Document resource
Introduction The magnitude of benefit reported across landmark endovascular thrombectomy (EVT) trials varies widely. While commonly attributed to procedural factors (time-to-reperfusion, advanced devices, operator experience, and others), this variability in reported effect size may primarily reflect stringent imaging-based patient selection.Methods We performed a trial-level meta-regression of 9 pivotal anterior circulation EVT randomized trials (MR CLEAN, ESCAPE, EXTEND-IA, SWIFT PRIME, REVASCAT, DAWN, DEFUSE 3, PISTE, and THRACE). Stringency in imaging-based patient selection was quantified using a structured index (Selection Enrichment Index; SEI) which incorporated infarct core thresholds, use of perfusion imaging, and collateral assessment. SEI was scored a priori from published eligibility criteria (range 0-4), with higher scores reflecting more restrictive physiologic gating. Primary outcome was odds ratio for functional independence (mRS 0-2 at 90 days).Results Reported treatment effects varied substantially across trials (OR range 1.57-6.25), and demonstrated a strong graded association with the trial’s SEI (β = 0.259 log OR per SEI point; 95% CI 0.153-0.365; weighted R 2 = 0.766). Trials with highest SEI showed up to 2.82 times higher OR, and better absolute risk reduction (β = 5.36% per increment; R2 = 0.820). Publication year alone did not significantly predict effect size (R2 = 0.234), and the SEI association persisted after controlling for year (partial R2 = 0.713). These findings were robust across all leave-one-out sensitivity analyses.Conclusions Rising effect sizes observed in landmark EVT trials were associated with more stringent imaging-based patient selection, not chronological improvement alone. Cross-trial comparisons without accounting for this selection enrichment bias may be fundamentally misleading.Disclosures R. Kamal: None. S. Amin: None. Z. Rosenstein: None. D. Sahlein: None. J. Saver: None. A. Chaudhari: None.Abstract E-031 Figure 1