Ann Thorac Surg 2008;86:1613-1619. doi:10.1016/j.athoracsur.2008.07.031
© 2008 The Society of Thoracic Surgeons
Original Articles: Pediatric Cardiac
Comparison of Two Cardioplegia Solutions Using Thermodilution Cardiac Output in Neonates and Infants
Pranava Sinha, MD,
David Zurakowski, PhD,
Richard A. Jonas, MD*
Department of Cardiovascular Surgery, Children's National Medical Center, Washington, DC
Accepted for publication July 11, 2008.
Abbreviations and Acronyms CAVC = complete atrioventricular canal; CP = cardioplegia; CPB = cardiopulmonary bypass; ICU = intensive care unit; IVS = intact ventricular septum; TA = truncus arteriosus; TGA = transposition of great arteries; TOF = tetralogy of Fallot; VSD = ventricular septal defect
* Address correspondence to Dr Jonas, Cardiovascular Surgery, Children's National Medical Center, 111 Michigan Ave NW, Washington, DC 20010 (Email: rjonas{at}cnmc.org).
 |
Abstract
|
|---|
Background: Information from clinical studies is limited regarding the optimal method for myocardial protection in immature hearts, and specifically the benefits of different cardioplegia (CP) formulations. We compared two CP techniques by evaluation of the cardiac index using thermodilution catheters.
Methods: The study cohort includes 102 neonates and infants (aged 2 to 269 days) undergoing biventricular repair surgery. A total of 52 patients were managed with commercially available crystalloid CP ("standard CP") and 50 had a custom mix of crystalloid CP with dilute blood ("custom CP"). Repeated-measures analysis of variance was applied to compare the cardiac index every 3 hours during the 24-hour postoperative period between the two groups and stratified by diagnosis. Adjustment for possible confounders was used to more objectively compare the groups.
Results: Standard crystalloid CP provided superior myocardial protection in patients who had transposition of great arteries (p < 0.001), and this advantage held after adjusting for age and cross-clamp time. Shorter ventilatory times and intensive care unit stays were also noted for the standard CP group (p = 0.01). Cardiac index after cardiopulmonary bypass was lower in patients who had transposition of great arteries and intact ventricular septum compared with the group who had transposition of great arteries and ventricular septal defect; and in both subgroups, the standard CP technique was superior to the custom CP solution. Age at operation was inversely correlated with the cardiac index.
Conclusions: Younger patients, particularly neonates, have a significantly higher postoperative cardiac index with standard CP than with custom CP. The advantage is not apparent beyond the neonatal period.
|
Introduction
|
|---|
The pediatric and adult myocardium have fundamental differences regarding substrate metabolism, glycogen content, insulin sensitivity, calcium handling, and enzyme activities that may account for the differences noted in ischemia tolerance between the adult and pediatric heart [1]. In general, immature myocardium has been shown to have greater tolerance to ischemia compared with the adult myocardium [2, 3], and laboratory studies on pediatric myocardial protection have produced conflicting results possibly because of species differences and differing endpoints [4, 5].
There have been remarkably few studies of myocardial protection methods in infants or children that have used clinical outcomes [6–8], and none in neonates that have used cardiac output as an endpoint. Clinical studies of myocardial protection in cyanotic infants using metabolic derangements as their endpoint suggest superior results with blood versus crystalloid cardioplegia (CP) [9–11] and equivocal results in acyanotic infants [10, 11]. Other clinical studies using ultrastructural and metabolic endpoints have shown differences in tolerance of ischemia in infants less than 3 months old compared with older infants [12], suggesting that neonates may need to be treated differently from infants.
The aim of the current study was to compare unique cardiac index data determined by a thermodilution technique between cold standard crystalloid CP and cold custom CP (4:1 dilution of custom crystalloid solution and blood) techniques in neonates and infants.
|
Patients and Methods
|
|---|
The study included a total of 272 patients who were enrolled in the Combined Boston Hematocrit Trials. The inclusion and exclusion criteria have been presented in previous publications [13, 14]. Of these, 102 patients had cardiac index measurements by thermodilution technique and formed the study group. Informed consent was obtained as per institutional guidelines.
All patients underwent biventricular repair under hypothermic cardiopulmonary bypass. Full-flow cardiopulmonary bypass (CPB [approximately 2.5 L · min–1 · m–2]) was used during cooling and rewarming. Some patients had periods of deep hypothermic circulatory arrest, and most had at least one period of reduced-flow CPB, for example, at approximately 0.75 L · min–1 · m–2 when at deep hypothermia (rectal temperature < 18°C). The pH-stat strategy was used during core cooling, low-flow hypothermic perfusion, and rewarming up to 30°C. Other aspects of anesthetic management and perfusion methods were identical in all patients. We used conventional ultrafiltration during CPB, but not modified ultrafiltration after CPB.
Myocardial protection was accomplished with either cold standard crystalloid CP or cold custom CP solutions. Choice of standard versus custom CP was at the discretion of the surgeon. The standard CP group received oxygenated crystalloid CP (Plegisol; Abbott Laboratories, North Chicago, Illinois), whereas the custom CP group received a custom mix of four parts of crystalloid (Baxter Compass, Edison, New Jersey) to one part blood. Details of CP composition and delivery are provided in the Appendix.
Cardiac output and index was determined using pediatric thermodilution catheters (Edwards Lifesciences, Irvine, California) placed in the pulmonary artery, at 3-hour intervals, for the first 24 hours after surgery. Two measurements were made at each time point through injection of 1 mL ice-cold saline in the right atrium, with a third being made only if the difference between the first two was greater than 10%. Postoperative clinical data including incidence of complications, level of inotropic support, duration of intubation, intensive care unit (ICU) stay, and duration of hospitalization were collected prospectively in all patients. Serum lactate levels were measured 60 minutes after CPB, as well as 3, 9, 12, 18, and 24 hours postoperatively.
Patients were categorized into three major subgroups according to their anatomical diagnosis: transposition of great arteries (TGA), tetralogy of Fallot or truncus arteriosus (TOF/TA), and ventricular septal defect or complete atrioventricular canal (VSD/CAVC) groups. The outcome differences in the TGA subgroup led us to analyze the TGA with intact ventricular septum (TGA/IVS) and the TGA with VSD (TGA/VSD) patients separately.
Statistical Analysis
Standard CP and custom CP groups were compared with respect to preoperative, perioperative and hemodynamic variables by the Student t test for continuous variables and the 
2 test for categorical variables including hematocrit level and diagnosis. Results of Hematocrit Study I [13] and II [14] were pooled as hematocrit levels for each study (ie, 20% and 30% for study I and 25% and 35% for study II) were found to be comparably distributed for the two CP groups. Therefore, standard CP and custom CP groups consisted of 52 and 50 patients, respectively.
Myocardial function, as measured by the cardiac index, was compared between the two CP groups using the repeated-measures, mixed-model analysis of variance (ANOVA) with the group-by-time interaction F test for comparing slopes during the first 24 hours after CPB [15]. Choice of covariance structure was based on the lowest value for the Akaike information criterion, and compound symmetry prevailed as the best-fitting structure for modeling the repeated measurements of the cardiac index over the 24-hour time course [16]. Subgroup analyses were performed to compare the cardiac index during 24 hours after CPB between standard CP and custom CP groups for each of three diagnoses: TGA, TOF/TA, and VSD/CAVC. Further subgroup analysis was performed to compare the cardiac index in the TGA group separated into the TGA/IVS and TGA/VSD groups, adjusted for age and cross-clamp time.
Variables not following a normal Gaussian-shaped distribution such as age, days intubated, and ICU and hospital days were compared with the nonparametric Mann-Whitney U test. Clinical outcomes including cardiac arrest, severe bleeding, reoperation, and death were compared between CP techniques by Fisher's exact test as well as by means of multivariable logistic regression analysis to assess differences in the risk of complications between the two CP groups after adjustment for possible confounders [17]. The nonparametric Spearman rho correlation coefficient was used to measure the association between postoperative cardiac index and lactate levels, as lactate generally was not normally distributed. Analysis of the data was performed using the SPSS package (version 16.0; SPSS, Chicago, Illinois). Power analysis indicated that the sample size of 102 patients with cardiac index measurements (52 standard CP, 50 custom CP) provided 80% power (β = 0.2, 
= 0.05) to detect a mean difference of 25% in the cardiac index during the 24 hours after CPB based on the F test in the repeated-measures ANOVA (version 7.0, nQuery Advisor; Statistical Solutions, Boston, Massachusetts) [18]. All reported p values are two-tailed with an alpha level of 0.05 used as the criterion for statistical significance.
|
Results
|
|---|
Preoperative and perioperative variables were similar for both CP groups with respect to age, weight, distribution of diagnoses, and hematocrit levels for each of the two studies of hematocrit (Table 1). Therefore, Hematocrit Study I and Hematocrit Study II were combined to increase statistical power for outcome comparisons. The CP intake was higher in the standard CP group compared with the custom CP group. As the CP delivery was pressure, and not volume or time, regulated, this could be accounted for by the difference in the viscosity of the two solutions. There was no difference in the cross-clamp times between the groups; however, the bypass times were longer in the custom CP group (p = 0.03). The intubation time and ICU stay were significantly less in the standard CP group (p = 0.01). Comparison of clinical outcomes between the two CP groups indicated no significant difference regarding the risk of any complication (Table 2).
As expected, the TGA group was younger at the time of operation, followed by the TOF group, with the VSD/CAVC group comprising the oldest patients. Comparing age at surgery subdivided by diagnosis revealed that TOF/TA patients were significantly older in the custom CP group (p=0.01) than in the standard CP group; however, there were no age differences in the other two diagnosis subgroups (Table 3).
With respect to hemodynamic variables, no group differences were observed on bypass, except lower tympanic membrane temperature on bypass and 60 minutes after bypass were noted along with a higher pO2 at the onset of low flow bypass in the standard CP group. In addition, hematocrit levels at the onset of low flow bypass and 60 minutes after bypass were slightly higher in patients who received standard CP compared with patients in the custom CP group (Table 4).
Cardiac Output
One hundred two patients (52 standard CP; 50 custom CP) had cardiac output and index measurements. No patient in the study had cardiac output measurements at fewer than four time points. Median number of measurements per patient was six. The total number of cases in each subgroup with the number of cardiac output measurements is detailed in Table 5.
Repeated-measures ANOVA confirmed that the cardiac index was comparable between the two CP groups during the first 24 hours after cardiopulmonary bypass based on all patients, and no significant differences were detected at any time point, although the trend was for the cardiac index to be higher in the standard CP group. The analysis indicated that both groups showed a time effect with a U-shaped pattern (p < 0.001), in which they initially decreased to a minimum between 6 and 12 hours after bypass and then rebounded at 15 hours after bypass, with a mean cardiac index higher at 24 hours compared with baseline. Thus, the nonsignificant group-by-time interaction (p = 0.26) revealed similar changes throughout the 24-hour period for both CP groups.
A two-way analysis of covariance confirmed that age negatively correlated with the cardiac index independent of cross-clamp time (p < 0.01). An ANOVA with Bonferroni group comparisons revealed that the cardiac index was significantly lower for babies with TGA compared with TOF/TA (p = 0.007) and VSD/CAVC (p < 0.001). The cardiac index was lower for babies with TOF/TA than for babies with VSD/CAVC, although nonsignificantly (p = 0.12).
Because the mixed-model ANOVA indicated a significant two-way interaction between diagnosis and CP group (p = 0.01), however, we modeled the cardiac index separately for each of the three diagnoses to test whether standard CP or custom CP demonstrated an advantage in patients with a particular cardiac diagnosis. Cardiac index in children with TGA was significantly higher for the standard CP group (n = 28) compared with the custom CP group (n = 80) at each of the eight time points throughout the 24-hour period after CPB (Fig 1A,
Table 6). The mixed-model ANOVA revealed a significant overall difference between the two CP groups (F = 12.46, p < 0.001), although the interaction of group by time confirmed that the slopes were parallel (p = 0.64). For children with TOF/TA, repeated-measures ANOVA indicated no significant differences in the cardiac index between custom CP (n = 56) and standard CP (n = 35; F = 0.69, p = 0.41) with no significant difference in slopes (p = 0.96; Fig 1B). Patients with VSD/CAVC, while generally having better cardiac function than patients with TGA or TOF/TA, also showed no significant differences between custom CP (n = 53) and standard CP (n = 19; F = 0.28, p = 0.60) and no differences in the slopes across the 24 hours after CPB (p = 0.75). Although the VSD/CAVC patients who received custom CP showed a slightly higher cardiac index during the first 12 hours, this was not statistically significant; and both CP groups demonstrated very comparable cardiac index values from 12 to 24 hours after CPB (Fig 1C).
View larger version (20K):
[in this window]
[in a new window]
|
Fig 1. (A) Cardiac index in transposition of great arteries patients for each of the two cardioplegia (CP) groups during the 24 hours after cardiopumonary bypass. Data represent the mean cardiac index at each 3-hour time point, and error bars denote standard errors. Repeated-measures analysis of variance indicated highly significant group differences at each time point (asterisks indicate p < 0.01; except at 3 hours and 15 hours, p = 0.03). The group-by-time interaction revealed comparable rates of change over the time course (ie, equal slopes). (B) The cardiac index in tetralogy of Fallot and truncus arteriosus malformation patients revealed no group differences, and both CP groups showed similar change in myocardial function over 24 hours. (C) The cardiac index in patients with ventricular septal defect and complete atrioventricular canal showed no group differences, and both CP groups showed similar change in myocardial function over 24 hours. (Open circles = standard cardioplegia; solid circles = custom cardioplegia.)
|
|
View this table:
[in this window]
[in a new window]
|
Table 6 Myocardial Function as Measured by Cardiac Index During 24 Hours After Cardiopulmonary Bypass for All TGA Patients and Two Subgroups
|
|
Owing to the age differences between the three diagnosis subgroups, Spearman rank-correlation performed showed an inverse correlation between age and cross-clamp time (r = –0.61), number of days in the ICU (r = –0.59), and hospital length of stay (r = –0.58), and a positive correlation with the postoperative cardiac index (r = 0.54; all p < 0.01).
Three-way ANOVA showed that the significant difference in the cardiac index between the standard CP and the custom CP groups (p < 0.001) remained for patients with TGA even after controlling for effects of age and cross-clamp time as covariates. The lack of significant differences between the two CP groups in the TOF and VSD/CAVC subgroups remained after adjustment for age and cross-clamp time.
After separating the TGA/IVS from the TGA/VSD group, further analysis revealed a lower cardiac index for the TGA/IVS group compared with the TGA/VSD group, irrespective of the CP strategy. Although not statistically significant (p = 0.22), the results were similar after age adjustment.
Comparing techniques, the cardiac index was significantly higher in the standard CP group compared with the custom CP group for patients with TGA/VSD at all time points (Table 6). The difference was less pronounced in the TGA/IVS group, although it approached significance (p = 0.07). After adjustment for age, differences in the cardiac index between CP techniques achieved significance for the TGA/IVS group (p < 0.05), but was less so for the TGA/VSD group (p = 0.09).
Lactate Levels
Inverse correlations were observed between average cardiac index during the entire 24 hour postoperative period and lactate levels, both at the onset of low flow (Spearman rho = –0.36, p < 0.01) and 60 minutes after bypass (Spearman rho = –0.45, p < 0.01). Cardiac index values at 24 hours negatively correlated with lactate values 60 minutes after bypass (Spearman rho = –0.41, p < 0.01) but were not significantly correlated with lactate levels at the onset of low flow (Spearman rho = –0.18, p = 0.20). Lactate levels 60 minutes after bypass for TGA patients tended to be lower in the standard CP group (p = 0.17).
|
Comment
|
|---|
This study of neonates and infants compares two CP groups using standard clinical outcomes, lactate levels, and direct measurement of postoperative cardiac output and index. In neonates with TGA (one third of the study group), cardiac output was significantly higher with the standard CP. There was a consistent trend for lactate levels to be lower with the standard CP solution, although it did not achieve statistical significance. Moreover, the ventilatory time and ICU stay were significantly shorter in the standard CP group.
As has been seen in two previous clinical trials enrolling neonates and infants [13, 14]0, this study found an initial decrease in cardiac output over the first 6 to 12 hours postoperatively, followed by an increase and rebound to higher than baseline cardiac indexes at 15 to 24 hours. Also consistent with previous studies, the cardiac output in the TGA group was significantly lower than in the other two groups, emphasizing the importance of the myocardial protection strategy selected for this population. Young age was also identified in this study as an independent risk factor for lower cardiac output.
This study reveals a significant advantage of using standard CP in patients with TGA, compared with a custom CP solution. Elimination of the confounding effects of age and cross-clamp times confirms an advantage of using standard CP in this younger group of patients. The subgroup of TGA/VSD had higher cardiac indexes than the TGA/IVS groups, but both groups had higher cardiac indexes with the use of standard CP. Although there was a significant age difference in the TGA/IVS subgroup between the two CP groups, the difference in the cardiac index favoring the standard CP technique in this diagnosis group was even more pronounced after correction for age.
Equivocal or slightly better cardiac indexes were seen in this analysis with the use of custom CP for the TOF/TA and VSD/CAVC groups. These groups are composed of older patients, and the results are in agreement with previous studies supporting the superiority of blood CP in this patient population [7, 9–11].
No previous studies have analyzed clinical outcomes that included direct cardiac index measurements in the neonatal age group. The only previous study in the pediatric population comparing CP techniques by direct cardiac output measurement lacked neonates in the study cohorts [7].
Although our study design can be criticized for being nonrandomized, with CP technique chosen by the surgeon, we tried to compare the groups as objectively and reliably as possible by controlling for potential confounders. Although one of the two CP solutions used in this study is a standard commercial crystalloid CP solution, the other solution studied is a custom crystalloid CP, mixed with blood in a 4:1 ratio (Appendix). At this dilution and with a custom CP base, the custom CP may not be representative of more widely used blood based CP solutions.
In conclusion, use of a standard crystalloid CP in neonates with TGA provides superior myocardial protection over custom CP. With increasing age and with other congenital anomalies, this advantage does not appear to hold.
|
Appendix
|
|---|
Cardioplegia Composition and Delivery
Standard crystalloid cardioplegia (CP) consisted of commercially available Plegisol (Abbott Laboratories, North Chicago, Illinois), mixed with 10 mL 8.4% sodium bicarbonate per liter, to adjust the pH. Each 100 mL of solution contains calcium chloride dihydrate 17.6 mg, magnesium chloride hexahydrate 325.3 mg, potassium chloride 119.3 mg, and sodium chloride 643 mg in water for injection. The electrolyte content per liter (not including ions for pH adjustment) are as follows: calcium 2.4 mEq, magnesium 32 mEq, potassium 16 mEq, sodium 110 mEq, and chloride 160 mEq. The osmolar concentration before sodium bicarbonate addition is 304 mOsmol/L with a pH 3.8 (3.5 to 3.9).
Custom CP was custom made by mixing 1 part blood with 4 parts of crystalloid solution (Baxter Compass, Edison, New Jersey). The crystalloid component consisted 16.3 mL 20% mannitol, 4 mL 50% magnesium chloride, 13 mEq 8.4% sodium bicarbonate, 13 mL 1% lidocaine, and 13 mL potassium Chloride (2 mEq/mL) added to 1 L plasmalyte A (Baxter Healthcare Corporation, Deerfield, Illinois). One liter of plasmalyte A has an ionic concentration of 140 mEq sodium, 5 mEq potassium, 3 mEq magnesium, 98 mEq chloride, 27 mEq acetate, and 23 mEq gluconate with an osmolarity of 294 mOsmol/L.
There was no difference in the delivery between the two CP solutions. Both CP solutions were recirculated through a coil immersed in ice, had similar temperatures (approximately 4° to 6°C), and were infused usually as a single dose of 20 mL/kg (usually over 1 to 2 minutes), with monitoring of the CP line pressure and adjustment of the infusion rate to maintain line pressure between 100 mm Hg and 200 mm Hg.
|
References
|
|---|
- Doenst T, Schlensak C, Beyersdorf F. Cardioplegia in pediatric cardiac surgery: do we believe in magic? Ann Thorac Surg 2003;75:1668-1677.[Abstract/Free Full Text]
- Pridjian AK, Bove EL, Bolling SF, Childs KF, Brosamer KM, Lupinetti FM. Developmental differences in myocardial protection in response to 5'-nucleotidase inhibition J Thorac Cardiovasc Surg 1994;107:520-526.[Abstract/Free Full Text]
- Jonas RA. Myocardial protection for neonates and infants Thorac Cardiovasc Surg 1998;46:288-291.[Medline]
- Baker JE, Boerboom LE, Olinger GN. Is protection of ischemic neonatal myocardium by cardioplegia species dependent? J Thorac Cardiovasc Surg 1990;99:280-287.[Abstract]
- Bolling K, Kronon M, Allen BS, Wang T, Ramon S, Feinberg H. Myocardial protection in normal and hypoxically stressed neonatal hearts: the superiority of blood versus crystalloid cardioplegia J Thorac Cardiovasc Surg 1997;113:994-1005.[Abstract/Free Full Text]
- Young JN, Choy IO, Silva NK, Obayashi DY, Barkan HE. Antegrade cold blood cardioplegia is not demonstrably advantageous over cold crystalloid cardioplegia in surgery for congenital heart disease J Thorac Cardiovasc Surg 1997;114:1002-1009.[Abstract/Free Full Text]
- Åmark K, Berggren H, Björk K, et al. Blood cardioplegia provides superior protection in infant cardiac surgery Ann Thorac Surg 2005;80:989-994.[Abstract/Free Full Text]
- Caputo M, Modi P, Imura H, et al. Cold blood versus cold crystalloid cardioplegia for repair of ventricular septal defects in pediatric heart surgery: a randomized controlled trial Ann Thorac Surg 2002;74:530-535.[Abstract/Free Full Text]
- Åmark K, Berggren H, Björk K, et al. Myocardial metabolism is better preserved after blood cardioplegia in infants Ann Thorac Surg 2006;82:172-178.[Abstract/Free Full Text]
- Modi P, Suleiman SM, Reeves BC, et al. Changes in myocardial free amino acids during pediatric cardiac surgery: a randomized controlled trial of three cardioplegic techniques Eur J Cardiothorac Surg 2006;30:41-48.[Abstract/Free Full Text]
- Modi P, Suleiman SM, Reeves BC, et al. Myocardial metabolic changes during pediatric cardiac surgery: a randomized study of 3 cardioplegic techniques J Thorac Cardiovasc Surg 2004;128:67-75.[Abstract/Free Full Text]
- Sawa K, Matsuda H, Shimazaki Y, et al. Ultrastructural assessment of the infant myocardium receiving crystalloid cardioplegia Circulation 1987;76:V141-V145.[Medline]
- Jonas RA, Wypij D, Roth SJ, et al. The influence of hemodilution on outcome after hypothermic cardiopulmonary bypass: results of a randomized trial in infants J Thorac Cardiovasc Surg 2003;126:1765-1774.[Abstract/Free Full Text]
- Wypij D, Jonas RA, Bellinger DC, et al. The effect of hematocrit during hypothermic cardiopulmonary bypass in infant heart surgery: results from the combined Boston hematocrit trials J Thorac Cardiovasc Surg 2008;135:355-360.[Abstract/Free Full Text]
- Laird NM, Ware JH. Random-effects models for longitudinal data Biometrics 1982;38:963-974.[Medline]
- Akaike H. Likelihood of a model and information criteria J Econometrics 1981;16:3-14.
- Katz MH. Multivariable analysis: a practical guide for clinicians2nd ed.. New York: Cambridge University Press; 2006. pp. 73-95.
- Piantadosi S. Clinical trials: a methodologic perspective2nd ed.. Hoboken, NJ: John Wiley & Sons; 2005. pp. 453-478.
Top
Abstract
Introduction
