Caries detector dyes - An in vitro assessment of some new compounds

Abstract

Previous studies have shown that the caries detector dyes, basic fuchsin and acid red, lack specificity. Accordingly, their clinical use can lead to the unnecessary removal of sound tissue. In the present study, the specificity of three further dyes, Carbolan Green, Coomassie Blue and Lissamine Blue was studied. Carious dentine was removed in vitro by means of rotary instruments until the cavities were deemed caries free by conventional clinical criteria. Experimental dyes were applied to the cavity floors, all of which became stained. Stained dentine was removed from half the cavity by means of a burr, the other half remaining as a control. Further stain was then applied and the procedure repeated until no further reduction of the staining of the cavity floor could be achieved. Light microscopy of ground sections of experimental teeth showed that sound tissue had been removed unnecessarily from the experimental half of the cavity due to the lack of specificity of these dyes. This lack of specificity of staining was similar to basic fuchsin and acid red. Only Carbolan Green showed possible differential staining between control and experimental sites, but this was not caries specific. If a clinically useful dye is to be developed, it would need to specifically stain either bacteria in infected dentine and/or the carious degradation products of dentine matrix.

Full text

Journal of Oral Rehabilitation 1999 26; 453–458
Caries detector dyes—an in vitro assessment of some new
compounds
G. ANSARI, J. A. BEELEY, J. S. REID & R. H. FOYE University of Glasgow Dental School,Glasgow,U.K.
repeated until no further reduction of the staining
SUMMARY
Previous studies have shown that the
of the cavity floor could be achieved. Light mi-caries detector dyes, basic fuchsin and acid red, lack
croscopy of ground sections of experimental teethspecificity. Accordingly, their clinical use can lead
to the unnecessary removal of sound tissue. In the showed that sound tissue had been removed unnec-
essarily from the experimental half of the cavitypresent study, the specificity of three further dyes,
due to the lack of specificity of these dyes. This lackCarbolan Green, Coomassie Blue and Lissamine
of specificity of staining was similar to basic fuchsinBlue was studied. Carious dentine was removed in
and acid red. Only Carbolan Green showed possiblevitro by means of rotary instruments until the cavi-
ties were deemed caries free by conventional clini- differential staining between control and experi-
cal criteria. Experimental dyes were applied to the mental sites, but this was not caries specific. If a
clinically useful dye is to be developed, it wouldcavity floors, all of which became stained. Stained
need to specifically stain either bacteria in infecteddentine was removed from half the cavity by means
of a burr, the other half remaining as a control. dentine and/or the carious degradation products of
dentine matrix.Further stain was then applied and the procedure
Introduction
Cavity preparation prior to restoration requires com-
plete removal of carious dentine. The process is nor-
mally deemed complete when the dentine surface
appears hard on probing and is stain free. Determining
what is and what is not remineralizable dentine is
basically a clinical judgement. Although tactile and
visual criteria are normally used, caries detector dyes
may also be employed. To be clinically useful, a caries
detector dye must selectively stain infected demineral-
ized dentine: if so it would act as a clinical guide for
removal of irreversibly decalcified tissue.
The possibility of caries detector dyes was originally
developed in the 1970s when basic fuchsin staining
was used as a guide to the removal of the outer layer of
infected unremineralizable dentine in carious lesions
(Fusayama & Terachima, 1972; Sato & Fusayama,
1976). There has always been concern about the safety
of disclosing agents in terms of carcinogenicity (Bonser,
Glayson & Jull, 1956) and 1% acid red in propylene
glycol was introduced as a safe and effective alternative
(Fusayama, Takatsu & Itoh, 1979). Both dyes have
been used clinically as caries detector dyes (Anderson
& Charbeneua, 1985; List et al., 1987), and both com-
pounds appeared to stain demineralized organic matrix
rather than bacteria (Boston & Graver, 1989).
Recent work has shown that neither dye is caries
specific, especially in deep cavities and in the vicinity of
the enamel–dentine junction (Yip, Stevenson &
Beeley, 1994). A large proportion of pulpal floors
judged sound by conventional clinical criteria were still
stainable with acid red and the deeper the cavity, the
greater the possibility of finding stainable dentine after
removal of caries (Kidd et al., 1989). Microbiological
studies on stained and unstained sites from cavities
deemed clinically sound resulted in the recovery of low
numbers of bacteria with no difference in the levels of
infection between them (Kidd, Joyston-Bechal &
Beighton, 1993). Povidone-iodine dyes produced simi-
© 1999 Blackwell Science Ltd 453

454 G. ANSARI et al.
lar results to acid red and also lacked specificity for
carious dentine (Maupome et al., 1995).
Circumpulpal dentine and the enamel–dentine junc-
tion in sound teeth could also be stained with these
dyes. Back scattered electron imaging showed that the
level of mineralization in these areas of the tooth was
lower than elsewhere even in sound teeth (Yip et al.,
1994). Staining would, therefore, appear to result from
reduced mineralization rather than demineralization
and caries.
This paper describes a study in which three further
dyes, Carbolan Green, Coomassie Blue and Lissamine
Blue were compared with basic fuchsin and acid red in
a further attempt to identify a caries detector dye
which might specifically stain carious dentine only and
therefore be clinically useful.
Materials and methods
Carbolan Green, Coomassie Blue (type G) and Lis-
samine Blue solutions were kindly supplied by Imperial
Chemical Industries (now Zeneca)*.
Seven dye solutions were tested: 1% (w/v) Carbolan
Green in propylene glycol, 1% (w/v) aqueous Carbolan
Green, 1% (w/v) Coomassie Blue G in propylene gly-
col, 1% (w/v) aqueous Coomassie Blue G, 1% (w/v)
aqueous Lissamine Blue, 1% (w/v) acid red in propy-
lene glycol and 0·5% (w/v) basic fuchsin.
Thirty-seven freshly extracted teeth with small class I
or class II cavities without pulpal involvement were
selected for the investigation. The teeth were placed in
phosphate buffered saline pH 7·2 (Dubelcco’s For-
mula
†
) immediately after extraction and experimental
procedures carried out within 2 h.
Carious material was removed from the cavities us-
ing a size 3–5 sterile burr with a slow speed handpiece.
Dentine was removed until the cavity was considered
to be clinically sound by the usual visual and tactile
criteria (Kidd et al., 1993). One of the dyes was then
applied to the floor of each prepared cavity using a
sterile cotton pledget for 10 s. Excess dye was thor-
oughly washed away with distilled water until no more
dye could be removed and the surface dried witha3in
1 syringe. Dye stained dentine was removed from one
half of the cavity with a sterile bur, the size of the bur
depending on the cavity size. The other half of the
cavity was left untouched as a control. The entire
cavity was then given a second application of dye, the
surplus removed by washing and drying, and a fresh
sterile bur used to remove stained dentine from the
experimental half of the cavity, the control area again
being left untouched. The procedure was repeated un-
til no further reduction in the staining of the dentine in
the experimental part of the cavity could be achieved.
A proforma for recording the different details of this
trial was filled in for each individual tooth and in-
cluded a diagram showing the outline of the prepared
cavity including the control and experimental sides. A
mark was then cut on the buccal surface of each tooth
to facilitate orientation of the specimens in subsequent
sectioning. On completion of the experiment, teeth
were fixed in 10% neutral buffered formalin solution.
Ground sections (150 mm thick) of the teeth were
prepared such that both control and experimental
halves of each cavity were included in each section.
The teeth were embedded in tanwax and sectioned
with a diamond wheel on a hard tissue saw (Microslice
2
‡
). The sections were then examined by light
microscopy.
Results
With all five dyes (seven solutions), areas of dentine on
the cavity floors were stained to some extent in all of
the cavities studied, even when deemed clinically
sound by conventional criteria. Mechanical removal of
stained dentine from half of the cavity with a bur
Table 1. Data from the in vitro study of caries detector dyes
Dye solution No. of teeth Mean no. of ap-
plications
9Lissamine Blue (aqueous) 3·8
3·24Coomassie Blue G (in propy-
lene glycol)
47Coomassie Blue G (aqueous)
4·2Carbolan Green (in propy- 4
lene glycol)
6 4·1Carbolan Green (aqueous)
Acid red (in propylene gly- 33
col)
4 4·2Basic fuchsin
* Grangemouth, Scotland.
†
ICN, Flow, CA, U.S.A.
‡
Malvern Instruments Ltd, Worcestershire, U.K.
© 1999 Blackwell Science Ltd, Journal of Oral Rehabilitation 26; 453–458

CARIES DETECTOR DYES 455
Fig. 1. Ground section prepared
from a tooth treated with aqueous
Carbolan Green. C, control half of
cavity; E, experimental half of
cavity. Arrows indicate areas of
increased removal of dentine (×30).
followed by reapplication of the dye led to some reduc-
tion in the staining intensity. Complete removal of
stainable dentine could not be achieved. The procedure
was repeated until no further reduction in intensity of
staining was possible. Approximately four dye applica-
tion/stained dentine removal cycles were needed to
achieve this (range 2–6) and the results are shown in
Table 1.
All five dyes (seven solutions) lacked specificity in
staining and required a similar number of applications
to achieve minimum staining. However, aqueous Car-
bolan Green did tend to show less staining of sound
dentine on cavity floors, but this was a subjective
finding and the difference between this and carious
dentine was not caries specific.
Light microscopy of the ground sections prepared
from the treated cavities showed that sound dentine
had been removed unnecessarily in all cases where
dentine removal had continued until dye staining was
minimal. Generally, the cavity floor was slightly
brighter where additional dentine had been removed
following dye application. A section from a cavity
treated with aqueous Carbolan Green is shown in Fig.
1. Figure 2 shows a similar section involving the use of
Coomassie Blue G in propylene glycol. All of the dyes
studied were clearly found to be non caries-specific. In
these ground sections, reactionary dentine beneath the
cavity floor is indicated by the presence of dead tracts.
These were found in areas close to the pulp and were
not related to the type of dye used. The experimental
side of the cavities was deeper in the experimental half
of the tooth compared to the control site indicating
unnecessary tissue removal. This was more evident
with teeth treated with Coomassie Blue G (Fig. 1) and
Lissamine Blue; when aqueous Carbolan Green was
used, less tissue was unnecessarily removed (Fig. 2)
possibly because of a reduction in intensity of staining
of sound dentine.
Discussion and conclusions
According to Fusayama et al. (1979), basic fuchsin
stains the outer layer of carious dentine but not the
inner. This outer layer is infected, highly degraded and
unremineralizable and therefore must be removed
prior to restoration. In contrast the inner layer is not
infected and has been invaded only by bacterial prod-
ucts. Collagen fibrils are not degraded and the structure
can be remineralized. Fusayama & Terachima (1972)
also separated lesions into acute and chronic in terms
of stainability. They postulated that in an acute lesion,
heavier staining occurred because of the lower dentine
hardness, whereas in a chronic lesion, lighter staining
is observed because of the harder dentine in the level
below.
© 1999 Blackwell Science Ltd, Journal of Oral Rehabilitation 26; 453–458

456 G. ANSARI et al.
Fig. 2. Ground section prepared
from a tooth treated with Coomassie
Blue G in propylene glycol.
Labelling is as in Fig. 1 (×30).
Use of a dye to selectively identify this outer layer is
therefore potentially extremely valuable in clinical
dentistry. Subsequent studies however, showed that
the use of such dyes (Boston & Graver, 1989) resulted
in the removal of tissue beyond the level of bacterial
invasion, whilst Anderson, Loesche & Charbeneau
(1985) and List et al. (1987) found that a significant
number of their specimens contained bacteria in the
unstained dentine. More recently (Kidd et al., 1989;
Kidd et al., 1993) found that not only was the staining
of dentine not caries specific, increased staining oc-
curred at the enamel–dentine junction and in deeper
areas of dentine in the layers closer to the pulp. When
carious dentine removal was deemed complete by nor-
mal clinical criteria, acid red stainable dentine still
remained but the levels of bacteria in this material
were both low and did not significantly differ from the
levels in non-stainable sound dentine. These findings
would suggest that acid red and basic fuchsin staining
is a function of level of mineralization rather than a
carious invasion. The findings of Yip et al. (1994) in-
volving the use of back scattered electron imaging
confirmed this hypothesis when they showed that the
level of mineralization in the enamel–dentine junction
and adjacent to the pulp was indeed reduced as com-
pared with the remainder of the dentine even in clini-
cally sound teeth. Staining with dyes such as basic
fuchsin and acid red would probably therefore seem to
be a function of availability of binding sites on the
organic (collagen) matrix.
In the present study, the lack of specificity of basic
fuchsin and acid red was confirmed and three further
dyes, Carbolan Green, Coomassie Blue (type G) and
Lissamine Blue, all of which predominantly stain
proteins, have been investigated. Their structures and
those of acid red and basic fuchsin are shown in Fig. 3.
These three new dyes have also been found to be
non-specific in their staining of dentine. Carbolan
Green, however, was found to cause less non-specific
staining than any of the other four dyes. Sections of
teeth stained with these dyes suggest that in addition
to the lower mineral content, staining intensity may
also increase as the pulp is approached because stain
can more easily penetrate the wider tubules.
Whilst the nature of the binding of dye molecules to
proteins is poorly understood, stacking of polyaromatic
residues, perhaps associated with proline residues, may
be involved. Carbolan dyes, however, have long hy-
drocarbon side-chains and are more hydrophobic than
the other four (Fig. 3). They are, therefore, likely to
bind less strongly to collagen which is hydrophilic and
have greater affinity for hydrophobic residues, such as
bacterial lipopolysaccharides. This may explain the ba-
sis of the finding that this reagent stained sound den-
tine less intensely, i.e. was less non-specific than the
other four.
© 1999 Blackwell Science Ltd, Journal of Oral Rehabilitation 26; 453–458

CARIES DETECTOR DYES 457
Fig. 3. Molecular structures of: A,
basic fuchsin; B, acid red-1; C,
Carbolan Green; D, Lissamine Blue;
and E, Coomassie Blue G.
However, the general conclusion from this study is
that none of the three new dyes studied offers any real
advantages over previous caries detector dyes and that
their use, like previous ones, would also lead to the
unnecessary removal of sound dentine.
If a specific caries detector dye is to be developed in
the future, it will probably need to be specifically di-
rected against cariogenic bacteria and/or degradation
products of dentine matrix.
References
A
NDERSON
,M.H.&C
HARBENEUA
, G.T. (1985) A comparison of
digital and optical criteria for detecting carious dentin. Journal
of Prosthetic Dentistry,53, 643.
© 1999 Blackwell Science Ltd, Journal of Oral Rehabilitation 26; 453–458

458 G. ANSARI et al.
A
NDERSON
, M.H., L
OESCHE
,W.J.&C
HARBENEAU
, G.T. (1985)
Bacteriologic study of a basic fuschin caries disclosing dye.
Journal of Prosthetic Dentistry,54, 51.
B
ONSER
, G.M., G
LAYSON
, D.B. & J
ULL
, J.W. (1956) The induction
of tumors of the subcutaneous tissues, liver and intestine in the
mouse by certain dye stuffs and their intermediates. British
Journal of Cancer,10, 653.
B
OSTON
,D.W.&G
RAVER
, H.T. (1989) Histological study of an acid
red caries-disclosing dye. Operative Dentistry,14, 186.
F
USAYAMA
, T., T
AKATSU
,T.&I
TOH
, K. (1979) New composition of
caries detector. Japanese Journal of Operative Dentistry,22, 37.
F
USAYAMA
,T.&T
ERACHIMA
, S. (1972) Differentiation of two
layers of carious dentine by staining. Journal of Dental Research,
51, 866.
K
IDD
, E.A., J
OYSTON
-B
ECHAL
,S.&B
EIGHTON
, D. (1993) The use of
a caries detector dye during cavity preparation: a microbiologi-
cal assessment. British Dental Journal,174, 245.
K
IDD
, E.A., J
OYSTON
-B
ECHAL
, S., S
MITH
, M.M., A
LLAN
, R., H
OWE
,
L.&S
MITH
, S.R. (1989) The use of a caries detector dye in
cavity preparation. British Dental Journal,167, 132.
L
IST
, G., L
OMMEL
, T.J., T
ILK
, M.A. & M
URDOCH
, H.G. (1987) Use of
a dye in caries identification. Quintessence International,18, 343.
M
AUPOME
, G., H
ERNANDEZ
-G
UERRERO
, J.C., G
ARCIA
-L
UNA
, M.,
T
REJO
-A
LVARADO
, A., H
ERNANDEZ
-P
EREZ
,M.&D
IEZ
-
DE
-B
ONILLA
,
J. (1995) In vivo diagnostic assessment of dentinal caries utilis-
ing acid red and povidone-iodine dyes. Operative Dentistry,20,
119.
S
ATO
,Y.&F
USAYAMA
, T. (1976) Removal of dentine by fuchsin
staining. Journal of Dental Research,55, 678.
Y
IP
, H.K., S
TEVENSON
,A.G.&B
EELEY
, J.A. (1994) The specificity
of caries detector dyes in cavity preparation. British Dental
Journal,176, 417.
Correspondence: Dr Josie A. Beeley, Oral Sciences, University of
Glasgow Dental School, 378 Sauchiehall Street, Glasgow, G2 3JZ,
U.K. E-mail: j.a.beeley@dental.gla.ac.uk.
.
© 1999 Blackwell Science Ltd, Journal of Oral Rehabilitation 26; 453–458
.