Abstract Body: Introduction: Reduction in infarct growth is shown to mediate reperfusion effect in patients with small strokes. However, exploration of reperfusion benefit and its mediation through infarct growth among patients with large strokes is limited.

Methods: In a secondary analysis of SELECT2 study, we employed a mediation analysis framework to quantify infarct growth mediation effect on successful reperfusion (mTICI≥2b) association with independent ambulation (mRS 0-3 at 90-day follow-up) using Causal Mediation approach.

Results: Of 180 patients who received EVT, 144 (80%) patients achieved successful reperfusion. Baseline clinical and imaging characteristics were largely similar between those achieving successful vs unsuccessful reperfusion. Median (IQR) infarct growth was 60 (20-128) ml among those with successful reperfusion and 99.5 (85-151.5) ml among those with unsuccessful reperfusion. Successful reperfusion was associated with reductions in infarct growth (adj β, −35.3; 95%CI, −64.6 to −6.0; p=0.018). Controlled direct effect – odds of independent ambulation with successful reperfusion, accounting for change in infarct growth was 2.63 (95%CI:0.93-9.24), whereas natural indirect effect – odds of independent ambulation with average reduction in infarct growth due to successful reperfusion was 1.49 (95% CI:1.09 to 2.47), resulting in total effect (controlled direct effect + natural indirect effect –odds of independent ambulation with cumulative effect of successful reperfusion and reduction in infarct growth associated with successful reperfusion) of 3.92 (95% CI:1.31 to 13.39). Reduction in Infarct growth associated with successful reperfusion explained 29% of cumulative effect of successful reperfusion (controlled direct effect of successful reperfusion and indirect effect - reduction in infarct growth due to successful reperfusion).

Conclusions: In this mediation analysis from SELECT2 trial, only a proportion (~29%) of successful reperfusion effect on functional outcome was mediated through reduction in infarct growth, suggesting a higher proportion of effect mediated through other possible mechanisms. Additional analyses are required to both confirm our findings, detect additional mediators to explain successful reperfusion residual effect and identify novel therapeutic targets to further enhance reperfusion benefits.