INTRODUCTION

Incipient white spot lesion in enamel (WSLs) is generally believed to be remineralized1), and to enhance the remineralization, plaque control and fluoride therapy are recommended. However, patients at high risk of caries, such as physically and mentally handicapped people, may develop more acute WSLs over a relatively short period. In these cases, effective treatment for arresting the WSLs is required.

Recently, the resin infiltration technique (ICON®) has been introduced2). The main chemical component is TEGDMA (triethylene glycol dimethacrylate)3). This product is designed to infiltrate resinous material within WSLs for inhibiting further lesion demineralization or for arresting the lesion3,4). Thus far, several in vitro studies have been carried out to compare the effects of acid-etching conditions5,6) and the resin penetration rate (or viscosity) on the degree of resin penetration into the lesion body7,8).

Paris et al. reported several studies on the inhibitory efficacy of the resin upon further demineralization, in vitro3,4,9), in vivo10), and in clinical trials2). Among them, certain studies demonstrated a long-term efficacy of the resin on anti-demineralization2,10), regardless of the absence of a functional group for chemical adsorption to the apatite surface in the molecule. The present authors assumed that the long-term efficacy would be attributed to a good adaptation of the resin at the interface between the resin and the lesion bottom surface (R–L interface) and resistance against the deterioration of the resin.

Traditionally, thermal cyclic treatment of resin-restored specimens has been employed to simulate physiological aging and predict the clinical durability of dental materials11,12). Although the mechanism of the artificial aging of dental materials by thermal cycling treatment in water storage is not well understood, it could be said that deterioration in the mechanical properties of the material will be caused by cycling of thermal expansion and contraction12). Cycling of water absorption and excretion may enhance the hydrolytic deterioration of the material and dissolution of water soluble components in the matrix12). It was reported that TEGDMA had a capability of high water sorption13); thus, degradation might have occurred during thermal cycling.

On the other hand, clinical diagnosis for the gap formation at the R–L interface would be valuable to monitor the integrity of the resin. This gap may allow acid penetration, leading to demineralization of enamel and dentin at the R–L interface. Swept-source optical coherence tomography (SS-OCT) is an emerging diagnostic method for obtaining cross-sectional images of internal biological structures in real-time without the use of X-ray irradiation14). For instance, studies on the detection and quantification of de-/remineralization in enamel and dentin have been performed both in vitro and in vivo14). Bakhsh et al. examined the resin-dentin interfacial micrometer gap by SS-OCT and validated the technology against confocal microscopy15). The principle for detecting the gap is based on the difference in light refractive index between two interfacing phases16).

To our knowledge, no study has reported the effect of artificial aging by thermal cyclic stress of the resin-

Effect of thermal cyclic stress on acid resistance of resin-infiltrated incipient enamel lesions in vitroNami TAKASHINO1, Syozi NAKASHIMA1, Yasushi SHIMADA1, Junji TAGAMI1 and Yasunori SUMI2

1 Department of Cariology and Operative Dentistry, Division of Oral Health Sciences, Graduate School of Medical and Dental Sciences, Tokyo Medical and Dental University, 1-5-45 Yushima, Bunkyo-ku, Tokyo 113-8549, Japan

2 Division of Oral and Dental Surgery, Department of Advanced Medicine, National Hospital for Geriatric Medicine, National Center for Geriatrics and Gerontology, 36-3 Gengo, Morioka, Obu-city, Aichi 474-8511, Japan

Corresponding author, Syozi NAKASHIMA; E-mail: nakashima_ japanjp@yahoo.co.jp

This study aimed to examine the effect of thermal cycling on gap formation at the interface between infiltrated resin (ICON®) and enamel lesion and on the durability of anti-demineralization efficacy to predict the future performance. SS-OCT technique was examined to determine whether it has the potential to detect the gap. Bovine enamel lesions were prepared, and the infiltrated resin was applied to the lesion. Resin-infiltrated lesion specimens were thermal cycled 10,000 cycles and further demineralized in pH 4.5 buffer for 7 days. Released Ca (mg/cm2) was quantified by Ca electrode. The SS-OCT technique was applied to detect the gap, and SEM observation was performed to determine the presence of the gap. There was no significant difference in the amount of Ca release before and after the thermal cycling, suggesting long-lasting anti-demineralization efficacy of the resin. SS-OCT and SEM observations indicated no apparent gap formation after the thermal cycling.

Keywords: Resin infiltration, Enamel lesions, SS-OCT, Thermal cycle, Acid resistance

Color figures can be viewed in the online issue, which is avail-able at J-STAGE.Received Oct 7, 2015: Accepted Dec 25, 2015doi:10.4012/dmj.2015-341 JOI JST.JSTAGE/dmj/2015-341

Dental Materials Journal 2016; 35(3): 425–431

Fig. 1 Flow chart demonstrating the experimental process.

infiltrated incipient enamel lesion in water storage upon the gap formation at the R–L interface and acid resistance. Thus, the current study aimed to examine the effect of thermal cyclic stress on the gap formation at the R–L interface by SS-OCT and SEM observations, as well as the durability of acid resistance, to predict future performance. The null hypotheses are that 1) the thermal cyclic stress affects the integrity of sealing at the R–L interface; 2) it affects the anti-demineralization efficacy; and 3) SS-OCT is not able to detect the gap.

MATERIALS AND METHODS

A flow chart of the experimental processes in this study is shown in Fig. 1.

Enamel specimen preparation and demineralizationApproximately 5×5×2-mm enamel blocks were prepared from freshly extracted bovine crown using a low speed diamond saw (Isomet, Buehler, IL, USA). The enamel surface was polished on silicon carbide polishing paper (Sankyo, Saitama, Japan) with a final grid of 2000, and then the blocks were resin embedded (Shade clear, UNIFAST 2, GC, Tokyo, Japan). The specimens were put into an ultrasonic bath for 3 min to remove debris attached to the specimen surface. The marginal area of the polished enamel surface was coated using nail varnish (Revlon Nail Enamel, Revlon, NY, USA) including the resin, leaving a window for treatment of approximately 2×2 mm. The enamel specimens (n=50) were immersed in 250 mL demineralization solution (CaCl2: 5.0 mmol/L, KH2PO4: 4.0 mmol/L, acetic acid: 50 mmol/L, NaN3: 0.02%, pH 4.8) for 19 days at 37°C to create a subsurface lesion in bovine enamel (Fig. 2). The solution was not renewed during the demineralization period.

ICON® treatmentBefore the ICON (DMG, Hamburg, Germany) treatment, the nail varnish, which might affect the resin penetration of ICON, was carefully removed by a hand scaler from the coated enamel surface. According to the instructions of the manufacturer, ICON treatment was performed on the demineralized enamel lesion. Briefly, hydrochloric acid (HCl) gel was placed on the demineralized enamel surface, including the surrounding sound enamel surface for 3 min and then rinsed by deionized water for 30 s followed by dehydration using alcohol. The ICON resin monomer was then applied and light-cured (Optilux 501, Kerr, Wisconcin, USA) for 40 s. The resin treatment was repeated twice. The top surface of uncured ICON resin was gently polished using abrasive containing non-fluoride paste (Pressage, Shofu, Kyoto, Japan) until the surface became harder. In the present study, no intensive air-blowing of uncured ICON resin was performed to avoid unnecessary removal of the resin. Clinically, this application would be conducted for treatment at the interproximal area17).

Thermal cyclic treatmentICON treated specimens were randomly divided into three groups (15 specimens per group): A) without thermal cycle; B) with 5,000; and C) with 10,000 thermal cycles. The specimens in group B and C were immersed in deionized water baths at 5 and 55°C for 30 s in each temperature and a transfer time of 2 s between baths (Cool Line CL200 and Cool Mate TE200, Yamato Scientific, Tokyo, Japan), during which the specimens in group A were kept moist in a refrigerator.

Assessment of anti-demineralization efficacy of ICON by Ca electrodeEach specimen in group A, B, and C was immersed in 5.0 mL of 50 mmol/L pH 4.5 acetic acid demineralizing solution at 37°C for 7 days to which 100 mmol/L KCl was added for ion concentration adjustment, for Ca assessment by the electrode (Laqua® 6583-10C, Horiba, Kyoto, Japan). Each group contained 12 specimens. The specimens without ICON treatment (n=10) were similarly demineralized and used as a control group. After the demineralization, the solution pH was adjusted to pH 5.5 by adding concentrated KOH solution, which was in the optimal range of pH for the Ca determination. The required KOH solution for pH adjustment was less than 0.1 mL in any specimen. The electrode was calibrated against Ca standard solutions of pH 5.5, which contained 0.1, 0.5, 1.0, and 5.0 mmol/L CaCl2, 50 mmol/L acetic acid, and 100 mmol/L KCl. The calibration showed a good linearity with R2=0.98–1.0. During this Ca measurement, the enamel surface around the ICON coated surface was carefully sealed using nail varnish, leaving only the ICON covered area unsealed to prevent Ca release from the ICON uncovered enamel surface.

Transverse Micro Radiographic (TMR) analysisTo confirm that the demineralized specimens before ICON treatment were subsurface lesions with a surface layer and that the surface of HCl gel treated specimens was removed, as well as for estimating the lesion depth of the uncoated (control) group after 7 days demineralization, TMR mineral assessment was independently

426 Dent Mater J 2016; 35(3): 425–431

Fig. 2 Transverse microradiography images of the enamel subsurface lesion a) before HCl etching treatment and b) after HCl etching treatment.

Surface layer is seen in a) but not in b).

Fig. 3 Graphs showing a) Ca2+ release (mg/cm²) and b) pH changes after the second demineralization. a), b) *p<0.0083, Wilcoxon rank sum test with Bonferroni correction.

performed using five specimen blocks. Thin sections of approximately 120 μm thickness were prepared from the blocks. The details of the TMR procedure is described elsewhere18). By use of a customized image processing software (ImageJ, version1.42q, Wayne Rasband, NIH, USA), mean depth profiles of the lesion images over approximately 1 mm of the lesion window were created, and the lesion depth, mineral loss (ΔZ: vol%•μm), and maximum surface mineral density were then measured.

SEM observation and energy dispersive X-ray spectrometry (EDS) analysisSEM observation was performed to examine whether the gap at the R–L interface was present. For this purpose, the block specimens were cross-sectionally cut into sections of approximately 2 mm thickness. The cut surface of the specimens was fine-polished using a polishing device (ML-160 A, Maruto, Tokyo, Japan) and diamond paste slurries of different particle sizes (decreasing serially from 6 μm down to 0.25 μm in

diameter), and the polished specimens were placed in a covered glass vial and dehydrated at room temperature for 1 day.

After osmium coating, the specimens were observed by Field-Emission SEM/EDS (S-4500, Hitachi, Hitachinaka, Japan) with an accelerating voltage of 20 kV, and the gap at the R–L interface was examined.

EDS analysis was performed to evaluate the resin penetration depth in the treated specimens. The specimens for SEM observation were used for EDS analysis, and elemental line analyses of Ca and carbon were carried out (group A: n=5, B: n=3, C: n=3).

SS-OCT observationThe SS-OCT system (Santec OCT-2000®, Santec, Komaki, Japan) was used to observe the gap at the R–L interface. The axial and lateral resolutions of the system in air were 11 and 17 μm, respectively. Two-dimensional cross-sectional images were created from serial backscatter (reflectivity) profiles (A-scans) along the surface18). To observe the same site on the specimen before and after thermal cyclic treatment, a line was drawn on the nail varnished surface as a marker. The specimen was placed at the same position on the metal stage as accurately as possible. The laser beam was applied at 90° to the tooth surface.

Statistical analysesIn terms of Ca release and pH, statistical analyses were performed for multiple comparisons among four groups, i.e., control without ICON, with ICON but no thermal cyclic treatment, and with thermal cyclic treatment (5,000 and 10,000 thermal cycles). Data were analyzed by the Wilcoxon rank sum test. To avoid accumulation errors due to multiple comparisons, the significance level was modified to p< 0.0083 by dividing 0.05 by six comparisons (Bonferroni Correction). All statistical procedures were analyzed using PASW Statistics 18 (SPSS, Chicago, IL, USA).

RESULTS

Figure 2 shows representative TMR images of the subsurface lesions used before (a) and after (b) the HCl etching treatment. In the lesions before etching (n=6),

427Dent Mater J 2016; 35(3): 425–431

Fig. 4 Optical coherence tomography images of resin infiltration with a) no thermal cycling (TC/0), b) at 5,000 cycles (TC/5,000) and c) at 10,000 cycles (TC/10,000).

No apparent gap is observed between the resin front and the lesion base. Several white dots were obviously seen at the resin surface in TC/10,000, which is indicated by arrows. These dots were confirmed as cleavages by SEM image.

Fig. 5 Scanning electron microscope images focusing on the interface between the resin front and lesion base to observe the adaptation at low magnification (×200), a) TC/0, b) TC/5,000, c) TC/10,000, and at high magnification (×1,000), d) TC/0, e) TC/5,000, f) TC/10,000. No apparent gap was observed at the interface.

a well-preserved surface layer was observed. The lesion depth, mineral loss (ΔZ), and the maximum surface mineral density were 135±34 μm, 2,065±503 vol%•μm, and 55±2 vol%, respectively. The corresponding lesion depth and ΔZ values after the HCl etching (n=7) were 177±32 μm and 2,902±342 vol%•μm, respectively. HCl

etching increased the lesion extent and removed the surface layer, facilitating resin penetration to the lesion body. These characteristics indicated that the lesions used were similar to a type of active lesion. On the other hand, the 7 days demineralized lesions (n=5) in the ICON uncoated control group were severely destroyed, showing cavitation without a surface layer and a lesion depth of 174±10 μm and ΔZ of 10,525±948 vol%•μm.

In Fig. 3, the amount of Ca (mg/cm2) released from the specimens (a) and pH values after the demineralization (b) are shown. In the current study, two specimens in the ICON coated group (5,000 and 10,000 thermal cycles) showed partial detachment of the nail varnish at the margin of the ICON coated surface and five- to ten-fold greater amount of Ca (mg/cm2) release. Thus, the two specimens were excluded from the statistical analyses. An additional experiment revealed that no detectable Ca was observed in the specimens whose enamel surface was completely coated by the nail varnish (data not shown).

There were significant differences in the released Ca (mg/cm2) between the three ICON coated groups: group A (n=10): 0.09±0.07 mg/cm2, group B (n=10): 0.08±0.05 mg/cm2, and group C (n=10): 0.09±0.05 mg/cm2 and the uncoated control group (n=10): 1.75±0.19 mg/cm2. Percentage mean reduction of Ca (mg/cm2) released in the three ICON coated groups as compared to the control group was 95.4%. However, no difference was noted among the three ICON coated groups. Figure 3b indicates pH changes after the demineralization. As compared to the control group (4.82±0.02), the ICON coated groups showed significantly lower pH values (4.52–4.53) which were closer to the original pH value of 4.50, indicating minimum demineralization.

428 Dent Mater J 2016; 35(3): 425–431

Fig. 6 Energy dispersive spectroscopy images of carbon (C) and calcium (Ca). Sharp differences were seen between the infiltrated resin layer (R) and enamel (E). Ca

was not detected in the resin layer.

Figure 4 shows representative SS-OCT images of the resin-infiltrated lesion with (a) no thermal cycling (TC/0; n=10), (b) 5,000 (TC/5,000; n=10), and (c) 10,000 cycles (TC/10,000; n=10). In TC/10,000 groups, several white dots were obviously seen at the resin surface, which is indicated by arrows. These dots were confirmed as cleavages by SEM image as shown below; however, no gaps were evident at the R–L interface, irrespective of the thermal cycles. Furthermore, after the second demineralization for 7 days, no gap as well as no obvious increase in refractive intensity at the R–L interface was apparent (image not shown).

SEM images of the cross-sectioned lesion focusing on the presence of a gap at the R–L interface were taken to further precisely observe the R–L interface. Figure 5a–c show low magnification images (×200). Cleavages in the ICON resin layer can be seen in TC/5,000 and TC/10,000, but not in the TC/0 groups; however, the cleavages did not reach the R–L interface. Evidently, no gap at the R–L interface was observed with and without the thermal cycling. Higher magnification (×1,000) shown in Figs. 5d–e revealed good adaptation of the resin.

Figure 6 shows representative EDS line profiles of carbon (C) and calcium (Ca) from the top surface to the sound enamel below the lesion body together with the corresponding SEM images. Higher carbon intensity in infiltrated resin area (R) and Ca intensity in sound enamel (E) are evident, indicating the ICON penetration depth being approximately 180±30 μm, which was close to the lesion depth (177±32 μm) determined by TMR measurement. On the other hand, no Ca intensity was observed at the ICON penetrated region.

DISCUSSION

In the current study, we employed considerably severe demineralization conditions, which would rarely occur in vivo, i.e., continuous demineralization at pH 4.5 for 7 days without inclusion of Ca and phosphate ions in the demineralization solution, as we considered ICON to be a stable resinous material over time to perform strong sealing efficacy against acid attack. In fact, TMR observation revealed severely destroyed lesions after the 7 days demineralization in the ICON uncoated control group, which lost surface layers completely.

ICON coating demonstrated almost perfect protection of enamel demineralization from acid attack, and the percentage reduction was calculated as 95.4% (Fig. 3). Paris et al. reported similar results of approximately 90% inhibition of further demineralization in the natural WSL in enamel as compared with the untreated control group after long-term immersion (400 days) of the enamel lesions coated by ICON, with higher penetration coefficient resin in a pH 4.95 acid buffer4). In addition, the same group found approximately 86% inhibition of further demineralization of artificial enamel caries lesions under cariogenic conditions in situ for 100 days10). Since there was no difference in Ca (mg/cm2) release between 5,000 and 10,000 thermal cycles in the current study, it was suggested that the efficacy of protection would be maintained for more than 10,000 thermal cycles. Gale and Darvell suggested a provisional estimate of approximately 10,000 thermal cycles per year19).

Significantly higher anti-demineralization efficacy

429Dent Mater J 2016; 35(3): 425–431

could be explained by the current findings that no apparent gap at the R–L interface was noted by SEM and SS-OCT observations. The gap would facilitate acid penetration. In more detail, all of the three ICON coated groups showed a minimum amount of Ca (mg/cm2) release, suggesting that ICON material might allow acid penetration to some extent during a long-term period. As mentioned above4,10), slight demineralization was noted after a long-term exposure to acid in the ICON coated group. This might be due to incomplete embedding of enamel minerals in the lesion body, or due to the resin shrinkage during light curing20). However, the limited acid penetration was considered sufficient for arresting the incipient enamel lesion in vivo.

It would be valuable to consider the reason that the ICON coated group exhibited no apparent gap formation and almost perfect protection from acid attack, from a view point of the chemical composition of the resin. Earlier studies found that the resin with high penetration coefficients could penetrate deeper inside the lesion20,21). The ICON resin monomer (TEGDMA) contains no functional group in the molecule to exert strong chemical bonding performance to the apatite surface such as esterized phosphate, e.g., methacryloyloxydecyl dihydrogen phosphate (MDP)22) and carboxy groups e.g., 4-methacryloxyethyl trimellitate anhydride (4-META)23). Thus, it is believed that TEGDMA does not have the ability to adhere to teeth24). One possible chemical interaction would be the presence of a carbonyl group (C=O) in the molecule. It is generally known that the carbonyl group forms a hydrogen bond25); thus, we considered that the hydrogen bond between the carbonyl group and the OH group in the apatite lattice or H2O at the stern layer on apatite surface26) likely occurred.

Another possibility is that micro adaptation of the ICON resin to the demineralized porous enamel surface with increased roughness by the demineralization occurred, as ICON preparation is designed to have a low viscosity to facilitate the resin penetration into the lesion4), leading to mechanical interlocking.

It has been reported that thermal cycling stress in water storage affects the integrity at the tooth and restorative material interface. De Munck et al. explained that the chemical degradation process at the interface occurs in two ways: 1) hot water may accelerate the hydrolysis of interface components and the subsequent uptake of water and extraction of breakdown products or poorly polymerized resin oligomer, or 2) due to the higher thermal contraction/expansion coefficient of the restorative material (as compared with that of tooth tissue), repetitive contraction/expansion stress is generated at the interface27). These stresses may lead to cleavages or gap formation at the R–L interface. In the present study, no apparent gap was observed, although the reason for this is unclear. We assumed that the difference in thermal expansion between the ICON resin and the enamel minerals was not significant at the interface, probably due to inclusion of enamel minerals in the lesion body into the resin, where the mineral density was distributed increasingly from the

R–L interface to the sound enamel bottom as shown in Fig. 2. This assumption might be supported by another study showing that a highly filled hybrid composite had a coefficient of thermal expansion closest to that of the tooth crown28). Furthermore, it is likely that TEGDMA contains no hydrolysable chemical bond in the molecule at the temperature range and under the acidic condition used; thus, this property might be partly attributed to no apparent gap formation.

The SS-OCT images of the ICON coated specimens did not show the apparent gap at the R–L interface, irrespective of the thermal cycles (Fig. 4), which was similar to outcomes of SEM observation. Gray and Shellis confirmed by SEM observation that artificial caries lesions could be infiltrated almost completely with the resin29). Moreover, after the second demineralization, no obvious increase in refractive intensity was seen at the R–L interface, suggesting no further demineralization30,31). In our study, under similar thermal cyclic conditions, gap formation between the dentin cavity floor and a flowable resin composite was detected by confocal laser scanning microscopy and SS-OCT32).Therefore, if the gap had been created, SS-OCT was assumed to have detected the gap in the current study. The minimum Ca (mg/cm2) release in the ICON coated group as compared with the control group may support this assumption.

In contrast, obvious cleavage was seen at the resin surface (Fig. 4). This might be explained by preferential deterioration and dissolution of the resin at the surface, in which uncured resin material might remain, even after gentle polishing. The preferential deterioration and dissolution might result in shrinkage of the water-adsorbed resin when the specimen was observed by SS-OCT under a dried condition. The cleavage was additionally evident in SEM images (Fig. 5), which would be due to drying of the specimens for the SEM preparation.

EDS profiles of carbon (C) and Ca indicated sharp intensity differences at the R–L interface (Fig. 6), and Ca intensity was not detected within the resin layer. However, as shown in Fig. 2, the residual minerals should have remained in the lesion body, even after removal of the surface layer by HCl etching, indicating that Ca should have been detected to some extent in the resin layer. This is probably due to the resin penetration into the porous residual mineral zone, resulting in embedding of the residual minerals by the resin.

The first and second null hypotheses, namely that the thermal cyclic stress affects the integrity of sealing at the interface and that it affects the anti-demineralization efficacy were rejected. On the other hand, the third hypothesis that SS-OCT is not able to detect the gap was not clearly rejected, since no apparent gap was created under the present experiment condition.

CONCLUSION

Thermal cyclic stress did not affect the sealing ability and the anti-demineralization efficacy of ICON. No clear

430 Dent Mater J 2016; 35(3): 425–431

conclusion was obtained regarding that SS-OCT was able to detect the gap, since no apparent gap was created in this study. Further study is therefore required.

ACKNOWLEDGMENT

The authors thank Yoshida Dental Trade Distribution. for providing the ICON®. This research was supported by Grant-in-Aid for Scientific Research no. 25462949 from the Japan Society for the Promotion of Science.

CONFLICTS OF INTEREST

The authors declare no potential conflicts of interest with respect to the authorship and/or publication of this article.

REFERENCES

1) Nicol AJ, Creanor S, Hall AF, Macpherson LM. A randomised, observer-blind, clinical study investigating the effect of fluoridated milk on artificially created early enamel caries using transverse microradiography. Caries Res 2008; 42: 305-311.

2) Paris S, Hopfenmuller W, Meyer-Lueckel H. Resin infiltration of caries lesions: an efficacy randomized trial. J Dent Res 2010; 89: 823-826.

3) Paris S, Schwendicke F, Seddig S, Muller WD, Dorfer C, Meyer-Lueckel H. Micro-hardness and mineral loss of enamel lesions after infiltration with various resins: influence of infiltrant composition and application frequency in vitro. J Dent 2013; 41: 543-548.

4) Paris S, Meyer-Lueckel H. Infiltrants inhibit progression of natural caries lesions in vitro. J Dent Res 2010; 89: 1276-1280.

5) Paris S, Meyer-Lueckel H, Kielbassa AM. Resin infiltration of natural caries lesions. J Dent Res 2007; 86: 662-666.

6) Neuhaus KW, Schlafer S, Lussi A, Nyvad B. Infiltration of natural caries lesions in relation to their activity status and acid pretreatment in vitro. Caries Res 2013; 47: 203-210.

7) Meyer-Lueckel H, Chatzidakis A, Naumann M, Dorfer CE, Paris S. Influence of application time on penetration of an infiltrant into natural enamel caries. J Dent 2011; 39: 465-469.

8) Paris S, Soviero VM, Seddig S, Meyer-Lueckel H. Penetration depths of an infiltrant into proximal caries lesions in primary molars after different application times in vitro. Int J Paediatr Dent 2012; 22: 349-355.

9) Gelani R, Zandona AF, Lippert F, Kamocka MM, Eckert G. In vitro progression of artificial white spot lesions sealed with an infiltrant resin. Oper Dent 2014; 39: 481-488.

10) Paris S, Meyer-Lueckel H. Inhibition of caries progression by resin infiltration in situ. Caries Res 2010; 44: 47-54.

11) Morresi AL, D’Amario M, Capogreco M, Gatto R, Marzo G, D’Arcangelo C, Monaco A. Thermal cycling for restorative materials: does a standardized protocol exist in laboratory testing? A literature review. J Mech Behav Biomed Mater 2014; 29: 295-308.

12) De Munck J, Van Landuyt K, Coutinho E, Poitevin A, Peumans M, Lambrechts P, Van Meerbeek B. Micro-tensile bond strength of adhesives bonded to Class-I cavity-bottom dentin after thermo-cycling. Dent Mater 2005; 21: 999-1007.

13) Sideridou I, Tserki V, Papanastasiou G. Study of water sorption, solubility and modulus of elasticity of light-cured dimethacrylate-based dental resins. Biomaterials 2003; 24: 655-665.

14) Mandurah MM, Sadr A, Shimada Y, Kitasako Y, Nakashima S, Bakhsh TA, Tagami J, Sumi Y. Monitoring remineralization of enamel subsurface lesions by optical coherence tomography. J Biomed Opt 2013; 18: 046006.

15) Bakhsh TA, Sadr A, Shimada Y, Tagami J, Sumi Y. Non-invasive quantification of resin-dentin interfacial gaps using optical coherence tomography: validation against confocal microscopy. Dent Mater 2011; 27: 915-925.

16) Turkistani A, Sadr A, Shimada Y, Nikaido T, Sumi Y, Tagami J. Sealing performance of resin cements before and after thermal cycling: evaluation by optical coherence tomography. Dent Mater 2014; 30: 993-1004.

17) Kielbassa AM, Muller J, Gernhardt CR. Closing the gap between oral hygiene and minimally invasive dentistry: a review on the resin infiltration technique of incipient (proximal) enamel lesions. Quintessence Int 2009; 40: 663-681.

18) Natsume Y, Nakashima S, Sadr A, Shimada Y, Tagami J, Sumi Y. Estimation of lesion progress in artificial root caries by swept source optical coherence tomography in comparison to transverse microradiography. J Biomed Opt 2011; 16: 071408.

19) Gale MS, Darvell BW. Thermal cycling procedures for laboratory testing of dental restorations. J Dent 1999; 27: 89-99.

20) Paris S, Meyer-Lueckel H, Colfen H, Kielbassa AM. Resin infiltration of artificial enamel caries lesions with experimental light curing resins. Dent Mater J 2007; 26: 582-588.

21) Meyer-Lueckel H, Paris S. Improved resin infiltration of natural caries lesions. J Dent Res 2008; 87: 1112-1116.

22) Yoshida Y, Nagakane K, Fukuda R, Nakayama Y, Okazaki M, Shintani H, Inoue S, Tagawa Y, Suzuki K, De Munck J, Van Meerbeek B. Comparative study on adhesive performance of functional monomers. J Dent Res 2004; 83: 454-458.

23) Van Landuyt KL, Snauwaert J, De Munck J, Peumans M, Yoshida Y, Poitevin A, Coutinho E, Suzuki K, Lambrechts P, Van Meerbeek B. Systematic review of the chemical composition of contemporary dental adhesives. Biomaterials 2007; 28: 3757-3785.

24) Wang T, Nikaido T, Nakabayashi N. Photocure bonding agent containing phosphoric methacrylate. Dent Mater 1991; 7: 59-62.

25) Nishiyama N, Suzuki K, Komatsu K, Yasuda S, Nemoto K. A 13C NMR study on the adsorption characteristics of HEMA to dentinal collagen. J Dent Res 2002; 81: 469-471.

26) Arends J, Jongebloed WL. The enamel substrate-characteristics of the enamel surface. Swed Dent J 1977; 1: 215-224.

27) De Munck J, Van Landuyt K, Peumans M, Poitevin A, Lambrechts P, Braem M, Van Meerbeek B. A critical review of the durability of adhesion to tooth tissue: methods and results. J Dent Res 2005; 84: 118-132.

28) Versluis A, Douglas WH, Sakaguchi RL. Thermal expansion coefficient of dental composites measured with strain gauges. Dent Mater 1996; 12: 290-294.

29) Gray GB, Shellis P. Infiltration of resin into white spot caries-like lesions of enamel: an in vitro study. Eur J Prosthodont Restor Dent 2002; 10: 27-32.

30) Turkistani A, Nakashima S, Shimada Y, Tagami J, Sadr A. Microgaps and demineralization progress around composite restorations. J Dent Res 2015; 94: 1070-1077.

31) Nazari A, Sadr A, Campillo-Funollet M, Nakashima S, Shimada Y, Tagami J, Sumi Y. Effect of hydration on assessment of early enamel lesion using swept-source optical coherence tomography. J Biophotonics 2013; 6: 171-177.

32) Makishi P, Shimada Y, Sadr A, Tagami J, Sumi Y. Non-destructive 3D imaging of composite restorations using optical coherence tomography: marginal adaptation of self-etch adhesives. J Dent 2011; 39: 316-325.

431Dent Mater J 2016; 35(3): 425–431