Neonates and Infants with Left Heart Obstruction and Borderline Left Ventricle Undergoing Biventricular Repair: What Do We Know about Long-Term Outcomes? A Critical Review

Background: The decision to perform biventricular repair (BVR) in neonates and infants presenting with either single or multiple left ventricle outflow obstructions (LVOTOs) and a borderline left ventricle (BLV) is subject to extensive discussion, and limited information is known regarding the long-term outcomes. As a result, the objective of this study is to critically assess and summarize the available data regarding the prognosis of neonates and infants with LVOTO and BLV who underwent BVR. Methods: In February 2023, we conducted a review study with three different medical search engines (the National Library of Medicine, Science Direct, and Cochrane Library) for Medical Subject Headings and free text terms including “congenital heart disease”, “outcome”, and “borderline left ventricle”. The search was refined by adding keywords for “Shone’s complex”, “complex LVOT obstruction”, “hypoplastic left heart syndrome/complex”, and “critical aortic stenosis”. Results: Out of a total of 51 studies, 15 studies were included in the final analysis. The authors utilized heterogeneous definitions to characterize BLV, resulting in considerable variation in inclusion criteria among studies. Three distinct categories of studies were identified, encompassing those specifically designed to evaluate BLV, those focused on Shone’s complex, and finally those on aortic stenosis. Despite the challenges associated with comparing data originating from slightly different cardiac defects and from different eras, our results indicate a favorable survival rate and clinical outcome following BVR. However, the incidence of reintervention remains high, and concerns persist regarding residual pulmonary hypertension, which has been inadequately investigated. Conclusions: The available data concerning neonates and infants with LVOTO and BLV who undergo BVR are inadequate and fragmented. Consequently, large-scale studies are necessary to fully ascertain the long-term outcome of these complex defects.


Background
The decision whether to perform biventricular repair (BVR) or univentricular palliation (UVP) in neonates and infants affected by multiple left heart obstructive lesions and borderline left ventricle (BLV) is always challenging [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16], and knowledge regarding the long-term prognosis of these patients remains scarce.Various scores utilizing echocardiography data have been proposed as a means of predicting the efficacy of BVR in BLV [2][3][4][5]9,14]; however, these are solely based on short-term results [3,4,14], with a mean follow-up period of 5-6 years.Furthermore, these scoring systems are not without limitations, which include heterogeneity 2 of 13 in the echocardiographic parameters and outcome measures evaluated, as well as the retrospective design during the score development [2][3][4][5]9,14].Additionally, significant variations in inclusion criteria were observed among the studies from differing authors [2][3][4][5]9,14].Moreover, the term "borderline left ventricle" encompasses a broad spectrum of complex cardiac defects characterized by one or multiple left-sided obstructions and diminutive left sections, with significant uncertainty regarding the optimal choice between BVR and UVP.The absence of a universally accepted definition of "borderline left ventricle" has resulted in the use of various definitions, such as critical left ventricular outflow (LVOT) stenosis, hypoplastic left heart complex, multiple left heart obstructive lesions, and small left heart structures.Additionally, neonates and infants with Shone's complex [17][18][19][20][21][22][23][24][25][26][27] have typically been evaluated separately, even though these congenital heart defects (CHDs) are also characterized by multiple left-sided obstructions and borderline left structures.Despite the high volume of literature on surgical and percutaneous aortic stenosis (AS) valvuloplasty in pediatric patients , studies have not specifically focused on BLV despite neonatal AS being typically associated with a diminutive left ventricle (LV).Data regarding the long-term outcomes of neonates and infants with Shone's complex [17][18][19][20][21][22][23][24][25][26][27] and critical neonatal AS  undergoing BVR are also limited.The aim of the present study is to systematically review data pertaining to the long-term outcomes of neonates and infants with one or multiple LVOT obstructions and BLV who have undergone a biventricular correction.

Methods
In February 2023, we conducted a review study within 3 different medical search engines (the National Library of Medicine, Science Direct, and Cochrane Library) for Medical Subject Headings and free text terms including "congenital heart disease", "outcome", and "borderline left ventricle".The search was refined by adding keywords for "Shone's complex", "complex LVOT obstruction", "hypoplastic left heart syndrome/complex", and "critical aortic stenosis".
The titles and abstracts of articles identified by this strategy were evaluated and excluded if (i) the reports were written in languages other than English (1 study); (ii) studies did not report a data follow-up of at least 3 years (13 studies); (iii) there were duplicate data (6 works); (iv) the reports mixed neonates and infants with older children (9 studies); (v) the reports evaluated only single lesions (e.g., isolated aortic stenosis, excluding multiple stenosis and borderline LV) (7 studies) (Figure 1).
observed among the studies from differing authors [2][3][4][5]9,14].Moreove "borderline left ventricle" encompasses a broad spectrum of complex car characterized by one or multiple left-sided obstructions and diminutive left s significant uncertainty regarding the optimal choice between BVR and UVP. of a universally accepted definition of "borderline left ventricle" has resulted various definitions, such as critical left ventricular outflow (LVOT) stenosis, left heart complex, multiple left heart obstructive lesions, and small left hea Additionally, neonates and infants with Shone's complex [17][18][19][20][21][22][23][24][25][26][27] have ty evaluated separately, even though these congenital heart defects (CHD characterized by multiple left-sided obstructions and borderline left structu the high volume of literature on surgical and percutaneous aortic st valvuloplasty in pediatric patients , studies have not specifically focu despite neonatal AS being typically associated with a diminutive left ventric regarding the long-term outcomes of neonates and infants with Shoneʹs com and critical neonatal AS  undergoing BVR are also limited.The aim o study is to systematically review data pertaining to the long-term outcomes and infants with one or multiple LVOT obstructions and BLV who have u biventricular correction.

Methods
In February 2023, we conducted a review study within 3 different me engines (the National Library of Medicine, Science Direct, and Cochrane Medical Subject Headings and free text terms including "congenital he "outcome", and "borderline left ventricle".The search was refined by addin for "Shone's complex", "complex LVOT obstruction", "hypoplastic syndrome/complex", and "critical aortic stenosis". The titles and abstracts of articles identified by this strategy were ev excluded if (i) the reports were written in languages other than English ( studies did not report a data follow-up of at least 3 years (13 studies); (iii duplicate data (6 works); (iv) the reports mixed neonates and infants with ol (9 studies); (v) the reports evaluated only single lesions (e.g., isolated ao excluding multiple stenosis and borderline LV) (7 studies) (Figure 1).
The included studies on aortic stenosis were all single-center with retrospective design [28,30,31,43,47].Here, mean sample sizes varied from 37 [30] to 84 subjects [28], while follow-up varied from a mean of 3.2 years [43] to >8 years [31], with a total duration of follow-up of 17.7 years in some cases [28].

Freedom from Reintervention
Reports on reintervention rates in borderline LV studies have been inconsistent or not reported at all [1].Even when reported, obtaining a real estimate is difficult due to very limited [7, 9,11] or varying [8] follow-up duration.Freedom from reintervention at 1 year varied from 50% [7] to 61% [9], while at 3 years, it ranged from 25% [11] to 50% [5].Multiple interventions were frequently required [4][5][6]10], with a series [10] of 72 hypoplastic left heart complexes with a mean follow-up of 5.9 years (range 2.0 to 12.1 years) requiring an estimated 1.9 interventions per patient [10].In the Cardiac Heart Surgery Society (CHSS) series, 19.2% required two reinterventions and 5.7% required three reinterventions [5].Studies pertaining to Shone's complex evaluated freedom from reintervention on longerterm follow-up, with freedom from reintervention estimates varying from 72% [22] to 61.3% [24] at 10 years.In all the studies, reinterventions on the mitral valve (repair or replacement) were the most common, along with subaortic membrane resection or more invasive interventions on the LVOT such as Konno or Ross operations [22][23][24].
Finally, in aortic stenosis freedom from reintervention varied from 21% [47] to 65% [28] at 10 years and from 18% [28] to 64% [31] at 15 years.The limited available data make it difficult to compare surgical and percutaneous valvuloplasty.However, a recent review and meta-analysis [54] demonstrated no significant difference in survival between surgical and percutaneous valvuloplasty for congenital aortic stenosis, although the incidence of reintervention was higher in percutaneous valvuloplasty (p < 0.001).

Clinical Outcome, Incidence of Pulmonary Hypertension, Need of Medications
Limited data are available on New York Heart Association (NYHA) class and the need for medication during follow-up.However, in the available studies, most patients were in class NYHA I at the last follow-up examination [6,7,10].In a small series of eight children with hypoplastic left ventricle and mitral stenosis with a mean follow up of 6.5 years (±4.5 years), most cases were in class NYHA I or II, with only three patients requiring medication.Another study [7] of over 39 hypoplastic left heart complexes, with a mean follow-up f of 34 months (range 177 months), reported that 74% of patients were in class NYHA I, while the remaining 26% were in class NYHA II, and only 19% of all patients required medications.
A study on a series of 43 patients with Shone's complex demonstrated that at 10 years, 82.6% of survivors were in NYHA I, with mild or less mitral regurgitation present in 66.7%, normal LV function was observed in 79.2%, and 91.7% were free of any LVOTO [24].Similarly, another series reporting on 27 Shone cases found that at 15 years, all survivors were either NYHA class I or II, 30% had moderate LV dysfunction, and 70% were free of residual obstruction [23].
The occurrence of pulmonary hypertension has been infrequently reported in the literature [1,6,7].In the included studies, the incidence varies from 7.1% [1] (24 hypoplastic left heart with a mean follow-up of 8 years) to 44.1% [22] (122 Shone's complex patients with a mean follow-up of 7.2 years), and even as high as 57.1% [6] (8 hypoplastic left ventricle+ mitral stenosis with a mean follow-up of 6.5 years).

Risk Factors for Poor Outcome and Reintervention
Several echocardiographic parameters have been evaluated to determine which ones are more predictive of poor outcomes following BVR.Studies have demonstrated that a lower aortic annulus Z-score [28,43,47], MV annulus Z-score [43,47], LV end-diastolic Z-scores [6,31,47], and LV dysfunction [28,47] were all predictors for successful BVR.Both the presence and the degree of endocardial fibroelastosis (EFE) were risk factors for poor outcome on univariate analysis [3,4,28,31].Moderate or severe EFE was either the only factor that remained significant on multivariate analysis [43] or the combination of moderate or severe EFE with either lower aortic valve Z-scores and younger age [3] or LVOT diameters < 4 mm [4].EFE has been found to be a predictor of death, even in cases of adequate LV function [4].Furthermore, EFE has been shown to be a predictor for reintervention [28] along with ventricular dysfunction [28], the presence of multilevel stenosis [31], a small aortic annulus Z-score [4,43], a lower aortic root, and LV end-diastolic diameters [43].In a study of 72 hypoplastic left heart complexes [10], a moderate to large ventricular septal defect (odds ratio-OR-= 0.22, p = 0.001) was found to be the strongest predictor of BVR failure, followed by a unicommisural valve (OR = 16, p = 0.006) and mitral valve Z-score (OR = 2.2, p = 0.002).
Discrepancies in the inclusion criteria among different studies hamper the possibility to compare data and to perform a meta-analysis.Furthermore, it is difficult to compare data originating from different eras, with time periods from the late 1990s [3,4,10,11] to the present day [1] accompanied by their different outcomes due to the improvement of diagnostic interventional strategies.Even Z-scores employed for echocardiographic diagnosis and disease severity estimation employed by older works [56][57][58] presented significant limitations that have been overcome by more recent works [59][60][61].Older Z-scores [58] tend to overestimate the disease severity [59].Thus, authors who utilized older nomograms [3,7,10] probably included even milder forms of LV obstructive diseases in the definition of "borderline left ventricle", which may have contributed to a more favorable outcome.
Neonates and infants with borderline LV who have undergone BVR continue to face a significant burden of reintervention.One or multiple reinterventions are often required [6][7][8][9][10][11], with up to 50% [7] or 60% [8,9] of patients requiring it within the first year of life.Early reintervention may indicate inappropriate decision making and is often associated with poor outcomes [5].In fact, a short interval after the first reintervention, particularly within the first 30 days, has been identified [5] as a risk factor for mortality.As stated previously, multiple reinterventions are often necessary [5,10] and the available data indicate that freedom from reintervention in long-term follow-up is limited to 20% of cases [47] and up to 18% at 15 years [28].
Incidence of pulmonary hypertension, which is a worrisome complication even in successful BVR without significant residual stenosis [34], was rarely described [1,6,7], varying from 44.1% [22] up to 57.1% [6] at medium-term follow-up.Despite the normalization of Z-scores of left sections that has been described by multiple authors [6][7][8][9]11,22], even in the lack of significant residual stenosis a moderate LV systolic [23,24] and diastolic dysfunction may persist and contribute to the maintenance or development of pulmonary hypertension.Unfortunately, however, little to no investigation has been conducted on LV diastolic function.

Strength and Limitations
This paper has several strengths.This is one of the first studies that has tried to evaluate together pathologies that have usually been considered separately (hypoplastic left heart complexes, Shone's anomalies, critical aortic stenosis) despite belonging to the same spectrum of the disease (e.g., one or multiple LVOT obstruction and BLV).This is also one of the few papers that has tried to address the long-term outcome of BLV, including some aspects that have been poorly evaluated so far (e.g., the clinical status, the presence of pulmonary hypertension).
This paper also has some limitations, including the lack of homogenous definitions of the same disease and the fact that the choice of different end points hampered the possibility to perform a metanalysis of the current data.Even comparisons among such different data (and coming from different eras) resulted in difficulties.Large, multicenter studies with clear and uniform disease definitions and well-defined outcomes are advised for a better understanding of the long-term outcomes of complex CHDs characterized by one or multiple LVOT obstruction and BLV.

Conclusions
Data on medium-and long-term outcomes of neonates and infants with multiple LVOT obstruction and borderline LV undergoing BVR are limited and fragmented.While survival rates have improved, patients who underwent BVR remain burdened by high reintervention rates.It is necessary to conduct large studies with standardized criteria for inclusion and exclusion, outcomes, and evaluation of clinical and echocardiographic variables, as well as with longer and more uniform follow-up intervals, to gain a better understanding of the long-term outcomes in this specific patient population.Our critical appraisal of the available literature may help guide clinicians in both parental counseling and clinical decision making, while also laying the foundation for future standardized studies regarding BVR for neonates and infants with BLV and multiple LVOT obstructions.

Table 1 .
(A) Borderline LV definition according to different authors.(B) Shone's complex definition according to major studies.

Table 2 .
(A) Major studies on outcomes in borderline LV attempting BVR.(B) Major studies on outcomes in Shone undergoing BVR.(C) Major studies on outcomes in neonates and infants with aortic stenosis undergoing either surgical/percutaneous valvuloplasty.