CONFIRMATION OF CT CRITERIA TO DISTINGUISH PATHOPHYSIOLOGIC SUBTYPES OF CEREBRAL INFARCTION

PURPOSE: To determine whether cerebral infarctions classified as embol ic or hemodynamic by their appearance on CT scans reflect distinct pat hophysiologic entities. METHODS: Cerebral infarctions were retrospecti vely classified into two groups according to their morphologic appeara nce on CT scans: territorial infarctions and watershed, or terminal su pply area, infarctions. Specific CO2 reactivity for both groups of pat ients was determined with the xenon-133 method and 32 stationary detec tors. Twenty-one patients with unilateral, supratentorial, ischemic ce rebral infarctions were selected. CT findings were highly suggestive o f a territorial infarction in 14 patients (mean age, 56 years) and of a watershed infarction in seven patients (mean age, 52 years). RESULTS : The initial slope index of the territorial and watershed infarction groups during CO2 inhalation was 55.1 +/- 2.4 sec(-1) and 52.0 +/- 1.9 sec(-1), respectively, in the infarcted hemispheres and 58.3 +/- 2.3 sec(-1) and 55.1 +/- 1.5 sec(-1), respectively, in the noninfarcted he mispheres. CO2 reactivity of the unaffected detectors was 1.75 +/- 0.3 sec(-1) mm Hg-1 and 1.51 +/- 0.2 sec(-1) mm Hg-1 for the territorial and watershed infarction groups, respectively. CO2 reactivity of the a ffected detectors was 1.75 +/- 0.3 sec(-1) mm Hg-1 and 1.27 +/- 0.2 se c(-1) mm Hg-1 for the two groups, respectively. The CO2 reactivity dif ference between affected detectors of the hemodynamic group and age-ma tched healthy control subjects was significant. CONCLUSIONS: The diffe rence in CO2 reactivity between the two groups supports the concept th at CT criteria can identify two pathophysiologic entities. In addition , we conclude that during the chronic stage, lower CO2 reactivity of t he watershed infarction indicates that the global hemodynamic situatio n in these infarcts is more severely compromised than in territorial i nfarctions.