Demographics, Risk Factors, and Etiology of Posterior Cerebral Artery Stroke Causing Homonymous Hemianopia

Abstract

Background: Posterior cerebral artery (PCA) strokes account for up to 10% of all ischemic strokes, often presenting with homonymous hemianopia. The proportion of these strokes attributed to various etiologies varies widely in previously published studies, owing largely to differing patient populations, definitions of stroke pathogenesis, and vascular territories involved. The Causative Classification System (CCS), an automated version of the Stop Stroke Study (SSS) Trial of Org 10,172 in Acute Stroke Treatment (TOAST) system, allows for a more rigorous assignment of stroke etiology. Methods: We excerpted clinical and imaging data on 85 patients who had PCA stroke with homonymous hemianopia examined at the University of Michigan. We compared the stroke risk factor profile of our PCA cohort with that of 135 patients with stroke in the distribution of the internal carotid artery (ICA) and middle cerebral artery (MCA) in an unpublished University of Michigan registry. We applied the CCS web-based calculator to our PCA cohort to determine stroke etiology. Results: In our PCA cohort, 80.0% had at least 2 conventional stroke risk factors and 30.6% had 4 risk factors, most commonly systemic hypertension. The risk factor profile of our PCA cohort resembled that of our ICA/MCA cohort except that the mean age of our PCA cohort was more than a decade younger and had a significantly lower frequency of atrial fibrillation (AF) than our ICA/MCA cohort. In nearly half of the patients with AF in our PCA cohort, AF was diagnosed after the stroke. Among stroke etiologies in our PCA cohort, 40.0% were of undetermined cause, 30.6% were from cardioaortic embolism, 17.6% were from other determined causes, and only 11.8% were from supra-aortic large artery atherosclerosis. Strokes after endovascular or surgical interventions were prominent among other determined causes. Conclusions: Most patients in our PCA cohort had multiple conventional stroke risk factors, a finding not previously documented. Mean age at stroke onset and AF frequency were lower than in our ICA/MCA cohort, in agreement with previous studies. As some other studies have found, nearly 1/3 of strokes were attributed to cardioaortic embolism. Within that group, AF was often a poststroke diagnosis, a finding not previously highlighted. Compared with earlier studies, a relatively high portion of strokes were of undetermined etiology and of other determined etiologies, including stroke after endovascular or surgical interventions. Supra-aortic large artery atherosclerosis was a relatively uncommon explanation for stroke.