The increasing use of normothermic cardiopulmonary

A Prospective, Randomized Study of Cardiopulmonary Bypass Temperature and Blood Transfusion Paul E. Stensrud, MD, Gregory A. Nuttall, MD, Maria A. de Castro, MD, Martin D. Abel, MD, Mark H. Ereth, MD, William C. Oliver, Jr, MD, Sandra C. Bryant, MS, and Hartzell V. Schaff, MD Departments of Anesthesiology, Surgery, and Biostatistics, Mayo Clinic, Rochester, Minnesota Background. We hypothesized that normothermic cardiopulmonary bypass (CPB) would be associated with decreased blood loss and allogeneic transfusion requirements relative to hypothermic CPB. Methods. After obtaining institutional review board approval and informed patient consent, we conducted a prospective, randomized study of 79 patients undergoing CPB for a primary cardiac operation at normothermic (37 C) (n 44) or hypothermic temperature (25 C) (n 35). Blood loss and transfusion requirements in the operating room and for the first 24 hours in the intensive care unit were determined. A paired t test and rank sum tests were used. A p value of less than 0.05 was considered significant. The increasing use of normothermic cardiopulmonary bypass (CPB) in patients undergoing cardiac operations has initiated a debate concerning the relative risks and benefits of normothermic and hypothermic CPB [1 10]. This debate has spawned a number of clinical studies reporting various measures of outcome, but little information has been published pertaining to differences in blood loss and allogeneic transfusion requirements between normothermic and hypothermic techniques [2 10]. Cardiopulmonary bypass has been demonstrated to be associated with major hemostatic defects arising from exposure of the blood to synthetic nonbiologic surfaces [11, 12]. These defects of hemostasis result in blood loss and frequent transfusion of allogeneic blood products to patients undergoing CPB [13 16]. Hypothermia alone is known to induce platelet dysfunction and inhibit coagulation, which may exacerbate the bleeding when hypothermia is used in conjunction with CPB [6 8]. Patients undergoing CPB are thus at risk for transfusion reaction and transmission of transfusionrelated disease [17 19]. In addition, the higher transfusion requirements place a burden on blood banks and increase the cost of care for these patients. These risks and costs may be increased if hypothermic CPB results in greater blood loss and transfusion requirements. Accepted for publication Sep 3, 1998. Address reprint requests to Dr Stensrud, Department of Anesthesiology, Mayo Clinic, Rochester, MN 55905; e-mail: stensrud.paul@mayo.edu. Results. The normothermic and hypothermic CPB groups did not differ in demographic variables, CPB or cross-clamp duration, heparin sodium or protamine sulfate dose, prothrombin time, or thromboelastogram results. There were no differences between the two CPB groups in blood loss or transfusion requirements. Conclusions. We found that when there was no difference in duration of CPB, normothermic and hypothermic CPB groups demonstrated similar blood loss and transfusion requirements even though other studies have shown hypothermia induces platelet dysfunction and alters the activity of the coagulation cascade. (Ann Thorac Surg 1999;67:711 5) 1999 by The Society of Thoracic Surgeons On the basis of clinical observations, we believed that patients undergoing normothermic CPB bled less than patients undergoing hypothermic CPB. We therefore hypothesized that cardiac surgical procedures using normothermic CPB would be associated with decreased blood loss and transfusion of allogeneic blood products compared with such procedures with hypothermic CPB. Material and Methods After we obtained institutional review board approval and informed patient consent, 79 patients undergoing primary median sternotomy for coronary artery bypass grafting, septal myotomy-myectomy, or single-valve repair or replacement were randomly assigned to receive normothermic (37 C) or hypothermic (25 C) CPB. Exclusion criteria included any antifibrinolytic therapy, weight of less than 45 kg, preoperative hemoglobin level of less than 10 g/dl, preoperative platelet count lower than 100 10 3 /mm 3, creatinine value greater than 2.0 mg/dl, preoperative administration of thrombolytic agents, preoperative warfarin sodium therapy, and known history of a bleeding disorder. Preoperative intake of aspirin, di- This article has been selected for the open discussion forum on the STS Web site: http://www.sts.org/section/atsdiscussion/ 1999 by The Society of Thoracic Surgeons 0003-4975/99/$20.00 Published by Elsevier Science Inc PII S0003-4975(99)00040-5

712 STENSRUD ET AL Ann Thorac Surg CPB TEMPERATURE AND TRANSFUSION 1999;67:711 5 pyridamole, and nonsteroidal antiinflammatory medications was noted. Intraoperative Management A single surgeon (H.V.S.) operated on all of the patients. A standardized moderate-dose opioid-based anesthesia technique supplemented with benzodiazepines, muscle relaxants, and inhalational anesthetic agents was used. Cardiopulmonary bypass was conducted with a Univox membrane oxygenator (Baxter Healthcare Corp, Bentley Laboratories Division, Irvine, CA) and a Sarns 9000 CPB machine (Sarns Inc, Ann Arbor, MI). The CPB circuit was primed with Plasmalyte, 1.5 L, sodium bicarbonate (NaHCO 3 ), 10 meq, and mannitol, 12.5 g. Porcine heparin was administered to patients as follows: The initial dose consisted of a bolus of 9,000 units body surface area (square meters) and a pump priming dose of 10,000 units. Additional heparin (5,000 units) was administered when the celite activated coagulation time (ACT) was less than 450 seconds (Hemochron 801; International Technidyne, Edison, NJ). After discontinuation of CPB, the initial protamine sulfate dose was 0.013 mg per unit of heparin administered. Heparin neutralization was considered adequate if the ACT result after protamine was within 10% of the baseline ACT value or, in the case of patients receiving preoperative heparin therapy, if the ACT after protamine was less than 150 seconds. Additional protamine (20 to 50 mg) was administered at the discretion of the attending clinician if the ACT had not returned to this range. Patient temperature was measured with a nasopharyngeal thermistor. Cardiopulmonary bypass temperature was documented as the lowest nasopharyngeal temperature achieved during CPB. Intraoperative autotransfusion and postoperative reinfusion of shed blood were used in all patients. Transfusion criteria for allogeneic blood products were standardized. Red blood cells were administered if the hemoglobin concentration was less than 7 g/dl during CPB and less than 8 g/dl after discontinuation of CPB. Fresh frozen plasma, platelets, or cryoprecipitate was given if the patient demonstrated nonsurgical bleeding, abnormal laboratory studies of hemostasis (thromboelastogram [TEG], platelet count, prothrombin time, activated partial thromboplastin time [aptt], and fibrinogen concentration), or both. Data Collection Preoperative hemoglobin concentration and platelet count were recorded. Activated clotting times were recorded before heparin administration, during CPB, and after heparin neutralization with protamine. Patient temperature was recorded as baseline, minimum temperature during CPB, and temperature on departure from the operating room. After protamine administration and return of ACT to within 10% of baseline, blood was collected for determination of hemoglobin and fibrinogen concentration, platelet count, prothrombin time, aptt, and TEG. Intraoperative blood loss after CPB was estimated by recording the volume of reinfused salvaged red blood Table 1. Perioperative Demographic Characteristics a,b cells, which had been collected in a dedicated suction canister. Drain output was recorded for the first 24 hours postoperatively. Shed mediastinal blood was salvaged and reinfused into the patient for the first 24 hours postoperatively. The amount of allogeneic and autologous red blood cells, platelets, fresh frozen plasma, and cryoprecipitate transfused in the operating room and for the first 24 hours postoperatively was recorded. Statistical Analysis Data were expressed as the median and the range. Wilcoxon rank sum tests were used to compare data between the hypothermic and normothermic groups. A p value of less than 0.05 was considered significant. For the outcome variables, 95% confidence intervals about the median differences between the two groups were based on distribution-free methods in Hollander and Wolfe [20]. Multiple regression modeling was used to assess the effect of temperature on blood loss and amount of blood transfused after adjusting for appropriate covariates. Results Hypothermic (25 C) CPB (n 35) Normothermic (37 C) CPB (n 44) p Value Age (y) 62, 40 79 64, 28 81 0.67 Female sex 11 (31) 13 (30) 1.00 Height (cm) 173, 152 192 172, 154 191 0.71 Weight (kg) 80, 50 106 83, 50 122 0.67 BSA (m 2 ) 2.0, 1.5 2.4 2.0, 1.5 2.6 0.74 Preop aspirin 19 (54) 31 (70) 0.16 Preop heparin 6 (17) 16 (36) 0.08 Preop nonsteroidal 1 (3) 3 (7) 0.63 drug Procedure 0.03 CABG 16 (46) 31 (70) Aortic valve repair 9 (26) 2 (5) Mitral valve repair 5 (14) 8 (18) Septal myectomy 3 (9) 1 (2) Other 2 (6) 2 (5) a Where applicable, data are shown as the median and the range. b Numbers in parentheses are percentages. BSA body surface area; CABG coronary artery bypass grafting; CPB cardiopulmonary bypass. The demographic characteristics of the study population are summarized in Table 1. There were no differences in these variables between the two groups, except that a greater percentage of patients in the normothermic group received preoperative heparin therapy ( p 0.08) and had a coronary artery bypass grafting procedure ( p 0.03). Perioperative surgical characteristics are summarized in Table 2. There were no differences between the two groups except the minimal temperature on CPB, temperature on arrival at intensive care unit (ICU), and initial ACT. The minimum patient temperature on CPB and the patient temperature on arrival at the ICU were lower in

Ann Thorac Surg STENSRUD ET AL 1999;67:711 5 CPB TEMPERATURE AND TRANSFUSION 713 Table 2. Perioperative Surgical Characteristics a Hypothermic (25 C) CPB (n 35) Normothermic (37 C) CPB (n 44) p Value CPB duration (min) 52, 28 91 49, 22 99 0.91 Aortic cross-clamp duration (min) 34, 18 68 32.5, 13 57 0.39 Temperature ( C) Initial 35.9, 35.1 36.6 36.0, 34.9 37.0 0.08 Minimal on CPB 25, 24.2 31.5 37.0, 35.0 38.0 0.0001 After CPB 37.3, 35.5 38.5 37.6, 35.8 39.3 0.14 On arrival to ICU 36.0, 34.6 37.1 37, 35.8 38 0.0001 Initial heparin dose (units) 18,000, 14,000 24,000 18,000, 10,000 25,000 0.76 Protamine dose (mg) 250, 180 300 250, 175 360 0.61 Initial ACT (s) 146, 111 180 157, 113 204 0.018 ACT after protamine (s) 136, 104 179 137, 107 215 0.66 a Data are shown as the median and the range. ACT activated clotting time; CPB cardiopulmonary bypass; ICU intensive care unit. the hypothermic CPB group. The baseline ACT was prolonged in the normothermic CPB group, but this group included a greater number of patients who received heparin infusions preoperatively. The initial temperature was also marginally higher in the normothermic group than in the hypothermic group, although not enough to reach significance ( p 0.08). Blood loss and coagulation test results are summarized in Table 3. There was no difference in perioperative blood loss between The aptt after CPB was prolonged in the hypothermic CPB group. The prothrombin time, platelet count, hemoglobin concentration, and fibrinogen concentrations after CPB did not differ between The amount of blood products transfused in the operating room and the ICU are shown in Figures 1 and 2, respectively. The blood product transfusions were non- Gaussian in distribution, and the results were thus expressed as median and interquartile ranges. There were no significant differences in blood product transfusion between the two groups when analyzed by the Wilcoxon rank sum test. There were no significant differences in either median ICU stay or median hospital stay between Because there were significant differences between the two groups in initial ACT and number of coronary artery bypass grafting procedures and nearly significant differences in preoperative heparin therapy and initial temperature, multiple regression modeling was performed to assess the effect of temperature on blood loss and amount of blood transfused. The dependent variables were intraoperative blood loss, 12-hour blood loss, amount of red blood cells transfused in the operating room, total blood transfused in the operating room, and amount of red blood cells transfused within 12 hours after leaving the operating room. After adjusting for the Table 3. Blood Loss and Coagulation Test Results Hypothermic (25 C) CPB (n 35) a Normothermic (37 C) CPB (n 44) a p Value Median Difference Between Groups (95% CI) Intraop blood loss after CPB (ml) 248, 0 744 248, 88 1,114 0.67 0 ( 17, 68) 12-Hour ICU blood loss (ml) 725, 410 3,250 775, 195 1,899 0.31 98.5 ( 50, 309) Preop hemoglobin (g/dl) 13.8, 10.9 16.1 14.2, 10.5 16.6 0.15 0.4 ( 1.1, 0.4) Preop platelet count (10 3 /mm 3 ) 217, 132 422 226, 113 579 0.90 1 ( 31, 27) Hemoglobin after 24 hours in ICU (g/dl) 10.6, 6.9 12.8 10.8, 7.8 14.5 0.46 0.3 ( 1, 0.5) Intraop PT after CPB (s) 14.3, 11.8 20.7 14.4, 11.0 21.0 0.76 0.1 ( 0.4, 0.8) Intraop aptt after CPB (s) 39, 28 224 36, 28 77 0.01 4.5 (1.2, 8.7) Intraop platelet count after CPB (10 3 /mm 3 ) 124, 66 219 134.5, 44 253 0.39 9 ( 25, 20) Intraop fibrinogen after CPB (mg/dl) 191, 69 331 194.5, 98 341 0.70 3 ( 22, 15) Thromboelastography R (mm) 15, 10 25 14, 8 27 0.48 0.5 ( 0.5, 3.0) Angle (degrees) 49.5, 26 68 50, 27 70 0.33 2 ( 5, 3) MA (mm) 54, 34 62 54, 40 69 0.31 1.5 ( 4, 2.5) MA 60 (mm) 44, 0 61 46.8, 0.5 63 0.84 0.45, ( 2, 11) a Data are shown as the median and the range. aptt activated partial thromboplastin time; CI confidence interval; CPB cardiopulmonary bypass; ICU intensive care unit; MA maximal amplitude; MA 60 amplitude 60 minutes after MA; PT prothrombin time; R reaction time.

714 STENSRUD ET AL Ann Thorac Surg CPB TEMPERATURE AND TRANSFUSION 1999;67:711 5 Fig 1. Allogeneic transfusion and intraoperative autotransfusion exposures in operating room after cardiopulmonary bypass (CPB) (median and interquartile ranges). Each dot represents 1 patient, and patients who did not receive that blood product are positioned under 0. There were no significant differences in median or mean transfusion requirements between groups. (cryo cryoprecipitate; FFP fresh frozen plasma; RBCs red blood cells.) covariates, there were no significant differences between Comment This prospective, randomized study demonstrated no difference in blood loss or transfusion requirements between a group of patients undergoing normothermic (37 C) CPB and a group having hypothermic (25 C) CPB. Table 3 indicates with 95% certainty that the betweengroup difference in intraoperative blood loss is no greater than 68 ml and 12-hour blood loss, no greater than 309 ml. A recent prospective, randomized study [5] of CPB Fig 2. Allogeneic transfusion exposures in the first 24 hours in intensive care unit (ICU) (median and interquartile ranges). Each dot represents 1 patient, and patients who did not receive that blood product are positioned under 0. There were no significant differences in median or mean transfusion requirements between groups. (cryo cryoprecipitate; FFP fresh frozen plasma; RBCs red blood cells.) temperature and bleeding and transfusion requirements showed the same results. All patients in our study were operated on by a single surgeon (H.V.S.). There was no difference in duration of CPB or aortic cross-clamp time between Anesthesia teams using standardized anesthesia techniques during a defined period of time cared for the patients. Transfusion practices in the operating room and the ICU were standardized. Signal features of this study include a prospective, randomized design, relatively large groups, similar CPB times, and use of the TEG and other coagulation tests to measure global hemostatic function. The intent of this study was to have equal numbers of patients in the normothermic and hypothermic groups. Medical record numbers were used to determine group assignment: even numbers received normothermic CPB and odd numbers, hypothermic CPB. This method of randomization resulted in the group sizes observed. Therefore, power analyses were performed to understand the power of the resulting study. For intraoperative blood loss, this study would have had 90% power to detect a 41% decrease in blood loss in the normothermic group and 80% power to detect a 35% drop in blood loss in the normothermic group. These power estimates were obtained using a pooled standard deviation estimate from both groups and the mean observed in the hypothermic group. Using similar techniques for 12-hour blood loss, similar power results were obtained. This study could have been improved by blinding the anesthesia and surgical personnel to the temperature. However, this would be impractical, as the surgeon might be able to detect the temperature difference in the operative field and the anesthesia team is obliged to monitor the patient s temperature during the surgical procedure. There was a difference in patient temperature on arrival to the ICU, but it was not clinically significant. We might have found different results if there had been a greater difference in patient temperature on arrival to the ICU [13]. ecause random differences were found in preoperative heparin therapy, initial ACT, and initial temperature between the two groups, these covariates along with other significant covariates were included in a multiple regression model to assess the effect of bypass temperature on blood use. One study [16] found previous heparin therapy was a predictor of increased red blood cell transfusions, but other studies [13, 15] reported that preoperative heparin therapy was not a significant determinant of blood transfusion in patients undergoing primary coronary artery bypass procedures. Further, in the recent prospective, randomized study by Dennehy and Nathan [5], there was no difference in preoperative heparin use and no difference in blood loss or transfusion requirements between normothermic and hypothermic groups. The CPB durations were short. We might have seen different results with longer CPB times. Finally, the study included only patients undergoing primary sternotomy and no more than single-valve replacement. Different results might have been noted with patients undergoing repeat sternotomy or more complex surgical procedures.

Ann Thorac Surg STENSRUD ET AL 1999;67:711 5 CPB TEMPERATURE AND TRANSFUSION 715 Some groups [7 10] have reported an increase in transfusion requirements in patients undergoing hypothermic CPB relative to those undergoing normothermic CPB. Other studies [3 6] demonstrated no difference in blood loss or transfusion requirements between normothermic and hypothermic CPB groups. In all these negative studies, there were no significant differences in length of CPB. An advantage of normothermic CPB is that it frequently allows shorter CPB durations, as the time to cool and rewarm is obviated [2]. Platelet activation and subsequent dysfunction increase with the duration of CPB. Comparative studies [1, 2, 13] have consistently shown an increased CPB time with hypothermic CPB. Therefore, temperature during CPB may be a less important factor than duration of CPB on blood loss and transfusion requirements. In other words, the effects of the CPB circuit and oxygenator on hemostasis over time appear to be much more important than the effect of temperature. Despite the fact that hypothermia is documented to inhibit platelet and coagulation factor function [6 8], this study demonstrated no difference in TEG results. The absence of difference in the TEG maximal amplitude, a coagulation test that is thought to measure platelet function, may indicate that temperature during CPB had little effect on postoperative platelet function. The TEG maximal amplitude is also affected by fibrinogen concentration and therefore may not be as sensitive as other measures of platelet function, such as bleeding time or platelet aggregometry. The only difference in coagulation test results after CPB between the normothermic and hypothermic groups in this study was a longer aptt in the hypothermic group. The prolonged aptt in this group is consistent with the finding in a previous report [20] of the effect of hypothermia on aptt. Another possible cause of the prolonged aptt after CPB may represent a subtle worsening of hemostatic function in this group that cannot be detected in relatively crude measures of blood loss and transfusion requirement. Blood loss and transfusion requirements are, however, the major outcome variables associated with patient complications and increased cost of care. The results of this study combined with those of earlier studies support the supposition that temperature during CPB may be a less critical determinant of perioperative blood loss and allogeneic transfusion requirements than duration of CPB. In effect, the hemostatic insult imposed by the CPB circuit and oxygenator may be more important than that imposed by hypothermia. In summary, this prospective, randomized study with similar, albeit relatively short, CPB durations demonstrated no difference in blood loss or transfusion requirements between normothermic and hypothermic CPB groups. Supported by the Mayo Foundation for Medical Education and Research. We thank Ms Malinda Woodward, Ms Stacy Hanf, and Ms Kimberly Smith for assistance with manuscript preparation. References 1. The Warm Heart Investigators. Randomised trial of normothermic versus hypothermic coronary bypass surgery. Lancet 1994;343:559 63. 2. Christenson JT, Maurice JM, Simonet F, Velebit V, Schmuziger M. Normothermic versus hypothermic perfusion during primary coronary artery bypass grafting. Cardiovasc Surg 1995;3:519 24. 3. Yau TM, Carson S, Weisel RD, et al. The effect of warm heart surgery on postoperative bleeding. J Thorac Cardiovasc Surg 1992;103:1155 63. 4. Nathan HJ, Polis T. The management of temperature during hypothermic cardiopulmonary bypass. II. Effects of prolonged hypothermia. Can J Anaesth 1995;42:672 6. 5. Dennehy KC, Nathan HJ. The effect of mild hypothermia on bleeding following coronary artery bypass graft surgery [Abstract]. Anesth Analg 1997;84:S70. 6. Mazer CD, Hornstein A, Freedman J. Platelet activation in warm and cold heart surgery. Ann Thorac Surg 1995;59: 1481 6. 7. Boldt J, Knothe C, Welters I, Dapper FL, Hempelmann G. Normothermic versus hypothermic cardiopulmonary bypass: do changes in coagulation differ? Ann Thorac Surg 1996;62:130 5. 8. Boldt J, Knothe C, Zickmann B, Bill S, Dapper F, Hempelmann G. Platelet function in cardiac surgery: influence of temperature and aprotinin. Ann Thorac Surg 1993;55:652 8. 9. Tönz M, Mihaljevic T, von Segesser LK, et al. Normothermia versus hypothermia during cardiopulmonary bypass: a randomized, controlled trial. Ann Thorac Surg 1995;59:137 43. 10. Arom KV, Emery RW, Northrup WF. Warm heart surgery: a prospective comparison between normothermic and tepid temperature. J Cardiac Surg 1995;10:221 6. 11. Khuri SF, Wolfe JA, Josa M, et al. Hematologic changes during and after cardiopulmonary bypass and their relationship to the bleeding time and nonsurgical blood loss. J Thorac Cardiovasc Surg 1992;104:94 107. 12. Woodman RC, Harker LA. Bleeding complications associated with cardiopulmonary bypass. Blood 1990;76:1680 97. 13. Despotis GT, Filos KS, Zoys TN, Hogue CW, Spitznagel E, Lappas DG. Factors associated with excessive postoperative blood loss and hemostatic transfusion requirements: a multivariate analysis in cardiac surgical patients. Anesth Analg 1996;82:13 21. 14. Sanguis Study Group. Use of blood products for elective surgery in 43 European hospitals. Transfusion Med 1994;4: 251 68. 15. Surgenor DM, Churchill WH, Wallace EL, et al. Determinants of red cell, platelet, plasma, and cryoprecipitate transfusions during coronary artery bypass graft surgery: the Collaborative Hospital Transfusion Study. Transfusion 1996; 36:521 32. 16. Ferraris VA, Gildengorin V. Predictors of excessive blood use after coronary artery bypass grafting. J Thorac Cardiovasc Surg 1989;98:492 7. 17. Schreiber GB, Busch MP, Kleinman SH, Korelotz JJ. The risk of transfusion-transmitted viral infections. The Retrovirus Epidemiology Donor Study. N Engl J Med 1996; 334:1685 90. 18. Sloand EM, Pitt E, Klein H. Safety of the blood supply. JAMA 1995;274:1368 79. 19. Donahue JG, Munoz A, Ness PM, et al. The declining risk of post-transfusion hepatitis C virus infection. N Engl J Med 1992;327:369 70. 20. Hollander M, Wolf D. Nonparametric statistical methods. New York: John Wiley & Sons, 1973:78 82. 21. Staab DB, Sorensen VJ, Fath JJ, et al. Coagulation defects resulting from ambient temperature-induced hypothermia. J Trauma 1994;36:634 8.