The Impact of Dominant Ventricular Morphology on the Early Postoperative Course After the Glenn Procedure

The dominant ventricular morphology affects both the early and late outcomes of the Fontan procedure, but its impact on the patients’ status immediately following the Glenn procedure is unknown. This study aims to evaluate the effect of the infants' dominant ventricular morphology on the immediate course after undergoing the Glenn procedure. This single-center, retrospective study included all patients who underwent the Glenn procedure between October 2003 and May 2016. The patients were divided into two groups according to their dominant ventricular morphology. Their postoperative records were reviewed and compared. Out of the 89 patients who underwent the Glenn procedure during the study period, 40 (44.9%) had dominant right ventricular morphology and 49 (55.1%) had left ventricular morphology. There were no significant group differences in baseline characteristics or operative data. The maximal postoperative vasoactive-inotropic score was significantly higher and the extent of ventricular dysfunction was significantly more severe in the dominant right ventricle group (P < 0.05). The length of hospitalization was slightly but not significantly longer in the hypoplastic LV group. It is concluded that patients with a dominant LV morphology had a superior ventricular function and required less inotropic support compared to that of a dominant RV morphology in the immediate postoperative course following the Glenn procedure. Survival was not affected by these differences. Further study to determine the pathophysiologic basis for these differences is warranted.


Introduction
Patients with single ventricular morphology represent one of the most challenging types of congenital heart defects (CHD) [1]. The ventricular morphology may be left, in a hypoplastic right ventricle, right, in a hypoplastic left ventricle, or undetermined. The currently accepted management of single-ventricle defects, regardless of the underlying cardiac anatomy, consists of a three-stage surgical palliation process, with the goal of achieving blood circulation in sequence [2]. The first-stage intervention depends upon the anatomy and associated physiology of the CHD, while the second and third stages, namely, the Glenn and Fontan procedures, are generally similar among all patients who are candidates for single-ventricle palliation. The bidirectional Glenn procedure (or Kawashima in cases of interrupted inferior vena cava (IVC)) involves the creation of an anastomosis between the superior vena cava (SVC) and the right pulmonary artery (RPA) [3].
Long-term outcomes of the bidirectional Glenn of patients with right ventricular (RV) morphology is reportedly associated with a higher risk of adverse events, such as atrioventricular valve regurgitation (AVVR), and late Fontan failure [4,5]. However, the effect of a single-ventricle morphology on the immediate postoperative course following the bidirectional Glenn procedure has not yet been reported. The objective of this study, therefore, was to evaluate the early postoperative course after the Glenn procedure according to the dominant ventricular morphology.

Study Design, Ethics, and Patients
This single-center retrospective observational study of prospectively collected data was performed at the Edmond J. Safra International Congenital Heart Center. The study was approved by the institutional review board of the Chaim Sheba Medical Center, Tel Hashomer. All of the patients who had undergone the Glenn procedure between October 2003 and May 2016 comprised the cohort. The sole exclusion criterion was undetermined ventricular morphology. Data were collected from the patients' digital records, and the patients were divided into two groups according to their dominant ventricular morphology. The hypoplastic left ventricle (HLV) group included patients with hypoplastic left heart syndrome (HLHS), mitral atresia, aortic atresia, unbalanced, left dominant atrioventricular canal (AVC) defect, transposition of the great arteries (TGA)-HLV, and double outlet right ventricle-HLV. In the hypoplastic right ventricle (HRV) group, the patients' CHD included tricuspid atresia, pulmonic atresia, TGA-HRV, unbalanced, right dominant AVC, and double inlet left ventricle-HRV. The follow-up period for each patient was defined as the time from the procedure until hospital discharge. All variables were collected within this time period, and the period of hospitalization prior to surgery was not included.
The crude effect of the ventricular morphology on the perioperative course was tested by continuous variables, including the length of hospitalization (LOH), duration of mechanical ventilation, maximal vasoactive-inotropic score (VIS: dopamine [mcg/kg/min] + dobutamine [mcg/kg/min] + 100 × epinephrine [mcg/kg/min] + 10 × milrinone [mcg/kg/min] + 10,000 × vasopressin [units/kg/min] + 100 × norepinephrine [mcg/kg/min]), and the blood oxygen saturation level at discharge [6]. Postoperative complications and echocardiographic features, including AVVR, pericardial effusion, flow through the Glenn shunt, and ventricular function at discharge, were also recorded. All preoperative and postoperative echocardiograms were independently and blindly interpreted by two pediatric cardiologists (YS and UP). A third cardiologist determined the diagnosis in cases of controversy.

Statistical Analysis
The distributions of variables were presented by means and standard deviations (SD) for continuous variables, and by frequencies for the categorical ones. LOH was transformed into log scale to improve the agreement with normal distribution. The effect of the ventricular morphology on the continuous variables was tested twice, first by a T-test with unequal variances and then by backward linear regression. The initial list of variables in the latter included all potentially important covariates.
A similar logic was applied to the binary outcomes. First, the relationship between the outcomes and the ventricular morphology was tested by the Fisher's exact test for the corresponding 2 × 2 frequency table. The backward logistic regression was then applied with the same list of potential covariates. The crude effect of ventricular morphology on discrete variables was calculated using the Fisher test rather than the chi-square test, due to the low occurrence of adverse outcomes in the study population. The variable of ventricular morphology was fixed in the model to determine its effect in each regression after adjustment for other covariates. Only the final concise model was presented for each outcome.
All tests were two-sided. The P values were calculated either by the Fisher test for discrete variables or the t-test for continuous variables, and considered significant when less than 0.05. Corrections for multiple comparisons were performed as indicated. All calculations were carried out with STATA SE software (StataCorp LLC, College Station, TX, USA). All results are described as mean (± SD) unless stated otherwise.

Results
A total of 89 patients (mean age 12.87 ± 9.29 months, 54 [60.7%] males) underwent the Glenn procedure during the study period, and they all met the study inclusion criteria. Forty patients (44.9%) had right ventricular morphology (the HLV group), and 49 (55.1%) had left ventricular morphology (the HRV group). There were no significant group differences in the preoperative and intra-operative characteristics ( Table 1).
Evaluation of the crude effect of the dominant ventricular morphology revealed a significant group difference in the maximal VIS: the maximal VIS was higher in the HLV group compared to that of the HRV group (10.96 vs 4.12, respectively, P = 0.011). The rate of occurrence of ventricular dysfunction was also higher in the HLV group (7 vs 1, P = 0.03). The LOH was slightly but not significantly longer in the HLV group (12.37 vs 9.45 days, P = 0.067). The rates of postoperative complications, such as the need for extracorporeal membrane oxygenation (ECMO), tracheostomy, or chylothorax, and any associated neurological deficits and infections, were similar for the two groups. There was no mortality in the HRV group and only one mortality case in the HLV group (Table 2).
Mechanical ventilation time was tested twice: first, as a continuous variable and then as a categorical variable of < 8 h or ≥ 8 h of postoperative ventilation. There were no significant group differences, although there was a trend toward inferiority of the HLV group as demonstrated by more patients having been ventilated for longer than 8 h (56% vs 42% for the HRV group, P = 0.1).
A logistic regression analysis revealed that the maximal VIS was significantly affected by the ventricular morphology, while the LOH was not. However, the LOH was affected by the type of Glenn procedure performed, as well as by the infant's age at the time of the procedure. Specifically, older infants and those undergoing a unilateral bidirectional Glenn had shorter hospitalizations than younger patients and those undergoing the Glenn that included the Damus-Kaye-Stansel (DKS) procedure (Table 3).

Discussion
The ventricular morphology has been reported to affect the long-term outcomes following a Fontan procedure for the palliation of single-ventricle CHD [7,8], and that right ventricular morphology is associated with a higher risk of Fontan failure compared to left ventricular morphology [5,[7][8][9]. The evidence for the immediate postoperative effect of ventricular morphology in Fontan patients, however, is controversial: while several studies reported increased mortality and prolonged ventilation and hospitalization for patients with dominant right ventricular morphology, others demonstrated no differences in outcomes in the early postoperative course [7,8,10]. The effects of the dominant ventricular morphology on the immediate postoperative outcomes following second-stage palliation, however, have not been reported to date.
The results of the present study demonstrated decreased ventricular function and a greater need for ionotropic support for the patients who had a dominant right ventricular morphology after the Glenn procedure compared to those who had left dominance. Interestingly, none of the patients in the entire cohort evidenced either moderate or severe ventricular dysfunction on preoperative echocardiography, although, despite that, those with a dominant right ventricular morphology had worse outcomes. A possible explanation for that finding is the fact that, embryonically, the right ventricle was not designed to withstand the high pressures of the systemic circulation [11], and that it could not recover from the stress of the operation as efficiently as the dominant left morphology ventricle. Furthermore, it has been reported that the right ventricle suffers from a higher fiber stress, a higher rate of ventricular dilatation, and lesser circumferential fiber shortening after Fontan procedures compared to the left ventricle, all of which are directly related to ventricular function [11]. It was also shown that after first-stage palliation, a single right ventricle demonstrated progressive reduction in longitudinal and circumferential function, which may  [12]. The second significant finding of this study was the greater need for ionotropic support among the HLV infants. The magnitude of the inotropic support is an important predictor of the patient's postoperative course, and the amount of cardiovascular support required during the first 48 h after procedures involving cardiopulmonary bypass (CPB) was also shown to predict eventual morbidity and mortality [13,14]. For that reason, the VIS is considered a good predictor of short-term outcomes where higher scores are associated with worse outcomes, such as prolonged intensive-care unit, total LOH, and the duration of mechanical ventilation [13,15]. Higher VIS could be biased by a longer CPB time [13], however, the CPB time did not differ significantly between the current study groups ( Table 1).
The LOH was not significantly affected by the ventricular morphology. While the crude effect analysis showed a possible trend for shorter LOH in patients with left ventricular morphology, the transformation into logistic regression analysis showed no difference whatsoever, calling for further research to determine whether such an effect actually exists.
Given the fact that patients after the Glenn procedure benefit hemodynamically from spontaneous negative pressure ventilation since the pulmonary blood flow is directly influenced by the venous return, these patients are usually extubated early [11,16,17]. As such it is difficult to assess Table 2 The crude effect of the ventricular morphology on the postoperative outcome of Glenn procedure Statistically significant differences are given in bold Statistically significant differences were shown in the requirement for ionotropic support, which was higher among the HLV group, and the higher rate of ventricular dysfunction in this group of patients. Other postoperative measures that were compared and present in this table were found in similar rates between the groups HLV Hypoplastic left ventricle, HRV hypoplastic right ventricle, VIS vasoactive-inotropic score, AVVR atrioventricular valve regurgitation  Lastly, given the low mortality rate among patients undergoing the Glenn procedure which ranged from 0 to 4% [5,18], and in this study 0.9% (n = 1) it was difficult to prove any significant differences between the groups. However, even though the effect of the ventricular morphology on mortality and other severe adverse outcomes could not be tested due to their low occurrence rate in this study population, there was a significant group difference in the maximal VIS, which may reflect a higher risk of mortality and severe morbidity among patients with right ventricular morphology compared to those with left ventricular morphology.

Limitations
The main limitation of our study was its retrospective design which subjected it to selection bias. Second, even though all of the patients underwent right heart catheterization prior to their Glenn procedure, the technical data of the procedure were unavailable and are therefore not a part of this study. Nonetheless, all of the patients who undergo the Glenn procedure in our institution must have pulmonary vascular resistance lower than 3 wood units and pulmonary artery pressure lower than 20 mmHg, leading us to assume that all of the study participants had met these criteria. Lastly, the postoperative course of the first-stage palliation was not considered in the description of the preoperative condition, which might have affected the group's homogeneity.

Conclusion
The findings of this study demonstrated a significant difference in the ventricular function between left and right ventricular morphologies, with worse cardiac function and higher requirement for inotropic support postoperatively in patients with right ventricle dominance. As for the length of hospitalization and duration of ventilation, due to the particular physiology of Glenn patients and the need for early extubation, no significant differences between the groups were found. Further study to determine the pathophysiologic basis for these differences is warranted.
Author contributions E.K is a resident assisting major part of the procedures and wrote the main manuscript, prepared the tables and figures.I.A collected database.D.M and A.S are the two senior surgeons who performed all the procedures.Y.S is the cardiologist who performed most of the TEE exams pre and postoperatively.U.P-conceptualization and PCICU attending.All authors reviewed the manuscript.
Funding None.
Data Availability Supporting data are available upon reasonable request.

Declarations
Competing interest The authors declare that they have no conflict of interest.

Ethical Approval
The study was approved by the hospital IRB. Approval number: 4527. Date of approval January 4, 2011. Last extension date: February 28, 2021. (patient' consent for participation-not relevant).
Informed Consent All authors consent for publication.