Balloon guide catheter outcomes by occlusion site in the randomized PROTECT-MT trial

BACKGROUND: Balloon guide catheters (BGCs) are used during endovascular thrombectomy, but whether their effect varies by occlusion site is uncertain. We evaluated effect modification by baseline occlusion site in the PROTECT-MT trial. METHODS: This exploratory secondary analysis of the randomized PROTECT-MT trial included patients with core laboratory-adjudicated anterior circulation occlusion site data. Patients were classified as having intracranial internal carotid artery (ICA), M1, or M2 occlusion. The primary outcome was the 90-day modified Rankin Scale (mRS) distribution. Treatment effect modification was assessed using adjusted ordinal logistic regression with a treatment-by-occlusion site interaction term. The parent plan prespecified internal carotid artery (ICA) versus middle cerebral artery (MCA) occlusions; the three-level comparison and post hoc Holm correction of 19 secondary interactions were exploratory. RESULTS: Among 329 patients, 88 (26.7%) had ICA occlusions, 206 (62.6%) M1 occlusions, and 35 (10.6%) M2 occlusions. The interaction for the 90-day mRS distribution was not statistically significant (P=0.072). Adjusted common ORs with BGC versus control were 0.38 (95% CI 0.17 to 0.85) for ICA occlusions, 0.60 (95% CI 0.36 to 0.99) for M1 occlusions, and 4.32 (95% CI 0.53 to 35.02) for M2 occlusions. No consistent interaction was observed for first-pass or final reperfusion. Secondary interaction analyses remained exploratory after post hoc multiplicity correction. CONCLUSIONS: This exploratory analysis did not provide conclusive evidence that the occlusion site modified the effect of BGC use on 90-day outcome. Imprecise estimates, particularly for M2, preclude site-specific guide catheter selection. TRIAL REGISTRATION NUMBER: NCT05592054.