AI-driven virtual histology with optical coherence tomography for comparing acute and short-trem performance of bioresorbable scaffolds and drug-eluting stents

Incomplete stent expansion is an important mechanism leading to adverse events. However, few detailed OCT studies have compared the immediate performance of bioresorbable scaffolds (BRS) and drug-eluting stents (DES). The aim of this study was to compare the acute and short-term performance differences between polymeric BRS and metallic DES, and to explore the underlying mechanisms of immediate stent underexpansion and eccentricity reduction. Our study consecutively enrolled 298 patients who underwent optical coherence tomography (OCT) pullbacks after BRS or DES implantation, with 178 patients in the BRS group and 120 patients in the DES group. To reduce confounding, propensity score matching generated 64 balanced pairs of patients. The core laboratory used AI-driven virtual histology technology to measure immediate vascular and plaque parameters. Clinical follow-up was conducted for 1 year post-procedure, and major adverse cardiac events (MACE) were recorded. After propensity score matching, the BRS group had similar luminal and stent parameters to the DES group, with lower mean and minimum stent eccentricity indices (SEC). Furthermore, the BRS and DES groups had similar characteristics of fibrous plaque (FP) and calcific plaque (CP). More severe CP and FP restricted adequate expansion of BRS and DES. The incidence of 1-year MACE did not differ significantly between the BRS and DES groups (log-rank p = 0.685). The BRS appears to have comparable acute performance and reduced SEC when compared to DES. The stent expansion and eccentricity were associated with the characteristics of CP and FP. BRS and DES showed no significant difference at 1-year follow-up.