Acute Ischemic Stroke with Very Early Clinical Improvement: A National Institute of Neurological Disorders and Stroke Recombinant Tissue Plasminogen Activator Stroke Trials Exploratory Analysis

BackgroundA high proportion of patients excluded from recombinant tissue plasminogen activator (rt-PA) treatment because of rapid improvement occurring before treatment decision had incomplete recovery. The National Institute of Neurological Disorders and Stroke (NINDS) rt-PA Stroke Trials dataset allows for systematic analyses of very early postrandomization improvement (VEPRIM) in stroke severity as a National Institutes of Health Stroke Scale (NIHSS) score was available for all subjects enrolled in the study at baseline (NIHSSB) and at 2 hours after randomization (NIHSS2H). We explored various definitions of VEPRIM to characterize predictive values for clinical outcomes.MethodsPost hoc analyses of the NINDS rt-PA Stroke Trials were conducted. VEPRIM was defined as the difference between the NIHSSB and the NIHSS2H scores using 3 approaches: raw, percent, and normalized change. We assessed the association between VEPRIM and 3-month favorable outcome (mRS score of 0-1), symptomatic intracerebral hemorrhage (sICH), and death.ResultsIn the 624 subjects, every VEPRIM definition was independently associated with an increased probability of favorable outcome: for each unit of change within the VEPRIM definitions, there were 2%-24% (all P < .05) relative increased probability of favorable outcome, 2%-15% (all P < .05) decreased likelihood of death, and 2%-13% (all P < .05) decreased likelihood of sICH. Adjusting for NIHSSB and prestroke mRS scores, there was a significant rt-PA treatment effect for improvement seen for all 3 VEPRIM definitions.ConclusionsVEPRIM predicted favorable outcomes independent of definition and treatment arm. Patients with VEPRIM by any definition, while doing better than patients without VEPRIM, also derived increased clinical benefit when treated with rt-PA compared to placebo. Even with VEPRIM, a substantial percentage of patients had unfavorable outcomes.