*Purpose: Circulating donor-derived cell free DNA (cfDNA; Allosure, CareDx, USA), a non-invasive test that could detect rejection in kidney allografts was validated using histologic diagnoses(HDx). The interpretation of these findings could be difficult as the inter and intra-observer agreement with regards to histologic diagnoses of rejection has been no higher than 60% in published studies. The Molecular microscope (MMDx, Edmonton, CA) might be more accurate than histology. In this prospective study, we compared the detection of rejection from cfDNA to both histology and MMDx reads.

*Methods: 50 indication kidney transplant biopsies were analyzed using HDx and MMDx. cfDNA was measured prior to biopsy. Gold standard diagnosis was based upon a clinical consensus(CDx) by 5 transplant nephrologists after assessing all clinical factors and results. Sensitivity and specificity were calculated using cfDNA cut-offs of <0.21% to rule-out rejection and >1% to rule-in rejection. Spearman’s correlation between cf-DNA and MMDx was calculated.

*Results: Based upon CDx, 17 (34%) had antibody-mediated rejection [AbMR; median cfDNA=1.2 (range: 0.23-6.9)], 3 (6%) had mixed rejection (median cfDNA=0.67 (range: 0.23-1.3)], 2 (4%) had T-cell mediated rejection [TCMR; median cfDNA=1.7 (range:1.2-2.2)], 29 (58%) had acute kidney injury (AKI; tubular injury, median cfDNA=0.23, range=<0.19-2.2)] and 1(2%) had BKV nephropathy (cfDNA=1.6). 7 cases (14%) had discrepant rejection diagnoses between MMDx and HDx. MMDx had a 100% agreement with CDx, compared with HDx (sensitivity: 86%; specificity: 93%). cfDNA>1% had a higher sensitivity and specificity for prediction of rejection when compared with MMDx than with histology. cfDNA<0.21% had a negative predictive value (NPV) of 93% to rule out rejection when compared to either HDx or MMDx. cfDNA scores showed a continuous positive correlation with histologic microvascular inflammation (G+PTC) scores (ρ=0.48; p<0.001), MMDx rejection score (ρ=0.64; p<0.0001), ABMR score (ρ=0.62; p<0.0001) and global disturbance score (ρ=0.54; p<0.0001). No such correlation was seen with GFR at biopsy or change in GFR from baseline.

*Conclusions: In this first report, we find that cfDNA>1% was better correlated with the Molecular microscope than it was with histology with excellent specificity for the prediction of rejection but poor sensitivity. cfDNA can be used as a screening test to rule out rejection (NPV~93% for cfDNA<0.21%) but cannot replace kidney biopsy to determine the type of rejection.

« Back to 2019 American Transplant Congress