Perfusion CT helps in predicting stroke outcome

Perfusion CT adds clinically relevant information for evaluating patients with suspected acute ischemic stroke, according to a study published online Feb. 12 in Stroke.

Physicians use noncontrast head CT and CT angiography (CTA) combined with clinical information to assess tissue fate and outcomes in patients with suspected acute ischemic stroke. But perfusion/penumbral imaging also may offer a method for quickly gaining qualitative and quantitative information that potentially helps physicians discriminate between ischemic but salvageable and infarcted tissue.

Guangming Zhu, MD, PhD, of the radiology department at the University of Virginia in Charlottesville, and colleagues observed that some sites use perfusion CT in routine workup of stroke patients but, because insufficient evidence exists that perfusion CT adds value, it is not yet part of standard care. They designed their study to fill that knowledge gap with two analyses: first, to assess whether perfusion CT infarct and perfusion CT penumbra can be predicted by noncontrast head CT, CTA and clinical variables; and second, whether information about perfusion CT infarct and perfusion CT penumbra in addition to noncontrast head CT, CTA and clinical variables predicts clinical outcome.

The researchers identified 165 patients retrospectively from two sites between 2003 and 2011 who had internal carotid artery or occlusion and no evidence of intracerebral hemorrhage; a complete stroke CT workup on admission with a time from onset to imaging was of less than 12 hours; completion of recanalization imaging between one and 48 hours; National Institutes of Health Stroke Scale (NIHSS) on admission and modified Rankin score (mRS) at 90 days.

Perfusion CT studies were done on 16- and 64-slice CT scanners and were used to calculate volumes of infarct core and ischemic penumbra on admission. They assessed recanalization status with follow-up imaging and used multivariate regression for the two analyses.

The patients were almost evenly split by gender (49.7 percent women) with a mean age of 65.4 years and mean NIHSS on admission of 15.4. The mean perfusion CT infarct volume was 44.6 mL and mean perfusion CT penumbra volume was 47.2 mL.

Zhu et al found that perfusion CT penumbra volume could not be predicted by noncontrast head CT, CTA and clinical variables. Clinical data, perfusion CT data and post-workup treatment and recanalization data were significantly associated with 90-day mRS outcome, but noncontrast head CT and CTA were not significantly associated with 90-day mRS outcome. Recanalization emerged as the most important predictor of 90-day mRS outcome, with perfusion CT penumbra volume also a significant predictor.

“In our large series of patients, we found that perfusion CT infarct core can indeed be inferred from clinical, noncontrast head CT and CTA data, or various combinations thereof,” they wrote. They added that although there was a significant association of noncontrast head CT, CTA and clinical variables with perfusion CT infarct core volume, the latter nonetheless provided additional information for assessments.

Their second analysis showed that perfusion CT penumbra was an independent predictor of 90-day mRS and provided unique and clinically relevant information, they continued. “Interestingly, penumbra was a hallmark of positive outcome only in the presence of recanalization, whereas it was associated with worst outcome in the absence of recanalization.”

They suggested that perfusion CT may assist in the initial evaluation of acute ischemic stroke patients in addition to other imaging, and recommended prospective studies to validate perfusion CT’s role in selecting patients for reperfusion therapy.

The study was limited by variability in imaging modalities for assessing recanalization. Also, by selecting a patient population in which noncontrast head CT and CTA are highly sensitive, they may have underestimated the value of perfusion CT.

The study was funded in part by Cardiomet CHUV.