Studies of collateral perfusion to canine middle cerebral artery territory.

Previous studies of intracranial collateral circulation have not distinguished between true "collateral" blood flow (flow to a region that occurs only when a primary artery is occluded) and "overlap" flow (flow to a region that is present under both normal and demand conditions). These experiments had three purposes: 1) to identify tissues that were truly collateral dependent, 2) to determine potential for true collateral flow in the absence of overlap flow, and 3) to determine whether an anatomical basis for overlap flow could be demonstrated. Branches (700-900 microns) of the dog middle cerebral artery (MCA) were perfused with autologous blood. The perfused region, which was the area at risk, was identified by intravenous injection of neutral red dye. Microspheres were used to measure regional cerebral blood flow (rCBF). Overlap flow was determined by perfusion of the artery with microsphere-free blood. True collateral flow (total rCBF minus overlap flow) was determined by analysis of rCBF to the risk area after cessation of vessel perfusion. Most of the risk area had substantial levels of overlap flow (about one-third of base line). In the center of the area at risk, the true collateral-dependent area was identified [mean overlap flow 4 +/- 1 (mean +/- SE) ml.min-1.100 g-1], which had high levels of perfusion from collateral vessels (102 +/- 14) within 30 s of vascular occlusion. Microfil injection into two adjacent MCA branches showed discrete borders between vascular territories, with no overlapping vessels.(ABSTRACT TRUNCATED AT 250 WORDS)