BORDERZONE BRAIN INFARCTS ON CT TAKING INTO ACCOUNT THE VARIABILITY IN VASCULAR SUPPLY AREAS

Although borderzone brain infarcts lack a uniform definition and are d ifficult to define because of significant variability in cerebral vasc ular supply, they are presumed to be related to haemodynamic factors s uch as carotid obstruction. In this study on 813 first-ever stroke pat ients, borderzone infarcts were defined not. only according to the cur rently used 'classic' definition, but also taking into account the var iability in cerebral vascular supply areas. Borderzone infarcts, defin ed in either way, were compared with remaining superficial infarcts wi th respect to vascular risk factors. We identified 5 patients with 'cl assic' borderzone infarcts (slit-like infarcts along the borders of th e different vascular territories). Four of these had a carotid occlusi on, which was significantly more frequent than in remaining superficia l infarcts (odds ratio, OR: 15.52; 95% confidence interval, CI: 2.01-1 19.68; p = 0.008). Borderzone infarcts defined accounting for variabil ity in vascular supply did not show any difference with regard to spec ific risk factors when compared with remaining superficial infarcts. C hronic obstructive pulmonary disease (COPD) and especially COPD combin ed with diabetes mellitus was more frequent among borderzone infarcts defined in either way (adjusted OR 21.3, 95% CI, 3.80-120, p = 0.0004, and adjusted OR 12.0, 95% CI 2.26-63.6, p = 0.003, respectively); thi s factor may increase the risk of infarction in case of reduced flow o r thrombo-embolism. In conclusion: 'classic' borderzone infarcts are a ssociated with ipsilateral carotid occlusive disease; this assocation is lost when expanding the definition, although COPD combined with dia betes mellitus then appears as a new risk factor associated with borde rzone infarcts.