Penetration of radiocalcium at the margins of resin and glass ionomer dentine bonding agents in primary and permanent teeth

Journal of Dentistry Journal of Dentistry 28 (2000) 481–486 www.elsevier.com/locate/jdent

Penetration of radiocalcium at the margins of resin and glass ionomer dentine bonding agents in primary and permanent teeth ¨ . Tulunoglu a, I. Tulunoglu b,*, T. Ulusu a, Y. Genc¸ c O a

Department of Pedodontics, Faculty of Dentistry, University of Gazi, Ankara, Turkey Department of Prosthodontics, Faculty of Dentistry, University of Hacettepe, Ankara, Turkey c Department of Biostatistics, Faculty of Medicine, University of Ankara, Ankara, Turkey

b

Received 13 October 1999; accepted 17 April 2000

Abstract Objectives: The purpose of this study was to compare the leakage of three resin dentine bonding agents (Prime and Bond, Scotchbond Multi-Purpose, Probond) and a glass ionomer dentine bonding agent (GC Fujibond LC), in cervical cavities prepared in primary and permanent molar teeth restored with a hybrid composite resin (Tetric). Methods: Cervical cavities without a bevel at the cavo-surface margins were prepared in the buccal and lingual surfaces of extracted primary and permanent molar teeth. After being restored, the teeth were stored for 1 week in a saline solution at 37⬚C and then thermally cycled between 5 and 55⬚C. Marginal leakage was determined subsequently using a radioactive isotope containing 45Ca and an autoradiographic technique. Results: The results revealed that there were no statistically significant differences in microleakage of the bond between permanent and primary teeth dentine and Fuji Bond LC and Probond dentine bonding agents. The difference between permanent and primary teeth groups for gingival values of the Prime and Bond 2.1 group U ˆ 22.5, p ˆ 0.0355 and the Scotchbond Multipurpose group U ˆ 24.0, p ˆ 0.0406 were statistically significant. There were no significant differences between the occlusal and gingival microleakage values in either primary or permanent teeth with Prime and Bond 2.1, Fuji Bond LC and Probond except the difference at Scotchbond Multipurpose in primary teeth. For primary teeth gingival margins, none of the bonding systems were significantly different from the control group. Conclusions: These results indicate that although no statistically significant differences were found between test and control group values, the use of Fuji II LC in cervical cavities with cementum margins in primary teeth would provide the best resistance to microleakage among the test materials while the use of Scotchbond Multi-Purpose would provide the best resistance to microleakage in cervical cavities with cementum margins in permanent teeth. 䉷 2000 Elsevier Science Ltd. All rights reserved. Keywords: Microleakage; Dentine bonding agents; Radioisotope detection

1. Introduction It has been known for many years that conventional restorative materials and techniques produce dental restorations that do not provide a complete marginal seal [1,2]. To achieve clinical success with cervical restorations at a dentine/cementum margin was very difficult as composite resin and bonding agents could not bond directly to these tooth structures. Then, newer dentine bonding agents have therefore been introduced that, it is claimed, allow micromechanical adhesion of composite resin to dentine [3]. Several studies [3,4] have indicated that dentine bonding agents are effective in reducing marginal leakages. * Corresponding author. Tel.: ⫹90-312-324-61-45; fax: ⫹90-312-31137-41.

Investigators [5] stated that Probond dental adhesive agent was effective for reducing microleakage when used as a liner in primary teeth. Other dental adhesives may be equally effective [6]. The Probond system does not remove the smear layer and establishes a resin infiltrated smear layer. This kind of elastic layer may reduce the polymerization stresses caused by the composite [5]. Modern dentine bonding agents have evolved from the original concept of increasing dental permeability and wettability and promoting bonding to smear layer, to the partial removal of the smear layer, and finally to the use of stronger etchants to modify or remove the smear layer and obtain some form of micromechanical retention. More recently, photocured resin bonding agents have been developed (Scotchbond Multipurpose); carboxylic acid groups that became available for attachment to dentine have been incorporated [7].

0300-5712/00/$ - see front matter 䉷 2000 Elsevier Science Ltd. All rights reserved. PII: S0300-571 2(00)00029-4

¨ . Tulunoglu et al. / Journal of Dentistry 28 (2000) 481–486 O

482 Table 1 The materials used in the study Test group no.

Dentine bonding system

Manufacturer

1 2 3 4 5

Prime and Bond 2.1(P and B) Scotchbond Multi-Purpose (SBMP) Pro-Bond (PRO) Fuji bond LC (FBLC) Control

LD Caulk/Dentsply, Milford, CA 3M, St Paul, MN LD Caulk/Dentsply, Milford, CA GC Corporation, Tokyo, Japan

The bonding mechanism of the fourth generation adhesive systems is a three-step process: condition, prime and bond. Conditioning removes the smear layer, opens the dentinal tubules, increases dentinal permeability and demineralizes the intertubular and peritubular dentine. Removal of hydroxyapatite crystals leaves a collagen meshwork that can collapse and shrink because of the loss of inorganic support [8]. For optimum bonding most current bonding systems require a basic two-step surface priming and resin bonding treatment after etching of enamel/dentine, before the application of composite restorations. The development of onestep systems was directed towards the simplification of the application process of composite bonding systems. These systems combine the two-step method of priming and bonding resin into a one-step procedure that makes them simpler and quicker to use [9,10]. It has been suggested that glass ionomer cements, because of their ability to renew broken chemical bonds, have better cavity sealing properties and resistance to microleakage over long periods. In addition, because of their ability to leach fluoride, they possess some cariostatic properties [7]. The search for a material that could cushion masticatory pressures like a shock-absorber and ensure adequate protection against caries led to the development of Fuji Bond LC, the first modern glass ionomer-based, dentine-enamel adhesive. Despite the vast number of research reports on the efficacy of bonding of resin adhesives to dentine of permanent teeth, very few have addressed resin bonding to primary dentine [11]. There is not a report in the literature comparing the microleakage properties of the bonding of latest generation bonding agents to primary and permanent teeth. The purpose of this in vitro study, therefore, was to compare the effects of several resin dentine bonding agents: Probond, Scotchbond Multi Purpose Plus (SBMP), Prime and Bond and Fuji Bond LC glass ionomer bonding agents on the marginal leakage of posterior composite restorations in cervical cavities in permanent and primary teeth.

2. Methods and materials Fifty caries-free freshly extracted human mandibular molar teeth—twenty-five primary and twenty-five permanent—were stored in deionized water with a bactericidal

Composite resin filling material

Tetric

agent, 0.2% sodium azide, until they were ready for use. Tissue remnants were removed mechanically with a no. U 15-30 periodontal scaler and the teeth rinsed thoroughly under running tap water for 15 min to remove the sodium azide solution. Cervical cavities were prepared using a high-speed handpiece and water spray and a round shaped no. 35 diamond bur on both buccal and lingual surfaces of the 50 molar teeth. Each cavity preparation was 1.5 mm deep, rectangular in shape with internal line angles rounded, measuring 2 × 4 mm parallel to the cemento–enamel junction (CEJ), and the gingival half of the preparations extended 0.5 mm below the CEJ. Cavosurface walls were finished to a butt joint with no. 55 tungsten carbide bur with slow speed. Cavity preparations were rinsed for 20 s with an air/water spray and gently air dried for 30 s. The materials used are shown in Table 1. The buccal and lingual surfaces of each tooth were numbered from 1 to 100, and the treatment sequence/ scheme for each surface was determined using a number table. The dentine bonding systems Scotchbond Multipurpose Plus (3M, St Paul MN, batch no. 19951108), Prime and Bond 2.1 (LD Caulk /Dentsply, Milford, DE, batch no. 107349/2), Probond (LD Caulk/Dentsply, Milford, CA, batch no. 941019) and GC Fuji Bond LC (GC Corporation, Tokyo, Japan, batch no. 111271) were applied according to the manufacturer’s directions. At the completion of the bonding procedure, composite resin filling material (Tetric, Vivadent, Liechtenstein) was applied to the treated dentine surface. The negative control preparations were rinsed and air dried and composite resin restorations were applied directly. The composite resin was applied in 1 mm increments and light cured for 40 s per increment to ensure complete polymerization. The restoration was then finished with a finishing bur followed by polishing disks (Soft-lex, 3M Dental Products, St Paul, MN). After restoration, the teeth were stored at 37⬚C until being thermocycled from 5 to 60⬚C for 20 s in each bath and a dwell time of 10 s in a resting bath at 34⬚C for 540 cycles. The apices of the specimens were sealed with Scotchbond Multi-Purpose/Tetric. All tooth surfaces were covered with two coats of fingernail polish to within approximately 1.0 mm of the tooth-restoration margin. The procedure used to determine microleakage has been previously described [12]. The marginal leakage of each specimen was determined by the presence of an isotope at the interface

¨ . Tulunoglu et al. / Journal of Dentistry 28 (2000) 481–486 O

483

Table 2 Distribution of isotope penetration scores at the enamel (occlusal) margins of permanent teeth

Table 3 Distribution of isotope penetration scores at the cementum (gingival) margins of permanent teeth

Dentine bonding system

Dentine bonding system

Prime and Bond 2.1 SBMP Probond Fuji Bond LC Control

Microleakage values

Median

0

1

2

3

8 7 3 7 0

0 2 5 1 3

2 0 1 1 4

0 1 1 1 3

0 0 1 0 2

of the tooth structure and the restorative materials, as shown on the autoradiograph. Each specimen was immersed for 2 h in 45Ca isotope solution. The concentration of the isotope was 0.1 mCi/ml of solution in the form of calcium chloride with the pH adjusted to 7. After the specimens were removed from the isotope solution, they were brushed with detergent and cleaned under running tap water. The teeth were embedded in an autopolymerizing polymethyl methacrylate resin and sectioned longitudinally through the center of each restoration using a slow speed diamond saw (Isomet, Buehler, USA). After being washed and dried, the tooth section surfaces were placed on ultraspeed films for 17 h to produce autoradiographs. The films were processed automatically. The microleakage of the specimens was evaluated separately for the enamel and dentine margins according to the following scale: 0 ˆ no evidence that the isotope penetrated the interface of the restoration and the tooth structure; 1 ˆ evidence that the isotope penetrated the interface of the restoration and the tooth up to 1 mm; 2 ˆ evidence that the isotope penetrated the interface of the restoration and the tooth beyond 1 mm but less than the axial wall; and 3 ˆ evidence that the isotope penetrated the interface of the restoration and the tooth beyond the axial wall of the preparation. The data obtained were analyzed statistically by the Kruskal–Wallis analysis of variance, Multiple Dunn test, Mann–Whitney U-test and Wilcoxon matched pair signed rank test.

3. Results 3.1. Permanent teeth At the occlusal-enamel area, the greatest median microleakage value was obtained with the control group (2.0) and this was followed consecutively by Probond (1.0), Fuji LC (0), SBMP (0) and Prime and Bond 2.1 (0) groups (Table 2). At the gingival area, the greatest median microleakage values was obtained with the control group (2.5) and this was followed consecutively by Probond (1.5), Fuji LC (1), Prime and Bond 2.1 (1) and SBMP (1) groups (Table 3).

Prime and Bond 2.1 SBMP Pro Bond Fuji Bond LC Control

Microleakage values

Median

0

1

2

3

3 4 2 4 0

4 5 3 2 1

3 1 4 3 4

0 0 1 1 5

1 1 1.5 1 2.5

3.2. Primary teeth At the occlusal-enamel area, the greatest median microleakage value was obtained with the control group (3) and this was followed consecutively by Fuji LC (1), Prime and Bond 2.1 (1), Probond (0.5) and SBMP (0) groups (Table 4). At the gingival area, the greatest median microleakage value was obtained with the control group (3) and this was followed consecutively by Probond (2.5), Prime and Bond 2.1 (2), SBMP (2), and Fuji LC (2) groups (Table 5). The median values of test groups are shown in Fig. 1. 3.3. Statistical analysis As the bonding agent groups were independent from each other and the number of samples in each group was lower than 30 they were compared with each other separately for each of the occlusal-enamel or gingival subgroups of primary and permanent teeth groups with Kruskal–Wallis analysis of variance (p ⬍ 0.05). As the results of the Kruskal–Wallis analysis of variance were indicating significant differences (p ⬍ 0.05) all the groups (primary teeth occlusal-enamel subgroups, primary teeth gingival-cementum subgroups, permanent teeth occlusal-enamel subgroups, permanent teeth gingival-cementum subgroups) were analyzed with Dunn tests in order to detect the groups causing differences. The results of this test revealed that: for occlusal/enamel permanent teeth groups, the differences between the Prime and Bond 2.1-control, Fuji Bond LCcontrol and SBMP-control subgroups were significant (p ⬍ 0.05); for gingival/cementum permanent teeth groups the differences between the Prime and Bond 2.1-control, SBMP-control, Probond-control and Fuji LC-control subgroups were significant (p ⬍ 0.05); for occlusal/enamel primary teeth groups, the differences between the Prime and Bond 2.1-control, Fuji Bond-control, Probond-control and SBMP-control were significant (p ⬍ 0.05); and for gingival/ cementum primary teeth groups, there were no differences. As the groups were independent from each other, the differences between the values obtained from permanent and primary teeth groups for each material were evaluated with Mann–Whitney U tests. For gingival values of Prime and Bond 2.1 group (U ˆ 22.5, p ˆ 0.0355) and SBMP group (U ˆ 24.0, p ˆ 0.0406) the differences between

¨ . Tulunoglu et al. / Journal of Dentistry 28 (2000) 481–486 O

484

Table 4 Distribution of isotope penetration scores at the enamel (occlusal) margins of primary teeth

Table 5 Distribution of isotope penetration scores at the cementum (gingival) margins of primary teeth

Dentine bonding system

Dentine bonding system

Prime and Bond 2.1 SBMP Pro Bond Fuji Bond LC Control

Microleakage values

Median

0

1

2

3

4 6 5 4 0

5 1 1 2 0

0 2 1 3 3

1 1 3 1 7

1 0 0.5 1 3

permanent and primary teeth groups were found to be statistically significant. For analyzing the differences between the enamel and cementum margin values of each material in permanent and primary teeth, as the groups were dependent in nature, sign tests were used. Generally for every group, gingival microleakage values were greater than occlusal microleakage values. Test results revealed that the differences between the occlusal and gingival microleakage scores of Scotchbond Multipurpose group (p ˆ 0.016) and Prime and Bond in primary teeth were statistically significant.

4. Discussion A number of methods are used for detecting microleakage including air-pressure, radioactive tracers, dye penetration, bacterial penetration, neutron activation analysis, electrochemical analysis, silver nitrate staining, direct observation with electron microscopy, fluorimetric assays, electrical conductivity and chemical tracers [13,14]. One common method for detecting microleakage patterns has involved the use of radioactive isotopes. In general, 45Ca in the form of calcium chloride at a concentration of 0.1 Ci/ ml has been the most popular isotope to be used because it has a low energy beta emitter and does not readily penetrate enamel [15]. Although the radioactive isotope method is very technique sensitive as the choice of isotope, the distance between the source and emulsion, the length of exposure and rinsing may affect the resolution of an autoradiograph [16]. The use of isotopes permits detection of minute amounts of leakage as the smaller isotope molecule measures only 40 nm compared to the smaller dye particles (120 nm) [17]. Dentinal tubules have an average diameter of 1.65 mm [2]. In a study, Going [17] compared the effects of 45 Ca and gentian violet dye solution and showed that both indicators would pass easily through dentinal tubules to the pulp, but that at the restoration margin the isotope is more likely to penetrate than the dye. Alani [14] showed that 45Ca showed deep penetration into defects and stated that radioactive isotopes have the advantage over tracers for their presence can be readily detected even in very small concentrations. Holtan et al. [18] reported dye penetration to one-third of

Prime and Bond 2.1 SBMP Pro Bond Fuji Bond LC Control

Microleakage values

Median

0

1

2

3

0 2 0 2 0

3 2 4 2 0

5 2 1 5 1

2 4 5 1 9

2 2 2.5 2 3

the cavosurface margins of cavities treated with Scotchbond Multipurpose Plus and filled with composite resin. In another dye penetration study [19], for Scotchbond Multipurpose Plus at the occlusal margin, nine specimens showed no microleakage and one specimen showed penetration into half the distance of the axial wall; and at gingival margin seven specimens showed no microleakage and three of the specimens showed penetration into one-half the distance of the axial wall. Yap et al. [9] reported that no significant differences in dye penetration scores were observed at both enamel and dentine margins for Scotchbond Multipurpose, Prime and Bond 2.0, Fuji Bond LC groups stored for one week in 37⬚C isotonic saline. However, they stated that after 500 thermal cycles between 5 and 65⬚C for 2 s in each bath, and a dwell time of 10 s in a resting bath at 34⬚C, Fuji Bond LC showed better sealing ability at dentine margins compared to other test groups and that thermal cycling significantly decreased leakage at the dentine margins of restorations bonded with Fuji Bond LC, in permanent teeth. The relatively greater microleakage values obtained in this study could be due to the method used to detect the microleakage. In a previous study, it has been found that there were no statistically significant differences in microleakage between buccal and lingual tooth enamel or cementum surfaces using both rank order scoring or measurement scoring [20]. A large number of microleakage studies [20–22] employ both buccal and lingual surfaces of teeth in their methodology. The assumption made by the research community is that for buccal and lingual enamel, dentine and cementum respond in a similar manner with respect to microleakage. The clinical assumption is that the buccal and lingual surfaces can be prepared and restored in the same manner. The results of this study: for primary teeth cementum margins none of the bonding systems were significantly different than the control groups, shows that there are still inadequacies in the bonding systems used in this study to prevent microleakage at gingival margins of primary molars. Asmussen [23] stated that the dentine and cementum bonds are the weak link since etching the less calcified dentine does not provide the degree of mechanical bond that can be obtained in enamel. However, more recent studies [24,25] using third generation dentine bonding agents have reported predictable bond strengths to dentine.

¨ . Tulunoglu et al. / Journal of Dentistry 28 (2000) 481–486 O

485

Fig. 1. Mean microleakage scores for different bonding systems.

Bachmann et al. [26] state that, in contrast to enamel, dentine has a much more complex structure and a successful bond on dentine can only be obtained if an optimal interlocking of an adhesive system can be achieved on the dentinal surfaces. Friedl et al. [5] reported that there were no significant differences in marginal gaps between the enamel- and dentine-composite interfaces in permanent teeth for Scotchbond Multipurpose and Probond groups after 5000 thermal cycles between 5 and 55⬚C. In this study, the differences between occlusal and gingival microleakage scores were not statistically significant (p ⬎ 0.05) in both permanent and primary teeth groups after 500 thermal cycles between 5 and 65⬚C for 2 s in each bath and a dwell time of 10 s in a resting bath at 34⬚C. It has been shown that the bond strength of composite resins to the dentine surface is lower in primary teeth than in permanent teeth [27]. In a previous study, the mineral content of dentine was measured with neutron activation analysis and lower concentrations of calcium and phosphorus were measured for primary teeth than for permanent, but this difference was not statistically significant [28] In another study performed with energy dispersive spectroscopy, the concentrations of calcium and phosphorus were shown to be decreased in both peritubular and intertubular dentine of primary teeth compared with permanent teeth [29]. Significant differences were found [29,30] regarding the effect of acidic dentine conditioners on smear layer removal, suggesting that the substrate produced when primary tooth dentine is conditioned, does not reproduce the substrate found in permanent teeth. The results obtained in No¨r’s study [30] rejects the

hypothesis that primary dentine reactivity is identical to permanent dentine. The acids used to condition the dentine surface removed smear layer more rapidly from primary teeth than from permanent teeth [30]. He stated that the composition of the smear layers related directly to the composition of the underlying dentine. Considering this fact, a reasonable explanation for the differences found between primary and permanent teeth is that they may have different chemical compositions or chemical reactivity. Another possible reason for differences in smear layer removal may be related to the number of dentine tubules present. The decreased dentine permeability of primary teeth is caused by smaller tubule concentration and diameter [30]. These authors [30] have found that the hybrid layer created at the resin–dentine interface of primary teeth is thicker than that observed in permanent teeth and stated that the dentine conditioners used in their study removed the smear layer present more effectively at the dentine surface of primary teeth than at the permanent. In this study, the microleakage values were greater in primary teeth than in permanent, however the differences were not statistically significant (p ⬎ 0.05). This can be due to the fact that the use of the materials according to the manufacturers’ directions could help to effectively remove the smear layer and thus predispose more integration between bonding agent and materials.

5. Conclusion As the highest microleakage values were obtained from

486

¨ . Tulunoglu et al. / Journal of Dentistry 28 (2000) 481–486 O

control groups, the use of dentine adhesive agents decrease microleakage under composite resin restorations in either primary or permanent teeth. However, as the differences were not statistically significant in primary molar gingival microleakage between each of the four bonding systems and the control group, the inadequacies to prevent microleakage with these bonding systems in primary teeth cementum margins must be remembered. All the bonding agents showed better microleakage properties in permanent teeth compared to primary, either in occlusal or gingival areas. The results revealed that there were no statistically significant differences in microleakage of the bond between permanent and primary teeth dentine and Fuji Bond LC and Probond dentine bonding agents, except for gingival values of Prime and Bond 2.1 group (U ˆ 22.5, p ˆ 0.0355) and Scotchbond Multipurpose group (U ˆ 24.0, p ˆ 0.0406) where the difference between permanent and primary teeth groups was found to be statistically significant. In all groups, gingival microleakage values were greater than occlusal microleakage values. There were no significant differences between the occlusal and gingival microleakage values in either primary or permanent teeth with Prime and Bond 2.1, Fuji Bond LC and Probond except for the difference with Scotchbond Multipurpose in primary teeth. References [1] Going RE, Massler M, Dute HL. Marginal penetration of dental restorations by different radioactive isotopes. Journal of Dental Research 1960;39:273–84. [2] Youngson CC, Glyn-Jones JC, Manogue M, et al. In vitro dentinal penetration by tracers used in microleakage studies. International Endodontic Journal 1998;31:90–9. [3] Saunders WP, Grieve AR, Russell EM, et al. The effect of dentine bonding agents on marginal leakage of composite restorations. Journal of Oral Rehabilitation 1990;17:519–27. [4] Gillette KE, Robinson BE, Blank LW, et al. A dentine bonding agent and microleakage below the cemento-enamel junction. Journal of Dental Research 1984;63:179 (abstract no. 73). [5] Freidel KH, Schmalz G, Hiller KA, et al. Marginal adaptation of composite restorations versus hybrid ionomer/composite sandwich restorations. Operative Dentistry 1997;22:21–9. [6] Royse MC, Ott NW, Mathieu GP. Dentine adhesive superior to copal varnish in preventing microleakage in primary teeth. American Academy of Pediatric Dentistry 1996;18:440–3. [7] Mc Lean JW. Dentinal bonding agents versus glass-ionomer cements. Quintessence International 1996;27:659–67. [8] Swift EJ. Bonding systems for restorative materials—a comprehensive review. American Academy of Pediatric Dentistry 1998; 20:80–4. [9] Yap AUJ, Ho KS, Wong KM. Comparison of marginal sealing ability of new generation bonding systems. Journal of Oral Rehabilitation 1998;25:666–71. [10] Swift EJ, Perdigao J, Heymann HO. Enamel bond strengths of one-

[11]

[12]

[13] [14] [15] [16]

[17] [18] [19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

bottle adhesives. American Academy of Pediatric Dentistry 1998;20:259–62. Araujo FB, Garcia-Godoy F, Issao MA. Comparison of three resin bonding agents to primary tooth dentin. American Academy of Pediatric Dentistry 1997;19:253–7. Fitchie JG, Reeves GW, Scarbrough AR, et al. Microleakage of two new dentinal bonding systems. Quintessence International 1990;21:749–52. Taylor MS, Lynch E. Microleakage. Journal of Dentistry 1992;20:3– 10. Alani AH, Toh CG. Detection of microleakage around dental materials: a review. Operative Dentistry 1997;22:173–85. Bauer JG, Henson JL. Microleakage: A measure of the performance of direct filling materials. Operative Dentistry 1984;9:2–9. Delivanis PD, Chapman KA. Comparison and reliability of techniques for measuring leakage and marginal penetration. Oral Surgery, Oral Medicine, Oral Pathology 1982;53:410–6. Going RE. Cavity liners and dentine treatment. Journal of American Dental Association 1964;69:415–22. Holtan JR, Nystrom GP, Rensch SE, et al. Microleakage of five dentinal adhesives. Operative Dentistry 1993;19:189–93. Ferrari M, Davidson CL. Sealing capacity of a resin-modified glassionomer and resin composite placed in vivo in Class 5 restoration. Operative Dentistry 1996;21:69–72. Barnes DM, Mc Donald NJ, Thompson VP, et al. Microleakage in buccal and lingual class 5 composite restorations: a comparison. Operative Dentistry 1994;19:133–7. Mc Connell TA, Richards ND, Mixson JM, et al. Microleakage in unprepared abrasion lesions vs prepared cervical lesions. Journal of Dental Research 1991;70:388 (abstract no. 985). Rigsby DF, Retief DH, Bidez MW, et al. Microleakage of Class V restorations subjected to temperature and load cycling. Journal of Dental Research 1991;70:388 (abstract no. 981). Asmussen E, Munksgaard CC. Adhesion of restorative resins to dentinal tissues: mechanisms accomplishments and expectations. In: Vanherle G, Smith DC, editors. Posterior resin dental restorative materials, Utrecht, The Netherlands: P Szule, 1985. p. 217–29. Øilo G, Olsson S. Tensile bond strengths of dentine adhesives: a comparison of materials and methods. Dental Materials 1990;6: 138–44. Wong T, Nakabayashi N. Effect of 2-methacryl oxyethyl phenyl hydrogen phosphate on adhesion to dentine. Journal of Dental Research 1991;70:59–66. Bachmann M, Paul SJ, Lu¨thy H, et al. Effect of cleaning dentine with soap and pumice on shear bond strengths of dentine bonding agents. Journal of Oral Rehabilitation 1997;24:433–8. Bordin-Aykroyd S, Sefton J, Davies EH. In vitro bond strengths of three current dentine adhesives to primary and permanent teeth. Dental Materials 1992;8:74–8. Lakomea EL, Ryto¨maa I. Mineral composition of primary and permanent teeth in Finland. Scandinavian Journal of Dental Research 1977;85:89–95. Hirayama, A.. Experimental analytical electron microscopic studies on the quantitative analysis of elemental concentrations in biological thin specimens and its application to dental science. Shikwa Gahuko 1990;90:1019–36. In: No¨r JE, Feigal RJ, Dennison JB, Edwards CA. Dentine bonding: SEM comparison of the dentine surface in primary and permanent teeth. American Academy of Pediatric Dentistry 1997;19:246–52. No¨r JE, Feigal RJ, Dennison JB, et al. Dentine bonding: SEM comparison of the dentine surface in primary and permanent teeth. American Academy of Pediatric Dentistry 1997;19:246–52.