A new canine model of focal cerebral ischemia has been developed employing intravascular delivery of microfibrillar collagen via femoral catheterization. In 13 dogs, dose-effect studies showed neurologic deficits (ranging from mild hemiparesis to death) related to the dose of microfibrillar collagen delivered. In another 10 dogs, 0.5 ml of 60 mg/ml microfibrillar collagen was injected into the common carotid artery; neurologic assessment over 48 hours revealed a survivable stroke syndrome in seven dogs, death at 40 hours in one dog and at <12 hours in another, and no clinical effect in one dog. The eight surviving dogs were sacrificed at 48 hours; nine of the 10 dogs had middle cerebral artery distribution infarcts (two grossly hemorrhagic and five grossly nonhemorrhagic) on histologic examination. Angiography in three dogs demonstrated no significant major vascular occlusion. All seven dogs with survivable strokes demonstrated a dense hemiparesis at 24 hours that improved to ambulatory status at 48 hours. O ne of the most widely used animal models of focal ischemia involves permanent or reversible occlusion of the middle cerebral artery (MCA).
-11 However, this model is compromised by the necessity of enucleation to gain access to the MCA, by violation of the dura, and by the need for sophisticated microsurgery. The physical manipulation of subarachnoid vessels and adjacent brain tissue as well as the risk of spinal fluid leakage further complicates this model. The fact that intracranial surgery is required raises questions as to whether the model is sufficiently analogous to embolic disease to allow extrapolation of physiologic data to the more common circumstance in humans. One embolic model in dogs 5 -6 ' 11 uses silicone cylinders introduced into the internal carotid artery to embolize the MCA, producing basal ganglia infarcts. However, this technique Received September 21, 1988; accepted March 17, 1989. also requires surgical access to the carotid arteries in the neck and results in persistent occlusion of the main trunk of the MCA. We have recently developed a modification of this model that results in reversible occlusion. 11 Primates would provide the best model for cerebrovascular disease amenable to study with noninvasive imaging modalities; however, both ethical and economic issues limit the widespread use of primates.Because their brains are large enough to image radiographically, dogs represent an acceptable alternative to primates as candidates for the study of cerebrovascular disease. Remote delivery of embolic material is desirable to study ischemia immediately following its onset without contamination by surgical artifacts. Using microfibrillar collagen (Avitene; Aicon Laboratories, Inc., Fort Worth, Texas), we have developed a model for cerebral ischemia in dogs that consistently creates infarcts in the MCA distribution and permits clinical-pathologic correlations and simultaneous physiologic studies. We describe the model and the attendant clinical, angiographic, and histopathologic result...