Low Preoperative Superior Vena Cava Blood Flow Predicts Bidirectional Cavopulmonary Shunt Failure

ABSTRACT Background To determine whether preoperative superior vena cava(SVC) blood flow measured by cardiac magnetic resonance (CMR) predicts physiology and clinical outcome after bidirectional cavopulmonary shunt(BCPS). Methods The retrospective single-center study includes 65(2012-2017) patients who underwent BCPS. Preoperative CMR, echocardiography, catheterization, and clinical outcomes were reviewed. SVC flow was measured by phase contrast CMR. The Kaplan-Meier method and Cox regression was used for BCPS takedown free survival and predictor analyses. Results The absolute and indexed SVC flow was 0.5(0.4-0.7) L/min and 1.7(1.4-2.0) L/min/mm2 respectively, which was comparable to the SVC blood flow volume previously measured. The median age and body weight at BCPS was 6.5(5.5-8.5) months and 6.9(6.0-7.7) kg. After follow-up, at a median of 17.1(7.9-41.3) months, 14 patients(21.5%) underwent the Fontan completion and 40(61.5%) with BCPS physiology were waiting for the Fontan completion. The 11 remaining patients(16.9%), include those that underwent takedown(n=7, 10.8%) or died with a BCPS(n=4, 6.2%). Severe hypoxia was the leading cause of mortality, directly accounting for two thirds of deaths(66.6%, 6/9). The BCPS takedown free survival was 96.8% at six month, 79.9% at 3 years. Preoperative SVC blood flow was significantly positively correlated with early post-BCPS arterial saturation(p=0.00). The multivariable analysis showed SVC flow was the only factor associated with BCPS failure (HR:0.186, P=0.04) among the predictors related to the pre-BCPS anatomy and physiology. Conclusions SVC blood flow may be as critically important as pulmonary artery anatomic and physiologic parameters in evaluation of BCPS candidacy in the single ventricle population.