EPITHET Results Show Efficacy of t-PA Beyond the Three-Hour Stroke Treatment Window
NEW ORLEANS — The three-hour treatment window for administration of t-PA may be safely and effectively extended to a full six hours after stroke onset, according to findings of the Echoplanar Imaging Thrombolysis Evaluation Trial (EPITHET). Stephen M. Davis, MD, reported that use of t-PA in the extended treatment period led to significantly increased reperfusion. However, the study’s primary outcome measure—reducing infarct growth—was not significant, although there was a strong trend toward infarct growth attenuation.
“We conclude from EPITHET that there is a biological rationale for using t-PA beyond three hours,” asserted Dr. Davis at the 2008 International Stroke Conference. He noted that the effect particularly applies to stroke patients with mismatched penumbra.
Penumbral Mismatch
Penumbral mismatch occurs when the rate of blood delivery to brain tissue differs from that of diffusion from the ischemic lesion, explained Dr. Davis. “In acute ischemic stroke, a large proportion of patients have a mismatch between a larger perfusion and a smaller diffusion lesion, typically associated with occlusion of the corresponding middle cerebral artery,” he said. This area, which can be viewed with magnetic resonance perfusion- and diffusion-weighted images, represents potentially salvageable tissue. “In patients with [magnetic resonance–detected] mismatch who go on to successfully reperfuse based on recanalization, [infarct] growth can be attenuated with minimal lesion growth and better neurologic and functional outcomes,” he added. Dr. Davis is a Professor of Neurology and Head of the Stroke Care Unit at the Royal Melbourne Hospital.
A total of 101 stroke patients from 15 centers in Australia, New Zealand, Belgium, and Scotland were enrolled in the phase II study. All participants had been screened to exclude hemorrhage and were then randomized to receive placebo or t-PA three to six hours after stroke onset. MRI was conducted to measure baseline perfusion and diffusion.
Reimaging occurred between days 3 and 5. Of the 91 participants included in the analysis, 86% had a mismatch—37 patients who received t-PA and 43 who received placebo. “Age, neurologic severity, time to treatment, median diffusion-weighted image volume, as well as perfusion-weighted image volume, mismatch volume, mismatch percentage, and those with the malignant profile—large baseline lesions—were well-matched between the two groups,” noted Dr. Davis.
t-PA and Infarct Growth
The EPITHET researchers reported that although there was 30% less relative infarct growth with t-PA than with placebo, the difference was not significant. “Median growth ratio achieved borderline statistical significance,” observed Dr. Davis. A strong trend for reduced median absolute infarct growth was also observed in the mismatch patients, at 4.1 mL of growth with t-PA versus 28.7 mL with placebo. Adjustment for postgrowth mismatch or classification of the patients based on treatment at three to 4.5 hours and 4.5 to six hours did not affect the significance of the outcomes.
In post hoc analysis, the researchers found that any infarct growth occurred in 54% of t-PA recipients and in 77% of placebo recipients. Once patients with small lesions (
Reperfusion and Recanalization
The major secondary outcome of reperfusion was significantly increased when t-PA was given between three and six hours, as 56% of those patients had 90% or greater reperfusion by days 3 to 5, versus 26% of placebo patients. Median reperfusion percentage was also significantly higher (91% vs 65%). “And reperfusion matters,” asserted Dr. Davis. “Reperfusion correlated with less geometric mean growth and better functional and neurologic outcomes.”
Measures of recanalization were less conclusive. “We measured recanalization at days 3 to 5 to catch a maximal infarct growth, but probably at too late a time to look at recanalization properly,” noted Dr. Davis. A higher proportion of patients were recanalized by day 5 in the t-PA group, however, and among those who achieved recanalization, there was typically less geometric mean growth and better functional and neurologic outcomes.
Differences in clinical outcome among the entire mismatch population were not statistically significant; however, as Dr. Davis pointed out, the study was not sufficiently powered to detect such differences. “But if you look at the [patients who achieved modified Rankin Scale scores of 0 to 1 at day 90], there is a strong trend with a 15% absolute improvement with t-PA compared to placebo,” he said.
“Given the strong link between reperfusion and improved clinical outcomes, phase III trials in [the three- to six-hour] time window are warranted,” concluded Dr. Davis. He also said that the trends observed among the clinical outcome measures suggested that fewer than 200 penumbral mismatch patients per arm would be required. He added that although the hypothesis that mismatch patients have the best chance of salvaging penumbral tissue cannot be directly validated, as there were too few nonmismatch patients, “the results suggest an enriched sample with the mismatch patients.”
