Non-invasive rejection surveillance in stable heart transplantation using local laboratory-run donor-derived cell-free DNA assay: two-centre European feasibility study

Abstract

Abstract Background Monitoring adverse events is crucial after heart transplantation (HTx). Allograft rejection is a major etiological factor of graft dysfunction, hospitalisation, and death posttransplant. Endomyocardial biopsy (EMB) is the standard method of monitoring rejection; however, there are risks and limitations associated with it. The ISHLT guidelines for the care of HTx recipients have included non-invasive surveillance such as testing for donor-derived cell-free DNA (dd-cfDNA). While major studies have been conducted with patient cohorts and centralised testing in the USA, hereby we explored a European two-site cohort where dd-cfDNA was quantified at local centres by next-generation sequencing-based CE-marked assay. Purpose Our objective was to assess feasibility of assay and evaluate dd-cfDNA values in clinically stable HTx recipients. Methods HTx patients were enrolled at their scheduled visits in this cross-sectional study in 2 European HTx centres. Serial data after HTx, including clinical status, maintenance immunosuppression therapy and drug levels, diagnostic tests (blood tests, echocardiograms, coronary angiograms, EMBs), and adverse events were collected. Venous blood specimens were drawn in Streck cfDNA blood collection tubes for quantification of dd-cfDNA levels. We analysed dd-cfDNA fraction in plasma using AlloSeq cfDNA assay. For analysis, clinical stability was defined as asymptomatic patients >28 days post-HTx with normal graft function, without cytomegalovirus infection, history of rejection, or allograft vasculopathy. Results Between July 2021 and December 2022, 50 HTx recipients were enrolled at a median of 274 (112–395) days posttransplant. The median age at enrolment was 55.5 years. Of the total 82 dd-cfDNA results, 34 samples were excluded as they did not meet stability criteria. The median dd-cfDNA value of stable patients was 0.11% (95% CI 0.08%–0.13%). Median levels of dd-cfDNA did not differ between samples drawn in the first year after HTx (n=40) and at later time (n=8) [0.11% vs. 0.11%; p=0.41]. Of the 48 dd-cfDNA samples, 7 had values at risk based on the dd-cfDNA cut-off values from previous clinical validation trials: 4 at the injury cut-off of ≥0.20% and 3 at the severe injury threshold of ≥0.35%. Comparing clinical variables of samples with dd-cfDNA ≥0.20% (n=7) and <0.20% (n=41), we observed a 3-fold, non-significant increase in rate of de novo DSA formation (33.3% vs. 10.8%; p=0.19). Conclusions Our data show feasibility to analyse dd-cfDNA at the local level. In our two-site European reference cohort, dd-cfDNA is present at low levels (median 0.11%) which is within former reported ranges for clinically stable HTx recipients. Levels of dd-cfDNA do not vary beyond the first post-HTx year. The fact that in patients with de novo DSA higher dd-cfDNA levels are found; although, not statistically significant, suggests that dd-cfDNA may be a marker of subclinical rejection and requires further study.