Background: While endovascular therapy (EVT) is effective for large vessel occlusions (LVO), most patients present to hospitals without EVT capability and are transferred for intervention, delaying treatment.
Objective: We evaluated outcomes in LVO patients treated with thrombectomy who were transferred compared to those presenting directly to EVT facilities.
Methods: In a large multicenter international prospective cohort study of LVO patients treated with stent retriever thrombectomy (TREVO Registry), patients were stratified by initial presentation into transferred (TNS) vs direct (DIR). 90 day mRS was the primary outcome (0-1 excellent, 0-2 good outcomes); sICH and reperfusion by mTICI were secondary outcomes. Outcomes were compared in the 2 groups (0-8 hrs onset to groin puncture (GP) then in time matched 3-8 hrs subgroups for validation). Logistic regression identified independent predictors of good outcome in TNS patients.
Results: We identified 540 patients (230 TNS; 310 DIR) (Fig 1). TNS patients were younger and had longer onset to GP times (4.6 vs 3.1 hrs; p<0.001) (Table 1). DIR achieved higher excellent outcomes (50.4 vs 38.7%; p<0.001) (Table 2). There were no significant differences in good clinical outcomes (61 DIR vs 57.4% TNS, OR 0.90, 95% CI 0.63-1.27; p=0.4) (Fig 2) and no difference in the time matched 3-8 hrs subgroups (59.2% DIR vs 56.3% TNS, p=0.6). Fig 3 plots good outcome probabilities over time, showing similar confidence interval bands. Younger age (OR 0.95), lower NIHSS (OR 0.90), glucose level < 170 mg/dL (OR 2.4), distal clot location (M2) (OR 1.7), excellent reperfusion (mTICI≥2b) (OR 2) and time to GP <5 hrs (OR 1.6) were independent predictors of good outcome in TNS patients.
Conclusion: While excellent outcomes were higher in directly-presenting patients, EVT-treated transfers may achieve similar good outcomes. The association between earlier EVT after transfer and better outcomes emphasizes the need to streamline the transfer process.