A prospective, randomized comparison of cerebral venous oxygen saturation during normothermic and hypothermic cardiopulmonary bypass

Transcription

1 A prospective, randomized comparison of cerebral venous oxygen saturation during normothermic and hypothermic cardiopulmonary bypass Recent reports have described cerebral venous oxygen desaturation during and after rewarming from hypothermic cardiopulmonary bypass. Additionally, patients undergoing normothermic cardiopulmonary bypass may be at higher risk for neurologic injury. This study was designed to determine whether patients undergoing normothermic cardiopulmonary bypass are at increased risk for sustained cerebral desaturation. Fifty-two patients undergoing first-time coronary artery bypass grafting were randomized to receive normothermic (37 0 C, n = 26) or hypothermic (27 0 C, n = 26) cardiopulmonary bypass. The anesthetic was standardized and alpha-stat ph management was used. A 4F oximetric catheter was placed in the jugular bulb and cerebral venous and radial arterial blood were sampled. Oxygen partial pressure and saturation were measured at six intervals from cerebral venous blood and from radial arterial blood. Patients receiving normothermic cardiopulmonary bypass had lesser values of oxygen partial pressure and saturation in cerebral venous blood than patients subjected to hypothermia during the first 40 minutes of bypass. Cerebral venous desaturation (oxygen saturation in cerebral venous blood of 50 % or less) was observed in 54 % of patients in the normothermic group and 12 % of patients in the hypothermic group during cardiopulmonary bypass. In the normothermic group, cerebral desaturation occurred primarily in early bypass (14 of 26). The three episodes of desaturation in the hypothermic group occurred during rewarming. During cardiopulmonary bypass, the arteriovenous oxygen content difference was greater in the normothermic group than in that in the hypothermic group, suggesting higher oxygen consumption. Differences in glucose utilization during early cardiopulmonary bypass between the groups was also detected. One patient in the hypothermic group had an embolic stroke and subsequently died. There were no other perioperative strokes or deaths in the study population. The present study demonstrates that patients undergoing normothermic cardiopulmonary bypass are at greater risk for cerebral desaturation. Because it is a global assessment, cerebral venous oxygen saturation may be insensitive to focal ischemic events. It remains to be seen whether these differences in cerebral physiologic states translate into differences in clinical outcome. (J THoRAc CARDIOVASC SURG 1994;107:1020-9) David J. Cook, MD, William C. Oliver, Jr., MD, ThomasA Orszulak, MD, and Richard C. Daly, MD, Rochester, Minn. Util recently, most cardiac surgical procedures that required cardiopulmonary bypass (CPB) were done in the presence of moderate systemic hypothermia (27 0 to 30 From the Department of Anesthesiology and the Section of Cardiothoracic Surgery-Department of Surgery, Mayo Clinic, Rochester, Minn. Read at the Nineteenth Annual Meeting of The Western Thoracic Surgical Association, Carlsbad, Calif., June 23-26, Address for reprints: David J. Cook, MD, Department of Anesthesiology, Mayo Clinic, 200 First St. SW, Rochester, MN by Mosby-Year Book, Inc /94 $ /6/52965 C). Hypothermia reduces tissue metabolic rate and oxygen demand and is used to assist in the protection of the myocardium and brain against potential ischemic insult. l 2 Because of reports of improved myocardial performance with warm cardioplegia 3,4 and systemic normothermia,s CPB with warm cardioplegia and systemic normothermia is being increasingly used at our institution and elsewhere. Patients undergoing normothermic bypass have a predictably higher cerebral oxygen demand 6 and may be at higher risk for prolonged cerebral desaturation and ischemic injury7 than has been reported with hypothermic CPB. 8 - l6 1020

2 Volume 107, Number 4 Cook et al Table I. Characteristics of the normothermic and hypothermic study groups (26 patients per group) Crossclamp CPB Age time time No. of (yr) (min) (min) grafts Normothermic group Mean SD Hypothermic group Mean SD SD, Standard deviation. There is evidence that the coupling of cerebral blood flow (CBF) to metabolism is largely preserved during hypothermic CPB using an alpha-stat carbon dioxide management scheme. 17, 18 Also when arterial oxygen tension is adequate, cerebral venous oxygen saturation (Sjv02) reflects the global balance of CBF and cerebral metabolic rate (CMR02) Three studies suggest that CMR02-CBF imbalance occurs during and after warming from hypothermic CPB, with jugular venous desaturation occurring in up to 23% of patients If we presume that CBF is relatively constant during CPB, cerebral desaturation may reflect a rise in cerebral oxygen demand during rewarming. The report of an inverse correlation between Sjv02 and nasopharyngeal temperature would support that conclusion. 22 Jugular venous desaturation may become clinically significant when saturation reaches approximately 50%. During carotid operations, previous investigations found transient neurologic defects with cerebral venous saturations of 50% and no defects when saturations were 60% or better Meyer and colleagues 27 correlated Sjv02 and electroencephalographic changes in normal volunteers and showed consistent electroencephalographic slowing when the cerebral venous partial pressure of oxygen fell to 19 mm Hg (corresponding approximately to an Sjv02 of 40%). Although neurophysiologic changes begin to occur at an Sjv02 of approximately 50%, it is not known how long cerebral desaturation must be maintained for cerebral injury to occur. There have been reports suggesting that systemic normothermia with CPB may be associated with a higher prevalence of neurologic injury,?,28 The aim of this study was to assess and compare the coupling of cerebral metabolism and CBF in patients undergoing normothermic and hypothermic CPB as indicated by cerebral venous oximetry. Methods After approval by our Institutional Review Board, 52 patients undergoing elective first-time coronary artery bypass grafting Fig. 1. Radiograph demonstraiing correct catheter position in jugular bulb. Injection of contrast medium obscures catheter but demonstrates venous anatomy. were studied after giving informed consent. Patients were equally randomized into two groups: hypothermic (27 0 C) and normothermic (37 0 C) CPB. Female patients of childbearing age; patients with clinical or laboratory evidence of cerebrovascular disease, increased intracranial pressure, or insulin-dependent diabetes mellitus; and those with uncontrolled hypertension, contrast allergy, or demonstrated or suspected bacteremia were excluded from the study. Routine carotid ultrasonography and oculoplethysmography were not done. The anesthetic was standardized. Patients were premedicated with diazepam (5 to 10 mg orally) 90 minutes before induction of anesthesia. On arrival of the patient in the operating room, a radial artery catheter was placed and anesthesia was induced with fentanyl citrate (30 j.lgjkg intravenously) and midazolam hydrochloride (0.1 mgjkg intravenously) and maintained by fentanyl-midazolam infusion. Muscular relaxation was achieved and maintained with vecuronium or pancuronium, and the trachea was intubated. The lungs were ventilated with 40% to 80% oxygen and arterial carbon dioxide tension (Paco2) was maintained within the normocapnic range, uncorrected for temperature. A pulmonary artery catheter was placed. After anesthetic induction, a 4F oximetry catheter (Opticath, Abbott Laboratories, North Chicago, Ill.) was passed to the jugular bulb retrogradely from the internal jugular vein for continuous monitoring of Sjv The catheter was calibrated in vitro before insertion and recalibrated in vivo if necessary. Proper catheter positioning in the jugular bulb was confirmed by fluoroscopy (radiographic image intensifier) and contrast injection (Fig. 1). Heart rate; arterial, pulmonary artery, and right atrial blood pressures; end tidal carbon dioxide, and nasopharyngeal temperature were continuously measured along with Sjv02.

3 1022 Cook et al. April 1994 Sternotomy min Period I II III IV V VI I X i i i X Onset CPR End CPR I, "warm CPB" ' 60 ' (min on CPB) "cold CPB" (degrees C) Fig. 2. Schema describing six study periods. Periods I, V, and VI were identical in the two groups. During CPB, periods II, III, and IV were defined by bypass time in normothermic group and by bypass temperature in hypothermic group.. CPB was established and a nonpulsatile pump flow rate of 2.0 to 2.5 L. min -\. m - 2 was maintained. A membrane oxygenator and arterial line filter were used. Paco2 was adjusted to normocapnic levels (35 to 40 mm Hg) without temperature correction by varying membrane oxygenator gas flow (alphastat regulation).33 The bypass pump was primed to maintain a hematocrit value of 0.20 or greater during the bypass period. Management of hypothermic and normothermic bypass did not differ with the exception of temperature (hypothermic 27 C, normothermic 37 C). Sodium nitroprusside or phenylephrine infusions were used during CPB to maintain mean arterial pressures of 45 to 65 mm Hg. Nasopharyngeal temperature was measured continuously as an indicator of brain temperature. Two surgeons with similar techniques operated on all patients. Coronary artery bypass grafting was done with aortic crossclamping, and intermittent antegrade and retrograde normothermic blood cardioplegia was administered between end distal anastomoses. Proximal anastomoses were done after all distal anastomoses and cardiac reperfusion during CPB. All patients were weaned from CPB without mechanical support and none required epinephrine infusion. Arterial and jugular bulb venous blood samples were obtained for determination of blood gases, oxygen saturation (IL4-286 Co-Oximeter Instrumentation Laboratories, Inc., Boston, Mass.), and glucose concentration during six periods (Fig. 2). In both groups, prebypass baseline measurements were made 15 minutes after sternotomy (period I) and postbypass measurements were made at 15 and 45 minutes after weaning from CPB (periods V and VI, respectively). In the hypothermic bypass group, samples were also taken at 32 C during cooling (period 11), during stable hypothermia at 27 C (period III), and at 32 C during rewarming (period IV). In the normothermic bypass group samples for periods II, III, and IV were taken at 20, 40, and 60 minutes of CPB (Fig. 2). Arterial-to-venous oxygen content difference (A VDOz) and arteriovenous glucose difference (A VDgl u ) were calculated from measured values. Equations are given as follows: AVD02= (CaO] - Cv02) mi dl l CX02 = Hb(SX02) (Pxo]) AVDglu = (Aglu - Vglu) mg dl- l where CX02 is blood oxygen content (venous or arterial), SX02 is blood oxygen saturation, PX02 is partial pressure of oxygen, Hb is hemoglobin concentration, and Aglu and V glu are arterial and venous glucose concentrations. Temperature and hemodynamic variables were monitored at the same six stages. Cardiac index was measured in each of the non bypass periods (I, V, VI) and pump flow was recorded in the CPB periods (II, III, IV). No formal neurologic or neuropsychologic testing was done to compare clinical outcome in the two groups. Protocol and statistical analysis. The 52 subjects were equally divided into normothermic and hypothermic CPB groups by a block randomization method. One subject in the normothermic group had an extracorporeal circulation time less than 60 minutes, so there are only 25 data points in study period IV in the normothermic group. Additionally, only 25 data points in study period V in the hypothermic group were available. Collection of the data as outlined in the preceding section allowed the following comparisons to be made: (I) Sjvoz, arterial glucose, hemoglobin, arterial-to-venous oxygen content difference, and arteriovenous glucose difference in normothermic and hypothermic groups in each of the six study periods; (2) the proportion of patients in each group demonstrating cerebral venous desaturation (Sjv02 less than or equal to 50%) during each of the three CPB periods; (3) examination of saturation trends in each group during CPB; and (4) correlation of physiologic variables (arterial hemoglobin, mean arterial pressure, cardiac index, and arterial carbon dioxide) and cerebral venous desaturation during each period in both normothermic and hypothermic CPB groups. Data are expressed as mean plus or minus the standard deviation of the mean. Differences from baseline were assessed by the Wilcoxon signed rank test. Comparison of mean values between normothermic and hypothermic groups was done by the Wilcoxon rank sum test. The Wilcoxon rank sum test was also used to assess whether Sjv02 differed with respect to physiologic variables. The prevalence of desaturation in the two groups during CPB was assessed with a comparison of two proportions test. Differences in mean values were considered significant at p ::s Results Groups did not differ with respect to age, crossclamp time, bypass time, or number of grafts done (Table I). In the preby-

4 Volume 107, Number 4 Cook et al Table II. Physiologic variables describing both patient groups during the six study periods Nasopharyngeal temperature (0 C) Hypothermic group 35.7 ± 0.5 Normothermic group 35.5 ± 0.7 Mean arterial pressure (mm Hg) Hypothermic group 84.6 ± 14.5 Normothermic group 84.3 ± 13.2 Cardiac index/pump flow (L. min-i. m- 2 ) Hypothermic group 2.5 ± 0.8 Normothermic group 2.5 ± 0.7 Hemoglobin (gm. dl- 1 ) Hypothermic group 12.0 ±'1.7 Normothermic group 12.1 ± 1.4 Glucose (mg. dl- 1 ) Hypothermic group 151 ± 36.2 Normothermic group 145 ± 32.6 }>ao2 (mm Hg) Hypothermic group 239 ± 105 Normothermic group 296 ± 120. Paco2 (mm Hg) Hypothermic group 34.5 ± 5.5 Normothermic group 32.2 ± 3.4 Values are mean ± standard deviation. *p < 0.01 between group difference (rank sum test). tp < I II 31.8 ± ± ± ± ± ± ± ± ± ± ± ± ± ± 3.5 CPS Study periods III IV V VI 27.3 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 41.4* 171 ± 51.9t 170 ± 52.0t 128 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 3.6 pass baseline period (period I), the normothermic and hypothermic groups did not differ as to mean arterial pressure, cardiac index, temperature, hemoglobin, glucose, Paco2, arterial oxygen tension (Pao2, or Sjv02 (Table II;p > 0.05 by Student's t test). During CPB, cooling of the hypothermic group was associated with steady increases in cerebral venous oxygen saturation as indicated by continuous oximetry (Fig. 3, A). Conversely, initiation of CPB in the normothermic group was frequently associated with either an unchanged or decreasing Sjv02 as recorded by continuous oximetry (Fig. 3, B). Although continuous oximetry provided useful trends (with episodes of desaturation on initiation of bypass typically lasting 10 to 20 minutes), the instantaneous output is subject to motion artifact so only saturations obtained by direct blood gas measurement are reported. Jugular bulb venous samples drawn during study period II confirmed the trends demonstrated by continuous oximetry. In the hypothermic group, cooling to 32 C was associated with an increase in saturation from 54% ± 12% at baseline (period I) to 68% ± 10% at 32 C (period II). Sjv02 in the normothermic and hypothermic groups differed (52% versus 68%) at a significance level of p < by the rank sum test (Fig. 4). In the normothermic group, although the decrease in saturation after 20 minutes of bypass (52% ± 13%) was not statistically differentfrom baseline levels (53.9% ± 12.6%), 14 of 26 patients had a measured Sjv02 of 50% or less in period II versus none of 26 hypothermic patients; p s by comparison of two proportions (Fig. 5). Patients in the hypothermic group, in study period III (27 C), showed a further increase in Sjv02 (73% ± to 8%) whereas patients in the normothermic group in period III (40 minutes of CPB) showed the beginning of a trend toward increasing Sjv02 (56% ± 10%; Fig. 4). In study period III, mean Sjv02 values in normothermic and hypothermic patients differed at the level of p < by rank sum test. During period III, 8 of 26 patients in the normothermic group had a measured S jv02 of 50% or less, versus none in the hypothermic group (p s 0.0 I by comparison of two proportions; Fig. 5). During rewarming at 32 C (period IV), the Sjv02 of patients in the hypothermic group declined (Fig. 3, A and Fig. 4), whereas the Sjv02 in the normothermic group continued to increase. In period IV, the mean Sjv02 values in the normothermic and hypothermic groups did not differ (Fig. 4). In the hypothermic group, the mean Sjv02 at 32 C during rewarming (62%) was lower than that at 32 C during cooling (68%; p s 0.05 by Wilcoxon signed rank test). Additionally, rewarming was the only period in which patients in the hypothermic group demonstrated cerebral venous desaturation (3/26). Three of 25 patients in the normothermic group also demonstrated cerebral venous oxygen desaturation during this period (Fig. 5). In the normothermic group during CPB, the proportion of patients who demonstrated cerebral desaturation (Sjv02 s 50%) decreased with the time from a 54% incidence at 20 minutes to a 31 % incidence at 40 minutes and a 12% incidence at 60 minutes (Fig. 5). The arterial-to-venous oxygen content difference between the groups paralleled the intergroup difference in Sjv02. Arterialto-venous oxygen content difference did not differ between groups in the baseline period, but differed significantly in the normothermic and hypothermic groups in period II (5.6 ± 1.6

5 Cook et al. April 1994 ONSET HYPOTHERMIC CPB TERMINATION BYPASS BYPASS E, OXIHETRIX E, 7:1. --:-+Ioo-----""'------Ioof l A :::~:::: ~... -'--'--'.. :.. : :.... ~.. i.. i'. 30 ~~t.o~~~~~~ ----t-- COOLING --+: ~Iil I '" j j I I 1 '~AiwiNG, : I il~~o~1 ~-~( 2~-;;9:3;:::~::;1:;0;:1. E, loam 11 AM ONSET BYPASS IOAI1 E, TERMINATION BYPASS 12PH ~ r ~--ioof l.~---- B.-----~---~---~----~ I : :. I 20 NORMOTHERMIC CPB Fig. 3. A, Continuous oximetry printout provided by oximetric computer in patient in hypothermic bypass group. E] designates onset of bypass and E2 termination of bypass. Note that saturation is inversely related to temperature. B, continuous oximetry printout provided by oximetric computer in patient in normothermic bypass group. Note fall in saturation with initiation of bypass and gradual improvement with time. and 3.9 ± 1.4 mi dl- 1 ) and period III (5.3 ± 1.4 and 3.5 ± 1.1 ml. dl- 1 ). In period IV, arterial-to-venous oxygen content difference did not differ between groups (Fig. 6). Arteriovenous glucose difference did not differ at baseline but did differ between the groups during CPB periods II and III. After bypass arteriovenous glucose difference differed between the hypothermic and normothermic groups during period V only (Fig. 7). There was a trend toward increasing blood glucose levels in the hypothermic group with the onset of cooling. Glucose levels in the hypothermic group remained elevated through the postbypass period. This was not seen in the normothermic group (Table 11). Physiologic variables (Paco2, hemoglobin, mean arterial pressure, pump flow rate, and Paco2) in patients demonstrating desaturation were compared with those in patients not demonstrating desaturation. No patient was acidotic when samples were drawn and no correlation was found between any physiologic variable to account for the low SjV02 values in patients with desaturation. Although the study was not designed to assess neurologic outcome, we note that one patient of the 52 had an embolic stroke and subsequently died. This patient underwent hypothermic CPB. There were no other perioperative strokes or deaths in either group. Discussion Normothermic CPB appears to have beneficial hemodynamic effects, but there is concern that this practice may be associated with a higher risk of neurologic injury.7,28 Hypothermia reduces cerebral oxygen demand

6 Volume 107, Number 4 Cook et al o 80 N 65 > (/) C o Q) ::E Normothermic,A.--,A. I Hypothermic 0-0 l------~ ~ 1 1 ) / : : ~~r::::: 1 50 ~ i I ~ j I ~ Pre-CPB 1-- CPB -I (minutes post CPB) Fig. 4. Mean Sjv02 during six study periods in hypothermic and normothermic bypass groups. Values are mean ± standard deviation (n = 26 per group). ***p < by Wilcoxon rank sum test...c ~ ~ I/) 0 +- Ll) c Q) N c.. 0 >. 0 (/) Z 14 Normothermic _ Hypothermic ~. 0 "worm CPS" 20 minutes 40 min. 60 min. "cold CPS" 32 cooling worming Fig. 5. Prevalence of cerebral venous desaturation in hypothermic and normothermic bypass groups in three CPB periods (periods II, III, and IV). and can protect the brain from ischemic insult. l, 2, 6 Although the prevalence of cerebral emboli may not be affected by bypass temperature, it is possible that emboli cause more damage in the warm brain. Additionally, any ischemia that might arise from CPB-induced flow-metabolism mismatch is potentially increased by normal brain temperature. There is evidence that CBF may be insufficient relative to rising cerebral oxygen demand during and after rewarming from hypothermic bypass This phenomenon could be sustained in patients undergoing normothermic CPB. Jugular bulb cannulation is a relatively simple technique and measurement of the oxygen saturation of cerebral venous blood can provide an estimate of CBF:metabolism coupling Properly placed, a catheter in the jugular bulb can accurately sample cerebral venous blood, inasmuch as less than 5% of the blood traversing the jugular bulb is contaminated from extracerebral sources. 34 The cerebral venous saturation differences between patients undergoing hypothermic bypass and those undergoing normothermic bypass and the prevalence of cerebral venous desaturation in the normothermic group were dramatic. Fifty-four percent of patients undergoing normothermic CPB had periods of cerebral desaturation, implying temporary relative cerebral flow insufficiency during early bypass. Potentially related findings were reported in an early article by Theye, Patrick, and Kirklin.35 These investiga-

7 Cook et al.,, "' E ~ ""-./ 0 0 N 6 > 4 «2,..-, Normothermic Hypothermic j\~-+-~=!~~ Pre-CPS / CPS-I (minutes post CPB) April 1994 Fig. 6. Mean arterial-to-venous oxygen content difference (AVD02) during six study periods in hypothermic and normothermic bypass groups. Values are mean ± standard deviation (n = 26 per group). **p < 0.01, ***p < by rank sum test. " )10 C 0 E Q) '--'" ~ ::J - 0) 6 0 > «2 Pre-CPS,..--,.. Normothermic 0-0 Hypothermic CPS-I (minutes post CPB) Fig. 7. Mean arteriovenous glucose difference (AVD g1u ) during six study periods in hypothermic and normothermic bypass groups. Values are mean ± standard error (n = 26 per group). *p < 0.05 by rank sum test. tors reported electroencephalographic changes suggestive of ischemia in 45 of 100 patients on the initiation of normothermic CPB. The authors postulated this may result from cerebral flow insufficiency, although a change in blood temperature related to an ambient-temperature pump prime was also considered. In our study, pump prime was warmed to 34 to 35 C before initiation of CPB. In the normothermic group, the cerebral circulation appeared to adapt to CPB flow conditions with time. The incidence of cerebral desaturation decreased and mean Sjv02 tended to increase through bypass. If we presume CMR02 is constant during normothermic CPB, as were pump flow and mean arterial pressure, the increase in cerebral venous saturation with time could be explained by microcirculatory adjustments to nonpulsatile flow. This circulatory change appears to be sustained for some period after CPB as demonstrated by persistent elevations in Sjv02 in the postbypass periods. The results in our hypothermic group are similar to those of Croughwell and associates 23 in that patients receiving hypothermic CPB demonstrated increasing Sjv02 and decreasing arterial-to-venous oxygen content difference with cooling. Furthermore, all episodes of desaturation in the hypothermic group were confined to the period of rewarming. We report a lower prevalence of cerebral desaturation during rewarming (3/26, 12%) than Croughwell and associates 23 (23%), probably because our measurements were taken at 32 C and that group's at 36 C; we may also have rewarmed more

8 Volume 107, Number 4 Cook et al slowly. During rewarming, temperature was allowed to stabilize at 32 0 C before samples were taken. The delay at 32 0 C may have allowed for some adjustments of the CBF to the rising CMR02. Hypothermia reduces CMR02 to a proportionately greater extent than CBF, resulting in "luxuriant perfusion."36,37 The increases in Sjv02 and decreases in arterial-to-venous oxygen content difference we report provide further physiologic evidence for this phenomenon. Though it may be naive to suggest that lower pump flow rates would reduce cerebral embolic events, evidence of "luxuriant perfusion" might support the safety of lower flow rates in patients undergoing hypothermic bypass. 37,38. It ultimately remains to be asked what the significance of these differences in cerebral physiologic states is. The one patient who had frank neurologic injury underwent hypothermic CPB and multiple embolic infarcts were documented by postoperative computed tomography. Although emboli may be responsible for most grossly apparent bypass-related brain injury, subtle neuropsychologic defects are more common and flow:metabolism mismatch should not be discounted, inasmuch as oxygen demand can be manipulated by perfusate temperature. Though desaturation did not correlate to gross neurologic injury, complex neuropsychologic testing might have been more revealing. Finally, the study excluded patient groups at high risk for cerebral events. Cerebral saturation assessment may have some role in these groups. Cerebral venous oxygen saturation assesses the balance between CBF and CMR02, but does not quantify either variable. Specific measurements of CBF and CMR02 during CPB would be more desirable, and hypotheses as to what is occurring in the cerebral circulation during bypass would be less speculative if direct measurements were made. In conclusion, patients undergoing normothermic CPB demonstrate significantly lower cerebral venous oxygen saturation than patients undergoing hypothermic CPB during extracorporeal circulation. Cerebral flow:metabolism matching is disturbed during early normothermic CPB, but with time the cerebral vasculature appears to adapt to the circulatory conditions. It remains to be seen whether these neurophysiologic abnormalities will translate into clinical outcome differences. We thank the Critical Care Division of Abbott Laboratories for provision of the oximetric catheters used in this study. REFERENCES 1. Berryessa RG, Tyndal CM Jr. Perfusion techniques that may decrease brain injury during cardiopulmonary bypass. In: Hilberman M, ed. Brain injury and protection during heart surgery. Boston: Martinus Nijhoff, 1988: 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: Kavanagh BP, Mazer CD, Panos A, Lichtenstein SV. Effect of warm heart surgery on perioperative management of patients undergoing urgent cardiac surgery. J Cardiothorac Vasc Anesth 1992;6: Lichtenstein SV, Ashe KA, el Dalati H, Cusimano RJ, Panos A, Slutsky AS. Warm heart surgery. J THORAC CARDIOVASC SURG 1991;101: Singh AK, Feng WC, Bert AA, Rotenberg FA. Warm body, cold heart surgery: clinical experience in patients. Eur J Cardiothorac Surg 1993;7: Steen PA, Newberg L, Milde JH, Michenfelder JD. Hypothermia and barbiturates: individual and combined effects on canine cerebral oxygen consumption. Anesthesiology 1983;58: Martin TD, Craver JM, Gott JP, et al. A prospective randomized trial of retrograde warm blood cardioplegia: myocardial benefit and neurologic threat. Presented at the Twenty-ninth Annual Meeting of the Society of Thoracic Surgeons, January 25-27, 1'993, San Antonio, Tex. 8. Kolkka R, Hilberman M. Neurologic dysfunction following cardiac operation with low-flow, low-pressure cardiopulmonary bypass. J THORAC CARDIOV ASC SURG 1980; 79: Slogoff S, Girgis KZ, Keats AS. Etiologic factors in neuropsychiatric complications associated with cardiopulmonary bypass. Anesth Analg 1982;61 : Furlan AJ, Breuer AC. Central nervous system complications of open heart surgery. Stroke 1984;15: Breuer AC, FurlanAJ, Hanson MR, etal. Central nervous system complications of coronary artery bypass surgery: prospective analysis of 421 patients. Stroke 1983; 14: Shaw PJ, Bates D, Cartlidge NEF, Heaviside D, Julian DG, Shaw DA. Early neurological complications of coronary artery bypass surgery. BMJ 1985;291: Aberg T, Ronquist G, Tyden H, Brunnkvist S, Bergstrom K. Cerebral damage during open-heart surgery. Scand J Thorac Cardiovasc Surg 1987;21: Sotaniemi KA. Brain damage and neurological outcome after open-heart surgery. J Neurol Neurosurg Psychiatry 1980;43: Sotaniemi KA, Mononen H, Hokkanen TE. Long-term cerebral outcome after open heart surgery: a five-year neuropsychological follow-up study. Stroke 1986; 17: Parker FB Jr, Marvasti MA, Bove EL. Neurologic complications following coronary artery bypass: the role of atherosclerotic emboli. Thorac Cardiovasc Surg 1985;33: Govier AV, Reves JG, McKay RD, et al. Factors and their influence on regional cerebral blood flow during nonpulsatile cardiopulmonary bypass. Ann Thorac Surg 1984; 38:

9 Cook et al. April Murkin JM, Farrar JK, Tweed AW, McKenzie FN, Guiraudon G. Cerebral autoregulation and flow/metabolism coupling during cardiopulmonary bypass: the influence ofpaco2. Anesth Analg 1987;66: Raichle ME, Grubb RL Jr, Gado MH, Eichling JO, Ter Pogossian MM. Correlation between regional cerebral blood flow and oxidative metabolism: in vivo studies in man. Arch NeuroI1976;33: Gibbs EL, Lennox WG, Nims LF, Gibbs FA. Arterial and cerebral venous blood: arterial-venous differences in man. J BioI Chem 1942;144: Robertson CS, Narayan RK, Gokaslan ZL, etal. Cerebral arteriovenous oxygen difference as an estimate of cerebral blood flow in comatose patients. J N eurosurg 1989;70: Nakajima T, Kuro M, Hayashi Y, Kitaguchi K, Uchida 0, Takaki O. Clinical evaluation of cerebral oxygen balance during cardiopulmonary bypass: on-line continuous monitoring of jugular venous oxyhemoglobin saturation. Anesth Analg 1992;74: Croughwell ND, Frasco P, Blumenthal JA, Leone BJ, White WD, Reves JG. Warming during cardiopulmonary bypass is associated with jugular bulb desaturation. Ann Thorac Surg 1992;53: Newman M, Croughwell N, Baldwin B, et al. Effects of thiopental and isoflurane on jugular bulb desaturation during nonpulsatile cardiopulmonary bypass [Abstract]. Anesthesiology 1992;77:A Lyons C, Clark LC Jr, McDowell H, McArthur K. Cerebral venous oxygen content during carotid thrombintimectomy. Ann Surg 1964;160: Clauss RH, Hass WK, Ransohoff J. Simplified method for monitoring adequacy of brain oxygenation during carotid artery surgery. N Eng! J Med 1965;273: Meyer JS, Gotoh F, EbiharaS, Tomita M. Effects of anoxia on cerebral metabolism and electrolytes in man. Neurology 1965;15: Martin TD, Tao X, Weintraub WS, Craver JM, Guyton RA. Warm blood versus cold crystalloid cardioplegia: a case matched comparison [Abstract]. Circulation 1992; 86(Suppl): Sheinberg M, Kanter MJ, Robertson CS, Contant CF, Narayan RK, Grossman RG. Continuous monitoring of jugular venous oxygen saturation in head-injured patients. J Neurosurg 1992;76: Garlick R, Bihari D. The use of intermittent and continuous recordings of jugular venous bulb oxygen saturation in the unconscious patient. Scand J Clin Lab Invest 1987;47(SuppI188): Cruz J, Miner ME, Allen SJ, Alves WM, Gennarelli TA. Continuous monitoring of cerebral oxygenation in acute brain injury: injection of mannitol during hyperventilation. J Neurosurg 1990;73: Gayle MO, Frewen TC, Armstrong RF, et al. Jugular venous bulb catheterization in infants and children. Crit Care Med 1989;17: Swain JA. Hypothermia and blood ph. Arch Intern Med 1988;148: Shenkin HA, Harmel MH, Kety SS. Dynamic anatomy of the cerebral circulation. Arch Neurol Psychiatry 1948; 60: Theye RA, Patrick RT, Kirklin JW. The electro-encephalogram in patients undergoing open intracardiac operations with the aid of extracorporeal circulation. J THORAC SURG 1957;34: Croughwell N, Smith LR, Quill T, et al. The effect of temperature on cerebral metabolism and blood flow in adults during cardiopulmonary bypass. J THORAC CARDIOV ASC SURG 1992;103: Henriksen L. Brain luxury perfusion during cardiopulmonary bypass in humans: a study of the cerebral blood flow response to changes in CO2, 02, and blood pressure. J Cereb Blood Flow Metab 1986;6: Rebeyka 1M, Coles JG, Wilson GJ, et al. The effect of low-flow cardiopulmonary bypass on cerebral function: an experimental and clinical study. Ann Thorac Surg 1987; 43: Discussion Dr. Richard J. Hurvitz (Los Angeles, Calif). This article addresses cerebral protection during normothermic and hypothermic CPB. The topic of normothermia versus hypothermia for myocardial protection is indeed one of the more controversial topics in cardiac surgery today. A large body of information, regarding the efficacy of hypothermia for protection of the heart, as well as other organs, dates back to Drs. Bigelow and Swan in the 1950s. The information regarding normothermic cardioplegia is both relatively recent and meager. There appear to be some compelling reasons for its use; however, it is a relatively more complex procedure and the "margin of safety" is reduced. There are at this time more questions than answers regarding normothermic cardioplegia. Our surgical group at the University of Southern California continues to rely on the protection of cold, in addition to the other modalities involved with the myocardial protection process. I would urge caution to persons who wish to begin the technique of warm cardioplegia and suggest the evaluation of further reports, like the one presented and that from the groups that have already gone through the learning curve on this technique, before jumping on the bandwagon. I have the following two questions. First, the authors state that some patients received nitroprusside, but do not specifically identify these patients. Because nitroprusside, by generating cyanide, can interfere with oxygen uptake at the cellular level, do the authors think it might have affected the results? Second, now that the authors have shown that there is a significant difference in the cerebral venous oxygen saturation in the two techniques, are they attempting to evaluate neurologic outcome between the two techniques? Dr. Cook. I did not report requirements for vasodilators or inotropic agents as part of the study. It is a relevant question, though. What we found was that patients undergoing hypothermia during bypass required significantly higher doses and a higher frequency of use of nitroprusside to maintain the blood pressures within the range defined by the protocol. Cyanide toxicity decreases oxygen use and would alter saturation values.

10 Volume 107, Number 4 Cook et al There was no evidence in either group that any cyanide toxicity developed. With the examination of both arterial and venous blood in the cerebral circulation, as well as systemically, there was no evidence that any alteration in cellular metabolism was occurring. Dr. Hurvitz. Now that you have shown there is a significant difference, are you attempting to evaluate neurologic outcome between the two techniques? Dr. Cook. Yes, we are. The Mayo Foundation and the American Heart Association are supporting a comprehensive project with Drs. Oliver, Orszulak, and Daly and me not only to directly quantitate CBF and metabolism but also to do intraoperative electroencephalography in patients undergoing normothermic bypass and finally to do complex studies of both shortand long-term neurologic outcome. Dr. John Opie (Phoenix, Ariz.). On the second conclusion on the discussion slides, the authors mentioned that there is an ischemic time associated with their findings. I would submit that it is not really an ischemic finding that has been produced but rather a desaturation finding. It is not an ischemic time at all and I do not think it should be defined as ischemic. We have just completed a study published in The Annals of Thoracic Surgery on about 358 patients on whom we operated with warm continuous blood cardioplegia. Like you, we found no strokes except in a comparison cold group that had one or two residual strokes. Dr. Cook. I think your point is well taken. It is a matter of definition. I think desaturation would be a better term in that context. Dr. Steve Gundry (Loma Linda, Calif). I noticed that the authors said the anesthetic management was controlled by protocol. Perhaps you can enlighten us as to whether flow rates were increased in patients receiving normothermic CPB and wheth- er dips in blood pressure that occurred during normothermia were treated with phenylephrine hydrochloride (Neo-Synephrine). Dr. Cook. The flow rates did not differ between the two groups; they were defined by the protocol. We kept the CPB flow rate for all patients at 2.2 to 2.4 L. min-i. m- 2. We thought that treating dips in saturation would potentially alter the results of the study and therefore they were not treated. Dr. Gundry. I mean dips in mean arterial pressure. Dr. Cook. Yes. Mean arterial pressure was maintained between 45 and 65 mm Hg rigorously in both groups. Dr. Gundry. How did you maintain it in the normothermic group? Dr. Cook. We frequently had to use phenylephrine to maintain pressure in those ranges. Dr. Gundry. Did you look at cerebral oxygen saturation during the times that phenylephrine was administered to keep up the mean arterial pressure? Dr. Cook. No, we did not directly isolate those patients receiving phenylephrine. In a few patients in pilot work that we did, I attempted to treat falls in saturation with phenylephrine infusion and there did seem to be transient increases or improvements in saturation as the mean arterial pressure was increased, but that finding is certainly only anecdotal. Dr. Gundry. My second question is that obviously a lot ofthe data refer to differences in oxygen saturation between inflow and outflow oxygen saturation chal'.ges. Did you correct for any effect that hypothermia had on shifting the oxygen hemoglobin dissociation curve? I think a lot of the findings are directly related to that shift in the patients undergoing hypothermic bypass, which would invalidate your conclusions. Dr. Cook. No, we did not make those corrections.