Circulating Cell-Free DNA as a Non-Invasive Marker of Pediatric Heart Transplant Rejection and Immunosuppressive Treatment

Abstract

Abstracts S75dd-cfDNA levels between 13 rejection, 14 post-rejection, and 217 non-rejection samples (R vs NR, p= 0.96), and no significant difference in dd-cfDNA levels between biopsy grades (132 OR, 103 1R, and 8 2R samples; p= 0.85 for OR vs ≥ 2R). Of interest was the observation that dd-cfDNA levels rose prior to rejection in 3/5 patients, and decreased post-rejection in 5/8 patients; and patients with only Grade 0 biopsies had consistently low dd-cfDNA levels. Conclusion: With approximately half of this pediatric cohort tested, we found no difference in dd-cfDNA levels between biopsy grades or in pediatric patients with treated acute rejection events. Additional studies are required to determine why the positive results reported in adults were not replicated in children.Purpose: Precise anticoagulation management is critical in minimizing the superimposed risks for bleeding and thromboembolism in CF-LVAD patients. Warfarin dose response is influenced by genetic polymorphisms for warfarin metabolism (Cytochrome P4502C9 [CYP2C9]) and the enzymatic target of vitamin K antagonists (vitamin K epoxide reductase complex 1 [VKORC1]). The purpose of this study is to validate several genomics-based warfarin dosing algorithms in CF-LVAD patients. Methods: This retrospective analysis included all CF-LVAD patients at a single academic institution with genotype data available for CYP2C9 (*1, *2, and *3 [rs1799853 and rs1057910]) and VKORC1 (-1639 G> A [rs9923231]) variants. Several dosing algorithms were tested, including those developed by Sconce et al., Anderson et al., Gage et al., and the International Warfarin Pharmacogenomics Consortium (IWPC), as well as the recommendations from the warfarin US package insert. To validate each algorithm, the predicted warfarin dose was compared against the actual dose at discharge from the index implantation. Actual daily dose was regressed on predicted daily dose, from which R2 values were derived. Results: 121 CF-LVAD patients with warfarin genotype data were identified. The mean age was 57.2 years, 20.7% female, 58.7% ischemic etiology, 56.2% Caucasian, and 32.2% African American. 22 patients (24.4%) carried at least one copy of CYP2C9 variant allele (*2, *3, or both), and 28 patients (31.1%) carried at least one copy of VKORC1 variant allele (A). In our cohort the warfarin dose recommended by the US package insert had the lowest correlation to the actual dose required upon discharge (R2 = 0.084). While the algorithms developed by Sconce and Anderson performed slightly better (R2 = 0.160 and 0.196, respectively), the doses suggested by the Gage and IWPC algorithms produced the highest R2 values (Figure). Conclusion: The Gage and IWPC algorithms were most accurate in predicting warfarin dose requirements in CF-LVAD patients.