Background and purpose: Stroke is one of the most common causes of mortality and morbidity in the Western world. It results from the occlusion of a cerebral artery followed by severe disturbances in blood supply through microvessels to brain tissue. Despite an extensive literature its pathophysiology is poorly understood, and this has severely impeded the logical development of therapy.

Methods: Brains were obtained from 10 patients aged 46 to 85 years with survival times of 5 to 92 days after their stroke. Infarcted areas and representative control tissues from the contralateral uninvolved brain hemisphere were collected. Microvessel density was measured microscopically. A total of 6520 microvessels were scored in 10,801 areas. The level of activation of the endothelial cells was studied by immunohistochemistry using three monoclonal antibodies, viz, E-9, raised against activated endothelial cells; IG11, recognizing vascular cell adhesion molecule-1; and anti-proliferating cell nuclear antigen. Angiogenic activity in tissue extracts was examined using an in vivo chicken chorioallantoic membrane assay.

Results: There was a statistically significant increase in the number of microvessels (Wilcoxon log-rank test; P < or = .01) in 9 of 10 infarcted brain tissues when compared with their contralateral normal hemisphere. In these patients the higher blood vessel counts correlated with longer survival, as ascertained by Spearman's p analysis (P < .02). The number of microvessels filled with blood cells was significantly lower in the infarcted hemispheres (P < .01). In contrast, statistically significant increased numbers of empty microvessels occurred in infarcted tissues compared with the contralateral hemisphere. Monoclonal antibody E-9 reacted weakly with normal-brain vascular endothelial cells; anti-proliferating cell nuclear antigen and IG11 were virtually negative. All three antibodies strongly stained the blood vessels of stroke tissues. The stroke tissues contained angiogenic activity, as shown by the induction of new blood vessels in a chorioallantoic membrane assay.

Conclusions: We have shown that stroke causes active angiogenesis that is more developed in the penumbra. Further experiments are needed to determine if this angiogenesis has beneficial effect.