Randomized Comparison Between Normothermic and Hypothermic Cardiopulmonary Bypass in Pediatric Open-Heart Surgery

Randomized Comparison Between Normothermic and Hypothermic Cardiopulmonary Bypass in Pediatric Open-Heart Surgery Massimo Caputo, MD, Simon Bays, FRCS, Chris A. Rogers, PhD, Ash Pawade, FRCS, Andrew J. Parry, FRCS, Saadeh Suleiman, PhD, and Gianni D. Angelini, FRCS Bristol Heart Institute, University of Bristol, Bristol Royal Infirmary, Bristol, United Kingdom Background. The purpose of this study is to investigate the effect of cardiopulmonary bypass () temperature on myocardial reperfusion injury, oxidative stress, and inflammatory response in pediatric open heart surgery. Methods. Fifty-nine children (median age 78 months; interquartile range, 39 130) undergoing correction of simple congenital heart defects were randomized to receive either hypothermic (28 C) or normothermic (35 37 C). Troponin I and 8-isoprostane, complement activation C3a, interleukin (IL) -6, -8, and -10, were measured preoperatively, on removal of the aortic cross clamp, 30 minutes, 6, and 24 hours postoperatively. Results. Troponin I and 8-isoprostane were significantly raised, compared to baseline, in both groups, and remained high at 24 hours. Overall, troponin I and 8-isoprostane levels were 37% and 84% higher in the hypothermic than in the normothermic group, respectively (ratio 1.37, 95% CI 1.00 to 1.88, p 0.053 and 1.84, 95% CI 1.22 to 2.78, p 0.0045, respectively), and there was no evidence to suggest the treatment effect changed significantly over the time points measured (p 0.63). Adjusting for aortic cross-clamp time reduced the effect of hypothermia on troponin (p 0.18) but not on 8-isoprostane levels (p 0.0028). The C3a, IL-6, and IL-8 release was similar in the two groups. The IL-10 release between the groups changed over time (p 0.059) and examining differences at individual time points highlighted a statistically significant difference at the end of the cross-clamp time (p 0.0079). Conclusions. Normothermic is associated with reduced oxidative stress compared with hypothermic, and similar myocardial reperfusion injury and whole body inflammatory response, in children undergoing open heart surgery. A larger study with clinical outcomes as primary end points is now warranted. (Ann Thorac Surg 2005;80:982 8) 2005 by The Society of Thoracic Surgeons In adults, normothermic (35 37 C) cardiopulmonary bypass () has been shown to lead to improved clinical outcome and less organ dysfunction when compared with hypothermic [1 3]. However, in pediatric open-heart surgery the use of hypothermia is still widely spread in the majority of cardiac surgical centers [4]. The main rationale for body cooling is to protect organs such as the brain, the kidneys, and the heart from ischemic injury by a reduction in metabolic rate, reflected in decreased oxygen consumption [4]. Furthermore hypothermia has been associated with a reduction in the whole body inflammatory response to [5 7]. Nevertheless, perfusion temperature strategies and their effects on organ functions are largely the result of work carried out on adult human [1 3] or animal models [5, 6] and the conclusions drawn from those studies cannot necessarily be applied directly to infants or children. Indeed, other groups [7, 8] have recommended the use of normothermic full-flow bypass for pediatric cardiac surgery despite the lack of prospective randomized evidence. The aim of this study was therefore to examine, in a randomized controlled trial, the effect of temperature on myocardial reperfusion injury, oxidative stress, and inflammatory response in children undergoing repair of simple congenital cardiac malformations. Material and Methods Fifty-nine children undergoing corrective cardiac surgery between November 2002 and November 2003 at the Bristol Royal Hospital for Children were randomized to receive either hypothermic (28 C) or normothermic (35 37 C). Neonates or patients requiring hypothermic circulatory arrest or complex repair of the pulmonary arterial system with periods of low flow were excluded from the study. The study was approved by the Accepted for publication March 16, 2005. Address reprint requests to Professor Angelini, Bristol Heart Institute, University of Bristol, Bristol Royal Infirmary, Bristol, BS2 8HW, UK; E-mail: G.D.Angelini@bristol.ac.uk. This article has been selected for the open discussion forum on the CTSNet Web Site: http://www.ctsnet.org.discuss 2005 by The Society of Thoracic Surgeons 0003-4975/05/$30.00 Published by Elsevier Inc doi:10.1016/j.athoracsur.2005.03.062

Ann Thorac Surg CAPUTO ET AL 2005;80:982 8 EFFECT OF TEMPERATURE IN PEDIATRIC OPEN HEART SURGERY Hospital Ethical Committee and parental informed consent was gained for all patients. Random treatment allocations, stratified by cyanosis, were generated by computer in advance of starting the study, using block randomization with varying block sizes. Allocation details were concealed in sequentially numbered, opaque sealed envelopes. Randomization was revealed to the surgeon after the start of the operation. Anesthetic and Surgical Technique Anesthetic technique was standardized as reported previously [9]. Temperature was monitored through nasopharyngeal, rectal, and skin probes. On establishment of, the patient was either cooled to a target nasopharyngeal temperature of 28 C or maintained at 35 37 C depending on randomization. St Thomas I crystalloid cardioplegia (4 to 6 C) was delivered in an antegrade fashion through the aortic root, as previously reported [10]. Topical cooling with cold saline solution was used in all patients. The flow was kept between 3.0 and 3.5 L m 2 min 1, and hematrocrit values between 25% and 30%. In the hypothermic group rewarming was started just prior to removal of the aortic cross clamp. Patients were rewarmed with a temperature difference of 8 C at the level of the heat exchanger between the blood and the rewarming fluid. Rewarming only took place in the normothermic group if the temperature was less than 36 C. A dopamine infusion was commenced in all patients at an initial rate of 5 mcg kg min prior to weaning from. Patients were weaned from once the nasopharyngeal temperature had reached 36 C. After the operation, all patients were admitted to the pediatric intensive care unit and were managed according to the unit protocols [9] by intensivists and pediatric cardiologists blinded to the randomization. Outcome Variables Primary end points were the release of troponin I and 8-isoprostane as measurements of myocardial reperfusion injury and oxidative stress, and the release of markers of the whole body inflammatory response (complement activation C3a; and interleukin (IL) -6, -8, and -10). The sample size calculated was based on our previous experience in similar studies on adult and pediatric patients [2, 3, 9, 10], where we have been able to demonstrate differences using groups of between 20 and 30 patients. Effect sizes corresponding to 0.8 to 1.1 standard deviations have been observed for the total area under curves describing the levels of biochemical markers during the first 24 hours after operation; eg, IL-6, IL-8, and troponin I. Therefore, a total sample size of 29 patients in each group was chosen in order to detect a difference between the groups equivalent to 0.75 standard deviations (ie, about the minimum observed in our previous studies) or greater, with 80% power at a 5% significance level (two-tailed). We also recorded all-cause in-hospital mortality and morbidity as previously reported [9, 10]. A full neurologic examination by a pediatric neurologist, blinded to the treatment allocation, was performed preoperatively and on a daily basis postoperatively. Sample Collection and Analysis Blood was drawn preoperatively, on removal of the aortic cross clamp, and 30 minutes, 6 hours, and 24 hours thereafter. This was immediately centrifuged at 4 C, at 4,000 rpm for 15 minutes. The resulting plasma was then immediately frozen in liquid nitrogen, prior to storage at 80 C. This was then analyzed for troponin I (enzymelinked immunosorbent assays (ELISA, Access, Beckman Coulter, Fullerton, CA), 8-isoprostane (EIA, Cayman Chemicals, Ann Arbor, MI), interleukin 6, 8, and 10 (ELISA, Amersham Biosciences, Amersham, UK), and complement activation C3a (ELISA, BD Biosciences Pharmingen, San Jose, CA). Statistical Analysis Operative characteristics were compared using the Wilcoxon rank sum test. Postoperative measurements of myocardial reperfusion injury, oxidative stress, and whole body inflammatory response were compared using regression analysis, with and without adjustment for aortic cross-clamp (ACC) time. All analyses were adjusted for preoperative readings. A variety of models describing the correlation structure between repeated measurements on the same patient were examined and the structure leading to the lowest value for the Schwarz s Bayesian information criterion was chosen in each case. All measurements followed a skewed distribution and were transformed to the logarithmic scale for analysis. Changes in treatment effect over time were assessed using the F-test and if statistically significant at the 10% level the treatment difference is reported separately at time points, otherwise an overall effect of treatment is given. Results are presented as geometric means and as a ratio of means with 95% confidence intervals, adjusted for differences in baseline response and ACC time. A post hoc analysis of 8-isoprostane, comparing the response when adjusting for (a) the preoperative reading and time and (b) the preoperative reading, ACC and times, was also carried out. The regression models were fitted using SAS version 8.2 (SAS Institute, Cary, NC). The rank correlation between myocardial reperfusion injury and oxidative stress (as measured by the area under the curve for troponin I [TnI] and 8-isoprostane) and clinical outcomes (duration of ventilation, inotropic support, and total postoperative hospital stay) was assessed for each treatment group separately and overall. Results 983 Thirty-one patients were randomized to the hypothermic group and 28 to the normothermic group. Overall, the median age was 78 months (interquartile range, 39 130). The median age was higher in the normothermic group, but the age distribution was similar (age range hypothermic: 4 191 months vs normothermic: 2 209 months). Overall, 42% of children were male. Reasons for operation included a predominance of atrial septal defect (ASD [37%]) and ventricular septal defect (VSD [33%]). There were 9 redo procedures (Table 1). The ACC and CARDIOVASCULAR

984 CAPUTO ET AL Ann Thorac Surg EFFECT OF TEMPERATURE IN PEDIATRIC OPEN HEART SURGERY 2005;80:982 8 Table 1. Baseline Characteristics Hypothermic (n 31) Normothermic (n 28) Age (months) 52 26 138 90.5 42 127 Male gender 16 52 9 32 Cyanotic 2 6 1 4 Indication ASD 10 32 12 43 VSD 12 39 7 29 AV valve repair 2 6 1 4 pavsd 4 13 1 4 RVOT repair 3 10 0 0 PA reconstruction 1 3 1 4 Pulmonary valvotomy 0 0 1 4 PAPVD 0 0 1 4 Sub AS 5 16 3 11 PVR 1 3 0 0 TVR 1 3 0 0 PDA 0 0 2 7 AoV repair 1 3 0 0 Redo 4 13 5 18 Body surface area (m 2 ) 0.69 0.50 1.2 0.90 0.65 1.12 Data are number and percentage or median and interquartile range. AoV aortic valve; ASD atrial septal defect; AV atrioventricular; cardiopulmonary bypass; PA pulmonary artery; PAPVD partial anomalous pulmonary venous drainage; pavsd partial atrioventricular septal defect; PDA patent ductus arteriosus; PVR pulmonary vascular resistance; RVOT right ventricular outflow tract; Sub AS subaortic stenosis; TVR tricuspid valve regurgitation; VSD ventricular septal defect. times were significantly longer in the hypothermic group (p 0.0096 and p 0.030, respectively) (Table 2). As anticipated, lowest body and reperfusion temperatures were both significantly lower in the hypothermic group (both p 0.0001). Clinical outcomes are reported in Table 2. Mortality and morbidity were similar across the two groups. Neurologic examination did not reveal any transient or permanent cerebrovascular event and no episodes of epilepsy were observed in either group. Troponin I and 8-Isoprostane Release Troponin I was released in a time-dependent fashion as previously described [9, 10] (Fig 1A). By 30 minutes after the release of the ACC, troponin levels were significantly raised compared to baseline, in both groups (Table 3). The levels continued to rise during the 6 hours after operation, and remained high (compared to baseline) at 24 hours. Overall, troponin I levels were 37% higher in the hypothermic group (ratio 1.37, 95% CI 1.00 to 1.88, p 0.053) (Table 3), and there was no evidence to suggest the treatment effect changed significantly over the time points measured (p 0.63). Adjusting for ACC time reduced the effect of hypothermia on troponin I (ratio 1.27, 95% CI 0.89 to 1.79, p 0.18). The release of 8-isoprostane was also time dependent (Fig 1B). In both groups levels rose significantly from baseline by the time of release of the ACC, and remained high 30 minutes later after which time levels declined. The pattern of response was the same in both groups. There was no evidence to suggest time-related differences between the two groups (p 0.62) but, throughout, 8-isoprostane levels were significantly higher in the hypothermic group. On average, 8-isoprostane levels were 84% higher in the hypothermic group (ratio 1.84, 95% CI 1.22 to 2.78, p 0.0045), and the difference remained significant after adjusting for ACC time (ratio 1.79, 95% CI 1.23 to 2.60, p 0.0028), time (ratio 1.94, 95% CI Table 2. Operative Characteristics and Clinical Outcomes Hypothermic (n 31) Normothermic (n 28) Cross-clamp time (min) 32 27 49 22 14.5 33 time (min) 62.0 47 79 48.5 39 66 Lowest body temperature ( C) 28.2 28 28.6 35.0 34.5 35.2 Reperfusion temperature ( C) 33.8 32 34.9 36.1 35.9 36.5 In-hospital mortality 0 0 Blood loss (ml) 130 95 240 160 127.5 267.5 Fluid balance (ml) 46 280 143 34.75 262.5 139 Red cell transfusion 8 29 8 29 ml/kg 9.52 6.78 16.6 9.58 6.76 19.7 Platelet/FFP transfusion 5 18 6 21 ml/kg 5.17 4.90 5.54 9.95 4.89 10.0 Inotropic support 24 77 17 61 Total ( g/kg/10 3 ) 18.15 11.75 38.25 15.0 10.0 20.5 Peak ( g/kg/10 3 ) 6.5 5 10 5 5 10 Duration (hours) 20 14.5 24 19 12 22 Postoperative ventilation time (hours) 8 6 23 8.5 6 21 Postoperative hospital stay (days) 5 5 7 5 5 6 Data are number and percentage or median and interquartile range. cardiopulmonary bypass; FFP fresh frozen plasma.

Ann Thorac Surg CAPUTO ET AL 2005;80:982 8 EFFECT OF TEMPERATURE IN PEDIATRIC OPEN HEART SURGERY 985 release (p 0.0079), a suggested difference 30 minutes later (p 0.072), but no difference thereafter. Adjusting for ACC time reduced the significance of the time-related differences between the groups for both markers (IL-8, p 0.13; IL-10, p 0.11). Differences in IL-10 response at ACC (p 0.0065) and 30 minutes later (p 0.018) were indicated, but otherwise, as before, no significant differences were found. The release of complement activation C3a was similar in the two groups and there was no evidence to suggest statistically significant time-related differences (p 0.15); the response was similar across the two groups throughout (p 0.16). Adjusting for ACC time did not change this conclusion; complement activation C3a remained similar in both groups (p 0.38). CARDIOVASCULAR Comment This study is a randomized trial comparing hypothermic and normothermic in pediatric heart surgery. The study shows that normothermic is associated with reduced oxidative stress, compared with hypothermic, and similar myocardial reperfusion injury and whole body inflammatory response. Fig 1. Time-related plasma changes in the geometric mean troponin I (A) and 8-isoprostane (B) in the hypothermic ( ) and normothermic ( ) groups. (ACC end of aortic cross clamp; SE standard error.) 1.29 to 2.89, p 0.0018), and both ACC and times (ratio 1.80, 95% CI 1.26 to 2.57, p 0.0017). The correlation between both TnI and 8-isoprostane release with ventilation time, duration of inotropic support, and length of hospital stay was weak; all correlations were less than or equal to 0.47 (Table 4). Inflammatory Markers Release The release of IL-6, IL-8, and IL-10 varied over time (Table 3). For IL-6 the pattern of response over time was similar in the two groups and there was no evidence to suggest statistically significant time-related differences (p 0.27); the response was similar across the two groups throughout (p 0.79) and adjustment for ACC time did not change this conclusion (p 0.59). In contrast, for IL-8 and IL-10, time-related differences between the two treatment groups were suggested (IL-8, p 0.069; IL-10, p 0.059) (Table 3), although there was no statistically significant difference at individual time points between the two groups. For IL-10 there was a higher release in the normothermic group over time and examining differences at individual time points highlighted a statistically significant difference in IL-10 response at ACC Oxidative Stress and Myocardial Reperfusion Injury Reperfusion injury is characterized by the formation of oxygen-free radicals such as superoxide and peroxide, which are responsible for cell membrane degradation through lipid peroxidation. Myocardial reperfusion injury has been studied in great detail, and it is now evident that oxidative stress is a major part of the cellular mechanism of the resulting myocardial damage [11 14]. The 8-isoprostane has been shown to be a reliable marker for the volume of myocardium exposed to oxidant stress during cardioplegic arrest and acute myocardial infarction [11 14]. The increased 8-isoprostane levels in the hypothermic group were associated with elevated TnI release, suggesting an effect of temperature on the degree of myocardial reperfusion injury. The effect on TnI, however, was reduced after adjusting for ACC time. The 8-isoprostane measured in peripheral blood is not a specific marker of myocardial reperfusion injury. It is most likely, therefore, that the higher levels of this marker in the hypothermic group reflect a combination of myocardial and whole body oxidative stress. We can only hypothesize as to the exact source, but as the isoprostanes are breakdown products of lipid peroxidation, any organ system rich in lipids could be responsible. This, we believe, warrants further investigation. The view that normothermic cardiopulmonary bypass may cause rewarming of the heart by collateral blood flow, and from contiguous structures resulting in less effective hypothermic cardioplegic myocardial protection, was not supported by our study. The differences in oxidative stress between the two groups were not translated into differences in clinical outcomes, as demonstrated by the weak correlations between TnI and 8-isoprostane release and ventilation time, inotropic support, and hospital stay.

986 CAPUTO ET AL Ann Thorac Surg EFFECT OF TEMPERATURE IN PEDIATRIC OPEN HEART SURGERY 2005;80:982 8 Table 3. Biochemical Markers: Effect Sizes Hypothermic Geometric mean a (SE) b Normothermic Geometric mean a (SE) b Ratio 95% CI p Value Troponin (ng/ml) Baseline reading c 0.01 (0.0 0.02) 0.01 (0.0 0.02) ACC 0.71 (1.17) 0.55 (1.14) 1.29 30 mins 2.55 (1.17) 1.82 (1.13) 1.40 6 hours 6.70 (1.15) 4.38 (1.13) 1.53 24 hours 4.71 (1.13) 3.70 (1.13) 1.27 Test for common ratio across time points p 0.64 Estimate of common ratio 1.37 1.00 to 1.88 0.053 8-isoprostane (pg/ml) Baseline reading c 1.95 (1.33 5.93) 1.91 (0.77 3.44) ACC 7.44 (1.23) 4.33 (1.17) 1.72 30 mins 4.79 (1.19) 2.63 (1.15) 1.82 6 hours 2.81 (1.18) 1.32 (1.20) 2.14 24 hours 1.42 (1.29) 0.83 (1.27) 1.72 Test for common ratio across time points p 0.62 Estimate of common ratio 1.84 1.22 to 2.78 0.0045 IL-6 (pg/ml) Baseline reading c 0.36 (0.0 1.70) 0.0 (0.0 0.69) ACC 0.31 (1.46) 0.24 (1.42) 1.29 30 mins 0.37 (1.40) 0.51 (1.42) 0.73 6 hours 8.02 (1.26) 11.39 (1.22) 0.70 24 hours 13.11 (1.34) 11.21 (1.22) 1.17 Test for common ratio across time points p 0.27 Estimate of common ratio 0.94 0.57 to 1.54 0.79 IL-8 (pg/ml) Baseline reading c 5.58 (2.39 7.75) 4.88 (0.0 6.85) ACC 8.97 (1.11) 10.86 (1.10) 0.83 0.61 to 1.10 0.19 30 mins 13.08 (1.12) 12.55 (1.13) 1.04 0.74 to 1.47 0.81 6 hours 16.70 (1.10) 14.94 (1.12) 1.12 0.82 to 1.52 0.47 24 hours 15.36 (1.10) 12.61 (1.13) 1.22 0.90 to 1.65 0.20 Test for common ratio across time points p 0.069 IL-10 (pg/ml) Baseline reading c 4.12 (1.20 11.8) 2.72 (1.20 4.90) ACC 8.97 (1.19) 20.52 (1.28) 0.44 0.24 to 0.80 0.0079 30 mins 114.0 (1.18) 175.3 (1.18) 0.65 0.41 to 1.04 0.072 6 hours 30.88 (1.20) 33.46 (1.19) 0.92 0.56 to 1.52 0.75 24 hours 7.68 (1.13) 6.72 (1.13) 1.14 0.81 to 1.62 0.45 Test for common ratio across time points p 0.059 C3 (ng/ml) Baseline reading c 1.31 (1.00 1.60) 1.27 (1.10 1.61) ACC 2.19 (1.05) 2.26 (1.04) 0.97 30 mins 2.85 (1.06) 2.57 (1.06) 1.11 6 hours 1.20 (1.06) 1.05 (1.04) 1.14 24 hours 1.11 (1.07) 1.05 (1.05) 1.06 Test for common ratio across time points p 0.15 Estimate of common ratio 1.07 0.97 to 1.17 0.16 a Adjusted for differences in baseline reading; b A 95% confidence interval for a geometric mean is given by (geometric mean) (SE) 1.96 ; c Median and interquartile range. ACC end of aortic cross-clamp; CI confidence interval; cardiopulmonary bypass; SE standard error.

Ann Thorac Surg CAPUTO ET AL 2005;80:982 8 EFFECT OF TEMPERATURE IN PEDIATRIC OPEN HEART SURGERY Table 4. Spearman s Rank Correlations Hypothermic Troponin (AUC) a Normothermic Overall Hypothermic 8-isoprostane (AUC) a Normothermic 987 Overall CARDIOVASCULAR Ventilation time 0.47 0.19 0.33 0.06 0.10 0.13 Duration of inotropic support 0.28 0.07 0.22 0.31 0.06 0.02 Length of hospital stay 0.21 0.07 0.22 0.08 0.24 0.15 a Missing values were replaced by the preceding value when calculating the area under the curve. AUC area under the curve; cardiopulmonary bypass. Inflammatory Response This study shows that the inflammatory response triggered by during low risk pediatric heart surgery is similar in the hypothermic and normothermic groups. The proinflammatory cytokines levels (IL-6 and IL-8) increased significantly 6 hours postoperatively and remained significantly elevated after 24 hours, indicating the induction of a systemic inflammatory response not temperature dependent. This was associated with a significant increase in complement activation (C3a) 30 minutes after the end of the ischemic time. A similar pattern of proinflammatory cytokines production was seen in neonates after arterial switch operation [15]. Our findings, however, contrast with previous animal data [5, 6], which have documented a reduction of the inflammatory response to by hypothermia compared with normothermia, mainly as a result of reduced white blood cell activation and disturbances of the microcirculation. In contrast to the release of IL-6 and IL-8, and in agreement with previous studies [16 18], IL-10 release showed a significant peak as early as 30 minutes after the end of the ischemic time, and tended to return toward preoperative levels after 24 hours. The immediate IL-10 release was significantly higher in the normothermic group and was associated with a lower IL-8 liberation 6 hours postoperatively compared with the hypothermic group. Interleukin 10 is an antiinflammatory cytokine that deactivates monocytes and macrophages and inhibits the synthesis of proinflammatory cytokines such as tumor necrosis factor (TNF)- and IL-8 [19]. Therefore, early IL-10 liberation during may down-regulate the production of proinflammatory cytokines, and play a protective role in pediatric patients undergoing open heart surgery. The view that normothermia may be associated with an exaggerated inflammatory response to bypass was therefore not supported in this study. Limitations of the Study One of the confounding factors in explaining the results of the study is the difference in ischemic and times between the two groups. While it is expected that the times would be shorter in the normothermic group due to the avoidance of time taken to cool and subsequently rewarm, it is surprising that the ACC times were significantly different. Troponin values are known to correspond to the duration of ischemia, and hence we would expect them to be lower in the group with the shorter ACC time. Hence we adjusted for ACC time and even though the difference in TnI levels was then no longer statistically significant, the values were still lower overall in the normothermic group. Although we did not investigate cognitive outcomes, all patients received a full preoperative and postoperative neurologic examination. Several studies in an adult population have addressed this question with conflicting results [20, 21]. To the best of our knowledge, no data are available in a pediatric population. The study excluded neonates and patients with complex congenital heart pathologies and was designed to detect changes in biochemical markers but not in clinical outcome. Using estimates for ventilation time, intensive therapy unit (ITU), and hospital stay from a previous study carried out at our institution [10], we calculate that a total sample size of 180 patients would be required to detect a 32% reduction in ventilation time (15 hours), a 23% reduction in length of ITU stay (26 hours), and a 17% reduction in length of hospital stay (1.8 days) with 90% power, assuming a 5% level of statistical significance and a two-tailed test. Conclusions In infants and children requiring correction of simple congenital heart defects, normothermic is associated with reduced oxidative stress and similar myocardial reperfusion injury and inflammatory response compared with hypothermic. A larger study with a sample size capable to detect differences in clinical outcome is now warranted. The British Heart Foundation, National Heart Research Fund, and Garfield Weston Trust supported this work. We would like to thank Mark Ginty and Svitlana Korolchuk for performing the biochemical analyses, and Prof Andrew Wolf, Dr Depesh Trivedi, and the pediatric nursing staff for their support. References 1. Birdi I, Regragui I, Izzat MB, Bryan AJ, Angelini GD. Influence of normothermic systemic perfusion temperature during coronary artery surgery: a randomized prospective study. J Thorac Cardiovasc Surg 1997;114:475 81. 2. Birdi I, Caputo M, Underwood M, Bryan AJ, Angelini GD. Influence of normothermic systemic perfusion temperature on cold myocardial protection during coronary artery bypass surgery. Cardiovasc Surg 1999;7:369 74.

988 CAPUTO ET AL Ann Thorac Surg EFFECT OF TEMPERATURE IN PEDIATRIC OPEN HEART SURGERY 2005;80:982 8 3. Birdi I, Caputo M, Underwood M, Bryan AJ, Angelini GD. The effects of cardiopulmonary bypass temperature on inflammatory response following cardiopulmonary bypass. Eur J Cardiothorac Surg 1999;16:540 5. 4. Edmunds LH. Inflammatory and immunological response to cardiopulmonary bypass. In: Cardiopulmonary bypass in neonates, infants and young children. Jonas RA, Elliot MJ, eds. Oxford: Butterworth-Heinemann; 1994:225 41. 5. Wagner FM, Schiller W, Dilg G, Depner C, Welz A, Lacour- Gayet F. Direct visualization of the influence of normothermic as opposed to hypothermic cardiopulmonary bypass on the systemic microcirculation in neonatal pigs. Cardiol Young 2001;11:532 8. 6. Anttila V, Hagino I, Zurakowski D, Lidov HG, Jonas RA. Higher bypass temperature correlates with increased white cell activation in the cerebral microcirculation. J Thorac Cardiovasc Surg 2004;127:1781 8. 7. Corno AF, von Segesser LK. Is hypothermia necessary in paediatric cardiac surgery? Eur J Cardiothorac Surg 1999;15: 110 1. 8. Durandy Y, Hulin S, Lecompte Y. Normothermic cardiopulmonary bypass in pediatric surgery. J Thorac Cardiovasc Surg 2002;123:194 5. 9. Imura H, Caputo M, Parry A, Pawade A, Angelini GD, Suleiman M-S. Age-dependent and hypoxia-related differences in myocardial protection during pediatric open heart surgery. Circulation 2001;103:1551 6. 10. Modi P, Reeves B, Pawade A, et al. Myocardial metabolic changes during pediatric open heart surgery: a randomized controlled study of three cardioplegic techniques. J Thorac Cardiovasc Surg 2004;128:67 75. 11. Mehlhorn U, Krahwinkel A, Geissler HJ, et al. Nitrotyrosine and 8-isoprostane formation indicate free radical-mediated injury in hearts of patients subjected to cardioplegia. J Thorac Cardiovasc Surg 2003;125:178 83. 12. Blasig IE, Grune T, Schonheit K, et al. 4-Hydroxynonenal, a novel indicator of lipid peroxidation for reperfusion injury of the myocardium. Am J Physiol 1995;269:H14 22. 13. Delanty N, Reilly MP, Pratico D, et al. 8-Epi PGF 2 generation during coronary reperfusion: a potential quantitative marker of oxidant stress in vivo. Circulation 1997;95:2492 9. 14. Reilly MP, Delanty N, Roy L, et al. Increased formation of the isoprostanes IPF 2 -I and 8-epi-prostaglandin F 2 in acute coronary angioplasty: evidence for oxidant stress during coronary reperfusion in humans. Circulation 1997;96:3314 20. 15. Hovels-Gurich HH, Vazquez-Jimenez JF, Silvestri A, et al. Production of proinflammatory cytokines and myocardial dysfunction after arterial switch operation in neonates with transposition of the great arteries. J Thorac Cardiovasc Surg 2002;124:811 20. 16. Alcaraz AJ, Sancho L, Manzano L, et al. Newborn patients exhibit an unusual patter of interleukin 10 and interferon serum levels in response to cardiac surgery. J Thorac Cardiovasc Surg 2002;123:451 8. 17. Sugita T, Watasida S, Katsuyama K, Nakajima Y, Yamamoto R, Mori A. Interleukin-10 concentration in children undergoing cardiopulmonary bypass. J Thorac Cardiovasc Surg 1996;112:1127 8. 18. Tarnok A, Hambsch J, Schneider P. Cardiopulmonary bypass-induced increase of serum IL-10 levels in children. J Thorac Cardiovasc Surg 1998;115:475 7. 19. Rennick D, Berg D, Holland G. Interleukin-10: an overview. Prog Growth Factor Res 1992;4:207 27. 20. Martin TD, Craver JM, Gott JP, et al. Prospective, randomized trial of retrograde warm blood cardioplegia: myocardial benefit and neurologic threat. Ann Thorac Surg 1994;57:298 302. 21. Grigore AM, Mathew J, Grocott HP, et al. Prospective randomized trial of normothermic versus hypothermic cardiopulmonary bypass on cognitive function after coronary artery bypass graft surgery. Anesthesiology 2001;95:1110 9.