Patients recovering from surgery for CHD can suffer from haemodynamic instability or low cardiac output in the immediate post-operative period, characterised by signs and symptoms of hypoperfusion, including delayed capillary refill, decreased urine output, weak pulses, metabolic acidosis, and systemic hypotension. Reference Chandler and Kirsch1 Haemodynamic instability or low cardiac output can lead to end-organ dysfunction, delayed recovery, and in severe cases progress to cardiovascular collapse and cardiac arrest. Paediatric cardiac surgeons, anaesthesiologists and intensivists caring for neonates, infants, and children who undergo surgery for CHD are therefore commonly focused on prevention and management of haemodynamic instability or low cardiac output in the post-operative period.
While the occurrence of low cardiac output following surgery for CHD requiring cardiopulmonary bypass has been well described, Reference Wernovsky, Wypij and Jonas2–Reference Cavigelli-Brunner, Hug and Dave4 data describing low cardiac output or haemodynamic instability following procedures in which cardiopulmonary bypass is uncommon are sparse. Pulmonary artery banding is a surgical technique utilised by paediatric cardiovascular surgeons for a variety of congenital heart lesions, most commonly to control pulmonary blood flow and prevent pulmonary over-circulation that can be caused by large left-to-right shunts or complex single-ventricle physiology. Reference Sharma5,Reference Angeli, Pace Napoleone, Turci, Oppido and Gargiulo6 Patients who undergo pulmonary artery banding are rarely exposed to cardiopulmonary bypass, yet low cardiac output or haemodynamic instability necessitating the use of inotropic support following this procedure has been reported in specific patient populations. Reference Wernovsky, Giglia, Jonas, Mone, Colan and Wessel7,Reference Li, Zhang and Benson8 We aimed to examine a heterogenous cohort of patients who underwent pulmonary artery banding at our institution, report the proportion of patients who required inotropic support for low cardiac output or haemodynamic instability post-operatively, and identify pre-operative or operative risk factors associated with its occurrence.
Materials and methods
Patient population
We performed a retrospective cohort study of neonates and infants less than 1 year of age who underwent banding of their main pulmonary artery at Riley Hospital for Children at Indiana University Health between January 2016 and July 2019. Patients who underwent banding of right and left pulmonary arteries were excluded. The primary outcome of interest was the need for inotropic support in the immediate post-operative period, which we defined as the initiation of a continuous inotropic infusion (e.g., dopamine, epinephrine, or milrinone) for depressed myocardial function, compromised perfusion, or systemic hypotension in the first 24 hours after surgery. Patients receiving inotropic support pre-operatively at the time of surgery were therefore also excluded. This study was approved by the Institutional Review Board at Indiana University.
Operative protocol
Riley Hospital for Children at Indiana University Health is a tertiary care referral centre at which four paediatric cardiovascular surgeons perform approximately 450–550 cardiothoracic operations annually. Pulmonary artery banding is performed for various indications at our institution, most commonly as the first procedure in the staged repair of complex left-to-right shunt lesions often accompanied by arch obstruction or in staged palliation of complex single-ventricle anatomy with unrestricted pulmonary blood flow. A left thoracotomy approach is used for the majority of cases except in the setting of l-transposition of the great arteries. Band material consists of umbilical tape or silastic membrane, with diameters typically ranging from 5 to 6 mm depending on patient size and desired degree of constriction. Band diameter is adjusted intraoperatively depending on observed peripheral oxygen saturation values. Intraoperative transthoracic echocardiography is often used to verify the initial band gradient. Cardiopulmonary bypass is not required for the pulmonary artery banding procedure but is occasionally implemented if additional surgical procedures are performed.
Anesthetic management for pulmonary artery banding is not protocolized at our institution. In most cases, a combination of a low inspired concentration of sevoflurane, low-dose intravenous anaesthetics (e.g., ketamine and fentanyl), and neuromuscular blockade (e.g., cisatracurium and rocuronium) are used for induction of anaesthesia. Maintenance of anaesthesia is typically achieved with intravenous ketamine or, more recently, intravenous dexmedetomidine in combination with a low inspired concentration of sevoflurane. Maintenance of anaesthesia is also often supplemented by intermittent bolus doses of intravenous anaesthetics such as fentanyl and ketamine and is accompanied by an infusion of a neuromuscular blocking agent, the most common of which is cisatracurium. Post-operatively, inotropic support is not implemented for prophylaxis against low cardiac output or hypotension but rather only implemented if signs or symptoms of low cardiac output or haemodynamic instability are observed. The decision to initiate inotropic support is also not protocolized and therefore at the discretion of the primary care team. Additionally, for some patients who remain endotracheally intubated after surgery, morphine infusion may be initiated for pain or sedation management at the discretion of the primary care team.
Data collected and statistical analysis
Pre-operative, intraoperative, and post-operative data were collected for all patients. Pre-operative shock was defined by the Society of Thoracic Surgeons Congenital Heart Surgery Database (STS-CHSD).[12] Vasoactive inotrope score was calculated as previously described: Reference Gaies, Jeffries and Niebler13 Dopamine dose (mcg/kg/min) + Dobutamine dose (mcg/kg/min) + 100 × Epinephrine dose (mcg/kg/min) + 10 × Milrinone dose (mcg/kg/min) + 10,000 × Vasopressin dose (units/kg/min) + 100 × Norepinephrine dose (mcg/kg/min). Data are represented as mean with standard deviation for normally distributed continuous variables, median with interquartile range for skewed continuous variables, and absolute counts with percentages for categorical variables. Bivariate analyses comparing patients who received inotropic support were compared to patients who did not receive inotropic support using Student’s t-tests, Mann–Whitney U tests, and Fisher’s exact tests as appropriate. Multivariable regression analysis was performed to identify independent factors associated with the need for inotropic support after pulmonary artery banding. All baseline, pre-operative, and operative variables collected were analysed; variables with p-values of less than 0.2 on univariate analysis were considered for the multivariate model. Variables with p-values less than 0.05 after multivariate analysis were identified as independent risk factors for the need for post-operative inotropic support.
Results
We reviewed 69 patients who underwent surgical banding of their main pulmonary artery. Eight patients receiving inotropic support pre-operatively at the time of surgery were excluded. Median age at surgery in the remaining 61 patients was 10 days (range: 3,111). Additional surgical procedures performed in 41 patients (68%), of which 5 required the use cardiopulmonary bypass. Cardiac lesions and additional surgical procedures are summarised in Table 1.
Cardiac lesions and surgical procedures for patients who underwent main pulmonary artery banding
AVC: atrioventricular canal defect; BAS: balloon atrial septostomy; BTT: Blalock–Taussing–Thomas shunt; CoA: coarctation of the aorta; CPB: cardiopulmonary bypass; DILV: double-inlet left ventricle; DORV: double-outlet right ventricle; HLH: hypoplastic left heart; HRH: hypoplastic right heart; IAA: interrupted aortic arch; PA: pulmonary artery; TGA: transposition of the great arteries; VSD: ventricular septal defect.
Inotropic support was initiated within 24 hours of surgery in 30 patients (49%), with initiation occurring in the operating room in 15 patients and in the ICU in 15 patients. The most common first-line therapy was dopamine, which was initiated in 21 patients (67%). For the other patients who received inotropic support, first-line therapy was milrinone in 6 patients, epinephrine in 2 patients, and dobutamine in 1 patient. Seven patients received an infusion of a second inotropic agent. Median peak vasoactive inotrope score during the first 24 hours for these patients was 5 (range: 3,11). Patients who received inotropic support had significantly greater peak base deficit and significantly lower urine output during the first 24 hours as compared to patients who were managed without inotropic support, as shown in Table 2. Patients who received inotropic support also had significantly longer duration of mechanical ventilation – median 73 hours (25%,75%: 46,144) versus 41 hours (25%,75%: 21,68), p < 0.001 – and ICU stay – median 7 days (25%,75%: 4,13) versus 4 days (25%,75%: 3,7), p < 0.001.
Baseline and pre-operative data of patients who did and did not receive inotropic support after pulmonary artery banding
Data are represented as median (25,75%) for continuous skewed variables and counts (%) for categorical variables.
a At any point in time pre-operatively.
b Per Society of Thoracic Surgeons Congenital Heart Database definition (pH <7.2, lactate >4).
c Any degree of depressed left, right, biventricular, or single-ventricle function on pre-operative echocardiogram.
Additional variables for patients who received inotropic support within 24 hours of surgery are compared to the rest of the cohort in Tables 2 and 3. Baseline and pre-operative characteristics were not significantly different between groups. Intraoperative anaesthesia consisted of intravenous fentanyl in 59 patients (97%), inhaled sevoflurane in 58 patients (95%), intravenous ketamine in 53 patients (87%), intravenous dexmedetomidine in 14 patients (23%), and intravenous propofol in 1 patient (2%). Patients who received inotropic support had significantly longer surgical duration [median time 3.9 hours (25%,75%: 3.2,4.8) versus 3.5 hours (2.6,4.0), p = 0.031], received significantly more intraoperative ketamine [median cumulative dose 4.0 mg/kg (25%,75%: 2.8,5.9) versus 1.8 mg/kg (0.9,4.5), p < 0.001], and tended to have shorter durations of sevoflurane exposure [median time 2.4 hours (25%,75%: 1.0,3.7) versus 3 hours (25%,75%: 2.2,4.0), p = 0.11]. Moreover, both the maximum and minimum doses of inhaled sevoflurane used during the operation were significantly lower in patients who received inotropic support. Box plots of total cumulative ketamine dose and maximum sevoflurane dose are depicted in Figures 1A and 1B.
Box plots comparing ( a ) cumulative ketamine dose and ( b ) maximum sevoflurane concentration used in patients who received inotropic support (n = 30) and patients who did not receive inotropic support (n = 31) after pulmonary artery banding. Patients who received inotropic support were exposed to greater cumulative doses of ketamine (p < 0.001) and lower maximum sevoflurane concentration (p = 0.005).
Operative data of patients who did and did not receive inotropic support after pulmonary artery banding
PA: pulmonary artery; PRBC: packed red blood cells.
Data are represented as median (25,75%) for continuous skewed variables and counts (%) for categorical variables.
a Peak band gradient measured at any point post-operatively, available for 56 patients (30 who received inotropic support, 26 who did not), obtained at median post-operative day #3 (range: 0,13).
b Peak band gradient measured within first 24 hours post-operatively, available for 24 patients (15 who received inotropic support, 9 without).
c Initiated prior to initiation of inotropic support when applicable.
In multivariable analysis, total cumulative ketamine dose remained significantly associated with inotropic support after pulmonary artery banding, independent of total surgery time (Table 4). More specifically, odds of inotropic support in patients who received greater than 2.5 mg/kg of ketamine intraoperatively were 5.5 times that of the rest of the cohort.
Multivariable analyses of risk factors for inotropic support after pulmonary artery banding surgery
a Cumulative ketamine dose as a continuous variable.
Discussion
In our heterogenous cohort of neonates and infants who underwent pulmonary artery banding, initiation of inotropic support for presumed low cardiac output syndrome or haemodynamic instability occurred in nearly half of the cases. Notably, total surgical durations for these procedures were relatively short (median duration 3.7 hours) and use of cardiopulmonary bypass was infrequent (8%), yet inotropic support was commonly implemented. Literature on low cardiac output or the use of inotropic support after pulmonary artery banding is limited. Nearly three decades ago, Wernovsky and colleagues published their seminal work describing the post-operative course following pulmonary artery banding and systemic-to-pulmonary artery shunt placement in patients with d-transposition of the great arteries, as a preparatory procedure preceding their arterial switch operation. Reference Wernovsky, Giglia, Jonas, Mone, Colan and Wessel7 In this patient population, transient episodes of low cardiac output, characterised by decreased ejection fraction, mild metabolic acidosis, low urine output, tachycardia, mild hypotension, and the need for inotropic support, were common within the first few days after the procedure. The authors postulated that the acute increase in left ventricular ejection impedance produced by the pulmonary artery band was likely one of the factors contributing to the development of low cardiac output, causing acute left ventricular pressure overload, consequent left ventricular dysfunction, and impaired right ventricular diastolic function due to ventricular interaction and alterations in the geometry of the ventricular septum. Similarly, Li and co-workers reported diminished cardiac output post-operatively in neonates with hypoplastic left heart syndrome who underwent bilateral pulmonary artery banding as part of their hybrid palliation and attributed this finding to the imposition of an acute pressure load on a volume overloaded right ventricle. Reference Li, Zhang and Benson8
Additional studies providing granular data describing the post-operative course of neonates and infants who undergo pulmonary artery banding for other indications, such as the patients examined in this report, are lacking. To our knowledge, this study is the first to report the frequency of inotropic support utilisation in a heterogenous cohort of patients who underwent pulmonary banding. Surprisingly, inotropic support was not more common in neonatal patients, patients with single-ventricle anatomy, or patients with pre-operative shock or ventricular dysfunction. Indeed, the only factor associated with the use of inotropic support for low cardiac output that could be identified in this study was the cumulative dose of ketamine anaesthetic used during surgery. This association was independent of duration of surgery, that is, patients did not receive more ketamine because they required longer operations. We also noted that patients who received post-operative inotropic support received lower absolute doses of inhaled sevoflurane and tended to be exposed to inhaled sevoflurane for shorter periods of time. Anaesthetic management, therefore, was variable, and patients who received post-operative inotropic infusions were more likely to receive an anaesthetic regimen that had greater reliance on ketamine for induction and maintenance of anaesthesia during their pulmonary artery banding procedure.
Ketamine is commonly utilised for anaesthesia during surgery for CHD and for procedural sedation during necessary procedures such as endotracheal intubation or chest tube thoracotomy in the cardiac ICU. Reference Loomba, Gray and Flores9,Reference Morray, Lynn, Stamm, Herndon, Kawabori and Stevenson10 Clinicians often opt to use ketamine in these clinical scenarios because evidence suggests that it may have less adverse haemodynamic effects as compared to other intravenous sedatives and anaesthetics. Reference Oklü, Bulutcu, Yalçin, Ozbek, Cakali and Bayindir11,Reference Sungur Ulke, Kartal, Orhan Sungur, Camci and Tugrul12 Ulke and co-workers compared the haemodynamic effects of ketamine and sevoflurane during induction of anaesthesia in a cohort of 47 children who underwent surgery for CHD. Reference Sungur Ulke, Kartal, Orhan Sungur, Camci and Tugrul12 In this study, mean systolic blood pressure was significantly lower in patients who received sevoflurane at 4, 6, and 8 minutes post-induction, though hypotension was not reported in either cohort. In a more recent study, Han and colleagues compared the effects of ketamine and sevoflurane for induction of anaesthesia in 44 children who underwent repair of ventricular septal defect. In contrast to the prior study, however, these investigators measured additional haemodynamic parameters and found that, while mean arterial blood pressures were higher in patients receiving ketamine, these patients had lower stroke volume index, cardiac index, and cardiac power output as compared to patients who received sevoflurane. Patients who received ketamine also had higher heart rates and systemic vascular resistance, the latter of which was likely most responsible for the observed differences in mean arterial blood pressure. Indeed, ketamine has been shown to have a dose-dependent direct negative inotropic effect on human heart muscles, Reference Han, Liu, Pan, Luo, Li and Ou-Yang14 and the favourable haemodynamic profile often associated with ketamine is related to inhibition of neuronal catecholamine uptake. Reference Sprung, Schuetz, Stewart and Moravec15 We cannot assert, from our data, that higher doses of ketamine were primarily responsible for the use of inotropic support in these patients. We speculate, however, that in some patients, especially those in which inotropic support was initiated in the operating room or early in the cardiac ICU, higher cumulative doses of ketamine could have contributed to the clinical symptomatology that prompted inotrope administration. A prospective clinical trial comparing clinical outcomes after PA banding using different anaesthetic regimens – no ketamine versus low-dose ketamine versus high-dose ketamine – should be pursued to confirm our findings.
This study is limited by its retrospective design. As a retrospective study, anaesthetic management was not protocolized and the rationale for the anaesthetic approach for each patient could not be gleaned from the medical record. It is therefore possible that the differences in cumulative doses of ketamine or inhaled anaesthetic used in some of the study patients were not arbitrary but rather chosen based on patient stability. Similarly, use of inotropic support was not strictly protocolized but rather at the discretion of the clinical team. Inotropic support for this patient population however is not implemented prophylactically at our institution but rather in response to clinical signs of low cardiac output or inadequate perfusion. Moreover, the cohort of patients receiving inotropic support had significantly greater base deficit and significantly lower urine output during the first 24 hours, providing some objective evidence that cardiac output was likely more compromised in these patients. Additionally, though pre-operative echocardiographic data were available for all patients, echocardiographic data prior to initiation of inotropic support were not available for most patients. We therefore are unable to effectively characterise the pathophysiology of each patient’s low cardiac output or haemodynamic instability, for example, ventricular dysfunction, atrioventricular valve regurgitation, high systemic vascular resistance, pulmonary over-circulation, or a combination of these factors. Lastly, these data are from a single centre and thus, due to the variation in surgical and anaesthetic management that exists between institutions, have limited generalizability.
In conclusion, low cardiac output or haemodynamic instability necessitating initiation of inotropic support following pulmonary artery banding was common and was not associated with age at surgery, underlying cardiac lesion, or pre-operative illness severity. We identified cumulative dose of ketamine greater than 2.5 mg/kg for induction and maintenance of anaesthesia to be associated with inotropic usage, independent of duration of surgery. Future research focused on ketamine and post-operative outcomes following surgery for CHD should be pursued.
Acknowledgements
None.
Financial support
This research received no specific grant from any funding agency, commercial, or not-for-profit sectors.
Conflicts of interest
None.
Ethical standards
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national guidelines on human experimentation (please name) and with the Helsinki Declaration of 1975, as revised in 2008, and have been approved by the institutional review board at Indiana University.
