| ||||||||||||||
|
|
|||||||||||||

Departments of *Anesthesiology (Division of Cardiothoracic Anesthesiology and Critical Care Medicine) and
Surgery (Division of Cardiothoracic Surgery), Duke University Medical Center, Durham, North Carolina
Address correspondence and reprint requests to Hilary P. Grocott, MD, Department of Anesthesiology, Duke University Medical Center, Box 3094, Durham, NC 27710. Address e-mail to h.grocott{at}duke.edu
| Abstract |
|---|
|
|
|---|
IMPLICATIONS: Cerebral hyperthermia during rewarming from cardiopulmonary bypass is associated with increased neurological injury. In this randomized, controlled study, we have shown that limiting the target rewarming temperature on bypass to 35°C, combined with continuous surface warming, can prevent cerebral hyperthermia without risking prolonged postoperative hypothermia.
| Introduction |
|---|
|
|
|---|
Cerebral ischemia, related either to global cerebral hypoperfusion or to macro- or micro-embolization, is a major pathogenic factor leading to neuronal dysfunction and death, with functionally apparent clinical damage after cardiac surgery (5). Numerous experimental models of cerebral ischemia have repeatedly demonstrated deleterious effects from hyperthermia and the protective effects of hypothermia (68). Therefore, hypothermia intuitively seems a logical technique to use in cardiac surgery during cardiopulmonary bypass (CPB). However, several studies have failed to confirm the putative protective effect of hypothermia during CABG (9,10). One possible explanation for this is that any neuroprotection afforded by hypothermia is counterbalanced by damage occurring during the rewarming process (11). Others and we have shown that brain temperature, as measured in the jugular bulb, frequently exceeds 38°C during rewarming (12,13). In addition, rewarming is associated with jugular bulb desaturation (13), which, in turn, has been associated with postoperative NCD (14,15). Cerebral hyperthermia can be prevented by either slowing the rewarming rate (at the cost of prolonging CPB time) or by limiting the rewarming temperature goal (at the risk of increasing postoperative hypothermia). This study sought to determine if the perioperative use of surface warming could allow limited rewarming from hypothermic CPB, hence reducing cerebral hyperthermia, while still avoiding prolonged postoperative hypothermia.
| Methods |
|---|
|
|
|---|
Anesthetic and surgical care was standardized between groups except for the main study intervention. After sedation with midazolam (12 mg) and fentanyl (50100 µg), invasive monitoring was established (radial and pulmonary artery cannulae). General anesthesia was induced with thiopental (35 mg/kg), fentanyl (710 µg/kg), and pancuronium (0.1 mg/kg) and maintained using isoflurane with additional doses of fentanyl and midazolam. Fluid warmers (HotlineTM; SIMS Inc, Rockland, MD) were used for infusing blood products and large amounts of fluid. Convective forced-air warming was not used. After systemic heparinization (to attain an activated clotting time >480 s), cannulation for CPB was performed via the ascending aorta, and CPB ensued using a nonpulsatile system with a membrane oxygenator. Patients were cooled to 33°C (nasopharyngeal temperature) during bypass. A 40-µm arterial line filter was routinely used. Crystalloid prime was used unless the preoperative hematocrit was <30%, when blood was added to maintain a hematocrit level during CPB
18%. Intermittent cold-blood cardioplegia was used for myocardial protection. During CPB, pump flows were kept at 22.4 L · min1 · m2 body surface area. Arterial blood pressure was maintained between 5090 mm Hg using phenylephrine, isoflurane, and sodium nitroprusside as required. Arterial partial oxygen pressure was maintained between 150250 mm Hg.
At the end of the operation, all patients were transported (approximate time, 5 min) to the intensive care unit (ICU) sedated (propofol 2050 µg · kg1 · min1), ventilated, and covered with a blanket. Hypothermic patients (either pulmonary artery catheter [PAC] or oral temperature <35.5°C) were warmed using convective forced-air warming blankets (Progressive Dynamics Medical, Inc, Marshall, MI). Acetaminophen (1000 mg orally or via nasogastric tube) was given to any patient with oral temperature more than 38°C. Fluid warmers were not used in the ICU.
Patients were randomized to either an active thermoregulation treatment group or a control group. Active thermoregulation consisted of surface cooling and warming using the Arctic Sun® thermoregulatory system (Medivance, Inc, Louisville, CO). This device works by circulating water through a set of heat-exchange pads placed in direct contact with the patients body. These specialized pads (Arctic Sun Energy transfer Pads®) are made of highly conductive biocompatible hydrogel material that adheres tightly to the patients skin. The control module of the Arctic Sun is servo-regulated by sensing both patient and water temperature and controls the water flow and temperature according to the set temperature and patient temperature. For the purpose of the current study, the device was connected to a separate nasopharyngeal temperature probe (Mallinkrodt Inc, St Louis, MO).
In patients randomized to the active treatment group, pads were applied to the back of the torso and posterolateral aspects of the upper legs (together comprising 25% of body surface area) before the induction of anesthesia and connected to the thermoregulatory control unit. The target patient nasopharyngeal temperature was set to 34°C to achieve a mild degree of hypothermia at the time of aortic cannulation. Cooling during CPB (to 33°C nasopharyngeal temperature) was identical for both groups, but during the rewarming phase, the control group was warmed to a nasopharyngeal temperature of 37°C using the heat exchanger intrinsic to the CPB apparatus, whereas the treatment group was warmed only to 35°C. Rewarming of the treatment-group patients continued through the first 4 h after separation from CPB using the Arctic Sun, with a set temperature goal of 36.8°C.
Near-infrared cerebral oximetry was monitored bilaterally (INVOS® Cerebral Oximeter; Somanetics Corporation, Troy, MI), and jugular bulb temperature and oxygen saturation were monitored with a 5.5F PAC (Oximetrix®, Abbott Inc, Abbott Park, IL) inserted retrogradely through the right internal jugular vein before the anesthesia induction. Jugular bulb temperature was used as a surrogate for brain temperature, and the measurements were used to define peak jugular temperature.
Nasopharyngeal and jugular bulb temperature data were manually recorded at specific milestone events (aortic cannulation, on and off CPB, aortic cross-clamp on and off, end of surgery, and 4 h after surgery). Additionally, nasopharyngeal temperature was measured continuously by the Arctic Sun device and recorded every minute from the start of surgery until disconnection of the device 4 h after surgery ended. These temperature data were integrated over the first 4 postoperative hours to quantify early postoperative hypothermia and hyperthermia by calculating the area under the curve for temperature <36°C and >37°C, respectively.
Core body temperature was measured during the first 24 h in the ICU using the PAC or, in its absence, by an oral thermometer. These data were used to define peak postoperative body temperature.
Data are presented as mean ± SD or median [IQR] as indicated. Groups were compared using Students t-test for normal continuous variables, Mann-Whitney U-test for non-normal continuous variables, and Fishers exact test for categorical dichotomous variables. Simple linear regression was employed to test the correlation between peak temperature and various continuous variables. P < 0.05 was considered significant, except for comparison of nasopharyngeal temperature between groups, where P < 0.006 was used to compensate for multiple (8) comparisons.
| Results |
|---|
|
|
|---|
|
|
|
|
|
| Discussion |
|---|
|
|
|---|
Two major mechanisms that are considered to contribute to neurological injury after cardiac surgery are macro- or micro-embolization and global cerebral hypoperfusion (5,16). Both embolization and hypoperfusion result in cerebral ischemia that, if persistent, leads to neuronal death and functionally apparent clinical damage. Many studies, both in experimental models of cerebral ischemia and in clinical stroke in humans, have shown deleterious effects of hyperthermia (68,17). The pathophysiological explanation for this deleterious effect probably relates to the existence of a zone of hypoperfused (and ischemic) but potentially viable tissue known as the penumbra. This area is vulnerable to secondary insults, and its fate (death or recovery) greatly influences the ultimate neurological outcome (18). Increased temperature is associated with an increased metabolic rate, potentially worsening the delicate balance between oxygen supply and demand existing in the ischemic penumbra. Hyperthermia is also linked with other potentially harmful effects as well, including accentuated release of toxic excitatory neurotransmitters, increased oxygen free-radical formation, blood-brain barrier disruption, cytoskeleton destruction, increased number of ischemic depolarizations, and increased neural intracellular lactic acidosis (19). Even mild degrees of hyperthermia can be harmful; animal models of cerebral ischemia found that temperature increase of as small as 1°C2°C was enough to increase damage (6). Against this background, it is plausible that cerebral hyperthermia, often accompanying the rewarming stage of CPB (12,13), may exacerbate neurological damage after cardiac surgery. Indeed, cerebral hyperthermia during CPB has been associated with jugular bulb desaturation (13), itself associated with postoperative NCD (14,15).
Theoretically, two approaches can be pursued to try to prevent rewarming-induced cerebral hyperthermia. One approach is to slow the rate of rewarming by limiting the difference between blood temperature and the water temperature in the CPB heat-exchanger apparatus. This approach reduced the incidence of jugular bulb desaturation (20) and postoperative NCD (11), although jugular or cerebral temperatures were not directly measured in these studies. A possible disadvantage of this approach is prolongation of the rewarming stage, which may increase overall CPB time with possible deleterious effects on morbidity after cardiac surgery.
The second approach to prevent cerebral hyperthermia is to limit the target rewarming temperature. Traditionally, separation from bypass is attempted after normothermia has been accomplished. Although this separation temperature varies among institutions, a nasopharyngeal temperature of 37°C would likely be a common temperature goal. However, brain temperature during the rewarming phase has been shown to be approximately 1°C3°C higher than the nasopharyngeal temperature (12,13), and a significant discrepancy remains well into the postoperative period (21). Limiting the rewarming nasopharyngeal temperature to 34°C35°C should therefore prevent cerebral hyperthermia. In fact, this approach was used in a randomized, controlled trial with a demonstrable improvement in neurocognitive function after CABG (22). The obvious drawback with this approach is the occurrence of mild hypothermia in the first several hours after separation from CPB, especially because separation from CPB is usually accompanied by a continued decrease in temperature (referred to as after-drop) of at least 0.5°C. This phenomenon was also observed in the control group in the present study (Fig. 1).
Perioperative hypothermia is associated with several adverse effects, including coagulopathy and an increase in myocardial oxygen consumption, both of which may be particularly harmful in the post-CPB period. Indeed, perioperative hypothermia in CABG patients has been specifically shown to be associated with increased mortality, time of mechanical ventilation, and transfusion requirements (23). Effective means of warming the patient gradually after separation from bypass and into the first several postoperative hours, as used in the current study, may ameliorate this hypothermia and its resulting complications. Separation from CPB at a lower body temperature is expected to result in a shorter bypass time, by itself a favorable consequence. Indeed, we found a trend towards a shorter CPB time in the treatment group (P = 0.06); however, the reduction in rewarming time was modest and not significantly different between groups. An alternative means to reduce CPB time would be to start the rewarming earlier. However, this strategy would likely require rewarming while the heart is still arrested and ischemic (before cross-clamp is removed), possibly increasing the risk of myocardial damage.
Whereas not the principal reason for doing the study, we undertook some measurements of cerebral oxygenation. Although cerebral venous desaturation has been repeatedly shown to coexist with cerebral hyperthermia during rewarming from hypothermic CPB, some studies suggest that the degree of desaturation is related to preoperative cerebral blood flow reserve (14) and anatomical brain abnormalities (24). Therefore, cerebral venous desaturation may be a sign of brain vulnerability and not a pathogenic factor such that its prevention may improve clinical outcome. We found neither a change in jugular bulb oxygen saturation accompanying the difference in jugular bulb temperature between the two groups, nor a difference in cerebral oximetry readings. However, the study was not powered to assess cerebral physiologic differences.
Our findings regarding safety end-points (Table 3) do not suggest any detrimental effect of the mild degree of hypothermia that occurred after separation from CPB and before reaching normothermia. Being a feasibility trial however, the current study was not powered to definitively prove or disprove the safety of the described approach. That said, both groups had similar body temperature at the time of admission to the ICU; therefore, the lack of difference in safety end-points is not unexpected. The blood transfusion rate was relatively high and time to tracheal extubation somewhat prolonged, although without a significant difference between the two groups. However, it should be noted that the patients enrolled in the current study were relatively low risk, being <70 years old, having only mildly reduced ejection fractions, and only requiring primary elective CABG. It is unknown if higher-risk patients may be more vulnerable to even short periods of mild hypothermia.
Possible adverse effects of the use of the Arctic Sun device itself include skin irritation, hypersensitivity reaction to the pads, or burns from the hot water circulating inside the pads. None of these was observed in our study, but more extensive experience is required to accurately quantify these theoretical risks.
Studies comparing hypothermic with normothermic bypass have shown conflicting results regarding neurocognitive outcome, with the most recent studies failing to show an advantage of hypothermic bypass (9,10,25). One possible explanation is that excessive rewarming after hypothermic bypass may induce cerebral hyperthermia, negating the beneficial effect of the previous hypothermic period. Limiting the rewarming temperature may thus allow the full realization of the theoretical benefits of hypothermic bypass. Combining this approach with efficient surface warming may prevent the risk of accompanying hypothermia. The current study has shown the feasibility of this approach; proving its safety, benefit, and applicability in the heterogeneous population of patients undergoing cardiac surgery awaits larger controlled clinical trials.
| Acknowledgments |
|---|
| Footnotes |
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
N. A. Nussmeier, W. Cheng, M. Marino, T. Spata, S. Li, G. Daniels, T. Clark, and W. K. Vaughn Temperature During Cardiopulmonary Bypass: The Discrepancies Between Monitored Sites Anesth. Analg., December 1, 2006; 103(6): 1373 - 1379. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. N. Djaiani Aortic arch atheroma: stroke reduction in cardiac surgical patients. Seminars in Cardiothoracic and Vascular Anesthesia, June 1, 2006; 10(2): 143 - 157. [Abstract] [PDF] |
||||
![]() |
A. M. Morariu, B. G. Loef, L. P. H. J. Aarts, G. W. Rietman, G. Rakhorst, W. van Oeveren, and A. H. Epema Dexamethasone: Benefit and Prejudice for Patients Undergoing On-Pump Coronary Artery Bypass Grafting: A Study on Myocardial, Pulmonary, Renal, Intestinal, and Hepatic Injury Chest, October 1, 2005; 128(4): 2677 - 2687. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. P. Grocott, H. M. Homi, and F. Puskas Cognitive Dysfunction After Cardiac Surgery: Revisiting Etiology Seminars in Cardiothoracic and Vascular Anesthesia, June 1, 2005; 9(2): 123 - 129. [Abstract] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|