THE USE OF SOMATOSENSORY-EVOKED POTENTIAL MONITORING TO PRODUCE A CANINE MODEL OF UNIFORM, MODERATELY SEVERE STROKE WITH PERMANENT ARTERIAL-OCCLUSION

TO DEVELOP A reliable canine model of cerebral infarction of moderate size, we compared infarctions caused by permanent occlusion of the following vessels in 42 dogs: 1) the middle cerebral artery (MCA), 2) the MCA and azygous anterior cerebral artery (ACA), 3) the MCA, azygous ACA, and posterior cerebral artery (PCA), and 4) sham-operated controls. The infarction volume was determined at 6 hours in half the animals and at 6 days in the others. Studies of somatosensory evoked potentials (SSEPs) and regional cerebral blood flow (rCBF) were performed before and after arterial occlusion, and good correlation was observed between the decrease in amplitude of the SSEPs and the decrease in rCBF observed after arterial occlusion. Only the groups in which the MCA and azygous ACA were occluded showed moderate infarctions of relatively consistent size. Analysis involving all groups revealed that the animals with SSEP amplitude preserved after vessel occlusion had only small infarctions; thus, preservation of SSEP amplitude after occlusion of the MCA and azygous ACA could in the future be used prospectively as a rejection criterion to improve the uniformity of infarction size. Conversely, animals with loss of SSEP amplitude after vessel occlusion had infarctions of moderate to large size; thus, loss of SSEP amplitude after MCA occlusion alone could in the future be used prospectively as a rejection criterion. When these rejection criteria were retrospectively applied to the groups in which both the MCA and azygous ACA were occluded, the resulting mean infarction volumes +/-1 SEM) for the acute and chronic subgroups were 20.3 +/- 2.8% and 38.2 +/- 4.5% of the hemisphere, respectively. To prospectively verify this model, chronic infarctions were created in 10 additional dogs using SSEP-based selection during occlusion of the MCA and azygous ACA. The mean infarction volume at 6 days was 28.3 +/- 2.8% of the hemisphere. We conclude that a reliable stroke model can be produced in the dog by simple two-vessel occlusion combined with selection based on the changes in SSEP amplitude immediately after occlusion.