Spinal cord ischemia. Development of a model in the mouse.

BACKGROUND AND PURPOSE Spinal cord ischemia with resulting paraplegia is a devastating complication of thoracoabdominal aortic surgery. Experimental models of spinal cord ischemia have been developed in primate, dog, pig, rabbit, and rat with variable reproducibility, but none has been developed in mouse. Because genetically engineered mice have become important to examine the impact of specific genes in ischemic pathophysiology, we sought to develop a reproducible mouse model of spinal cord ischemia. METHODS C57BL/6NCrlBR mice were subjected to cross-clamping of the aortic arch, left subclavian artery, and internal mammary artery for 9 minutes (group A; n=8) or 11 minutes (group B; n=29) followed by reperfusion for 24 or 48 hours. Mean distal arterial blood pressure (left femoral artery) and lumbar (L1) spinal cord blood flow (laser-Doppler flowmetry) were measured for the duration of the procedure. The arterial blood supply of the spinal cord was visualized by intravascular perfusion of carbon black ink. We evaluated motor function in the hind limbs at 0, 1, 3, 6, and 24 hours after reperfusion using a rating scale of 0 (normal function) to 6 (total absence of movement). Spinal cord histopathology was evaluated after 24 and 48 hours of reperfusion by Luxol fast blue-hematoxylin and eosin. RESULTS The vascular anatomy of the mouse and human spinal cord appeared similar in that blood was supplied by 1 anterior and 2 posterior spinal arteries and heterosegmental radicular arteries. During combined occlusion of aortic arch and left subclavian artery, mean distal arterial blood pressure dropped to 10+/-5 mm Hg, and spinal cord blood flow at the L1 level decreased to 27+/-7% of baseline. All animals recovered from anesthesia with acute paraplegia. Animals in the 9-minute group (group A) showed steady recovery of hind limb function over the ensuing 24 hours, whereas the majority (80%) in the 11-minute group (group B) remained paralyzed with maximum deficit throughout the postoperative period. Mortality was 0% and 21% in groups A and B, respectively. Maximal ischemic damage was observed at the lower thoracic and higher lumbar spinal levels in both groups. In group A (9 minutes), tissue damage was mild, affecting predominantly dorsal horns and intermediate gray matter, whereas ventral horns were minimally involved. All mice in group B (11 minutes) showed extensive gray matter lesions particularly involving dorsal horns and intermediate areas; in ventral horns, >50% of motor neurons died. White matter lesions were present in the most severely damaged cords only. CONCLUSIONS Spinal cord ischemia caused by aortic arch plus left subclavian artery cross-clamping provides a mouse model useful for the study of spinal cord injury and of potential relevance to the complications following thoracoabdominal aortic surgery in humans.