Pro: Prevention of neurologic dysfunction associated with cardiac surgery requires pharmacologic brain protection

Abstract

C ARDIAC SURGERY has become the most commonly performed major surgery in the United States over the past four decades. In large part, this has been attributed to success in decreasing mortality rates and improvements in cardiac, hematologic, and other morbidity rates. In contrast to these improvements, the incidence of neurologic dysfunction has not diminished, with cerebral injury accounting for an increased proportion of deaths after cardiac surgery. 1 These complications consume a tremendous amount of resources, both during as well as after hospitalization, and may have profound impact on the lives of patients and their families. 2 Although prevention of neurologic dysfunction has proven to be an elusive goal, recent scientific advances provide a basis for hope. Although the best-studied type of surgery is coronary artery bypass grafting (CABG), neurologic complications also occur frequently with valve replacement and other types of openchamber heart surgery as well as with repair of congenital heart disease. The preponderance of evidence suggests most neurologic dysfunction associated with cardiac surgery results from cerebral embolization. Sources of emboli include atheromatous debris from the aorta, thrombus, air, valvular debris, and emboli from the cardiopulmonary bypass circuit. However, global ischemia or watershed-type injuries may also result from circulatory arrest or severe hemodynamic instability. Clearly, the cascade of events leading to neurologic injury associated with cardiac surgery may be interrupted at several points. The argument has been advanced that neurologic dysfunction continues to occur with a disturbingly high incidence because of the increasing age of the patient population and the presence of other comorbid conditions. Indeed, numerous studies have identified several preoperative risk factors for adverse neurologic outcomes, including advanced age, previous stroke or transient ischemic attacks, hypertension, diabetes mellitus, carotid atherosclerosis, and pulmonary disease. 2,3 Identification of these risk factors has allowed the development of preoperative stroke risk indices that quantitate a patient's relative risk of sustaining a perioperative stroke) ,4 Therefore, the simplest, although probably least acceptable, strategy would be not to operate on patients at high risk. Although this approach would undoubtedly decrease the incidence of cerebral injury, it would also deprive many patients of a chance of prolonging and improving the quality of their lives. Clinicians should discuss the risk-to-benefit ratio very frankly with these patients, because the benefit of prolonged survival may be offset by the likelihood of substantial postoperative disability. These may also be the patients most likely to benefit from therapeutic interventions, and future interventional trials should be focused on this subset of patients presenting for cardiac surgery. It must also be emphasized that the clinician's ability to predict these complications remains limited, and even low-risk patients may sustain neurological injury after cardiac surgery. Another potential method of decreasing neurologic dysfunction associated with cardiac surgery is alteration of surgical techniques to decrease the incidence of cerebral embolization. Currently, much attention has been focused on the role of proximal aortic atherosclerosis in the development of neurologic dysfunction after surgery. 1,2,s,6 Although patients with diseased aortae are clearly at increased risk for sustaining neurological insults, it is much less clear which, if any, therapeutic maneuvers may lessen the chance for such an insult. Techniques that have been touted to decrease neurologic injury in these patients include alterations in the aortic cross-clamping site, changes in the site of the proximal anastomoses, use of only arterial grafts, performing all grafts with a single crossclamp, and even graft replacement of the diseased portion of the aorta using hypothermic circulatory arrest. 5,6 Unfortunately, these nonrandomized studies have potential for selection bias and have insufficient power to permit drawing conclusions. It is critical that such trials examine all sources of morbidity and long-term outcomes, especially for aggressive interventions that may carry significant morbidity of their own. Management of cardiopulmonary bypass may also decrease cerebral embolization. Pugsley et al7 showed use of an arterial filter decreases the number of emboli, and membrane oxygenators have also been associated with decreased rates of embolization. Again, no trial has demonstrated neurologic benefit associated with these techniques. One exception is the use of c~-stat blood gas management that has been shown to decrease neuropsychological dysfunction, but not neurologic dysfunction in patients undergoing CABG. 8 Whether this translates into