Effects of Sevoflurane Inhalation During Cardiopulmonary Bypass on Pediatric Patients

A Randomized Controlled Clinical Trial

Abstract

The effects of sevoflurane inhalation during cardiopulmonary bypass (CPB) on postoperative courses and serum cardiac troponin I (cTnI) concentrations in pediatric patients undergoing cardiac surgery have not been extensively investigated. In this single-center, prospective, randomized trial, an anesthetic regimen containing 2% sevoflurane used throughout the CPB process was compared with a total intravenous anesthesia (TIVA) regimen. One hundred and three patients undergoing congenital heart defect repair with CPB were included in this prospective randomized controlled study. They were randomized into two groups: the sevoflurane group, who received 2% sevoflurane during CPB via an oxygenator, and the control group, who received only an oxygen–air mixture. The pre- and intra-operative parameters were comparable between the two groups. There was a slight but significant increase of arterial diastolic pressure in the sevoflurane group immediately after CPB compared with control patients (46.9 ± 9.3 mm Hg vs. 43.6 ± 8.9 mm Hg; p = 0.033). There was no death in either group. The postoperative ventilation time (in mean [95% confidence interval]) was shorter in the sevoflurane group than that in the control group (26.1 [19.2, 33.0] h vs. 37.7 [24.4, 50.9] h; p = 0.014). The postoperative ICU time, hospital days, and serial serum cTnI concentrations were not significantly different between the two groups. Inhalation of 2% sevoflurane during CPB is beneficial to the recovery of pediatric patients undergoing cardiac surgery but has no significant effect on postoperative cTnI release.

A recent international consensus conference identified only 12 drugs, techniques, or strategies associated with a reduction in perioperative mortality, and the only anesthetic drugs included in this short list were volatile agents.¹ Volatile agents have documented pharmacological but nonanesthetic properties that confer cardiac protection and influence perioperative^2–4 and long-term clinically relevant outcomes.^5,6 Five studies suggested that the beneficial effect of volatile agents (desflurane, isoflurane, and sevoflurane) might translate into reduced mortality rates when compared with total intravenous anesthesia (TIVA) in cardiac surgery.^2–4,6,7 A recently published network meta-analysis conducted by Landoni et al.⁸ implied that anesthesia with volatile agents appears to reduce mortality after cardiac surgery when compared with TIVA, especially when sevoflurane or desflurane are used. However, all 3,996 patients enrolled in the 38 trials were adults.

Less information has been obtained regarding the benefits of volatile anesthetics in pediatric patients undergoing congenital heart defect repair. Thus far, only a few randomized clinical trials have reported the effects of volatile anesthetics on clinical outcomes and/or perioperative biomarkers. However, their results were inconsistent. In this single-center, prospective, randomized trial, an anesthetic regimen containing 2% sevoflurane used throughout the cardiopulmonary bypass (CPB) process was compared with a TIVA regimen in pediatric patients undergoing congenital heart defect repair. Our hypothesis was that sevoflurane via an oxygenator might provide some benefits regarding the clinical outcome and regarding myocardial protection with serial plasma troponin I (cardiac troponin I [cTnI]) content as the marker.

Materials and Methods

The study protocol was approved by the ethics committee of Xijing Hospital (Approval Number: XJ20110911) and registered in the ClinicalTrials.gov Protocol Registration System (NCT01450956, release date Oct 7th, 2011); the principle investigator was Dr. Zhen-xiao Jin. Informed consent was obtained from the parents before enrollment in the study. This randomized controlled trial was reported in accordance with the Consolidated Standards of Reporting Trials (CONSORT) checklist (version 2001). From December 1, 2011 to March 8, 2012, consecutive pediatric patients with body weights of less than 10 kg who were scheduled for congenital heart defect surgical repair in our hospital were eligible for recruitment. Exclusion criteria were redo-surgery, preoperative instability, and emergency surgery. Patients with isolated atrial septal defect (ASD), isolated pulmonary stenosis, or bidirectional cavopulmonary shunt undergoing Fontan completion were also excluded because bypass times are relatively short and the duration of stay in the intensive care unit (ICU) is often less than 24 h. Patients with chromosomal defects, airway and parenchymal lung disease, and immunodeficiency or blood disorders were also excluded. The enrolled patients were randomly allotted to the sevoflurane group or the control group using computer-generated random numbers. The randomization was achieved by messages in sealed envelopes given to the perfusionist and the anesthesiologist just before the operation. The staffs involved in clinical care, including surgeons, intensive care physicians, and members of the study group obtaining functional data, were blinded to the randomization for the period of data acquisition and analysis. Group allocation was not revealed until the final statistical analysis was completed. The trial profile is shown in Figure 1.

Figure 1: Trial profile. CPB, cardiopulmonary bypass.

Anesthesia

Patients received premedication consisting of oral atropine (0.02 mg·kg⁻¹) and midazolam (0.5 mg·kg⁻¹) 30 min before induction of anesthesia, which was induced with a bolus of sufentanil (1 mg·kg⁻¹) and pancuronium (0.2 mg·kg⁻¹). For maintenance of anesthesia during CPB, patients received a continuous infusion of midazolam (0.2 mg·kg⁻¹·h⁻¹, for the control group) or an end-tidal concentration of sevoflurane (Abbott Laboratories, Abbott Park, IL) at 2% (for the sevoflurane group). Sevoflurane inhalation was achieved via an Oxygenator (CAPIOX RX05 Baby RX, Terumo, Japan) during the period of CPB. Each patient received a continuous infusion of sufentanil (2 mg·kg⁻¹·h⁻¹) throughout the operation. The lungs were mechanically ventilated with a mixture of oxygen and air until the bivenocaval clamping and restarted after clamp removal. The bispectral index (BIS) was measured in all the patients and maintained at <50 throughout the surgery to make the two anesthetic schemes equivalent.

Cardiopulmonary Bypass

Anticoagulation was achieved with IV heparin 3 mg·kg⁻¹ to maintain the target activated clotting time of 480 seconds. After aortic and bivenocaval cannulations, CPB was instituted with a Terumo hollow fiber oxygenator with a blood flow between 150 and 300 ml·kg⁻¹·min⁻¹. The prime volume, 370–750 ml according to the patient’s weight, contained lactate-free Ringer’s solution, albumin, mannitol, blood, and heparin. Mean arterial pressure was maintained between 30 and 50 mm Hg by adjusting pump flow or by injecting 50 mcg phenylephrine boluses if required. Venous oxygen saturation was maintained at 65–75% by adjusting the blood flow rate and fraction of inspired oxygen (FiO₂) (50–100%). Body temperature during bypass was maintained at approximately 28°C except for patients undergoing circulatory arrest, which were cooled to approximately 20°C during the period of circulatory arrest. To arrest the heart, intermittent cold blood cardioplegia was given. Arterial blood gas analysis was performed every 20 min. Hematocrit was maintained at higher than 20%, and blood gases were managed according to the a-stat principle throughout CPB. On rewarming, 5 ng·kg⁻¹·min⁻¹ IV alprostadil and 5 µg·kg⁻¹·min⁻¹ dopamine were started to enable weaning from CPB. Before coming off CPB, patients were rewarmed to a rectal temperature of 35°C, and pacing was instituted whether there was partial or complete heart block or the heart rate was less than 80 beats per min. All patients underwent modified ultrafiltration at the end of the bypass. Anticoagulation was reversed with 3 mg/kg IV protamine sulfate. After completion of surgery, the patients were transferred to the ICU. They were weaned from ventilatory support as soon as they were hemodynamically stable, normothermic, and had achieved an adequate level of consciousness.

Blood Samples and Postoperative Assessments

Blood samples were collected in dry glass tubes before surgery and 0, 3, 6, 12, and 24 h after surgery. Serum was immediately separated and stored at −70°C for later quantitative determination of cTnI with a one-step enzyme immunoassay based on electrochemiluminescence technology (Beckman Instruments Inc., Fullerton, CA). Blood gas parameters of arterial blood samples were analyzed with a blood gas analyzer (ABL800, Radiometer, Denmark) according to the relevant ICU protocol. Serial oxygenation indexes (OI) were calculated as partial pressure arterial oxygen (PaO₂)/FiO₂. Fluid intake (including crystalloids, colloids, and blood products), output (urine, blood, and serous fluid loss), and fluid balance were recorded over a 36 h period after admission to the ICU. Inotropic support at each time point was quantified by calculating the inotropic score as described previously.^9–13 The inotropic score was calculated as (dopamine dose × 1) + (dobutamine dose × 1) + (adrenaline dose × 100) + (noradrenaline dose × 100), where all doses are expressed as microgram per kilogram per minute.

Primary and Secondary Outcomes

The primary outcome of this study was perioperative plasma cTnI concentrations. The secondary outcome was all-cause hospital death and major postoperative complications. Other variables including postoperative ventilation time, length of ICU stay, and postoperative hospital stay were also recorded and compared.

Statistics

In a retrospective analysis of 15 patients, the mean postoperative (SD) cTnI was 56.8 (25.0) ng·ml⁻¹. A power analysis based on these findings showed that we would need 102 patients to detect a difference in cTnI of 15 ng·ml⁻¹ with a = 0.05 and a power of 85%. The data were analyzed with the statistical package GraphPad Prism 4.0 for Windows software (GraphPad Software Inc, La Jolla, CA). The data with normal distribution were summarized as the mean and standard deviation and analyzed with ANOVA with Bonferroni correction for posthoc analysis. The data with non-normal distribution were expressed as the mean and 95% confidence intervals and were analyzed with the Mann–Whitney test. Because cTnI concentrations were not normally distributed, the data were first subjected to a natural logarithmic transformation before analysis by repeated measures ANOVA with Bonferroni correction. The categorical data were analyzed using the χ² test. The correlation coefficients between variables were calculated using the Pearson test for the normally distributed data and the Spearman test for the not normally distributed data. Values of p < 0.05 were considered statistically significant.

Results

Demographics Data

The groups were comparable with respect to sex, age, weight, type of surgery and CPB, aortic cross clamp, circulatory arrest times, cardioplegia solution dosages, hemocrit values during CPB, and ultrafiltration volumes. The number of cyanotic patients and those undergoing a ventriculotomy were well matched in each group (Table 1).

Table 1: Distribution of Characteristics and Congenital Heart Defects Between the Two Groups

Postoperative Clinical Data

There was no in-hospital death in either group. Two patients in the control group and one patient in the sevoflurane group were reintubated because of respiratory dysfunction; they recovered with no complications. One patient in the sevoflurane group who underwent Taussig-Bing anomaly repair was readmitted to the hospital because of pericardial effusion 15 days after discharge. This patient recovered after pericardial drainage. The urine volume during operation in the sevoflurane patients was statistically larger than that of the control patients (25.9 ± 22.6 ml vs. 21.9 ± 26.0 ml; p = 0.031; Table 2). Before the start of CPB, the blood pressure was similar between the two groups. However, after separation from CPB, the arterial diastolic pressure in the sevoflurane group was significantly higher than that in the control group (46.9 ± 9.3 mm Hg vs. 43.6 ± 8.9 mm Hg; p = 0.033) with the comparable inotropic scores. The systolic and mean arterial pressures were comparable between these two groups (Table 3). After admission to the ICU, all patients underwent mechanical ventilation. The serial OI of sevoflurane group were higher than that of control patients at the same time points of ICU 0, 3, 6, 12, and 24 h, although only the values at ICU 6 h reached statistically significance (Table 4). The overall ventilation time in the sevoflurane patients was significantly shorter than that in the control patients (26.1 [19.2, 33.0] h vs. 37.7 [19.2, 50.9] h; p = 0.014), but the postoperative 24 h urine, drainage volume, and inotropic scores were similar; the ICU time, the postoperative hospital days, and the perioperative blood use were also comparable (Table 4).

Table 2: Operative Measurements of the Two Groups, Mean ± SD

Table 3: Arterial Blood Pressure Measurements Before and After CPB (mm Hg)

Table 4: Postoperative Variables for the Two Groups, Mean (5%, 95% Confidence Interval)

Serial postoperative serum cTnI concentration measurements

In the first postoperative 24 h, serum concentrations of cTnI were determined at multiple time points. On admission to the ICU, serum cTnI concentrations increased significantly and reached their peaks at 3 h after admission to the ICU, then decreased slowly. However, they were still at a higher level at 24 h when compared with preoperative values. There were no significant differences between the two groups when the serum concentrations of cTnI were compared at the same time point (Table 5). Univariate analysis of variance showed that cTnI concentrations at 3 h after ICU arrival did not correlate with group treatment or cyanosis but correlated with right ventriculotomy (r = 0.30 [0.10, 0.48]; p = 0.0018). The peak value of cTnI concentration in cyanotic patients (61.89 [32.54, 91.24] ng·ml⁻¹) was not significantly different from that of noncyanotic patients (45.97 [33.77, 58.17] ng·ml⁻¹; p = 0.178), but the peak cTnI concentration was significantly higher in patients who underwent right ventriculotomy (73.39 [42.42, 112.4] ng·ml⁻¹) than in those who did not (43.45 [31.90, 55.00] ng·ml⁻¹; p =0.002).

Table 5: Serial Postoperative Serum Cardiac Troponin I Concentrations (ng·ml⁻¹) of the Two Groups (Mean [5%, 95% Confidence Interval])

Discussion

Although volatile anesthetics have been widely used in pediatric cardiac surgery, the potential beneficial effects on clinical outcomes and vital organ protection have not been extensively studied. A recent study from Singh et al.¹⁴ reported that 5 min treatment with one minimum alveolar concentration (MAC) of isoflurane, sevoflurane or desflurane after commencement of CPB provided some level of myocardial protection in children who underwent selective ventricular septal defect repair, with CK-MB as the myocardial damage marker. This might be the first double-blind, placebo-controlled clinical study to evaluate the preconditioning effect of three common clinically used volatile anesthetic agents in pediatric cardiac surgery against myocardial injury. They also found that preconditioning with volatile agents may also reduce the need for inotropic support, duration of mechanical ventilation, and total ICU stay. However, Malagon et al.⁹ reported that midazolam, propofol, and sevoflurane provided equal myocardial protection in pediatric cardiac surgery when using cTnT as a marker of myocardial damage; their postoperative clinical outcomes were also comparable.

Our study found no significant difference for the postoperative serum cTnI content between the sevoflurane and control groups, but we found that some minor clinical outcomes were improved in the sevoflurane patients. This was different from the studies by Singh et al. and Malagon et al. When we carefully reviewed these studies, we found that there were a few differences in their anesthetic protocols. Singh et al. only used volatile anesthetics for 5 min before cardioplegia-induced cardiac arrest. They did not use volatile anesthetics during anesthetic induction or other period of surgery, and their participants were only children with ventricular septal defects. Therefore, the characteristics were much different with our study and that of Malangon et al. Malagon et al. used sevoflurane in all patients for anesthetic induction. We already know that sevoflurane can mimic the cardioprotective effects of ischemic preconditioning.^10–12,15 Sevoflurane induction might provide some level of organ protective effects and blunt the sensitivity of the study design, and this might explain why Malagon et al. did not observe any beneficial effects for anesthesia maintenance with sevoflurane during the whole period of surgery. Taken together, these three studies might imply that sevoflurane provides some benefits in the surgical repair of congenital heart defects in pediatric patients. De Hert et al.¹³ implied that, at least in adult patients, the beneficial effects of volatile anesthetics were related to the method (dosage, duration, and timing) used. They demonstrated that the cardioprotective effects of sevoflurane in coronary artery bypass grafting (CABG) patients were clinically most apparent when sevoflurane was administered throughout the operation. However, in pediatric patients, the optimal protocol of volatile anesthetic use still needs to be established.

Although there was a report¹⁶ that the incidence of acute kidney injury after valvular heart surgery was higher with sevoflurane anesthesia that with propofol anesthesia, other studies demonstrated at least equivalent or better perioperative renal function with sevoflurane use in adult^17–19 and pediatric²⁰ patients who underwent cardiac surgery with CPB. We found that the intraoperative urine volume in sevoflurane patients was slightly but significantly larger than that in control patients, and the urine volumes during the first postoperative 24 h were similar between the two groups, which agree with most of the previously mentioned studies. The creatinine before surgery and at ICU arrival, postoperative day (POD)1 and POD3 were collected also. The creatinine has the surgery-related increase until POD3–POD5. However, the creatinine has no significant difference at each time point between two groups.

A phenomenon similar to reperfusion injury happens at the start of CPB in cyanotic patients. Allen et al.²¹ took biopsies of myocardial tissue in acyanotic and cyanotic patients before and 10 min after initiating bypass to measure antioxidant reserve capacity. In contrast to acyanotic patients, abrupt reoxygenation of cyanotic patients resulted in a significant depletion of endogenous tissue antioxidants. However, in our study, univariate analysis of variance showed that cTnI concentrations were not related to cyanosis, although the peak value of cTnI in cyanotic patients was higher than that in acyanotic patients, but the difference was not statistically significant. This study also found that the higher peak value of cTnI was correlated with right ventriculotomy. This result was similar to our previous study.²²

Rodig et al.²³ examined the dose-related effects of sevoflurane on systemic vascular resistance during CPB in patients undergoing cardiac surgery. They found that 1.0–2.0% sevoflurane was helpful in maintaining a stable systemic vascular resistance during hypothermic CPB, whereas SVRI increased at 15 to 20 min in patients who did not receive inhalational anesthetics. Ueda et al.²⁴ found that regulation of perfusion pressure during CPB using sevoflurane was safe and could easily maintain adequate SVRI, and the doses of norepinephrine needed at the end of CPB were significantly lower than in the chlorpromazine group. These studies showed that sevoflurane used during CPB had potential benefits in the regulation of systemic vascular resistance, but all the subjects studied were adult patients, and pediatric patients have not been extensively studied. Our results demonstrated that sevoflurane could significantly increase the diastolic arterial pressure at the end of CPB, although the difference between the groups was only 4 mm Hg. Because we did not measure the SVRI during the bypass period, whether sevoflurane is helpful for maintaining vascular reactivity and provides further benefits for the postoperative recovery (such as a shorter postoperative ventilation time) need further investigation.

Thus far, most relevant clinical trials demonstrated that the application of inhalational anesthetics in cardiac surgery with CPB is beneficial to the postoperative recovery.²⁵ Some research found that this beneficial effect was accompanied by the decreased release of serum biomarkers for myocardial injury.^{13,18,26–29} In addition, some studies demonstrated that the beneficial effects of sevoflurane on the postoperative recovery were not related to the decrease of these biomarkers. Piriou et al.³⁰ found that sevoflurane (1 MAC) pretreatment before CPB had no significant influence on the serum cTnI concentration and the protein content of protective signaling transduction in myocardial tissue, but the morbidity of low cardiac output index was significantly lower in sevoflurane patients. De Hert et al.⁶ also found that sevoflurane did not decrease the postoperative serum cTnT concentrations in patients who underwent CABG but significantly decreased the postoperative hospital time and 1 year mortality. These studies are the basis of the recent consensus that inhalational anesthetics can decrease the perioperative mortality of cardiac surgery.³¹

According to our results, we did not observe that sevoflurane inhalation during cardiac surgery in pediatric patients decreases the early postoperative serum cTnI release. However, we found that there were some beneficial effects of sevoflurane on the clinical parameters such as higher diastolic arterial pressures at the end of CPB, higher postoperative OI and a shorter postoperative ventilation time. It was reported in several small sample clinical investigations that sevoflurane inhalation during cardiac surgery related to reduced inflammatory responses which were reflected by decreased serum concentrations of interleukin 6 and matrix metalloproteinase 9 in Lindholm’s study³²and decreased plasma kynurenic acid content in Kotlinska-Hasiec’s study.³³ Because we did not investigate these inflammatory factors in this small clinical trial, we could not make certain that the benefit of sevoflurane inhalation during CPB on postoperative pulmonary function is the result of inflammatory inhibition. A larger, multicenter-clinical trial is warranted to determine the beneficial effects of sevoflurane, and extensive experimental investigations should also be conducted to elucidate the underlying mechanisms of sevoflurane application in pediatric patients undergoing cardiac surgery with CPB.