Introduction: Lung transplant recipients (LTRs) face risks like acute cellular rejection (ACR) as a predisposing risk factor for the development of chronic allograft dysfunction. Donor-derived cell-free DNA (dd-cfDNA) may be a promising biomarker for detecting ACR without the need for invasive transbronchial biopsies (TBB). We aimed to validate a clinical plasma dd-cfDNA assay for ACR detection and explore its clinical value. The aim of the abstract is to analyze dd-cfDNA  levels for "minimal" cellular rejection (grade A1) or higher that have been adjudicated by cryobiopsy since TBB obtained using the forceps may not be adequate for a confident diagnosis
Methods: In this prospective single-center cohort, dd-cfDNA fraction was measured using a novel single-nucleotide polymorphism-based assay in LTRs undergoing surveillance bronchoscopy with cryobiopsies 2, 4, 6, and 12 months after transplantation (and when indicated). Performance characteristics were calculated for LTRs without ACR (stable cohort) and LTRs with ACR (defined as signs of ACR based on pathological assessment of the cryobiopsies ≥A1 regardless of grade B).
Results: The study included 44 samples from 39 individual LTRs and the incidence of ACR was 15.9%. Median dd-cfDNA fraction was similar for the stable cohort and the cohort with ACR (median 0.41% [0.15% to 0.72%] vs. 0.56% [0.10% to 3.07%], p=0.630). Regression analysis suggested that neither C-reactive protein (p=0.245), nor indication-bronchoscopy (p=0.561) was associated with the incidence of ACR. We detected ACR grade A1 (N=2), grade A2 (N=3), grade A3 (N=1), and no grade A4 or antibody mediated rejection (AMR). The area under the receiver operator characteristic curve for ACR was 59.3% (95% CI 38.3%-80.3%). Using a ≥1% dd-cfDNA fraction threshold (≥0.5% for single lung transplantations), negative predictive value for ACR was 85.3% (95% CI 74.8%-95.8%), positive predictive value 20.0% (95% CI 8.2%-31.8%), sensitivity 28.6% (95% CI 15.2%-41.9%), and specificity 78.4% (95% CI 66.2%-90.5%). In the sensitivity analysis, altering the ACR category (≥A1 vs. ≥A2) or the dd-cfDNA threshold > 0.85% did not produce significant changes in the outcomes.
Conclusion: The incidence of ACR (≥A1 or ≥A2) was not associated with the fraction of dd-cfDNA (independent of the threshold) in our cohort. Possible reasons for these findings include the use of cryobiopsies, which allow for more sensitive detection of minimal ACR, a small sample size, and an overall low incidence of ACR. Further study will focus on long term outcomes vis-à-vis CLAD development associated with “minimal” ACR without significant elevation in this injury biomarker.