Postoperative Course and Hemodynamic Profile After the Arterial Switch Operation in Neonates and Infants
A Comparison of Low-Flow Cardiopulmonary Bypass and Circulatory Arrest
Background The neurological morbidity associated with prolonged periods of circulatory arrest has led some cardiac surgical teams to promote continuous low-flow cardiopulmonary bypass as an alternative strategy. The nonneurological postoperative effects of both techniques have been previously studied only in a limited fashion.
Methods and Results We compared the hemodynamic profile (cardiac index and systemic and pulmonary vascular resistances), intraoperative and postoperative fluid balance, and perioperative course after deep hypothermia and support consisting predominantly of total circulatory arrest or low-flow cardiopulmonary bypass in a randomized, single-center trial. Eligibility criteria included a diagnosis of transposition of the great arteries and a planned arterial switch operation before the age of 3 months. Of the 171 patients, 129 (66 assigned to circulatory arrest and 63 to low-flow bypass) had an intact ventricular septum and 42 (21 assigned to circulatory arrest and 21 to low-flow bypass) had an associated ventricular septal defect. There were 3 (1.8%) hospital deaths. Patients assigned to low-flow bypass had significantly greater weight gain and positive fluid balance compared with patients assigned to circulatory arrest. Despite the increased weight gain in the infants assigned to low-flow bypass, the duration of mechanical ventilation, stay in the intensive care unit, and hospital stay were similar in both groups. Hemodynamic measurements were made in 122 patients. During the first postoperative night, the cardiac index decreased (32.1±15.4%, mean±SD), while pulmonary and systemic vascular resistance increased. The measured cardiac index was <2.0 L · min−1 · m−2 in 23.8% of the patients, with the lowest measurement typically occurring 9 to 12 hours after surgery. Perfusion strategy assignment was not associated with postoperative hemodynamics or other nonneurological postoperative events.
Conclusions After heart surgery in neonates and infants, both low-flow bypass and circulatory arrest perfusion strategies have comparable effects on the nonneurological postoperative course and hemodynamic profile.
The dramatic reduction in surgical mortality associated with repair of congenital heart disease in recent decades has been accompanied by a growing recognition of adverse neurological sequelae in some survivors. In 1989, we undertook a prospective clinical trial to assess the short- and long-term neurological sequelae of infant heart surgery, comparing deep hypothermic circulatory arrest to continuous low-flow cardiopulmonary bypass.1 2
A related aim was to examine the short-term hemodynamic effects of the different perfusion strategies during the perioperative period. Patients undergoing the arterial switch operation for TGA were chosen as study subjects because of the relative homogeneity of patients with TGA (eg, clinical presentation, anatomy, and age at repair), low operative mortality,3 favorable long-term hemodynamic results,4 and the ability to perform surgery with equal facility using either perfusion technique.
This purposes of this report are to examine the postoperative course and short-term hemodynamic profile after the arterial switch operation in a large patient population and to compare the nonneurological perioperative effects of deep hypothermic circulatory arrest versus low-flow cardiopulmonary bypass in neonates and infants.
Eligibility criteria for entry were diagnosis of TGA with either an IVS or VSD, planned arterial switch at <3 months of age, no prior cardiac surgery (eg, coarctation repair), and informed consent from the infant’s parents. Exclusion criteria included birth weight <2.5 kg, recognizable syndrome of congenital anomalies, and associated extracardiac anomalies of more than minor severity (eg, arch hypoplasia or coarctation). The study was approved by the Institutional Human Investigation Committee.
We randomly assigned participating patients to receive either predominantly circulatory arrest or predominantly low-flow (50 cm3 · kg−1 · min−1) bypass, with stratification according to diagnosis (IVS versus VSD) and each individual surgeon. The Coordinating Center developed randomization schemes for each diagnosis-surgeon stratum using a permuted blocks design. Low-flow (rather than full-flow) cardiopulmonary bypass was chosen as the alternative to circulatory arrest to minimize blood return during surgery and to facilitate the repair.
The surgical techniques used in our institution for various aspects of the arterial switch were reported previously.5 The four study surgeons used previously reported2 standardized techniques with respect to the time sequence of the operation, including cannulation, cardioplegia (see below), transection of the great arteries, coronary mobilization, completion of the coronary and aortic anastomoses, closure of atrial and (if present) VSDs, patch closure of the coronary donor sites and completion of the pulmonary anastomosis, and decannulation. Either one or two glutaraldehyde-treated autologous pericardial patches were used for the coronary donor sites at the discretion of the individual surgeon.
Anesthesia and Extracorporeal Perfusion Protocol
Anesthetic management was standardized for all patients and was reported previously.2 Pump prime consisted of various amounts of Normosol-R and whole blood, depending on estimated blood volume, hematocrit, and total priming volume used (Cobe Variable Prime Membrane Oxygenator). No calcium was added to the priming solution. Hypothermic myocardial protection was provided by core cooling at flow rates of 150 to 200 mL · kg−1 · min−1 (2.0 to 3.0 kg) or 100 to 150 mL · kg−1 · min−1(3.0 to 5.0 kg) to rectal and esophageal temperatures of ≤18°C, followed by aortic cross clamping and infusion of oxygenated St Thomas cardioplegia in a single dose of 20 cm3/kg.
Once profound hypothermic temperatures were reached, either deep hypothermic circulatory arrest or continuous low-flow cardiopulmonary bypass was instituted during completion of the aortic and coronary anastomoses. Circulatory arrest was used in all cases for closure of the atrial septal defect and/or VSD. Core rewarming was instituted during completion of the pulmonary anastomosis. Mean perfusion pressures were maintained between 30 and 70 mm Hg during rewarming by use of phentolamine 100 μg/kg or phenylephrine 5 μg/kg, as needed. An additional 25 μg/kg of fentanyl and 100 μg/kg of pancuronium were given to maintain anesthesia. All patients were weaned from cardiopulmonary bypass with at least 5 μg · kg−1 · min−1 dopamine support after the rectal temperature reached 35°C. Lactated ringers, fresh whole blood, blood products, and increased inotropic support were given as necessary to maintain normal filling pressures and a systolic perfusion pressure of at least 60 mm Hg.
Intraoperative Fluid Balance
Intraoperative fluid intake was assessed by totaling the intravenous fluids, cardioplegia volume, and pump volume (total volume added during cardiopulmonary bypass plus reservoir volume at the start of cardiopulmonary bypass minus reservoir volume at the termination of cardiopulmonary bypass). Intraoperative fluid output was the total of urine produced on cardiopulmonary bypass, chest tube drainage before the patient left the operating room, and estimated blood loss (obtained by weighing intraoperative sponges). Intraoperative net fluid balance was the difference between fluid intake and output. In addition, patients were weighed before the induction of anesthesia and again (if hemodynamically stable) just before leaving the operating room.
The anesthetic period was extended through the first postoperative night in all patients by use of a continuous fentanyl infusion, typically 10 μg · kg−1 · h−1, and neuromuscular blockade. Routine continuous postoperative monitoring included the surface ECG, transcutaneous pulse oximetry, pulmonary arterial and right and left atrial pressures (through transthoracic catheters), and systemic arterial pressure. Inotropic, chronotropic, and afterload reducing agents were used as clinically indicated. Volume infusions (usually packed red blood cells or 5% albumin) were given to maintain adequate filling pressures with systolic perfusion pressures of at least 50 mm Hg. Diuretics (usually furosemide 1 to 2 mg/kg per dose, two to four times daily) were begun on the first postoperative morning or earlier if the patient was oliguric (<1 cm3 · kg−1 · h−1). Maintenance doses (5 μg/kg per dose, once or twice daily) of Digoxin (without digitalizing doses) were also instituted on the first postoperative morning.
Continuous fentanyl infusions and neuromuscular blockade were discontinued on the first postoperative morning in the hemodynamically stable patient or continued for longer periods as dictated by the clinical status of the patient. The rate of weaning of mechanical ventilation was determined by the patient’s fluid balance and gas exchange as indicated by arterial blood sampling, pattern of breathing, and daily radiographic findings.
Postoperative data were collected prospectively by the study team from the day of surgery until hospital discharge. Laboratory data included daily measurements of arterial blood gases, serum glucose, calcium, electrolytes, hematologic parameters, BUN, and creatinine. In cases of repeated measurements in the same 24-hour period, the values obtained between 5 and 7 am each postoperative day were used for this analysis. Each day, we recorded significant medical events, the total volume and type of fluid intake, urine output, and chest tube drainage. The total hours of mechanical ventilation and the days in the CICU and the hospital were recorded.
Cardiac output was obtained by thermodilution technique. The pulmonary artery catheter was a 3.5F double lumen (American-Edwards) equipped with a radiopaque thermistor. Triplicate determinations of cardiac output were made over 1 to 2 minutes using 1-cm3 injections into the right atrial line of iced 5% dextrose in water. Measurements were made at 3, 6, 9, 12, 18, and 24 hours after removal of the aortic cross clamp. Mean systemic arterial, pulmonary arterial, and right and left atrial pressures were recorded. Cardiac index was calculated by dividing the average cardiac output at each time period by the body surface area (using the preoperative weight and height and standard nomogram). Systemic and pulmonary vascular resistances were determined (Wood units corrected for body surface area, U · m2) with standard formulas. The doses of inotropic, chronotropic, and afterload reducing agents at the time of the cardiac output measurements were recorded. For between-group comparisons, a total inotrope dose was calculated by adding the doses of dopamine and dobutamine in micrograms per kilogram per minute and assigning an arbitrary equivalent value of 10 μg · kg−1 · min−1 inotrope for each 0.1 μg · kg−1 · min−1 epinephrine.
Hemodynamic data were excluded from later analysis if there was evidence of residual left-to right shunt from physical examination or visual inspection of the thermodilution curve for a recirculation phase or evidence of a residual atrial septal defect or VSD on early postoperative echocardiography (performed at discharge in all patients).6 7 Data were considered for analysis if at least four of the six measurements were obtained during the first 24 postoperative hours.
Treatment groups were compared in intention-to-treat analyses in which the strategy consisting predominantly of circulatory arrest was compared with the strategy consisting of predominantly low-flow bypass. Diagnostic groups were also compared, with the diagnosis of TGA/IVS compared with that of TGA/VSD.
Perioperative outcomes included continuous and categorical variables. Pearson correlation coefficients, t tests, linear regression, and graphical methods were used to analyze continuous outcome variables. When normality assumptions about the data were suspect, Wilcoxon’s rank-sum test was used to compare groups. Fisher’s exact tests and stratified exact tests were used to analyze categorical outcome variables. All probability values are two-tailed.
We enrolled patients from January 1989 through February 1992. Of the 246 patients who had an arterial switch in this time period, 174 infants met study criteria, and 166 (95.4%) were enrolled. The arterial switch was performed in 157 of 166 infants (94.6% of those enrolled); the remaining 9 patients were found at operation to have unsuitable coronary artery anatomy for the arterial switch (these 9 patients had an alternative operation performed and thus were ineligible for continued study). Fourteen additional infants met eligibility criteria, were enrolled and randomized according to the study protocol, and underwent the arterial switch between April 1988 and August 1988 (before the onset of the funded enrollment phase of the trial). Inclusion of these infants was approved by the Safety and Data Monitoring Board of the NIH, yielding a total of 171 study subjects.
Of the 129 infants with TGA/IVS, 66 were randomized to receive predominantly circulatory arrest and 63 to receive predominantly low-flow bypass. Of the 42 infants with TGA/VSD, 21 were randomized to receive predominantly circulatory arrest and 21 to receive predominantly low-flow bypass. Intraoperative data, including perfusion and circulatory arrest times, were reported previously2 and are summarized in Table 1⇓. Total support times (sum of circulatory arrest and total bypass times) and myocardial ischemic times were similar for both circulatory arrest and low-flow bypass groups within each diagnostic category (IVS or VSD), supporting the notion that either perfusion strategy can be used with equal facility during the arterial switch. Patients with an associated VSD had longer periods of both total support and myocardial ischemia compared with patients with TGA/IVS. With adjustment for diagnosis (IVS or VSD), there were no significant differences in other perfusion variables (eg, pH, base deficit, perfusion pressures, and ionized calcium) between treatment groups.
There were 3 early deaths (1.8%), 2 in the early postoperative period and 1 shortly after discharge from the hospital. A 5-day-old child with TGA/VSD had a single coronary artery arising from the right, posterior-facing sinus8 that supplied the entire myocardium. Transfer of this vessel produced kinking and inadequate filling of the left coronary artery. The patient could not be weaned from cardiopulmonary bypass.
Another 5-day-old child with a large, anterior malalignment type of VSD had moderate tricuspid stenosis and right ventricular hypoplasia that was not recognized preoperatively. The postoperative course was marked by a low output state with progressive oliguria, hepatic failure, and anasarca. Cardiac catheterization was undertaken on the seventh postoperative day with the child in grave condition; death occurred during attempted tricuspid valve dilation.
A 4-day-old child with TGA/IVS had an uncomplicated arterial switch and was discharged 8 days after surgery. Frequent emesis with mild tachypnea was noted, and a presumed diagnosis of gastroesophageal reflux was made. An echocardiogram 7 days postoperatively suggested near-systemic right ventricular pressure without pulmonary anastomosis obstruction, possibly caused by pulmonary hypertension. Four days after discharge, the patient died suddenly after feeding, presumably of aspiration. At autopsy, aspiration was confirmed, and narrowing of the right main stem bronchus was noted. The pulmonary arteries were mildly thickened. The great vessel and coronary anastomoses were widely patent. Early postoperative measurements of cardiac output (see below) were not made in these patients.
For the 169 hospital survivors, the median duration of mechanical ventilation was 2.9 days (4.2±5.5 days [mean±SD]; range, 22 hours to 61 days), and the median stays in the intensive care unit and the hospital were 5 and 9 days (mean, 6.7±6.4 and 11±8 days; ranges, 3 to 65 and 5 to 71 days, respectively). Eight patients (4.7%) were mechanically ventilated for >2 weeks.
In 19 patients (11.2%) with severe myocardial edema or poor hemodynamics, the sternotomy was not closed primarily at the time of the arterial switch; in one additional patient, the sternum was reopened in the CICU. The wound was covered with a Silastic sheet, and the sternum was subsequently closed in the intensive care unit (n=19) or operating room (n=1) at a median of 3 days (range, 1 to 5 days) after surgery.
Table 2⇓ summarizes the significant postoperative events. Early postoperative cardiac catheterization was performed in 5 patients (2.9%) for hemodynamic and/or electrophysiological studies (n=3), pulmonary artery dilation (n=1), and neoaortic valve dilation (n=1). Other significant postoperative problems included clinical seizures (n=11, 6.5%), hemidiaphragm paresis (n=5, 2.9%), chylothorax (n=3, 1.8%), suspected necrotizing enterocolitis (n=2, 1.2%), and mediastinitis (n=1, 0.6%).
Other than the incidence of seizures (which was related to the duration of circulatory arrest, older age at surgery, and an associated VSD),2 the only other postoperative complication or event associated with perfusion strategy was acidosis; infants randomized to low-flow bypass were more likely to have had a pH <7.25 than those randomized to circulatory arrest (P=.05). In particular, perfusion strategy did not influence the frequency of delayed sternal closure, duration of mechanical ventilation, or duration of CICU or hospital stay.
Cardiac index was determined in 122 of 170 patients (71.8%) who returned to the CICU from the operating room. In these 122 patients, six measurements were made in 99, five were made in 19, and four were made in 4 patients. Complete data were not obtained in 48 patients because of no pulmonary artery catheter (n=27), dysfunctional pulmonary arterial (n=9) or right atrial (n=3) catheter, residual atrial (n=1) or ventricular (n=5) left-to-right shunt, or “other” (n=3).
The 48 patients without complete (ie, fewer than four measurements) hemodynamic data differed slightly from the 122 operation survivors with complete hemodynamic measurements. These 48 patients had a higher lowest-documented pH preoperatively (7.30 versus 7.26 pH, P=.04), were more likely to have TGA/IVS (P=.001) and to have been randomized to circulatory arrest (P=.04), and had a longer mean duration of circulatory arrest (46±20 versus 32±21 minutes, P<.001). They were also more likely to have a longer myocardial ischemic time (87±18 versus 79±13 minutes, P=.006) and longer total support time (154±37 versus 140±28 minutes, P=.02). Finally, these 48 patients tended to receive slightly more pressor support in the immediate postoperative period (although this did not reach statistical significance). Thus, compared with the 48 patients without complete hemodynamic data, the 122 patients with complete data were slightly “less sick.” However, intraoperative fluid balance, weight gain, and most preoperative (eg, birth weight, Apgar scores, and lowest Po2) and postoperative (eg, duration of mechanical ventilation, inotropic support received, and total CICU and hospital stays) variables were similar between the groups with and without complete hemodynamic measurements.
Fig 1A⇓ shows the mean values for the groups at each time period. When analyzed as a group, cardiac index reached a nadir at 9 to 12 hours postoperatively (P<.001 versus baseline measurement at 3 hours), returning to baseline by 24 hours after surgery. During this period, the mean dose of inotropic support did not change (Fig 1A⇓). When individual patients were reviewed, the measured cardiac index was <2.0 L · min−1 · m−2 at least once in 29 of 122 patients (23.8%). The lowest measured cardiac index was first recorded at 3 hours in 12 patients (9.8%), at 6 hours in 22 patients (18.0%), at 9 hours in 30 patients (24.6%), at 12 hours in 40 patients (32.8%), at 18 hours in 13 patients (10.7%), and at 24 hours in 5 patients (4.1%). For the 113 patients whose lowest measured cardiac index occurred after the baseline measurement at 3 hours, the average fall in cardiac index was 32.1±15.4% during the first postoperative night.
Perfusion strategy (circulatory arrest or low-flow bypass) did not have an impact on postoperative hemodynamics. Patients with TGA/VSD had slightly better cardiac indexes and received slightly less inotropic support during the first postoperative night (Fig 1B⇑) than TGA/IVS patients. Longer total bypass and myocardial ischemic times were associated with higher doses of inotropic support and slightly higher cardiac index measurements in the first postoperative night. Fig 2⇓ shows other hemodynamic parameters.
In addition to dopamine (n=168), some patients received pressors and/or vasodilators during the study period at the discretion of the treating physicians; these included epinephrine (n=14), dobutamine hydrochloride (n=4), isoproterenol (n=11), sodium nitroprusside (n=26), and amrinone lactate (n=20). The use of additional inotropic agents or vasodilators was not related to perfusion strategy.
Weight Gain, Fluid Balance, and Mechanical Ventilation
During surgery, patients randomized to receive predominantly low-flow bypass gained significantly more weight (P=.01) and had more positive fluid balance (P<.001) than those randomized to receive circulatory arrest. When analyzed as a continuous variable, longer bypass time was positively associated with greater weight gain (r=.23, P=.01) and positive fluid balance (r=.46, P<.001; Fig 3⇓).
All patients received diuretics beginning on the first postoperative day. Urine output rose from 3.2±1.5 cm3 · kg−1 · h−1 during the first postoperative night to 7.1±1.8 cm3 · kg−1 · h−1 on the third postoperative day (Fig 4⇓), with no significant difference between randomized groups. However, fluid balance (input minus output) tended to be more negative on the second (P=.17) and third (P=.18) postoperative days in patients assigned to low-flow bypass (Fig 4⇓). As patients lost their total body fluid load, serum levels of BUN rose, while serum creatinine remained stable (Fig 4⇓).
Patients randomized to low-flow bypass were mechanically ventilated for a median of 71 hours (range, 22 hours to 19 days), whereas those randomized to circulatory arrest were ventilated for a median of 68.5 hours (range, 32 hours to 65 days; P=.47). Thus, although patients assigned to low-flow bypass had increased total body water immediately after surgery, the increased net fluid output the first 2 days after surgery resulted in similar net fluid balance, so the duration of mechanical ventilation was similar in both groups.
Sinus node dysfunction, treated with mechanical pacing (through either the esophagus or transthoracic temporary atrial wires), was seen in 23 patients (13.5%). Supraventricular tachycardia was seen in 23 patients (13.5%), was frequently brief and self-limited, but was treated with burst atrial pacing in 3 patients and cardioversion in 1 patient. Three patients (1.8%) had transient complete heart block. Sustained ventricular tachycardia was seen in 3 patients (1.8%; all underwent early postoperative catheterization [see above], which documented normal coronary perfusion). Perfusion strategy was not related to the incidence of postoperative arrhythmia. A more detailed analysis of arrhythmia after the arterial switch operation in a larger cohort of patients is reported elsewhere.9
After surgery, there was an initial elevation in the white blood cell count, percent neutrophils, and percent band forms in both treatment groups, which fell into the normal range by the second postoperative day (Fig 5⇓). The platelet count reached a nadir on the second postoperative morning; although there was a comparable fall in both treatment groups, patients with TGA/VSD had significantly lower platelet counts than those with TGA/IVS (Fig 6⇓). The number of patients with platelet counts <50 000 (Table 2⇑) and the number of patients receiving platelet transfusions (n=36 [21.2%]: low-flow, n=19; circulatory arrest, n=17) were similar in both treatment groups. The duration of cardiopulmonary bypass was not associated with lowest platelet count or the need for platelet transfusion.
Data from this study suggest that both low-flow bypass and circulatory arrest perfusion strategies have comparable effects on the nonneurological postoperative course in neonates and small infants after corrective surgery for TGA. As previously reported in this study population, prolonged periods of circulatory arrest were associated with a higher incidence of clinical and EEG seizures,2 longer recovery time to the first reappearance of EEG activity,2 a greater release of the brain isoenzyme of creatine kinase,2 and a higher prevalence of neurological and cognitive abnormalities at 1 year of age.1
We found that a longer duration of cardiopulmonary bypass (ie, shorter periods of circulatory arrest) was associated with increased weight gain and positive fluid balance immediately after surgery. Despite the increased total body edema, however, a strategy consisting predominantly of circulatory arrest, compared with one consisting predominantly of low-flow cardiopulmonary bypass, had no significant effect on postoperative hemodynamics, laboratory studies, the duration of mechanical ventilation, or the length of stay in the intensive care unit or hospital in this patient population.
The postoperative course in most patients was uneventful and rather predictable. The first postoperative night was typically characterized by a significant fall in cardiac index (similar to findings reported by Fontan et al10 in adults), coupled with a rise in the calculated systemic and pulmonary vascular resistance (Figs 1⇑ and 2⇑). This finding was consistent in both treatment and diagnostic groups. The overall cardiac index was in the low-normal range in our patients (despite receiving inotropic support); values were similar to those in previously reported studies in neonates and infants after other forms of cardiac surgery.11 12 13 14 Most patients received moderate (5 to 10 μg · kg−1 · min−1) inotropic support with dopamine; afterload reduction, when indicated, was typically given as amrinone (if cardiac index was low) or sodium nitroprusside. The amount and duration of inotropic support, ordered by the bedside physicians as determined by clinical examination and postoperative monitoring, remained fairly constant during the first 24 hours after surgery and was not related to treatment assignment. Thus, the fall in cardiac index seen during the first postoperative night was not related to withdrawal or reduction of inotropic medications.
This finding of a significant fall in the measured cardiac index during the first postoperative night, despite otherwise acceptable hemodynamics, is consistent with the observation that blood pressure may be maintained in neonates and young infants despite a falling cardiac output. The physiological mechanism most likely responsible for the maintained blood pressure is an elevation of systemic vascular resistance, which rose an average of 20% in our patients (Fig 2⇑). This is also consistent with the conventional teaching that hypotension is a late finding in neonates with low cardiac output.
The mechanism of the fall in cardiac index with a rise in systemic and pulmonary vascular resistances remains undefined but may be related to ischemia-reperfusion injury.15 16 17 18 19 20 21 22 23 24 Studies with monoclonal antibodies to leukocyte adhesion molecules have suggested a prominent role for inflammation in the response of the heart to ischemia and reperfusion.25 The actual time course of expression of the adhesion molecules in an intact organism is unknown, but in vitro studies showed that expression occurs hours after the inciting stimulus; we recently described how the mRNA of two endothelial adhesion molecules (e-selectin and intercellular adhesion molecule–1) is induced within hours of the institution of cardiopulmonary bypass.26 It is therefore tempting to speculate that this type of inflammation-mediated event is associated with the fall in cardiac index observed in this group of patients, but additional experimental work is clearly necessary to confirm this hypothesis.
We found a significant relation between the duration of cardiopulmonary bypass and edema (weight gain and positive fluid balance) in our patients. Although the observed weight gain and positive fluid balance seen in these patients are also potentially related to inflammation, this clinical finding is most likely multifactorial. It is a common clinical observation that young infants, particularly neonates, are prone to significant edema formation during cardiopulmonary bypass. The permeability of capillaries is known to be greater in immature than in mature individuals,27 28 and this microvascular permeability is enhanced by cardiopulmonary bypass.29 In addition, cardiopulmonary bypass is known to activate multiple potent vasoactive substances, including the serum anaphylatoxins C3a and C5a,30 tumor necrosis factor,16 31 and oxygen-derived free radicals,32 33 and stimulate the release of lysosomal hydrolases.33 34 All these factors combine to cause the nearly 30% increase in body weight in these neonates and infants. The weight gain was sufficient in 20 of our patients to warrant delayed sternal closure35 in the CICU after diuresis and loss of myocardial and chest wall edema.
The next 2 to 4 days after surgery were typically characterized by significant diuresis (augmented by diuretics in all cases), negative fluid balance, weaning from mechanical ventilation, and extubation. Patients randomized to low-flow bypass were typically more edematous after surgery but reduced the fluid load sufficiently through diuresis to achieve extubation at a time similar to that of the patients randomized to circulatory arrest. In our study conditions (hypothermia to 12°C to 18°C and relatively short periods of renal ischemia),36 37 there did not appear to be any significantly different effect on renal function caused by the different perfusion strategies.
The effects of cardiopulmonary bypass and hypothermia on platelet number38 39 and function40 41 42 43 were described previously in adult patients and laboratory studies. In our patients, the significant fall in the platelet count seen on the second postoperative morning is consistent with the hypothesis that platelets may be partially degranulated, misshapen, and aggregated after exposure to the prosthetic material of cardiopulmonary bypass. Radionuclide studies44 suggested that these dysfunctional platelets are sequestered primarily in the liver during the first 48 hours after cardiopulmonary bypass and that the mean survival time of these damaged, but still circulating, platelets was 58±8 hours.
Despite the prospective nature of this randomized trial, the hemodynamic profile measured in these infants does not necessarily reflect the natural history of myocardial function after cardiopulmonary bypass, aortic cross clamping, and deep hypothermic circulatory arrest. Although the mean values for the entire group fell consistently during the first postoperative night, there was considerable patient-to-patient variation, so the nadir of cardiac index for an individual patient may have been at any point during the measurement sequence. In addition, the clinicians caring for the patients were aware of the clinical status and the measured hemodynamic profile and were likely to respond to a falling cardiac output by adjusting support. This increased support in an individual patient might then be reflected in higher measurements of cardiac output later in the study period, which might confound the analysis of the effects of prolonged myocardial ischemic time or cardiopulmonary bypass time on hemodynamic measurements. Although we did not find a significant relation between the duration of aortic cross clamping or cardiopulmonary bypass time with postoperative hemodynamics, the intervening medical management may have masked a true relation. There also may have been other similar effects of prolonged cardiopulmonary bypass or myocardial ischemia on other postoperative issues, such as thrombocytopenia, that were not identified with this study design.
Although the data presented here represent a typical cross section of patients undergoing the arterial switch operation in our institution, the results may not be broadly generalizable to neonatal and early infant repair of other forms of congenital heart disease or to the effects of deep hypothermic circulatory arrest and low-flow cardiopulmonary bypass in older infants, children, and adults. The arterial switch is predominantly an extracardiac operation, so either circulatory arrest or low-flow bypass strategies can be used with equal facility. However, the use of low-flow bypass in other types of neonatal and infant surgeries may be associated with a longer total support time than that achievable with circulatory arrest, which may result in an increase in total body fluid overload and expose the patient to pump-related sources of brain injury.
Cardiovascular surgical teams will thus need to balance on an individual basis the technical advantages of circulatory arrest in facilitating a complete repair against the potential neurological risks of prolonged circulatory arrest. Postoperative nonneurological results are likely to be similar with either perfusion strategy but must be studied prospectively.
Selected Abbreviations and Acronyms
|BUN||=||blood urea nitrogen|
|CICU||=||cardiac intensive care unit|
|IVS||=||intact ventricular septum|
|TGA||=||transposition of the great arteries|
|VSD||=||ventricular septal defect|
This work was supported by grant HL-41786 from the NHLBI, NIH, Bethesda, Md. We would like to thank the medical and nursing staffs of the Cardiac Intensive Care Unit at Children’s Hospital for their help in the care of these patients and compliance with the data collection and study protocol; the members of the Safety and Data Monitoring Committee (Julien I.E. Hoffman, MD, chairman) appointed by the NIH; the perfusionists, Willis G. Gieser, CCP, Robert A. LaPierre, BS, CCP, Robert J. Howe, BS, CCP, David M. Farrell, MA, CCP, and Bettina Archilla, BS, CCP; Kristin C. Lucius, RN, for help with data acquisition; Ludmila Kyn for database and statistical programming; Donna M. Donati, Donna M. Duva, and Lisa-Jean Buckley for data management; Kathleen M. O’Brien for project coordination; and Matthew Martin and Tannis Bolton for help with manuscript preparation.
Reprint requests to David L. Wessel, MD, Director, Cardiac Intensive Care Unit, Farley 653, Children’s Hospital, 300 Longwood Ave, Boston, MA 02115.
Presented in part at the 65th Scientific Sessions of the American Heart Association, November 16-19, 1992, New Orleans, La.
- Received December 27, 1994.
- Revision received March 29, 1995.
- Accepted May 10, 1995.
- Copyright © 1995 by American Heart Association
Bellinger DC, Jonas RA, Rappaport LA, Wypij D, Wernovsky G, Kuban KCK, Barnes PD, Holmes GL, Hickey PR, Strand RD, Walsh AZ, Helmers SL, Constantinou JE, Carrazana EJ, Mayer JE Jr, Hanley FL, Castañeda AR, Ware JH, Newburger JW. A comparison of the developmental and neurologic status at one year of children who underwent heart surgery using hypothermic circulatory arrest or low-flow cardiopulmonary bypass. N Engl J Med. 1995;332:549-555.
Newburger JW, Jonas RA, Wernovsky G, Wypij D, Hickey PR, Kuban KCK, Farrell DM, Holmes GL, Helmers SL, Constantinou J, Carrazana E, Barlow JK, Walsh AZ, Lucius KPC, Share JC, Wessel DL, Hanley FL, Mayer JE Jr, Castañeda AR, Ware JH. A comparison of the perioperative neurologic effects of hypothermic circulatory arrest versus low-flow cardiopulmonary bypass in infant heart surgery. N Engl J Med. 1993;329:1057-1064.
D-transposition of the great arteries. In: Castañeda AR, Jonas RA, Mayer JE Jr, Hanley FL, eds. Cardiac Surgery of the Neonate and Infant. Philadelphia, Pa: WB Saunders Co; 1994:409-438.
Colan SD, Trowitzsch E, Wernovsky G, Sholler GF, Sanders SP, Castañeda AR. Myocardial performance after arterial switch operation for transposition of the great arteries with intact ventricular septum. Circulation. 1988;78:132-141.
Wernovsky G, Hougen TJ, Walsh EP, Sholler GF, Colan SD, Sanders SP, Parness IA, Keane JF, Mayer JE Jr, Jonas RA, Castañeda AR, Lang P. Midterm results after the arterial switch operation for transposition of the great arteries with intact ventricular septum: clinical, hemodynamic, echocardiographic, and electrophysiologic data. Circulation. 1988;77:1333-1344.
Fontan F, Madonna F, Naftel DC, Kirklin JW, Blackstone EH, Digerness S. Modifying myocardial management in cardiac surgery: a randomized trial. Eur J Cardiothorac Surg. 1992;6:127-137.
Parr GVS, Blackstone EH, Kirklin JW. Cardiac performance and mortality early after intracardiac surgery in infants and young children. Circulation. 1975;51:867-874.
Parr GVS, Blackstone EH, Kirklin JW, Pacifico AD, Lauridsen P. Cardiac performance early after interatrial transposition of venous return in infants and small children. Circulation. 1974:49;50(suppl II):II-2-II-8.
Braunwald E, Kloner RA. Myocardial reperfusion: a double-edged sword? J Clin Invest. 1985;76:1713-1719.
Brett J, Gerlach H, Nawroth P, Steinberg S, Godman G, Stern D. Tumor necrosis factor/cachectin increases permeability of endothelial cell monolayers by a mechanism involving regulatory G proteins. J Exp Med. 1989;169:1977-1991.
Engler RL, Dahlgren MD, Morris DD, Peterson MA, Schmid-Schonbein GW. Role of leukocytes in response to acute myocardial ischemia and reflow in dogs. Am J Physiol. 1986;251:H314-H323.
Kawata H, Aoki M, Mayer JE Jr. Nitroglycerin improves functional recovery of neonatal lamb hearts after 2 hours of cold ischemia. Circulation. 1993;88(suppl II):II-366-II-371.
Romson JL, Hook BG, Kunkel SL, Abrams GD, Schork MA, Lucchesi BR. Reduction of the extent of ischemic myocardial injury by neutrophil depletion in the dog. Circulation. 1983;67:1016-1023.
Wilson I, Gillinov AM, Curtis WE, DiNatale J, Burch RM, Gardner TJ, Cameron DE. Inhibition of neutrophil adherence improves postischemic ventricular performance of the neonatal heart. Circulation. 1993;88(suppl II):II-372-II-379.
Yamazaki T, Seko Y, Tamatani T, Miyasaka M, Yagita H, Okumura K, Nagai R, Yazaki Y. Expression of intercellular adhesion molecule-1 in rat heart with ischemia/reperfusion and limitation of infarct size by treatment with antibodies against cell adhesion molecules. Am J Pathol. 1993;143:410-418.
Rosenthal SM, LaJohn LA. Effect of age on transvascular fluid movement. Am J Physiol. 1975;228:134-140.
England MD, Cavarocchi NC, O’Brien JF, Solis E, Pluth JR, Orszulak TA, Kaye MP, Schaff HV. Influence of antioxidants (mannitol and allopurinol) on oxygen free radical generation during and after cardiopulmonary bypass. Circulation. 1986;74(suppl III):III-134-III-137.
Harker LA, Malpass TW, Branson HE, Hessel EA, Slichter SJ. Mechanism of abnormal bleeding in patients undergoing cardiopulmonary bypass: acquired transient platelet dysfunction associated with selective a-granule release. Blood. 1980;56:824-834.