Trueness and marginal fit of implant-supported complete-arch fixed prosthesis frameworks made of high-performance polymers and titanium: An explorative in-vitro study.

PURPOSE To investigate the trueness and marginal fit of computer-aided design-computer-aided manufactured (CAD-CAM) complete-arch implant-supported screw-retained fixed prosthesis (CAISFP) made of polyetheretherketone (PEEK), polyetherketoneketone (PEKK) and titanium (Ti) MATERIAL AND METHODS: A typodont model with four implants, their multiunit abutments (MUAs), and MUA scanbodies were digitized by using a laboratory scanner. The generated CAD was used to mill CAISFP frameworks in Ti, PEEK, or PEKK (each n = 10). The frameworks were digitized with an industrial light scanner to superimpose resulting standard tessellation language (STL) file with the CAD file. Deviations at five points at the abutment-framework interface of each of the four abutment sites (1:left first molar, 2:left canine, 3:right canine, 4:right first molar sites) were calculated (trueness). Marginal gaps were measured using the triple scan technique. A nonparametric repeated measures ANOVA by Brunner and Puri with factors being abutment location and material was performed to assess the mean deviations for trueness and mean marginal gaps, followed by Mann-Whitney or exact Wilcoxon Signed-Rank tests (alpha=.05). RESULTS Material type significantly affected the trueness (p<0.0001). PEEK had the lowest deviations (0.039 +/-0.01mm) followed by PEKK (0,049 +/-0.009mm), and Ti (0.074 +/-0.011mm). For marginal gaps, only abutment location's effect was significant (p = 0.003). Within PEKK, gaps at abutment 4 were significantly larger, compared with abutments 2 (p = 0.04) and 3 (p = 0.02). CONCLUSIONS The trueness of PEEK, PEKK, and Ti frameworks was different after milled. PEEK had the highest trueness. However, the marginal fit of the frameworks was similar and smaller than 90 µm in average. CLINICAL RELEVANCE PEEK, PEKK, and Ti complete-arch frameworks had clinically acceptable gaps and may therefore be recommended when their fit is considered. Higher trueness after milling did not result in better marginal fit.