Abstract
BACKGROUND: Plasma percent donor-derived cell-free DNA (%dd-cfDNA) has been investigated as a biomarker of allograft injury after lung transplantation. We sought to determine the clinical utility of %dd-cfDNA as a screen for acute cellular rejection (ACR) and respiratory infections (RIs) among lung transplant recipients (LTRs). METHODS: We retrospectively analyzed %dd-cfDNA results from 95 plasma samples collected from 81 bilateral LTRs >45 days after transplant with a paired transbronchial biopsy performed within 24 hours after sample collection. We calculated sensitivity, specificity, negative predictive value (NPV), and positive predictive value of %dd-cfDNA to detect ACR and RIs and used a generalized estimating equation model to compare %dd-cfDNA between groups. RESULTS: A dd-cfDNA threshold of 0.5% had low sensitivity to detect ACR among LTRs (41.67%), as did a 70% increase in %dd-cfDNA (50.00%). The NPV was high (88.89% and 87.50%, respectively) but driven by the low prevalence of ACR (12/95 [12.6%]). The area under the receiver operating characteristic curve (AUC) was 0.499 (95% confidence interval [CI] [0.326-0.672]) and 0.360 (95%CI [0.132-0.588]) for the detection of ACR and ACR grade ≥A2, respectively. The adjusted mean %dd-cfDNA trended higher in LTRs with a definite or possible RI (1.218, 95%CI [0.671-2.212]) than in LTRs without microbial isolation (0.731, 95%CI [0.525-1.017], p = 0.059), but was not significantly different from those with microbial colonization (0.873, 95%CI [0.538-1.415], p = 0.390). The AUC for the detection of allograft dysfunction due to ACR and/or RI was 0.573 (95%CI [0.431-0.716]). CONCLUSIONS: %dd-cfDNA may have limited utility as a screening tool to detect ACR and/or RI among LTRs.