Abstract 155: Critical Care Decisions After Large Core Cerebral Infarctions: A Secondary Analysis From the SELECT2 Trial

Abstract

Background/hypothesis: The SELECT2 trial randomly assigned patients with LVO and large ischemic cores to either endovascular thrombectomy (EVT) or medical management (MM). This population is at high risk for cerebral edema and other complications, often leading to critical decisions about decompressive hemicraniectomy (DHC) or early withdrawal of care (WOC). We hypothesized that patients initially treated with EVT were more likely to get life-sustaining care regardless of recanalization success, presumably because initial treatment with EVT led toward an expectation of aggressive care, while those treated with MM expected futility. Methods: We analyzed the full SELECT2 study population using the as-treated principle, comparing the use of DHC and early (within 7 days from randomization) transition to comfort measures/WOC. We also compared these decisions based on recanalization success in those receiving EVT. We further tested baseline characteristics for association with these outcomes. Results: Patients treated with EVT were as likely to undergo DHC (aRR:1.19 [0.75-1.88], p=0.46) or WOC (aRR:0.94 [0.66-1.34], p=0.72) as those given MM (Table). Time to DHC was also similar (EVT 47[19-74] vs. MM 36[27-61] hours, p=0.95). Patients with successful (mTICI≥2b) recanalization were numerically less likely to undergo DHC than those with unsuccessful (TICI 0-2a) recanalization (aRR:0.66 [0.33-1.3], p=0.23), while WOC was similar (Table). Larger estimated core volumes were associated with both DHC and WOC, with DHC used more in younger and WOC more in older patients. Conclusions: In the SELECT2 trial of patients with large infarct cores, DHC was performed in ~1 in 6 and WOC in ~1 in 5, without an observed difference based on treatment with EVT or MM. The similar distribution of decisions to proceed with DHC or to change goals of care to acute palliative measures provides reassurance that the overall trial outcomes were not biased by open-label treatment allocation.