Angle Orthod 2009;79:5-11.
21. Kokich VO Jr, Kiyak HA, Shapiro PA. Comparing the perception of
dentists and lay people to altered dental esthetics. J Esthet Dent
1999;11:311-24.
August 2021  Vol 160  Issue 2
ORIGINAL ARTICLE

Biomechanical effect of selective

osteotomy and corticotomy on
orthodontic molar uprighting
Xin She,a Toru Deguchi,b Hai Yao,c and Jing Zhoud
Charleston and Clemson, SC, and Columbus, Ohio

Introduction: Uprighting mesially tipped molars is often a necessary step before implant placement. However,
the orthodontic treatment can be lengthy and discourage patients from choosing implant prostheses. Periodon-
tally accelerated osteogenic orthodontics is reported to facilitate molar movements. This study aimed to evaluate
the biomechanical effects of various corticotomy and osteotomy approaches on the uprighting of a mesially tip-
ped mandibular second molar in a 3-dimensional finite element analysis model. Methods: The initial tooth
displacement and periodontal ligament (PDL) strain in 9 finite element analysis models with various corticotomy
and osteotomy simulations were compared under 3 intended tooth movement scenarios: distal crown tipping,
mesial root movement with restraints, and mesial root movement without restraints. Results: Corticotomy or os-
teotomy approaches altered the tooth displacement and the PDL strain in all 3 intended molar uprighting sce-
narios. The 2 most extensive surgical approaches, the combined mesial and distal osteotomy with horizontal
corticotomy and the circumferential corticotomy at root apex level, resulted in increased tooth movement but
had a distinct impact on PDL strain. Conclusions: It was revealed that different combinations of corticotomy
and osteotomy had a biomechanical impact on orthodontic molar uprighting movements. (Am J Orthod
Dentofacial Orthop 2021;160:292-301)

M
olars facing a mesial edentulous space are
likely to tip mesially over a long time.1,2 Tipped
molars can cause occlusal and periodontal
problems such as overeruption of opposing teeth and
angular loss of an alveolar bone level. It is a common sit-
uation among adult patients seeking comprehensive or
limited orthodontic treatment before bridge or implant
restoration for the premature loss of permanent molars.
It is often necessary to upright the tipped molar ortho-
dontically to develop proper edentulous span and root
a
Clemson-MUSC Joint Bioengineering Program, Department of Bioengineering,
Clemson University, Clemson, SC.
b
Division of Orthodontics, College of Dentistry, Ohio State University, Columbus,
Ohio.
c
Clemson-MUSC Joint Bioengineering Program, Department of Bioengineering,
Clemson University, Clemson, SC; Department of Oral Health Sciences, James B.
Edwards College of Dental Medicine, Medical University of South Carolina,
Charleston, SC.
d
Private practice, and Department of Orthodontics, James B. Edwards College of
Dental Medicine, Medical University of South Carolina, Charleston, SC.
All authors have completed and submitted the ICMJE Form for Disclosure of Po-
tential Conflicts of Interest, and none were reported.
Address correspondence to: Jing Zhou, Department of Orthodontics, James B.
Edwards College of Dental Medicine, Medical University of South Carolina,
1493 Appling Dr, Mt Pleasant, SC 29464; e-mail, jingzhouortho@gmail.com.
Submitted, October 2019; revised, May 2020; accepted, June 2020.
0889-5406/$36.00
Ó 2021.
https://doi.org/10.1016/j.ajodo.2020.06.034
angles for the dental prosthesis. However, conventional
or miniscrew-assisted orthodontic treatment to upright
a severely tipped molar can be time-consuming,3 espe-
cially at the mandibular posterior region with thick
cortical bones.4 The lengthy treatment time might
discourage patients from accepting the adjunctive or-
thodontic or restorative plan.
Periodontally accelerated osteogenic orthodontics
were reported to facilitate orthodontic movement for
uprighting tipped molars or intruding elongated mo-
lars.5-8 In these limited data, the authors believed that
the treatment time was significantly reduced. In a
recent clinical case report,9 2 severely inclined
second molars were uprighted with different selective
osteotomy-assisted orthodontic treatment as part of
the implant site development. It was revealed that the
selective osteotomy facilitated the distal tipping move-
ment of a mandibular molar with reduced treatment
time. The biological effects of the corticotomy are
believed to be associated with the regional acceleratory
phenomenon (RAP),10,11 which is the local and transitory
demineralization and remineralization in the alveolar
bone during the wound healing period. The mechanical
effect of the surgical interruption on the continuity of
the alveolar bone and the en-bloc movement of the

292
She et al
bone segment is another theory of the accelerated tooth
movement12,13 that has not been well studied.
Finite element analysis (FEA) is a numeric technique
for simulating a mechanical process in a physical system,
such as in an orthodontic tooth movement model. It is a
valid study method to analyze the immediate tooth
movement and the deformation in the dentoalveolar
structures under certain mechanical loadings without
using animal or clinical samples.14,15 FEA models have
been applied to study the mechanical effect of cortico-
tomy- and osteotomy-assisted canine retraction16 and
molar uprighting.9 It was revealed in these studies that
the simulated surgical alternation of the alveolar bone
structure can affect the immediate tooth movement
and periodontal ligament (PDL) deformation.
Therefore, this pilot study is designed to evaluate the
immediate mechanical effects of various corticotomy
and osteotomy approaches on the uprighting of a mesi-
ally tipped mandibular molar in a 3-dimensional (3D)
FEA model. Clinically there are 2 ways to achieve molar
uprighting: distal crown movement and mesial root
movement. The distal crown movement can be achieved
with an uprighting spring delivering distal tipping and
intrusion force. The mesial root movement can be
achieved with a moment built in the spring, with or
without the wire cinched distal to the molar bracket. In
this study, 3 mechanical settings were designed for these
intended tooth movement scenarios: distal crown
tipping, mesial root movement with restraint, and mesial
root movement without restraint (a fixed pivoting point
for the moment, mimicking the effect of the distally
cinched wire). The immediate mechanical effects on
tooth movement and the PDL deformation of the force
systems under the 9 simulated surgical interventions of
corticotomy and osteotomy were studied. Although the
biological effects of the RAP cannot be replicated in
this FEA model, a better understanding of the mechan-
ical effects of various surgical interventions on ortho-
dontic tooth movement may provide insights for
developing future study models or clinical protocols
that maximize the biological benefits of the RAP effect.
MATERIAL AND METHODS
With appropriate institutional review board approval,
the 3D geometric model of a mandible was recon-
structed using Amira image processing software (version
5.4; Amira, Thermo Fisher Scientific, Waltham, Mass)
from an existing pretreatment cone-beam computed to-
mography scanning of a previous patient (Planmeca 3D
Max; Planmeca, Helsinki, Finland; voxel dimensions:
0.35 3 0.35 3 0.35 mm3). This patient is a healthy
46-year-old African American male with bilaterally
American Journal of Orthodontics and Dentofacial Orthopedics
293
missing first molars and mesially tipped second molars.
There were no craniofacial anomalies or history of
temporomandibular disorder reported from this patient.
The treatment for the tipped molars was described in a
previous case report.9 Briefly, the mandibular
second molars were uprighted with selective
osteotomy-assisted orthodontics in 5 months.
The 3D geometric models of the mandibular right
second molar, the mandibular cortical, and trabecular
bones were reconstructed through segmentation of
cone-beam computed tomography images using Amira
software (Fig 1). The PDL surrounding this molar was
generated using Geomagic Studio designing software
(version 12; 3D Systems, Rock Hill, SC) with an average
thickness of 0.2 mm17 (Fig 1). The contour of a
stainless-steel bracket was created at the center of the
buccal surface of the molar crown using Solidworks (Sol-
idworks 2016; Solidworks, V
elizy-Villacoublay, France)
to indicate the location of the applied orthodontic force
loadings (Fig 1).
Volumetric meshes were generated separately for the
geometric components, including the mandibular
cortical and trabecular bones, the molar, and the PDL
in preprocessing software Hypermesh (version 11.0;
Altair Engineering, Troy, Mich). The numbers of ele-
ments and nodes for each component are listed in the
Table. The whole meshed models were then imported
into FEA software Abaqus (version 6.14; Abaqus, Inc,
V
elizy-Villacoublay, France) for further analysis. The
Fig 1. Geometric model of a mandible with a mesially tip-
ped second molar. Transparent blue and orange, the
cortical and trabecular bones; solid blue, the tipped
second molar; red, PDL; red triangles, the boundary con-
ditions.
August 2021  Vol 160  Issue 2
294 She et al

Table. Material properties and element numbers

Geometric components Young's modulus (GPa) Poisson's ratio Element types Nodes (n) Elements (n) References
Cortical bone 10.7 0.3 C3D10 393612 255626 Aversa et al (2009)27
Trabecular bone 0.97 0.3 C3D10 264710 181355 Aversa et al (2009)27
Molar 20.7 0.3 C3D10 19895 13515 Lee and Baek (2012)28
PDL 0.05 0.45 C3D20R 54231 11665 Lin et al (2013)29

mandible, molar, and PDL were modeled to be an
isotropic linearly elastic material. The Young's modulus
and Poisson's ratio were based on previous studies listed
in the Table.
The FEA models were constrained with fixed-
displacement boundary conditions along the bilateral
mandibular sigmoid notches, the condylar heads, and
the coronoid processes to prohibit the free movement
of the mandible on force loading (Fig 1). To prevent sep-
aration between geometric models, interfaces including
mandibular cortical bone to trabecular bone, mandib-
ular cortical and trabecular bones to the PDL, and the
PDL to the tooth were constraint to be rigidly bonded.
Three mechanical settings were designed for these
intended tooth movement scenarios: distal crown
tipping, mesial root movement with restraint, and mesial
root movement without restraint. For the distal crown
movement, 1 N (about 102 g or 3.6 oz) of distalizing
force and 0.5 N (about 51 g or 1.8 oz) of intrusion force
were loaded (Fig 2, A). For the mesial root movement,
5 N$mm (about 510 g$mm) of mesial root movement
moment was loaded with and without a fixed pivoting
point, mimicking orthodontic uprighting springs with
and without cinching back at the distal end (Fig 2, B
and C). The forces and moments were directly loaded
at the center point of the buccal crown surface. The force
and moment values are the same as in a clinical case
report9 and within the range of force and moment values
described in previously published orthodontic molar up-
righting studies.18,19
Each type of orthodontic loadings was applied to 9
different simulated models (1 control and 8 surgical
models): (a) control; (b) mesial osteotomy; (c) distal os-
teotomy; (d) combined mesial and distal osteotomy; (e)
mesial osteotomy combined with buccal horizontal cor-
ticotomy 10 mm below cementoenamel junction (CEJ);
(f) distal osteotomy combined with buccal horizontal
corticotomy 10 mm below CEJ; (g) combined mesial
and distal osteotomy with buccal horizontal corticotomy
10 mm below CEJ; (h) circumferential corticotomy with
buccal-lingual cuts at 10 mm below CEJ; and (i) circum-
ferential corticotomy with buccal-lingual cuts at root
apex level, approximately 13 mm below CEJ.
Shear bands representing the periodontal surgical
approaches were created using computer-assisted
design software Solidworks. The osteotomy shear bands
were all placed approximately 1 mm away from the
molar root with the dimension of 2 mm wide and a depth
through the cortical and trabecular bone. The cortico-
tomy cuts were 1 mm wide and only through the cortical
bone. The occlusal and buccal views and 3D models of all
cutting patterns were shown in Figure 3.
The FEA was conducted on the basis of 27 simula-
tions (9 simulation models by 3 loading configurations).
Displacement of the molar in each simulation was
measured on a node-by-node basis. Contour plots of
the magnitude of molar displacement were generated
to show the tooth movement pattern. The maximum
displacement of the molar was determined to evaluate
the influence of different cutting approaches on tooth
movement. Vector plots of molar displacement for
representative models with distal osteotomy cut under
3 mechanical settings were also created to characterize
the direction of tooth movement. Correspondingly, the

Fig 2. Three types of orthodontic loads were applied to the center of the buccal surface of the molar
crown, where the bracket was bonded: A, combined 1 N (about 102 g or 3.6 oz) distal pushing force
and a 0.5 N (about 51 g or 1.8 oz) intrusion force; B, 5 N$mm (about 510 g$mm) counterclock moment;
C, 5 N$mm counterclock moment with a fixed pivot (red triangle).

August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
She et al 295

Fig 3. Occlusal and buccal views (row 1 and 2) and the 3D models (row 3) of the shear bands
mimicking the osteotomy and corticotomy cuts around the tipped molar. a, control; b, mesial osteotomy;
c, distal osteotomy; d, combined mesial and distal osteotomy; e, mesial osteotomy combined with
buccal horizontal corticotomy 10 mm below CEJ; f, distal osteotomy combined with buccal horizontal
corticotomy 10 mm below CEJ; g, combined mesial and distal osteotomy with buccal horizontal cortico-
tomy 10 mm below CEJ; h, circumferential corticotomy with buccal-lingual cuts at 10 mm below CEJ; i,
circumferential corticotomy with buccal-lingual cuts at root apex level, approximately 13 mm below
CEJ. The asterisk (*) at the distal side of model i indicates that the circumferential corticotomy cut
was at root apex level, different from model h. Solid black lines: 2 3 10-mm osteotomy cuts 1 mm
away from the roots; dashed lines: the 1 mm corticotomy cuts 1 mm away from the roots.

strain distribution and maximum strain in PDL sur-
rounding the molar were quantified in each simulation
and compared between different cutting approaches.
RESULTS
In the control model of distal crown movement
(Fig 4, A; model a), the contour plot showed an initial
displacement pattern comparable to a bodily movement
with a similar amount of displacement of crown and
roots in the mesiodistal direction. The mesial osteotomy
with or without buccal corticotomy (Fig 4, A; models b
and e) did not change the initial displacement pattern
much. The combined mesial and distal osteotomy
(model d), the distal osteotomy combined with midroot
corticotomy (model f), as well as the circumferential cor-
ticotomy at root apex level (model i) had a similar impact
on tooth displacement (Fig 4, B). However, the com-
bined mesial-distal osteotomy resulted in the least PDL
strain among these 3 approaches (Fig 4, C and D; model
d). The most tipping movement resulted from the com-
bined mesial-distal osteotomy with horizontal cortico-
tomy (model g), which also generated the most strain
in PDL (Fig 4, D).
In the scenario of mesial root movement without a
fixed pivoting point, all simulation models demon-
strated most displacement at the crown and least
displacement at the distal apex area (Fig 5, A). These
models indicated that the application of a moment
without cinching distal end wire might not result in
mesial root movement. The most tipping movement
resulted from the combined mesial-distal osteotomy
with horizontal corticotomy (Fig 5, A and B; model g).
This model also generated the most strain in PDL (Fig
5, C and D). The combined mesial and distal osteotomy
(model d) and the circumferential corticotomy at root
apex level (model i) showed considerable impact on
crown displacement with minimum strain in PDL (Fig
5, C and D).
The application of the mesial root movement
moment with a fixed pivoting point resulted in the
desired mesial root movement (Fig 6, A). The combined
mesial-distal osteotomy with horizontal corticotomy
(model g) resulted in most root displacement and PDL
strain around root apex (Fig 6, C and D). The combined
mesial-distal osteotomy (model d) had a similar impact
on tooth displacement with the circumferential cortico-
tomy at root apex level (model i), but a different influ-
ence on the strain in PDL (Fig 6). The circumferential
corticotomy at root apex level resulted in the least strain
in PDL among all surgical approaches (Fig 6, D).
The tooth movement patterns in the 3 different
loading conditions were summarized in Figure 7. The
vector plot arrows indicated the relative tooth movement
amount and directions. The center of resistance of the
distal crown tipping, mesial root movement with a piv-
oting point, and mesial root movement without a pivot-
ing point was displayed. The 3 different tooth movement
scenarios yielded effects of a wide range of magnitudes,
but the range was comparable to other FEA studies on
orthodontic tooth movements.16,20 To visually compare

American Journal of Orthodontics and Dentofacial Orthopedics August 2021  Vol 160  Issue 2
296 She et al

Fig 4. FEA results of the distal crown movement models: A, buccal view of the initial tooth displace-
ment patterns; B, comparison of the maximum initial displacements of the molar in the models; C, strain
distribution patterns in the molar PDL (mesial view) in the distal crown movement models; D, compar-
ison of the maximum strain in PDL.

the tooth movement patterns, we adjusted the value
range display in each scenario.
DISCUSSION
Currently, there are very limited studies on the
biomechanical effect of the surgical bony cuts of corti-
cotomy and osteotomy on orthodontic tooth movement.
The mechanical effect of 2 osteotomy approaches on
molar uprighting has been studied in a previous clinical
case report,9 which correspond to models b and c in the
current study. It was observed that the center of rotation
on the molar under distalizing and intrusive forces
changed with the performed distal osteotomy (Fig 8).
In the current study, 27 simulations of control and sur-
gical patterns for molar uprighting movements were
analyzed with FEA. It was demonstrated that the cortico-
tomy or osteotomy approaches have altered the tooth
displacement amount and patterns, as well as the PDL
strain in all 3 mechanical scenarios.
The center of resistance of the distal crown tipping,
mesial root movement without a fixed pivoting point,
and mesial root movement with a fixed pivoting point
are near the mesial root apex (Fig 7, A), near the distal
root apex (Fig 7, B), and at the crown center (Fig 7, C),
respectively. The 3 different tooth movement scenarios
yielded effects of a wide range of magnitudes (Fig 7,

August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
She et al 297

Fig 5. FEA results of the mesial root movement models without a fixed pivoting point: A, buccal view of
the initial tooth displacement patterns; B, comparison of the maximum initial displacements of the molar
in the models; C, strain distribution patterns in the molar PDL (distal view) in the distal crown movement
models; D, comparison of the maximum strain in PDL.

A-C). The maximum initial displacement of the root in
Figure 7, A (red area) cannot be meaningfully compared
with the maximum movement in the crowns in Figure 7,
A and B under the same scale. However, the values of the
root movement in Figure 7, B (blue area) and C (red
area) are comparable. This is in accordance with the clin-
ical observations that the mesial root movement seemed
to be much harder to achieve than the distal crown
tipping movement. In Figure 7, B and C, the root goes
up against more resistance from the dentoalveolar struc-
ture than the unrestrained dental crowns. On the same
force delivery system, the crown moves more than the
apex in an uncinched scenario even when the desired
movement was the mesial movement of the root. There-
fore, the application of a moment without cinching the
distal end of the wire may not result in mesial root move-
ment clinically. This might result in a slight eruption of
the molar crowns, as shown in the arrows pointing up
in the red areas in Figure 7, B, which is an undesired
side effect that will cause interference in occlusion or
opening of the bite.
The general trend seems to be that more extensive
surgical cuts resulted in more tooth displacement. The
circumferential corticotomy at root apex level (model i)
facilitated even more efficient initial tooth movements
than corticotomy at 10 mm below CEJ (model h) in all

American Journal of Orthodontics and Dentofacial Orthopedics August 2021  Vol 160  Issue 2
298 She et al

Fig 6. FEA results of the mesial root movement models with a fixed pivoting point: A, buccal view of the
initial tooth displacement patterns; B, comparison of the maximum initial displacements of the molar in
the models; C, strain distribution patterns in the molar PDL (distal view) in the distal crown movement
models; D, comparison of the maximum strain in PDL.

3 tooth movement scenarios (Figs 4-6). Combined
mesial and distal osteotomy (model g) resulted in more
crown or root movement in all 3 scenarios than
circumferential corticotomy (models h and i; Figs 4-6).
The extent of the surgical interruption of the
dentoalveolar structure continuity seemed to be
associated with the amount of resistance during tooth
movement. This trend is in agreement with the results
reported by Yang et al,16 the extent of the corticotomy
can affect the mechanical responses of dentoalveolar
structures.
It was also revealed that the position of the osteot-
omy or corticotomy could affect the immediate tooth
movement. In the mesial root movement scenarios,
the distal osteotomy models (models c and f; Figs 5
and 6) resulted in more root displacement than the
mesial osteotomy (models b and e; Figs 5 and 6),
even when the desired movement was the mesial move-
ment of the root. In a previously reported patient,9 it
was also shown that mesial osteotomy only minimally
improved mesial root movement. A possible explana-
tion for this effect is that the distal osteotomy released
the tissue constraint on the molar roots from the distal
dentoalveolar structures, especially from the thicker
distal cortical bone after the buccal oblique line in
the posterior mandible.

August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
She et al 299

Fig 7. Vector plots of molar models with circumferential corticotomy cuts under the distal crown move-
ment model (A), the mesial root movement moment loading without a fixed pivoting point (B), and with a
fixed pivoting point (C). White arrows, the direction of displacement of nodes throughout the molar;
large black arrows, the rotation pattern of the molar.

Fig 8. A clinical case report of a tipped molar was uprighted with osteotomy-assisted orthodontics.9 A,
Surgical procedures of the distal osteotomy. B, Mechanical settings for the molar uprighting. C, Clinical
observations of tooth movements in 5 months. D, The FEA vector plot of the distal crown movement
with osteotomy-assisted molar uprighting.

However, different surgical approaches may result in in the surgical patterns d, f, and i in the distal tipping
a similar impact on initial tooth movement. For example, scenario (Fig 4, A and B). In contrast, the same surgical
a similar impact on crown movement was demonstrated designs may have distinct effects in various tooth

American Journal of Orthodontics and Dentofacial Orthopedics August 2021  Vol 160  Issue 2
300
movement configurations. An example is the circumfer-
ential corticotomy at root apex level resulted in a notably
different amount of mesial root movement with or
without cinching back. Thus, the clinicians should also
consider the invasiveness and accessibility of the surgery
for different patients to minimize surgical risks and
complications.
Regarding the PDL strain, the 2 most extensive surgi-
cal approaches, models g and i resulted in increased
tooth movement but had a distinct impact on PDL strain.
The corticotomy models h and i resulted in relatively
lower PDL strain than most of the osteotomy models.
A low and evenly distributed strain in PDL is considered
to be a favorable biomechanical response,20 because
light and continuous orthodontic force lead to frontal
bone resorption.21 A surgical cut through the cortical
bone, referred to as corticision in the study by Peron
et al,22 might reduce the hyalinization during tooth
movement. Lower levels of initial stresses in PDL may
also relate to less external root resorption in orthodontic
tooth movement.23 These findings may provide indirect
support for the claim that corticotomy may result in
faster tooth movement and less root resorption.24
In contrast, the osteotomy approaches resulted in
increased PDL strain with increased initial tooth move-
ment. Biologically, a significantly increased PDL strain
may result in a cutoff of blood flow and cell death in
the compressed area.21 Therefore, the increased tooth
movement might be explained by the bone block move-
ment theory proposed by Kole13 or the bond-bending ef-
fect or a greenstick fracture of the alveolar bone under
orthodontic force.25 This theory is not contradictory to
the RAP theory proposed by Wilcko et al11 in 2001, which
is referred to as a phenomenon of local and transitory
demineralization and remineralization that was first
described by Frost10 in 1983. Instead, we believe although
the RAP might contribute to the accelerated tooth move-
ment during the later wound healing process, the en-bloc
movement of the bone segment might facilitate the initial
tooth movement, especially in the posterior mandible
area with thick and dense cortical bone.
This study was based on an FEA with a simulated or-
thodontic tooth movement model. There are several lim-
itations in the nature of a numerical technique compared
with a real physical system. First, the FEA is based on the
assumption that the material in the modeling system is
homogenous. However, bone, PDL, and teeth are com-
plex nonhomogenous structures that are simplified in
FEA to be adapted for calculations. Therefore, the study
results should be interpreted with caution with aware-
ness of the limitations of a geometric model, bearing
in mind that every finite element is based on an assumed
shape function expressing internal displacements as
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
She et al
functions of nodal displacements.26 Second, the fixed
pivoting point of the forces and moments in the FEA
model may not represent the exact clinical condition re-
sulted from a cinched distal end wire. The limitations of
the cinched angle, the wire flexibility, and the chewing
motion may cause the wire to travel slightly in the molar
tube. Therefore, an FEA model may not completely repli-
cate exact clinical force conditions. Finally, the FEA
model is used only to study the immediate mechanical
effect of the force system. It does not replicate the bio-
logical effects of RAP, which takes a brief period to be
initiated. If the dimension of time or in vivo animal
models can be developed in future study models, it
will provide more evidence of the clinical applications
of the surgery interference.
An individualized FEA for each patient's molar up-
righting procedure is not feasible with current technolo-
gies. However, the FE models we analyzed in this study
revealed the key mechanical effects of several common
surgical options for clinicians. These results could help
clinicians evaluate the optimal surgical plans for peri-
odontally accelerated osteogenic orthodontics on the
basis of tooth movement directions, patient dentoalveo-
lar ridge anatomy, accessibility, and side effects of the
surgery. Clinicians should always evaluate the extent of
the surgical injury and the benefit of reduced treatment
time for osteotomy and corticotomy-assisted orthodon-
tics. The possible complications and risks from the sur-
geries and the additional cost and possible pain for the
patients should always be discussed during treatment
presentation. The authors only recommend the careful
application of the surgical interventions to suitable
limited or adjunct patients because of the transitory
effect of the RAP effect.
CONCLUSIONS
It was demonstrated that different combinations of
corticotomy and osteotomy had an immediate biome-
chanical impact on orthodontic molar uprighting move-
ments. The 2 most extensive surgical approaches, the
combined mesial and distal osteotomy with horizontal
corticotomy and the circumferential corticotomy at
root apex level, resulted in increased tooth movement
but had distinct impact on PDL strain. These results
may provide insights for clinicians to design proper sur-
gical approaches for a certain type of mechanical setting
in molar uprighting procedures.
AUTHOR CREDIT STATEMENT
Xin She contributed to data collection and manu-
script preparation and editing; Toru Deguchi contrib-
uted to study design and manuscript preparation and
She et al
editing; Hai Yao contributed to study design and manu-
script preparation and editing; and Jing Zhou contrib-
uted to study design, orthodontic treatment, data
collection, and manuscript preparation.
ACKNOWLEDGMENTS
This study was supported by the Scholar Award from
the Medical University of South Carolina College of
Dental Medicine and National Institutes of Health grants
P20GM121342 and R01DE021134.
REFERENCES
1. Lindskog-Stokland B, Hakeberg M, Hansen K. Molar position asso-
ciated with a missing opposed and/or adjacent tooth: a follow up
study in women. Swed Dent J 2013;37:97-104.
2. Lindskog-Stokland B, Hansen K, Tomasi C, Hakeberg M,
Wennstr€ om JL. Changes in molar position associated with missing
opposed and/or adjacent tooth: a 12-year study in women. J Oral
Rehabil 2012;39:136-43.
3. Magkavali-Trikka P, Emmanouilidis G, Papadopoulos MA.
Mandibular molar uprighting using orthodontic miniscrew im-
plants: a systematic review. Prog Orthod 2018;19:1.
4. Holmes PB, Wolf BJ, Zhou J. A CBCT atlas of buccal cortical bone
thickness in interradicular spaces. Angle Orthod 2015;85:911-9.
5. Kim SH, Kook YA, Jeong DM, Lee W, Chung KR, Nelson G. Clinical
application of accelerated osteogenic orthodontics and partially
osseointegrated mini-implants for minor tooth movement. Am J
Orthod Dentofacial Orthop 2009;136:431-9.
6. Moon CH, Wee JU, Lee HS. Intrusion of overerupted molars by cor-
ticotomy and orthodontic skeletal anchorage. Angle Orthod 2007;
77:1119-25.
7. Oliveira DD, de Oliveira BF, de Ara
ujo Brito HH, de Souza MM,
Medeiros PJ. Selective alveolar corticotomy to intrude overerupted
molars. Am J Orthod Dentofacial Orthop 2008;133:902-8.
8. Wang CW, Chou MY, Chen R, Rowe T, Masoud M, Kim DM, et al.
Simultaneous ridge augmentation and accelerated molar upright-
ing for implant site development: two case reports with a split-
mouth design. Int J Periodontics Restorative Dent 2017;37:
423-30.
9. Zhou J, She X, Miller PD, Yao H. Selective osteotomy-assisted
molar uprighting and simultaneous ridge augmentation for
implant site development. Am J Orthod Dentofacial Orthop
2019;156:846-57.
10. Frost HM. The regional acceleratory phenomenon: a review. Henry
Ford Hosp Med J 1983;31:3-9.
11. Wilcko WM, Wilcko T, Bouquot JE, Ferguson DJ. Rapid orthodon-
tics with alveolar reshaping: two case reports of decrowding. Int J
Periodontics Restorative Dent 2001;21:9-19.
12. Chung KR, Oh MY, Ko SJ. Corticotomy-assisted orthodontics. J
Clin Orthod 2001;35:331-9.
13. Kole H. Surgical operations on the alveolar ridge to correct occlusal
abnormalities. Oral Surg Oral Med Oral Pathol 1959;12:277-288
contd.
American Journal of Orthodontics and Dentofacial Orthopedics
301
14. Kojima Y, Fukui H. A numerical simulation of tooth movement
by wire bending. Am J Orthod Dentofacial Orthop 2006;130:
452-9.
15. Kojima Y, Mizuno T, Fukui H. A numerical simulation of tooth
movement produced by molar uprighting spring. Am J Orthod
Dentofacial Orthop 2007;132:630-8.
16. Yang C, Wang C, Deng F, Fan Y. Biomechanical effects of cortico-
tomy approaches on dentoalveolar structures during canine retrac-
tion: a 3-dimensional finite element analysis. Am J Orthod
Dentofacial Orthop 2015;148:457-65.
17. Wang C, Han J, Li Q, Wang L, Fan Y. Simulation of bone remodel-
ling in orthodontic treatment. Comput Methods Biomech Biomed
Engin 2014;17:1042-50.
18. Largura LZ, Argenta MA, Sakima MT, Camargo ES, Guariza-Filho O,
Tanaka OM. Bone stress and strain after use of a miniplate for molar
protraction and uprighting: a 3-dimensional finite element anal-
ysis. Am J Orthod Dentofacial Orthop 2014;146:198-206.
19. Raveli TB, Raveli DB, de Mathias Almeida KC, Pinto ADS. Molar up-
righting: A considerable and safe decision to avoid prosthetic
treatment. Open Dent J 2017;11:466-75.
20. McCormack SW, Witzel U, Watson PJ, Fagan MJ, Gr€ oning F. The
biomechanical function of periodontal ligament fibres in ortho-
dontic tooth movement. PLoS One 2014;9:e102387.
21. Khouw FE, Goldhaber P. Changes in vasculature of the periodon-
tium associated with tooth movement in the rhesus monkey and
dog. Arch Oral Biol 1970;15:1125-32.
22. Peron APLM, Johann ACBR, Papalexiou V, Tanaka OM, Guariza-
Filho O, Ign
acio SA, et al. Tissue responses resulting from tooth
movement surgically assisted by corticotomy and corticision in
rats. Angle Orthod 2017;87:118-24.
23. Viecilli RF, Kar-Kuri MH, Varriale J, Budiman A, Janal M. Effects of
initial stresses and time on orthodontic external root resorption. J
Dent Res 2013;92:346-51.
24. Bahammam MA. Effectiveness of bovine-derived xenograft versus
bioactive glass with periodontally accelerated osteogenic ortho-
dontics in adults: a randomized, controlled clinical trial. BMC
Oral Health 2016;16:126.
25. Chung KR, Kim SH, Lee BS. Speedy surgical-orthodontic treatment
with temporary anchorage devices as an alternative to orthog-
nathic surgery. Am J Orthod Dentofacial Orthop 2009;135:
787-98.
26. Chopade SR, Madhav VV, Palaskar J. Finite element analysis: new
dimension in prosthodontic research. J Dent Allied Sci 2014;3:
85-8.
27. Aversa R, Apicella D, Perillo L, Sorrentino R, Zarone F, Ferrari M,
et al. Non-linear elastic three-dimensional finite element analysis
on the effect of endocrown material rigidity on alveolar bone re-
modeling process. Dent Mater 2009;25:678-90.
28. Lee NK, Baek SH. Stress and displacement between maxillary pro-
traction with miniplates placed at the infrazygomatic crest and the
lateral nasal wall: a 3-dimensional finite element analysis. Am J
Orthod Dentofacial Orthop 2012;141:345-51.
29. Lin TS, Tsai FD, Chen CY, Lin LW. Factorial analysis of variables
affecting bone stress adjacent to the orthodontic anchorage
mini-implant with finite element analysis. Am J Orthod Dentofa-
cial Orthop 2013;143:182-9.
August 2021  Vol 160  Issue 2
DIGITAL ORTHODONTICS

Three-dimensional assessment of virtual

bracket removal for orthodontic
retainers: A prospective clinical study
Kaitlin Marsh,a Andre Weissheimer,a Kaifeng Yin,b Alexandra Chamberlain-Umanoff,c Hongsheng Tong,a
and Glenn T. Sameshimaa
Los Angeles and Redlands, Calif, and Houston, Tex

Introduction: Computer-aided design and manufacturing of orthodontic retainers from digitally debonded
models can be used to facilitate same-day delivery. The purpose of this prospective clinical study was to
validate a novel technique for virtual bracket removal (VBR) in-office, comparing the accuracy with 2
orthodontic laboratories that use VBR for retainer fabrication in the digital workflow. Methods: The sample con-
sisted of 40 intraoral scans of 20 patients. Four groups were compared. The scans without brackets were used
as a control group. VBR was performed by 3 groups: In-office VBR (Software Meshmixer, version 3.5.474;
Autodesk, San Rafael, Calif), Orthodent Laboratory (ODL; Buffalo, NY), and New England Orthodontic
Laboratory (NEOLab; Andover, Mass). The virtually debonded models were superimposed onto the control
models using surface-based registration. Regional 3-dimensional Euclidean distances between surface
points of superimposed models were calculated for comparative analysis of surface changes after VBR using
Vector Analysis Module (Canfield Scientific, Fairfield, NJ) software. Results: The accuracy of VBR using the
Meshmixer did not differ significantly from the VBR protocols used by the 2 laboratories. However, there was
a statistically significant difference between the 2 laboratories, with ODL showing lower accuracy than NEOLab.
Although some differences were statistically significant, they were very small and not considered clinically rele-
vant. There was also a statistically significant difference between the 3 tooth segments (incisors, canines/pre-
molars, and first molars), with VBR of the first molars and second premolars showing the least accuracy.
Conclusions: The VBR techniques using the in-office Meshmixer, ODL, and NEOLab were considered
accurate enough for the clinical use of orthodontic retainers fabricated from printed models. (Am J Orthod
Dentofacial Orthop 2021;160:302-11)

O
ne of the greatest orthodontic challenges is
maintaining tooth position after debonding,
and thus, ensuring timely manufacturing of re-
tainers is key to the success and longevity of orthodontic
treatment.1 Orthodontic retainers should be placed
immediately after the removal of the appliances because
some relapse may occur in a few hours.2,3 Traditionally,
retainer fabrication workflow has involved taking an
alginate impression before bracket removal or after
a
Advanced Orthodontic Program, Herman Ostrow School of Dentistry, University
of Southern California, Los Angeles, Calif.
b
c
Private practice, Houston, Tex.
Private practice, Redlands, Calif.
All authors have completed and submitted the ICMJE Form for Disclosure of Po-
tential Conflicts of Interest, and none were reported.
Address correspondence to: Andre Weissheimer, Department of Orthodontics,
Herman Ostrow School of Dentistry, University of Southern California, 925 W
34th St, Los Angeles, CA 90089; e-mail, weisshei@usc.edu.
Submitted, May 2020; revised, August 2020; accepted, September 2020.
0889-5406/$36.00
Ó 2021 by the American Association of Orthodontists. All rights reserved.
https://doi.org/10.1016/j.ajodo.2020.09.027
bracket removal, pouring the impression up in stone,
physically carving the brackets off if the impression was
taken before bracket removal, and fabricating the
retainer on the stone model. Problems during the stone
model fabrication may occur, requiring the patient to
come back to the office for a new impression. With the
introduction of intraoral scanners and software for
computer-aided design (CAD) and computer-aided
manufacturing, orthodontic appliance fabrication tech-
niques have evolved and have become digital.4,5 Advan-
tages of 3-dimensional (3D) digital scanning include
simplicity, accuracy, longevity,6 reduced patient discom-
fort, elimination of impression material in inventory,
reduced storage issues, and minimization of cross-
contamination.7 Moreover, the intraoral scan can be 3D
printed in resin and used for appliance fabrication.
With 3D printing gaining traction in the orthodontic
community, many private practices have been investing
in in-office digital laboratories for 3D-printing models
and fabrication of appliances to increase efficiency and

302
Marsh et al
reduce the number of appointments for patients.7-10
Using the digital workflow, high-quality retainers can
be fabricated from 3D printed models.7-11
In the digital workflow, the retainer fabrication involves
the acquisition of the patient’s intraoral scan, postprocess-
ing of the digital models in stereolithography (STL) file
format, virtual bracket removal (VBR) procedure in CAD
software, model 3D-printing, and fabrication of the
retainer before the debonding appointment. The first
step in this workflow is to acquire an accurate intraoral
scan, which is critical. There are many intraoral scanners
in the market that possess trueness and precision sufficient
for orthodontic applications.12 Among them, TRIOS 3
(3Shape, Copenhagen, Denmark) is a popular intraoral
scanner that has shown good trueness and precision even
when scanning arches with bonded buccal brackets.11
Studies have shown that digital impressions and models
accomplish equal or higher precision than some conven-
tional impression materials and stone models.14-17 After
the acquisition of the intraoral scan, digital model
postprocessing is necessary to remove artifacts such as
noise, outliers, holes, or ghost geometry,18,19 preparing
the models for the next step: VBR, which is a new procedure
in orthodontics in which the brackets are digitally selected
and removed from the tooth surface to produce a digital
model without brackets. VBR can be performed using
many different CAD software programs, such as Meshmixer
(Autodesk, San Rafael, Calif) and OrthoAnalyzer (3Shape).
VBR may take 4-5 minutes per arch, depending on the soft-
ware used and the operator’s skill. Once VBR is performed,
the digital model can be 3D printed to serve as a physical
model for retainer fabrication.8 VBR before the debonding
appointment has the important advantage of same-day
delivery of a well-fitting fixed or removable retainer, with
the added advantage of eliminating an office visit.
VBR can be performed either in-office or by an ortho-
dontic laboratory that provides this new digital service in
lieu of physical carving of brackets from traditional plas-
ter models. Because VBR is a novel technique, there are
no studies that have tested this procedure to produce ac-
curate 3D-printed models for retainer fabrication. Thus,
this prospective clinical study aimed to validate a novel
technique for VBR in-office, comparing the accuracy
with 2 orthodontic laboratories that use VBR for re-
tainers fabrication in the digital workflow.
The hypothesis is VBR can be performed accurately
enough to be used for orthodontic retainers, indepen-
dently of the software program or laboratory.
MATERIAL AND METHODS
The University Park Institutional Review Board of the
University of Southern California (USC) approved this
American Journal of Orthodontics and Dentofacial Orthopedics
303
study. Informed consent and assent were obtained
from the legal guardians and patients, respectively.
The inclusion criteria for this prospective clinical study
were as follows: (1) patients starting or finishing ortho-
dontic treatment, allowing for intraoral scans done on
the same day as bonding or debonding, (2) full fixed
labial appliances bonded at least from maxillary left first
molar to maxillary right first molar, and (3) at least one
tooth per segment (incisors, canine/premolars, and first
molar). In this study, 2 patients were excluded because
of poor scan quality and the presence of a band on a first
molar. The sample consisted of 40 maxillary dentition
intraoral scans of 20 patients of the USC Advanced Or-
thodontic Clinic. Two maxillary intraoral scans of each
patient, one with brackets and one without brackets,
were acquired at the same appointment during either
the beginning (before and after bonding, 2 of 20 pa-
tients) or at the completion of orthodontic treatment
(before and after debonding, 18 of 20 patients) using
the TRIOS 3 intraoral scanner. The scanner’s software
postprocessed the intraoral scans, exporting them as
digital models in the STL format. Four groups were
compared (all subject participants were included in
each group). Group 1 was the control group that
consisted of intraoral digital models without brackets
(postdebonding scans/prebonding scans). Group 2 was
post-VBR digital models from the in-office Meshmixer
VBR protocol. Group 3 was post-VBR digital models
from Orthodent Laboratory (ODL; Buffalo, NY), and
group 4 was post-VBR digital models from New England
Orthodontic Laboratory (NEOLab; Andover, Mass)
(Fig 1).
The Meshmixer VBR protocol used in this study was
developed and validated in vitro in a previous prelimi-
nary typodont study.20 Meshmixer is freeware software
that can be downloaded at www.meshmixer.com. After
importing the STL files into the Meshmixer software,
the first step was the model preparation, which includes
digitally removing scan artifacts connected to brackets
before VBR. Per the VBR protocol, surface lasso selection
mode allows surface faces to be selected without “paint-
ing.”21 These selection boundaries can be refined using
the smooth boundary tool. Once the boundary around
the bracket is smoothed, the erase and fill tool is selected
to virtually erase the bracket. Figure 2 details the proto-
col for VBR using Meshmixer, which was performed from
maxillary right first molar to maxillary left first molar.
There are a few outside orthodontic laboratories that
offer digital bracket removal. Two of these laboratories
were selected on the basis of their advertised capability
to digitally remove brackets: ODL and NEOLab. Each sub-
ject participant’s digital model was coded with a number
as not to reveal the subject’s name (eg, VBR #2). All
August 2021  Vol 160  Issue 2
304 Marsh et al

Fig 1. Workflow for VBR clinical study.

Fig 2. VBR with Meshmixer: A, intraoral scan with brackets; B, bracket selection using surface lasso
tool; C, smooth boundary tool (hotkey B) was applied; D, The irregular selection boundary around the
bracket was refined; E, virtual removal of the bracket using the erase and fill operation (hotkey F) set to
property panel defaults; F, visualization of a tooth after VBR.

digital models of the bracketed maxillary arch for each
subject participant were sent to both laboratories. These
laboratories performed digital bracket removal using Or-
thoAnalyzer software with their own VBR protocols and
attached a digital model in STL format for each patient
after bracket removal. To avoid the risk of bias, neither
laboratory was aware of the study.
For 3D evaluation of VBR accuracy, the digital
models of all groups were imported into the 3-matic
3D modeling software (Materialise, Leuven, Belgium)
for 3D superimposition. Although STL models do not
contain any volumetric data, their triangulated surface
data can be used for the superimposition of 3D surface
data of digital models.22-25 The scans of groups 2-4
were superimposed onto the group 1 control models
using the surface-based registration technique, which
provides the best fit between the models (Fig 3). The su-
perimposition accuracy on the areas not affected by the
VBR was evaluated by the iterative closest point algo-
rithm and color-coded maps (6 300 mm visualization
range). To complement the visual inspection, the mea-
sure analysis locally tool was used to quantify the super-
imposition error in the stable areas (green color) where
surface changes were not expected (Fig 3, D). If the

August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Marsh et al 305

Fig 3. Superimposition onto control: A, all digital models registration using surface superimposition; B,
digital models of group 3 (pink) and control group 1 (purple) before superimposition; C, after superim-
position; D, superimposition accuracy confirmation using color-coded maps (green hues indicates no
surface changes; blue or red indicates surface changes $ 0.3 mm in different directions).

registration error ranged from 0 to 0.05 mm, the model’s
superimposition accuracy was confirmed. Then, the dig-
ital models were exported as STL files and transferred to
the Vector Analysis Module (VAM; Canfield Scientific,
Fairfield, NJ) for 3D assessment of the VBR accuracy.
The entire VBR procedure was performed once by
each group. All the measurements for VBR accuracy
assessment (40 maxillary intraoral scans) were per-
formed by 2 separate investigators using VAM to ensure
interexaminer reliability. The investigators were third-
year orthodontic residents previously trained by 1 expe-
rienced orthodontist (AW) on how to use the software for
VBR evaluation. Each investigator repeated all of the
measurements twice (40 maxillary intraoral scans) after
a 2-week interval to ensure intraexaminer reliability.
Four groups of maxillary arch digital models were thus
identified for the measurements: (1) a control group of
clinically debonded or prebonding, (2) VBR performed
in-office by Meshmixer protocol, (3) VBR performed by
ODL, and (4) VBR performed by NEOLab. Regional 3D
Euclidean distances between surface points of the super-
imposed control and debonded models were measured
using VAM software. The superimposed control and
virtually debonded models were measured on the labial
surface using an iterative closest point algorithm for
comparative analysis of surface changes after VBR. The
color surface by distance tool (6 300 mm visualization
range) was used to display and to indicate the areas
with surface changes to be included in the measurement.
For each tooth, the paint area selection tool was used to
select the area where the bracket was virtually removed,
previously indicated by the color maps (Fig 4). A regional
color-coded map (6 300 mm visualization range) of the
selected area illustrated the linear surface changes after
VBR. VAM automatically calculated the minimum,
maximum, root mean square (RMS), and mean values
with a standard deviation for each selected area. Com-
parisons were made between the 3 VBR techniques
(Meshmixer, ODL, and NEOLab), the tooth segments,
and individual teeth. An overview of the whole workflow
for the VBR process and 3D evaluation is shown in
Figure 5.
The 3 models that underwent VBR in each sample
were randomized and coded by 1 of the authors to
ensure blinding of the examiners performing the mea-
surements. The 3 VBR models in each sample were given
the codes U, X, or Y, each corresponding to a specific lab-
oratory that was unknown by the examiner performing
the measurements. The coding system was revealed
only after the statistical analysis was complete.
Statistical analysis
The RMS values, which best represented the overall
magnitudes of surface change irrespective of the direc-
tion of change of the 3 techniques, were compared.
Descriptive statistical analysis was performed with
SPSS (version 20.0; IBM, Armonk, NY) for linear surface
changes because of VBR. Interexaminer (Cronbach
alpha) and intraexaminer (Cronbach alpha) reliability
were determined. The Shapiro-Wilk test was used to
evaluate the normality of the data. Multiple linear
regression analysis with tooth/segment and laboratory

American Journal of Orthodontics and Dentofacial Orthopedics August 2021  Vol 160  Issue 2
306 Marsh et al

Fig 4. Regional evaluation of VBR using VAM software: A, semitransparent original model showing the
brackets position in relation to the VBR displayed by the color maps; B, same color maps showing the
superimposed control and post-VBR scan to indicate the area to be measured; C, paint area tool over
the greatest surface changes indicated by the color maps; D, regional color maps of the painted area
and the linear surface changes measurements of the VBR. Color map (6 300 mm): Red hues indicate a
negative value where unintentional tooth surface removal occurred. Blue hues indicate positive values
where insufficient bracket removal occurred.

as independent variables was first attempted. Because

the laboratory was identified to contribute insignifi-
cantly to the total variance, 3 separate 1-way analysis
of variance (ANOVA) tests26 were used to detect the po-
tential differences between 3 VBR protocols/labora-
tories, separate teeth, and tooth segments (incisors,
canines/premolars, and first molars). Statistical differ-
ences between the test groups were further analyzed
with Scheff
e post-hoc test (⍺ 5 0.05).27

Fig 5. Overview of the VBR measurement workflow: A,
digital model of the control group; B, digital model with
brackets before VBR; C, digital model after VBR; D,
VBR model and its corresponding control model superim-
posed; E, models superimposition confirmation; F, selec-
tion of area for the VBR; G, regional color map of linear
surface changes after VBR. Color map (6 300 mm):
Red hues indicate a negative value where unintentional
tooth surface removal occurred; blue hues indicate posi-
tive values where insufficient bracket removal occurred.
RESULTS
Interexaminer and intraexaminer reliability were
determined to be high (.0.9). Because there were no
significant interexaminer and intraexaminer differences
in the VBR measurement reliability, the RMS values
from both operators and measurement trials were aver-
aged by tooth for each VBR group. One-way ANOVA
used averaged RMS surface changes by tooth, from cen-
tral incisor to first molar (1-6), and significance was
determined at P \0.05. RMS surface changes ranged
from 0.10 mm to 0.30 mm, with first molars exhibiting
the greatest amount of surface change and central inci-
sors exhibiting the least amount of surface change re-
sulting from VBR (Fig 6). The first molars and the
second premolar showed the greatest distribution of sur-
face change, indicating the largest error in VBR. Post-
hoc analysis with Scheff
e test (Table I) showed a pairwise

August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Marsh et al 307

Fig 6. Boxplot illustrating RMS surface changes distribu-
tions by tooth: (1) central incisor, (2) lateral incisor, (3)
canine, (4) first premolar, (5) second premolar, and (6)
first molar. The distribution of RMS surface change
values was greatest for tooth 5 and 6, with tooth 6
showing the greatest distribution and highest median.
o
an outlier; *an extreme outlier.
Fig 7. Boxplot illustrating RMS surface change values
distribution by tooth segment: (0) incisors segment; (1)
canine/premolars segment; (2) first molars. First molars
displayed the greatest distribution of RMS values for sur-
face changes and the greatest median value or greatest
error. oan outlier; *an extreme outlier.

and first molars) is illustrated in Figure 7. The first molar

statistically significant difference (P \0.05) in averaged segment showed the greatest distribution of surface
RMS values between the second premolar and all other change, indicating the largest error in VBR. Post-hoc
teeth in the arch as well as the first molar and all other analysis with Scheff
e test (Table II) showed a pairwise
teeth in the arch. statistically significant difference (P \0.05) in averaged
The 1-way ANOVA of the tooth segments (incisors, RMS values between the first molar segment and the in-
canine/premolars, and first molar) using averaged RMS cisors as well as the first molar segment and the canine/
surface changes and a significance determined at premolar segment.
P \0.05 showed RMS surface changes ranging from In evaluating the differences between the in-office
0.12 mm to 0.33 mm, with the first molars exhibiting Meshmixer method and 2 laboratories, a 1-way ANOVA
the greatest amount of surface change and the incisors was performed using averaged RMS surface changes by
exhibiting the least amount of surface change resulting laboratory (Meshmixer, ODL, and NEOLab). Significance
from VBR. The pattern of distribution of the surface was determined at P \0.05. VBR produced surface
changes for each segment (incisors, canine/premolars, changes that ranged from 0.15 mm to 0.19 mm, with
NEOLab exhibiting the least amount of surface changes
and ODL exhibiting the greatest amount of surface
Table I. Scheff

e post-hoc analysis of RMS by tooth change (RMS). The pattern of distribution of the surface
Pair Mean difference Standard error P value change is illustrated by a boxplot in Figure 8. The in-
1 vs 2 0.03 0.01 0.405 office method and the 2 laboratories showed a similar
1 vs 3 0.01 0.01 0.930 distribution of surface change which indicates a similar
1 vs 4 0.01 0.01 0.984
1 vs 5 0.09 0.01 \0.001*
1 vs 6 0.23 0.01 \0.001*
2 vs 3 0.01 0.01 0.944 Table II. Scheff

e post-hoc analysis of RMS by tooth
2 vs 4 0.02 0.01 0.897 segment
2 vs 5 0.06 0.01 0.001*
2 vs 6 0.20 0.01 \0.001* Mean Standard
3 vs 4 0.00 0.01 1.000 Pair difference error P value
3 vs 5 0.08 0.01 \0.001 Incisors vs canine/ 0.03 0.01 0.013*
3 vs 6 0.21 0.01 \0.001* premolar
4 vs 5 0.08 0.01 \0.001* Incisor vs first molar 0.21 0.01 \0.001***
4 vs 6 0.22 0.01 \0.001* Canine/premolar vs first 0.19 0.01 \0.001***
5 vs 6 0.14 0.01 \0.001* molar

Note. Values in millimeters. The mean difference is significant at the Note. Values in millimeters. The mean difference is significant at the
0.05 level. Tooth 1-6: central incisor to first molar. 0.05 level.
*P #0.001. *P #0.05; ***P #0.001.

American Journal of Orthodontics and Dentofacial Orthopedics August 2021  Vol 160  Issue 2
308
Fig 8. Boxplot illustrating RMS values distribution by lab-
oratory: (0) Meshmixer VBR; (1) ODL VBR; (2) NEOLab
VBR. oan outlier; *an extreme outlier.
error in VBR. Post-hoc analysis with the Scheff
e test
showed a pairwise statistically significant difference
(P \0.05) in averaged RMS values between ODL and
NEOLab (Table III). No statistically significant difference
was found between the Meshmixer and ODL as well as
Meshmixer and NEOLab.
The net negative and positive values from the raw
data measurements were added up, and the percentage
of negative and positive surface changes was calculated
for each laboratory (Table IV). Of the total sample of 223
teeth, NEOLab exhibited the greatest percentage of un-
intentional tooth surface removal (29.7%), followed by
ODL (13.7%) and Meshmixer (8.9%). Meshmixer showed
the greatest percentage of insufficient bracket removal
(91.1%), followed by ODL (86.3%) and NEOLab (70.3%).
DISCUSSION
As orthodontics moves from analog to digital, some
orthodontic laboratories have started offering VBR as a
digital service. However, there is no scientific evidence
regarding the accuracy of VBR, which has been recently
introduced and used for 3D printed models for same-day
retainer delivery. In our previous preliminary study on ty-
podonts, VBR was shown to be accurate and reproduc-
ible with in-office protocols established using either
Table III. Scheff

e post-hoc analysis of RMS by
laboratory
Pair Mean difference Standard error
Meshmixer vs ODL 0.02 0.01
Meshmixer vs NEOLab 0.02 0.01
ODL vs NEOLab 0.03 0.01
Note. Values in millimeters. The mean difference is significant at the
0.05 level.
*P #0.05.
August 2021  Vol 160  Issue 2
Marsh et al
Table IV. Percentage negative and positive tooth sur-
face change after VBR per laboratory
Tooth surface change Meshmixer ODL NEOLab
Negative surface change (% of 223) 8.9 13.7 29.7
Positive surface change (% of 223) 91.1 86.3 70.3
Note. Values are %. Negative surface changes mean unintentional
tooth surface removal. Positive surface changes mean insufficient
bracket removal.
Meshmixer or OrthoAnalyzer software.20 In this clinical
study, the in-office VBR with Meshmixer was compared
with 2 orthodontic laboratories that used OrthoAnalyzer.
Meshmixer was chosen because it is a widely-used soft-
ware, user-friendly, and free-to-download, whereas
OrthoAnalyzer is available at a high cost to be used for
in-office VBR for retainer fabrication.
VBR is performed on digital models obtained by any
intraoral scanner. Although several commercially avail-
able intraoral scanners generate reliable digital
models,16,17,28-31 TRIOS 3 was used in this clinical
study because it was the intraoral scanner available at
the USC orthodontic clinic. Some studies evaluated the
influence of scanning technique.28,32,33 TRIOS 3 is less
likely to be influenced by scanning technique.28,33
Moreover, it has also demonstrated reliable scanning ac-
curacy of dentitions with fixed labial appliances,12,13,34
which was critical for this study.
The accuracy of VBR processes is evaluated best by
superimposing the digital models using surface-based
registration and calculating the distances between sur-
face points, further illustrated by color map-
ping.14,22-25,35,36 The ability to make meaningful
conclusions about the error due to VBR relied on the
3D superimposition accuracy. Therefore, 3D surface-
based superimposition of the control and virtually
bonded models was performed using 3-magic software,
and the superimposition accuracy verified using color-
coded maps, ensuring any change on the labial surface
of the teeth would be the result of VBR and not because
of superimposition errors between the digital models. If
the registration error was greater than 0.05 mm, quanti-
fied by the measure analysis locally tool, the model’s su-
perimposition was repeated. When the registration
accuracy was confirmed (error ranging from 0 mm to
0.05 mm), the digital models were exported as STL files
P value and transferred to VAM software. The accuracy of VBR
0.341 was assessed 3-dimensionally using VAM’s color surface
0.356 by distance tool (6 300 mm visualization range) to
0.015* display and measure linear surface changes between
the superimposed digital models. The paint area selec-
tion tool, which is designed for identifying regions of
dimensional differences,37 was used to generate regional
American Journal of Orthodontics and Dentofacial Orthopedics
Marsh et al
color-coded maps on the labial surfaces where the
brackets were virtually removed for quantitative analysis
of the minimum values, maximum values, RMS values,
and standard deviations. To ensure reproducibility, the
paint area selection tool remained the same shape and
size for each bracket area selection. This procedure was
supported by the results, which showed a high degree
of intraexaminer and interexaminer reliability. The red
hues in the VAM software indicated negative values,
which represented zones of unintentional tooth surface
removal, and blue hues indicated positive values, which
represented zones of insufficient bracket removal. Aver-
aged RMS values, and not the average of mean values,
were used because RMS represents the overall magni-
tude of surface changes, regardless of the direction of
change. Conversely, the greatest value of minimum or
maximum for each tooth detailed the net direction of
change. RMS was used in accordance with a previous
study that measured surface changes between superim-
posed condyles using VAM software.37
The results of this study show that the accuracy of VBR
decreases from the anterior teeth to the posterior teeth. All
3 segments of teeth showed a statistically significant dif-
ference when compared with each other. The first molar
segment displayed the highest RMS value of 0.33 mm,
whereas the incisors showed the smallest RMS value of
0.12 mm, demonstrating that VBR is the least accurate
in the posterior segment. Furthermore, color mapping
showed a higher prevalence of blue hues in the posterior
segments, implying that there is a tendency to have insuf-
ficient bracket removal in the posterior teeth when
compared with the anterior teeth. The proximity between
the posterior brackets and the gingival margins may have
interfered with the algorithm used by the software when
computing the selected area and removing the bracket.
Canines and incisors have brackets positioned in the mid-
dle of the crown and away from the gingival margins,
which may explain the better results for VBR accuracy.
In addition, the greatest error in the first molar region
could also be explained by the presence of the buccal
groove and the inability of the virtual removal software
to navigate big changes in surface curvature and remove
adhesive inside the groove.20,36 Studies have shown that
intraoral scanning with TRIOS 3 is least accurate in the
molar region14,28,29,38; however, the scanning error is
less than 0.10 mm at the distal tooth, which is not signif-
icant.15 Because scanning accuracy was verified to be
within 0.10 mm, surface changes seen at all 3 segments
were likely because of VBR error and not because of scan-
ning inaccuracy. Although the VBR surface changes (RMS
values) for all 3 tooth segments were greater than 0.1 mm,
they were all within the reported accuracy for orthodontic
casts, which is 0.3-0.5 mm.39
American Journal of Orthodontics and Dentofacial Orthopedics
309
The comparison of the in-office Meshmixer with the
2 laboratories showed that ODL had the highest RMS
value of surface change (0.19 mm), followed by Mesh-
mixer (0.17 mm) and NEOLab (0.15 mm). Despite the
statistically significant difference between ODL and
NEOLab, the surface changes due to VBR are small and
may not be clinically significant (Table III). Regarding
the direction of surface changes after VBR, whether un-
intentional tooth surface removal or insufficient bracket
removal, all 3 VBR techniques showed a higher percent-
age of insufficient bracket removal, being 91.1% for
Meshmixer, 86.3% for ODL, and 70.3% for NEOLab
(Table IV). Minor insufficient bracket removal ranging
from 0.1 mm to 0.3 mm can be more favorable from a
clinical and from a retainer manufacturing standpoint
than unintentional tooth surface removal, which can
make the retainer insertion difficult. NEOLab displayed
the highest percentage of unintentional tooth surface
removal with 29.7%, followed by ODL with 13.7% and
the in-office Meshmixer with 8.9%. Although uninten-
tional tooth surface removal may be unfavorable
because vacuum-formed thermoplastic retainers may
not adapt well, these retainers have a certain elasticity,
and this may not be an issue. In the current study, the
ODL laboratory provided a thermoform retainer with
each VBR model delivered to patients. The clinical
assessment, although subjective, showed a proper fit of
the retainer on the maxillary dentition. However, future
studies should evaluate and measure the fit of the re-
tainers fabricated from VBR models, using methods of
fit evaluation outlined by previous studies.39 One limita-
tion of this clinical study was the possible variations that
may have occurred during the debonding procedure,
such as unintentional removal of tooth structure or the
presence of some residual adhesive. Using only prebond-
ing digital models as control would have the advantage
of no residual adhesive. However, using postbonding
digital models allowed for clinical assessment of the
retainer fit in the patient’s maxillary arch. Because of
the novelty of this research, there is a lack of published
articles to compare and discuss the results. The only
available data for comparison was our preliminary
in vitro typodont study in 3D printed models.20 In the
current clinical study, the VBR was less accurate in the
posterior segment with a tendency toward average pos-
itive change, which indicates insufficient bracket
removal across all groups. Conversely, the in vitro typo-
dont study found a negative change in the first molars
and incisors, indicating unintentional bracket removal
but positive changes for the premolars and canines.20
The difference between the in vitro and the clinical
studies may be attributed to the sample evaluated. The
former used 3D printed models in gray resin, whereas
August 2021  Vol 160  Issue 2
310
Table V. Multifactorial comparison between Meshmixer, ODL, and NEOLab
Factor Meshmixer
VBR software used Meshmixer
Cost Free software 1 3D-printed model
$2, retainer clear material
$2 1 labor work 1 other
variables. Estimation: $10-15
Turnaround time ~ 10 min for VBR 1 time to send
out to print or perform in-office
Practical application VBR only 1 print in-office for
retainer or send out to print
the latter used real tooth surfaces where the enamel was
the surface evaluated.
Regarding the accuracy of VBR techniques using in-
office Meshmixer, ODL, or NEOLab, although some dif-
ferences were statistically significant, they were very
small and not considered clinically relevant. Despite
the similar accuracy of the 3 techniques for VBR, other
attributes may also be important for the decision-
making regarding the technique of choice. Table V
outlines some features of the in-office VBR and the 2
laboratories, including the cost analysis. In-office VBR
using Meshmixer has the advantage of being free soft-
ware, and it has quicker turnaround time if the 3D-print
models and retainers are fabricated in-office as well.
ODL has a faster turnaround time than NEOLab and is
comparable in price. Both ODL and NEOLab use Ortho
Analyzer to perform VBR.
Future studies could focus on VBR of other bracket
materials not evaluated in this study, such as clear
ceramic brackets, lingual brackets,40 and aligner attach-
ments. Moreover, orthodontic bands with tubes were not
utilized in this study.
CONCLUSIONS
In this prospective clinical study, the hypothesis that
VBR is accurate to be used for orthodontic retainers,
independently of the software or laboratory, was
confirmed. The VBR techniques using the in-office
Meshmixer VBR protocol or by the orthodontic labora-
tories, ODL and NEOLab, were considered accurate
enough for the clinical use of orthodontic retainers
fabricated from 3D-printed models.
AUTHOR CREDIT STATEMENT
Kaitlin Marsh: investigation and writing original
draft; Andre Weissheimer: conceptualization, methodol-
ogy, investigation, writing, reviewing, editing, and visu-
alization; Kaifeng Yin: formal analysis; Alexandra
Chamberlain-Umanoff: preliminary typodont study;
August 2021  Vol 160  Issue 2
Marsh et al
ODL NEOLab
OrthoAnalyzer OrthoAnalyzer
$48.50 (1 arch VBR 1 printed $45 single arch package or $90
model 1 Essix) full arch package (includes VBR,
3D models, and Essix retainer)
~ 7-10 d ~ 14 d
Offers VBR 1 3D-printed Offers VBR only or VBR 1 3D-
model 1 Essix retainer as a printed model 1 Essix retainer
package
Hongsheng Tong: reviewing and editing; and Glenn T.
Sameshima: supervision, reviewing, and editing.
ACKNOWLEDGMENTS
This study was approved by the University of South-
ern California Institutional Review Board (no. HS-19-
00590). This study did not receive any specific grant
from funding agencies in the public, commercial, or
not-for-profit sectors.
SUPPLEMENTARY DATA
Supplementary data associated with this article can
be found, in the online version, at https://doi.org/10.
1016/j.ajodo.2020.09.027.
REFERENCES
1. Kravitz ND, Groth C, Jones PE, Graham JW, Redmond WR. Intrao-
ral digital scanners. J Clin Orthod 2014;48:337-47.
2. Reitan K. Principles of retention and avoidance of posttreatment
relapse. Am J Orthod 1969;55:776-90.
3. Brain WE. The effect of surgical transsection of free gingival fibers
on the regression of orthodontically rotated teeth in the dog. Am J
Orthod 1969;55:50-70.
4. Jheon AH, Oberoi S, Solem RC, Kapila S. Moving towards precision
orthodontics: an evolving paradigm shift in the planning and de-
livery of customized orthodontic therapy. Orthod Craniofac Res
2017;20(Suppl 1):106-13.
5. Vaid NR. Digital technologies in orthodontics–an update. Semin
Orthod 2018;24:373-5.
6. Hazeveld A, Huddleston Slater JJR, Ren Y. Accuracy and reproduc-
ibility of dental replica models reconstructed by different rapid
prototyping techniques. Am J Orthod Dentofacial Orthop 2014;
145:108-15.
7. Vasudavan S, Sullivan SR, Sonis AL. Comparison of intraoral 3D
scanning and conventional impressions for fabrication of ortho-
dontic retainers. J Clin Orthod 2010;44:495-7.
8. Groth C, Kravitz ND, Shirck JM. Incorporating three-dimensional
printing in orthodontics. J Clin Orthod 2018;52:28-33.
9. Tarraf NE, Ali DM. Present and the future of digital orthodontics✰.
Semin Orthod 2018;24:376-85.
10. Nakano H, Kato R, Kakami C, Okamoto H, Mamada K, Maki K.
Development of biocompatible resins for 3D printing of direct
aligners. J Photopol Sci Technol 2019;32:209-16.
American Journal of Orthodontics and Dentofacial Orthopedics
Marsh et al
11. Tahir NM, Wan Hassan WN, Saub R. Comparing retainers con-
structed on conventional stone models and on 3D printed models:
a randomization crossover clinical study. Eur J Orthod 2018;2019:
370-80.
12. Claus D, Radeke J, Zint M, Vogel AB, Satravaha Y, Kilic F, et al.
Generation of 3D digital models of the dental arches using optical
scanning techniques. Semin Orthod 2018;24:416-29.
13. Park JM, Choi SA, Myung JY, Chun YS, Kim M. Impact of ortho-
dontic brackets on the intraoral scan data accuracy. BioMed Res
Int 2016;2016:5075182.
14. Ender A, Attin T, Mehl A. In vivo precision of conventional and dig-
ital methods of obtaining complete-arch dental impressions. J
Prosthet Dent 2016;115:313-20.
15. Burzynski JA, Firestone AR, Beck FM, Fields HW Jr, Deguchi T.
Comparison of digital intraoral scanners and alginate impressions:
time and patient satisfaction. Am J Orthod Dentofacial Orthop
2018;153:534-41.
16. Sousa MVS, Vasconcelos EC, Janson G, Garib D, Pinzan A. Accu-
racy and reproducibility of 3-dimensional digital model measure-
ments. Am J Orthod Dentofacial Orthop 2012;142:269-73.
17. Wiranto MG, Engelbrecht WP, Tutein Nolthenius HE, van der
Meer WJ, Ren Y. Validity, reliability, and reproducibility of linear
measurements on digital models obtained from intraoral and
cone-beam computed tomography scans of alginate impressions.
Am J Orthod Dentofacial Orthop 2013;143:140-7.
18. Weyrich T, Pauly M, Keiser R, Heinzle S, Scandella S, Gross M. Post-
processing of scanned 3D surface data. In: Alexa M, Gross M,
Pfister H, Rusinkiewicz S, editors. Eurographics Symposium on
Point-Based Graphics. Aire-la-Ville, Switzerland: The Euro-
graphics Association; 2004. p. 83-94.
19. 3Shape. Ortho System 2020.1. Orthodontic Solution. Technical
Documentation. 3Shape A/S.. OS-85.0-B-EN. 3Shape; 2020.
Available at: www.3shape.com. Accessed June 2 2021.
20. Chamberlain-Umanoff A. Assessment of 3D surface changes
following virtual bracket removal [thesis]. Los Angeles: University
of Southern California; 2019.
21. Available at: http://help.autodesk.com/view/MSHMXR/2019/
ENU/?guid5GUID-6EE3D65F-F5E2-4097-87A5-3BA0750AFE
C5. Accessed June 2 2021.
22. Gkantidis N, Schauseil M, Pazera P, Zorkun B, Katsaros C,
Ludwig B. Evaluation of 3-dimensional superimposition tech-
niques on various skeletal structures of the head using surface
models. PLoS One 2015;10:e0118810.
23. Brown LG. A survey of image registration techniques. ACM Com-
put Surv 1992;24:325-76.
24. Ghoneima A, Cho H, Farouk K, Kula K. Accuracy and reliability of
landmark-based, surface-based and voxel-based 3D cone-beam
computed tomography superimposition methods. Orthod Cranio-
fac Res 2017;20:227-36.
American Journal of Orthodontics and Dentofacial Orthopedics
311
25. Lin HH, Chiang WC, Lo LJ, Wang CH. A new method for the inte-
gration of digital dental models and cone-beam computed tomog-
raphy images. Annu Int Conf IEEE Eng Med Biol Soc 2013;2013:
2328-31.
26. Kao LS, Green CE. Analysis of variance: is there a difference in
means and what does it mean? J Surg Res 2008;144:158-70.
27. Scheff
e H. The Analysis of Variance. New York: Wiley; 1959.
28. Anh JW, Park JM, Chun YS, Kim M, Kim M. A comparison of the
precision of three-dimensional images acquired by 2 digital intrao-
ral scanners: effects of tooth irregularity and scanning direction.
Korean J Orthod 2016;46:3-12.
29. Fl€
ugge TV, Schlager S, Nelson K, Nahles S, Metzger MC. Precision
of intraoral digital dental impressions with iTero and extraoral
digitization with the iTero and a model scanner. Am J Orthod Den-
tofacial Orthop 2013;144:471-8.
30. Gan N, Xiong Y, Jiao T. Accuracy of intraoral digital impressions for
whole upper jaws, including full dentitions and palatal soft tissues.
PLoS One 2016;11:e0158800.
31. Lim JH, Park JM, Kim M, Heo SJ, Myung JY. Comparison of
digital intraoral scanner reproducibility and image trueness
considering repetitive experience. J Prosthet Dent 2018;119:
225-32.
32. Ender A, Mehl A. Influence of scanning strategies on the accuracy
of digital intraoral scanning systems. Int J Comput Dent 2013;16:
11-21.
33. Michelinakis G, Apostolakis D, Tsagarakis A, Kourakis G,
Pavlakis E. A comparison of accuracy of 3 intraoral scanners: a
single-blinded in vitro study. J Prosthet Dent 2020;124:581-8.
34. Jung YR, Park JM, Chun YS, Lee KN, Kim M. Accuracy of four
different digital intraoral scanners: effects of the presence of or-
thodontic brackets and wire. Int J Comput Dent 2016;19:203-15.
35. Besl PJ, McKay ND. A method for registration of 3-D shapes. IEEE
Trans Pattern Anal Mach Intell 1992;14:239-56.
36. Zitov
a B, Flusser J. Image registration methods: a survey. Image
Vis Comput 2003;21:977-1000.
37. Claus JDP, Koerich L, Weissheimer A, Almeida MS, Belle de
Oliveira R. Assessment of condylar changes after orthognathic sur-
gery using computed tomography regional superimposition. Int J
Oral Maxillofac Surg 2019;48:1201-8.
38. Rudolph H, Luthardt RG, Walter MH. Computer-aided analysis
of the influence of digitizing and surfacing on the accuracy
in dental CAD/CAM technology. Comput Biol Med 2007;37:
579-87.
ufekçi E. Evaluation of fit
39. Cole D, Bencharit S, Carrico CK, Arias A, T€
for 3D-printed retainers compared with thermoform retainers. Am
J Orthod Dentofacial Orthop 2019;155:592-9.
40. Tong H, Weissheimer A, Pham J, Lee R, Redmond WR. Lingual or-
thodontics redefined with automation and friction-free me-
chanics. J Clin Orthod 2019;53:214-24.
August 2021  Vol 160  Issue 2
CASE REPORT

Treatment of facial asymmetry and severe

midline deviation with orthodontic
mini-implants
Min-Ho Jung
Seoul, South Korea

This case report describes the treatment of a 29-year-old woman with facial asymmetry and 2 hopeless teeth.
Her lower dental midline was shifted to the left side, and the mandibular left second molar would need to be ex-
tracted because of severe caries. The maxillary right second premolar was root rest, and the upper dental midline
was shifted to the right side. Because of the patient's asymmetry and Class III skeletal pattern, a severe Class III
relationship in the right canine region and lingual crossbite in the left side was observed. She did not want jaw
surgery. The mandibular right first premolar, 2 hopeless teeth, and maxillary left second premolar were ex-
tracted, and orthodontic mini-implants were used to correct the dental midline, crossbite, and crowding. The
mandibular left third molar was moved to the second molar extraction space by using orthodontic mini-
implant anchorage. Adequate functional and esthetic results were obtained. Correction of the crossbite on the
left side could improve facial asymmetry by changing the drape of the overlying lips. (Am J Orthod
Dentofacial Orthop 2021;160:312-24)

A
symmetry is a common finding in the craniofa-
cial region1 and often causes occlusal problems
like midline deviation and crossbite. If the de-
gree of facial asymmetry is severe, it is better to improve
the problem through orthodontic decompensation and
asymmetric jaw surgery.2 In comparison, if the asymme-
try is mild, it may be difficult for the patient to agree to
surgery, and occlusal problems may need to be corrected
through orthodontic treatment alone. To achieve proper
occlusion while conducting the necessary asymmetrical
movement, anchorage control is very important. In
most patients, it is difficult to improve facial asymmetry
without orthognathic surgery, but in some patients, it
can be improved significantly by orthodontic treatment
alone.3,4
Some orthodontic patients have a missing
second molar because of severe caries. If the patient's
third molar is healthy, protraction of the third molar
From the Department of Orthodontic, Dental Research Institute and School of
Dentistry, Seoul National University, Seoul, South Korea; Private practice, Seoul,
Korea
All authors have completed and submitted the ICMJE Form for Disclosure of
Potential Conflicts of Interest, and none were reported.
Address correspondence to: Min-Ho Jung, HONORS Orthodontics, 3rd Fl, Tae-
nam B/D, Cham-won Ro 3Gil 40, Seo-cho Gu, Seoul 06510, South Korea;
e-mail, fortit@chol.com.
Submitted, February 2020; revised and accepted, May 2020.
0889-5406/$36.00
Ó 2021 by the American Association of Orthodontists. All rights reserved.
https://doi.org/10.1016/j.ajodo.2020.05.019
312
into the missing space may be considered rather than
implant placement and third molar extraction. In such
a case, particularly when the retraction of anterior
teeth is not required, anchorage becomes a significant
problem.
The introduction of orthodontic mini-implants
(OMIs) made it possible to obtain absolute anchorage.5
Asymmetric tooth movement or molar protraction can
be achieved without patient cooperation,6,7 and the
scope of orthodontic treatment has become much wider.
This case report describes asymmetric mandibular
tooth movement in a patient with facial asymmetry
who was treated with OMIs to improve the midline devi-
ation and crossbite. The hopeless mandibular left
second molar was extracted, the third molar was pro-
tracted, and the patient's occlusal problem was success-
fully corrected with adequate OMI anchorage.
DIAGNOSIS AND ETIOLOGY
A 29-year-old woman sought treatment for her chief
complaints of facial asymmetry and poor occlusion. Her
chin was deviated to the left side because of overgrowth
of the right side, and lip canting was observed (Figs 1-3).
In the posteroanterior (PA) cephalogram, based on the
crista galli–anterior nasal spine line, menton was located
to the left of the midline about 4.7 mm (Table I, Fig 4, B).
The mandibular ramus and body lengths were signifi-
cantly different from one another, but the occlusal plane
Jung
Fig 1. Pretreatment photographs.
canting was mild (0.5 ; Fig 4, B). The interpupillary line
and commissure line were not parallel to each other
(Fig 4, C).
Lingual crossbite was observed in the left premolar
area because of jaw asymmetry and Class III tendency.
As the mandibular left first molar was tilted lingually
by natural dental compensation, there was no crossbite
in the left first molar area. Because of the patient's asym-
metry and Class III skeletal pattern, a severe Class III rela-
tionship in the right canine region was also observed. In
the right molar area, the Class I molar relationship was
shown because of the forward movement of the maxil-
lary right first molar toward the second premolar space.
The patient did not have any temporomandibular joint
symptoms. The bilateral manual manipulation tech-
nique was used to evaluate condyle position,8 and the
American Journal of Orthodontics and Dentofacial Orthopedics
313
incisor relationship did not show significant lateral devi-
ation during the manipulation procedure.
The maxillary right second premolar and mandibular
left second molar were root rest because of severe caries.
In the maxillary right second premolar area, the alveolar
bone recession was observed on the panoramic radio-
graph. The upper dental midline was shifted to the right
side because the anterior teeth were moved toward the
root rest. Separately, the lower midline was shifted to
the left because of mandibular asymmetry. The nose
dorsum was slightly curved. In addition, nasion, subna-
sale, and labrale superius were not on a straight line, and
it was difficult to evaluate the amount of lateral
displacement of the dental midline. In a frontal facial
photograph with the soft tissue nasion–subnasale line,
the upper dental midline was located 1.5 mm to the right
August 2021  Vol 160  Issue 2
314
Fig 2. Pretreatment dental casts.
side (Fig 4, D). But in a PA cephalogram with a crista
galli–anterior nasal spine line, it was located about
0.8 mm to the right side (Fig 4, B).
The upper midline was located at the interproximal
surface of the mandibular right central and lateral inci-
sors. Overbite and overjet were within normal limits
(2.5 mm and 2.5 mm), and a mild to moderate degree
of crowding was observed in both arches. The patient
did not have any specific medical or dental history.
Her vertical skeletal pattern was normal (Bjork sum,
399.1 ; Facial height ratio, 60.5) (Table II), and her
sagittal pattern was mild Class III (ANB, 0.6 ). Maxillary
and mandibular incisors showed a slight lingual inclina-
tion. Mild lower lip protrusion with lip incompetency
was noted on the facial photograph (lower lip to the
esthetic line, 2.1 mm).
On a panoramic radiograph, a horizontally impacted
mandibular right third molar and mesially inclined
mandibular left third molar were observed. The maxillary
right second premolar space was almost closed by
tipping and migration of adjacent teeth.
TREATMENT OBJECTIVES
Because the patient's main concern was facial asym-
metry and poor occlusion, the objectives of treatment
were to improve facial asymmetry and lingual crossbite,
correct the midline deviation, establish a Class I canine
relationship, improve the crossbite on the left side, and
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Jung
close the space of the 2 root rest areas. Correction of
crowding and improvement of the lip profile were also
required.
TREATMENT ALTERNATIVES
To solve the patient's chief complaint of asymmetry,
2-jaw surgery combined with orthodontic treatment was
deemed the best option. At the initial examination, she
displayed 3-dimensional asymmetry. Her mandibular
ramus and body length were longer on the right side,
and significant lip canting was observed. These skeletal
discrepancies cannot be corrected by orthodontic treat-
ment alone. However, she refused jaw surgery because of
fear and concern about the side effects. It was explained
and emphasized that, without jaw surgery, her facial
asymmetry cannot be improved, but she did not change
her mind. Thus, after further consultation with the pa-
tient, it was decided to attempt to improve the occlusion
and midline without facial asymmetry correction using
orthodontic treatment alone.
Extraction of the 2 root rest was inevitable. To correct
the upper dental midline and crowding, maxillary left
second premolar extraction was required. In the mandib-
ular arch, more than 3 mm of incisor movement to the
right side and improvement of moderate crowding
(arch length discrepancy, 4.0 mm) were necessary. In
other words, a large amount of tooth movement to the
right was required in the mandibular anterior teeth.
Jung 315

Fig 3. Pretreatment radiographs.

Implant placement in the mandibular left

Table I. Pretreatment posteroanterior cephalometric
second molar space was recommended because a large
measurements
amount of space was left in that region, and the mandib-
Pretreatment ular anterior teeth had to be moved to the right, which
Rt Co-Ag 72.6 could have made it more difficult to close the extraction
Lt Co-Ag 67.5 space. However, the patient wanted to close the
Rt Ag-Me 60.1
second molar space through a third molar protraction.
Lt Ag-Me 52.9
Rt Ag-MSL 49.2 To accomplish such an effect, an additional OMI was
Lt Ag-MSL 49.0 needed to protract the third molar (Fig 5). It was ex-
MSL-Me 4.7 plained to the patient that the treatment plan was very
Rt, right; Lt, left; Co, condylion; Ag, antegonial notch; Me, menton;
complicated even without third molar protraction and
MSL, midsagittal line (crista galli–anterior nasal spine). would be too time-consuming if third molar protraction
was included. The patient finally elected to undergo or-
Mandibular right first premolar extraction with thodontic space closure of the second molar space by the
anchorage reinforcement seemed to be indicated. outlined molar protraction.

American Journal of Orthodontics and Dentofacial Orthopedics August 2021  Vol 160  Issue 2
316 Jung

Fig 4. A, Tracing of pretreatment lateral cephalogram. B, Tracing of posteroanterior cephalogram. The

mandibular ramus (Co, condylion; Ag, antegonion) and body (Ag; menton, Me) length, the distance
from Ag to the midsagittal reference line (MSL; CG, crista galli; ANS, anterior nasal spine), and MSL
to Me distance were measured (Table I). C, The angle between soft tissue nasion–subnasale line
(NSL, red) and interpupillary line (yellow) was 92.0 . The angle between NSL and commissure line
(blue) was 87.4 . D, Upper dental midline was located 1.5 mm to the right side of NSL (blue).

Table II. Changes in cephalometric measurements

Pretreatment Posttreatment Postretention
Bjork sum ( ) 399.1 397.3 397.6
Facial height ratio (%) 60.5 61.7 61.6
ANB ( ) 0.6 0.6 0.5
A to N perpendicular (mm) 1.6 1.6 1.6
Pog to N perpendicular (mm) 2.9 1.2 1.2
U1 to FH ( ) 109.8 105.7 105.7
U1 to SN ( ) 100.1 96.0 95.9
L1 to A pog (mm) 4.1 0.2 0.3
IMPA ( ) 83.2 79.7 79.5
Interincisal angle ( ) 135.9 145.7 145.8
Nasolabial angle ( ) 102.2 103.5 103.8
Upper lip to Esthetic line (mm) 0.5 3.1 3.4
Lower lip to Esthetic line (mm) 2.1 1.4 1.4

Note. Bjort sum, Saddle angle + articular angle + gonial angle.

August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Jung 317

Fig 5. Extraction of the 2 root rest (green, maxillary right second premolar and mandibular left
second molar) was unavoidable. To correct crowding and midline deviation, maxillary left second pre-
molar and mandibular right first premolar extractions (red) were required. To reinforce the anchorage,
OMIs were planned in the lower right posterior and lower left anterior alveolar bone. Green arrow, the
direction of tooth movement; dotted line, nasion–subnasale line.

Fig 6. Posttreatment photographs.

American Journal of Orthodontics and Dentofacial Orthopedics August 2021  Vol 160  Issue 2
318
Fig 7. Posttreatment dental casts.
TREATMENT PROGRESS
Because the patient wanted to use ceramic brackets,
0.022-in preadjusted edgewise ceramic brackets (Clarity;
3M Dental Products, Monrovia, Calif) were placed on
the labial surface of the whole dentition. The archwire
sequence progressed from 0.014-in nickel-titanium wire
to 0.019 3 0.025-in stainless steel working wire. Acrylic
bonded bite planes were used in the central incisors to pre-
vent overbite change and minimize the occlusal interfer-
ence during uprighting and protraction of the mandibular
left third molar. Occlusal equilibration in the third molar
was also performed during leveling. Root rest and premolar
extractions were conducted during initial leveling.
At the final stage of leveling, 2 OMIs (Mplant U2;
diameter, 1.5 mm; length, 7.1 mm; BioMaterials Korea,
Inc, Seoul, South Korea) were placed on the buccal alve-
olar bone between the mandibular right first and
second molars and between the mandibular left canine
and first premolar. After a working wire was placed, a
force was applied to close all the extraction space.
During protraction of the mandibular left third
molar, tip backbend, and low-level force (about 100 g
of force),9 were used to minimize mesial tipping. After
correcting the midline and closing all extraction spaces,
archwire adjustments were conducted for finishing and
detailing purposes. To prevent lingual crossbite on the
left side, additional torque was given in the lower left
posterior region. In addition, a narrow mandibular arch-
wire and a wide maxillary archwire were used. The
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Jung
number of missed appointments was 3, and the total
treatment duration was 30 months.
Before bracket removal, maxillary and mandibular
lingual fixed retainers were placed. After bracket
removal, the patient was given a circumferential retainer
for the maxillary arch and a Hawley retainer for the
mandibular arch. The instructions to the patient
included full-time retainer use for 6 months after de-
bonding and then nights only for 2 years.
TREATMENT RESULTS
Posttreatment intraoral photographs (Fig 6) showed
adequate overbite and overjet, a Class I canine relation-
ship, corrected midline deviation, and well-aligned teeth.
Mild gingival inflammation and white spot lesions result-
ing from poor oral hygiene were observed. The maxillary
right first molar showed gingival recession because of
pretreatment bone recession in this area (Figs 7 and 8).
The patient was recommended to visit a periodontist
for consultation. Lip incompetency was improved, and
appropriate canine guidance was obtained. All the
extraction spaces were successfully closed. On the left
side, a Class II molar relationship was achieved. Maxillary
molars moved forward by the extraction space closure,
but the intermolar width was maintained (51.2 mm) us-
ing a wide form archwire. In the mandibular arch, inter-
premolar width (second premolar, 40.9 mm to 39.6 mm)
and intermolar width (44.6 mm to 44.3 mm) were
decreased by additional torque and the narrow archwire.
Jung
Fig 8. Posttreatment radiographs.
On the cephalometric superimposition tracing, the
incisors were retracted, and the upper and lower lip pro-
files were changed because of incisor retraction (Fig 5,
A). The mandibular plane was closed about 1.8 , and
Pogonion moved forward about 1.7 mm.
Although the amount of the menton deviation was
not changed significantly in the PA cephalogram (Fig 9,
B; 4.7-4.6 mm), lip canting was decreased (Fig 9, C),
and the dental midline was improved (Fig 9, D). To
evaluate lip asymmetry, according to a previous study,4
using a vertical line from the midpoint of the base of
the nose (the middistance between the inner outline
of the nostrils) through the midpoint of the philtrum,
lower lip asymmetry was evaluated by the percentage
of area of the right and left lower lip. Pretreatment
American Journal of Orthodontics and Dentofacial Orthopedics
319
lower left segment area was 53.20% (Fig 9, E) and
posttreatment lower left segment area was 50.57%
(Fig 9, F).
After 4 years and 6 months of retention, she re-
turned for a check-up. The occlusion was stable,
and a slight relapse of the midline deviation was
observed (Figs 10-12). Although a mild gingival
recession was noted, the overbite, overjet, and
buccal occlusion were well-maintained. In the lateral
cephalogram, there were no significant changes dur-
ing the retention period (Table II).
DISCUSSION
On the lower right side, a large amount of mandibular
incisor movement was required to treat midline
August 2021  Vol 160  Issue 2
320 Jung

Fig 9. A, Craniofacial, maxillary, and mandibular superimposition composite tracing of lateral cephalo-
grams at pretreatment (black) and posttreatment (red). B, Midsagittal reference line (MSL) to menton
(Me) distance was not changed significantly (4.7-4.6 mm). C, After treatment, the angle between NSL
(red) and interpupillary line (yellow) was 92.1 , and the angle between NSL and commissure line (blue)
was 89.6 . Lip canting was decreased. D, The upper and lower midline coincides, and it fits well with
NSL. E, To evaluate the lip asymmetry, using the midpoint of the base of the nose–philtrum line,4 the left
and right areas of the lower lip were measured. Pretreatment lower left segment area (yellow) showed
53.20%. F, Posttreatment lower left segment area (yellow) was 50.57%.

deviation and crowding. Absolute anchorage was
needed. On the contrary, the lower left side required
third molar protraction without distal movement of
the first molar. After introducing OMIs as an orthodontic
anchorage source, many studies have been published
regarding their effects in various procedures, including
incisor retraction,10,11 whole-arch retraction,12,13 intru-
sion,14 and molar protraction.15
In this patient, the OMIs were positioned between the
mandibular right first and second molars and between
the mandibular left canine and first premolar. On the
right side, the buccal shelf area was used because this
area does not have the risk of root contact,16,17 whereas
interradicular alveolar bone was used on the left side.
Mandibular anterior teeth retraction and third molar
protraction were completed without significant
anchorage loss. The severely deviated dental midline
was successfully corrected without jaw surgery. Without
the use of OMIs, the success of such tooth movements is
unlikely.
Protraction of the second molar into the first molar
extraction space has been reported several times,7,15,18
but protraction of the third molar into the
second molar extraction space is rare. Regardless, the
possible problems during molar protraction are similar
and include tipping, vertical change, and width change.
Such side effects occur because the force vector
cannot pass the center of resistance. When pulling the
second molar forward, some patients have sufficient
attached gingiva to easily position the force near the

August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Jung
Fig 10. Postretention photographs, 4.5 years into retention.
center of resistance. Because such treatment is impos-
sible in third molar protraction, molar tipping or
whole-arch rotation can be produced (Fig 13, A).19
Because the occlusal force acting on the posterior teeth
helps prevent molar tipping and extrusion if you use a tip
backbend and light force, side effects can be prevented
in most patients.
On the lower right side, incisor retraction using OMI
was performed. Because such a movement sometimes
produces a counterclockwise rotation (Fig 13, B),20 the
acrylic bite planes were bonded in central incisors (the
so-called bite turbo) at the initial leveling stage at the
level of the pretreatment overbite.20 Biteplanes also
reduce occlusal interference that can occur during molar
uprighting.
Usually, premolar extraction treatment does not
change the mandibular plane significantly, but some
of the patients may show mandibular plane rotation. A
recent systematic review showed that in extraction
American Journal of Orthodontics and Dentofacial Orthopedics
321
patients, the standard deviation of the mandibular plane
angle change during treatment was from 1.7 to 9.8 .21
In this patient, about 1.8 of counterclockwise mandib-
ular plane rotation was observed. The exact cause is un-
known, but it seems necessary to remember that slight
rotational changes may occur during treatment.
When the current patient refused to undergo jaw sur-
gery, she was informed that her facial asymmetry could
not be improved without such a procedure. However,
after debonding, a significant improvement of facial
asymmetry was observed. Given that orthodontic treat-
ment cannot change the size or shape of the jaw bone,
it is clear that these improvements are not because of
skeletal changes.
To correct the crossbite on the left side, the mandib-
ular canine and premolars moved inward, and maxillary
premolars moved outward by archwire adjustment. In
this regard, one can easily see the difference in arch
form by comparing the pretreatment and posttreatment
August 2021  Vol 160  Issue 2
322
Fig 11. Postretention dental casts.
occlusal intraoral photographs (Figs 1 and 6). Previous
research showed that correction of the unilateral cross-
bite improved lip asymmetry.3 In this study, the lower
lip surface area of the crossbite side was significantly
improved (from 56.85% to 52.12%) by orthodontic
treatment. The improved amount of lip asymmetry was
smaller (from 53.20% to 50.57%) in this patient, but it
seems to be because the lip asymmetry before treatment
was smaller. As the retraction of incisors produces lip
vermilion height decrease,22,23 outward position of teeth
in the crossbite side can make the lower lip larger, the
asymmetry more intense, and conversely, an inward po-
sition will make the lower lip smaller. It can be assumed
that the improvement of asymmetry by orthodontic
treatment can only be found in patients with unilateral
crossbite and mainly in lip vertical measurements. A
slight improvement in lip canting was also shown in
this patient, which is also thought to be due to vertical
lip changes.
Patient final occlusion can be established as a Class II
molar relationship. To obtain a proper Class I molar rela-
tionship, maxillary molar tubes usually have a 10 first-
order (molar offset) prescription.24 However, to acquire a
Class II molar relationship, 0 molar offset is necessary to
provide good occlusion.
Because the mandibular second molar tube of MBT
prescription has a 0 first-order bend and 10 torque,25,26
it can be an adequate substitute for the maxillary molars in
Class II molar finishing in patients. When different tubes
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Jung
are required for the right and left molars, more attention
should be paid to the bonding procedure to prevent poor
posttreatment occlusion or longer treatment because of re-
bonding. The posttreatment maxillary occlusal intraoral
photograph shows 10 molar offset on the right side and
0 molar offset on the left side.
CONCLUSIONS
This patient demonstrated successful correction of
midline discrepancy and third molar protraction using
OMI anchorage. Careful adjustment and use of appro-
priate biomechanics can successfully achieve a large
amount of asymmetric tooth movement. Correction of
the crossbite in the buccal segment may improve facial
asymmetry by changing the overlying lips.
REFERENCES
1. Vig PS, Hewitt AB. Asymmetry of the human facial skeleton. Angle
Orthod 1975;45:125-9.
2. Schwartz HC. Efficient surgical management of mandibular asym-
metry. J Oral Maxillofac Surg 2011;69:645-54.
3. Gazit-Rappaport T, Gazit E, Weinreb M. Quantitative evaluation of
lip symmetry in skeletal asymmetry. Eur J Orthod 2007;29:345-9.
4. Gazit-Rappaport T, Weinreb M, Gazit E. Quantitative evaluation of
lip symmetry in functional asymmetry. Eur J Orthod 2003;25:
443-50.
5. Young KA, Melrose CA, Harrison JE. Skeletal anchorage systems in
orthodontics: absolute anchorage. A dream or reality? J Orthod
2007;34:101-10.
Jung
Fig 12. Postretention radiographs and facial midline.
6. Jung MH. Asymmetric extractions in a patient with a hopeless maxil-
lary central incisor, followed by treatment with mini-implant
anchorage. Am J Orthod Dentofacial Orthop 2018;153:716-29.
7. Baik UB, Kook YA, Bayome M, Park JU, Park JH. Vertical eruption
patterns of impacted mandibular third molars after the mesialization
of second molars using miniscrews. Angle Orthod 2016;86:565-70.
8. Okeson JP. Orthodontic therapy and the patient with temporo-
mandibular disorder. In: Graber TM, Vanarsdall RLJ, Vig WL, edi-
tors. Orthodontics: Current Principles and Techniques. 4th ed. St
Louis: Elsevier; 2005. p. 331-44.
9. Theodorou CI, Kuijpers-Jagtman AM, Bronkhorst EM,
Wagener FADTG. Optimal force magnitude for bodily orthodontic
tooth movement with fixed appliances: a systematic review. Am J
Orthod Dentofacial Orthop 2019;156:582-92.
American Journal of Orthodontics and Dentofacial Orthopedics
323
10. Park CS, Yu HS, Cha JY, Mo SS, Lee KJ. Effect of archwire stiffness
and friction on maxillary posterior segment displacement during
anterior segment retraction: a three-dimensional finite element
analysis. Korean J Orthod 2019;49:393-403.
11. Barthelemi S, Desoutter A, Souare F, Cuisinier F. Effectiveness of
anchorage with temporary anchorage devices during anterior
maxillary tooth retraction: A randomized clinical trial. Korean J Or-
thod 2019;49:279-85.
12. Jung MH. A comparison of second premolar extraction and mini-
implant total arch distalization with interproximal stripping. Angle
Orthod 2013;83:680-5.
13. Bechtold TE, Kim JW, Choi TH, Park YC, Lee KJ. Distalization
pattern of the maxillary arch depending on the number of ortho-
dontic miniscrews. Angle Orthod 2013;83:266-73.
August 2021  Vol 160  Issue 2
324 Jung

Fig 13. A, Because the pulling force vector cannot pass the center of resistance (CR) of the third molar,
molar tipping or whole-arch rotation can be produced during protraction. Occlusal force reduces the
possibility of these side effects. Blue circle, CR of the mandibular arch; red circle, CR of mandibular
molar; green arrow, force and moment acting on teeth by protraction force; blue arrow, moment by
tip backbend; yellow arrow, occlusal force. B, Incisor retraction using OMI may cause counterclockwise
rotation of the mandibular occlusal plane. Anterior biteplane reduces the possibility of bite deepening.
Blue circle, CR of the mandibular arch; red circle, CR of mandibular anterior teeth; green arrow, force
and moment acting on teeth by retraction force; yellow arrow, occlusal force by anterior biteplane.

14. Jung MH. Vertical control of a Class II deep bite malocclu-
sion with the use of orthodontic mini-implants. Am J Or-
thod Dentofacial Orthop 2019;155:264-75.
15. Baik UB, Chun YS, Jung MH, Sugawara J. Protraction of mandib-
ular second and third molars into missing first molar spaces for a
patient with an anterior open bite and anterior spacing. Am J Or-
thod Dentofacial Orthop 2012;141:783-95.
16. Nucera R, Lo Giudice A, Bellocchio AM, Spinuzza P, Caprioglio A,
Perillo L, et al. Bone and cortical bone thickness of mandibular buccal
shelffor mini-screwinsertionin adults. Angle Orthod 2017;87:745-51.
17. Jung MH. Total arch distalization with interproximal stripping in a
patient with severe crowding. Korean J Orthod 2019;49:194-201.
18. Kravitz ND, Jolley T. Mandibular molar protraction with temporary
anchorage devices. J Clin Orthod 2008;42:351-5: quiz 40.
19. Jung MH. Occlusal plane rotation by molar protraction using ortho-
dontic mini-implant. Clin J Korean Assoc Orthod 2019;9:83-96.
20. Jung MH, Kim TW. Biomechanical considerations in treatment
with miniscrew anchorage. Part 1: the sagittal plane. J Clin Orthod
2008;42:79-83.
21. Kouvelis G, Dritsas K, Doulis I, Kloukos D, Gkantidis N. Effect of
orthodontic treatment with 4 premolar extractions compared
with nonextraction treatment on the vertical dimension of the
face: a systematic review. Am J Orthod Dentofacial Orthop 2018;
154:175-87.
22. Perkins RA, Staley RN. Change in lip vermilion height during ortho-
dontic treatment. Am J Orthod Dentofacial Orthop 1993;103:
147-54.
23. Liu ZY, Yu J, Dai FF, Jiang RP, Xu TM. Three-dimensional changes in
lip vermilion morphology of adult female patients after extraction and
non-extraction orthodontic treatment. Korean J Orthod 2019;49:
222-34.
24. Andrews LF. Straight Wire: The Concept and Appliance. San Diego,
CA: LA Wells; 1989.
25. Bennett JC, McLaughlin RP. Orthodontic Treatment Me-
chanics and the Preadjusted Appliance. St Louis: Mosby;
1993.
26. McLaughlin RP, Bennett JC, Trevisi HJ. Systemized Orthodontic
Treatment Mechanics. St Louis: Mosby; 2001.

August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
LITIGATION AND LEGISLATION
Gumballs
Laurance Jerrold
Woodbury, NY
Y
ou know what gumballs are, right? They're those
round things that you get out of a vending ma-
chine that tastes great for about 30 seconds, but
if you chew on them long enough can pull out your fill-
ings or extract deciduous teeth. Well, they're a nuisance
of another type, too, as was noted in Henderson v St.
Francis Community Hospital, 399 SE2d 767 (Sup. Ct.,
S. Car.; Oct 22, 1990). At trial, the plaintiff was awarded
$75,000 for the injuries she sustained, at which point the
defendants moved for a directed judgment in their favor,
notwithstanding the verdict. The trial judge granted the
motion. The plaintiff now appealed, and the appeals
court affirmed the ruling in favor of the defendant.
The plaintiff appealed again, this time to the supreme
court.
The facts of the case reveal the following. The plain-
tiff went to visit her friend in the hospital. On the way
back to her car, she stepped on an accumulation of sweet
gumballs. The unsteady footing caused her ankle to
twist; hence she fell, the result of which was that she
fractured her wrist and sustained other lesser injuries.
It seems that some 20 years earlier when the parking
lot was designed, sweet gum trees were planted next
to the walkway. Roughly 10 years later, an additional
parking lot was added. The design firm at that time rec-
ommended that the gum trees be removed because the
debris they exfoliated created a cleanup nuisance. The
hospital decided not to undertake the architectural
recommendation. In the end, a stairway was created be-
tween the 2 parking lots, which ran adjacent to the gum
trees. Enter the plaintiff, followed by the falling and
accumulation of gumballs, leading to both the injuries
and the lawsuit.
Section 343 of the Restatement (Second) of Torts
(West Publishing, 1965), entitled “Dangerous Conditions
Known to or Discoverable by Possessor,” notes that one
who possesses land that is subsequently used by an
invitee is subject to liability for any physical harm caused
to that invitee by a condition on the land if and only if
the landowner or possessor
Associate Editor for Litigation and Legislation, Woodbury, NY
Am J Orthod Dentofacial Orthop 2021;160:325-6
0889-5406/$36.00
Ó 2021 by the American Association of Orthodontists. All rights reserved.
https://doi.org/10.1016/j.ajodo.2021.05.002
(a) knows or by the exercise of reasonable care would
discover the condition, and should realize that it in-
volves an unreasonable risk of harm to such invitees,
and (b) should expect that they will not discover or
realize the danger or will fail to protect themselves
against it, and (c) fails to exercise reasonable care
to protect them against the danger.
In other words, the hospital, as the operator of the
parking lot, although not an insurer of the safety of
those using the lot, is under an obligation (has a duty)
to keep the premises in a relatively safe condition. The
degree of care required to maintain this level of safety
has to be commensurate with the particular circum-
stances relating to its use. It also has to take into account
the characteristics of the invitee, such as whether or not
the person has any infirmities that might inhibit the
invitee's ability to protect himself.
The evidence presented at trial reflects that gum trees
produce little gumballs that are hard and round and
eventually fall from the tree. Once on the ground, they
take years to deteriorate and are dangerous to pedes-
trians because they create an unevenness to the ground
and, hence, result in unsafe footing. The evidence also
showed that when the additional parking lot was added,
the hospital was advised to remove the trees because
they produced this debris, and the hospital, aware of
this situation, did not have a maintenance program to
remove the nuisance.
Given the evidence presented, it was not unreason-
able for a jury to decide that the hospital breached its
duty of care owed to the plaintiff and that it should be
held liable for her injuries, which is exactly what they
found. The court noted that it was a mistake on the
part of the trial judge and the appellate court to grant
the defendant's motion that the verdict be disregarded
and the judgment set aside. The supreme court reversed
the lower courts and remanded the case for a new trial.
COMMENTARY
Look, we all live in the real world, and the real world is
not perfect. Let's look at some of these imperfections as
they might apply to some of us. Many of you own your
own free-standing office building that has an adjacent
parking lot. What is the condition of the lot? Are there
325
326
any small cracks or potholes that may cause an invitee
(patient, parent, grandparent, or whomever) to lose
footing or twist an ankle, subsequently slip and fall,
resulting in an injury?
Maybe there is a stairway, even just a stoop, or an in-
clined walkway that in icy conditions could prove haz-
ardous. Is there a railing? Maybe, it's the walkway or
sidewalk from the parking lot or street parking which
you are ultimately responsible for that has an uneven
surface that you have been made aware of. If you own
a corner lot, maybe it’s the big bush that blocks a motor-
ist's view of another driver or cyclist coming from a
different direction that is the potential source of liability.
Move inside for a minute. Got worn or loose
carpeting that may contribute to a trip? No, not that
kind of trip; the one that lands someone in a hospital.
Maybe there is a little dollop of water spilled on a tile
floor in a hallway that creates a slippery condition. It
could be that there is a dental chair that is not locked
in position resulting in a not so minor slip and fall.
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
Litigation and legislation
Liability might also result from the lack of safety goggles
for the visitor in the operatory. The list goes on.
We are a personal service business. Our places of
business are places of public accommodation and are
open to all who are invited to enter our premises.
Whether we own or rent, we owe those invites the
assurance that they will not suffer an injury because
we have allowed unsafe conditions that we know, or
should know of, to exist within those premises. If we
are aware of such conditions but cannot immediately
mitigate them, we have a duty to advise those invitees
to take heed of and use special care concerning their
use of those areas. Rope it off, put up stations or
post signs, create other paths of ingress and egress.
You see these mitigations all the time in supermarkets,
parking lots, walkways, you know, all those places in
the real world.
We are exposed to potential liability on many fronts.
Professional negligence, malpractice, is just one. Pre-
mises liability is another biggie. No gumballs for me.
INTERNATIONAL ITEMS OF INTEREST
A recently released report by a major
scientific academy proposes significant
changes in understanding and managing
temporomandibular disorders
Charles S. Greene,a John W. Kusiak,b Terrie Cowley,b
and Allen W. Cowley, Jr.c
Chicago, Ill, and Milwaukee, Wis
A recent consensus study report1 of the National Aca-
demies of Sciences, Engineering, and Medicine (NASEM)
recommends that the field of temporomandibular disor-
ders (TMDs) must be totally changed in terms of bio-
medical research, professional education and training,
and patient care. Recent research demonstrates that
TMDs are complex multisystem disorders, which points
to the need for a different, patient-centered, interprofes-
sional approach to TMD research and treatment. There-
fore, traditional dental-centric approaches to research
and treatment of TMDs must be modernized to align
with insights gained from new scientific discoveries.
The NASEM Study Report was produced by a committee
on “Temporomandibular Disorders: Priorities for Re-
search and Care (2020)” to address improved care and
new research directions for TMD. This announcement
summarizes the NASEM committee's recommendations
and reports on ongoing efforts to implement these rec-
ommendations. These efforts are led by the TMJ Associ-
ation (TMJA), a nonprofit patient advocacy
organization, the National Institutes of Health (NIH),
and the TMJ Patient-Led RoundTable under the auspi-
ces of the United States Food and Drug Administration
and its Medical Device Epidemiology Network (MDEpi-
Net).
In March 2020, the committee released a consensus
study report1 which included 11 recommendations cov-
ering research, treatments, training, and education. The
report proposed near-term and medium-to long-term
recommendations to address gaps and opportunities in
each area.
The first 4 recommendations focus on developing a
national research consortium and setting priorities for
basic and translational research, public health research,
a
University of Illinois at Chicago, Chicago, Ill.
b
The TMJ Association, Milwaukee, Wis.
c
Medical College of Wisconsin, Milwaukee, Wis.
Address correspondence to: Charles S. Greene, University of Illinois at Chicago,
College of Dentistry, Department of Orthodontics, 1041 Ridge Rd, Apt 304,
Wilmette, IL 60091; e-mail, cgreene@uic.edu.
Am J Orthod Dentofacial Orthop 2021;160:327-8
0889-5406/$36.00
Ó 2021 by the American Association of Orthodontists. All rights reserved.
https://doi.org/10.1016/j.ajodo.2021.05.003
the burden of disease, and bolstering clinical research;
all 4 priorities form the basis for improving patient-
centered care. In response, the NIH and the National In-
stitute of Dental and Craniofacial Research established
the temporomandibular joint disorders (TMJs) multi-
council working group to review the NASEM report
and the recommendations, and then to develop strat-
egies for NIH to better support research efforts in this
area. The work of this group is underway. The MDEpiNet
Patient-Led RoundTable developed a research plan that
is in accord with the NASEM report. In a patient-driven
initiative, members of the TMJA are seeking additional
funds for TMJ research.
Recommendations 5 and 6 aim to improve the quality
of care for TMD patients through improved disease risk
assessment and stratification, diagnostics, and dissemi-
nation of clinical practice guidelines and metrics of
care. A current focus of the Food and Drug Administra-
tion and its MDEpiNet Initiative is the Coordinated
Registry Network, which gathers real-world evidence
data on patients’ health status and care in several linked
registries to be used in health care decision making and
postmarket monitoring of approved devices and other
treatments. Acting as the catalyst to advance the needs
of patients with TMJ implants, the TMJA developed
the TMJ Patient-Led RoundTable—the first patient-
centered, public-private collaboration among the federal
government, scientists, clinicians, dentists, advocates,
manufacturers, and others. The patient-led registry for
TMD that will be part of the Coordinated Registry Net-
work and addresses recommendations 5 and 6 by pro-
viding a large dataset to be used in determining risk
assessments for various TMD treatments and establish-
ing clinical guidelines for the care of patients suffering
from TMD disorders.
Recommendation 7 focuses on improved reimburse-
ment and access to assessment, treatment, and manage-
ment of TMDs. These recommendations will be
addressed in a more long-term fashion, pending out-
comes of results from other recommendations.
Recommendations 8-10 are centered on improving
TMD patient treatments and propose to develop “Cen-
ters of Excellence for TMDs and Orofacial Pain Treat-
ment” (recommendation 8), improve professional
school education (recommendation 9), and expand spe-
cialized continuing education for health care providers
(recommendation 10). To improve dental education,
the TMJA and the American Association of Orofacial
Pain recommended to the Committee on Dental Accred-
itation that TMJ must be included in the dental school
curriculum. That recommendation was approved and
will be implemented in 2022, and as a result, the
TMJA and the American Association of Orofacial Pain
327
328
are developing TMD predoctoral core curriculum out-
lines. Separately, the TMJA has established a Working
Group on Interprofessional Models of TMD Care. This
group is exploring ways to develop a new multidiscipli-
nary model of TMD care involving expertise across med-
icine, dentistry, nursing, physical therapy, psychological
therapies, and other relevant areas of health care.
Recommendation 11 addresses patient education
and awareness about TMDs and reducing the stigma
of the disease. The NASEM recommends that the
TMJA and TMJ Patient-Led RoundTable members
work together with the American Medical Association
Education group, American Dental Education Associa-
tion, and the National Institute of Dental and Craniofa-
cial Research Office of Communications and Health
Education to develop educational materials for TMDs,
based on the current understanding of this disorder as
summarized in the NASEM report. These materials will
include brochures, videos, and virtual educational work-
shops addressing many aspects of TMD management
and care, access to quality treatments, and approaches
to stigma reduction.
The actions described above are only the first steps
in a major effort to carry out a much-needed paradigm
shift directed toward TMD research, treatment, and
education. The NASEM report and its recommenda-
tions are a direct call to the biomedical research and
health care communities to advance 21st-century
science-based research and treatments of TMDs. The
new research results strongly suggest that the current
August 2021  Vol 160  Issue 2 American Journal of Orthodontics and Dentofacial Orthopedics
International Items of Interest
dental-focused treatments for TMDs must be recon-
ceived toward a multidisciplinary, interprofessional
team approach involving specialists within the broader
medical community.2 The treatments must be patient-
centered and evidence-based, and when necessary,
the use of any implanted devices must undergo
rigorous premarketing evaluation and postmarketing
surveillance.
New models of disease treatment and novel research
hypotheses are desperately required to revolutionize the
scientific and clinical approaches toward these condi-
tions. Research expertise not presently represented in
the TMD portfolio is essential to uncover new infor-
mation that will form the basis for interprofessional
care, clinical guidelines, and disease and treatment
risk assessments. Equipped with these new concepts,
multidisciplinary teams of health care providers will
be able to diagnose, treat, and manage TMDs in a
professional, patient-centered, and caring way that
will improve the health and lives of TMD patients.
REFERENCES
1. National Academies of Sciences. Engineering, and Medicine. Tem-
poromandibular Disorders: Priorities for Research and Care. Wash-
ington, DC: The National Academies Press; 2020. Available at:
https://www.nap.edu/catalog/25652/temporomandibular-disorders-
priorities-for-research-and-care.
2. Greene CS, Manfredini D. Treating temporomandibular disorders in
the 21st century: can we finally eliminate the “Third Pathway”? J
Oral Facial Pain Headache 2020;34:206-16.
CONTINUING EDUCATION
August 2021
Instructions: To submit your answers to this test and earn 3 hours of CE
credit, go to www.aaoinfo.org and log in as a member. Select the Edu-
cation tab, then AJO-DO Tests. Purchase the current test, or the test for
any issue published in the preceding 12 months. The fee for each test is
$20. You will take the test online but can download and print a PDF
version of the test. Results are tabulated immediately. To earn 3
AUGUST 2021 LEARNING OBJECTIVES:
After completing this course, the participant will have:
1. An understanding of the mechanical behavior of a newly designed
closing loop using the finite element method to evaluate.
2. Familiarity with the use of the Liou-alternate rapid maxillary expan-
sion/constriction (Alt-RAMEC) technique in patients with skeletal
Class III malocclusions.
3. An appreciation of how laypeople and orthodontists view alterations
in the profile of a female subject with retrognathia.
4. Awareness of the accuracy of virtual bracket removal (VBR) tech-
niques in fabricating orthodontic retainers from 3-dimensional-
printed models.
Article 1: Simulation of orthodontic tooth movement during
activation of an innovative design of closing loop using the
finite element method, by Tuan Nguyen Anh et al
1. The objectives of this study were to clinically evaluate orthodontic
tooth movement during the activation of a newly designed closing
loop combined with a gable bend and to investigate the optimal
activation conditions to achieve the desired tooth movement.
1. True
2. False
2. The basic design of the closing loop examined in this study was a
teardrop that was 10 mm in height and 2 mm in width, bent from
a 0.019 3 0.025-in stainless steel archwire.
1. True
2. False
3. The authors reported that a gable bend of 15 in the archwire
demonstrated lingual root tipping of the central incisor of 1.14 .
1. True
2. False
4. The authors concluded that torque control of the anterior teeth
and anchorage control of the posterior teeth could be carried
out effectively and simply by reducing by half the thickness of
the teardrop loop as described to a distance of 3 mm from its
apex and by incorporating various degrees of a gable bend into
the loop.
1. True
2. False
Article 2: Long-term follow-up of late maxillary orthopedic
advancement with the Liou-Alternate rapid maxillary expan-
sion/constriction technique in patients with skeletal Class III
malocclusion, by Maria Costanza Meazzini et al
CE Editor: Dr Allen H. Moffitt
hours of CE credit, you must answer 75% of the questions correctly.
If you do not receive a passing score the first time, you can take the
test again, free of charge, until you pass. Upon successful completion
of the test, your CE credits will be added to the AAO's online CE
Credit Manager, and you can use the CE Credit Manager to print a
certificate.
5. The objective of this study was to evaluate short and long-term re-
sults of the application of the Alt-RAMEC technique in patients with
skeletal Class III malocclusion.
1. True
2. False
6. In the experimental group, after completion of the expansion/
constriction cycles, maxillary protraction was delivered using a
reverse pull headgear appliance.
1. True
2. False
7. The authors reported an average maxillary advancement for the
experimental study group of 5.9 mm 1 1.7 mm.
1. True
2. False
8. The authors concluded that the Liou-Alt-RAMEC technique should
be performed before the pubertal growth spurt.
1. True
2. False
Article 3: Attractiveness assessment by orthodontists and
laypeople judging female profile modifications of Class II
Division 1 malocclusion, by Kadriye Kalin et al
9. This study aimed to investigate and compare the judgments of
laypeople and orthodontists on the effect of different soft tissue al-
terations on the profile of a female patient with a Class II Division 1
malocclusion with mandibular retrognathia, produced by software
simulated camouflage or mandibular advancement treatment.
1. True
2. False
10. The study population comprised 162 laypeople and 44 orthodon-
tists.
1. True
2. False
11. The authors reported that laypeople found the camouflage treat-
ment less attractive than did the orthodontists.
1. True
2. False
12. The authors concluded that the untreated baseline profile of a fe-
male subject with mandibular retrognathia was judged least attrac-
tive by both laypeople and orthodontists.
1. True
2. False
329.e1
Article 4: Three-dimensional assessment of virtual bracket
removal for orthodontic retainers: A prospective clinical
study, by Kaitlin Marsh et al
13. This prospective clinical study aimed to validate a novel technique
for virtual bracket removal (VBR) in-office, comparing its retainer
fabrication accuracy to traditional bracket removal with impressions
and 2 orthodontic laboratories that use VBR for retainer fabrication.
1. True
2. False
14. All measurements for VBR accuracy assessment were performed by
2 separate investigators using the vector analysis module to ensure
interexaminer reliability.
1. True
2. False
329.e2
15. The authors reported a statistically significant difference
between the 2 laboratories, with Orthodent Laboratory
showing lower accuracy than New England Orthodontic
Laboratory.
1. True
2. False
16. The authors concluded that the VBR techniques using the
in-office Meshmixer VBR protocol or the 2 orthodontic labora-
tories were considered accurate enough for clinical use of
orthodontic retainers fabricated from 3-dimensional-printed
models.
1. True
2. False
DIRECTORY
AAO Officers and Organizations
American Association of Orthodontists
Annual Session June 25-27, 2021,
Virtual
President, J Kendall Dillehay,
Wichita, KS
President-Elect, Norman J. Nagel,
Simi Valley, CA
Secretary-Treasurer, Myron D. Guymon,
Logan, UT
Speaker of the House, Dennis C. Hiller,
Thornton, NH
Editor-in-Chief, Rolf G. Behrents,
Overland Park, KS
Executive Director, Lynne Thomas Gordon,
401 N Lindbergh Blvd,
St Louis, MO 63141;
telephone, 800-424-2841; e-mail,
lthomasgordon@aaortho.org
Great Lakes Association of Orthodontists
President, Anthony Puntillo, Crown Point, IN
President-Elect, Valerie Martone, Beaver, PA
Vice President, J. Martin Palomo, Cleveland,
OH
Secretary-Treasurer, Scott Schulz, Traverse
City, MI
Trustee, Michael W. Sherman, Thornhill, ON,
Canada
Executive Director, Debbie Nunner, 400 W
Wilson Bridge RD Ste 120, Worthington, OH
43085; telephone, 877-274-6420; fax, 614-
221-5720; e-mail, debbie@assnoffices.com
Middle Atlantic Society of Orthodontists
President, Francis Picon, San Juan, Puerto Rico
President-Elect, Madeleine Goodman,
Potomac, MD
Secretary, Alireza O. Rajael, Waldorf, MD
Treasurer, Dave Harmon, Mitchellville, MD
Trustee, Steven Siegel, Glen Burnie, MD
Executive Director, Jane Treiber, 400 W.
Wilson Bridge Rd, Ste 120, Worthington,
OH 43085; telephone, 866-748-6276;
fax, 614-221-1989; email,
Jane@assnoffices.com
Midwestern Society of Orthodontists
President, Scott Arbit, Mequon, WI
President-Elect, Mark Dake, West Plains,
MO
Secretary-Treasurer, Nellie Kim-Weroha,
Rochester, MN
Trustee, Michael G. Durbin, Des Plaines, IL
Executive Director, Kristi Burmeister, 2131
Meadow Valley Dr, Innsbrook, MO 63390;
telephone, 636-745-3008; fax, 636-745-
3012; e-mail, association.info@icloud.com
Northeastern Society of Orthodontists
President, Dan Stuart, Dartmouth, Nova Scotia
President-Elect, Kenneth Webb, Westborough,
MA
330
Secretary, Lisa Howard, Scarborough, ME
Treasurer, Mariana De Deus Haughey,
Lansing, MI
Trustee, John Callahan, Syracuse, NY
Executive Director, Kristen E. Dunn,
2929 Arch St, Ste 1700, Philadelphia, PA
19104; telephone, 1-800-981-0476; e-mail,
info@neso.org
Pacific Coast Society of Orthodontists
President, Marie Lathrop, Portland, OR
President-Elect, Tom Merrill, East Wenatchee,
WA
Secretary-Treasurer, John Wachtel,
Scottsdale, AZ
Trustee, Norman J. Nagel, Simi Valley, CA
Executive Director, Callie Castro, CAE,
15621 W 87th St. #267, Lenexa, KS 66219;
telephone, 833-621-7276; e-mail,
ccastro@pcsortho.org
Rocky Mountain Society of Orthodontists
President, Anil Idiculla, Lone Tree, CO
President-Elect, Daniella Phillis, Durango, CO
Vice-President, Kristen Lowe, Aurora, CO
Secretary-Treasurer, Morris L. Poole, Logan,
UT
Trustee, Myron D. Guymon, Logan, UT
Executive Director, Jennifer Bennett, 15621
W 87th St #267, Lenexa, KS 66219;
telephone/fax, (833) 226-7676; e-mail,
info@rmso.org
Southern Association of Orthodontists
President, Sims Tompkins, Columbia, SC
President-Elect, Mark W. Dusek, Savannah, GA
Secretary-Treasurer, Preston Miller, Jackson,
TN
Trustee, Richard A Williams, Southaven, MS
Executive Director, Heather Hunt, 32 Lenox
Pointe NE, Atlanta, GA 30324-3169;
telephone, 404-261-5528; fax 844-214-1224;
e-mail, hhunt@saortho.org
Southwestern Society of Orthodontists
President, R. Bryn Cooper, Houston, TX
President-Elect, Jesse Teng, El Paso, TX
Vice President, Joe Moon, Overland Park, KS
Secretary-Treasurer, Onur Kadioglu,
Shawnee, OK
Trustee, Steven Robirds, Austin, TX
Interim Executive Director, Cindy Metcalf, 6708
Menchaca Road, Unit 30, Austin, TX 78745;
telephone, 918-960-2666; fax, 404-521-4180;
email, swsoexecdir@gmail.com
American Board of Orthodontics
President, David G. Sabott, Lafayette, CO
President-Elect, Patrick F. Foley,
Lake Zurich, IL
Secretary-Treasurer, Timothy Trulove,
Montgomery, AL
Director, Jae Hyun Park, Mesa, AZ
Director, Roberto Hernandez-Orsini,
San Juan, PR
Director, Stephen McCullough, Yukon, OH
Director, P. Emile Rossouw, Rochester, NY
Director, Anthony M Puntillo, Crown Point, IN
Executive Director, Carole Newport,
401 N Lindbergh Blvd, Suite 300,
St Louis, MO 63141; telephone,
314-432-6130; e-mail,
info@americanboardortho.com
American Association of Orthodontists
Foundation Board of Directors
President, Orhan C. Tuncay (MASO)
Secretary-Treasurer ex officio, Lynne Thomas
Gordon
National Planned Giving Chair, Robert James
Bray (MASO)
National Endowment Campaign Chair, Eric R.
Nease, SAO
PARC Board Liaison, Wanda Claro (SAO)
Executive Director, Jackie Bode
Email: jbode@aaorthod.org
College of Diplomates of the American Board
of Orthodontics
President, Daniel J. Rejman, Castle Rock, CO
President-Elect, Linda Rigali, Northampton,
MA
Secretary, Ashok Kothari, Countryside, IL
Treasurer, Sandy Bigman, San Ramon, CA
Councilors, Paul Sproul, Madison, AL;
Matthew Ng, Katy, TX
Editor, Howard Fine, Golden Bridge, NY
Historian, John William M. Carter, Leawood, KS
Parliamentarian, Perry Opin, Milford, CT
Executive Director, Kristi Burmeister, 2131
Meadow Valley Dr, Innsbrook, MO 63390
Email: association.info@icloud.com
Phone 888-217-2988
Fax 636-745-3012
College of Diplomates of the American Board
of Orthodontics Foundation
President, Robert Vaught, Savannah, GA
President-Elect, Daniel Rejman, Castle
Rock, CO
Secretary-Treasurer, Linda Rigali,
Northampton, MA
Trustees: Eric Dellinger, Angola, IN;
Rodney Hyduk, Troy, MI; Michael Guess,
El Dorado Hills, CA; Robert Moss, Albany,
GA; Kenneth Hrechka, Oxon Hill, MD;
Paul E. Miller, Quincy, IL; and Terry Sobler,
New City, NY
Executive Director, Dave Burmeister, 2131
Meadow Valley Dr. Innsbrook, MO 63390
Email: association.info@icloud.com
Phone 888-217-2988
Fax 636-745-3012
 Founded in 1915 Volume 160 Number 2 August 2021
Copyright Ó 2021 by the American Association of Orthodontists

CONTENTS
COVER
On the cover: Traditionally, retainer fabrication workflow has involved taking an alginate
impression, pouring the impression up in stone, and fabricating the retainer on the stone
model. Problems during fabrication can occur, requiring the patient to return to the office
for a new impression. With the intraoral scans and CAD-CAM software, orthodontic
appliance fabrication techniques have evolved and have become digital. The brackets
can be removed digitally using software. Then a new model can be printed in 3D and the
retainer fabricated before the debonding appointment. Drs. Marsh and Weissheimer and
colleagues at the University of Southern California performed a clinical study to validate
an innovative technique for virtual bracket removal. They also compared the accuracy
with that of 2 digital orthodontic laboratories that also use virtual bracket removal for
retainer fabrication. The images on the cover show the virtual bracket removal process
with Meshmixer software for retainer fabrication before the debonding appointment.
The smiling patient on the cover is Dawson High. His orthodontic treatment was provided
by James Klarsch, St. Louis, Mo.

GUEST EDITORIAL
Business science and evidence-based decision making 159
Donald J. Rinchuse, Irwin, Pa

READERS' FORUM
Hypoxia-inducible factor-1a may be the first host response in orthodontic tooth 163
movement
Stephanos Kyrkanides, Rochester, NY

Author’s response 164

Ching-Chang Ko, Columbus, Ohio

Correcting skeletal open bite with clear aligners and miniscrews 165
Robert Waxler, St. Louis, Mo

Author’s response 165

Teresa Pinho, Paredes, Portugal

Skeletal open bite treated with clear aligners and miniscrews 166
Yash Agarwal, Manish Goyal, Mukesh Kumar, and Amandeep Kaur, Jharkhand and Uttar Pradesh,
India

Author’s response 167

Teresa Pinho, Paredes, Portugal

Video available with article online at www.ajodo.org

The American Journal of Orthodontics and Dentofacial Orthopedics (ISSN 0889-5406) is published monthly by Elsevier Inc., 230 Park
Avenue, Suite 800, New York, NY 10169. Periodicals postage paid at New York, NY and additional mailing offices. POSTMASTER: Send address
changes to Elsevier, Journal Returns, 1799 Highway 50 East, Linn, MO 65051.

American Journal of Orthodontics and Dentofacial Orthopedics/August 2021

 CONTENTS continued

ETHICS IN ORTHODONTICS
The foundation 169
Peter M. Greco, Philadelphia, Pa

SYSTEMATIC REVIEW
Machine learning and orthodontics, current trends and the future opportunities: 170
A scoping review
Hossein Mohammad-Rahimi, Mohadeseh Nadimi, Mohammad Hossein Rohban, Erfan Shamsoddin,
Victor Y. Lee, and Saeed Reza Motamedian, Tehran, Iran, and New York, NY

ORIGINAL ARTICLES
Biomonitoring of children and adolescents using orthodontic appliances made of 193
acrylic resins through micronucleus testing of exfoliated buccal and palatal mucosa
cells
Joao Pedro Pedrosa Cruz, Nilton Cesar Nogueira dos Santos, Matheus Melo Pithon, and
Eneida de Morais Marc
ılio Cerqueira, Jequi
e, Bahia, and Rio de Janeiro, Brazil

Bias in a blink: Shedding light on implicit attitudes toward patients with a cleft lip 200
Rany M. Bous, Anthony Lyamichev, Ashleigh Kmentt, and Manish Valiathan, Cleveland, Ohio

Evaluation of antimicrobial potential and surface morphology in thin films of titanium 209
nitride and calcium phosphate on orthodontic brackets
Licia Pacheco Teixeira, Leonardo Cabral Gontijo, Adonias Ribeiro Franco J

unior,
Monalessa F
abia Pereira, Ricardo Pinto Schuenck, and Juliana Malacarne-Zanon, Vit
oria, Esp
ırito
Santo, Brazil

A clinically friendly viscoelastic finite element analysis model of the mandible with 215
Herbst appliance
Zahra Heidari Zadi, Amir J. Bidhendi, Ali Shariati, and Eung-Kwon Pae, Dumfries and Montclair,
Va, Montreal, Qu
ebec, Canada, and Baltimore, Md

Long-term follow-up of late maxillary orthopedic advancement with the Liou-Alternate 221
C rapid maxillary expansion-constriction technique in patients with skeletal Class III
E malocclusion
Maria Costanza Meazzini, Camilla Torre, Alessandro Cappello, Roberto Tintinelli, Elena De Ponti,
and Fabio Mazzoleni, Monza and Milan, Italy

Diagnostic accuracy of lateral cephalograms and cone-beam computed tomography 231

for the assessment of sella turcica bridging
Ashley Marie Acevedo, Manuel Lagravere-Vich, and Thikriat Al-Jewair, Bridgeport, CT, Edmonton,
Alberta, Canada, and Buffalo, NY

Simulation of orthodontic tooth movement during activation of an innovative design of 240

C closing loop using the finite element method
E Tuan Nguyen Anh, Ryo Hamanaka, Sachio Jinnai, Hiroya Komaki, Satoshi Yamaoka,
Jun-ya Tominaga, Yoshiyuki Koga, and Noriaki Yoshida, Nagasaki, Japan

American Journal of Orthodontics and Dentofacial Orthopedics/August 2021

 CONTENTS continued

Efficacy of Invisalign attachments: A retrospective study 250

Theresa Karras, Maharaj Singh, Emelia Karkazis, Dawei Liu, Ghada Nimeri, and Bhoomika Ahuja,
Milwaukee, Wis, and Chicago, Ill

Biomechanical analysis for total distalization of the maxillary dentition: A finite 259
element study
Jun Kawamura, Jae Hyun Park, Yukio Kojima, Naohiko Tamaya, Yoon-Ah Kook, Hee-Moon Kyung,
and Jong-Moon Chae, Gifu, Nagoya, and Fukui, Japan, and Mesa, Ariz, and Seoul, Daegu, and
Iksan, South Korea

Influence of the hyrax expander screw position on displacement and stress 266
distribution in teeth: A study with finite elements
Let
ıcia Chaves Fernandes, Robert Willer Farinazzo Vitral, Pedro Yoshito Noritomi,
Gustavo Silva Maximiano, and Marcio Jos
e da Silva Campos, Juiz de Fora, Minas Gerais, and
Campinas, S~ao Paulo, Brazil

Attractiveness assessment by orthodontists and laypeople judging female profile 276

C modifications of Class II Division 1 malocclusion
E Kadriye Kalin, Sebahat Yesim Iskender, and Reinder Kuitert, Utrecht and Amsterdam, The
Netherlands

The influence of incisor inclination and anterior vertical facial height on facial 283
attractiveness in an Asian woman
Umair Shoukat Ali, Rashna Hoshang Sukhia, Mubassar Fida, Adeel Tahir Kamal, and
Ausjah Abbas, Karachi, Pakistan

Biomechanical effect of selective osteotomy and corticotomy on orthodontic molar 292

uprighting
Xin She, Toru Deguchi, Hai Yao, and Jing Zhou, Charleston and Clemson, SC, and Columbus, Ohio

DIGITAL ORTHODONTICS
Three-dimensional assessment of virtual bracket removal for orthodontic retainers: A 302
C prospective clinical study
E Kaitlin Marsh, Andre Weissheimer, Kaifeng Yin, Alexandra Chamberlain-Umanoff,
Hongsheng Tong, and Glenn T. Sameshima, Los Angeles and Redlands, Calif, and Houston, Tex

CASE REPORT
Treatment of facial asymmetry and severe midline deviation with orthodontic 312
mini-implants
Min-Ho Jung, Seoul, South Korea

LITIGATION AND LEGISLATION

Gumballs 325
Laurance Jerrold, Woodbury, NY

American Journal of Orthodontics and Dentofacial Orthopedics/August 2021

 CONTENTS continued

INTERNATIONAL ITEMS OF INTEREST

A recently released report by a major scientific academy proposes significant changes 327
in understanding and managing temporomandibular disorders
Charles S. Greene, John W. Kusiak, Terrie Cowley, and Allen W. Cowley, Jr., Chicago, Ill, and
Milwaukee, Wis

CONTINUING EDUCATION
Instructions 329

DIRECTORY: AAO OFFICERS AND ORGANIZATIONS 330

READER'S SERVICE
Information for authors—on the Internet, go to www.ajodo.org, For Authors, Author
Information

American Journal of Orthodontics and Dentofacial Orthopedics/August 2021

 Official publication of the American Association of Orthodontists,
its constituent societies, the American Board of Orthodontics, and
the College of Diplomates of the American Board of Orthodontics

Editors
Rolf G. Behrents, DDS, MS, PhD, Editor-in-Chief
St Louis, MO; behrents@gmail.com
Jae Hyun Park, DMD, MSD, MS, PhD, Deputy Editor
Mesa, AZ; jpark@atsu.edu
David L. Turpin, DDS, MSD, Editor Emeritus
Seattle, WA; DLTurpin@aol.com
Wayne G. Watson, DDS, Editor Emeritus
La Jolla, CA; wwatson2@san.rr.com
Chris Burke, Managing Editor
Seattle, WA; ckburke@aol.com
Associate Editors
David Covell, Jr, DDS, MSD, PhD, Biology
Buffalo, NY; dacovell@buffalo.edu
Theodore Eliades, DDS, MS, Dr Med Sci, PhD, FIMMM, FRSC, FInstP, Dental Materials
Zurich, Switzerland; Theodore.Eliades@zzm.uzh.ch
Padhraig S. Fleming, MSc, PhD, MOrth, RCS, FDS (Orth), FHEA, Evidence-based Dentistry
London, United Kingdom; padhraig.fleming@gmail.com
Dan Grauer, DDS, MSD, PhD, Residents' Journal Review
Santa Monica, CA; drgrauer@yahoo.com
Peter M. Greco, DMD, Ethics in Orthodontics
Philadelphia, PA; pgrecodmd@gmail.com
Demetrios J. Halazonetis, DDS, MS, Imaging
Kifissia, Greece; dhal@dhal.com
Mark G. Hans, DDS, MSD, Sleep and Breathing
Berea, OH; mark.hans@case.edu
Laurance Jerrold, DDS, JD, Litigation and Legislation
Brooklyn, NY; drlarryjerrold@gmail.com
Sanjivan Kandasamy, BDSc, BScDent, DocClinDent, MOrthRCS, FRACDS, FDSRCS, TMD and Function
Midland, WA, Australia; sanj@kandasamy.com.au
Christos Katsaros, DDS, Dr med dent, Odont Dr/PhD, Craniofacial Anomalies/Cleft Lip and Palate
Bern, Switzerland; christos.katsaros@zmk.unibe.ch
Allen H. Moffitt, DMD, Continuing Education
Murray, KY; AHMGM@aol.com
J. Martin Palomo, DDS, MSD, Digital Orthodontics
Cleveland, OH; palomo@case.edu
Nikolaos Pandis, DDS, MS, Dr med dent, MSc, PhD, Statistics and Research Design
Bern, Switzerland, and Corfu, Greece; npandis@yahoo.com
Mario Polo, DMD, MS, Orthodontic Treatment
San Juan, Puerto Rico; drmariopolo@mariopolo.com
Zongyang Sun, DDS, MS, MSD, PhD, Biology
Columbus, OH; sun.254@osu.edu
Leslie A. Will, DMD, MSD, Growth and Development
Boston, MA; willla@bu.edu
Jae Hyun Park, DMD, MSD, MS, PhD, American Board of Orthodontics
Mesa, AZ; jpark@atsu.edu
Elsevier
Jane Ryley, Publisher
Elsevier Inc. - 314-447-9226; j.ryley@elsevier.com
Stacey Kauffman, Journal Manager
Elsevier Inc. - 215-239-3819; s.kauffman@elsevier.com

American Journal of Orthodontics and Dentofacial Orthopedics/August 2021

 Official publication of the American Association of Orthodontists,
its constituent societies, the American Board of Orthodontics, and
the College of Diplomates of the American Board of Orthodontics
Information for readers
The American Journal of Orthodontics & Dentofacial Orthopedics
(ISSN: 0889-5406) is published monthly by Elsevier Inc., 230 Park
Avenue, Suite 800, New York, NY 10169. Periodicals postage paid
at New York, NY and additional mailing offices.
POSTMASTER: Send address changes to Elsevier, Journal Returns,
1799 Highway 50 East, Linn, MO 65051.
Communications regarding editorial management should be
addressed via email to Rolf G. Behrents, Editor-in-Chief,
behrents@gmail.com. Send U.S. mail to Chris Burke, Managing
Editor, The American Journal of Orthodontics and Dentofacial
Orthopedics, University of Washington, Department of Orthodon-
tics, D-569, HSC Box 357446, Seattle, WA 98195-7446; tel: (206)
221-5413; E-mail: ckburke@aol.com.
Information for authors is available at www.ajodo.org, For
Authors, Author Information.
Customer Service (orders, claims, online, change of address).
Please visit our Support Hub page https://service.elsevier.com
for assistance.
Address changes must be submitted four weeks in advance.
Yearly Subscription Rates. United States and possessions:
Individual $423.00; Student/Resident $226.00. All other
countries (prices include airspeed delivery): Individual $511.00;
Student/Resident $252.00. To receive student/resident rate,
orders must be accompanied by name of affiliated institution,
date of term, and the signature of program/residency coordinator
on institution letterhead. Orders will be billed at the individual
rate until proof of status is received. Current prices are in effect
for back volumes and back issues.
Further information on this journal is available from the Publisher
or from this journal's website (http://www.AJODO.org). Informa-
tion on other Elsevier products is available through Elsevier's
website (http://www.elsevier.com).
Advertising information. Advertising orders and inquiries can
be sent to: USA, Canada, and South America, Allen L. Schwartz,
allen@ssmediasol.com, tel: (503) 784-8919, fax: (503) 961-0445;
Bill Kittredge, bill@ssmediasol.com, tel: (714) 264-7386; Jim
Shavel, jim@ssmediasol.com, tel: (215) 449-7342, fax: (215)
369-4381. Classified advertising orders and inquiries can
be sent to Ariel Medina, a.medina@elsevier.com, tel: (212)
633-3689, fax: (212) 633-3850. Europe and the rest of the world,
Sarah Ellis, s.ellis@elsevier.com, tel: (144) 20 7424 4538.
Author inquiries. For inquiries relating to the submission of
articles (including electronic submission where available), please
visit http://www.elsevier.com/track-submission. Visit http://
www.elsevier.com/trackarticle to track accepted articles and set
up e-mail alerts to inform you of when an article's status has
changed, as well as obtain detailed artwork guidelines, copyright
information, frequently asked questions, and more. Contact
details for questions arising after acceptance of an article,
especially those relating to proofs, are provided after registration
of an article for publication. You are also welcome to contact Cus-
tomer Support via http://support.elsevier.com.
American Journal of Orthodontics and Dentofacial Orthopedics/August 2021
N The paper used in this publication meets the requirements of
s
ANSI/NISO Z39.48-1992 (Permanence of Paper).
Reprints. For queries about author offprints, e-mail authorsupport@
elsevier.com. To order 100 or more reprints for educational,
commercial, or promotional use, contact Derrick Imasa at 212-
633-3874, Elsevier Inc., 230 Park Ave, Suite 800, New York, NY
10169. Fax: 212-462-1935; e-mail reprints@elsevier.com.
Reprints of single articles available online may be obtained by pur-
chasing Pay-Per-View access for $31.50 per article on the journal
website, http://www.AJODO.org. Microform edition available from
ProQuest Information and Learning, 300 N Zeeb Rd, Ann Arbor, MI
48106-1346. Ó 2021 American Association of Orthodontists. All
rights reserved. This journal and the individual contributions con-
tained in it are protected under copyright by the American Associ-
ation of Orthodontists, and the following terms and conditions
apply to their use: Photocopying. Single photocopies of single ar-
ticles may be made for personal use as allowed by national copy-
right laws. Permission of the Publisher and payment of a fee is
required for all other photocopying, including multiple or system-
atic copying, copying for advertising or promotional purposes, re-
sale, and all forms of document delivery. Special rates are available
for educational institutions that wish to make photocopies for non-
profit educational classroom use. Permissions may be sought di-
rectly from Elsevier's Global Rights Department in Oxford, UK:
tel: (11) 215 239 3804 or 144 (0)1865 843830, fax: 144 (0)1865
853333. Requests may also be completed online via the Elsevier
site (http://www.elsevier.com/authors/obtaining-permission-to-re-
use-elsevier-material). In the USA, users may clear permissions
and make payments through the Copyright Clearance Center, Inc.,
222 Rosewood Drive, Danvers, MA 01923, USA; tel: (11) (978)
750-8400, fax: (11) (978) 750-4744, and in the UK through the
Copyright Licensing Agency Rapid Clearance Service (CLARCS),
90 Tottenham Court Road, London W1P 0LP, UK; tel: (144) 20
7631 5555; fax: (144) 20 7631 5500. Other countries may have
a local reprographic rights agency for payments. Derivative Works.
Subscribers may reproduce tables of contents or prepare lists of
articles including abstracts for internal circulation within their insti-
tutions. Permission of the Publisher is required for resale or distribu-
tion outside the institution. Permission of the Publisher is required
for all other derivative works, including compilations and transla-
tions. Electronic Storage or Usage. Permission of the Publisher is re-
quired to store or use electronically any material contained in this
journal, including any article or part of an article. Except as outlined
above, no part of this publication may be reproduced, stored in
a retrieval system or transmitted in any form or by any means, elec-
tronic, mechanical, photocopying, recording or otherwise, without
prior written permission of the Publisher. Address permissions
requests to: Elsevier Rights Department, at the fax noted above.
Notice. No responsibility is assumed by the Publisher or the Ameri-
can Association of Orthodontists for any injury and/or damage to
persons or property as a matter of products liability, negligence or
otherwise, or from any use or operation of any methods, products,
instructions or ideas contained in the material herein. Because of
rapid advances in the medical sciences, in particular, independent
verification of diagnoses and drug dosages should be made.
Although all advertising material is expected to conform to ethical
(medical) standards, inclusion in this publication does not constitute
a guarantee or endorsement of the quality or value of such product or
of the claims made of it by its manufacturer.