Safety of Rapid Local Ischemic Postconditioning After Thrombectomy in Acute Stroke: A Dose-Finding Trial (RAPID SAVE)

BACKGROUND: Rapid local ischemic postconditioning may protect the brain after acute ischemic stroke, but its safety and optimal dosing in successfully reperfused patients after mechanical thrombectomy remain undefined. METHODS: This investigator-initiated, prospective, adaptive, multicenter phase I single-arm dose-finding trial employed a Bayesian Optimal Interval (Bayesian Optimal Interval Phase I/II) design. Patients with anterior circulation large-vessel occlusion and modified Thrombolysis in Cerebral Infarction 2b/3 reperfusion were enrolled without randomization. Within 5 minutes of recanalization, rapid local ischemic postconditioning was delivered via a balloon-guiding catheter positioned at the ipsilateral C1-intracranial internal carotid artery, alternating inflation/deflation to interrupt antegrade flow. Six dose levels were prespecified by inflation/deflation durations and cycles: 15/15s ×5; 1/1, 2/2, 3/3, 4/4, and 5/5 minutes ×4. The dose-limiting toxicity (including malignant infarction, procedure-related complications requiring treatment, or other procedure-attributable serious adverse events) threshold was 15%. The efficacy target (absence of infarct growth >10 mL at 72 hours) was 60%. Doses were eliminated if the posterior probability that toxicity exceeded 15% was ≥0.95 or efficacy <60% was ≥0.90. The dose with the highest utility meeting these criteria was selected. RESULTS: Five cohorts (n=25, 5 each) were enrolled. Four cohorts received 2/2 minutes×4 (n=20): 14 met the efficacy end point (posterior probability true efficacy <60%, ≈0.15), and 1 had a dose-limiting toxicity due to large infarction growth (probability true toxicity >15%, ≈0.16). One cohort received 3/3 minutes×4 (n=5): 3 met the efficacy end point (probability true efficacy <60%, ≈0.31) and 2 had dose-limiting toxicities due to large infarction growth (probability true toxicity >15%, ≈0.95). This triggered the predefined safety rule, preventing further testing at 3/3-minute and higher doses. Bayesian Optimal Interval Phase I/II selected 2/2 minutes×4 as the optimal regimen with a favorable efficacy-toxicity profile. CONCLUSIONS: Rapid local ischemic postconditioning initiated immediately after thrombectomy was feasible. The 2/2 minutes×4 regimen met prespecified safety and efficacy thresholds and warrants evaluation in a larger, definitive trial. REGISTRATION: URL: https://www.clinicaltrials.gov; Unique identifier: NCT06526429.