Vertical marginal discrepancy of posterior fixed partial denture frameworks across material-specific CAD-CAM workflows and prosthesis designs: an in vitro study

To compare vertical marginal discrepancy of posterior 3-unit fixed partial denture frameworks fabricated with cobalt–chromium (Co-Cr), zirconia, or polyetheretherketone (PEEK) computer-aided design and computer-aided manufacturing (CAD-CAM) workflows in fixed-fixed and cantilever designs. Sixty frameworks were allocated to 6 design–workflow groups ( n = 10). After standardized cementation, 3600 scanning electron microscopy measurements from 4 sites per framework were averaged. Normal-distribution generalized linear models with an identity link evaluated design, material-specific workflow, and their interaction. Pairwise comparisons used Sidak adjustment. Abutment- and surface-specific analyses were exploratory (α = 0.05). Mean overall vertical marginal discrepancies ranged from 34.17 to 57.37 μm. The workflow main effect was significant (Wald χ² = 14.258, df = 2, p < 0.001), whereas design ( p = 0.995) and the design × workflow interaction (Wald χ² = 1.354, df = 2, p = 0.508) were not significant. Within the cantilever design, Co-Cr had lower discrepancy than zirconia (mean difference = − 23.21 μm; 95% CI, − 43.06 to − 3.36; adjusted p = 0.009). No other design-stratified overall workflow contrast was significant. The cantilever Co-Cr workflow yielded smaller vertical marginal discrepancies than the zirconia workflow; a design-dependent workflow effect was not demonstrated. Group means below the contextual 150-µm benchmark do not establish clinical acceptability. Findings apply to the sampled sites and complete workflows, including geometry, and do not establish equivalence, intrinsic material superiority, or clinical suitability.