HONOLULU—Penumbral imaging does not appear to aid the selection of patients who will most benefit from endovascular therapy within eight hours of acute ischemic stroke onset, according to study results presented at the 2013 International Stroke Conference. Outcomes of embolectomy and standard medical care may be similar, regardless of whether a patient has a penumbral pattern, said Chelsea Kidwell, MD, Professor of Neurology at Georgetown University Medical Center in Washington, DC.
The study results, which were published in the New England Journal of Medicine, did not change after the investigators adjusted for age, which was the only significant univariate predictor. Patients with a penumbral pattern, however, had significantly improved outcomes overall, regardless of treatment assignment, compared with patients who did not have a penumbral pattern.
Patients Were Randomized to Embolectomy or Medical Therapy
Dr. Kidwell and colleagues studied 118 patients with acute ischemic stroke in the Mechanical Retrieval and Recanalization of Stroke Clots Using Embolectomy (MR RESCUE) trial, a multicenter, randomized, controlled, blinded-outcome study conducted at 22 sites in North America. The researchers screened patients within eight hours of symptom onset with multimodal MRI or CT. If the vessel imaging showed a distal internal carotid artery occlusion or a possible middle cerebral artery occlusion, patients were randomized to embolectomy or to standard medical care. Embolectomies were performed with the Merci Retriever or the Penumbra System. Randomization was stratified by either a favorable penumbral pattern or nonpenumbral pattern.
Patients’ mean age was 65.5, and their median NIH Stroke Scale score was 17. Mean time to enrollment was 5.5 hours. More than half (58%) of patients had a favorable penumbral pattern. A total of 37% of patients received IV t-PA, and 80% were screened with MRI. The study’s mortality rate was 21%, and the symptomatic hemorrhage rate was 4%. Neither of these rates differed among the four patient subgroups. Of the patients in the embolectomy arm, 67% had partial or complete recanalization following the procedure.
Early Reperfusion Was Associated With Good Outcomes
A shift analysis of the day 90 modified Rankin scores revealed no significant interaction between treatment assignment and penumbral pattern, said Dr. Kidwell. Patients with a penumbral pattern who received standard care were more likely to have good outcomes, as defined by a modified Rankin score of 0 to 2, than the other three patient groups, but this result did not reach statistical significance. A receiver–operator curve analysis failed to identify a threshold of predicted core volume that would have yielded a significant difference in outcomes, based on treatment assignment and favorable penumbral pattern.
Patients with substantial reperfusion or partial or complete revascularization at day seven had improved clinical outcomes, as measured by day 90 modified Rankin scores, and attenuated infarct growth, regardless of treatment assignment or penumbral pattern. “Had we not had a control arm, we would not have been able to show that the benefit from revascularization was not an acute embolectomy treatment effect,” observed Dr. Kidwell.
First-Generation Devices May Have Influenced the Study Results
Dr. Kidwell proposed various factors that could explain the study results. For example, the investigators used only first-generation devices that have relatively low recanalization rates, compared with new generation stent retriever devices. The team also introduced two imaging techniques that may have led to different penumbral prediction patterns. Patients with a penumbral pattern may have had a higher chance of favorable outcomes, regardless of the treatment assignment, because they had the support of collateral vessels until spontaneous recanalization occurred. “It’s possible that the penumbral imaging selection hypothesis, as currently conceived, may be flawed,” said Dr. Kidwell.
“MR RESCUE underscores the importance of confirming hypotheses in randomized, controlled trials prior to implementing treatment approaches in current clinical practice,” she continued. “We believe that further randomized, controlled trials with new-generation stent-retriever devices are needed to test the full spectrum of the penumbral-imaging selection hypothesis and to test the clinical efficacy of the new-generation stent-retriever devices.”
Unanswered Questions
In an editorial published online in the February 8 issue of the New England Journal of Medicine, Marc I. Chimowitz, MB, ChB, analyzed Dr. Kidwell’s theory that a favorable penumbral pattern may identify patients who are likely to have a good outcome regardless of treatment assignment. This hypothesis “does not explain why only 14 to 23% of patients with a favorable penumbral pattern had a good outcome in MR RESCUE,” said Dr. Chimowitz, who is a professor of neurology at the Medical University of South Carolina in Charleston. “It may be because the penumbral pattern lacks specificity as a marker of durable tissue viability—in other words, only a subgroup of patients with a favorable penumbral pattern have ischemic brain tissue that can recover with later reperfusion.”