Class II resin composite restorations—tunnel vs. box-only in vitro and in vivo

PURPOSE: In a combined in vitro/in vivo approach, tunnel vs. box-only resin composite restorations should be evaluated using thermomechanical loading (TML) in vitro and a restrospective clinical trial in vivo. MATERIALS AND METHODS: For the in vitro part, box-only and tunnel cavities were prepared in 32 extracted human third molars under simulated intraoral conditions in a phantom head. Specimens were randomly assigned to four groups (n = 8; 16 box-only/16 tunnel) and received bonded resin composite restorations with Amelogen Plus (box A/tunnel A) or lining with Ultraseal and Amelogen plus (box B/tunnel B) both bonded using PQ1 (all Ultradent). Specimens were subjected to a standardized aging protocol, 1-year water storage (WS) followed by TML (100,000 × 50 N; 2500 × + 5/+ 55 °C). Initially and after aging, marginal qualities were evaluated using replicas at × 200 magnification (SEM). For the corresponding in vivo observational study, 229 patients received 673 proximal resin composite restorations. From 371 tunnel restorations, 205 cavities were filled without flowable lining (tunnel A), and 166 tunnels were restored using UltraSeal as lining (tunnel B). A total of 302 teeth received conventional box-only fillings. Restorations were examined according to modified USPHS criteria during routine recalls up to 5 years of clinical service. RESULTS: In vitro, all initial results showed 100% gap-free margins when a flowable lining was used. Tunnels without lining exhibited some proximal shortcomings already before TML and even more pronounced after TML (p < 0.05). After TML, percentages of gap-free margins dropped to 87–90% in enamel with lining and 70–79% without lining (p < 0.05). In vivo, annual failure rates for box-only were 2.2%, for tunnel A 6.1%, and for tunnel B 1.8%, respectively (p < 0.05). Tunnels had significantly more sufficient proximal contact points than box-only restorations (p < 0.05). Flowable lining was highly beneficial for clinical outcome of tunnel-restorations (p < 0.05). CONCLUSIONS: With a flowable lining, tunnel restorations proved to be a good alternative to box-only resin composite restorations. CLINICAL RELEVANCE: Class II tunnel restorations showed to be a viable alternative for box-only restorations, however, only when flowable resin composite was used as adaptation promotor for areas being difficult to access.