121 Clinical Report 121 In Vitro-evaluation of Secondary Caries Formation around Restoration Ricardo Coelho Okida, Fernando Mandarino*, Renato Herman Sundfeld, Rodrigo Sversut de Alexandre** and Maria Lúcia M. M. Sundefeld*** Department of Restorative Dentistry, Araçatuba School of Dentistry, UNESP *Department of Restorative Dentistry, Ribeirão Preto School of Dentistry, USP **Department of Restorative Dentistry, Piracicaba School of Dentistry, UNICAMP ***Department of Biostatistics, Araçatuba School of Dentistry, Paulista State University, UNESP Received 4 October, 2006/Accepted for publication 14 May, 2008 Abstract The objective of this in vitro study was to evaluate demineralization around restora- tions. Class V preparations were made on the buccal and lingual surfaces of each tooth. TPH (Group 1), Fuji II LC (Group 2), Tetric (Group 3), Dyract (Group 4), GS 80 (Group 5) and Chelon Fil (Group 6) were randomly placed in equal numbers of teeth. The teeth were submitted to a pH-cycling model associated with a thermocycling model. Sections were made and the specimens were examined for the presence of demineralization under polarized light microscopy. Demineralization was significantly reduced with Chelon Fil (Group 6). Furthermore, a similar inhibitory effect on the development of demineralization was observed in Groups 2, 4 and 5. Key words: Fluoride—Demineralization—Polarization microscopy in restorative dentistry1,6). Fractures, insufficient marginal adaptation and excess restorative material, failures that are frequently observed in restorations, result in microleakage. Openings permit penetra- tion of oral fluids and microbiological agents along the interface between dental tissue and restoration, which may lead to the develop- ment of secondary caries5). It is currently believed that the cariostatic effect of fluoride is enhanced in lower, yet permanent, concentrations in the oral envi- ronment, so incorporating fluoride into Introduction The development of marginal alterations and secondary caries around composite resin restorations has been documented9,10,25) and is considered a major cause of restorative failure and replacement7,20,25). The key factors in pre- venting secondary caries around restorations are marginal integrity of the restoration14), durable adhesion15), the physical properties of the restorative materials themselves, and oral hygiene13). However, replacement of restora- tions due to secondary caries is still a problem Bull Tokyo Dent Coll (2008) 49(3): 121–128 122 Okida RC et al. restorative materials is of special interest19,22). Fluoridated materials offer a potential source of this element, which would expand the spec- trum of prevention in restorative dentistry. The use of fluoride to demineralize and remineralize early enamel carious lesions directly interferes with the evolution of caries1,6,19). It is accepted that part of the fluo- ride present in restorative materials may be directly released onto areas of risk, such as restoration margins, where secondary caries may develop1,19,22). Thus, the use of fluoride may be considered an additional method of preventing caries. Glass ionomer cement, initially described in the early 1970s, is regarded as the material of choice in cases where cavity sealing and pre- vention of secondary caries are desirable, due to its properties of adhesion to dental struc- tures and its high rate of fluoride release16,17,22). However, the predominance of resin com- posites among esthetic restorative materials is evident, especially due to their highly satisfactory esthetics and easy manipulation. The greatest difficulty with this material is the occurrence of secondary carious lesions adjacent to the restoration, which is observed less frequently in teeth restored with glass ionomer1,13,17,22). In an attempt to obtain an ideal material that would combine the good properties of both resin composites and glass ionomer cements, new materials have been developed, such as resin-modified glass ionomers and polyacid-modified resin composites. Fluoride-releasing restorative materials are important in inhibiting secondary carious lesions, especially in patients at high risk1,22) and those with high caries activity. Therefore, a comparative evaluation of the cariostatic action of such materials together with that of conventional resin composites is required to enhance our knowledge of their behavior. The aim of this study was to evaluate marginal demineralization around restorations using polarized light microscopy. Materials and Methods Thirty extracted human third permanent molars, which had been cleaned and stored in 2% formaldehyde solution (pH7.0) at room temperature, were utilized. Class V preparations were made in the middle third of both buccal and lingual surfaces of each tooth with a water-cooled high-speed handpiece and a # 16F diamond bur (KG Sorensen, Barueri, SP, Brazil). The approximate dimensions of the 60 formed cavities were: 1.5 mm in depth and 1.2mm in diameter. The bur was discharged after every 5 cavities. No bevel was made on the cavo- surface angle. The tested restorative materials placed in the prepared teeth are shown in Table 1. Prior to restoration placement, the teeth were individually immersed in deionized water and randomized into 6 treatment groups of 10 specimens each (Table 2). The technique for application of all materials followed the manufacturers’ instructions. Teeth restored with composite resin (TPH) and Polyacid modified composite resin (Dyract) after acid Table 1 Materials used in each group Material Group Manufacturer Lot # Composite resin (TPH) 1 Light-cured Dentsply C830475 Resin modified glass-ionomer (Fuji II LC) 2 Light-cured G.C. American 310351 Fluoride-containing composite resin (Tetric) 3 Light-cured Vivadent 580356 Polyacid modified composite resin (Dyract) 4 Light-cured Dentsply C940756 GS 80 5 — SDI 091316 Glass-ionomer (Chelon Fil) 6 Self-cured ESPE Z098 123Secondary Caries around Restoration etching (37% phosphoric acid), received 2 applications of adhesive system Prime & Bond (Dentsply Caulk, Milford, DE, USA). The incremental technique was used for these restorative materials. Only specific pre- treatment was applied prior to the restorative procedure (Table 2). All restored teeth were stored in a humid environment for 48hrs to allow completion of reactions such as polymerization, crystalliza- tion and gelation. Restorations in groups 1, 2, 3, 4 and 6 were polished with Sof-Lex Pop On (3M/ESPE Dental Products St. Paul, USA) disks; amal- gams were finished with a low-speed hand- piece and flame-shaped burs, followed by abrasive rubber (KG Sorensen, Barueri, SP, Brazil). Round segments (4 mm in diameter) of adhesive tape (3M/ESPE Dental Products, St. Paul, USA) were placed on the restorations. Then, all tooth surfaces were coated with acid-resistant varnish. A nylon wire was fixed at each tooth to facilitate its handling. The tape was removed from the tooth as soon as the varnish dried, leading to exposure of the restoration as well as of 1mm of dental tissue around the restorative material. Each group of teeth was immersed in 100ml synthetic acid demineralizing solution (2.0mmol l�1 Ca, 2.0mmol l�1 P, in 75mmol l�1 acetate buffer, pH 4.3) for 4 hrs. The teeth were then immersed in 20 ml remineralizing solution (1.5 mmol l�1 Ca, 0.9mmol l�1 P, 0.1 buffer, pH 7.0) for 20hrs. All teeth were rinsed thoroughly with deionised water and dried with absorbent paper before and after the dematerializing period. Continuous cycles of demineralization and reminerali- zation were carried out for 28 days. During the in vitro demineralization/remineralization cycling model, the teeth in each group were subjected to thermocycling for 100 cycles. A complete cycle consisted of 90 sec at 37°C, 90 sec at 55°C and 90sec at 5°C. After 28 days, the teeth were individually fixed in acrylic resin blocks and sectioned to a thickness of about 500�m using a diamond sectioning saw (Isomet 2000-Buehler UK LTD, Lake Bluff, USA). The sections were then ground to a thickness of approximately 100�m. After 48 hours of imbibition in deionised water, the sections were examined and photographed using polarized light microscopy (Axiophot-ZEISS DSM-940 A, Oberkochen, Germany). Readings were taken at the R1 and R2 regions, around the occlusal and cervical margins of the restoration, along the interface between the dental tissue and the restoration (P1) and at 100�m (P2), 200�m (P3) and 300�m (P4) from the interface between the Table 2 Groups according to treatment, time and materials Group Number of specimens Pre-treatment Time Material 1 10 Enamel/Dentin 15 sec Prime&Bond 37% Phosphoric Acid (gel) TPH 2 10 Enamel/Dentin 15 sec Fuji II LC G.C. Conditioner/Poliacrilic Acid 3 10 Enamel/Dentin 15 sec Tetric 37% Phosphoric Acid (gel) 4 10 Enamel/Dentin 30 sec Prime & Bond Dyract Primer PSA Dyract 5 10 Enamel/Dentin 60 sec GS-80 1.23% APF 6 10 Enamel/Dentin 15 sec Chelon Fil 25% Poliacrilic Acid 124 dental tissue and the restoration. Demineral- ization was determined by depth (Fig. 1). Results Mean depths of demineralization were cal- culated (Fig. 1) for each group and position (Table 3). Application of the Tukey test at the 5% level, as shown in Table 3, revealed that interaction between group 6 and any of the positions analyzed clearly demonstrated a higher resis- tance to the development of demineralization, since the interactions observed presented the lowest mean depths of demineralization on the enamel surface for all positions investi- gated. Groups 2 and 4, where fluoride release materials were used, showed the strongest effect in the P1 position. However, no effect was observed in Group 1 in any position, with this group exhibiting the highest mean depth of demineralization. Furthermore, P1 was deeper than P4 in Group 1, which was not the case in the other groups. Discussion Control of carious lesions is mainly related to the presence of fluoride in the oral cavity, and may not be considered as directly depen- dent on the amount incorporated by the tooth6,17), since the main mechanism of action of fluoride is dynamic, inhibiting demineral- ization and enhancing remineralization4,19). Therefore, the constant presence of fluoride in the oral cavity, in saliva or oral fluids, dental plaque and enamel, may control or even inhibit the appearance of secondary carious lesions, as well as lead to arrest of existing lesions, preventing progression of incipient lesions to formation of cavity6,22). It should be noted that the model of caries development adopted in this in vitro study is similar to that advocated by Featherstone et al.4), which assumes a correlation with the onset and progression of carious lesions in vivo where there is a high risk of caries. However, in terms of the cariogenic chal- lenge employed in the present study, besides utilization of pH cycling, as suggested by Featherstone et al.4), thermal cycling was also Table 3 Mean values (standard deviation) of demineralization depth (�m) according each positions Groups 0�m (P1) 100�m (P2) 200�m (P3) 300�m (P4) 1 190.5 (65.21)a 176.5 (43.72)a 173.0 (40.84)a 158.0 (31.02)a 2 44.5 (36.76)c 77.5 (34.99)d 95.5 (25.19)c 96.5 (23.95)b 3 126.0 (21.32)b 132.0 (24.06)bc 136.0 (24.59)ab 136.0 (24.36)a 4 69.5 (31.04)c 102.0 (20.17)cd 115.5 (14.62)bc 121.0 (13.08)ab 5 147.0 (39.33)b 154.5 (37.66)ab 158.0 (25.76)a 157.5 (21.99)a 6 18.5 (30.37)d 32.5 (38.17)e 38.6 (39.98)d 40.1 (41.41)c Minimum significant difference at the 5% level. Means with different letters indicate difference between means. Fig. 1 Positions P1, P2, P3, P4 in terms of demineralization/remineralization development in regions R1 and R2 Okida RC et al. 125 conducted to approach the real conditions of the oral cavity. Polarized light microscopy revealed the depth of demineralization in the dental enamel. Table 3 shows the mean values for group and position. The values presented in Table 1 represent interaction at the P1 posi- tion (interface) in each group, and show that the material employed in Group 6 (CHELON FIL) performed best in control of demineral- ization, followed by the materials in Groups 2 (FUJI II LC), 4 (DYRACT), 3 (TETRIC) and 5 (fluoride solution was applied as pre- treatment before amalgam restoration), and finally Group 1 (TPH), which exhibited the worst performance. Similarly, as shown in Table 3, interaction in each group at positions P2, P3 and P4 demon- strated the same performance as the material employed in Group 6 (CHELON FIL) for con- trol of demineralization. However, the other groups revealed a tendency toward similar outcomes in terms of cariostatic potential, proportional to the distance from the tooth/ restoration interface25). These findings fur- ther reinforce the belief that the amount of fluoride present in the material, as well as its concentration and release, are important aspects in the reduction of demineralization. This leads to the assumption that the cariostatic property of CHELON FIL (Group 6) may be explained by the intensive fluoride release of this material17,19,25), and the amount of fluoride released depends on its concentra- tion in the material; in addition, ionic fluoride is present in this material, which favors its release17). This is corroborated by the mean values observed in this study, which demonstrated that the material employed in Group 6 (CHELON FIL), with high fluoride release and bonding to the tooth structure, pre- sented a better performance for control of demineralization. Furthermore, these results for Group 6 (CHELON FIL) (Fig. 2) are in agreement with the findings of Hicks et al.8) Fig. 2 Longitudinal sections of enamel-resin interface (A) Photomicrograph (40�) of interface of one specimen from Chelon Fil restoration illustrating integrity of restoration surface. (B) Photomicrograph (40�) of Dyract restoration illustrating marginal caries formation (P2). (C) Photomicrograph (40�) of GS-80 restoration illustrating marginal caries formation. (E) Enamel. Secondary Caries around Restoration 126 ride by the tooth or even reduce the mineral loss close to restorations; thus the present out- comes are in agreement with those of Pin et al.19). Group 5 (GS 80-Fig. 2) revealed that the amalgam did not totally inhibit demineraliza- tion, being similar to the performance pre- sented by FUJI II LC (Group 2), DYRACT (Group 4) and TETRIC (Group 3). Similarly, the results demonstrated that the ability of topical fluoride application before insertion of amalgam was not significant enough to prevent the occurrence of caries when com- pared to CHELON FIL (Group 6). Within this context, Pimenta et al.18) (1998) also evaluated the effectiveness of applica- tion of acidulated phosphate fluoride before accomplishment of amalgam restorations compared to the bonded amalgam technique and observed that application of this solution in the cavity was unable to effectively inhibit demineralization, yet reduced its formation at the tooth/restoration interface. Group 1 (TPH) revealed higher values of depth of demineralization. Since it repre- sented the control group, its performance is in agreement with the findings of Purton and Rodda21), who observed that non-fluoridated composite resins do not present any potential to inhibit demineralization. Another factor, besides fluoride release, should be considered. Thermal cycling can induce interface degradation in materials that use adhesive technique12). This process can increase microleakage with some factors. The principal factor is the difference in the thermal expansion coefficient between restor- ative material and tooth. This difference can overload the interface during thermal cycling and lead to gap formation12). Therefore, Groups 1, 3 and 4 may have been influenced by thermal cycling, principally with composite resin, which showed a deeper P1 than P4 (Table 3). Possibly, the fluoride release capac- ity contributed to reduced demineralization in Groups 3 and 4 on P1 distance when they are compared to the Group 1 (Table 3). The present findings agree with Ten Cate24) (1990), who stated that enamel remineraliza- and Purton & Rodda21) in studies using polar- ized light microscopy which revealed forma- tion of mild demineralization and that the establishment and progression of demineral- ization in these cases was reduced, possibly due to the precipitation of calcium and phos- phate triggered by the high fluoride release of these materials. This may be explained by the intensive fluoride release of this material, which depends on its concentration in the material and especially on the presence of ionic fluoride, which enhances its release. In Group 2 (FUJI II LC), its performance was inferior to that in Group 6 (CHELON FIL), thereby demonstrating that resin-modified glass ionomers also present an anticariogenic action, but are inferior when compared to conventional glass ionomer. Therefore, the results obtained in this study for Group 2 (FUJI II LC) are in agreement with the findings of Croll et al.2) and Pin et al.19), who achieved similar results, observing a signifi- cant reduction in demineralization, assigned to the significant fluoride release of this mate- rial11), because of the spontaneous acid-base reaction, which leaves free fluoride ions in the bulk to be released. The performance presented by the poly- acid-modified resin composite in Group 4 (DYRACT-Fig. 2) also demonstrated a moder- ate ability to inhibit demineralization, which may be explained by the different fluoride release of polyacid-modified resin composite compared to conventional glass ionomers, or even to resin-modified glass ionomers3,22,23). According to the manufacturer, curing of the material with posterior water absorption is required for the occurrence of fluoride release, since it favors an acid-base reaction and fluoride ion release23). However, this late acid-base reaction may considerably limit fluoride release, even in the demineralization process. The material TETRIC in Group 3 demon- strated that composites are not effective in caries inhibition19). Even though the fluori- dated resins currently available present fluo- ride release, they do not maintain this pattern to favor a considerable incorporation of fluo- Okida RC et al. 127 tion requires the presence of partially dem- ineralized hydroxyapatite crystals and is dependent on the degree of saturation of the area. This also corroborates the results found for material CHELON FIL (Group 6), which may be related to its higher fluoride release compared to the other materials, presenting a more effective action in the process of inhibi- tion and/or progression of demineralization. It should be noted that materials presenting low fluoride release, such as FUJI II LC and DYRACT (Groups 2 and 4), were not effective at distant sites, but were effective on the tooth/restoration interface, confirming that efficiency for control of distant lesions requires topical fluoride application and utilization of fluoridated mouth rinses and dentifrices, which allow the constant presence of low concentrations of fluoride in the oral cavity, which is more effective than fluoride release by materials. These results demonstrated that control of demineralization depends on the ability of materials to release fluoride ions; however, the clinical individuality of each patient should be considered for indication and application of a material or restorative technique. Conclusion The present results indicate the following: (1) Conventional glass ionomer was more effective against progression of demineral- ization on the enamel surface at all dis- tances and depths analyzed when compared to fluoridated resin materials such as resin modified glass ionomer and polyacid modified composite resin, which were only effective in the initial position (P1). (2) The highest mean depths of demineraliza- tion were observed after utilization of non- fluoridated resin material, demonstrating the inefficiency of its potential for inhibi- tion of demineralization. (3) Application of acidulated phosphate fluo- ride solution on the cavity before insertion of amalgam was unable to effectively inhibit the occurrence of demineralization. References 1) Attar N, Onen A (2002) Fluoride release and uptake characteristics of aesthetic restorative materials. J Oral Rehabil 29:791–798. 2) Croll TP, Helpin ML, Donly KJ (2000) Vitremer restorative cement for children: three clinicians’ observations in three pediat- ric dental practices. ASDC J Dent Child 67: 374, 391–398. 3) de Araujo FB, Garcia-Godoy F, Cury JA, Conceicao EN (1996) Fluoride release from fluoride-containing materials. Oper Dent 21: 185–190. 4) Featherstone JDB, O’Really MM, Shariati M, Brugler S (1986) Enhancement of reminerali- zation in vitro and in vivo; Factors Relating to Demineralization and Remineralization of the Teeth, Leach SA, 1st ed., pp.23–34, Oxford. 5) Fontana M, Dunipace AJ, Gregory RL, Noblitt TW, Li Y, Park KK, Stookey GK (1996) An in vitro microbial model for studying secondary caries formation. Caries Res 30:112–118. 6) Fontana M, González-Cabezas C, Haider A, Stookey GK (2002) Inhibition of secondary caries lesion progression using fluoride var- nish. Caries Res 36:129–135. 7) Hickel R, Manhart J (2001) Longevity of resto- rations in posterior teeth and reasons for fail- ure. J Adhes Dent 3:45–64. 8) Hicks MJ, Flaitz CM, Silverstone LM (1986) Secondary caries formation in vitro around glass ionomer restorations. Quintessence Int 17:527–532. 9) Kidd EA, Beighton D (1996) Prediction of sec- ondary caries around tooth-colored restora- tions: a clinical and microbiological study. J Dent Res 75:1942–1946. 10) Kidd EA, Toffenetti F, Mjor IA (1992) Second- ary caries. Int Dent J 42:127–138. 11) Kotsanos N (2001) An intraoral study of caries induced on enamel in contact with fluoride- releasing restorative materials. Caries Res 35: 200–204. 12) Mitsui FH, Peris AR, Cavalcanti AN, Marchi GM, Pimenta LA (2006) Influence of thermal and mechanical load cycling on microtensile bond strengths of total and self-etching adhe- sive systems. Oper Dent 31:240–247. 13) Mjor IA, Toffenetti F (2000) Secondary caries: a literature review with case reports. Quintes- sence Int 31:165–179. 14) Nakabayashi N (1985) Bonding of restorative materials to dentine: the present status in Japan. Int Dent J 35:145–154. 15) Okuda M, Pereira PN, Nakajima M, Tagami J, Pashley DH (2002) Long-term durability of resin dentin interface: nanoleakage vs. micro- Secondary Caries around Restoration 128 tensile bond strength. Oper Dent 27:289–296. 16) Okuda M, Pereira PN, Nikaido T, Tagami J (2003) Evaluation of in vitro secondary caries using confocal laser scanning microscope and X-ray analytical microscope. Am J Dent 16: 191–196. 17) Pereira PN, Inokoshi S, Tagami J (1998) In vitro secondary caries inhibition around fluo- ride releasing materials. J Dent 26:505–510. 18) Pimenta LA, Fontana UF, Cury JA, Serra MC, Elderton RJ (1998) Inhibition of demineral- ization in vitro around amalgam restorations. Quintessence Int 29:363–367. 19) Pin ML, Abdo RC, Machado MA, da Silva SM, Pavarini A, Marta SN (2005) In vitro evaluation of the cariostatic action of esthetic restorative materials in bovine teeth under severe cari- ogenic challenge. Oper Dent 30:368–375. 20) Prati C, Chersoni S, Suppa P, Breschi L (2003) Resistance of marginal enamel to acid solu- bility is influenced by restorative systems: an in vitro scanning electron microscopic study. Clin Oral Investig 7:86–91. 21) Purton DG, Rodda JC (1988) Artificial caries around restorations in roots. J Dent Res 67: 817–821. 22) Serra MC, Cury JA (1992) The in vitro effect of glass-ionomer cement restoration on enamel subjected to a demineralization and reminerali- zation model. Quintessence Int 23:143–147. 23) Suljak JP, Hatibovic-Kofman S (1996) A fluo- ride release-adsorption-release system applied to fluoride-releasing restorative materials. Quintessence Int 27:635–638. 24) Ten Cate JM (1990) In vitro studies on the effects of fluoride on de- and remineraliza- tion. J Dent Res 69. Spec No:614–619; discus- sion 634–636. 25) Tsuchiya S, Nikaido T, Sonoda H, Foxton RM, Tagami J (2004) Ultrastructure of the dentin- adhesive interface after acid-base challenge. J Adhes Dent 6:183–190. Reprint requests to : Prof. Ass. Dr. Ricardo Coelho Okida Department of Restorative Dentistry, Araçatuba School of Dentistry, Paulista State University, UNESP, Rua José Bonifácio 1193, CEP: 16015 050, Araçatuba, São Paulo, Brazil Fax: +55-18-3636-3349 E-mail: rcokida@foa.unesp.br Okida RC et al.