Received: 16 June 2019 | Accepted: 19 June 2019

DOI: 10.1111/jocd.13073

ORIGINAL CONTRIBUTION

Subperiosteal injections during facial soft tissue filler

injections—Is it possible?

Tatjana Pavicic MD1 | Mariya Yankova MD1 | Thilo L. Schenck MD, PhD2 |
Konstantin Frank MD2 | David L. Freytag2 | Jonathan Sykes MD3 |
Jeremy B. Green MD4 | Hassan Hamade BSc5 | Gabriela Casabona MD6 |
Sebastian Cotofana MD, PhD, PhD5,7

1
Private Practice, Munich, Germany
2 Abstract
Department for Hand, Plastic and Aesthetic
Surgery, Ludwig—Maximilian University Background: It can be hypothesized that safety during soft tissue filler injection
Munich, Munich, Germany
could be enhanced if the product could be positioned between the periosteum and
3
Facial Plastic and Reconstructive
Surgery, UC Davis Medical Center,
the bone surface i.e. subperiosteal.
Sacramento, CA, USA Aim: This study investigated the feasibilityof subperiosteal injections.
4
Skin Associates of South Florida, Coral Patients/Methods: We analysed 126 injection procedures performed in seven
Gables, FL, USA
5 Caucasian body donors (4 males, 3 females) with a mean age of 75.29 ± 4.95 years
Department of Medical Education, Albany
Medical College, Albany, NY, USA [range: 70 – 87] and a mean body mass index of 23.53 ± 3.96 kg/m2 [range: 16.46 –
6
Ocean Clinic, Marbella, Spain 32.23]. The injection procedures were performed in the forehead, scalp, zygomatic
7
Department of Surgery, Division of Plastic
arch and the mandible bilaterally. Injection procedures were conducted using 25G,
Surgery, Albany Medical Center, Albany,
NY, USA 27G and 30G sharp‐tip needles (TSK Laboratory, Tochigi, Japan) and at various angles
from the bone surface measured with a goniometer: 90 degrees (perpendicular), 45
Correspondence
Sebastian Cotofana, Department of Medical degrees and 10 degrees (as tangential as possible to bone surface).
Education, Albany Medical College, 47
Results: Cadaveric dissections of the injection sites showed that no product was lo‐
New Scotland Avenue MC‐135, Albany, NY
12208, USA. cated deep to the periosteum in any of the investigated regions. This indicates that all
Email: cotofas@amc.edu
performed injection procedures positioned the product superficial to the periosteum
(100%) with a zero‐success rate of subperiosteal injections.
Conclusion: In the setting of this cadaveric investigation, despite varying needle size
and injection angle, subperiosteal injection could not be achieved. This indicates
that the product can spread uncontrolled into more superficial layers yielding an in‐
creased risk for adverse aesthetic and vascular events.

KEYWORDS
facial anatomy, injections, layers of the face, soft tissue fillers, subperiosteal

1 | I NTRO D U C TI O N in 2018, which represents a threefold increase vs to the year 2000.1

The number of soft tissue filler injections performed in the United
States is constantly increasing. According to the annual report
released by the American Society of Plastic Surgery, a total of
2 676 907 soft tissue filler injections (unspecified) were performed
Fillers are traditionally injected into the facial soft tissues using ei‐
ther a sharp‐tip needle or a blunt‐tip cannula.
Cannulas are commonly referred to as atraumatic because the tip
is blunt and rounded and the opening for product delivery is on the
side. Having a device with a blunt tip is thought to reduce the risk of

J Cosmet Dermatol. 2019;00:1–6. wileyonlinelibrary.com/journal/jocd

© 2019 Wiley Periodicals, Inc. | 1
2 | PAVICIC et al

perforating and/or penetrating nerves and blood vessels during the

injection procedure. This is assumed to occur due to the increased
pressure needed for a blunt‐tipped device to penetrate a structure
and a resulting “sliding” phenomenon around the respective neuro‐
vasculature. This purportedly reduces the risk of nerve injuries and
the risk of intra‐arterial injection of the product. 2-6
In contrast, it is thought that needles deliver product more
precisely and accurately despite creating more trauma to tissue.7
However, recent studies have shown that using a needle results in
less precise product placement, if precision is defined as the filler
material remaining in the plane of intended implantation.4,5 For in‐
stance, fluoroscopy X‐ray analyses5 revealed that needle‐injected
filler material can run retrograde along the created canal which dis‐
tributes product in all fascial planes, potentially increasing risk for
intra‐arterial placement.
It can be hypothesized that safety during soft tissue filler injec‐
tion could be enhanced if the product could be positioned between
the periosteum and the bone surface, ie, subperiosteal. Positioning
the product in this deep plane might create a safe space for filler
injection where the product is prevented from spreading into more
superficial layers by the overlying dense fibrous periosteum. To test
this hypothesis, the investigators conducted a cadaveric simulation
and dissection study injecting cephalic specimens with different
sized needles (25G, 27G, 30G) at various injection angles (90 de‐
grees, 45 degrees, 10 degrees) to ascertain whether subperiosteal
injections are possible. F I G U R E 1 Facial three‐dimensional reconstruction of a CT scan
of a human body donor showing the circled investigated (forehead,
scalp, zygomatic region and mandible) and the insertion angle of the
2 | M ATE R I A L A N D M E TH O DS needle at the scalp (90 degrees, 45 degrees, 10 degrees)

2.1 | Study sample

We analyzed 126 injection procedures performed in seven

Caucasian body donors (4 males, 3 females) with a mean age of
75.29 ± 4.95 years [range: 70‐87] and a mean body mass index of
23.53 ± 3.96 kg/m2 [range: 16.46‐32.23]. The cephalic specimens
were screened and not included into this analysis if previous surgery,
trauma, or diseases disrupted the integrity of the facial anatomy.
Each body donor had given informed consent while alive for the use
of his or her body for medical, scientific, and educational purposes.
All aspects of the study conform to the laws of the country where
the study was conducted.

2.2 | Injection procedure

The injection procedures were performed in the forehead, scalp,

zygomatic arch, and the mandible bilaterally. Injection procedures
were conducted using a 25G, 27G, and 30G sharp‐tip needle (TSK
Laboratory) at different angles when measured with a goniometer
from the bone surface: 90 degrees (perpendicular), 45 degrees, and
10 degrees (as tangential as possible to bone surface). The 10‐de‐
gree injection was performed in a bevel‐down position to ensure
closest product placement to the bone surface (Figures 1-3).
F I G U R E 2 Photograph showing the scalp of a cadaveric
specimen after layer‐by‐layer dissection including dissection of the
periosteum. Injections with the 25G needle have been performed
in the left area, while a 27G needle was used in the middle and a
30G needle on the right. Increasing angles have been used during
injection (cranial: 10 degrees, middle: 45 degrees, caudal: 90
degrees). Note how the injected product was deposited above the
periosteum within the loose areolar tissue
PAVICIC et al | 3

10 degree
TA B L E 1 Table showing the number of injection procedures performed using different sized needle (25G, 27G, 30G) at varying injection angles (90 degrees, 45 degrees and 10 degrees)

3
3
3
45 degree

3
3
3
90 degree
Mandible

3
3
3
10 degree

2
2
2
45 degree

2
2
2
Zygomatic arch
F I G U R E 3 Photograph showing the injection procedure

90 degree
performed at a cadaveric specimen in the zygomatic arch region.
The 27G needle is inserted at 10 degrees

2
2
2
The product injected was commercially available hair gel of high
viscosity (Henkel AG & Co. KGaA) mixed with regular food coloring

10 degree
(Dr Oetker GmbH). The mixed product was tested manually for consis‐
tency to ensure similar viscoelastic properties to available soft tissue

2
2
2
fillers used during daily clinical use. A series of 9 injection procedures
(0.2 cc of material per needle size and per injection angle) was con‐ 45 degree
ducted in the same location to ensure consistency between results. 2
2

2.3 | Anatomical dissection 2

90 degree

Layer‐by‐layer dissections were performed to verify the location

Scalp
per investigated region: forehead, scalp, zygomatic arch, mandible

of the injected product. Specifically, the periosteum was detached

2
2
2

from the bone surface to identify whether colored material was lo‐
cated in the subperiosteal plane.
10 degree

2.4 | Statistical analyses

7
7
7

The position of the product was recorded and related to the dif‐
45 degree

ferent sized needles (25G vs 27G vs 30G) and injection angles (90
degrees, 45 degrees, 10 degrees). Chi‐squared test was run to calcu‐
7
7
7

late influences of the injector size and the injection angle using SPSS
Statistics 23 (IBM). Results were considered statistically significant
90 degree
Forehead

at a probability level of ≤0.05 to guide conclusions.

7
7
7

3 | R E S U LT S
Needle Size

A total of 126 injection procedures were performed. Sixty‐three in‐

30G
25G
27G

jection procedures were performed in the forehead, 18 in the scalp,

4 |
F I G U R E 4 Layered arrangement consisting of dermis, superficial fatty layer, SMAS, deep fatty layer and periosteum. The periosteum
has been reflected in this photograph. Note how the injected product is dispersed within the loose areolar tissue and within the connective
tissue
F I G U R E 5 Photograph showing the forehead and scalp of a
cadaveric specimen captured from posterior to anterior. Vascular
structures run above the periosteum within the respective layers.
The periosteal layer can be considered avascular
18 along the zygomatic arch, and 27 along the mandible. Of those,
42 injection procedures were conducted with a 25G needle, 42 with
a 27G needle, and 42 with a 30G needle. Forty‐two injection proce‐
dures were conducted at an injection angle of 90 degrees, 42 at 45
degrees, and 42 at 10 degrees (Table 1, Figure 1).
PAVICIC et al
Cadaveric dissections of the injection sites showed that no prod‐
uct was located deep to the periosteum in any of the investigated
regions. This indicates that all performed injections positioned the
product superficial to the periosteum (100%) with a zero success
rate of subperiosteal placement. There was no difference in the abil‐
ity to position the product deep to the periosteum between the dif‐
ferent sized needles (25G, 27G, 30G) or injection angles (90 degrees,
45 degrees, 10 degrees, despite the bevel being down for the latter).
(Figures 2,4,5).
In the mandibular region, with all performed injection proce‐
dures the product was identified inside the masseter, ie, intramus‐
cular. After detachment of the muscle from bone, no product was
identified in direct contact with the bone.
4 | D I S CU S S I O N
This cadaveric study was designed to investigate whether subperi‐
osteal soft tissue filler injection procedures are possible. A total of
126 injection procedures were conducted using different sized nee‐
dles (25G, 27G, 30G) and at varying injection angles (90 degrees, 45
degrees, 10 degrees). Dyed injection material was utilized to evalu‐
ate outcomes. Upon dissection irrespective of needle size or angle
of approach, no injected material was located in the subperiosteal
plane.
One strength of this study is the number of injection procedures
performed (n = 126). This large number enables one to provide valid
data on the claimed hypothesis, especially as none of the proce‐
dures positioned the product subperiosteally. Another strength is
that the size of the utilized needle varied between 25G and 30G.
Each of the three sizes of needles is commonly used daily in an in‐
jectable filler practice. Moreover, it was assumed that with a smaller
needle diameter subperiosteal injections might be more feasible.
However, the 30G needles (n = 42) yielded an identical success rate
PAVICIC et al
of subperiosteal filler placement to the larger needles, zero. Another
strength of this study is that the injections were performed at three
distinct angles of approach: 90 degrees, 45 degrees, and 10 degrees
(when measured form bone surface). The authors theorized that the
angle most approaching parallel to the bone surface (10 degrees)
when performed with a downward facing bevel would have the
greatest chance of success. However, in none of the performed in‐
jection procedures was product found deep to the periosteum.
A limitation of this investigation is that a cadaveric model was
used to test the hypothesis. The tissue properties are different
in non‐living tissue due to the absence of blood pressure, tissue
turgor pressure, muscular contraction, and the different tempera‐
tures. These factors could influence product distribution inside
the tissue which would ultimately affect result interpretation.
However, no confirmatory dissections can be performed in living
individuals nor could the amount of injections (9 injections per lo‐
cation; each needle size and each injection angle) be performed
in the same individual. Another limitation of the cadaveric model
is that the periosteum in a non‐living individual could adhere dif‐
ferently to the underlying bone as compared to living patients.
Alterations in this connection would influence the results; a less
firm connection in living patients would increase the likelihood
of subperiosteal injection success, whereas a firmer connection
in living patients would yield results consistent with those pre‐
sented. To date, there are no data available that allow reliable con‐
clusions on this topic.
This investigation sought to determine whether subperiosteal
injections are possible to potentially increase safety during facial
soft tissue filler injections. Safety is of paramount importance, and
the number of soft tissue filler‐associated cases of blindness has
increased. 8 Between the years 2015 and 2018, 48 cases were pub‐
lished 8 vs 98 cases for the period between 1906 and 2015.9 The
increasing incidence of this catastrophic event can be correlated
1
to the increased number of soft tissue filler injections performed.
The underlying pathophysiologic mechanism of this dreaded com‐
plication was recently simulated utilizing a perfused cadaveric
model.10 The authors were able to identify an association between
the intra‐arterial introduction of the filler material to the supra‐
trochlear artery and the embolization of retinal arteries, which con‐
firmed current theories regarding the occurrence of blindness after
facial soft tissue filler injections. 8-10 Additionally, the authors re‐
ported that a mean pressure of 166.7 mm Hg [range 160‐180] was
10,11
needed to embolize the ophthalmic artery As hyaluronic acid‐
based filler material is a non‐Newtonian fluid,11 a higher pressure
at the beginning of each injection procedure is needed to initiate
product flow. This increased pressure at the onset of the injection
procedure could propel the filler embolus into the ophthalmic ar‐
tery circulation. Another study reported that the average volume
of soft tissue filler necessary to fill the supratrochlear artery from
the glabella to the bifurcation of the ophthalmic and central reti‐
nal arteries was 0.085 cc [range 0.04‐0.12].12This small amount of
product combined with the high (above systolic cranial pressure)
initial injection pressure could potentially result in intra‐arterial
| 5
application of filler material leading to tissue loss, and in the worst
cases blindness.10
Using sharp‐tip needles for the delivery of filler material has
been shown to be less precise, if precision is defined as the filler
material remaining in the plane of intended implantation.4,5 These
two independent studies used 27G and 30G needle tips in contact
with bone and found that the injected product runs retrograde along
the created injection canal. This classic behavior of needle injection
distributes the material in every fascial layer the tip has penetrated
prior to contact with bone. This uncontrolled product spread can
cause non‐aesthetically appealing results like surface irregularities
or discoloration.9,13 Moreover, assuming a through‐and‐through
phenomenon, ie, the tip of the needle has penetrated both sides of
an artery and is located extra‐arterial when touching the bone, the
risk for retrograde product flow and accidental intra‐arterial product
deposition is real.
5 | CO N C LU S I O N
The results of the present study revealed that irrespective of needle
size (25G, 27G, 30G) or injection angle (90 degrees, 45 degrees, 10
degrees) the product cannot be positioned deep to the periosteum.
This indicates that product can spread uncontrolled into more su‐
perficial layers potentially resulting in an increased risk for adverse
aesthetic and vascular complications. A careful selection of the uti‐
lized injector (needle size, needle or cannula) and injection technique
(here: injection angle) should be performed before applying facial
soft tissue fillers to avoid adverse events.
D I S C LO S U R E
None of the other authors listed have any commercial associations
or financial disclosures that might pose or create a conflict of inter‐
est with the methods applied or the results presented in this article.
ORCID
Konstantin Frank https://orcid.org/0000-0001-6994-8877
REFERENCES
1. American Society of Plastic Surgeons. 2018 National plastic sur‐
gery statistics. 2018. https​://www.plast​icsur​gery.org/docum​ents/
News/Stati​s tics/​2018/plast​ic-surge​r y-stati​s tics-report-2018.pdf.
Accessed April 14, 2019.
2. Frank K, Koban K, Targosinski S, et al. The anatomy behind ad‐
verse events in hand volumizing procedures. Plast Reconstr Surg.
2018;141(5):650e‐662e.
3. Tansatit T, Apinuntrum P, Phetudom T. A dark side of the cannula in‐
jections: how arterial wall perforations and emboli occur. Aesthetic
Plast Surg. 2017;41(1):221‐227.
4. van Loghem J, Humzah D, Kerscher M. Cannula versus sharp needle
for placement of soft tissue fillers: an observational cadaver study.
Aesthetic Surg J. 2017;38(1):73‐88.
6 | PAVICIC et al

5. Pavicic T, Frank K, Erlbacher K, et al. Precision in dermal filling: a 12. Khan TT, Colon‐Acevedo B, Mettu P, DeLorenzi C, Woodward JA.
comparison between needle and cannula when using soft tissue An anatomical analysis of the supratrochlear artery: considerations
fillers. J Drugs Dermatol. 2017;16(9):866‐872. in facial filler injections and preventing vision loss. Aesthetic Surg J.
6. Pavicic T, Webb KL, Frank K, Gotkin RH, Tamura B, Cotofana S. 2017;37(2):203‐208.
Arterial wall penetration forces in needles versus cannulas. Plast 13. Urdiales‐Gálvez F, Delgado NE, Figueiredo V, et al. Treatment of
Reconstr Surg. 2019;143(3):504e‐512e. soft tissue filler complications: expert consensus recommenda‐
7. Huang H, Lin Y‐S. A novel needle structure that can avoid intravascular tions. Aesthetic Plast Surg. 2018;42(2):498‐510.
injection of any filler. Plast Reconstr surgery Glob open. 2017;5(4):e1285.
8. Beleznay K, Carruthers J, Humphrey S, Carruthers A, Jones D.
Update on avoiding and treating blindness from fillers: a recent re‐
How to cite this article: Pavicic T, Yankova M, Schenck TL, et
view of the world literature. Aesthetic Surg J. 2019;39(6):662‐674.
al. Subperiosteal injections during facial soft tissue filler
9. Beleznay K, Carruthers J, Humphrey S, Jones D. Avoiding and treat‐
ing blindness from fillers. Dermatologic Surg. 2015;41(10):1097‐1117. injections—Is it possible? J Cosmet Dermatol. 2019;00:1–6.
10. Cho K‐H, Dalla Pozza E, Toth G, Bassiri Gharb B, Zins JE. https​://doi.org/10.1111/jocd.13073​
Pathophysiology study of filler‐induced blindness. Aesthetic Surg J.
2019;39(1):96‐106.
11. Ilyin SO, Kulichikhin VG, Malkin AY. The rheological characterisa‐
tion of typical injection implants based on hyaluronic acid for con‐
tour correction. Rheol Acta. 2016;55(3):223‐233.