Outcomes and Predictors of Failure of Systemic-to-Pulmonary Shunts: Experience of a Single Institution Over 14 Years

BackgroundSystemic artery-to-pulmonary artery (SA-PA) shunts provide effective palliation for complex congenital heart disease (CHD) but carry a risk for morbidity and mortality. We aimed to comprehensively analyze our experience with SA-PA shunts.MethodsOur institutional Society of Thoracic Surgeons (STS) database was queried to identify patients who underwent SA-PA shunts from 2009 to 2022, excluding those who underwent the Norwood procedure, right ventricle-PA shunt, or fenestrated patch. Definitions from the STS Congenital Heart Surgery Database Specifications were used. Shunt failure included dysfunction secondary to thrombosis, obstruction, stenosis, or outgrowth requiring intervention.ResultsA total of 287 patients met inclusion criteria. Shunts were placed at a median of 15.0 days (interquartile range 7.0-39.5). A thoracotomy approach was used in 178 out of 287 patients (62.0%), and cardiopulmonary bypass was employed in only 46 of 287 cases (16.0%). Survival to the next stage was 89.5% (246/275), with in-hospital mortality of 6.3% (18/287) and interstage mortality of 4.2% (12/287). Shunt failure occurred in 54/287 (18.8%), and 77/287 (26.8%) required reintervention for shunt-related complications. On multivariable analysis, poorer shunt failure-free survival was associated with any syndrome, left-sided arch vessel shunt origin, concurrent complex repairs, competitive flow from a patent ductus arteriosus, and delayed antiplatelet initiation. The thoracotomy approach was protective. Risk factors for worse survival to the next stage included shunt thrombosis and perioperative platelet transfusion.ConclusionsWhile complications remain common, our contemporary results demonstrate that SA-PA shunts remain a reliable palliation for CHD with insufficient pulmonary blood flow. Risk reduction may involve careful management of competitive pulmonary blood flow and prompt initiation of antiplatelet therapy.