Continuous Cerebral Perfusion for Aortic Arch Repair: Hypothermia Versus Normothermia

ORIGINAL ARTICLES: PEDIATRIC CARDIAC

PEDIATRIC CARDIAC SURGERY: The Annals of Thoracic Surgery CME Program is located online at http://cme.ctsnetjournals.org. To take the CME activity related to this article, you must have either an STS member or an individual non-member subscription to the journal. PEDIATRIC CARDIAC

Continuous Cerebral Perfusion for Aortic Arch Repair: Hypothermia Versus Normothermia Mohamed Ly, MD, François Roubertie, MD, Emre Belli, MD, Oswin Grollmuss, MD, Minh Thanh Bui, MD, Regine Roussin, MD, Emmanuel Lebret, MD, André Capderou, MD, and Alain Serraf, MD, PhD Departments of Congenital Heart Surgery, and Congenital Intensive Care Unit, Marie Lannelongue Hospital, University ParisSud, Le Plessis-Robinson; and Faculty of Medicine, University Paris-Sud, U999 INSERM, Orsay, France

Background. Deep hypothermia with circulatory arrest (CA) is routinely used for aortic arch repair. Antegrade selective cerebral perfusion (ASCP) has been proposed to avoid neurologic complications. The optimal temperature during aortic arch repair with ASCP is not well established. We therefore compared early outcomes of patients undergoing aortic arch repair associated with intracardiac repair with ASCP either with hypothermia (< 28°C, group I; n ⴝ 70) or normothermia (> 34°C, group II; nⴝ40). Methods. From 2002 to 2008, 110 consecutive patients with interrupted aortic arch (n ⴝ 40) or coarctation of the aorta (n ⴝ 70) and cardiac anomalies underwent intracardiac and aortic arch repair without CA. Median age at repair was 12 days. Full cardiopulmonary bypass (CPB), high hematocrit, and high rates of flow were used. ASCP flow was adjusted to maintain arterial pressure at greater than or equal to 50 mm Hg. ASCP was achieved either

through a Gore-Tex (W. L. Gore & Associates, Inc., Elkton, MD) graft to the innominate artery (n ⴝ 36) or by direct cannulation (n ⴝ 74). An electroencephalogram (EEG) was continuously monitored and 30 patients were monitored by near-infrared spectroscopy (NIRS). Results. Preoperative data were similar in both groups. Group II demonstrated higher ASCP flows (p < 0.001). Time to extubation, stay in the intensive care unit (ICU), and postoperative urine output and lactate levels were similar between groups. Mortality was 8.5% in group I versus 10% in group II. During the postoperative course, there were no clinical or electrical neurologic events in either group. Conclusions. Antegrade selective cerebral perfusion can safely avoid CA. No worse, early, or long-term effects of ASCP with normothermia were found. (Ann Thorac Surg 2011;92:942– 8) © 2011 by The Society of Thoracic Surgeons

T

not yet been identified, we report our experience with 110 consecutive aortic arch reconstructions performed under normothermic or hypothermic temperature conditions.

he objectives of every pediatric cardiac operation are technical success and absence of iatrogenic injury from inadequate myocardial or brain protection. Although clinical cardiac surgery generally has become safe and effective with warm blood myocardial protection techniques [1–3], aortic arch reconstruction has frequently required deep hypothermic circulatory arrest (DHCA) that might be associated with adverse neurologic outcome [4, 5]. Antegrade selective cerebral perfusion (ASCP) has been proposed to avoid DHCA. Asou and colleagues described a technique of SCP during aortic arch repair in neonates [6]. The optimal temperature for the aortic arch repair remains unclear. As the optimal temperature strategy for aortic arch repair has

Accepted for publication March 7, 2011. Presented at the Forty-sixth Annual Meeting of The Society of Thoracic Surgeons, Fort Lauderdale, FL, Jan 25–27, 2010. Address correspondence to Dr Ly, Department of Congenital Heart Surgery, Marie Lannelongue Hospital, 133 avenue de la Résistance 92350 Le Plessis-Robinson, France; e-mail: [email protected].

© 2011 by The Society of Thoracic Surgeons Published by Elsevier Inc

Material and Methods Patients Between 2000 and 2008 110 consecutive patients underwent anatomic reconstruction of the aortic arch with intracardiac repair at our institution. We retrospectively reviewed all preoperative, intraoperative, and postoperative data. Preoperative parental consent was obtained by interview, and the study was approved by the local ethics committee. Seventy patients (group I) underwent operation under hypothermia, which was defined as rectal temperature equal to or lower than 28°C. Forty patients (group II) underwent operation under normothermia, which was defined as a rectal temperature equal to or higher than 34°C (Table 1). All operations were performed by 2 senior surgeons. The presence of 1 or more associated cardiac 0003-4975/$36.00 doi:10.1016/j.athoracsur.2011.03.006

Ann Thorac Surg 2011;92:942– 8

LY ET AL CEREBRAL PERFUSION HYPOTHERMIA VS NORMOTHERMIA

943

Median age (days) Median weight (kg) Prostaglandin E1 Preoperative ventilation Premature DiGeorge syndrome Diagnosis and associated lesions Interrupted aortic arch Coarctation Taussig-Bing malformation Truncus arteriosus Transposition of the great arteries Subaortic stenosis Aortic valve bicuspid Multiple ventricular septal defect Pulmonary artery window Double-outlet right ventricle

Hypothermia (ⱕ 28°C), N ⫽ 70

Normothermia (ⱖ 34°C), N ⫽ 40

p Value

12, range (3–180) 3.05, range (2–6.5) 58 4 3 9

13, range (4–660) 3.29, range (2.1–7.2) 27 5 1 8

0.43 0.3 0.1 0.3 0.9 0.4

23 47 6 7 17 4 5 2 7 1

17 23 2 1 7 2 3 4 1 1

0.4 0.4 0.75 0.2 0.5 0.7 0.9 0.637 0.377 ⬎9

tions, and pulse oximetry. The cerebral monitoring consisted of electroencephalography for all patients and NIRS in 30 patients (Table 2). Anesthesia was induced with sevoflurane and maintained with intravenous midazolam and ketamine. Pancuronium was used for muscle relaxation. Analgesia was ensured by sufentanil infusion. Median sternotomy was performed for all patients and the aortic arch and cerebral vessels were extensively prepared and dissected. The main pulmonary artery and its branch arteries were mobilized. After resection of all ductal tissues and mobilization of the descending aorta, aortic arch repair was performed under selective cerebral perfusion. Cardiopulmonary bypass (CPB) was established by bicaval cannulation and a single aortic cannula placed in the right side of the distal ascending aorta advanced into the innominate artery during the aortic arch reconstruction (74 patients) or through a 3.5-mm polytetrafluoroethylene (PTFE) graft anastomosed to the right innominate artery (36 patients). The aortic anomalies were corrected first. During a limited period of ASCP

anomalies was the rule, and specific lesions are summarized in Table 1. Preoperatively the diagnosis of aortic arch anomalies was confirmed by clinical assessment and 2-dimensional echocardiography with Doppler ultrasonography. Prostaglandin E1 was administered in 85% of patients to improve perfusion of the distal aorta through the ductus arteriosus. Balloon atrial septostomy (BAS) was performed in 17 patients presenting with transposition of the great arteries and an intact or restrictive interventricular septum.

Operative Technique and Cardiopulmonary Bypass Management Intensive preoperative management played an important role in the care of these infants, improving their hemodynamic status before surgery. The following were monitored continuously: invasive arterial and central venous pressures, electrocardiogram, rectal and nasopharyngeal temperatures, inspiratory and expiratory gas concentraTable 2. Bilateral Neurologic NIRS Monitoring During Procedure Incision

Initiate CPB

Start ASCP

Unclamping of Aorta

End of CPB

Left signal rSO2i G I rSO2i G II

50 (42–64) 52 (46–60)

68 (62–74) 68 (56–78)

58 (50–78) 56 (48–70)

55 (50–71) 55 (50–65)

60 (54–78) 62 (58–80)

Right signal rSO2i G I rSO2i G II

52 (46–60) 52 (47–61)

69 (62–78) 67 (51–78)

53 (47–73) 56 (50–72)

59 (50–69) 57 (50–68)

62 (54–74) 66 (58–78)

Note: Results are expressed as median and ranges. Regional cerebral oxygen saturation index (rSO2i) expressed as percentage. ASCP ⫽ antegrade selective cerebral perfusion; group; NIRS ⫽ near-infrared spectroscopy.

CPB ⫽ cardiopulmonary bypass;

G I ⫽ hypothermia group;

G II ⫽ normothermia

PEDIATRIC CARDIAC

Table 1. Population Characteristics

944

LY ET AL CEREBRAL PERFUSION HYPOTHERMIA VS NORMOTHERMIA

Ann Thorac Surg 2011;92:942– 8

PEDIATRIC CARDIAC

gistic regression model was expressed as odds ratio (OR) and 95% confidence intervals (CI). Event-free survival and freedom from reoperation were calculated by actuarial method with 95% CI and comparison between groups by the log-rank Cox-Mantel test. P values less than 0.05 were considered statistically significant.

the zone of coarctation was resected with all ductal tissues. The descending aorta was anastomosed to the arch by an extended end-to-end anastomosis. In the presence of severe aortic arch or ascending aorta hypoplasia, or both, patch enlargement was performed. Subsequently the associated cardiac lesions were repaired after return to full-flow CPB. The intracardiac repair was performed by previously described techniques for each specific malformation. In normothermia the blood prime was warmed to 35°C to 37°C and this temperature was maintained during the entire bypass time. Arterial flow was maintained at 150 to 200 mL/kg/min. The hematocrit was 32% to 35% during CPB time. Our bypass circuit was miniaturized by using minimal priming volume. During the aortic arch repair the cerebral vessels were snared and the flow was lowered to one third of the full flow of CPB. The blood pressure was measured in the right radial artery. ASCP flow was adjusted to maintain an arterial pressure level of 40 to 50 mm Hg.

Results Intraoperative data are summarized in Table 3.

Postoperative Courses and Complications We uniformly found shorter cross-clamp and CPB times in patients who underwent the operation under normothermia. The difference was statistically significant between the 2 groups. No patient required peritoneal dialysis. Postoperative data are shown in Table 4. Lactate levels were similar between the 2 groups at the end of CPB. However the duration of normalization after 24 hours was shorter in the normothermia group (Fig 1). Analysis by ANOVA with correction for repeated measures revealed no significant difference between the 2 groups (p ⫽ 0.35). Diuresis and renal function were similar during the early postoperative period. Two patients who subsequently required extracorporeal membrane oxygenation (ECMO) support died intraoperatively of uncontrollable bleeding. However there was a significantly higher rate of reopening for bleeding in patients who underwent operation under deep hypothermia (6 patients; p ⫽ 0.14). Incidence of supraventricular arrhythmias, right and left ventricular function, duration of mechanical ventilation, and intensive care unit stay were similar in the 2 groups. Complications are listed in Table 5.

Statistics Follow-up was achieved for all survivors by means of telephone calls and letters with the referring pediatric cardiologists. All patients underwent neurologic examination and routine postoperative cerebral ultrasonographic studies. Data were collected retrospectively and analyzed for all patients. Statistical analysis was performed using Stat View 5 (SAS Institute, Cary, NC). Continuous variables were expressed as mean and standard deviation, and noncontinuous variables were expressed as medians and ranges. Categoric variables were compared using the ␹2 test and continuous variables were compared with the Student’s t test, the MannWhitney U test, or by analysis of variance (ANOVA) repeated measures as appropriate. The multivariate lo-

Table 3. Intraoperative Data

Cerebral perfusion PTFE graft Aortic cannulation Techniques of repair Direct anastomosis RV/PA conduit Pericardial patch Pulmonary artery patch Subclavian artery patch Status at end of operation Aortic clamp (min) CPB time (min) Flow of cerebral perfusion (mL/kg/min) Perfusion pressure (mm Hg) Duration of perfusion (min) Delayed sternal closure (%) CPB ⫽ cardiopulmonary bypass;

PA ⫽ pulmonary artery;

Hypothermia (ⱕ 28°C), N ⫽ 70

Normothermia (ⱖ 34°C), N ⫽ 40

4 66

27 13

13 30 6 20 1

20 6 7 6 1

87.9 ⫾ 31.2 173 ⫾ 71.3 47.9 ⫾ 23.3 53.7 ⫾ 8.3 32 ⫾ 11.9 75

61.7 ⫾ 29.3 112.3 ⫾ 64.8 52.7 ⫾ 27.9 49.3 ⫾ 8.6 26.3 ⫾ 9.1 65

PTFE ⫽ polytetrafluoroethylene;

RV ⫽ right ventricular.

p Value

0.0001 0.0001 0.18 0.012 0.0015 0.32

Ann Thorac Surg 2011;92:942– 8

945

LY ET AL CEREBRAL PERFUSION HYPOTHERMIA VS NORMOTHERMIA

Revision for bleeding (%) Left ventricular failure (%) Extracorporeal membrane oxygenation (n) Supraventricular arrhythmia (%) Atrioventricular block (n) Time to extubation (hr) Intensive care unit stay (days) Diuresis first day (mL/kg) Diuresis after 2 days (mL/kg) Diuresis after 3 days (mL/kg) Lactate levels (mmol/L) After CPB Lactate level (mmol/L) 12 hr after CPB Lactate level (mmol/L) 24 hr after CPB Creatinemia ␮m/L day 1 Creatinemia (␮m/L) day 2 Creatinemia (␮m/L) day 3

Hypothermia (ⱕ 28°C) N ⫽ 70

Normothermia (ⱖ 34°C) N ⫽ 40

p Value

6 32 1 17 1 181.9 ⫾ 173 9.4 ⫾ 7.8 36.6 ⫾ 22 74.5 ⫾ 31.1 123.7 ⫾ 45.2 5.3 ⫾ 2.3

0 22.5 1 7 1 181.2 ⫾ 200 10 ⫾ 9.3 37.6 ⫾ 25.3 85.8 ⫾ 40.9 119 ⫾ 45.9 5.3 ⫾ 2.5

0.14 0.3 ⬎.9 0.5 ⬎.9 0.6 0.7 0.8 0.2 0.5 0.9

3⫾1 2.5 ⫾ 5.3 58 ⫾ 15.5 64 ⫾ 23 66 ⫾ 21

2.9 ⫾ 1.5 1.7 ⫾ 0.9 61 ⫾ 15.1 68 ⫾ 2.3 66 ⫾ 28

0.12 0.35 0.12 0.38 0.35

CPB ⫽ cardiopulmonary bypass.

Reoperation Causes of reoperation are shown in Table 6. During the follow-up period, a total of 28 reoperations were performed for repair of associated anomalies (reoperation rate, 25%). Freedom from reoperation was 67% (95% CI, 76% to 93%) at 1 year and 67% (95% CI, 57% to 79%) at 3 years in the hypothermic group and 77% (95% CI, 63% to 90%) at 1 year and 66% (95% CI, 49% to 82%) at 3 years in the normothermic group. Persistent aortic arch obstruction or restenosis was seen in 9 patients (8%). In our series the use of a direct anastomosis was associated with a higher incidence of this complication (21% of patients). Eight patients with this complication underwent balloon dilation. One patient underwent surgery to enlarge the aortic arch with a homograft patch. Severe

subaortic stenosis with bicuspid valve required surgical treatment in 3 patients. One of these patients had a Ross-Kono procedure. The other 2 patients underwent membrane resection.

Survival In the early postoperative period, hospital death accounted for 6 patients in group I (8.5%), and 4 patients in group II (10%) (p ⬎ 0.9). Cardiac-related events caused early death in 10 patients (Table 7). Risk factor analysis of the preoperative and postoperative variables within the 2 groups revealed only CPB time as a risk factor. ASCP duration did not affect early mortality in either group (OR, 1,026; 95% CI [1,006 –1,046]; p ⫽ 0.01.The mean follow-up time was 38.8 ⫾ 19.5 months (range, 2 to 80 months). There was no late mortality. All patients were free from neurologic symptoms and at follow-up were growing and developing normally.

Comment The cerebral protection methods in aortic arch operations have been well defined in recent reports. In the past few years, we have assisted in the resurgence of interest in the ASCP technique because of its favorable postopTable 5. Complication Data

Fig 1. Evolution of lactate levels during initial 24 hours after cardiopulmonary bypass (p⫽ 0.35).

Pulmonary infection Diaphragm paralysis Bronchial compression Pulmonary hypertension Mediastinitis

Group I

Group II

1 4 3 2 1

3 1 1 3 1

PEDIATRIC CARDIAC

Table 4. Postoperative Data

946

LY ET AL CEREBRAL PERFUSION HYPOTHERMIA VS NORMOTHERMIA

Table 6. Causes of Early and Late Reoperation

PEDIATRIC CARDIAC

Recurrent coarctation Subaortic membrane resection Bronchial compression Replacement of RV-PA conduit Mediastinitis Diaphragm paralysis PA ⫽ pulmonary artery;

Group I

Group II

5 2 4 3 1 4

4 1 1 0 2 1

RV ⫽ right ventricle.

erative outcomes. This cerebral protection method depends on bihemispheric perfusion of the brain through the circle of Willis. Therefore perfusion of 1 or more aortic arch branches would perfuse the brain. Nevertheless the appropriate flow rate for the brain and temperature for perfusion remains to be defined. Several studies revealed good results using ASCP with deep or moderate hypothermia [7–9]. However there is no prospective and randomized study concerning ASCP in normothermia. Touati and colleagues [10] proposed carrying out ASCP in a totally normothermic aortic arch replacement with perfusion of the lower part of the body during the time of the aortic arch repair by means of a balloon cannula placed in the descending aorta through the arch opening,

Ann Thorac Surg 2011;92:942– 8

with satisfactory results. Hyperthermia during the rewarming phase of hypothermic CPB has been shown to be associated with adverse neurologic outcomes. Kaukuntla and associates [11] demonstrated that during rewarming all peripheral sites underestimate brain temperature, and caution is required to avoid hyperthermic arterial inflow, which may result in brain hyperthermia. Wong and associates [12] demonstrated after using a battery of neuropsychological tests on 34 patients randomized to warm (⬎ 34°C) or hypothermic CPB (28°C) that no benefit was seen with the hypothermic technique. Rasmussen and colleagues [13] found no difference in the release of brain-specific proteins (S-100␤ protein) between patients undergoing normothermic CPB versus patients undergoing hypothermic CPB, but blood loss was higher after normothermic CPB. Jacobs and coworkers [14] in their series with ASCP under moderate hypothermia (28° to 30°C) found the transient neurologic deficit rate to be 4% and concluded that there was no correlation between long SCP duration and neurologic injury. Neonates with congenital heart disease may have both preoperative and postoperative brain injury. Indeed previous studies have shown that after cardiac operations with CPB 36% to 73% of neonates had new intraparenchymal brain injury observed on magnetic resonance

Table 7. Causes of Early Mortality Pt

Sex

Age Weight (days) (kg)

Diagnosis

Cardiopulmonary Bypass Temperature

Complications

Causes of Death and (Days Op)

1 Female

17

2.7

IAA ⫹ truncus arteriosus

Hypothermia

2 Male 3 Male,

40 7

3.6 3.2

Hypothermia Hypothermia

Left ventricular failure, Major LV dysfunction RV/PA Contegra (45) tube thrombosis AMI, ECMO Bleeding (1) AV block ARDS (53)

4 Female

10

2

Hypothermia

Myocardial ischemia

Normothermia

Coronary dissection during cardioplegia Myocardial ischemia by RV/PA conduit Shelhigh compression LV dysfunction , balloon valvuloplasty at 28 days AV block

5 Male

5

2.7

IAA ⫹ truncus arteriosus IAA ⫹ Taussig-Bing malformation CoA ⫹ VSD ⫹ LV failure ⫹ ventilation preoperatively CoA ⫹ TGV ⫹ VSD

6 Male

8

2

IAA ⫹ truncus arteriosus

Hypothermia

7 Female

6

2.7

IAA ⫹ subaortic stenosis ⫹ bicuspid aortic valve

Normothermia

8 Male

11

2

IAA ⫹ Taussig-Bing malformation

Normothermia

9 Male 10 Male

16 10

2.6 3.4

CoA ⫹ VSD IAA ⫹ subaortic stenosis

Hypothermia Normothermia

Enterocolitis LCO, Ross-Kono procedure, Coronary compression by RV/ PA conduit, refection and ECMO

Major LV dysfunction (3) Heart failure (0) Heart failure (1)

LV failure (51)

CA during PM implantation by thoracotomy (30) Septicemia (55) Major LV dysfunction Bleeding (4)

AMI ⫽ acute myocardial infarction; ARDS ⫽ acute respiratory distress syndrome; AV block ⫽ atrioventricular block; CA ⫽ circulatory arrest; CoA ⫽ coronary artery disease; ECMO ⫽ extracorporeal membrane oxygenation; IAA ⫽ interrupted aortic arch dissection; LCO ⫽ low cardiac output; LV ⫽ left ventricular; PA ⫽ pulmonary artery; PM ⫽ pacemaker; RV ⫽ right ventricular; TGV ⫽ transposition of the great vessels; VSD ⫽ ventricular septal defect.

imaging (MRI) [15, 16]. Absence of detectable preoperative or postoperative neurologic symptoms in our series is probably due to the fact that we did not evaluate our patients with MRI. One of the major limitations of ASCP management is that the safe reduced CPB flow remains poorly defined. Some authors suggest that the higher cerebral perfusion flow rates maintain cortical oxygen tension [17] and that a normothermic procedure is safe for aortic arch surgery and may ensure a more physiologic autoregulation of cerebral blood flow [10]. In our study the cerebral blood flow was higher in the normothermic group (52.7 ⫾ 27.9 versus 48.7 ⫾ 23.7 in the hypothermic group) but the difference was not significant. Andropoulos and colleagues [15] found a 36% incidence of new brain injury with mean ASCP flow rate equal to 57 mL/kg/min. The ASCP flow rate in Dent and associates’ [16] study was less than half in 73% of patients who had new lesions. That is why more sophisticated cerebral monitoring methods need to be developed to allow select bypass strategies that carry safety conditions providing acceptable flow rates and temperature. Electroencephalographic evaluation was used in all patients. However limitations of this technique were reported by Lozano and Mossad [18]. Electroencephalography requires trained personnel; and artifacts and interference from leakage currents, electrocautery, roller pumps, and cardiac pacers preclude adequate recording. Electroencephalographic slowing is not specific for ischemia or hypoxia and does not differentiate among the effects of anesthetic agents, hypothermia, hypoglycemia, hemodilution, Pco2 changes, and extreme hypotension. NIRS tends to be used most frequently in our practice (the last 30 patients). It allowed us to verify the quality of total brain perfusion during ASCP. This technique is a noninvasive means of continuous monitoring of regional cerebral oxygen saturation and perfusion. It may be more sensitive in detecting cerebral perfusion in low-flow states [18, 19]. Dent and colleagues [16] demonstrated that prolonged low postoperative cerebral oximetry readings (⬍45% for 180 minutes) were associated with the development of ischemic lesions detected by MRI. Visceral organs and spinal cord were perfused to some extent through collaterals and protected from ischemia during SCP. It is well known that blood actually returns to the descending aorta through collaterals during ASCP [20]. This collateral flow provides important protection for the visceral organs and the spinal cord from ischemia when there is no active perfusion to the lower body [21]. The most common malformations concerning the aortic arch are usually accompanied by the formation of collaterals. Postoperative renal and hepatic function tests showed a significant difference when compared with the preoperative values but were reported to be within normal limits for routine cardiac operations [22]. The CA time of the lower body organs corresponds to the duration of ASCP and was shorter in our normothermic group (26.3 ⫾ 9.1 minutes). Lactate is a marker of an anaerobic metabolic state with a clear response to hypoperfusion. Our study demonstrated that until the end of CPB the

LY ET AL CEREBRAL PERFUSION HYPOTHERMIA VS NORMOTHERMIA

947

serum lactate level did not show a highly important rise in the 2 groups. Delayed sternal closure rate was high in our series (71%). It’ may be the 1 reason for prolonged inotropic support, time to extubation, and ICU stay. The univariate analysis of adverse events did not conclude significantly that perioperative morbidity and mortality was reduced in patients operated on under normothermia. In examining risk factors influencing early mortality, the only variables that reached statistical significance were CPB time between the 2 groups. There are some limitations of the present study. It is nonrandomized and therefore subject to all the inherent limitations of retrospective studies. The late neurologic status was assessed by routine clinical means only. Evaluation of late intellectual, cognitive, and psychometric development is necessary. In light of this information, the aortic arch repair in normothermia is feasible and safe but should be validated by prospective studies.

References 1. Durandy Y, Hulin S. Intermittent warm blood cardioplegia in the surgical treatment of congenital heart disease: clinical experience with 1400 cases. J Thorac Cardiovasc Surg 2007; 133:241– 6. 2. Caputo M, Bays S, Rogers CA, et al. Randomized comparison between normothermic and hypothermic cardiopulmonary bypass in pediatric open-heart surgery. Ann Thorac Surg 2005;80:982– 8. 3. Pouard P, Mauriat P, Ek F, et al. Normothermic cardiopulmonary bypass and myocardial protection for neonatal arterial switch operation, Eur J Cardiothorac Surg 2006;30: 695–9. 4. Amir G, Ramamoorthy C, Riemer RK, Reddy VM, Hanley FL. Neonatal brain protection and deep hypothermic circulatory arrest: pathophysiology of ischemic neuronal injury and protective strategies. Ann Thorac Surg 2005;80:1955– 64. 5. Wypij D, Newburger JW, Rappaport LA, et al. The effect of duration of deep hypothermic circulatory arrest in infant heart surgery on late neurodevelopment: the Boston Circulatory Arrest Trial. J Thorac Cardiovasc Surg 2003;126:1397– 1403. 6. Asou T, Kado H, Imoto Y, et al. Selective cerebral perfusion technique during aortic arch repair in neonates. Ann Thorac Surg 1996;61:1546 – 8. 7. Küçüker SA, Özatik MA, Sarıtas¸ A, Tas¸demir O. Arch repair with unilateral antegrade cerebral perfusion. Eur J Cardiothorac Surg 2005;27:638 – 43. 8. Oppido G, Napoleone CP, Turci S, et al. Moderately hypothermic cardiopulmonary bypass and low-flow antegrade selective cerebral perfusion for neonatal aortic arch surgery. Ann Thorac Surg 2006;82:2233–9. 9. Harrington DK, Fragomeni F, Bonser RS. Cerebral perfusion. Ann Thorac Surg 2007;83:S799 – 804. 10. Touati GD, Marticho M, Farag M, et al. Totally normothermic aortic arch replacement without circulatory arrest. Eur J Cardiothorac Surg 2007;32:263– 8. 11. Kaukuntla H, Harrington D, Bilkoo I, Clutton-Brock T, Jones T, Bonser RS. Temperature monitoring during cardiopulmonary bypass— do we undercool or overheat the brain? Eur J Cardiothorac Surg 2004;26:580 –5. 12. Wong BI, McLean RF, Naylor CD, et al. Central-nervoussystem dysfunction after warm or hypothermic cardiopulmonary bypass. Lancet 1992;339:1383– 4. 13. Rasmussen LS, Sztuk F, Christiansen M, Elliott MJ. Normothermic versus hypothermic cardiopulmonary bypass during repair of congenital heart disease. J Cardiothorac Vasc Anesth 2001;15:563– 6.

PEDIATRIC CARDIAC

Ann Thorac Surg 2011;92:942– 8

948

LY ET AL CEREBRAL PERFUSION HYPOTHERMIA VS NORMOTHERMIA

PEDIATRIC CARDIAC

14. Jacobs MJ, de Mol BA, Veldman DJ. Aortic arch and proximal supraaortic arterial repair under continuous antegrade cerebral perfusion and moderate hypothermia Cardiovasc Surg 2001;9:396 – 402. 15. Andropoulos DB, Hunter JV, Nelson DP, et al. Brain immaturity is associated with brain injury before and after neonatal cardiac surgery with high-flow bypass and cerebral oxygenation monitoring. J Thorac Cardiovasc Surg 2010;139:543–56. 16. Dent CL, Spaeth JP, Jones BV, et al. Brain magnetic resonance imaging abnormalities after the Norwood procedure using regional cerebral perfusion. J Thorac Cardiovasc Surg 2006;131:190 –7. 17. DeCampli WM, Schears G, Myung R, et al. Tissue oxygen tension during regional low-flow perfusion in neonates. J Thorac Cardiovasc Surg 2003;125:472– 80. 18. Lozano S, Mossad E. Cerebral function monitors during pediatric cardiac surgery: can they make a difference? J Cardiothorac Vasc Anesth 2004;18:645–56.

Ann Thorac Surg 2011;92:942– 8

19. Pouard P, Collange V. Neuromonitoring par la spectroscopie dans le proche infrarouge en chirurgie cardiaque pédiatrique: Neuromonitoring by near infrared spectroscopy in paediatric cardiac surgery. IRBM 2007;28(Suppl 1):9 –14. 20. Miyamoto Y, Fukui S, Kajiyama T, et al. Analysis of collateral blood flow to the lower body during selective cerebral perfusion: is three-vessel perfusion better than two-vessel perfusion? Eur J Cardiothorac Surg 2009;35:684 – 8. 21. Roerick O, Seitz T, Mauser-Weber P, Palmaers T, Weyand M, Cesnjevar R. Low-flow perfusion via the innominate artery during aortic arch operations provides only limited somatic circulatory support. Eur J Cardiothorac Surg 2006; 29:514 –24. 22. Kucuker SA, Ozatik MA, and A. Saritas A, Tas¸demir O. Arch repair with unilateral antegrade cerebral perfusion. Eur J Cardiothorac Surg 2005;27:638 – 43.

DISCUSSION DR WILLIAM DECAMPLI (Orlando, FL): Dr That was a very nice presentation. I have just one comment and would appreciate your response. You show that the outcomes of performing selective cerebral perfusion under conditions of near normothermia were not significantly different from the group in which you used at least moderate hypothermia (and I assume on some of them your temperatures were even considerably less than 28 degrees, because that’s how your abstract reads). Given that result, one might infer that the method to choose is the one that provides the most margin of safety against rare but catastrophic technical events such as those that disrupt perfusion for a period of time. Is this a reasonable inference? I suppose the question is whether the shorter cooling and warming times with near “normothermia” are worth the tradeoff of safety. DR LY: Since 2002 in our center we practice regularly CBP with normothermia and we have not identified specific problems related to this method. In this study it was shown that this method is feasible. However it is well known that surgery of the aortic arch is frequently performed with cerebral perfusion under hypothermia or under deep hypothermic circulatory arrest. We need more follow-up time and we opted for a prospective study with complete study of the regional brain perfusion with MRI, NIRS, and a specific biological assay.

DR MARSHALL L. JACOBS (Cleveland, OH): Very nice presentation. In your method slide, I believe that in describing the process of cerebral perfusion with near normothermia your target was an arterial blood pressure of 40 to 50 mm Hg; is that correct? DR LY: Yes, we maintained our arterial blood pressure around 40 to 50 mm Hg. DR JACOBS: And was that the same target when you used hypothermia? And what were the relative flow levels that were usually required to achieve that level of pressure in the two different temperature groups? DR LY: We haven’t shown a difference between the two groups concerning the rate of flow perfusion. It was a retrospective study, which therefore had no particular protocol. During the selective cerebral perfusion, the rate flow is always reduced to one third of the full flow and this was valid in both techniques. Therefore it is necessary to conduct a more detailed prospective study taking into account consequences and possible neurologic complications. In conclusion I want to say that the use of the normothermic cardiopulmonary bypass and selective cerebral perfusion in aortic arch surgery is safe and it’s feasible. Evaluation of late intellectual, cognitive, and psychometric development is necessary.