Abstract 19792: Prediction of Right Ventricular Failure After Cardiac Transplantation : A Recipient Transcriptomic Study

Abstract

Introduction: Right ventricle failure (RVF) is a frequent and severe complication after cardiac transplantation. However, risk stratification for RVF is poorly achieved. Development of transcriptomic biomarkers for outcome prediction in cardio-vascular diseases is promising. Hypothesis: Our aim was to identify right ventricular gene expression signature associated to RVF and to define a transcriptomic biomarker that could predict post-transplantation RVF. Methods: Recipient RV myocardial samples of 44 patients transplanted from February 1998 to November 2002 in our center were collected. We retrospectively identified patients with (RVF group) and without (CTL group) post-transplantation RVF. A 4035-gene expression profile was obtained for all patients. Differentially expressed genes between RVF and CTL groups were identified and a molecular RVF predictor was used to determine for each patient a RVF prediction score (RVFs). Results: 9 (20%) and 18 (41%) patients were classified in RVF and CTL groups respectively. As compared to CTL group, RVF patients showed higher pre-operative bilirubin level and higher post-operative death rate. Molecular RVF predictor included 75 differentially expressed genes. Using this predictor, risk for post-transplantation RVF was 2.8-fold greater if RVFs was >0.5 (CI 95%: 1.243-6.305). Sensitivity and specificity of RVFs were 0.778 and 0.889, respectively. Using receiver operating characteristic analysis, RVFs area under curve (AUC) was significantly greater than AUC of commonly used RVF predictors (pulmonary vascular resistance, trans-pulmonary gradient). Conclusions: Gene expression profiling of recipient right ventricle could be used to predict post-transplantation RVF. Transcriptomic biomarkers should be further investigated as a new tool for selection of cardiac transplant candidates.