Donor-derived cell-free DNA is a valuable monitoring tool after single lung transplantation: Multicenter analysis

Abstract

Donor-derived cell-free DNA (dd-cfDNA) is a nonspecific plasma biomarker for tissue injury which has been validated for monitoring of acute rejection (AR) after lung transplantation (LT). However, no studies to date have focused specifically on single lung transplantation (SLT). Herein we report the performance of dd-cfDNA in detecting AR in SLT from six academic centers who implemented this biomarker surveillance in their standard of practice (SOP). Dd-cfDNA test results were corrected for SLT by an algorithm in the CLIA-laboratory to permit comparison against the same 1.0% threshold used in double-lung transplant. Investigators reviewed patient SOP EMR clinical data to assign test results into cohorts based on clinical allograft health status. To avoid ambiguity in interpretation, samples drawn after a prior acute rejection or infection event or without histopathologic confirmation of acute rejection, were excluded from further analysis. Diagnostic cohorts included: Acute Rejection (AR, N=25 samples), healthy (STABLE, N=137), allograft-related Infection (INFXN, N=41), Chronic Lung Allograft Dysfunction (CLAD, N=7), and “OTHER” (N=12) types of graft injury. The study included a total of 257 dd-cfDNA results from 103 SLT patients while one patient was excluded due to active cancer. Samples were drawn a median of 233 days (IQR: 96-489) after SLT. Analysis of laterality for SLT (R vs L) and median dd-cfDNA fraction in AR and STABLE cohort was not statistically different. The median dd-cfDNA fraction was elevated with AR (1.8%) and INFXN (1.1%) vs STABLE (0.46%; p<0.0001). dd-cfDNA with CLAD was also significantly higher than STABLE cohort (p=0.0155). The Area Under Receiver Operator Characteristics (AUROC) Curve was 0.850 (95% CI: 0.72 - 0.95, p<0.0001) for AR vs STABLE cohort. Applying the dd-cfDNA threshold ≥ 1.0% for AR, yielded a Sensitivity= 77.8%, Specificity= 84.6%, Positive (PPV)= 38.31%, and Negative Predictive Value (NPV)= 96.83%. These multi-center data, incorporating real-world experiences, support the clinical validity and utility of dd-cfDNA monitoring of SLT recipients. Additional studies of the impact of biomarker surveillance on clinically meaningful outcomes should be forthcoming from robust, prospective, clinical trials already in progress.