Objectives: This study aimed to determine whether surface wear, identified through the superimposition of intraoral scans (IOS), can predict subsurface damage progression detected by optical coherence tomography (OCT) during fatigue testing of computer-aided design/computer-aided manufacturing (CAD/CAM) composite crowns. Methods: Monolithic CAD/CAM composite crowns (Brilliant Crios; $n=8$) were adhesively luted to standardized prepared human teeth and artificially aged by cyclic loading in a mouth-motion simulator (50–500 N, 2 Hz, 37 °C). Under phantom-head condition, IOS (surface wear) and handheld swept-source (SS)-OCT (subsurface damage) were performed before loading and after every 250,000 cycles. OCT crack depth/width were normalized to local thickness and cusp-tip distance; correspondence between IOS- and OCT-derived metrics at each timepoint was assessed with Spearman’s rank correlation coefficient ($\rho$) to test whether surface wear can predict subsurface damage under the given conditions. Results: All specimens survived without catastrophic failure, and both modalities revealed progressive damage from the earliest observation interval. OCT consistently showed higher defect percentages and larger dispersion (e.g., mean vertical defects (25.47 ± 4.97)% OCT vs. (4.36 ± 0.91)% IOS at T1 and (66.79 ± 19.53)% OCT vs. (7.78 ± 3.19)% IOS at T5). Across all timepoints, no statistically significant associations between IOS and OCT were observed (p = 0.146 to 0.955). Conclusions: Within the limitations of this exploratory, single-material in vitro study, restricted to a CAD/CAM composite (Brilliant Crios), surface-based monitoring alone did not reliably reflect subsurface damage progression. Clinically, this suggests that surface wear assessment may underestimate subsurface fatigue damage. Intraoral OCT may provide complementary, non-invasive information alongside routine IOS for individualized monitoring, but its added value needs to be confirmed in larger studies and other CAD/CAM composite materials and additional restorative material classes.

This study extended a previously established intraoral scanning (IOS) and optical coherence tomography (OCT) dual-modality monitoring workflow for computer-aided design/computer-aided manufacturing (CAD/CAM) restorations to three additional crown material classes alongside a resin composite (RECO) reference. Four material classes were investigated ($n=8$ each): RECO, polymer-infiltrated ceramic network (PICN), lithium disilicate ceramic (LDSC), and zirconia-reinforced lithium silicate ceramic (ZLSC). Monolithic crowns were adhesively luted to standardized human molar abutment teeth and aged by cyclic loading (50500$\mathrm{N}$, 2$\mathrm{Hz}$, 37 2${}^{\circ }\mathrm{C}$, up to 1250000 cycles) in a mouth-motion simulator. IOS and handheld OCT were performed at baseline and after every 250000 cycles under phantom-head conditions; correspondence was assessed using Spearman’s rank correlation coefficient (exploratory, uncorrected for multiple comparisons). OCT consistently showed higher defect extents than IOS across all material classes and timepoints. While no significant IOS-OCT associations were found for RECO and the PICN, OCT detected full-thickness vertical subsurface damage propagation from the earliest timepoint in LDSC and ZLSC, with IOS-derived surface wear remaining markedly lower. Surface-based monitoring alone did not reliably reflect subsurface damage propagation, a dissociation most pronounced in the vertical dimension and silicate-based materials. Intraoral OCT may provide complementary, non-invasive subsurface information to support individualized recall scheduling and minimally invasive repair decisions.