Context: The prevalence and clinical significance of early ischemic changes (EICs) on baseline computed tomography (CT) scan of the head obtained within 3 hours of ischemic stroke are not established.
Objective: To determine the frequency and significance of EIC on baseline head CT scans in the National Institute of Neurological Disorders and Stroke (NINDS) rt-PA (recombinant tissue plasminogen activator) Stroke Trial.
Design and setting: The original study, a randomized controlled trial, took place from January 1991 through October 1994 at 43 sites, during which CT images were obtained within 3 hours of symptom onset and prior to the initiation of rt-PA or placebo. For the current analysis, detailed reevaluation was undertaken after October 1994 of all baseline head CT scans with clinical data available pretreatment (blinded to treatment arm).
Patients: Of 624 patients enrolled in the trial, baseline CT scans were retrieved and reviewed for 616 (99%).
Main outcome measures: Frequency of EICs on baseline CT scans; association of EIC with other baseline variables; effect of EICs on deterioration at 24 hours (>/=4 points increase from the baseline National Institutes of Health Stroke Scale [NIHSS] score); clinical outcome (measured by 4 clinical scales) at 3 months, CT lesion volume at 3 months, death at 90 days; and symptomatic intracranial hemorrhage (ICH) within 36 hours of treatment.
Results: The prevalence of EIC on baseline CT in the combined rt-PA and placebo groups was 31% (n = 194). The EIC was significantly associated with baseline NIHSS score (rho = 0.23; P<.001) and time from stroke onset to baseline CT scan (rho = 0.11; P =.007). After adjusting for baseline variables, there was no EIC x treatment interaction detected for any clinical outcome, including deterioration at 24 hours, 4 clinical scales, lesion volume, and death at 90 days (P>/=.25), implying that EIC is unlikely to affect response to rt-PA treatment. After adjusting for NIHSS score (an independent predictor of ICH), no EIC association with symptomatic ICH at 36 hours was detected in the group treated with rt-PA (P>/=.22).
Conclusions: Our analysis suggests that EICs are prevalent within 3 hours of stroke onset and correlate with stroke severity. However, EICs are not independently associated with increased risk of adverse outcome after rt-PA treatment. Patients treated with rt-PA did better whether or not they had EICs, suggesting that EICs on CT scan are not critical to the decision to treat otherwise eligible patients with rt-PA within 3 hours of stroke onset.