Variability in Donor-Derived Cell-Free DNA Scores to Predict Mortality in Heart Transplant Recipients – A Proof-of-Concept Study

BACKGROUND: Over the last decade, expanding use of molecular diagnostics in heart transplantation has allowed implementation of non-invasive surveillance strategies for monitoring allograft health. The commercially available HeartCare platform combines the AlloMap gene expression profiling assay and the AlloSure donor-derived cell-free DNA test (dd-cfDNA). Beyond their established use for assessment of rejection, evidence is building for predictive utility, with the longitudinal AlloMap Variability score previously shown to correlate with the risk of future rejection, graft dysfunction, re-transplantation, or death. In this single-center, retrospective pilot study, we evaluated the performance of a novel AlloSure Variability metric in predicting mortality in a cohort of heart transplant recipients. METHODS: Seventy-two adult heart transplant recipients with at least 3 concurrent AlloMap/AlloSure results were included. Demographic, clinical, imaging, and laboratory parameters were captured. Variability was defined as the standard deviation of longitudinal AlloMap/AlloSure results. A Cox multivariable adjusted proportional hazards model was used to evaluate the variability metrics as predictors of mortality. Associations between AlloMap/AlloSure variability and donor specific antibody (DSA) status were also assessed. RESULTS: A total of 5 patients (6.9%) died during a median follow-up of 480 days. In a univariate Cox proportional hazards model, higher AlloSure variability (HR 1.66, 95%CI 1.14 – 2.41), but not AlloMap variability or the cross-sectional AlloSure/AlloMap results was associated with increased mortality risk. Longitudinal AlloSure variability was also higher among patients with both preformed DSA and those developing de novo DSA. CONCLUSION: Our results suggest that increased variability of dd-cfDNA in heart transplant patients is associated with both mortality risk and the presence of donor specific antibodies. These findings highlight the added value of longitudinal data in the interpretation of AlloMap/AlloSure scores in this population and open the door to larger studies investigating the utility of these metrics in shaping post-transplant clinical care paradigms.