Marginal adaptation of composite restorations performed in vivo with different acid-etch restorative procedures.

243 experimental Class V restorations using a chemically cured composite resin were inserted in human third molars with the conventional acid-etch restorative procedure and with eight modifications of this technique. The teeth were extracted following an observation period of 4 months. Sections from each cavity were stained and examined for bacteria on the cavity floor. Bacteria in more than 10% of the sections was used as the criterion for marginal leakage along a restoration. Results showed that non-composite resin applied as an intermediary layer in etched cavities or as a surface coating of immediately finished and re-etched restorations, beveling of the margins of preparations, and treatment of the etched cavities with the surface-active comonomer NPG-GMA before filling all significantly decrease the leakage occurrence in the oral environment. The observed differences between the modifications varied considerably but none of them was significant, and no additive effects were found.