Microleakage of Class 5 composite resin restorations: a comparison between in vivo and in vitro

The purpose of this study was to determine if a difference exists in microleakage between in vivo and in vitro class 5 composite resin restorations using two variations of a bonding resin (Universal Bond 2 and Universal Bond 3). Class 5 cavities were prepared in 24 matched pairs of teeth on the buccal and lingual surfaces. One tooth of each pair was prepared and restored in vivo and the other in vitro. After preparation and enamel etching, Universal Bond 2 was randomly applied to one surface of each tooth in the pair and Universal Bond 3 to the other surface. Composite resin (Prisma AP.H) was placed in each preparation, light cured, and finished using a standard technique. The in vivo samples were extracted approximately 6 weeks after placement of the restorations. The in vitro samples were thermocycled for 540 cycles (5 to 55 degrees C/1-minute dwell time). All teeth were stained with silver nitrate and longitudinally divided. The teeth were scored on a ranking system of 0 being no leakage to 4 being leakage to the pulpal wall of the preparation. A Wilcoxon matched pairs signed rank test was performed on the data (P < or = 0.05). There were no statistically significant differences in microleakage between restorations using Universal Bond 2 and Universal Bond 3. There were statistically significant differences when comparing the in vivo and in vitro restorations. More microleakage occurred in the in vitro restorations.