journal of dentistry 39 (2011) 154–162

available at www.sciencedirect.com

journal homepage: www.intl.elsevierhealth.com/journals/jden

Caries-removal effectiveness and minimal-invasiveness

potential of caries-excavation techniques: A micro-CT
investigation

Aline de A. Neves, Eduardo Coutinho, Jan De Munck, Bart Van Meerbeek *

Leuven BIOMAT Research Cluster, Department of Conservative Dentistry, School of Dentistry, Oral Pathology and Maxillo-Facial Surgery,
Catholic University of Leuven, Kapucijnenvoer 7, B-3000 Leuven, Belgium

article info abstract

Article history: Objectives: To determine the caries-removal effectiveness (CRE) and minimal-invasiveness
Received 12 July 2010 potential (MIP) of contemporary caries-removal techniques.
Received in revised form Methods: Carious molars were scanned using micro-CT, after which dentine caries was
16 November 2010 removed by 9 contemporary caries-removal techniques. The micro-CT was repeated and
Accepted 17 November 2010 CRE was determined on basis of the relative volume of residual caries and the mineral
density (MD) at the cavity floor. MIP was determined by measuring the cavity size relative to
the initial size of the caries lesion.
Keywords: Results: CRE and MIP varied most for the Er:YAG laser (Kavo) despite its laser-induced
Caries-excavation fluorescence (LIF) feedback system. Whilst some specimens revealed much residual caries,
Caries-removal techniques others showed over-excavation into sound dentine. With the highest Relative Cavity Size, the
Minimal-invasive dentistry Er:YAG laser presented the lowest MIP. Rotary/oscillating instruments revealed a more favour-
Micro-CT able CRE with some tendency towards over-excavation, except for CeraBur (Komet-Brasseler)
Er:YAG laser and Cariex (Kavo) that typically left caries at the cavity floor and cavity walls, respectively.
Carisolv Chemo-mechanical excavation aided by conventional metal excavators (Carisolv, MediTeam;
Caries Detector exp. SFC-V and SFC-VIII, 3M-ESPE) combined best CRE with MIP. When however a plastic
CeraBur excavator was used along with exp. SFC-VIII, caries was less completely removed.
Cariex Significance: Er:YAG-laser aided by LIF resulted in non-selective caries removal. Rotary/
Tungsten-carbide bur oscillating caries removal may lead to over-excavation, especially when burs are combined
with Caries Detector (Kuraray). This risk for over-excavation is reduced when a tungsten-
carbide bur is solely used. On the contrary, Cariex (Kavo) and CeraBur showed a tendency for
under-preparation. Chemo-mechanical methods were most selective in removing caries,
whilst preserving sound tissue.
# 2010 Elsevier Ltd. All rights reserved.

1. Introduction tissue.1 As soft and wet carious dentine lesions harbour

The widespread concept of ‘‘minimal-invasive dentistry’’
implies that heavily infected and irreversibly denatured
dentine should be removed selectively in order to preserve
as much as possible sound or potentially remineralisable tooth
significantly more bacteria than hard and/or dry lesions,2
clinicians are commonly advised to remove carious dentine to
the level where it is ‘firm’.3 Doing so, they may possibly leave
at the cavity floor demineralised dentine that is judged to still
possess some remineralisation/healing potential.4

* Corresponding author. Tel.: +32 16 337587; fax: +32 16 332752.

E-mail address: bart.vanmeerbeek@med.kuleuven.be (B. Van Meerbeek).
0300-5712/$ – see front matter # 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jdent.2010.11.006
 journal of dentistry 39 (2011) 154–162
Nevertheless, the exact endpoint of caries removal can
clinically not easily be defined. This may be explained by the
fact that dentists have mainly been trained in relying on the
hardness of dentine as felt with a dental probe, whilst other
more subjective characteristics, such as the colour and
moisture of the excavated dentine, are often neglected.
A first attempt to more objectively define the caries-removal
endpoint involved the use of a staining agent consisting of an 1%
acid-red solution in a propylene glycol base (Caries Detector,
Kuraray, Osaka, Japan).5 It was first recommended to remove all
the red-stained dentine because the staining solution was
considered to stain the irreversibly denaturated and thus not
remineralisable collagen.6 After more recent research showed
that removal of all red-stained dentine led invariably to over-
excavation,7 one nowadays instruct to retain the ‘pink’ or ‘light-
pink’ stained tissue, because this usually contains only slightly
demineralised dentine that potentially can remineralise.8
Unfortunately, objective interpretation of what is tissue stained
‘red’, ‘pink’ or even ‘light pink’ is still a major issue of dispute.
Moreover, the tendency of Caries Detector to over-stain sound
tissue cannot be ignored.9
In light of minimal-invasive tooth preparation and with the
intention to simplify and standardise caries-removal proce-
dures, so-called ‘self-limiting’ caries-excavation techniques
have more recently been introduced. For instance, new rotary
cutting instruments include round burs made of an alumina-
based ceramic material intended for slow-speed caries
excavation (CeraBur K1SM, Komet-Brasseler, Lemgo,
Germany),10 or oscillating sono-abrasion tungsten-carbide
tips (Cariex system, Kavo, Biberach, Germany).11 Other self-
limiting alternatives include dentine solubilising agents used
to selectively dissolve carious dentine, such as the sodium-
hypochlorite-based Carisolv (MediTeam, Göteborg, Sweden)12
or the new experimental enzyme-based caries-removal gels
(exp. SFC-V and VIII, 3M-ESPE, Seefeld, Germany).13 The latter
consist of pepsin in a phosphoric acid/sodium biphosphate
buffer that is claimed to more selectively remove carious
tissue.14 According to the manufacturer, phosphoric acid
dissolves the inorganic component of carious dentine, allow-
ing pepsin to access the organic part of the caries biomass,
thereby selectively dissolving the denatured dentine collagen
that lost the triple-helix structure. The chemically softened
tissue can then be removed mechanically using either a metal
or plastic hand instrument.
An Er:YAG laser equipped with a laser-induced fluores-
cence (LIF) feedback system (Key III, Kavo, Biberach, Germany)
also claims to possess self-limiting caries-removal potential.
The fluorescence emitted by the bacterial metabolites present
in the carious tissue is continuously measured during the
caries-removal process, and if a pre-selected threshold is
exceeded, the laser device is activated and the carious tissue
ablated.15 The clinical use of a LIF threshold level of 7 has been
shown to result in an endpoint of cavity preparation where the
bacterial viability within the remaining tissue was very low,15
or where histologically bacteria at the cavity floor could no
longer be detected.16 Moreover, this technique was demon-
strated to be more minimally invasive (smaller cavity size)
than a conventional tungsten-carbide bur.17
Micro-CT is a non-destructive research technique that
allows hard tissues to be measured volumetrically and
155
evaluated on their mineral density.18 This technique is
increasingly becoming popular in dental research, as it
enables to collect detailed full-quantitative data of the
substrate before and after a specific substrate treat-
ment.13,19–22 Its application to study caries-excavation tech-
niques has recently been demonstrated.20 Following that
approach, the aim of this study is to determine, using micro-
CT, the caries-removal effectiveness (CRE) and minimal-
invasiveness potential (MIP) of 9 contemporary caries-exca-
vation techniques.
2. Materials and methods
2.1. Selection of teeth and micro-CT scanning procedures
From a bulk of extracted molars, stored in aqueous chloramine
for less than 6 months, those presenting occlusal carious
lesions were selected. After ultrasonic cleaning of plaque,
calculus and other debris, a radiograph was taken (MiniRay,
Soredex, Tuusula, Finland) with the aid of a CCD-detector
(Vista Ray CCD Systems, Dürr Dental, Bietigheim-Bissingen,
Germany), so that teeth without dentine-caries involvement
and those, of which the carious lesion was less than 1 mm
remote from the pulp chamber, were excluded. The teeth were
mounted by the roots in gypsum for ease of manipulation
(n = 63).
A first micro-CT scan of the occlusal part of each tooth was
performed using a Skyscan 1172 desktop micro-CT (Skyscan,
Kontich, Belgium) and the resulting 3D volume was assigned
to the ‘Baseline’ stack (Fig. 1a1). The acquisition settings
employed were 100 mA, 100 kV, 14.6 mm pixel size and a
rotation step of 0.78. A flat-field reference was taken before the
first scan and the random-movement amplitude was set to 30
lines to reduce ring artefacts. To improve signal-to-noise ratio,
32-frame averaging was applied during the acquisition phase.
These settings were standardised following a methodological
study done before,20 during which a polychromatic source
micro-CT technique was validated to study dentine caries-
excavation methods. During the scanning procedure, desic-
cation of the tooth was prevented by wrapping the specimen
in Parafilm (Pechiney Plastic Packaging, Menasha, WI, USA),
together with a cotton pellet soaked in chloramine enclosed.
After caries excavation, each tooth was scanned again
using micro-CT following the same acquisition settings
described above. The resulting 3D volume obtained from the
cross-section images was assigned to the ‘Excavated Caries’
stack (Fig. 1b1). The micro-CT projections obtained for both the
‘Baseline’ and ‘Excavated Caries’ stacks were reconstructed
(NRecon, Skyscan, Kontich, Belgium) based on a modified
Feldkamp algorithm. Specific reconstruction settings included
a 5th order polynomial beam-hardening correction and input
of optimal contrast limits, as described previously.20
2.2. Caries-excavation procedures
The teeth were then randomly assigned to 9 different
contemporary caries-excavation techniques, including 6
commercially available and 3 experimental methodologies.
All caries-removal techniques were employed by one experi-
156 journal of dentistry 39 (2011) 154–162
[()TD$FIG]

Fig. 1 – 3D-volume renderings obtained from micro-CT cross-section slices of a tooth before and after caries excavation,
based on a previously defined cut-off point for dentine caries.23 (a1) 3D volume of a tooth before caries removal (‘Baseline’).
(b1) 3D volume of the same tooth after caries removal (‘Excavated Caries’). (c1) Result of the subtraction of the ‘Excavated
Caries’ (b1) from ‘Baseline’ (a1). (a2) 3D volume of initial caries (IC) after application of the cut-off point (1.11 g/cm3 HAp) at
‘Baseline’. (b2) 3D volume of residual caries (RC) after application of the cut-off point (1.11 g/cm3 HAp) at ‘Excavated Caries’.
(c2) Resulting 3D-volume of the prepared cavity (PC) after subtracting ‘Excavated Caries’ (b1) from ‘Baseline’ (a1).

enced operator (AAN). If a certain degree of subjectivity was
expected during assessment of the caries-removal endpoint,
some guidelines were used. The relative hardness of the cavity
floor, felt upon gentle pressure with a blunt explorer, was
considered as the caries-removal endpoint for tungsten-
carbide bur excavation and the Cariex system, whilst a visible
residual pink staining was considered the endpoint of caries
removal after application of the Caries Detector solution
followed by thorough rinsing with water.
Regarding the cavity preparation, first, all enamel overhangs
from each carious lesion were minimal invasively removed
with a cylinder diamond bur (Komet-Brasseler, Lemgo,
Germany) in a high-speed air turbine under water cooling until
the underlying dentine lesion was exposed. Five teeth were
excluded due to a too small dentinal caries lesion and five other
teeth because of pulp exposure after caries-excavation. The
caries-removal techniques employed are described below:
(a) Tungsten-carbide round bur (Komet-Brasseler, Lemgo,
Germany): Different bur sizes (n.10–23) were used, depend-
ing on the size of the carious lesion, along with a low-speed
contra-angle with an approximate speed of 1500 rpm,
without water-cooling. The caries-removal endpoint was
reached when a ‘hard’ cavity floor was felt upon gentle
pressure with a blunt dental explorer (n = 7).
(b) Tungsten-carbide round bur aided by Caries Detector
(Kuraray, Osaka, Japan): Dentine stained with Caries
Detector was removed with a tungsten-carbide bur, as
described above. The caries-removal endpoint was reached
when the residual dentine stained ‘light-pink’ (n = 6).
(c) CeraBur (K1SM, Komet-Brasseler, Lemgo, Germany): Dif-
ferent bur sizes (n.10–23) were used, depending on the size
of the carious lesion, along with a low-speed contra-angle
with an approximate speed of 1500 rpm, without water-
cooling. The caries-removal endpoint was established by
the self-cutting ability of the instrument (n = 6).
(d) Cariex (Kavo, Biberach, Germany): An airscaler (Sonicflex
2003L, Kavo) to which tungsten-carbide oscillating tips (TC
tips n.71 and n.72) were coupled, was employed with an
oscillation of >6.5 kHz under water cooling. The caries-
removal endpoint was reached when a ‘hard’ cavity floor
was felt upon gentle pressure with a blunt dental explorer
(n = 6).
(e) Carisolv (MediTeam, Göteborg, Sweden): After dispensing
the gel with the auto-mix syringe system, a drop of the
solution was placed in the cavity. After 30 s, the mace-tips
Carisolv instruments (n.2–5) were used to scrape off the
carious tissue. This procedure was repeated until the
caries-removal endpoint based on the self-limiting capac-
ity of the solution was reached (n = 7).
 journal of dentistry 39 (2011) 154–162
(f) Experimental SFC-V (3M-ESPE, Seefeld, Germany) aided by
a conventional metal excavator: After mixing the two
separate gels according to the manufacturer’s instruction,
a drop of the mixture was placed in the cavity. After 30 s, a
metal spoon excavator was used to remove the caries. This
procedure was repeated until the caries-removal endpoint
based on the self-limiting capacity of the solution was
reached (n = 5).
(g) Experimental SFC-VIII (3M-ESPE, Seefeld, Germany) aided
by a conventional metal excavator: Same procedure as
under (f) (n = 5).
(h) Experimental SFC-VIII (3M-ESPE, Seefeld, Germany) aided
by a prototype plastic excavator (Star v.1.3, 3M-ESPE):
Same procedure as under (f) and (g), but instead of a metal
excavator, a prototype, disposable, star-shaped plastic
excavator was used to remove the carious tissue. The
plastic instrument was replaced during the caries-removal
procedure if it became blunt (n = 6).
(i) Er:YAG laser (Kavo): The substrate was irradiated using a
non-contact handpiece (n.2060, Kavo) at a working
distance of approximately 15 mm, whilst the irradiated
area was continuously cooled by a water spray (1 ml/min)
(n = 5). The output settings for dentine ablation were
250 mJ/pulse and 4 pulses/s repetition. The laser was
equipped with a LIF-feedback system, which emitted light
with a wavelength of 655 nm (red light). When the
measured LIF value of dentine was above a pre-selected
threshold, the laser was activated. For the present study,
the LIF threshold level was set to 7, thereby following
previous studies.16,17
2.3. Caries-removal effectiveness (CRE)
The effectiveness of caries removal was evaluated by means of
two parameters measured after caries excavation: (1) the
(mean) relative volume of residual caries and (2) the mean
mineral density (MD) at the bottom of the cavity. First, two 8-
mm diameter HAp phantoms made of fine calcium hydroxy-
apatite powder embedded in epoxy resin, resulting in two
different mineral densities (0.25 g/cm3 and 0.75 g/cm3), were
obtained from the micro-CT manufacturer (Skyscan, Kontich,
Belgium). A third phantom was produced by cutting a sintered
HAp block with a higher mineral density (3.14 g/cm3; Pentax
Lifecare Division, Tokyo, Japan) into a circular 8-mm cross-
section slab. Grey values obtained by micro-CT were converted
into MD of hydroxyapatite (HAp) by scanning and reconstruct-
ing the HAp phantoms using the same micro-CT acquisition
parameters as for the teeth, and eventually by defining a
calibration curve.20 Second, a cut-off point corresponding to a
MD value of 1.11 g/cm3 HAp (as determined previously by
correlating micro-CT grey values with hardness values of
carious dentine23) was used as a cut-off point to segment each
tooth into a ‘sound’ and ‘carious’ volume, this both in the
‘Baseline’ and ‘Excavated Caries’ stacks. Next, the two CRE
parameters were calculated as follows:
(1) The mean relative residual caries volume or RC/IC ratio
was obtained as the ratio of the volume of ‘carious’ tissue
segmented in the ‘Excavated Caries’ stack (residual caries
or RC, Fig. 1b2), over the volume of ‘carious’ tissue
157
[()TD$FIG]
Fig. 2 – Correlation between IC volume (volume of ‘carious’
tissue, segmented in ‘Baseline’) and RC volume (volume of
‘sound’ tissue, segmented in ‘Excavated Caries’). Pearson
moment correlation coefficient (PMCC) was statistically
significant (0.68, p < 0.05), indicating a positive correlation
between the two variables.
segmented in the ‘Baseline’ stack (initial caries or IC,
Fig. 1a1). Consequently, the lower the RC/IC ratio is, the
more effective the carious lesion was excavated. This
normalisation procedure by dividing RC by IC was needed
for two reasons: (a) to exclude the probability of larger
lesions having larger areas of residual caries, and (b) to
correct/compensate the volume of residual ‘carious’ tissue
in the ‘Excavated Caries’ stack for a partial volume effect
(PVE). PVE is an artefact inherent to the micro-CT
technique24; in this case it accounts for low grey-value
pixels (and thus low MD values, similar to that of the
carious lesion) at the edge between the air (background)
and tooth tissue. In this study, the higher the IC was, the
higher the inner cavity area was and the higher the PVE
was. In fact, a statistically significant correlation was found
between RC and IC (Pearson’s coefficient = 0.68, p < 0.05),
as seen in Fig. 2.
(2) The MD at the cavity floor was calculated from a selected
volume of interest (VOI) with a thickness of 70 mm at the
deepest dentine part of the cavity in the ‘Excavated Caries’
volumes, as depicted in Fig. 3. This VOI was obtained by
first defining the deepest point of the prepared cavity.
Further, the VOI was expanded sideways towards the
cavity walls and next to a depth of 70 mm underneath the
cavity floor. Besides registering the absolute mean MD,
each caries-excavated tooth was further classified as
‘sound’ if the mean MD at the cavity floor was above the
dentine-caries cut-off point (1.11 g/cm3 HAp), or as ‘cari-
ous’ if the mean MD was lower than the cut-off point.
2.4. Minimal-invasiveness potential (MIP)
The minimal-invasiveness potential of the different caries-
excavation techniques was evaluated by means of the
158 journal of dentistry 39 (2011) 154–162
[()TD$FIG]
Fig. 3 – (a) Occlusal view of a 3D-volume rendering obtained
from micro-CT cross-section slices of a tooth after caries
removal (‘Excavated Caries’), showing the selected volume
of interest (VOI), where MD was measured (rendered in red
colour). (b) Micro-CT mesio-distal cut view from the same
tooth in (a) showing the depth of the VOI used for MD
measurement. (For interpretation of the references to
colour in this figure legend, the reader is referred to the
web version of this article.)
‘Relative Cavity Size’ or PC/IC ratio.25 From each tooth,
the ‘Excavated Caries’ stack was subtracted from the
‘Baseline’ volume (Fig. 1c1), by which the volume of the
prepared cavity (PC, Fig. 1c2) was obtained. This PC was
divided by IC to provide the ‘Relative Cavity Size’. Conse-
quently, a ‘Relative Cavity Size’ near ‘1’ indicated a perfect
MIP, since the volume of the removed tissue (PC) then
corresponded to the volume of the initial carious lesion (IC
volume).
2.5. Statistical analysis
The two CRE parameters and the MIP parameter measured for
the 9 caries-excavation techniques were compared by one-
way analysis of variance with Fisher’s least significant
difference (LSD). The difference in percentage of ‘sound’
versus ‘carious’ teeth for the 9 caries-excavation techniques
was statistically assessed by the Fisher’s exact test. The
significance level was set to 5% for all analyses.
3. Results
The CRE in terms of both the RC/IC ratio and MD parameter is
shown for the 9 caries-excavation methods investigated in
Fig. 4a and b. The percentage of teeth exhibiting ‘sound’ versus
‘carious’ dentine according to MD is depicted in Fig. 4c. The
MIP of the different caries-excavation techniques investigated
is depicted in Fig. 5.
Er:YAG laser guided by the LIF-feedback system resulted in
the most varying results and this for all parameters evaluated
(high confidence intervals). Whilst some specimens showed a
high RC/IC ratio (Fig. 4a), others did not and in fact exhibited
over-excavation into sound dentine (0.22  0.26). This high
variability is confirmed by the also highly varying MD
parameter (1.05  0.57) for the Er:YAG laser (Fig. 4b). The
mean Relative Cavity Size of the Er:YAG laser was statistically
significantly higher than that of all other caries-excavation
techniques, thereby having resulted in a statistically signifi-
cant lower MIP (19.16  28.58), except for the tungsten-carbide
bur when used along with Caries Detector (Fig. 5, ANOVA with
Fisher’s LSD test, p < 0.05).
CRE improved when rotary/oscillating caries-excavation
techniques were employed, with the tungsten-carbide bur, in
combination with Caries Detector or not, scoring the lowest
RC/IC ratio (0.05  0.05 and 0.06  0.06, respectively; Fig. 4a).
Although not statistically significant, a clear tendency
towards over-excavation was, however, noted especially
when Caries Detector was used to guide caries excavation
(with 1.34  0.14 as the highest MD in Fig. 4b, and 8.5  8.45 as
the lowest but one MIP in Fig. 5). Regarding CRE (in terms of
‘Mean Relative Residual Caries Volume’), Cariex and CeraBur
presented relatively high mean RC/IC ratios (0.16  0.15 and
0.15  0.08, respectively; Fig. 4a). For CeraBur, as depicted in
Fig. 4b, a statistically significantly lower MD at the cavity floor
was measured (0.89  0.19, ANOVA with Fisher’s LSD test,
p < 0.05), as compared to that of the other rotary/oscillating
caries-excavation techniques (1.34  0.14, 1.2  0.22,
1.2  0.22 for tungsten-carbide bur with or without Caries
Detector, and Cariex, respectively). All specimens prepared
with CeraBurs still revealed residual caries at the cavity floor
(Fig. 4c). On the contrary, for Cariex the mean MD at the cavity
floor remained above the threshold (1.2  0.22), indicating
that dentine caries was mostly completely removed (Fig. 4b).
Regarding MIP of rotary/oscillating instruments, both
CeraBur and Cariex showed a tendency towards more
conservative cavity preparation (1.85  1.43 and 2.11  1.46,
respectively; Fig. 5), although as mentioned above this cannot
be interpreted as favourable for CeraBurs that always left
caries at the cavity floor (Fig. 6).
Chemo-mechanical caries-removal techniques combined
best CRE with MIP. In terms of MD at the cavity floor, caries was
removed up to a level most closely approximating the micro-
CT determined dentine-caries cut-off point (Fig. 4b), when the
exp. SFC-VIII and Carisolv were employed (1.13  0.09 and
1.13  0.28, respectively). The exp. SFC-V revealed a slight
tendency for over-excavation (1.26  0.21). The MIP of chemo-
mechanical caries-removal techniques was statistically sig-
nificantly better than that of the Er:YAG laser (ANOVA with
Fisher’s LSD test, p < 0.05, Fig. 5). However when a plastic
excavator was used in conjunction with exp. SFC-VIII, the CRE
 journal of dentistry 39 (2011) 154–162 159
[()TD$FIG]

Fig. 4 – The parameter caries-removal effectiveness (CRE) measured in terms of the ‘Mean Relative Residual Caries Volume’
or RC/IC ratio in (a) and in terms of the ‘Mean MD at the cavity floor’ in (b). (a) Means (*) and confidence intervals (vertical
whiskers) for the ‘Mean Relative Residual Caries Volume’. The dotted line indicates the most optimal RC/IC (best score = 0),
indicating that all carious tissue was effectively removed. Horizontal bars connect groups without statistical significance,
as disclosed by ANOVA followed by Fisher’s LSD test. (b) Means (*) and confidence intervals (vertical whiskers) for MD at
the cavity floor. The dotted line indicates the cut-off point for residual caries. Horizontal bars connect groups without
statistical significance, as disclosed by ANOVA followed by Fisher’s LSD test. (c) Percentage of teeth exhibiting ‘sound’
versus ‘carious’ dentine, as determined on basis of MD. Horizontal bars connect groups without statistical significance, as
disclosed by Fisher’s exact test.

(in terms of RC/IC ratio) was clearly less favourable (0.23  0.22
compared to 0.11  0.05, 0.08  0.06 and 0.11  0.09 obtained
for Carisolv, SFC-V and SFC-VIII, respectively), and even in
range with that of the Er:YAG laser (Fig. 4a).
4. Discussion
The ideal caries-excavation technique would be the one that
selectively removes the irreversibly destroyed tissue, but
leaves the potentially remineralisable tissue at the cavity
floor. This is however hardly achievable clinically, because
even when the currently available caries-excavation tech-
niques were specific enough to remove solely the irrevers-
ibly destroyed carious tissue, some ‘sound’ or at least
potentially remineralisable tissue is still sacrificed in order
to provide instrumental access to the internal dentine caries
lesion.
In this study, Er:YAG-laser excavation resulted in both a
highly variable CRE and MIP, by which it cannot be considered
160 journal of dentistry 39 (2011) 154–162
[()TD$FIG]

Fig. 5 – The minimal-invasiveness potential parameter (MIP). Means (*) and confidence intervals (vertical whiskers) for the
Relative Cavity Size for the 9 caries-excavation methods investigated. The green dotted line represents the most optimal
MIP (best score = 1), indicating that the volume of removed tissue was equal to the volume of the initial carious tissue.
Horizontal bars connect groups without statistical significance, as disclosed by ANOVA followed by Fisher’s LSD test. (For
interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

as a selective caries-removal technique. This is confirmed by
the fact that the Er:YAG laser presented with the lowest MIP,
which is in contrast with what is often claimed by laser
manufacturers and users. These unfavourable results were
recorded despite the laser investigated in this study was
equipped with a LIF-feedback system. Previous research23,26
has shown that the staining status of residual dentine to a
large extent affected LIF. As we used a clinically representative
pool of carious lesions, including both active and inactive
lesions, we therefore hypothesise that excessive dentine may
have been removed in inactive, stained lesions, having thus
increased the ‘Relative Cavity Size’ and even lead to over-
excavation into sound dentine in some specimens. On the
[()TD$FIG]
other hand, active, less stained carious lesions were most
likely not appropriately (completely) excavated, and thus
more residual caries was left at the cavity floor (lower MD) in
other specimens.
A low coincidence between the size of the actual carious
lesion and that of the prepared cavity for the Er:YAG laser
excavation was also previously reported, typically combining
deep areas of over-excavation together with widely underpre-
pared zones.27 The results of this study are, however, not in
agreement with the study from Eberhard et al.17 They also
used an Er:YAG laser equipped with a LIF-feedback system at
the same threshold of ‘7’, as was used in our study. A smaller
‘Relative Cavity Size’, indicating better minimal-invasiveness

Fig. 6 – Summary of results obtained for caries-removal effectiveness (CRE) and minimal-invasiveness potential (MIP) for the
caries-removal techniques tested.
 journal of dentistry 39 (2011) 154–162
potential, was found for a non-contact Er:YAG laser than for a
conventional tungsten-carbide bur excavation.
When tungsten-carbide bur excavation was guided by
Caries Detector, caries was removed until dentine solely
stained ‘light-pink’, which is definitely more tissue-preserving
than the original ‘red’ staining threshold.8 Both this technique
and the solely use of a tungsten-carbide-bur resulted in the
best CRE in terms of ‘Mean Relative Residual Caries Volume’
(Fig. 4a). However, Caries Detector involves a non-negligible
risk on over-excavation into sound dentine, as was confirmed
by the highest MD of residual dentine measured at the cavity
floor and the lower MIP (or higher mean ‘Relative Cavity Size’).
This over-excavation risk corroborates other studies that even
suggested that the ‘light-pink’ threshold still leads to a too
invasive removal of tissue.5,7 Indeed, solely using a tungsten-
carbide bur appeared more conservative in our study, which is
also in agreement with another study.28
For the sono-abrasion Cariex, the mean MD at the cavity
floor was higher than the cut-off point for dentine caries
(Fig. 4b), indicating that sound dentine was exposed and no
carious tissue was left. However, CRE in terms of ‘Mean
Relative Residual Caries Volume’ (RC/IC) indicated that
residual caries was found, which thus must have remained
at areas different from the cavity floor. It is not unreasonable
that the sono-abrasion was effective at the cavity floor, but not
at the cavity walls, since indeed most residual caries was
found at the latter location. Although in literature no data is
available regarding tungsten-carbide sono-abrasion, also
diamond-coated sono-abrasion previously revealed a tenden-
cy towards underpreparation.28 Furthermore, sono-abrasion
was reported to induce a ‘compacting’ effect on carious
dentine,28 which may be an alternative explanation for the
increased MD at the cavity floor measured in this study.
For CeraBur, the lowest MD at the cavity floor was
measured (Fig. 4b). In addition, all specimens still presented
‘carious’ dentine (Fig. 4c). Different from Cariex, most carious
dentine was remained at the cavity floor itself. The only study
reporting on CeraBur showed that, although not statistically
significant, this ceramic bur tended to leave more caries at the
cavity floor as compared to a tungsten-carbide bur.10 Never-
theless, such a ceramic bur is not expected to become blunt as
easily as its precursor made of plastic material (SmartBurs,
SSWhite, Lakewood, NJ, USA). Possibly, a learning curve with
respect to the force applied during caries removal is involved
and thus may improve its CRE. Both CeraBur and Cariex
resulted, however, in the best MIP (lowest Relative Cavity Size),
meaning that these instruments most likely do not easily cut
into sound dentine tissue.
The chemo-mechanical excavation methods most closely
removed carious tissue up to the level of the actual MD cut-off
point for dentine caries (Fig. 4b). The exp. SFC-V, however,
resulted in the highest mean MD at the cavity floor, indicating
that it may remove caries less selectively. According to the
manufacturer, this difference could have been caused by some
instability of the thickening agent present in the exp. SFC-V at
room temperature; when the solution becomes more fluid, it
must penetrate deeper and dissolve more demineralised
dentine.
A previous micro-CT study has revealed MDs at the cavity
floor of 0.85 and 0.9 g/cm3, when caries was removed with
161
Carisolv or the exp. SFC-V, respectively.13 Both methods made
use of the same prototype plastic excavator, as used in this
study with the exp. SFC-VIII. Therefore, these relatively low
MDs probably resulted from the use of the plastic excavator,
which when employed along with the exp. SFC-VIII in this
study, also resulted in the highest ‘Mean Relative Residual
Caries Volume’ (Fig. 4a). In fact, during caries excavation, the
plastic instrument became easily blunt and had to be replaced
at least 3 times during treatment of one single specimen. The
use of a conventional metal spoon excavator with both
experimental SFC versions improved the CRE, as proven by
a lower ‘Mean Relative Residual Caries Volume’ compared to
that of the exp. SFC-VIII used along with the plastic excavator
(Fig. 4a). In general, chemo-mechanical methods resulted in
the best compromise, entailing a rather ‘complete’ caries
removal and adequate MIP.
5. Conclusion
The caries-removal effectiveness and minimal-invasiveness
potential varied amongst the contemporary caries-removal
techniques tested in this study.
Er:YAG-laser excavation guided by the LIF feedback system
resulted in low CRE and MIP, with highly variable results.
Whilst some specimens presented the highest volume of
residual caries, others excessively removed tissue into sound
dentine. Thus, the Er:YAG-laser excavation could not be
considered a selective caries-removal technique.
The use of rotary/oscillation caries-excavation techniques
resulted in a better CRE, with the exception of CeraBur and
Cariex, which although amongst the most conservative caries-
removing methods, showed a clear tendency to leave residual
caries at the cavity floor and walls, respectively. Tungsten-
carbide bur excavation guided by Caries Detector, resulted in
over-excavation into sound dentine, which was slightly less
the case when a tungsten-carbide bur was solely used.
Regarding chemo-mechanical methods, whilst the plastic
excavator associated with the exp. SFC-VIII resulted in a low
CRE, the other methods (exp. SFC-VIII, exp. SFC-V, and
Carisolv), when used along with a metal excavator, resulted
in the better compromise between effective and selective
caries removal.
references
1. Mount GJ. Minimal intervention dentistry: rationale of
cavity design. Operative Dentistry 2003;28:92–9.
2. Kidd EA, Joystonbechal S, Beighton D. Microbiological
validation of assessments of caries activity during cavity
preparation. Caries Research 1993;27:402–8.
3. Kidd EA. How ‘clean’ must a cavity be before restoration?
Caries Research 2004;38:305–13.
4. Ten Cate JM, Van Duinen RN. Hypermineralization
of dentinal lesions adjacent to glass-ionomer cement
restorations. Journal of Dental Research
1995;74:1266–71.
5. Kidd EA, Joystonbechal S, Smith MM, Allan R, Howe L, Smith
SR. The use of a caries detector dye in cavity preparation.
British Dental Journal 1989;167:132–4.
162 journal of dentistry 39 (2011) 154–162
6. Kuboki Y, Liu CF, Fusayama T. Mechanism of differential
staining in carious dentin. Journal of Dental Research
1983;62:713–4.
7. Yip HK, Stevenson AG, Beeley JA. The specificity of caries
detector dyes in cavity preparation. British Dental Journal
1994;176:417–21.
8. Sunago M, Nakashima S, Tagami J. Association between
staining by caries detector dye and the corresponding
mineral density in dentin caries. American Journal of Dentistry
2009;22:49–54.
9. Boston DW, Liao J. Staining of non-carious human coronal
dentin by caries dyes. Operative Dentistry 2004;29:280–6.
10. Dammaschke T, Vesnic A, Schäfer E. In vitro comparison of
ceramic burs and conventional tungsten carbide bud burs in
dentin caries excavation. Quintessence International
2008;39:495–9.
11. Yazici AR, Ozgunaltay G, Dayangac B. A scanning electron
microscopic study of different caries removal techniques on
human dentin. Operative Dentistry 2002;27:360–6.
12. Beeley JA, Yip HK, Stevenson AG. Chemochemical
caries removal: a review of the techniques and
latest developments. British Dental Journal
2000;188:427–30.
13. Clementino-Luedemann TN, Ilie AD, Hickel R, Kunzelmann
KH. Micro-computed tomographic evaluation of a new
enzyme solution for caries removal in deciduous teeth.
Dental Materials Journal 2006;25:675–83.
14. Ahmed AA, Garcia-Godoy F, Kunzelmann KH. Self-limiting
caries therapy with proteolytic agents. American Journal of
Dentistry 2008;21:303–12.
15. Dommisch H, Peus K, Kneist S, Krause F, Braun A, Hedderich
J, et al. Fluorescence-controlled Er:YAG laser for caries
removal in permanent teeth: a randomized clinical trial.
European Journal of Oral Sciences 2008;116:170–6.
16. Eberhard J, Eisenbeiss AK, Braun A, Hedderich J, Jepsen S.
Evaluation of selective caries removal by a fluorescence
feedback-controlled Er:YAG laser in vitro. Caries Research
2005;39:496–504.
17. Eberhard J, Bode K, Hedderich J, Jepsen S. Cavity size
difference after caries removal by a fluorescence-controlled
Er:YAG laser and by conventional bur treatment. Clinical Oral
Investigations 2008;12:311–8.
18. Elliott JC, Davis GR, Anderson P, Wong FSL, Dowker SE,
Mercer CE. Application of laboratory microtomography to
the study of mineralised tissues. Anales de Quimica
1997;93:S77–82.
19. Lee HS, Berg JH, Garcia-Godoy F, Jang KT. Long-term
evaluation of the remineralization of interproximal caries-
like lesions adjacent to glass-ionomer restorations: a micro-
CT study. American Journal of Dentistry 2008;21:129–32.
20. Neves AA, Coutinho E, Vivan-Cardoso M, Jaecques S, Van
Meerbeek B. Micro-CT based quantitative evaluation of
caries-excavation. Dental Materials 2010;26:579–88.
21. Willmott NS, Wong FSL, Davis GR. An X-ray
microtomography study on the mineral concentration of
carious dentine removed during cavity preparation in
deciduous molars. Caries Research 2007;41:129–34.
22. Dowker SE, Elliott JC, Davis GR, Wassif HS. Longitudinal
study of the three-dimensional development of subsurface
enamel lesions during in vitro demineralisation. Caries
Research 2003;37:237–45.
23. Neves AA, Coutinho E, Vivan-Cardoso M, Van Meerbeek B.
Residual caries determination by diagnodent after different
caries-excavation methods. Journal of Dental Research
2010;89. abstr 2038.
24. Davis GR, Elliott JC. Artefacts in X-ray microtomography of
materials. Materials Science & Technology 2006;22:1011–8.
25. Lennon AM, Attin T, Buchalla W. Quantity of remaining
bacteria and cavity size after excavation with face, caries
detector dye and conventional excavation in vitro. Operative
Dentistry 2007;32:236–41.
26. Karlsson L, Johansson E, Tranaeus S. Validity and reliability
of laser-induced fluorescence measurements on carious
root surfaces in vitro. Caries Research 2009;43:397–404.
27. Celiberti P, Francescut P, Lussi A. Performance of four
dentine excavation methods in deciduous teeth. Caries
Research 2006;40:117–23.
28. Banerjee A, Kidd EA, Watson TF. In vitro evaluation of five
alternative methods of carious dentine excavation. Caries
Research 2000;34:144–50.