Experimental evidence of cerebral injury from profound hypothermia during cardiopulmonary bypass.

Abstract

Choreoathetosis, seizures, and impaired mental development continue to occur in children undergoing cardiopulmonary bypass (CPB) and profound hypothermia with or without circulatory arrest. Although there is some evidence that the hypothermia itself may be causing these neurologic problems, skepticism remains because of lack of evidence from experimental studies simulating the clinical setting. In this experimental study, we examined the effect of profound and moderate hypothermia on the brain while maintaining normal flow rates during CPB. Ten adult mongrel dogs equally divided into two groups were anesthetized and subjected to CPB and varying levels of hypothermia (group 1, < or = 15 degreesC; group 2, < or = 2 degreesC). Both groups were kept at the desired temperature for 1 hour prior to rewarming and discontinuation of CPB. The dogs were euthanized 4-6 weeks later and neuropathologic studies were performed. The mean CPB flow rates during cooling and at the desired rectal temperature were comparable in both groups: group 1, 108 +/- 10 ml/kg/min versus 106 +/- 7 ml/kg/min in group 2 (p = NS) and 95 +/- 12 ml/kg/min in group 1 versus 101 +/- 5 ml/kg/min in group 2 (p = NS). Because of the difference in temperature between the two groups, the mean cooling time (onset of CPB to desired rectal temperature) was longer in group 1 (70 +/- 14 minutes) than in group 2 (28 +/- 11 minutes, p = 0.007). Hence, the total mean CPB time was also longer in group 1 (198 +/- 25 minutes) than in group 2 (143 +/- 13 minutes, p = 0.002). The lowest mean blood and rectal temperature achieved in group 1 were 11 +/- .9 degreesC and 12 +/- 1 degreesC versus 29 +/- .4 degreesC (p < 0.001) and 30 +/- .6 degreesC (p = 0.001), respectively, in group 2 (p = 0.001). Neuronal loss and degeneration was noted in all dogs in group 1 ranging from 2 to 8 cells per 1000 cells counted compared to none in group 2 (p = 0.05). These lesions occurred in both the basal ganglia and the cortex, although they were more marked in the caudate when compared to the cortex and cerebellum. Both in the cortex and in the caudate, neuronal loss was more marked around the capillaries. We conclude that the use of profound hypothermia of < or =15 degreesC and maintenance of normal flow rates during cooling at this temperature for 1 hour produces neuronal loss and degeneration in the brain. These lesions being more marked around capillaries points to the vulnerability of the neurons, probably because of their high lipid content to injury from the cold perfusate.