Selection criteria for endovascular therapy for acute ischaemic stroke: Are patients missing out?

Endovascular clot retrieval (ECR) following intravenous thrombolysis is superior to intravenous thrombolysis alone for acute stroke with large vessel occlusion. However, trial selection criteria may exclude potentially salvageable patients. We investigated the impact of published selection criteria on the different proportions of patients excluded and clinical outcome.

We included patients with anterior circulation stroke treated with ECR from a single centre. Selection criteria from five trials (REVASCAT, EXTEND IA, MR CLEAN, SWIFT PRIME, ESCAPE) and American Stroke Association (ASA) guidelines were applied. We calculated the proportion of patient's ineligible for ECR according to different selection criteria. Clinical benefit and harm were quantified as the number of patients benefiting per 1 patient harmed (NB1H) for each of the 6 applied selection criteria.

One hundred and seventy-eight patients were included. Mean age was 74 (SD 14) years, 60.1% were male, median baseline NIHSS was 17 (IQR 13-21). Patients were hypothetically excluded from ECR: REVASCAT 35.4%, EXTENDA IA 86%, SWIFT PRIME 86%, MR CLEAN 2.3%, ESCAPE 93.3% and ASA 29.2%. The NB1H for included and excluded patients respectively in decreasing order of magnitude: EXTEND IA >100 vs 3, ESCAPE >100 vs 3.4, SWIFT PRIME 10 vs 3.3, REVASCAT 4.4 vs 2.9, MR CLEAN 3.7 vs >100, and ASA 3.7 vs 3.9.

We found that criteria from MR CLEAN, ASA and REVASCAT excluded the lowest proportion of patients with comparable NB1H. We believe that these criteria would be reasonable to be utilised for ECR selection.