Abstract 13837: Evaluating Predictors of Pulmonary Valve Replacement Failure in Congenital Heart Disease Patients

Abstract

Introduction: We aimed to determine predictors of reintervention after pulmonary valve replacement (PVR) at our center with the novel consideration of valve angulation within the outflow tract. Methods: Patients with a history of PVR between 06/2007 and 06/2019 and subsequent cardiac catheterization, cardiac MRI, or chest CTA since PVR were eligible for inclusion, as stored images allowed for valve angle measurement. The angle relative to the patient’s anterior-posterior (horizontal) plane was measured using lateral angiography and sagittal cross-sectional images. Valve orientation was analyzed as follows: 1.) the measured angle, 2.) the difference in that angle from the mean of 40 control patients, and 3.) whether that angle was outside of 2 standard deviations from the mean. Other variables included: original cardiac anatomy, valve type, true internal valve diameter indexed for weight at PVR, and immediate post-operative peak systolic gradient and valve insufficiency. Freedom from reintervention was evaluated using the log-rank test and Cox proportional-hazards with multiple imputation of modest missing echocardiographic data. Results: Of 168 patients who underwent PVR during that period, 52 patients met inclusion criteria. Over a median follow-up time of 6.0 years (bootstrapped 95% CI [4.6, 7.4]), 14 of those 52 (26.9%) underwent reintervention (valvuloplasty or valve replacement). In both univariate analysis as well as bivariate analysis controlling for indexed valve diameter, no measure of valve orientation was associated with freedom from reintervention. Potential associations of valve type with the outcome disappeared in bivariate analysis. Higher internal valve diameter indexed for weight at PVR was consistently associated with earlier reintervention. Results for freedom from valve revision were similar to those for any reintervention, while acknowledging that we may be underpowered to detect certain associations. Conclusions: Valve orientation does not appear to impact freedom from reintervention following PVR; however, higher indexed internal valve diameter is associated with decreased time to reintervention. An augmented sample size through multicenter collaboration may allow for a more robust multivariate analysis.