The clinical applications for periods of hypothermic

Transcription

1 SESSION 4: AORTIC ARCH II Cerebral Metabolic Suppression During Hypothermic Circulatory Arrest in Humans Jock N. McCullough, MD, Ning Zhang, MD, David L. Reich, MD, Tatu S. Juvonen, MD, PhD, James J. Klein, MD, David Spielvogel, MD, M. Arisan Ergin, MD, PhD, and Randall B. Griepp, MD Department of Cardiothoracic Surgery, Mount Sinai Medical Center, New York, New York Background. Hypothermic circulatory arrest (HCA) is used in surgery for aortic and congenital cardiac diseases. Although studies of the safety of HCA in animals have been carried out, the degree to which metabolism is suppressed in patients during hypothermia has been difficult to determine because of problems with serial measurements of cerebral blood flow in the clinical setting. Methods. To quantify the degree of metabolic suppression achieved by hypothermia, we studied 37 adults undergoing operations employing HCA. Cerebral blood flow was estimated using an ultrasonic flow probe on the left common carotid artery, and cerebral arteriovenous oxygen content differences were calculated from jugular venous bulb and arterial oxygen saturations. Cerebral metabolic rates while cooling were then ascertained. The temperature coefficient, Q 10, which is the ratio of metabolic rates at temperatures 10 C apart, was determined. Results. The human cerebral Q 10 was found to be 2.3. The cerebral metabolic rate is still 17% of baseline at 15 C. If one assumes that cerebral blood flow can safely be interrupted for 5 min at 37 C, and that cerebral metabolic suppression accounts for the protective effects of hypothermia, the predicted safe duration of HCA at 15 C is only 29 min. Conclusions. The safe intervals calculated from measured cerebral oxygen consumption suggest that shorter intervals and lower temperatures than those currently used may be necessary to assure adequate cerebral protection during hypothermic circulatory arrest. (Ann Thorac Surg 1999;67:1895 9) 1999 by The Society of Thoracic Surgeons The clinical applications for periods of hypothermic circulatory arrest (HCA) during cardiovascular surgical procedures as well some neurosurgical operations have increased with time. In operations on the thoracic aorta, an open distal anastomosis has become an essential part of many techniques and, by definition, requires circulatory exclusion of the cerebrovascular tree [1], which in most instances can efficiently be accomplished by an interval of HCA. Despite a long history of successful clinical use, probably originating with Charles Drew in 1959 [2] and popularized by Griepp and associates [3], the technique carries with it the threat of severe complications from somewhat unpredictable and as yet poorly defined neurological injury. It has been shown that a relationship exists between length of circulatory arrest and temporary neurological dysfunction in adults and intelligence quotient in children [1]. As a consequence of technical problems with reliable determination of cerebral blood flow [4] and other methodological limitations, the physiology of HCA in humans, and especially in adults, has not been well characterized, particularly over profoundly hypothermic temperature ranges. Presented at the Aortic Surgery Symposium VI, April 30 May 1, 1998, New York, NY. Address reprint requests to Dr McCullough, Department of Cardiothoracic Surgery, Mount Sinai Medical Center, Box 1028, One Gustave L. Levy Place, New York, NY The exact mechanism of cerebral protection during hypothermia is not well understood [5]. It is widely assumed that at least a portion of the protection afforded by hypothermia is secondary to metabolic suppression. Because the brain is dependent upon oxygen, at all times requiring aerobic glycolysis [7], the measurement of the cerebral metabolic rate for oxygen (CMRO 2 ) gives useful information as to the metabolic state of the brain. The temperature coefficient Q 10 is a well-described physiologic variable describing the temperature-dependent decrease in cerebral metabolism: the Q 10 for CMRO 2 is defined as the ratio of two CMRO 2 measurements differing by 10 C [6]. The Q 10 thus allows estimation of the oxygen requirement at a given hypothermic temperature, and is useful in enabling assessment of whether oxygen demand has been reduced to a level low enough to permit an extended interval of absent cerebral blood flow [6]. This review of our clinical cerebral blood flow and metabolism database was undertaken to characterize the CMRO 2 in an adult population undergoing operations on the thoracic aorta utilizing profoundly hypothermic temperatures and an interval of HCA. The study was designed to identify the Q 10 for adult human brains, in order to clarify the temperature dependence of cerebral metabolism and relate it to the clinical use of deep hypothermic circulatory arrest (DHCA) by The Society of Thoracic Surgeons /99/$20.00 Published by Elsevier Science Inc PII S (99)

2 1896 McCULLOUGH ET AL Ann Thorac Surg CEREBRAL METABOLISM DURING HYPOTHERMIA 1999;67: Patients and Methods Clinical Protocol Thirty-seven patients undergoing nonemergent operations on the ascending aorta or arch utilizing an interval of DHCA between January 31, 1996 and April 16, 1997 were included in the analysis. Anesthesia was induced and maintained with high-dose opioids, 100% oxygen, and muscle relaxant technique, supplemented with isoflurane and midazolam as needed to maintain hemodynamic stability. Intraoperative monitoring included an intraarterial catheter, a pulmonary artery catheter placed via an internal jugular vein, a3frjugular bulb venous catheter, capnometry, and esophageal and bladder temperature probes. After median sternotomy and initial dissection, an ultrasonic flow probe (Transonic Systems Inc, Ithaca, NY) was placed around the left common carotid artery. Baseline samples were drawn before the institution of cardiopulmonary bypass with the mean arterial pressure maintained at 80 mm Hg by infusion of volume if needed. The subjects were then placed on total cardiopulmonary bypass (CPB) utilizing a membrane oxygenator with nonpulsatile pump flows of L/min, and cooled to an esophageal temperature of 15 C. While on CPB, arterial tension of carbon dioxide (PaCO 2 ) was managed without correction for hypothermic temperatures, the -stat strategy. Mean arterial blood pressure (MAP) was maintained between 40 and 70 mm Hg using sodium nitroprusside or phenylephrine infusions as needed. The patients were maintained at 15 C for variable intervals of time while necessary surgical repairs were completed. A period of cooling with a blood perfusate temperature of 10 C was begun (hard cooling) once the majority of the proximal repair had been completed and the interval of circulatory arrest was less than 30 minutes away. Physiological measurements were recorded just before circulatory arrest, and at several points during reperfusion and systemic rewarming. After separation from CPB, protamine was administered and volume was used to maintain a stable MAP. Final physiologic measurements were made while off CPB and stable. Cerebral physiologic data, including cerebral blood flow (CBF), arterial blood gases (ABG), mixed venous gases (MVG), jugular venous gases (JVG), and cardiac output were obtained at nine points: 1) at baseline, before CPB and cooling (baseline); 2) 10 minutes after the onset of CPB and systemic cooling (cooling-10 minutes); 3) at an esophageal temperature of 15 C (at 15 C); 4) before the onset of circulatory arrest (before DHCA); 5) 10 minutes after the resumption of CPB and systemic rewarming; 6) 30 minutes after the resumption of CPB and systemic rewarming; 7) 45 minutes after the resumption of CPB and systemic rewarming; 8) 60 minutes after the resumption of CPB and systemic rewarming; and 9) after discontinuation of CPB. Calculations The cerebral metabolic rate for oxygen (CMRO 2 ) was estimated with the equation: CMRO 2 CBF cerebral Table 1. Demographic Data Variable arteriovenous oxygen content difference/100. The Q 10 was then derived for the population: 1) The natural logarithm for the CMRO 2 (lncmro 2 ) was plotted against temperature while cooling for each patient, and the slope of the resultant line was found. 2) The mean slope for the population (m) was then identified. 3) Therefore, for any patient p, lncmro 2(p) a (p) m(t), where t temperature and a (p) patient unique intercept. 4) The Q 10 can then be solved as the ratio of the CMRO 2 at a temperature (t) and at t 10. Values for CBF and CMRO 2 were transformed as proportions of baseline values for further analysis by assuming their value at 37 C to represent 100%. Statistical analysis was performed with a Friedman repeated measures analysis of variance on ranks followed by Dunn s test as appropriate. All values are reported as means standard deviation. A p value 0.05 was considered significant. Results Value Age (yrs) Cross-clamp time (min) Cardiopulmonary bypass (min) Circulatory arrest (min) Minimum esophageal temperature ( C) Male (%) 73% Discharged alive (%) 92% The demographic characteristics of the population are reported in Table 1. An interval of selective cerebral perfusion was used in 5 (14%) patients for an average of 37 minutes. A brief interval of retrograde perfusion to flush air and debris from the open aorta was used in 6 (16%) patients for an average interval of 7 3 min. A more prolonged period of retrograde cerebral perfusion was used in 13 (35%) of patients, with an average duration of 25 7 min. The physiologic variables and cerebral metabolic results are reported in Tables 2 and 3. While cooling, CBF did not change significantly from baseline values until just before circulatory arrest, and was consistently higher, although not significantly different from baseline, during rewarming (Tables 2 and 3). The CMRO 2 was significantly lower than baseline throughout the cooling interval, and did not return to baseline values until after the first determination during rewarming, between 10 and 30 min (Tables 2 and 3). The CMRO 2 also showed a consistent but statistically insignificant increase over baseline values during the later intervals of rewarming. The average slope of the population for the plot of lncmro 2 versus temperature was , which yields a temperature coefficient (Q 10 ) of 2.3, with 95% confidence intervals of The Q 10 was also calculated for two different cooling phases: the Q 10 for cooling

3 Ann Thorac Surg McCULLOUGH ET AL 1999;67: CEREBRAL METABOLISM DURING HYPOTHERMIA 1897 Table 2. Physiologic Variables During Cooling Variable Baseline Cooling: 10 min At 15 C Before DHCA Esoph Temp ( C) a a a MAP (mm Hg) a 47 8 a 50 8 a Hematocrit a 24 4 a 26 4 a CO (L/min) a a a CBF (% Baseline) a CMRO 2 (% Baseline) a a a a Identifies significant change from baseline values by Friedman repeated measures analysis of variance followed by Dunn s test as appropriate ( p 0.05). Before DHCA are values measured just before the start of deep hypothermic circulatory arrest (DHCA). CBF cerebral blood flow expressed as proportion of baseline value; CMRO 2 cerebral metabolic rate for oxygen expressed as a proportion of value extrapolated back to 37 C; CO cardiac output when off cardiopulmonary bypass and pump flow while on cardiopulmonary bypass; Esoph Temp esophageal temperature; MAP mean arterial pressure. from baseline to 15 C was 2.05, and from 15 C through the period of hard cooling to a mean esophageal temperature of C, it was 3.5. This difference in Q 10 was not significant by Wilcoxon test ( p 0.38), or t test ( p 0.19). Using the Q 10 derived from this patient population, and assuming that CBF can be safely interrupted for 5 min at 37 C (and that the only mechanism for cerebral protection during hypothermia is metabolic suppression), a calculated safe duration of HCA can be derived for any temperature, as shown in Table 4. Comment We found Q 10, the metabolic coefficient for the adult human brain during cooling to profoundly hypothermic temperatures, to be 2.3. The relationship between CMRO 2 and temperature best fit a log-linear model over the temperature range investigated. The theoretical safe duration for hypothermic circulatory arrest derived from our data fit closely with observed clinical experience [1, 8], implying that hypothermic metabolic suppression is a major protective mechanism for the brain during intervals of circulatory arrest. Assuming that the optimal relationship between flow and metabolism is that which prevails at baseline, there appears to be an uncoupling of the relationship between cerebral blood flow and metabolism both during cooling to and rewarming from profoundly hypothermic temperatures; this phenomenon has previously been noted [9, 10]. One remarkable finding of the current study is that the CMRO 2 is still 24% of baseline values at 20 C, and 16% of baseline at 15 C, temperatures often utilized for HCA [12, 13]. As shown in Table 4, temperatures in this range may be adequate for durations of circulatory arrest less than 30 min, but a lower metabolic rate is required for more prolonged periods of HCA, mandating the use of lower temperatures when long intervals of interrupted antegrade circulation are anticipated. It should be noted that the average time elapsed from the onset of cooling until DHCA was 75 min; we agree strongly with others [14] that prolonged cooling to ensure uniform cerebral hypothermia is critical to the successful use of DHCA. The use of a jugular bulb catheter to quantify the cerebral mixed venous oxygen saturation is being reported with increasing frequency [9, 11, 13, 15, 16]. We feel this to be the best currently available clinical guide to cerebral metabolic activity and therefore the best indicator of the adequacy of cooling. Another technique commonly used to determine whether cerebral metabolism has been suppressed is the electroencephalogram (EEG) [17], but we have noted disappearance of cerebral electrical activity while jugular bulb mixed venous values were still low [11], implying significant residual cerebral Table 3. Physiologic Variables During Warming RW: 10 min RW: 30 min RW: 45 min RW: 60 min Off CPB Esoph Temp ( C) a a MAP (mm Hg) a a Hematocrit 25 3 a 28 3 a a a CO (L/min) a a CBF (% Baseline) CMRO 2 (% Baseline) a a Identifies change from baseline value by Friedman repeated measures analysis of variance followed by Dunn s test as appropriate ( p 0.05). CBF cerebral blood flow expressed as proportion of baseline value; CPB cardiopulmonary bypass; CMRO 2 cerebral metabolic rate for oxygen expressed as a proportion of value extrapolated to 37 C; CO cardiac output when off cardiopulmonary bypass and pump flow while on cardiopulmonary bypass; Esoph Temp esophageal temperature; MAP mean arterial pressure; RW elapsed time since resumption of cardiopulmonary bypass and systemic rewarming after circulatory arrest.

4 1898 McCULLOUGH ET AL Ann Thorac Surg CEREBRAL METABOLISM DURING HYPOTHERMIA 1999;67: Table 4. Calculated Safe Duration of Hypothermic Circulatory Arrest Temperature ( C) Cerebral Metabolic Rate (% of baseline) Safe Duration of HCA (min) (52 60) 9 (8 10) (33 42) 14 (12 15) (21 29) 21 (17 24) (13 20) 31 (25 38) (8 14) 45 (36 62) Calculations based on assumption that there is a 5-min tolerance for circulatory arrest at 37 C. Values in parenthesis are 95% confidence intervals. HCA hypothermic circulatory arrest. activity, and therefore we question the accuracy of EEG silence in identifying adequate suppression of cerebral metabolism before embarking on a period of circulatory arrest. The observation of a delay in return of cerebral metabolism to baseline values during rewarming, in contrast to the very prompt recovery of cerebral blood flow, is worthy of comment. This phenomenon has previously been reported during rewarming after CPB with moderate hypothermia without circulatory arrest [9]. During neurosurgical procedures involving an interval of HCA [5, 10], cerebral tissue PO 2 was observed to decrease to near zero after 30 minutes of circulatory arrest [10], with a significant delay in the return of cerebral tissue PO 2 after the onset of rewarming after circulatory arrest, despite the return of cerebral blood flow and oxygen delivery to normal values within a few minutes after the onset of rewarming [5, 10]. These observations suggest that there may be a period of cerebral vulnerability after circulatory arrest while tissue PO 2 and cerebral metabolic rate for oxygen slowly return to baseline after a severe oxygen deficit. An alternative possibility is that cerebral blood flow is inappropriately high for brain temperature during early perfusion, a phenomenon that has been termed luxury perfusion, and that the metabolic rate is temperature appropriate. The identification of the Q 10 for adult humans as 2.3 is similar to what was found in a canine experimental model, in which a Q 10 of 2.23 was calculated while cooling from 37 C to 27 C [6]. In the canine study, the investigators found the relationship between temperature and CMRO 2 to be nonlinear, with a Q 10 of 4.54 between 27 C and 14 C. We noted a similar change in the slope of the relationship between CMRO 2 and temperature for the range from 37 C to 15 C compared with 15 C to 11 C, but this difference was not significant; the Q 10 for the warmer range was 2.05, and for the cooler temperatures was 3.5. Our studies in a pig model using two different microsphere methods to calculate cerebral blood flow over a temperature range from 37 C to 8 C resulted in calculations of Q 10 of 2.3 and 2.7, very close to the values of the current study. Higher values have been found by others, however. In one report, using a monkey model and a temperature range from 37 C to 19 C, a Q 10 of 3.53 was calculated [18], and in a clinical study in patients aged 1 day to 14 years cooled to 18 C, a Q 10 of 3.65 was found [13]. Taken together, it seems possible from these data that a different Q 10 value may exist at the cooler extremes of temperature, and that this difference might have emerged as significant in our study if a larger population had been examined. But if the relationship between metabolism and temperature is linear, and metabolic suppression is the major mechanism for cerebral protection during cooling before HCA, then the tolerance for circulatory arrest based on a Q 10 of 2.3 would be only 26 min at 18 C, assuming a safe duration for circulatory arrest of 5 min at 37 C. Although, on the basis of clinical neurological recovery, we believe that tolerance for HCA may be somewhat higher than these safety estimates, recent studies using sophisticated neuropsychological testing suggest that measurable although possibly transient damage does occur after 30 min of HCA as usually implemented clinically [19, 20]. The fact that our experience with circulatory arrest using deep hypothermia for longer intervals has generally been favorable may be a reflection of the fact that there is somewhat more profound metabolic suppression during cooling below 15 C than is estimated using this Q 10 derived at higher temperatures. In conclusion, our data support the use of truly profound hypothermia before a period of circulatory arrest to achieve maximal metabolic suppression and cerebral protection. We believe that the duration of HCA should be less than 21 min at an esophageal temperature of 20 C, and that the patient should be cooled toward an esophageal temperature of 10 C to 11 C if a period of HCA exceeding 30 min is anticipated. While this represents a more profound level of hypothermia than is often clinically employed, we believe its use is supported by our recent findings showing impaired performance on sophisticated neuropsychiatric testing after hypothermic circulatory arrest. References 1. Ergin MA, Griepp EB, Lansman SL, Galla JD, Levy M, Griepp RB. Hypothermic circulatory arrest and other methods of cerebral protection during operations on the thoracic aorta. J Card Surg 1994;9: Dobell AR, Bailey JS. Charles Drew and the origins of hypothermic circulatory arrest. Ann Thorac Surg 1997;63: Griepp RB, Stinson EB, Hollingsworth JF, Buehler D. Prosthetic replacement of the aortic arch. J Thorac Cardiovasc Surg 1975;70: Michenfelder JD, Messick JM, Theye RA. Simultaneous cerebral blood flow measured by direct and indirect methods. J Surg Res 1968;8: Hoffman WE, Charbel FT, Munoz L, Ausman JI. Comparison of brain tissue metabolic changes during ischema at 35 and 18 C. Surg Neurol 1998;49: Michenfelder JD, Milder JH. The relationship among canine brain temperature, metabolism, and function during hypothermia. Anesthesiology 1991;75: Powers WJ, Grubb RL, Darriet D, Raichle ME. Cerebral blood flow and cerebral metabolic rate of oxygen requirements for cerebral function and viability in humans. J Cereb Blood Flow Metab. 1985;4:600 8.

5 Ann Thorac Surg McCULLOUGH ET AL 1999;67: CEREBRAL METABOLISM DURING HYPOTHERMIA Ergin MA, Galla JD, Lansman SL, et al. Hypothermic circulatory arrest in operations on the thoracic aorta. Determinations of operative mortality and neurologic outcome. J Thorac Cardiovasc Surg 1994;107: Cook DJ, Oliver WC, Orszulak TA, et al. Cardiopulmonary bypass temperature, hematocrit, and cerebral oxygen delivery in humans. Ann Thorac Surg 1995;60: Hoffman WE, Charbel FT, Edelman G, Ausman JI. Brain tissue oxygen pressure, carbon dioxide pressure, and ph during hypothermic circulatory arrest. Surg Neurol 1996;46: Griepp RB, Ergin MA, McCullough JN, et al. Use of hypothermic circulatory arrest for cerebral protection during aortic surgery. J Card Surg 1997;12: Ausman JI, McCormick PW, Stewart M, et al. Cerebral oxygen metabolism during hypothermic circulatory arrest in humans. J Neurosurg 1993;79: Greeley WJ, Kern FH, Ungerleider RM, et al. The effect of hypothermic cardiopulmonary bypass and total circulatory arrest on cerebral metabolism in neonates, infants, and children. J Thorac Cardiovasc Surg 1991;101: Greeley WJ, Kern FH, Meliones JN, Ungerleider RM. Effect of deep hypothermia and circulatory arrest on cerebral blood flow and metabolism. Ann Thorac Surg 1993;56: Cook DJ, Oliver WC, Orszulak TA, Daly RC. A prospective, randomized comparison of cerebral venous oxygen saturation during normothermic and hypothermic bypass. J Thorac Cardiovasc Surg 1994;107: Silvay G, Reich DL, Vela-Cantos F. Jugular bulb oxygen saturation in surgery requiring deep hypothermic circulatory arrest. (abst) Brit J Anaesth 1995;74(Suppl. 2): Coselli JS, Crawford ES, Beall AC, et al. Determination of brain temperature for safe circulatory arrest during cardiovascular operation. Ann Thorac Surg 1988;45: Bering EA. Effect of body temperature change on cerebral oxygen consumption of the intact monkey. Am J Physiol 1961;200: Reich DL, Ergin MA, Uysal S, et al. Neuropsychological outcome after deep hypothermic circulatory arrest. J Thorac Cardiovasc Surg 1998;117: Oates RK, Simpson JM, Turnbull JAB, Cartmill TB. The relationship between intelligence and duration of circulatory arrest with deep hypothermia. J Thorac Cardiovasc Surg 1995;110: