CT Perfusion in Acute Stroke Questioned

The extra time needed for CT perfusion imaging in patients with an acute ischemic stroke may not be warranted, based on a retrospective analysis of 418 patients treated at nine U.S. tertiary stroke centers.

The analysis showed that the outcomes in patients assessed using CT perfusion (CTP) were very similar to those in patients worked up with noncontrast CT (NCCT), and that CTP added an average of 48 minutes to the time elapsed between the start of imaging and the completion of the reperfusion procedure, Dr. Rishi Gupta said at the annual International Stroke Conference.

Dr. Rishi Gupta
“Additional imaging did not translate into better clinical outcomes or reduced hemorrhage rates, raising the question of whether NCCT is good enough,” said Dr. Gupta, a neurologist at Emory University and director of the acute stroke network at Grady Health System, both in Atlanta.

Because the analysis was retrospective, the next step is a prospective, randomized study to compare the impact of NCCT and CT perfusion, Dr. Gupta noted. Despite this limitation, he said he and his associates at Grady Health are convinced by the findings and have already scaled back their use of CTP to rely more on NCCT.

“CTP better demarcates the area of irreversible brain damage, but we stopped doing a lot of CTP because we were cognizant of the time delay. We didn’t feel we got additional, helpful information. We cut back quite a lot on our CTP based on these findings,” although a majority of centers that perform endovascular perfusion on acute stroke patients “tolerate the delay and get CTP,” he said in an interview. “We need to do a randomized study” to settle the question, he added.

The data set compiled from the nine participating U.S. centers included 418 eligible patients who underwent imaging prior to endovascular reperfusion therapy between September 2009 and December 2011. Of these, 227 (54%) had CTP and 191 (46%) had NCCT.

The study included consecutive patients with an occlusion of the middle cerebral or internal carotid artery treated within 8 hours of symptom onset. The analysis excluded patients with a posterior circulation stroke, those who underwent MRI, and those with a thrombus in their anterior cerebral or distal middle cerebral artery. The patients’ average age was 67 years, and their average National Institutes of Health Stroke Scale score was 18.

The analysis showed successful reperfusion in 65% of the NCCT patients and in 71% of those imaged with CTP. A good clinical outcome – a modified Rankin Scale score of 0-2 at 90 days after hospitalization – was achieved in 37% of the NCCT and 38% of the CTP patients. Mortality was 23% in the NCCT and 21% in the CTP patients. The rates of symptomatic and asymptomatic hemorrhage were also similar in the two subgroups. None of these between-group differences were statistically significant. In a multivariate analysis, the use of CTP was not a significant determinant of a good clinical outcome.

Average time from the start of CT imaging to reperfusion was 175 minutes in the NCCT patients and 223 minutes in the CTP group – an average of 48 minutes additional time with CTP.

Dr. Gupta presented two additional analyses designed to compare outcomes using the two imaging methods in closely matched subgroups. In one analysis, he focused exclusively on the 291 patients in the database who had occlusions at the M1 site of the middle cerebral artery. In these patients, NCCT saved an average of 40 minutes, compared with CTP, and outcomes were not significantly better with CTP. In the second analysis, he categorized patients by their Alberta Stroke Program Early CT Score (ASPECTS). In the subgroup of 198 patients with an ASPECTS of more than 7 on a scale of 10, with 10 being normal, CTP did not lead to significant improvement in outcomes and took an average of 45 minutes longer to reperfusion, compared with NCCT.

Dr. Gupta said that he has been a consultant to Concentric Medical, CoAxia, Rapid Medical, and Codman.

LEAVE YOUR COMMENT

Your email address will not be published.