Perioperative Temperature and Cardiac Surgery

Perioperative thermoregulation encompasses a very broad field in anesthesia and surgery. This review focuses principally on issues surrounding intraoperative temperature management, with particular emphasis on cardiopulmonary bypass (CPB) temperature issues. The implications of temperature during the postoperative period and its relationship to outcome after cardiac surgery will also be addressed. Temperature in the setting of cardiac surgery has been a major research focus for a number of decades. The judicious use of hypothermia remains a mainstay of perioperative management in the cardiac surgical patient with its putative, though far from definitively proven, global organ protective effects having led to its continued use. Although it has effects on most organ systems, this review will principally focus on the effects of temperature on the brain.

Although hypothermia has a defined and measurable effect of suppressing cerebral metabolism (approximately 6%–7% decline per °C) (1), it is likely that its other neuroprotective effect(s) may be mediated by non-metabolic actions. In the ischemic brain, for example, moderate hypothermia has been demonstrated to block the release of excitotoxic glutamate (2), reduce calcium influx (3), hasten recovery of protein synthesis (4), diminish membrane-bound protein kinase C activity (5), slow time to onset of depolarization (6), reduce formation of reactive oxygen species (7), and suppress nitric oxide synthase activity (8). It is likely that some or all of these effects in combination convey some of the neuroprotective effects of hypothermia. Although experimental demonstrations of this are abundant, clinical examples of hypothermia neuroprotection, until recently (9,10), have been elusive (11,12).

The past ten years have seen some definitive work in the field of temperature management during CPB. Although much of this work is centered on the effects of temperature on the brain, this work was initiated because of a developing body of evidence focused on optimizing temperature-mediated myocardial preservation during CPB. In the late 80’s and early 90’s, the judicious use of warm CPB was used because of its putative myocardial salvaging effects when used with continuous warm cardioplegia (13). However, because CPB was being carried out at a higher temperature than what was considered conventional, the implications on the brain were also studied. Several large studies were undertaken to elucidate the effects of temperature management on cerebral outcome after cardiac surgery. These were the Warm Heart Investigators trial (13), a trial performed at Emory University (14), and a later trial at Duke University (15). Although there were several differences between these trials, there were some very similar results with respect to neurocognitive outcome (16,17), but some very divergent results with respect to stroke. In short, none of the studies demonstrated any neuroprotective effect of hypothermia on neurocognitive outcome after cardiac surgery. What the Emory trial did demonstrate, however, was an apparent injurious effect (as manifest by a worse stroke outcome) of what was most likely mild degrees of cerebral hyperthermia during CPB. Neither the Warm Heart Investigators trial nor the Duke trial showed any effect of temperature on stroke per se. However, one of the issues raised from these trials related to how temperature was defined, monitored, and managed-of paramount importance when trying to understand the effects of temperature on the brain. The Warm Heart Investigators trial was in actual fact a comparison of moderate hypothermia to mild hypothermia, the Emory trial compared mild hypothermia to mild hyperthermia, and the Duke trial examined hypothermia vs. normothermia. In addition to these differences in temperature management, there were some other fundamental differences with respect to cardioplegia management and intraoperative glucose management.

Although there are numerous sites for monitoring temperature during cardiac surgery, several warrant special consideration. The take home message from the three warm vs. cold trials above, as well as other information regarding temperature gradients between the bypass circuit, nasopharyngeal and brain (18), is that it is very important to monitor (and use as a target) a temperature site relevant to the organ of interest. If it is the body, then a core temperature measured in the bladder, rectum, pulmonary artery, or esophagus would be appropriate. However, if one wants to measure the temperature of the brain, barring implantation of a thermistor directly into the brain (which has been done) (19), one needs to look at surrogates of brain temperature. These include nasopharyngeal temperature as well as tympanic membrane temperature. However, more invasive surrogates of brain temperature have also been obtained using a jugular bulb thermistor (18,20). What is clear from these different temperature sites is that vast gradients appear across the body and across the brain with respect to temperature. It is likely that during periods of rapid flux (such as during rewarming), that these temperature gradients are maximal.

Just as hypothermia likely has some protective effects on the brain, hyperthermia, in an opposite and disproportionate fashion, has some injurious effects. Although one can argue whether hypothermia in the setting of cardiac surgery has any definitive neuroprotective effect, there is emerging evidence that whatever neuroprotection can be afforded by hypothermia may be negated by the obligatory rewarming period that must ensue (21). Indeed, Grigore et al. (21), demonstrated that when compared to conventional “fast” rewarming, slower rewarming resulted in a lower incidence of neurocognitive dysfunction six weeks after cardiac surgery. These lower rewarming rates led to lower peak cerebral temperatures during rewarming, consistent with past observations that rapid rewarming can lead to an overshoot in cerebral temperature resulting in inadvertent cerebral hyperthermia. By reducing this rewarming rate, one reduces the overshoot in temperature and prevents the negative effects of cerebral hyperthermia. Consistent with the concept that preventing some of the rewarming may be protective was a study by Nathan et al. (22) that similarly demonstrated an avoidance of cerebral hyperthermia during rewarming and improvement of cognitive outcome after cardiac surgery.

There are several mechanisms by which hyperthermia may adversely affect the brain. Metabolically, cerebral hyperthermia has been demonstrated to increase intracellular acidosis after ischemic reperfusion; the recovery of ATP and other high-energy phosphates is also attenuated by hyperthermia (23). Sternau et al. have demonstrated that the release of neurotransmitters (in excitotoxic quantities) is accentuated by hyperthermia (24). A greater increase in oxygen-derived free radical production after hyperthermic reperfusion after global ischemia has also been demonstrated (7). Exaggerated increases in blood-brain barrier permeability occur during ischemia under hyperthermic conditions compared with normothermia (25). Additionally, hyperthermia during ischemia increases ischemic depolarizations in the peri-infarct region and, as a consequence, increases infarct size (26). Lastly, the cytoskeleton is senstitive to hyperthermia, with decreases in microtubule-associated protein (a cytoskeletal protein) that are due to calpain degradation (27).

Although a great deal of work has focused on trying to avoid hypothermia in the perioperative period, to prevent some of its associated complications (28), several studies have demonstrated not only the safety of hypothermia but also a neurologic benefit (22). This was demonstrated in the study from the Ottawa Heart Institute by Nathan et al. (22) which demonstrated a neurocognitive benefit for patients who were maintained between 34 and 36° for a prolonged (12 hours) period postoperatively. Studies by Kurz et al. (28) have eloquently outlined the potential serious side effects of hypothermia, such as perioperative infection. This may be particularly important in the setting of cardiac surgery where mediastinitis can be a very serious complication leading to increased morbidity and mortality. However, no cardiac surgery studies have addressed any links to perioperative temperature and infection. Most of the links between perioperative hypothermia and infection have been in the setting of general surgery such as bowel resection (28). Another concern regarding postoperative hypothermia relates to the potential for increased blood loss. Interestingly, despite an intuitive argument to the contrary, this has not been thus far demonstrated (29).

Although a great deal of interest and focus has been placed on intraoperative temperature, one must not ignore postoperative temperature. Several studies have linked postoperative hyperthermia to adverse outcomes. We have previously demonstrated that maximum postoperative temperature within 24 hours after cardiac surgery was associated with adverse neurocognitive outcome six weeks after cardiac surgery (30). In addition, a recent study has shown a further link between adverse outcomes and the peak temperature on admission to the intensive care unit (31). Although these studies suggest that postoperative hyperthermia is important, what they do not demonstrate is whether intervening to prevent postoperative hyperthermia can prevent some of the adverse complications seen with cardiac surgery.

In summary, much has been learned in the recent decades regarding optimizing perioperative thermoregulation. We have learned that although hypothermia may not be as beneficial as was once thought, hyperthermia is particularly detrimental. The likely best approach in this regard is to allow for some mild hypothermia but to prevent aggressive rewarming which may lead to an overshoot in cerebral temperature. In addition, additional work needs to be done to determine whether intervening to prevent postoperative hyperthermia can be protective. Perioperative thermoregulation remains a fruitful area for perioperative research.