Effect of different impression techniques on marginal integrity of CAD-CAM milled all-on-four mandibular frameworks: an in vitro study.

Background: To guarantee a passive fit, full arch implant supported prostheses require scrupulous impressions. The accuracy of conventional and digital impressions is still up for debate, despite several studies comparing both acquisition techniques. The present study aimed to compare mandibular full arch implant impressions by assessing the vertical misfit of implant supported frameworks obtained through conventional and digital impressions.

Methods: To simulate the "All-on-4" scenario, a completely edentulous epoxy mandibular reference model was prepared with the installment of two straight implants in the anterior region and two 30-degree angled implants in the posterior region. Two acquisition techniques were evaluated: the conventional impression technique (CI group, n = 11) with open tray splinted impression copings using vinyl siloxane ether (VSXE) impression and the digital impression technique (DI group, n = 11) using Medit i-700 intraoral scanner (IOS). To create virtual models, the Medit T-Series laboratory scanner was used to scan the models created by the CI group. Scans obtained from both groups were saved as STL files for framework design. Screw retained bars (n = 22) were designed on the virtual models and then machined in cobalt chromium. The frameworks fabricated using both impressions were screwed to the reference model, evaluated using the Sheffield test, and the vertical misfits were analyzed under a stereomicroscope at 80× magnification. Comparisons between the two study groups were performed using independent samples t-test, and the average vertical misfits of each multi-unit abutment in each group were compared by using the ANOVA test followed by a Post Hoc test (adjusted Bonferroni) for pairwise comparison. At P <.05, statistical significance was assessed.

Results: When tightening the screw at multi-unit abutment #45, the vertical misfits of the frameworks manufactured by DI group (82.34 ± 5.05 μm) were lower than those of the CI group (91.09 ± 6.29 μm) with significant difference at P =.002, while no statistical significant difference was reported in the average vertical misfit between the CI group (43.60 ± 11.93 μm) and the DI group (43.90 ± 5.31 μm) (P =.940) while securing the screw at multi-unit abutment #35.

Conclusions: Achieving a passive fit for implant supported frameworks in completely edentulous patients is quite challenging. A fully digital workflow offers a steadfast alternative to conventional methods with vertical misfits that differ based on the impression technique, though these differences are typically not statistically significant.