Surgical Decision-Making in Neonates with Borderline Left Ventricular Hypoplasia without Significant Aortic or Mitral Valve Stenosis using Cardiovascular Magnetic Resonance

Abstract

AbstractBackgroundMost patients with significant left ventricular (LV) hypoplasia undergo single ventricle (SV) palliation, but biventricular (Bi-V) repair is viable in some patients with borderline LV hypoplasia. We sought to identify CMR (cardiovascular magnetic resonance) criteria predictive of successful primary Bi-V repair in neonates with borderline LV hypoplasia without significant stenosis of the mitral valve (MV) and aortic valve (AV), and to determine reasons for reintervention after successful Bi-V repair.MethodsThis retrospective study included all patients with borderline LV hypoplasia who underwent CMR from 2003-2024 for surgical decision-making. Patients with abnormal segmental connections, atrioventricular septal defects, unrestrictive ventricular septal defects, those with more than mild MV stenosis (mean Doppler flow gradient > 5 mmHg) and/or more than mild AV stenosis (peak Doppler flow gradient >20 mmHg) were excluded. Patients were divided into two groups based on initial intervention - primary Bi-V repair and hybrid/ other staging procedure. Outcomes were categorized as successful primary Bi-V repair, successful staged Bi-V repair and failure to achieve Bi-V repair (hybrid followed by SV palliation, transplant, death). Fisher exact test and Mann-Whitney U test was utilized to explore potential relationships. ROC curves were used to test diagnostic accuracy of parameters to predict successful primary Bi-V repair.ResultsAmong 37 patients meeting the inclusion criteria, 23 (62%) patients underwent successful primary Bi-V repair, 8 (22%) underwent staged Bi-V repair, 6 (16%) failed to achieve Bi-V repair. Patients who underwent successful primary/ staged Bi-V repair had higher values for left ventricular diastolic volume index (LVEDVi 28 mL/m2vs. 17.4.00 mL/m2; p <0.002), higher blood flow volume through the ascending aorta (QAo:1.99 L/min/m2vs. 0.97 L/min/m2, p <0.012), and QAo/ superior vena cava (QSVC) flow ratio (1.44 vs. 0.85, p =0.034) compared to those who had failure to achieve Bi-V repair. CMR LVEDVi cutoff of CMR 27 mL/m², had 87% sensitivity and 79% specificity with an AUC of 87.6% and QAothreshold of 1.9 L/min/m2had 65.2% sensitivity and 92.9% specificity (AUC: 86.0%) to predict successful primary Bi-V repair. Of 31 patients with primary or staged Bi-V repair, 7 (22%) underwent reinterventions for LVOT obstruction followed by mitral stenosis.ConclusionsCMR plays a critical role in pre-operative evaluation, surveillance and decision-making in patients with borderline LV hypoplasia. In patients with borderline LV hypoplasia without MV/AV stenosis, successful primary Bi-V repair can be achieved when the CMR-derived LVEDVi is >27 mL/m2and QAois > 1.99 L/min/m2.