Original Paper Received: May 20, 2014 Accepted after revision: January 6, 2015 Published online: March 26, 2015
Caries Res 2015;49:259–265 DOI: 10.1159/000371897
In vitro Induction of Residual Caries Lesions in Dentin: Comparative Mineral Loss and Nano-Hardness Analysis Falk Schwendicke a Kerrin Eggers b Hendrik Meyer-Lueckel d Christof Dörfer b Alexander Kovalev c Stanislav Gorb c Sebastian Paris a
a
Department of Operative and Preventive Dentistry, Charité – Universitätsmedizin Berlin, Berlin, b Clinic for Conservative Dentistry and Periodontology and c Functional Morphology and Biomechanics, Zoological Institute, Christian Albrecht University, Kiel, and d Department of Operative Dentistry, Periodontology and Preventive Dentistry, RWTH Aachen University, Aachen, Germany
Abstract Artificially inducing dentinal lesions mimicking those remaining after selective excavation should allow to investigate the effects and limits of such selective excavation, for example regarding the mechanical properties of treated teeth or the remineralisation of sealed residual lesions. Such analyses might otherwise be limited by the variability of natural lesions or ethical and practical concerns. This study compared different demineralisation protocols for their suitability to induce lesions similar to natural residual caries. Twelve natural deep lesions were excavated until leathery dentin remained, and analysed for their mineral loss (ΔZ), lesion depth (LD), mineral loss ratio (R), the slope of the mineral gradient and their nano-hardness profile. Artificial lesions were induced using four different demineralisation protocols (acetic acid pH = 4.95; 0.1 M lactic acid gel pH = 5.0; 0.5 M ethylenediaminetetraacetic acid pH = 7.2; Streptococcus mutans biofilms) and their depths monitored over differ-
© 2015 S. Karger AG, Basel 0008–6568/15/0493–0259$39.50/0 E-Mail [email protected] www.karger.com/cre
ent demineralisation times. Lesions with depths most according to those of natural lesions were analysed using transversal microradiography. Lesions induced by acetic acid solution did not significantly differ with regards to LD, ΔZ, R and mineral profile. Seven dentin specimens were subsequently submitted to a moderately acidic (pH = 5.3) methylhydroxydiphosphonate-buffered acetate solution for 12 weeks. Natural and artificial residual lesions were similarly deep (mean ± SD: LD = 626 ± 212 and 563 ± 88 μm), demineralised (R = 19.5 ± 4.7 and 29.8 ± 4.1%), showed a flat and continuous mineral gradient (slope = 0.10 ± 0.05 and 0.13 ± 0.06 vol%/μm) and did not significantly differ regarding their nano-hardness profile. The described protocol induces lesions with mineral content and mechanical properties similar to natural residual lesions. © 2015 S. Karger AG, Basel
Whilst there are several methods to artificially induce caries lesions in enamel, there are only few studies investigating the induction of dentinal lesions [Marquezan et al., 2009; Moron et al., 2013]. Of these, none evaluated the suitability of different demineralisation protocols for creating residual lesions, i.e. lesions remaining after selective Dr. Falk Schwendicke Department of Operative and Preventive Dentistry Charité – Universitätsmedizin Berlin Aßmannshauser Str. 4–6, DE–14197 Berlin (Germany) E-Mail falk.schwendicke @ charite.de
Downloaded by: James H Quillen Coll of Med. 198.101.234.89 - 5/26/2015 4:55:26 PM
Key Words Artificial · Caries · Demineralisation · Excavation · Incomplete · Microradiography · Nano-hardness · Partial · Removal
Schwendicke et al., 2014a], we aimed at screening different protocols for lesion induction. 82 dentin specimens (5 × 4 × 4 mm) were prepared from 20 sound human teeth of the second dentition as previously described [Schwendicke et al., 2014a]. Samples were covered using acid-resistant nail varnish (Rossmann, Burgwedel, Germany), with a window of 2 × 2 mm left uncovered, and randomly submitted to one of four demineralisation protocols: (1) Acetic acid solution (n = 4 for 7, 14, 21, 28 and 35 days, total n = 20): Samples were stored in 1 litre of a demineralising solution [Buskes et al., 1985] containing 50 mM acetic acid, 3 mM CaCl2 × H2O, 3 mM KH2PO2 and 6 mM methylhydroxydiphosphonate (MHDP) at pH 4.95. The pH of the solution was monitored daily (InLab micro; Mettler-Toledo, Gießen, Germany) and adjusted if necessary using 10% HCl (Roth, Karlsruhe, Germany) or 10 M KOH. The solution was renewed weekly. (2) Carboxymethylcellulose-based lactic acid gel (n = 4 for 7, 14, 21, 28 and 35 days, total n = 20): The gel was prepared using 0.1 M lactate, 5 M KOH and 3% CMC, with pH = 5 being adjusted using 10% HCl or 10 M KOH [Kawasaki et al., 2000]. Samples were covered by 0.5 cm gel (equalling constant volumes of 1.5 ml per sample), which was renewed weekly. (3) Ethylenediaminetetraacetic acid (EDTA, n = 4 for 24, 48, 72, 96 and 120 h, total n = 20): 0.5 M EDTA was adjusted to pH 7.2 using 10% citric acid, with no renewal during the demineralisation period [Kawasaki et al., 2000]. If not given otherwise, chemicals were purchased from Sigma (St. Louis, Mo., USA). All specimens were placed into an orbital shaker (Incubator Shaker; New Brunswick Scientific, Enfield, Conn., USA) at 37 ° C. (4) In addition, we used a continuous-culture bacterial biofilm model (n = 6 for 7, 14 and 21 days, total n = 18) [Schwendicke et al., 2014a]. Briefly, Streptococcus mutans ATCC25175 (DSMZ, Braunschweig, Germany) was cultured in a selective medium [Takada and Hirasawa, 2005] containing 5 mg bacitracin, 5 mg valinomycin and 1.25 mg amphotericin per litre of brain heart infusion supplemented with 2% sucrose (BHI-S). Overnight cultures were inoculated onto sterilized dentin specimens and biofilms aerobically cultured at 37 ° C. BHI-S was supplied for 1 min per hour at a flow rate of 1.25 ml/min. We did not extend cultivation beyond 21 days to reduce the risk of contamination. One sample was lost during microradiographic analysis and could not be evaluated. Specimens were constantly stored under moist conditions. Lesion depth (LD) over time was monitored microscopically [Almqvist et al., 1990]. All lesions from one time point which best accorded with natural lesions (see below) regarding their depth were analysed using microradiography and compared with natural lesions for their LD (μm), integrated mineral loss (ΔZ, vol% × μm), average mineral loss ratio (R = ΔZ/LD, vol%) and the slope of the demineralisation front (vol%/μm). The demineralisation front was defined as the lesion part with mineral loss between 10 and 40 vol%. Since demineralisation using acetic acid was found to create lesions with the smallest differences of these parameters compared with natural residual lesions, we varied both the pH and the demineralisation time to induce lesions most similar to natural residual lesions. Demineralisation using acetic acid solution at pH 5.3 for 12 weeks was found suitable to create such lesions. Seven further specimens were submitted to this protocol and analysed in more detail regarding both their mineral content and their nanohardness profile.
Materials and Methods Study Design Within this explorative study, we performed a stepwise comparative analysis of different demineralisation protocols. First, we evaluated natural lesions remaining after selective excavation (‘residual caries lesions’), assessing both their mineral content and their nano-hardness profiles. Second, we aimed at identifying a suitable demineralisation protocol for inducing lesions resembling those remaining after selective excavation. The artificial lesions were supposed to result from demineralisation only, i.e. no excavation was attempted, since this was supposed to increase the variability, given the difficulties of standardising caries removal [Banerjee et al., 2000]. Lesions resulting from different protocols were monitored for their depth over different demineralisation times, and those according best with the depth of natural lesions were analysed further regarding their mineral content characteristics. Third, the most suitable demineralisation protocol was optimized and resulting lesions compared with natural residual lesions regarding their mineral loss and nano-hardness characteristics. Natural and Artificial Lesions To evaluate the characteristics of natural residual caries, 12 extracted, fully developed human teeth with deep caries, i.e. ICDAS score 6 [Ismail et al., 2007] were obtained under an ethics-approved protocol (D444/10). Caries was excavated as described by one experienced operator (F.S.), with only leathery, slightly moist carious dentin remaining in proximity to the pulp; the excavation was controlled by a second calibrated operator (K.E.). Based on existing studies in the field [Kawasaki et al., 2000; Magalhães et al., 2009; Marquezan et al., 2009; Moron et al., 2013;
260
Caries Res 2015;49:259–265 DOI: 10.1159/000371897
Schwendicke/Eggers/Meyer-Lueckel/ Dörfer/Kovalev/Gorb/Paris
Downloaded by: James H Quillen Coll of Med. 198.101.234.89 - 5/26/2015 4:55:26 PM
(incomplete) caries removal [Ricketts et al., 2013; Schwendicke et al., 2013]. Considering that leaving different amounts of carious dentin beneath restorations might differently affect the mechanical properties of teeth and restorations [Hevinga et al., 2010; Schwendicke et al., 2014b] or the remineralisation and arrest of the lesions [Weerheijm et al., 1992; Oong et al., 2008; Alves et al., 2010], there is need for such protocols. By artificially creating lesions, researchers can, for example, control both the depth and mineral loss of lesions. Thus, it is possible to predict the characteristics of the induced lesions, which is difficult when using natural lesions due to their variability, the difficulties of standardised excavation and the challenges of assessing these mineral or hardness properties a priori. The present study assessed different protocols of lesion induction and compared the resulting lesions with natural residual lesions regarding both mineral content and mechanical properties. We aimed to validate a protocol for inducing artificial residual lesions suitable for future studies of the mechanical effects of selective excavation, i.e. leaving caries, or the remineralisation of residual lesions with different depths.
Mineral loss (vol% × μm)
50 40 30 20 10 100
300 500 700 Depth (μm)
900
100
300 500 700 Depth (μm)
900
100
300 500 700 Depth (μm)
900
100
300 500 700 Depth (μm)
900
100
300 500 700 Depth (μm)
900
a
Mineral loss (vol% × μm)
50 40 30 20 10
b
Mineral loss (vol% × μm)
50 40 30 20 10
c
Mineral loss (vol% × μm)
50
Fig. 1. Natural (a) and artificial (b–e) caries
30 20 10
d 50 Mineral loss (vol% × μm)
40 30 20 10
e
Transversal Microradiography and Nano-Hardness Measurement LD, integrated mineral loss (ΔZ), average mineral loss ratio (R = ΔZ/LD) and mineral loss profile were evaluated using transversal microradiography as described [Paris and Meyer-Lueckel, 2010]. Mechanical properties were assessed by dynamic nano-indentation using continuous stiffness measurement mode (Nano
Indenter SA2; MTS Nano Instruments, Oak Ridge, Tenn., USA), which allows for testing of soft biological tissues [Enders et al., 2004]. Plan-parallelized, polished cross-sections of specimens were embedded in a metal chamber serving as sample holder, which was covered using Parafilm, with only the surface to measure being exposed to avoid desiccation during measurements. Indents were made in intertubular dentin and aligned perpendicu-
Artificial Residual Caries
Caries Res 2015;49:259–265 DOI: 10.1159/000371897
261
Downloaded by: James H Quillen Coll of Med. 198.101.234.89 - 5/26/2015 4:55:26 PM
lesions. We assessed artificial lesions induced by four different protocols: acetic acid solution for 35 days (b), lactic acid gel for 35 days (c), EDTA solution for 96 h (d) and bacterial biofilm for 21 days (e). Left panel: Exemplary microradiographs of lesions. The dotted white line indicates the lesion surface. Right panel: Mineral profiles of evaluated lesions. The mineral content (in %) was plotted along the depth of the dentin. The separate line plots of all assessed lesions (dotted light grey) and the mean plot (solid black) are shown. Note the variability especially of natural lesions with regards to depth, mineral loss and slope.
40
Table 1. Mineral content properties of natural and artificial caries lesions
Sample size Exposure time ΔZ, vol% × μm LD, μm R, vol% Slope, vol%/μm
Natural lesions
Acetic acid solution
Lactic acid gel
EDTA solution
Bacterial biofilm
12 – 12,437 ± 5,834a 626 ± 212a 19.5 ± 4.7a 0.10 ± 0.05a
4 35 days 9,072 ± 1,536a 357 ± 41a 26.9 ± 2.8a, b 0.25 ± 0.09a
4 35 days 17,147 ± 5,444a 572 ± 117a 29.4 ± 4.3b 0.17 ± 0.1a
4 96 hours 28,615 ± 3,129b 655 ± 68a 43.8 ± 0.4c 1.71 ± 0.67b
6 21 days 8,664 ± 2,014a 407 ± 165a 23.1 ± 6.5a, b 0.80 ± 0.52c
Artificial lesions were created using different demineralisation protocols for various time periods. All lesions from one time point with depths according best with those of natural lesions were evaluated using transversal microradiography. Means ± standard deviations are presented. Different superscript letters within rows indicate statistically significant difference between groups (t test/Bonferroni).
Eeff = (π1/2 × S) / (2β × Ac1/2) and H = Pmax / Ac [Oliver and Pharr, 1992], Eeff being the effective elasticity modulus, S the contact stiffness, β the correction factor for the indenter form, H the hardness and Pmax the maximal load. Evaluation Statistical analysis was performed using SPSS 22 (IBM, Armonk, N.Y., USA). Data were controlled for normal distribution using the Shapiro-Wilk test. Mineral loss characteristics between groups were compared using two-sided independent-samples t test with Bonferroni correction. The slope of the demineralisation front was calculated by linearly regressing the mineral content within the demineralisation front on the respective depth [Schwartz et al., 2012]. To statistically compare hardness profiles, analysis of variance (ANOVA) was used. Graphical comparison was performed using reference bands [Bowman and Young, 1996]. Briefly, these bands have widths of two pooled standard errors (se) and are centred at the average of the two means of each curve per time point. Significant differences between the plots are thus indicated by the curves exceeding this band according to y1j – y2j > 2sej (y1j – y2j) and sej (y1j – y2j) = 2sdj √ (1/n1j – 1/n2j), with y1j and y2j being the mean of each curve (i) at time point j, sdj being the pooled standard deviations, and nij being the number of observations per time point. Statistical significance was assumed if p < 0.05.
Results
Lesions resulting from different protocols were compared with natural residual lesions (fig. 1) regarding LD, ΔZ, R and the slope of the demineralisation front (ta262
Caries Res 2015;49:259–265 DOI: 10.1159/000371897
ble 1). Lesions induced using acetic acid solution for 35 days did not significantly differ from natural lesions (p > 0.05, t test/Bonferroni), whilst all other lesions differed in one or more parameters compared with natural residual lesions (p < 0.05). Seven specimens were submitted to further demineralisation using acetic acid solution at pH 5.3 for 12 weeks. Resulting lesions showed a mean ± SD ΔZ of 16,593 ± 2,270 vol% × μm, a LD of 563 ± 88 μm, a R of 29.8 ± 4.1 vol% and a slope of 0.13 ± 0.06 vol%/μm. The nanohardness profiles of both lesion types were not found to be significantly different using both statistical comparison with ANOVA (F = 0.168, p = 0.752, ω2 = 0.144) and graphical analysis (fig. 2).
Discussion
The present study assessed different demineralisation protocols for their suitability to induce caries lesions similar to natural residual caries, and aimed to validate one of these methods with regards to the mineral loss and mechanical properties of resulting lesions being similar to those of natural lesions. Artificial lesions induced using acetic acid demineralisation solutions accorded in depth, mineral loss, slope of their demineralisation front and nano-hardness profile with natural residual lesions. Using this protocol one could induce artificial residual lesions of, for example, different depths to investigate the effects of leaving different amounts of carious dentin on the mechanical properties of the tooth, or the remineralisation of the remaining lesions [Kinney et al., 1995; Kawasaki et al., 2000]. It should be highlighted that this protocol was not validated against other parameters – such Schwendicke/Eggers/Meyer-Lueckel/ Dörfer/Kovalev/Gorb/Paris
Downloaded by: James H Quillen Coll of Med. 198.101.234.89 - 5/26/2015 4:55:26 PM
larly to the lesion floor in intervals of 35 μm. The samples were loaded under constant strain rate conditions with a rate of 0.02/s to a maximum penetration depth of 10 μm. The elastic properties were assessed using force displacement curves following
Natural lesions, n = 12 Reference band
Hardness (MPa)
800 700 600 500 400 300 200 100 100
200
300
400 500 600 Depth (μm)
700
800
Fig. 2. Cross-sectional nano-hardness (MPa) of natural (mean:
continuous line; SD: mid-grey shadow) and artificial lesions (mean: dotted line; SD: dark-grey shadow). Artificial lesions were induced by 12 weeks of demineralisation using acetic acid solution (pH 5.3). The constructed reference band (light grey) has a width of two pooled standard errors and is centred on the pooled means of the plots. None of the plots exceeds the band, indicating no significant difference of nano-hardness between different lesions. Arrows: LD as indicated by microradiography for natural (continuous line) and artificial lesions (dotted line). Note that hardness is still moderately increasing within dentin identified as ‘sound’ via microradiography.
as bacterial load or surface properties – which are certainly of interest as well: here other protocols, for example using bacterial biofilms, might be better suited. Similarly, excavating until only ‘leathery’ dentin remains is not the only possible criterion when aiming at selective excavation, and other criteria might yield lesions of different depths, mineral or mechanical characteristics. If aiming at mimicking these different residual lesions, the protocols would require validation once more. Last, certain questions – especially those related to pulpal reactions – with regards to selective excavation may not be investigable at all in vitro. Here, clinical studies will be required. By using different demineralisation protocols, one can control the properties of the resulting lesions, e.g. their LD via the demineralisation time, the pH or the temperature, their average mineral loss ratio and the gradient of mineral loss via the pH, the degree of mineral saturation or the use of dissolution inhibitors like MHDP, and their presence of a surface layer via fluoride addition to the medium or creation of a diffusion barrier on the lesion surface [Theuns et al., 1984; Buskes et al., 1985; Damen et al., 1998; Magalhães et al., 2009]. Acetic acid solution used at Artificial Residual Caries
moderately low pH and containing a dissolution inhibitor (MHDP) was found to induce lesions with moderate depths and mineral loss and nano-hardness and mineral profiles similar to natural lesions. The absence of a surface layer in natural residual lesions – as can be expected after excavation – was also found in artificial lesions. Both natural and artificial lesions showed a gradual, albeit slightly differently steep slope of the mineral front. For natural reasons, this indicates a slow, presumably diffusion-limited demineralisation process [Gao et al., 1993], whilst the found variability of this slope indicates that lesions might have been differently active, i.e. demineralisation occurred at different speeds. The gradual increase of natural lesions explains the difficulty of standardising the excavation when using tactile criteria, since a clear threshold cannot easily be identified. Mimicking lesions with such slope via demineralisation allows to omit the excavation step and also removes lesion activity as a potential confounder, and thereby standardises the resulting lesions to a certain degree. It should be highlighted that the induced lesions showed a higher degree of mineral loss, i.e. a higher mineral loss ratio, than natural lesions. Further modifications of the protocol might help to improve the similarity between lesions even further, but demineralisation will be slower, and practical aspects might eventually limit how far one can approximate the parameters of artificial and natural lesions. Other protocols, for example using bacterial biofilms, induced lesions of similar depth, mineral loss and mineral loss ratio, but significantly steeper mineral gradients and well-established surface layers. The latter might be the product of the biofilm acting as a diffusion barrier, and it might be speculated that using lactobacilli or bacterial consortia as well as altering the cariogenicity of the cultivation milieu could alter lesion characteristics to better resemble natural residual caries [Clarkson et al., 1984; Shu et al., 2000; Schwendicke et al., 2014a]. This system would have the additional benefit of creating infected lesions, with inherent activity of both host and bacterial proteinases and the option of microbiological analysis, for example regarding arrest of differently infected lesions. Lesions induced by lactic acid gels might also be suitable for certain applications, whilst their mineral loss and mineral loss ratio were higher than those of natural lesions. This is in contrast to previous studies [Moron et al., 2013], but might be due to longer demineralisation and the specimens being agitated, which could have increased diffusion of minerals out of the lesion and the gel, impeding mineral saturation of the gel. This could also explain Caries Res 2015;49:259–265 DOI: 10.1159/000371897
263
Downloaded by: James H Quillen Coll of Med. 198.101.234.89 - 5/26/2015 4:55:26 PM
Acetic acid, n = 7
264
Caries Res 2015;49:259–265 DOI: 10.1159/000371897
have been sealed for certain time periods, i.e. have been remineralised, inactivated and re-hardened, might be possible as well after amending the protocol. In conclusion, demineralisation with MHDP-buffered, moderately acidic acetate solution induced lesions with similar mineral loss and nano-hardness properties as selectively excavated lesions. Both lesion types were relatively deep, moderately demineralised and showed constant and gradually increasing mineral content and nano-hardness along the demineralisation front. Modifying the presented protocol might broaden its further applicability in vitro, whilst certain investigations will require natural lesions or clinical settings.
Acknowledgements We thank Regina Marquardt, Conservative Dentistry and Periodontology, Christian Albrecht University Kiel, for her assistance with microradiographic analyses. The study was funded by the authors and their institutions.
Author Contributions Development of the protocol: F. Schwendicke, S. Paris, S. Gorb. Laboratory investigations: F. Schwendicke, A. Kovalev, K. Eggers. Analysis and interpretation of data: F. Schwendicke. Manuscript preparation: F. Schwendicke, S. Paris, H. Meyer-Lueckel, C. Dörfer, S. Gorb.
Disclosure Statement The authors declare no conflict of interest. The study was funded by the authors and their institutions.
References
Almqvist H, Wefel JS, Lagerlöf F: Root hard-tissue demineralization rate measured by 125I absorptiometry: comparison with lesiondepth measurements. J Dent Res 1990; 69: 1519–1521. Alves LS, Fontanella V, Damo AC, Ferreira de Oliveira E, Maltz M: Qualitative and quantitative radiographic assessment of sealed carious dentin: a 10-year prospective study. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2010;109:135–141. Balooch M, Wu-Magidi IC, Balazs A, Lundkvist AS, Marshall SJ, Marshall GW, Siekhaus WJ, Kinney JH: Viscoelastic properties of demineralized human dentin measured in water with atomic force microscope (AFM)-based indentation. J Biomed Mater Res 1998; 40: 539–544.
Schwendicke/Eggers/Meyer-Lueckel/ Dörfer/Kovalev/Gorb/Paris
Downloaded by: James H Quillen Coll of Med. 198.101.234.89 - 5/26/2015 4:55:26 PM
the absence of a surface layer, which was found in previous studies when using lactic acid gel demineralisation [Moron et al., 2013]. Similarly, the protocol using highly concentrated chelating agents like EDTA induced very deep lesions with extremely steep mineral gradients, no surface layer and nearly complete mineral eradication within the lesion body. This protocol is easy to perform and requires extremely short times for demineralisation, but seems unsuitable for the induction of artificial lesions mimicking those remaining after selective excavation. Our study is the first one to comparatively report the nano-hardness in both natural and artificial dentinal lesions. Whilst we corroborate findings from a previous study regarding the magnitude and cross-sectional profile of nano-hardness along carious dentin [Joves et al., 2014], it was relevant to compare this nano-hardness with that of artificial lesions, since mineral content in caries lesions was found to not readily translate into mechanical properties [Balooch et al., 1998; Kinney et al., 2003]. This was confirmed in our study, with nano-hardness increasing in areas identified as sound via microradiography. As discussed for the mineral content, the nano-hardness of natural lesions will further depend on both the original lesion properties (active or inactive lesions, older or younger teeth) and the used excavation criterion, resulting in the variability we found with regards to the mechanical properties of natural lesions [Zheng et al., 2003; Senawongse et al., 2006]. Given the hierarchical structure of dentin and the influence of specimen hydration on mechanical properties [Kinney et al., 1995; Balooch et al., 1998], nano-hardness measurements might generally yield more variable results than micro-hardness, but have the advantage of a higher resolution [Magalhães et al., 2009]. Given the focus of this study, we did not address all aspects related to the induction of artificial residual dentin lesions. We did, for example, not investigate whether the adhesive properties of artificial lesions are similar to those of natural, caries-affected or -infected dentinal lesions. The adhesive properties might be further modified by the presence (natural lesions) or absence (artificial lesions) of a smear layer, and studies in that direction might need to modify the presented protocol. Bacterial infection and enzymatic degradation of the dentin were not mimicked by the used acid-based demineralisation, but might be relevant with regards to lesion arrest or remineralisation of infected dentin. Again, modifications of our protocol (e.g. pre-demineralisation via acetic acid solutions followed by biofilm-based lesion progression and bacterial invasion) might be viable. Mimicking lesions which
Artificial Residual Caries
Kawasaki K, Ruben J, Tsuda H, Huysmans MC, Takagi O: Relationship between mineral distributions in dentine lesions and subsequent remineralization in vitro. Caries Res 2000;34: 395–403. Kinney JH, Balooch M, Haupt DL, Marshall SJ, Marshall GW: Mineral distribution and dimensional changes in human dentin during demineralization. J Dent Res 1995; 74: 1179– 1184. Kinney JH, Habelitz S, Marshall SJ, Marshall GW: The importance of intrafibrillar mineralization of collagen on the mechanical properties of dentin. J Dent Res 2003;82:957–961. Magalhães AC, Moron BM, Comar LP, Wiegand A, Buchalla W, Buzalaf MAR: Comparison of cross-sectional hardness and transverse microradiography of artificial carious enamel lesions induced by different demineralising solutions and gels. Caries Res 2009;43:474–483. Marquezan M, Correa FN, Sanabe ME, Rodrigues Filho LE, Hebling J, Guedes-Pinto AC, Mendes FM: Artificial methods of dentine caries induction: a hardness and morphological comparative study. Arch Oral Biol 2009; 54:1111–1117. Moron BM, Comar LP, Wiegand A, Buchalla W, Yu H, Buzalaf MAR, Magalhães AC: Different protocols to produce artificial dentine carious lesions in vitro and in situ: hardness and mineral content correlation. Caries Res 2013; 47: 162–170. Oliver WC, Pharr GM: An improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments. J Mater Res 1992; 7: 1564– 1583. Oong EM, Griffin SO, Kohn WG, Gooch BF, Caufield PW: The effect of dental sealants on bacteria levels in caries lesions. J Am Dent Assoc 2008;139:271–278. Paris S, Meyer-Lueckel H: Infiltrants inhibit progression of natural caries lesions in vitro. J Dent Res 2010;89:1276–1280.
Ricketts D, Lamont T, Innes NP, Kidd E, Clarkson JE: Operative caries management in adults and children. Cochrane Database Syst Rev 2013;3:CD003808. Schwartz AG, Pasteris JD, Genin GM, Daulton TL, Thomopoulos S: Mineral distributions at the developing tendon enthesis. PLoS One 2012;7:e48630. Schwendicke F, Dörfer C, Kneist S, Meyer-Lueckel H, Paris S: Cariogenic effects of probiotic Lactobacillus rhamnosus GG in a dental biofilm model. Caries Res 2014a;48:186–192. Schwendicke F, Dörfer CE, Paris S: Incomplete caries removal: a systematic review and metaanalysis. J Dent Res 2013;92:306–314. Schwendicke F, Kern M, Meyer-Lueckel H, Boels A, Doerfer C, Paris S: Fracture resistance and cuspal deflection of incompletely excavated teeth. J Dent 2014b;42:107–113. Senawongse P, Otsuki M, Tagami J, Mjör I: Agerelated changes in hardness and modulus of elasticity of dentine. Arch Oral Biol 2006; 51: 457–463. Shu M, Wong L, Miller JH, Sissons CH: Development of multi-species consortia biofilms of oral bacteria as an enamel and root caries model system. Arch Oral Biol 2000;45:27–40. Takada K, Hirasawa M: A novel selective medium for isolation of Streptococcus mutans. J Microbiol Methods 2005;60:189–193. Theuns HM, van Dijk JW, Driessens FC, Groeneveld A: The effect of undissociated aceticacid concentration of buffer solutions on artificial caries-like lesion formation in human tooth enamel. Arch Oral Biol 1984; 29: 759– 763. Weerheijm KL, de Soet JJ, van Amerongen WE, de Graaff J: Sealing of occlusal hidden caries lesions: an alternative for curative treatment? J Dent Child 1992;59:263–268. Zheng L, Hilton JF, Habelitz S, Marshall SJ, Marshall GW: Dentin caries activity status related to hardness and elasticity. Eur J Oral Sci 2003; 111:243–252.
Caries Res 2015;49:259–265 DOI: 10.1159/000371897
265
Downloaded by: James H Quillen Coll of Med. 198.101.234.89 - 5/26/2015 4:55:26 PM
Banerjee A, Kidd EAM, Watson TF: In vitro evaluation of five alternative methods of carious dentine excavation. Caries Res 2000; 34: 144– 150. Bowman A, Young S: Graphical comparison of nonparametric curves. J R Stat Soc Ser C Appl Stat 1996;45:83–98. Buskes JA, Christoffersen J, Arends J: Lesion formation and lesion remineralization in enamel under constant composition conditions. A new technique with applications. Caries Res 1985;19:490–496. Clarkson BH, Wefel JS, Miller I: A model for producing caries-like lesions in enamel and dentin using oral bacteria in vitro. J Dent Res 1984;63:1186–1189. Damen JJ, Buijs MJ, ten Cate JM: Fluoride-dependent formation of mineralized layers in bovine dentin during demineralization in vitro. Caries Res 1998;32:435–440. Enders S, Barbakadse N, Gorb SN, Arzt E: Exploring biological surfaces by nanoindentation. J Mater Res 2004;19:880–887. Gao XJ, Elliott JC, Anderson P: Scanning microradiographic study of the kinetics of subsurface demineralization in tooth sections under constant-composition and small constantvolume conditions. J Dent Res 1993; 72: 923– 930. Hevinga MA, Opdam NJ, Frencken JE, Truin GJ, Huysmans MC: Does incomplete caries removal reduce strength of restored teeth? J Dent Res 2010;89:1270–1275. Ismail A, Sohn W, Tellez M, Amaya A, Sen A, Hasson H, Pitts N: The International Caries Detection and Assessment System (ICDAS): an integrated system for measuring dental caries. Community Dent Oral Epidemiol 2007;35:170–178. Joves GJ, Inoue G, Sadr A, Nikaido T, Tagami J: Nanoindentation hardness of intertubular dentin in sound, demineralized and natural caries-affected dentin. J Mech Behav Biomed Mater 2014;32:39–45.
Caries Res 2015;49:259–265 DOI: 10.1159/000371897
In vitro Induction of Residual Caries Lesions in Dentin: Comparative Mineral Loss and Nano-Hardness Analysis Falk Schwendicke a Kerrin Eggers b Hendrik Meyer-Lueckel d Christof Dörfer b Alexander Kovalev c Stanislav Gorb c Sebastian Paris a
a
Department of Operative and Preventive Dentistry, Charité – Universitätsmedizin Berlin, Berlin, b Clinic for Conservative Dentistry and Periodontology and c Functional Morphology and Biomechanics, Zoological Institute, Christian Albrecht University, Kiel, and d Department of Operative Dentistry, Periodontology and Preventive Dentistry, RWTH Aachen University, Aachen, Germany
Abstract Artificially inducing dentinal lesions mimicking those remaining after selective excavation should allow to investigate the effects and limits of such selective excavation, for example regarding the mechanical properties of treated teeth or the remineralisation of sealed residual lesions. Such analyses might otherwise be limited by the variability of natural lesions or ethical and practical concerns. This study compared different demineralisation protocols for their suitability to induce lesions similar to natural residual caries. Twelve natural deep lesions were excavated until leathery dentin remained, and analysed for their mineral loss (ΔZ), lesion depth (LD), mineral loss ratio (R), the slope of the mineral gradient and their nano-hardness profile. Artificial lesions were induced using four different demineralisation protocols (acetic acid pH = 4.95; 0.1 M lactic acid gel pH = 5.0; 0.5 M ethylenediaminetetraacetic acid pH = 7.2; Streptococcus mutans biofilms) and their depths monitored over differ-
© 2015 S. Karger AG, Basel 0008–6568/15/0493–0259$39.50/0 E-Mail [email protected] www.karger.com/cre
ent demineralisation times. Lesions with depths most according to those of natural lesions were analysed using transversal microradiography. Lesions induced by acetic acid solution did not significantly differ with regards to LD, ΔZ, R and mineral profile. Seven dentin specimens were subsequently submitted to a moderately acidic (pH = 5.3) methylhydroxydiphosphonate-buffered acetate solution for 12 weeks. Natural and artificial residual lesions were similarly deep (mean ± SD: LD = 626 ± 212 and 563 ± 88 μm), demineralised (R = 19.5 ± 4.7 and 29.8 ± 4.1%), showed a flat and continuous mineral gradient (slope = 0.10 ± 0.05 and 0.13 ± 0.06 vol%/μm) and did not significantly differ regarding their nano-hardness profile. The described protocol induces lesions with mineral content and mechanical properties similar to natural residual lesions. © 2015 S. Karger AG, Basel
Whilst there are several methods to artificially induce caries lesions in enamel, there are only few studies investigating the induction of dentinal lesions [Marquezan et al., 2009; Moron et al., 2013]. Of these, none evaluated the suitability of different demineralisation protocols for creating residual lesions, i.e. lesions remaining after selective Dr. Falk Schwendicke Department of Operative and Preventive Dentistry Charité – Universitätsmedizin Berlin Aßmannshauser Str. 4–6, DE–14197 Berlin (Germany) E-Mail falk.schwendicke @ charite.de
Downloaded by: James H Quillen Coll of Med. 198.101.234.89 - 5/26/2015 4:55:26 PM
Key Words Artificial · Caries · Demineralisation · Excavation · Incomplete · Microradiography · Nano-hardness · Partial · Removal
Schwendicke et al., 2014a], we aimed at screening different protocols for lesion induction. 82 dentin specimens (5 × 4 × 4 mm) were prepared from 20 sound human teeth of the second dentition as previously described [Schwendicke et al., 2014a]. Samples were covered using acid-resistant nail varnish (Rossmann, Burgwedel, Germany), with a window of 2 × 2 mm left uncovered, and randomly submitted to one of four demineralisation protocols: (1) Acetic acid solution (n = 4 for 7, 14, 21, 28 and 35 days, total n = 20): Samples were stored in 1 litre of a demineralising solution [Buskes et al., 1985] containing 50 mM acetic acid, 3 mM CaCl2 × H2O, 3 mM KH2PO2 and 6 mM methylhydroxydiphosphonate (MHDP) at pH 4.95. The pH of the solution was monitored daily (InLab micro; Mettler-Toledo, Gießen, Germany) and adjusted if necessary using 10% HCl (Roth, Karlsruhe, Germany) or 10 M KOH. The solution was renewed weekly. (2) Carboxymethylcellulose-based lactic acid gel (n = 4 for 7, 14, 21, 28 and 35 days, total n = 20): The gel was prepared using 0.1 M lactate, 5 M KOH and 3% CMC, with pH = 5 being adjusted using 10% HCl or 10 M KOH [Kawasaki et al., 2000]. Samples were covered by 0.5 cm gel (equalling constant volumes of 1.5 ml per sample), which was renewed weekly. (3) Ethylenediaminetetraacetic acid (EDTA, n = 4 for 24, 48, 72, 96 and 120 h, total n = 20): 0.5 M EDTA was adjusted to pH 7.2 using 10% citric acid, with no renewal during the demineralisation period [Kawasaki et al., 2000]. If not given otherwise, chemicals were purchased from Sigma (St. Louis, Mo., USA). All specimens were placed into an orbital shaker (Incubator Shaker; New Brunswick Scientific, Enfield, Conn., USA) at 37 ° C. (4) In addition, we used a continuous-culture bacterial biofilm model (n = 6 for 7, 14 and 21 days, total n = 18) [Schwendicke et al., 2014a]. Briefly, Streptococcus mutans ATCC25175 (DSMZ, Braunschweig, Germany) was cultured in a selective medium [Takada and Hirasawa, 2005] containing 5 mg bacitracin, 5 mg valinomycin and 1.25 mg amphotericin per litre of brain heart infusion supplemented with 2% sucrose (BHI-S). Overnight cultures were inoculated onto sterilized dentin specimens and biofilms aerobically cultured at 37 ° C. BHI-S was supplied for 1 min per hour at a flow rate of 1.25 ml/min. We did not extend cultivation beyond 21 days to reduce the risk of contamination. One sample was lost during microradiographic analysis and could not be evaluated. Specimens were constantly stored under moist conditions. Lesion depth (LD) over time was monitored microscopically [Almqvist et al., 1990]. All lesions from one time point which best accorded with natural lesions (see below) regarding their depth were analysed using microradiography and compared with natural lesions for their LD (μm), integrated mineral loss (ΔZ, vol% × μm), average mineral loss ratio (R = ΔZ/LD, vol%) and the slope of the demineralisation front (vol%/μm). The demineralisation front was defined as the lesion part with mineral loss between 10 and 40 vol%. Since demineralisation using acetic acid was found to create lesions with the smallest differences of these parameters compared with natural residual lesions, we varied both the pH and the demineralisation time to induce lesions most similar to natural residual lesions. Demineralisation using acetic acid solution at pH 5.3 for 12 weeks was found suitable to create such lesions. Seven further specimens were submitted to this protocol and analysed in more detail regarding both their mineral content and their nanohardness profile.
Materials and Methods Study Design Within this explorative study, we performed a stepwise comparative analysis of different demineralisation protocols. First, we evaluated natural lesions remaining after selective excavation (‘residual caries lesions’), assessing both their mineral content and their nano-hardness profiles. Second, we aimed at identifying a suitable demineralisation protocol for inducing lesions resembling those remaining after selective excavation. The artificial lesions were supposed to result from demineralisation only, i.e. no excavation was attempted, since this was supposed to increase the variability, given the difficulties of standardising caries removal [Banerjee et al., 2000]. Lesions resulting from different protocols were monitored for their depth over different demineralisation times, and those according best with the depth of natural lesions were analysed further regarding their mineral content characteristics. Third, the most suitable demineralisation protocol was optimized and resulting lesions compared with natural residual lesions regarding their mineral loss and nano-hardness characteristics. Natural and Artificial Lesions To evaluate the characteristics of natural residual caries, 12 extracted, fully developed human teeth with deep caries, i.e. ICDAS score 6 [Ismail et al., 2007] were obtained under an ethics-approved protocol (D444/10). Caries was excavated as described by one experienced operator (F.S.), with only leathery, slightly moist carious dentin remaining in proximity to the pulp; the excavation was controlled by a second calibrated operator (K.E.). Based on existing studies in the field [Kawasaki et al., 2000; Magalhães et al., 2009; Marquezan et al., 2009; Moron et al., 2013;
260
Caries Res 2015;49:259–265 DOI: 10.1159/000371897
Schwendicke/Eggers/Meyer-Lueckel/ Dörfer/Kovalev/Gorb/Paris
Downloaded by: James H Quillen Coll of Med. 198.101.234.89 - 5/26/2015 4:55:26 PM
(incomplete) caries removal [Ricketts et al., 2013; Schwendicke et al., 2013]. Considering that leaving different amounts of carious dentin beneath restorations might differently affect the mechanical properties of teeth and restorations [Hevinga et al., 2010; Schwendicke et al., 2014b] or the remineralisation and arrest of the lesions [Weerheijm et al., 1992; Oong et al., 2008; Alves et al., 2010], there is need for such protocols. By artificially creating lesions, researchers can, for example, control both the depth and mineral loss of lesions. Thus, it is possible to predict the characteristics of the induced lesions, which is difficult when using natural lesions due to their variability, the difficulties of standardised excavation and the challenges of assessing these mineral or hardness properties a priori. The present study assessed different protocols of lesion induction and compared the resulting lesions with natural residual lesions regarding both mineral content and mechanical properties. We aimed to validate a protocol for inducing artificial residual lesions suitable for future studies of the mechanical effects of selective excavation, i.e. leaving caries, or the remineralisation of residual lesions with different depths.
Mineral loss (vol% × μm)
50 40 30 20 10 100
300 500 700 Depth (μm)
900
100
300 500 700 Depth (μm)
900
100
300 500 700 Depth (μm)
900
100
300 500 700 Depth (μm)
900
100
300 500 700 Depth (μm)
900
a
Mineral loss (vol% × μm)
50 40 30 20 10
b
Mineral loss (vol% × μm)
50 40 30 20 10
c
Mineral loss (vol% × μm)
50
Fig. 1. Natural (a) and artificial (b–e) caries
30 20 10
d 50 Mineral loss (vol% × μm)
40 30 20 10
e
Transversal Microradiography and Nano-Hardness Measurement LD, integrated mineral loss (ΔZ), average mineral loss ratio (R = ΔZ/LD) and mineral loss profile were evaluated using transversal microradiography as described [Paris and Meyer-Lueckel, 2010]. Mechanical properties were assessed by dynamic nano-indentation using continuous stiffness measurement mode (Nano
Indenter SA2; MTS Nano Instruments, Oak Ridge, Tenn., USA), which allows for testing of soft biological tissues [Enders et al., 2004]. Plan-parallelized, polished cross-sections of specimens were embedded in a metal chamber serving as sample holder, which was covered using Parafilm, with only the surface to measure being exposed to avoid desiccation during measurements. Indents were made in intertubular dentin and aligned perpendicu-
Artificial Residual Caries
Caries Res 2015;49:259–265 DOI: 10.1159/000371897
261
Downloaded by: James H Quillen Coll of Med. 198.101.234.89 - 5/26/2015 4:55:26 PM
lesions. We assessed artificial lesions induced by four different protocols: acetic acid solution for 35 days (b), lactic acid gel for 35 days (c), EDTA solution for 96 h (d) and bacterial biofilm for 21 days (e). Left panel: Exemplary microradiographs of lesions. The dotted white line indicates the lesion surface. Right panel: Mineral profiles of evaluated lesions. The mineral content (in %) was plotted along the depth of the dentin. The separate line plots of all assessed lesions (dotted light grey) and the mean plot (solid black) are shown. Note the variability especially of natural lesions with regards to depth, mineral loss and slope.
40
Table 1. Mineral content properties of natural and artificial caries lesions
Sample size Exposure time ΔZ, vol% × μm LD, μm R, vol% Slope, vol%/μm
Natural lesions
Acetic acid solution
Lactic acid gel
EDTA solution
Bacterial biofilm
12 – 12,437 ± 5,834a 626 ± 212a 19.5 ± 4.7a 0.10 ± 0.05a
4 35 days 9,072 ± 1,536a 357 ± 41a 26.9 ± 2.8a, b 0.25 ± 0.09a
4 35 days 17,147 ± 5,444a 572 ± 117a 29.4 ± 4.3b 0.17 ± 0.1a
4 96 hours 28,615 ± 3,129b 655 ± 68a 43.8 ± 0.4c 1.71 ± 0.67b
6 21 days 8,664 ± 2,014a 407 ± 165a 23.1 ± 6.5a, b 0.80 ± 0.52c
Artificial lesions were created using different demineralisation protocols for various time periods. All lesions from one time point with depths according best with those of natural lesions were evaluated using transversal microradiography. Means ± standard deviations are presented. Different superscript letters within rows indicate statistically significant difference between groups (t test/Bonferroni).
Eeff = (π1/2 × S) / (2β × Ac1/2) and H = Pmax / Ac [Oliver and Pharr, 1992], Eeff being the effective elasticity modulus, S the contact stiffness, β the correction factor for the indenter form, H the hardness and Pmax the maximal load. Evaluation Statistical analysis was performed using SPSS 22 (IBM, Armonk, N.Y., USA). Data were controlled for normal distribution using the Shapiro-Wilk test. Mineral loss characteristics between groups were compared using two-sided independent-samples t test with Bonferroni correction. The slope of the demineralisation front was calculated by linearly regressing the mineral content within the demineralisation front on the respective depth [Schwartz et al., 2012]. To statistically compare hardness profiles, analysis of variance (ANOVA) was used. Graphical comparison was performed using reference bands [Bowman and Young, 1996]. Briefly, these bands have widths of two pooled standard errors (se) and are centred at the average of the two means of each curve per time point. Significant differences between the plots are thus indicated by the curves exceeding this band according to y1j – y2j > 2sej (y1j – y2j) and sej (y1j – y2j) = 2sdj √ (1/n1j – 1/n2j), with y1j and y2j being the mean of each curve (i) at time point j, sdj being the pooled standard deviations, and nij being the number of observations per time point. Statistical significance was assumed if p < 0.05.
Results
Lesions resulting from different protocols were compared with natural residual lesions (fig. 1) regarding LD, ΔZ, R and the slope of the demineralisation front (ta262
Caries Res 2015;49:259–265 DOI: 10.1159/000371897
ble 1). Lesions induced using acetic acid solution for 35 days did not significantly differ from natural lesions (p > 0.05, t test/Bonferroni), whilst all other lesions differed in one or more parameters compared with natural residual lesions (p < 0.05). Seven specimens were submitted to further demineralisation using acetic acid solution at pH 5.3 for 12 weeks. Resulting lesions showed a mean ± SD ΔZ of 16,593 ± 2,270 vol% × μm, a LD of 563 ± 88 μm, a R of 29.8 ± 4.1 vol% and a slope of 0.13 ± 0.06 vol%/μm. The nanohardness profiles of both lesion types were not found to be significantly different using both statistical comparison with ANOVA (F = 0.168, p = 0.752, ω2 = 0.144) and graphical analysis (fig. 2).
Discussion
The present study assessed different demineralisation protocols for their suitability to induce caries lesions similar to natural residual caries, and aimed to validate one of these methods with regards to the mineral loss and mechanical properties of resulting lesions being similar to those of natural lesions. Artificial lesions induced using acetic acid demineralisation solutions accorded in depth, mineral loss, slope of their demineralisation front and nano-hardness profile with natural residual lesions. Using this protocol one could induce artificial residual lesions of, for example, different depths to investigate the effects of leaving different amounts of carious dentin on the mechanical properties of the tooth, or the remineralisation of the remaining lesions [Kinney et al., 1995; Kawasaki et al., 2000]. It should be highlighted that this protocol was not validated against other parameters – such Schwendicke/Eggers/Meyer-Lueckel/ Dörfer/Kovalev/Gorb/Paris
Downloaded by: James H Quillen Coll of Med. 198.101.234.89 - 5/26/2015 4:55:26 PM
larly to the lesion floor in intervals of 35 μm. The samples were loaded under constant strain rate conditions with a rate of 0.02/s to a maximum penetration depth of 10 μm. The elastic properties were assessed using force displacement curves following
Natural lesions, n = 12 Reference band
Hardness (MPa)
800 700 600 500 400 300 200 100 100
200
300
400 500 600 Depth (μm)
700
800
Fig. 2. Cross-sectional nano-hardness (MPa) of natural (mean:
continuous line; SD: mid-grey shadow) and artificial lesions (mean: dotted line; SD: dark-grey shadow). Artificial lesions were induced by 12 weeks of demineralisation using acetic acid solution (pH 5.3). The constructed reference band (light grey) has a width of two pooled standard errors and is centred on the pooled means of the plots. None of the plots exceeds the band, indicating no significant difference of nano-hardness between different lesions. Arrows: LD as indicated by microradiography for natural (continuous line) and artificial lesions (dotted line). Note that hardness is still moderately increasing within dentin identified as ‘sound’ via microradiography.
as bacterial load or surface properties – which are certainly of interest as well: here other protocols, for example using bacterial biofilms, might be better suited. Similarly, excavating until only ‘leathery’ dentin remains is not the only possible criterion when aiming at selective excavation, and other criteria might yield lesions of different depths, mineral or mechanical characteristics. If aiming at mimicking these different residual lesions, the protocols would require validation once more. Last, certain questions – especially those related to pulpal reactions – with regards to selective excavation may not be investigable at all in vitro. Here, clinical studies will be required. By using different demineralisation protocols, one can control the properties of the resulting lesions, e.g. their LD via the demineralisation time, the pH or the temperature, their average mineral loss ratio and the gradient of mineral loss via the pH, the degree of mineral saturation or the use of dissolution inhibitors like MHDP, and their presence of a surface layer via fluoride addition to the medium or creation of a diffusion barrier on the lesion surface [Theuns et al., 1984; Buskes et al., 1985; Damen et al., 1998; Magalhães et al., 2009]. Acetic acid solution used at Artificial Residual Caries
moderately low pH and containing a dissolution inhibitor (MHDP) was found to induce lesions with moderate depths and mineral loss and nano-hardness and mineral profiles similar to natural lesions. The absence of a surface layer in natural residual lesions – as can be expected after excavation – was also found in artificial lesions. Both natural and artificial lesions showed a gradual, albeit slightly differently steep slope of the mineral front. For natural reasons, this indicates a slow, presumably diffusion-limited demineralisation process [Gao et al., 1993], whilst the found variability of this slope indicates that lesions might have been differently active, i.e. demineralisation occurred at different speeds. The gradual increase of natural lesions explains the difficulty of standardising the excavation when using tactile criteria, since a clear threshold cannot easily be identified. Mimicking lesions with such slope via demineralisation allows to omit the excavation step and also removes lesion activity as a potential confounder, and thereby standardises the resulting lesions to a certain degree. It should be highlighted that the induced lesions showed a higher degree of mineral loss, i.e. a higher mineral loss ratio, than natural lesions. Further modifications of the protocol might help to improve the similarity between lesions even further, but demineralisation will be slower, and practical aspects might eventually limit how far one can approximate the parameters of artificial and natural lesions. Other protocols, for example using bacterial biofilms, induced lesions of similar depth, mineral loss and mineral loss ratio, but significantly steeper mineral gradients and well-established surface layers. The latter might be the product of the biofilm acting as a diffusion barrier, and it might be speculated that using lactobacilli or bacterial consortia as well as altering the cariogenicity of the cultivation milieu could alter lesion characteristics to better resemble natural residual caries [Clarkson et al., 1984; Shu et al., 2000; Schwendicke et al., 2014a]. This system would have the additional benefit of creating infected lesions, with inherent activity of both host and bacterial proteinases and the option of microbiological analysis, for example regarding arrest of differently infected lesions. Lesions induced by lactic acid gels might also be suitable for certain applications, whilst their mineral loss and mineral loss ratio were higher than those of natural lesions. This is in contrast to previous studies [Moron et al., 2013], but might be due to longer demineralisation and the specimens being agitated, which could have increased diffusion of minerals out of the lesion and the gel, impeding mineral saturation of the gel. This could also explain Caries Res 2015;49:259–265 DOI: 10.1159/000371897
263
Downloaded by: James H Quillen Coll of Med. 198.101.234.89 - 5/26/2015 4:55:26 PM
Acetic acid, n = 7
264
Caries Res 2015;49:259–265 DOI: 10.1159/000371897
have been sealed for certain time periods, i.e. have been remineralised, inactivated and re-hardened, might be possible as well after amending the protocol. In conclusion, demineralisation with MHDP-buffered, moderately acidic acetate solution induced lesions with similar mineral loss and nano-hardness properties as selectively excavated lesions. Both lesion types were relatively deep, moderately demineralised and showed constant and gradually increasing mineral content and nano-hardness along the demineralisation front. Modifying the presented protocol might broaden its further applicability in vitro, whilst certain investigations will require natural lesions or clinical settings.
Acknowledgements We thank Regina Marquardt, Conservative Dentistry and Periodontology, Christian Albrecht University Kiel, for her assistance with microradiographic analyses. The study was funded by the authors and their institutions.
Author Contributions Development of the protocol: F. Schwendicke, S. Paris, S. Gorb. Laboratory investigations: F. Schwendicke, A. Kovalev, K. Eggers. Analysis and interpretation of data: F. Schwendicke. Manuscript preparation: F. Schwendicke, S. Paris, H. Meyer-Lueckel, C. Dörfer, S. Gorb.
Disclosure Statement The authors declare no conflict of interest. The study was funded by the authors and their institutions.
References
Almqvist H, Wefel JS, Lagerlöf F: Root hard-tissue demineralization rate measured by 125I absorptiometry: comparison with lesiondepth measurements. J Dent Res 1990; 69: 1519–1521. Alves LS, Fontanella V, Damo AC, Ferreira de Oliveira E, Maltz M: Qualitative and quantitative radiographic assessment of sealed carious dentin: a 10-year prospective study. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2010;109:135–141. Balooch M, Wu-Magidi IC, Balazs A, Lundkvist AS, Marshall SJ, Marshall GW, Siekhaus WJ, Kinney JH: Viscoelastic properties of demineralized human dentin measured in water with atomic force microscope (AFM)-based indentation. J Biomed Mater Res 1998; 40: 539–544.
Schwendicke/Eggers/Meyer-Lueckel/ Dörfer/Kovalev/Gorb/Paris
Downloaded by: James H Quillen Coll of Med. 198.101.234.89 - 5/26/2015 4:55:26 PM
the absence of a surface layer, which was found in previous studies when using lactic acid gel demineralisation [Moron et al., 2013]. Similarly, the protocol using highly concentrated chelating agents like EDTA induced very deep lesions with extremely steep mineral gradients, no surface layer and nearly complete mineral eradication within the lesion body. This protocol is easy to perform and requires extremely short times for demineralisation, but seems unsuitable for the induction of artificial lesions mimicking those remaining after selective excavation. Our study is the first one to comparatively report the nano-hardness in both natural and artificial dentinal lesions. Whilst we corroborate findings from a previous study regarding the magnitude and cross-sectional profile of nano-hardness along carious dentin [Joves et al., 2014], it was relevant to compare this nano-hardness with that of artificial lesions, since mineral content in caries lesions was found to not readily translate into mechanical properties [Balooch et al., 1998; Kinney et al., 2003]. This was confirmed in our study, with nano-hardness increasing in areas identified as sound via microradiography. As discussed for the mineral content, the nano-hardness of natural lesions will further depend on both the original lesion properties (active or inactive lesions, older or younger teeth) and the used excavation criterion, resulting in the variability we found with regards to the mechanical properties of natural lesions [Zheng et al., 2003; Senawongse et al., 2006]. Given the hierarchical structure of dentin and the influence of specimen hydration on mechanical properties [Kinney et al., 1995; Balooch et al., 1998], nano-hardness measurements might generally yield more variable results than micro-hardness, but have the advantage of a higher resolution [Magalhães et al., 2009]. Given the focus of this study, we did not address all aspects related to the induction of artificial residual dentin lesions. We did, for example, not investigate whether the adhesive properties of artificial lesions are similar to those of natural, caries-affected or -infected dentinal lesions. The adhesive properties might be further modified by the presence (natural lesions) or absence (artificial lesions) of a smear layer, and studies in that direction might need to modify the presented protocol. Bacterial infection and enzymatic degradation of the dentin were not mimicked by the used acid-based demineralisation, but might be relevant with regards to lesion arrest or remineralisation of infected dentin. Again, modifications of our protocol (e.g. pre-demineralisation via acetic acid solutions followed by biofilm-based lesion progression and bacterial invasion) might be viable. Mimicking lesions which
Artificial Residual Caries
Kawasaki K, Ruben J, Tsuda H, Huysmans MC, Takagi O: Relationship between mineral distributions in dentine lesions and subsequent remineralization in vitro. Caries Res 2000;34: 395–403. Kinney JH, Balooch M, Haupt DL, Marshall SJ, Marshall GW: Mineral distribution and dimensional changes in human dentin during demineralization. J Dent Res 1995; 74: 1179– 1184. Kinney JH, Habelitz S, Marshall SJ, Marshall GW: The importance of intrafibrillar mineralization of collagen on the mechanical properties of dentin. J Dent Res 2003;82:957–961. Magalhães AC, Moron BM, Comar LP, Wiegand A, Buchalla W, Buzalaf MAR: Comparison of cross-sectional hardness and transverse microradiography of artificial carious enamel lesions induced by different demineralising solutions and gels. Caries Res 2009;43:474–483. Marquezan M, Correa FN, Sanabe ME, Rodrigues Filho LE, Hebling J, Guedes-Pinto AC, Mendes FM: Artificial methods of dentine caries induction: a hardness and morphological comparative study. Arch Oral Biol 2009; 54:1111–1117. Moron BM, Comar LP, Wiegand A, Buchalla W, Yu H, Buzalaf MAR, Magalhães AC: Different protocols to produce artificial dentine carious lesions in vitro and in situ: hardness and mineral content correlation. Caries Res 2013; 47: 162–170. Oliver WC, Pharr GM: An improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments. J Mater Res 1992; 7: 1564– 1583. Oong EM, Griffin SO, Kohn WG, Gooch BF, Caufield PW: The effect of dental sealants on bacteria levels in caries lesions. J Am Dent Assoc 2008;139:271–278. Paris S, Meyer-Lueckel H: Infiltrants inhibit progression of natural caries lesions in vitro. J Dent Res 2010;89:1276–1280.
Ricketts D, Lamont T, Innes NP, Kidd E, Clarkson JE: Operative caries management in adults and children. Cochrane Database Syst Rev 2013;3:CD003808. Schwartz AG, Pasteris JD, Genin GM, Daulton TL, Thomopoulos S: Mineral distributions at the developing tendon enthesis. PLoS One 2012;7:e48630. Schwendicke F, Dörfer C, Kneist S, Meyer-Lueckel H, Paris S: Cariogenic effects of probiotic Lactobacillus rhamnosus GG in a dental biofilm model. Caries Res 2014a;48:186–192. Schwendicke F, Dörfer CE, Paris S: Incomplete caries removal: a systematic review and metaanalysis. J Dent Res 2013;92:306–314. Schwendicke F, Kern M, Meyer-Lueckel H, Boels A, Doerfer C, Paris S: Fracture resistance and cuspal deflection of incompletely excavated teeth. J Dent 2014b;42:107–113. Senawongse P, Otsuki M, Tagami J, Mjör I: Agerelated changes in hardness and modulus of elasticity of dentine. Arch Oral Biol 2006; 51: 457–463. Shu M, Wong L, Miller JH, Sissons CH: Development of multi-species consortia biofilms of oral bacteria as an enamel and root caries model system. Arch Oral Biol 2000;45:27–40. Takada K, Hirasawa M: A novel selective medium for isolation of Streptococcus mutans. J Microbiol Methods 2005;60:189–193. Theuns HM, van Dijk JW, Driessens FC, Groeneveld A: The effect of undissociated aceticacid concentration of buffer solutions on artificial caries-like lesion formation in human tooth enamel. Arch Oral Biol 1984; 29: 759– 763. Weerheijm KL, de Soet JJ, van Amerongen WE, de Graaff J: Sealing of occlusal hidden caries lesions: an alternative for curative treatment? J Dent Child 1992;59:263–268. Zheng L, Hilton JF, Habelitz S, Marshall SJ, Marshall GW: Dentin caries activity status related to hardness and elasticity. Eur J Oral Sci 2003; 111:243–252.
Caries Res 2015;49:259–265 DOI: 10.1159/000371897
265
Downloaded by: James H Quillen Coll of Med. 198.101.234.89 - 5/26/2015 4:55:26 PM
Banerjee A, Kidd EAM, Watson TF: In vitro evaluation of five alternative methods of carious dentine excavation. Caries Res 2000; 34: 144– 150. Bowman A, Young S: Graphical comparison of nonparametric curves. J R Stat Soc Ser C Appl Stat 1996;45:83–98. Buskes JA, Christoffersen J, Arends J: Lesion formation and lesion remineralization in enamel under constant composition conditions. A new technique with applications. Caries Res 1985;19:490–496. Clarkson BH, Wefel JS, Miller I: A model for producing caries-like lesions in enamel and dentin using oral bacteria in vitro. J Dent Res 1984;63:1186–1189. Damen JJ, Buijs MJ, ten Cate JM: Fluoride-dependent formation of mineralized layers in bovine dentin during demineralization in vitro. Caries Res 1998;32:435–440. Enders S, Barbakadse N, Gorb SN, Arzt E: Exploring biological surfaces by nanoindentation. J Mater Res 2004;19:880–887. Gao XJ, Elliott JC, Anderson P: Scanning microradiographic study of the kinetics of subsurface demineralization in tooth sections under constant-composition and small constantvolume conditions. J Dent Res 1993; 72: 923– 930. Hevinga MA, Opdam NJ, Frencken JE, Truin GJ, Huysmans MC: Does incomplete caries removal reduce strength of restored teeth? J Dent Res 2010;89:1270–1275. Ismail A, Sohn W, Tellez M, Amaya A, Sen A, Hasson H, Pitts N: The International Caries Detection and Assessment System (ICDAS): an integrated system for measuring dental caries. Community Dent Oral Epidemiol 2007;35:170–178. Joves GJ, Inoue G, Sadr A, Nikaido T, Tagami J: Nanoindentation hardness of intertubular dentin in sound, demineralized and natural caries-affected dentin. J Mech Behav Biomed Mater 2014;32:39–45.
