One-hour pulsatility index as an independent predictor of futile recanalization after successful mechanical thrombectomy: A prospective observational study

Futile recanalization (FR) - poor functional outcome despite successful mechanical thrombectomy - affects approximately 40% of patients with anterior circulation large vessel occlusion. Reliable bedside predictors of FR at the time of recanalization remain lacking.

This single-center prospective observational study enrolled 122 consecutive patients with successful thrombectomy (mTICI ≥2b; all mTICI 2b cases corresponded to eTICI 2b67) between 1 October 2025 and 12 April 2026. Pre-procedural optic nerve sheath diameter (ONSD) was measured by ultrasound, and post-procedural transcranial color-coded Doppler measurements were obtained by an investigator who was not involved in the endovascular procedure and was blinded to the thrombectomy technique and angiographic reperfusion findings. FR was defined as 90-day mRS 3-6. Discriminative performance was assessed by ROC analysis with DeLong comparison; independent predictors by multivariable logistic regression with bootstrap validation (5000 resamples).

FR occurred in 48 patients (39.3%). The 1-hour symptomatic-hemisphere pulsatility index (PI) was markedly elevated in FR patients (1.23 vs. 0.81; p < 0.001; AUC 0.922, 95% CI 0.861-0.984). At a threshold > 0.99, sensitivity was 91.7% (95% CI: 80.4-96.7%) and specificity 89.2% (95% CI: 80.1-94.4%).Preoperative ONSD was also higher in FR patients (5.2 vs. 4.7 mm; p < 0.001; AUC 0.785). The combined model reached AUC 0.934, without significant incremental gain over PI alone (p = 0.062). On multivariable analysis, 1-hour PI (OR 2.929; p < 0.001) and admission NIHSS (OR 1.457; p = 0.009) were independent predictors. Predictive accuracy was preserved in the local anesthesia subgroup (AUC 0.895).

The 1-hour symptomatic PI was independently associated with futile recanalization and showed high discriminatory performance after thrombectomy. Preoperative ONSD may provide complementary risk stratification. Multicenter external validation is warranted.