The effect of internal bevel on marginal leakage at the approximal surface of Class 2 composite restorations.

The aim of this study was to assess the effect of intentionally leaving undermined enamel (internal bevel) along the cervical margins of class 2 composites on marginal leakage. Conventional MO and DO cavity preparations were prepared in 25 extracted permanent premolars. In each tooth the cervical margin was a butt-joint for the control group and an internal bevel for the experimental group. A transparent celluloid matrix was adapted and the teeth were restored with scotchbond Multi-Purpose and increments of Z100. A U-shaped increment was attached to the buccal and lingual walls and cervical floor of the box, leaving a gap for a middle increment. A third increment filled the occlusal part of the cavity. Following thermocycling the teeth were immersed in basic fuchsin for 24 hours, and 0.5 mm-wide sections were cut in a mesiodistal direction. Dye penetration was scored: 0 = no penetration; 1 = dye along enamel tooth interface; 2 = dye along the gingival floor; 3 = dye along pulpal wall; 4 = dye penetration into dentinal tubules. The experimental group showed 17 teeth (68%) with no dye penetration, compared to eight (32%) in the control group. Severe dye penetration (score 4) was found in only four teeth (16%) of the experimental group and 12 (48%) of the controls. The difference between the groups was statistically significant (Wilcoxon matched-pairs signed-rank test; P < 0.01). This study showed that an internal bevel at the cervical margins of class 2 composite restorations reduces marginal leakage when compared to a conventional butt-joint cavity preparation.