Pharmacologically induced preconditioning with diazoxide: A novel approachto brain protection

Background. Ischemic preconditioning is an endogenous mechanism whereby bri ef periods of ischemia render neurons resistant to subsequent lethal insult s. This protection appears to alter cellular apoptosis and can be induced b y potassium channel openers acting on the inner membrane of the mitochondri a (mitoK(ATP)). To test the hypothesis that pharmacologic preconditioning c ould provide neuroprotection, the mitoKATP opener diazoxide was used in a c anine model of brain injury induced by hypothermic circulatory arrest (HCA) . Methods. Seventeen dogs were placed on cardiopulmonary bypass (CPB) and coo led to 18 degreesC. After 2 hours of HCA, animals were rewarmed and weaned from CPB. Six dogs received intravenous diazoxide (2.5 mg/kg bolus 15 minut es prior to CPB, then 0.5 mg/min until circulatory arrest, then restarted f or the first hour of rewarming). Six animals received vehicle only. Five re ceived diazoxide and the mitoK(ATP) blocker 5-hydroxydecanoate (5-HD). Usin g a modified Pittsburgh Canine Neurological Scoring System (0 = normal, 500 = brain death), animals were evaluated every 24 hours for 3 days. The brai ns were removed and histologic sections of four regions characteristically injured in this model were scored (0 = no injury, 4 = infarction) by a neur opathologist in a blinded fashion. Results. Clinical scoring showed marked improvement in the diazoxide group at 48 hours (101 +/- 10.5 vs 165 +/- 14.8, p < 0.01) and 72 hours (54 +/- 9 .3 vs 137 +/- 12.1, p < 0.01). This neuroprotection was attenuated when 5-H D was concomitantly administered. Three of four brain regions typically inj ured in this model (cortex, hippocampus, and entorhinal cortex) had signifi cant neuron preservation in the diazoxide group. Likewise, combined region scores were significantly improved in the treatment group (1.18 +/- 0.2 vs 2.46 +/- 0.2, p < 0.01). Conclusions. Pretreatment with diazoxide resulted in significant improvemen t in both clinical neurologic scores and histopathology in our model of HCA . This suggests that pharmacologic preconditioning with the mitoK(ATP) chan nel opener diazoxide may offer effective neuroprotection during HCA. (C) 20 01 by The Society of Thoracic Surgeons.