Donor-derived Cell-free DNA as a New Biomarker for Cardiac Allograft Rejection: a Prospective Study (FreeDNA-CAR).

Abstract

Background: There is a long-standing need for a non-invasive biomarker that allows monitoring of cardiac allograft rejection, avoiding the need for periodic endomyocardial biopsies (EMB).

Methods: Multicenter, observational, prospective study, performed between 2019 and 2023 (NCT04973943). All patients underwent 7 per-protocol surveillance EMB during the first post-HT year. Donor-derived cell-free DNA (dd-cfDNA) levels were determined before each EMB, using Next Generation Sequencing Technology (Allonext™ assay, Eurofins Genome). The primary endpoint was the association between dd-cfDNA levels and the presence of acute cellular rejection (ACR) in EMB.

Results: The study included 206 patients from 12 centers, with 1090 pairs of EMB/dd-cfDNA determinations available for analysis. EMB with ACR (n=49) were associated to dd-cfDNA levels significantly higher than those without, median 0.189% (IQR 0.05-0.70) vs 0.095% (0.04-0.23), p=0.013. A dd-cfDNA threshold of 0.10% showed a negative predictive value for ACR of 97%. A statistically significant association between NTproBNP and dd-cfDNA was also found, with an increase of 0.007% dd-cfDNA (95%CI 0.003-0.011) for every 500 units of NTproBNP, p <0.001. The combination of both biomarkers for diagnosis of ACR showed an area under the ROC curve of 0.681, and this combined approach was significantly better than dd-cfDNA alone (area under the ROC curve 0.603), p = 0.016. Using a cut-off point of 0.10% for dd-cfDNA and 1000 UI/ml for NTproBNP, negative predictive value increased to 98.1%.

Conclusions: Dd-cfDNA may be a useful biomarker to rule out significant ACR in a low-risk population. However, a dd-cfDNA value above normal threshold does not correlate robustly with the presence of disease. The combination with NTproBNP, a readily available biomarker, increased the discrimination power of dd-cfDNA alone.