Original Paper

Caries Res 2000;34:144150

Received: September 2, 1999 Accepted after revision: October 12, 1999

In vitro Evaluation of Five Alternative Methods of Carious Dentine Excavation

A. Banerjee E.A.M. Kidd T.F. Watson
Division of Conservative Dentistry, Guys, Kings and St. Thomas Dental Institute, KCL, London, UK

Key Words Air-abrasion Bur Carisolv Confocal microscopy Dentine caries Excavation Fluorescence Sonoabrasion

ventional hand excavation appeared to offer the best combination of efficiency and effectiveness for carious dentine excavation within the parameters used in this study.
Copyright 2000 S. Karger AG, Basel

Abstract This in vitro, split-tooth study aimed to evaluate the efficiency (time taken) and effectiveness (quantity of dentine removed) of four techniques of carious dentine excavation (bur, air-abrasion, sono-abrasion and Carisolv gel) compared to conventional hand excavation. Eighty freshly extracted human molars were assigned to four experimental groups (n = 20), sectioned longitudinally through occlusal lesions and pre-excavation colour photomicrographs obtained. Using the natural autofluorescence of carious dentine (detected using confocal laser scanning microscopy) as an objective and reproducible guide, carious dentine removal was assessed in each half of the split tooth sample, comparing hand excavation to the test method. The time taken to reach a cavity floor that was hard to a dental probe was noted and final colour photomicrographs were taken. From the results, it was concluded that bur excavation was quickest but overprepared cavities relative to the autofluorescent signature, whereas Carisolv excavation was slowest but removed adequate quantities of tissue. Sono-abrasion tended to underprepare whereas airabrasion was more comparable to hand excavation in both the time and amounts of dentine removed. Con-

When operative treatment is indicated in deep dentinal lesions, quantities of softened carious tissue are removed with the ultimate aims of eliminating the highly infected biomass of tissue to prevent further lesion progression and leaving a cavity which can be suitably restored for strength and function [Elderton, 1984; Kidd et al., 1993; Frencken et al., 1994; Bjrndal et al., 1997; Mertz-Fairhurst et al., 1998; Weerheijm et al., 1999]. Currently, the physical criterion used most commonly by dental practitioners to guide clinical excavation of this infected, demineralised dentine is the hardness/texture of the tissue, although some dentists may take into account its colour and may use caries detector dyes [van de Rijke, 1991; Kidd et al., 1993, 1996]. All of these criteria suffer from the subjectivity inherent between dentists in clinical practice, which is likely to result in variations in the quality and quantity of dentine removed during operative intervention. These variations may have clinical implications including differences in the size of the cavities produced, the pulpal health beneath prepared cavities and the strength of the remaining tooth structure. It seems sensible, therefore, to develop an objective marker for excavatable carious dentine. This objectivity is also needed in the research laboratory as well as the clinic. These are interest-

2000 S.Karger AG, Basel 00086568/00/03420144 $17.50/0 Fax +41 61 306 12 34 E-Mail karger@karger.ch www.karger.com Accessible online at: www.karger.com/journals/cre

Dr. Avijit Banerjee Floor 26, Guys Tower, Guys, Kings and St. Thomas Dental Institute London Bridge, London SE1 9RT (UK) Tel. +44 (0)207 955 5000, Ext 3604, Fax +44 (0)207 955 8740 E-Mail avijit.banerjee@kcl.ac.uk

ing times with alternative techniques including air-abrasion and Carisolv gel becoming more widely available to practitioners [Goldstein and Parkins, 1994; Renson, 1995; Banerjee, 1999; Ericson et al., 1999]. These alternative excavation techniques require laboratory as well as clinical evaluation but at present there are no clear, objective and reproducible histological markers available to delineate the heavily infected dentine that ought to be removed and thus allow accurate comparisons between techniques to be made. The use of natural fluorescence or autofluorescence (AF) to delineate the carious dentine requiring excavation is under investigation. In vitro studies have shown that carious dentine exhibits natural AF detectable using confocal laser scanning microscopy [van der Veen and ten Bosch, 1996; Banerjee and Boyde, 1997, 1998; Banerjee, 1999]. The methodology used in these studies allowed the AF signal to be accurately related to the overall structure of the carious lesion in the laboratory. In laboratory studies mimicking clinical caries removal, a correlation was found between the amount of carious dentine excavated using a hand excavator and the distribution of the AF signal [Banerjee et al., 1997, 1998; Banerjee, 1999]. In this instance, tissue excavation was guided by the hardness of dentine; excavation was terminated at the level where hard dentine was first detected using a dental probe. Results implied that the AF was associated with the relative hardness of carious dentine as determined by a dentist during excavation [Banerjee et al., 1998; Banerjee, 1999]. Indeed, the relationship between the hardness of carious dentine and its AF has been examined further in vitro; results from Knoop hardness measurements indicated a positive correlation between the relative hardness of the carious tissue and the distribution of the AF [Banerjee et al., 1999]. The AF of carious dentine was used as the histological validation method for carious dentine requiring excavation as studies have indicated that it corresponds to the zone of tissue in which the high numbers of bacteria present have interacted with the surrounding dentine matrix. This interaction has resulted in the generation or dequenching of a chromophore a fluorescent moiety residing in the dentine matrix which is therefore responsible for the detectable AF signal [Banerjee and Boyde, 1998; Banerjee, 1999; Banerjee et al., 1999]. There are now a number of techniques available to the dentist that may remove the reversibly demineralised dentine at the depths of lesions as well as the irreversibly demineralised and heavily infected tissue more superficially. These techniques include the two conventionally used techniques, hand excavation and the slow-speed round bur, as well as air-abrasion (using a novel alumina-hydroxyapatite

powder mix), sonic abrasion (or sono-abrasion) using diamond-coated oscillating tips and chemo-mechanical excavation using Carisolv gel. The primary aim of this study was to evaluate the efficiency (time taken) and the effectiveness (amount of dentine removed) of these methods compared to the gold standard of the universal and simple process of hand excavation. In order to gain clinical objectivity, this in vitro, split-tooth study used the AF as the objective criterion for the irreversibly demineralised and infected, carious dentine requiring removal.

Materials and Method

Eighty extracted carious human molars, stored in distilled water for no longer than 2 weeks after extraction, were allocated to four different technique groups (n = 20). The teeth were sectioned longitudinally through the occlusal lesions using a diamond-impregnated circular saw (Labcut 1010, Agar Scientific Ltd., Essex, UK). The cut surfaces of the two pieces of each sample were visually examined; if there was a discrepancy between the areas of the lesions on each section (i.e. the lesion had not been sectioned through its midline into two equal halves), the sample was discarded. After grid reference lines were lightly scored onto the cut face of all sections using a scalpel blade, the samples were mounted on putty on a glass slide and kept hydrated throughout all the experiments. The autofluorescent (AF) signatures of the lesions were captured using a Noran Odyssey confocal laser scanning microscope (Middleton, Wisc., USA) operating in fluorescence mode. A 2.5/0.08 dry objective was used in conjunction with a 488-nm excitation filter (allowing blue light on to the sample) and a c515 nm emission filter (to allow yellow-green light into the detection optics of the microscope). The output of the 488-nm argon ion laser line was 1.8 mW and was operated at 50% intensity. A 15-m confocal slit was selected and standardised gain and contrast settings were used throughout. The 512E512E8 bit digitised grey scale images (with 256 frame averaging) were captured using Metamorph image analysis software (v 1.1, Universal Imaging Corp., Pa., USA) and stored on a computer hard disk. Colour photomicrographs of all samples were obtained on film (Kodak Elite Ektachrome, ASA 400) at the same magnification and were processed onto Photo CD (Kodak Ltd., Rochester, N.Y., USA) [Banerjee et al., 1999]. Any overlying carious enamel was then removed using a water-cooled, tungsten carbide fissure bur in an air-turbine handpiece. A line running parallel to the cut edge of the section and separated from it by 1.52 mm, depending on the size of each individual lesion buccolingually, was then scored onto the occlusal surface of the carious dentine, using a sharp scalpel blade. Hand Excavation In each technique group, one half of each tooth was excavated by hand and this served as the control reference. Hand excavation using a straight spoon excavator (1.30 mm diameter, Ash G5, Claudius Ash Ltd., Potters Bar, Herts., UK) was commenced from the occlusal aspect of the hydrated lesion using the scored line as an excavation boundary buccolingually. The excavation procedure was terminated when clinically hard dentine was detected using a dental probe at the base of the cavity. The time taken for caries excavation was noted.

Excavation of Carious Dentine

Caries Res 2000;34:144150

145

Carbon-Steel Bur A size 3 (ISO No. 012), Ash, round, carbon-steel bur was used in a slow-speed, Micromega 40E contra-angled handpiece (Claudius Ash Ltd.). The maximum cutting dimension of the bur was 1.21 mm. The dentine was cut from the occlusal aspect using the scored line as a boundary and tissue removal was terminated when the dentine was hard to the probe. The time taken to complete this procedure, including any pauses in which the cavity surface was washed, dried and checked, was recorded. Carisolv Gel The gel was used as per the manufacturers instructions (MediTeam, Gothenburg, Sweden). Carisolv consists of two carboxymethylcellulose-based gels: (1) a red gel containing 0.1 M amino acids (glutamic acid, leucine and lysine), NaCl, NaOH, erythrosine and purified water, and (2) a clear liquid of sodium hypochlorite (NaOCl0.5% w/v). The two separate gels, which had been stored at 4C before use, were allowed to regain room temperature (the laboratory temperature was 31.5 C) and were then mixed in their interconnecting vials until homogenised. The round-head excavator with a maximum diameter of 1.32 mm was selected, placed in the gel and a drop of the solution carried to the hydrated surface of the carious dentine. After 30 s firm, cyclical hand abrasion was performed until a clinically hard cavity surface was attained. During the preparation, the gel was washed away periodically prior to checking the hardness of the dentine surface with a probe. If further excavation was deemed necessary, more gel was placed onto the surface and the excavation continued. Again, the time taken was recorded in seconds. Sono-Abrasion The Sonicsys micro-oscillating, hemispherical diamond-coated tip was selected (diameter 1.5 mm) and attached to the KaVo Sonicflex 2000 L/N airscaler handpiece (KaVo Dental Ltd., Amersham, Bucks., UK) which provided sonic frequency oscillations (d6.5 kHz). Again, using the scored line as a boundary, carious dentine removal was instigated with this system and the excavation was terminated when hard dentine remained at the prepared surface, tested with a dental probe drawn across the surface. Copious water irrigation was used and the time taken was recorded. Air-Abrasion A Micrograver II Precision Micro-sandblaster (Danville Eng Ltd., San Ramon, Calif. USA) was used at an operating pressure of 80 pounds per square inch. The powder was a pre-prepared mixture of aluminium oxide and hydroxyapatite in a volume ratio of 3:1 with a particle size ranging from 3 to 60 m. The internal nozzle diameter was measured at 0.75 mm and this was used at an operating distance d5 mm. Carious dentine excavation was deemed complete when hard dentine was detected using a clinical probe and the time taken to achieve this was measured in seconds. Post-excavation, the final cavities were again imaged on colour slide transparency film (Kodak Elite Ektachrome, ASA 400). In order to numerically analyse the extent of caries removal, the final prepared cavity outline was compared to the digitally superimposed AF signature outline of the original lesion. A straight line (AB, fig. 1) was projected onto the post-excavation photomicrograph joining two EDJ points on the cavity margin (i.e. points A and B). From its halfway point (X), a perpendicular line was dropped until it reached the cavity border (point 1). The distance X1 was measured in micrometres; after superimposing the AF signal outline, the distance between point X

Fig. 1. Image showing the data measurement points after excavation. The line A213B is the cavity outline viewed from the flat, longitudinal cut surface of the sample and the criss-crossing, diagonal lines below are the grid reference lines (GRL) used for image superimposition. The enamel (E) and EDJ can be seen on the left of the image (fieldwidth = 3 mm).

and the AF signal border on the corresponding axis was also measured in micrometres and any difference between the two values noted as a percentage of the original AF signature depth. Two further measurements were used, X2 and X3, which bisected the 90 angles, AX1 and BX1. In this way, the cavity was designated as being overprepared (i.e. negative % value) or underprepared (i.e. a positive % value) relative to the AF signature at the three points being measured.

Results

Statistical analysis was performed using PC-based software, Stata v. 5.0 and StatXact v. 3.02 (Cytel Software Co., Cambridge, USA).
Efficiency Table 1 shows the mean of the 20 excavation times taken in each of the four technique groups. A non-parametric, Wilcoxon signed rank test showed statistically significant differences between the method excavation times compared to those of the corresponding hand excavation times within each of the four different technique groups (p d0.05). The results of one-way analysis of variance (ANOVA) showed there to be significant differences between the four methods (p d0.001). Bonferroni analysis (table 2) indicated significant differences in the mean excavation times between various method combinations except for sono-abrasion and air-

146

Caries Res 2000;34:144150

Banerjee/Kidd/Watson

Fig. 2. Box and whisker plot of excavation times using the four methods relative to hand excavation (0). Negative times (left) indicate the method is slower than hand excavation and positive times, quicker. ANOVA tests indicated statistically significant differences between the test methods and hand excavation (p d0.001).

Table 1. Mean excavation times and standard deviations (SD) of the 20 samples in each of the four technique groups, comparing the technique to the control hand excavation

Technique group

Cavity preparation time, s mean SD 34.4 17.4 32.5 35.0 46.9 92.5 38.8 41.4

abrasion. A box and whisker plot (fig. 2) showed clearly that the bur method was the quickest of all the gel excavation technique the slowest. Sono-abrasion and air-abrasion seemed to remove the carious dentine with similar clinical efficiency to hand excavation.
Effectiveness The numerical data for the five different excavation techniques used can be seen in table 3. Statistical analysis using one-way, Bonferroni multiple comparison of the mean differences showed there to be significant statistical differences between some of the five excavation techniques (p d0.05, table 4, shaded cells). The gel, air-abrasion and hand excavation methods seemed to prepare cavities of a similar extent when related to the AF signatures of the lesions, whereas the bur technique tended to overprepare cavities and the sono-abrasion tended to underprepare its cavities.

Bur Sonic Carisolv Air-abrasion

hand bur hand sono-abrasion hand gel hand air

107.2 57.1 86.0 98.0 93.3 216.5 81.1 99.6

Table 2. Bonferronis multiple comparison of the difference in the mean excavation times for each group combination

Bur Sonic Gel Air-abrasion 62.15 secs 163.05 66.45

Sonic 100.90 4.30

Gel 96.60

Discussion

Negative values indicate the methods in the rows were slower than those in the columns and vice versa. Shaded cells indicate those method combinations with significant differences in their mean excavation times (pd0.05).

In order for studies assessing clinical operative technique to have clinical relevance, the carious dentine lesions should be naturally produced as opposed to the in vitro induced lesions created by either acid demineralisation [van der Veen and ten Bosch, 1996] or the in vitro microbial caries system [Gilmour et al., 1990, 1997]. However, variations will still occur in the size, shape, depth of penetration, natural history and state of activity of in vivo lesions in the

Excavation of Carious Dentine

Caries Res 2000;34:144150

147

same patient. Therefore, a split-tooth methodology was used in order to minimise these variables as the source of the carious dentine would be exactly the same in each hemisected sample, thus allowing comparisons to be made between different, paired excavation methods. The samples were stored whole in distilled water for a short post-extraction period as past studies indicated that there are no significant changes in the structural, biochemical and mechanical components of dentine when using this storage protocol [Jameson et al., 1993, 1994; Banerjee, 1999]. In this study, the primary objective was to compare four alternative mechanical excavation techniques to hand excavation, in terms of their efficiency and effectiveness of carious removal. In order to exclude interoperator subjectivity, a single operator was used after technique practice was performed in order to gain familiarity with each method. Indeed, in the case of Carisolv gel, instruction was provided by the manufacturer. In order to maintain the clinical realism of the study, it was decided that clinical criteria used by dentists were to be followed: the detection of dentine that was hard to a hand instrument as it was passed across the prepared cavity surface [Kidd et al., 1993, 1996].
Efficiency In this instance, the main factor to be controlled was the volume of tissue to be excavated. Comparative results between the technique and control would be meaningless if grossly different volumes of tissue were being removed from the samples. Therefore, the volumes were controlled as closely as possible by using the split-tooth method where the lesion was sectioned through its midline into two equal halves and by limiting the depth of the excavation buccolingually using a scored line as reference and using similar diameter cutting tips with the different techniques. The results indicated that bur excavation was the quickest method out of all those tested and the gel method was the slowest (with a mean difference in excavation times between the techniques approaching 3 min, table 1). The latter finding was probably due to the fact that the gel method was essentially a form of gel-assisted hand abrasion as opposed to actual excavation. An earlier form of chemomechanical caries removal, the Caridex system (GK-101E), a water-based hand excavation method [Goldman and Kronman, 1976] from which Carisolv was developed, had some similarities with this newer technique. Interestingly, Zinck et al. [1988] showed that the Caridex system took between 4 and 10 min longer than a bur to prepare clinically acceptable cavities, but this was not performed as a split-tooth comparative study so the variables mentioned earlier may have affected the results.

Table 3. Mean percent differences between the cavity and AF outline measured at the three reference points on each sample for the five excavation groups, with their associated SEM and 95% CI

Technique

Point

Differences, % mean SEM 0.61 0.63 0.56 1.82 1.28 2.02 1.48 1.48 1.63 1.19 0.95 1.04 1.90 2.63 2.21 95% CI 5.588.00 6.178.68 6.168.38 19.7812.12 20.4815.12 17.94 9.51 5.1911.40 7.0313.22 5.5112.34 16.0921.10 18.2822.27 18.1022.45 6.6914.66 0.5310.48 1.3310.57

Hand excavation (control) 1 2 3 Carbon-steel bur 1 2 3 1 2 3 1 2 3 1 2 3

6.79 7.43 7.27 15.98 17.80 13.73 8.30 10.13 8.93 18.60 20.28 20.28 10.68 4.98 5.95

Carisolv gel

Sono-abrasion

Air-abrasion

Negative percent values indicate overprepared cavities relative to the AF signature outline.

Table 4. Bonferronis multiple comparison of the difference in the mean percent values from the three reference points between each group combination

Hand excavation Bur 22.77 25.23 20.99 11.81 12.85 13.01 1.51 2.70 1.66 3.89 2.45 1.32

Bur

Sonic

Gel

Sonic

34.58 38.08 34.00 24.28 27.93 22.65 26.65 22.78 19.68 10.30 10.15 11.35 7.93 15.30 14.33 2.38 5.15 2.98

Gel

Abrasion

Shaded cells indicate combinations with significant differences (pd0.05) in their extent of tissue removal relative to the AF signal. Negative values indicate that techniques in the rows remove more dentine (relative to AF) than the techniques in the columns.

148

Caries Res 2000;34:144150

Banerjee/Kidd/Watson

Effectiveness The tripod data analysis configuration used to measure the relative cavity depths prepared by the various techniques allowed the corresponding points to be measured in each sample without introducing bias into the analysis. The fact that the bur method of excavation tended to overprepare cavities could be explained by the lack of sensitivity of the tactile feedback when using this method. This resulted in gross, rapid removal of tissue with reduced control over the whole process. Thus, it was not always apparent to the operator when the true clinical endpoint was reached so the excavation procedure continued into healthier dentine leading to eventual overpreparation. Interestingly, the sono-abrasive method tended to underprepare cavities relative to their AF signatures, significantly more than any of the techniques tested. It is possible that the diamond-coated tip oscillation was being translated to vibration at the dentine surface and this may have had a compacting effect upon the carious dentine substrate. This would lead to an overall increase in apparent hardness of the cavity surface so giving the operator a false indication of the clinical endpoint having been reached. It should be noted that although the methods of carious dentine removal differed, the method of assessing the excavation endpoint (by using a probe) remained the same throughout the experiments. The air-abrasive system, even though producing numerical results that implied a good correlation with the AF marker, also had the greatest range in the 95% confidence interval values at the three reference points (ranging from 0.53 to 14.66%, table 3). This indicated a degree of variation in the effectiveness of the technique and more samples would be required to validate this initial finding. The specifically formulated alumina-hydroxyapatite powder mixture used in this experiment did allow the softened carious dentine to be removed more selectively, however, than conventional alumina particles alone, which are ineffective in removing softened dentine [Horiguchi et al., 1998; Banerjee, 1999]. The Carisolv gel excavation, even though slower than the other methods, did appear to remove the required amount of dentine, as indicated by the AF signal. There is, at present, little published scientific literature on the method of action and efficacy of this technique in removing carious dentine. However, a few studies assessing the performance of the Caridex system did produce some interesting findings which might be applicable, in principle, to this gel system. Barwart et al. [1991] performed an in vitro study which showed there to be no significant difference in carious dentine removal between the Caridex system and an equivalent

water-only control system. This finding implied that the mechanical action of the technqiue was more relevant to its function than the chemical component. Schutzbank et al. [1978] performed an in vitro study which concluded that the Caridex system was significantly better than a saline control system at removing caries in medium-hard lesions, but not in the cases of soft-medium lesions. However, limited details were given in this 20-year-old study as to how the samples were subjectively classified into these groups. It is arguable that, when compared to the current accepted concepts of distinguishing carious dentine, these groups would not nowadays be differentiated.

Conclusions

The primary objectives of this experiment were to assess the efficiency and the effectiveness of carious dentine removal using four different chemo-mechanical methods of excavation compared to hand excavation. The results showed: (a) in terms of efficiency (i.e. time taken to prepare the cavities): bur hand excavation a sono-abrasion a airabrasion Carisolv gel, bur excavation being shortest and most efficient, Carisolv gel longest; (b) in terms of effectiveness (i.e. extent of carious dentine removed relative to AF signature): bur hand excavation a Carisolv gel a airabrasion sono-abrasion, bur excavation tending to overprepare, Carisolv giving adequate preparation and being most effective, and sono-abrasion tending to underprepare. In terms of technique selectivity for carious dentine removal, it appeared from this study that the bur excavation method was the least selective method for removing carious dentine and would therefore rely upon the clinical skill of the operator to a significant degree. Hand excavation and Carisolv gel-assisted excavation to a degree would appear, from these results, to be the most selective methods which might allow more consistent cavities to be prepared.

Acknowledgements
This study has been kindly supported by the Medical Research Council, UK (Clinical Training Fellowship G84/4339) and MediTeam, Sweden, who supplied the Carisolv gel and instruments. Thanks also to Mr. P. Pilecki for his technical assistance.

Excavation of Carious Dentine

Caries Res 2000;34:144150

149

References
Banerjee A: Applications of Scanning Microscopy in the Assessment of Dentine Caries and Methods for Its Removal; PhD thesis University of London, 1999. Banerjee A, Boyde A: Comparison of autofluorescence and mineral content of carious dentine using scanning electron and optical microscopies. Caries Res 1997;31:284. Banerjee A, Boyde A, Watson TF, Kidd EAM: A new confocal fluorescence imaging technique for assessment of carious dentine and its clinical removal. Cell Vision 1997;4:112113. Banerjee A, Boyde A: Autofluorescence and mineral content of carious dentine: Scanning optical and backscattered electron microscopic studies. Caries Res 1998;32:219226. Banerjee A, Watson TF, Kidd EAM: Relation between the autofluorescence and excavation of carious dentine. J Dent Res 1998;77:632. Banerjee A, Sherriff M, Kidd EAM, Watson TF: A confocal microscopic study relating the autofluorescence of carious dentine to its microhardness. Br Dent J 1999;187:206210. Barwart O, Moschen I, Graber A, Pfaller K: In vitro study to compare the efficacy of N-monochloro-D,L-2-aminobutyrate (NMAB, GK-101E) and water in caries removal. J Oral Rehabil 1991;18:523529. Bjrndal L, Larsen T, Thylstrup A: A clinical and microbiological study of deep carious lesions during stepwise excavation using long treatment intervals. Caries Res 1997;31:411417. Elderton RJ: New approaches to cavity design with special reference to the class II lesion. Br Dent J 1984;157:421427. Ericson D, Zimmerman M, Raber H, Gtrick B, Bornstein R, Thorell J: Clinical evaluation of efficacy and safety of a new method for chemo-mechanical removal of caries. Caries Res 1999;33:171177. Frencken JE, Songpaisan Y, Phantumvanit P, Pilot T: An atraumatic restorative treatment (ART) technique: Evaluation after one year. Int Dent J 1994;44:460464. Gilmour ASM, Edmunds DH, Dummer PMH: The production of secondary caries-like lesions on cavity walls and the assessment of microleakage using an in vitro microbial caries system. J Oral Rehabil 1990;17:573578. Gilmour ASM, Edmunds DH, Newcombe RG: Prevalence and depth of artificial caries-like lesions adjacent to cavities prepared in roots and restored with a glass ionomer or a dentinbonded composite material. J Dent Res 1997; 76:18541861. Goldman M, Kronman JH: A preliminary report on a chemomechanical means of removing caries. J Am Dent Assoc 1976;93:11491153. Goldstein RE, Perkins FM: Air-abrasive technology: Its new role in restorative dentistry. J Am Dent Assoc 1994;125:551557. Horiguchi S, Yamada T, Inokoshi S, Tagami J: Selective caries removal with air abrasion. Op Dent 1998;23:236243. Jameson MW, Hood JAA, Tidmarsh BJ: The effects of dehydration and rehydration on some mechanical properties of human dentine. J Biomech 1993;26:10551065. Jameson MW, Tidmarsh BJ, Hood JAA: Effect of storage media on subsequent water loss and regain by human and bovine dentine and on mechanical properties of human dentine in vitro. Arch Oral Biol 1994;39:759767. Kidd EAM, Joyston-Bechal S, Beighton D: Microbiological validation of assessments of caries activity during cavity preparation. Caries Res 1993,27:402408. Kidd EAM, Ricketts DNJ, Beighton D: Criteria for caries removal at the enamel-dentine junction: A clinical and microbiological study. Br Dent J 1996;180:287291. Mertz-Fairhurst EJ, Curtis JW, Ergle JW, Rueggeberg FA, Adair SM: Ultraconservative and cariostatic sealed restorations: Results at year 10. J Am Dent Assoc 1998;129:5566. Renson CE: Back to the future in cavity preparation. Dent Update 1995;22:9395. Schutzbank SG, Galaini J, Kronman JH, Goldman M, Clark RE: A comparative in vitro study of GK-101 and GK-101E in caries removal. J Dent Res 1978;57:861864. van de Rijke JW: Use of dyes in cariology. Int Dent J 1991;41:111116. van der Veen MH, ten Bosch JJ: The influence of mineral loss on the auto-fluorescent behaviour of in vitro demineralised dentine. Caries Res 1996;30:9399. Weerheijm KL, Kreulen CM, de Soet JJ, Groen HJ, van Amerongen WE: Bacterial counts in carious dentine under restorations: 2-year in vivo effects. Caries Res 1999;33:130134. Zinck JH, McInnes-Ledoux P, Capdeboscq C, Weinberg R: Chemomechanical caries removal: A clinical evaluation. J Oral Rehabil 1988;15:2333.

150

Caries Res 2000;34:144150

Banerjee/Kidd/Watson