Short-term changes in cerebral activity in on-pump and off-pump cardiac surgery defined by functional magnetic resonance imaging and their relationship to microembolization.
Abu-Omar Y., Cader S., Guerrieri Wolf L., Pigott D., Matthews PM., Taggart DP.
OBJECTIVE: Cognitive dysfunction is common early after cardiac surgery. We previously reported that functional magnetic resonance imaging of the brain can detect subclinical changes in prefrontal cortical activation after coronary artery bypass grafting. In this study, we used functional magnetic resonance imaging to contrast perioperative prefrontal activation in patients undergoing on-pump and off-pump coronary artery bypass grafting and to relate differences to cerebral microembolic load. METHODS: Functional images of the brain were acquired in 25 patients undergoing cardiac surgery (13 off-pump and 12 on-pump) before surgery and 4 weeks after surgery during performance of a verbal memory task of increasing complexity (n-back task). Continuous intraoperative transcranial Doppler scanning was performed to quantify the number of cerebral microemboli. Perioperative changes in task-associated prefrontal activation were compared between the 2 groups and were then correlated with the number of microemboli recorded during surgery. RESULTS: The median (interquartile range) number of detected microemboli was 35 (21-63) in the off-pump group and 254 (116-397) in the on-pump group (P < .005). Functional imaging performed before surgery demonstrated increased activity in the prefrontal regions with increasing task complexity. After surgery, there was a significant reduction in task-associated prefrontal activation in the on-pump, but not in the off-pump, group (P < .05). There was a negative correlation between the perioperative signal changes in the prefrontal region and the total number of microemboli (r = -0.63; P < .01). CONCLUSIONS: Patients undergoing on-pump, but not off-pump, surgery have a significant relative reduction in prefrontal activation, which correlates with intraoperative cerebral microembolic load. We hypothesize that this reduction in activation is related to subclinical functional impairments and that microembolic load is an important mechanism of perioperative cerebral insult.