Donor-derived cell-free DNA versus left ventricular global longitudinal strain in noninvasive detection of heart allograft injury

Abstract Introduction Invasive endomyocardial biopsy (EMB) is still considered the gold standard method for monitoring heart transplant (HTx) rejection. However, it is associated with potentially serious complications, and its histopathologic interpretation has high interobserver variability. Thus, noninvasive methods for surveillance of the transplanted organ remains of great interest. The noninvasive donor-derived cell-free DNA is a marker of graft injury which can indicate acute rejection, even before the histopathological diagnosis of rejection detected by EMB. The assay has a high negative predictive value for acute rejection. Left ventricular global longitudinal strain (LV GLS) is an emerging marker of subclinical myocardial injury which plays a crucial role in the evaluation of several cardiac diseases; however, its role in HTx injury has so far been poorly investigated. Purpose We aimed to investigate the efficacy of a local laboratory-run dd-cfDNA assay-based heart transplant rejection surveillance programme, and the prognostic and discriminatory performance of LV GLS for subsequent heart allograft injury. Methods HTx recipients without a history of allograft rejection requiring intensification of immunosuppressive therapy were transitioned from our EMB-based surveillance protocol to dd-cfDNA-based rejection surveillance. We assessed the percentage of dd-cfDNA to total cell-free DNA. Injury cut-off was defined by a dd-cfDNA level ≥0.20%, while the severe injury threshold was at ≥0.35%. LV ejection fraction was calculated using 2D Simpson’s method. To derive LV GLS, we analyzed apical 4-chamber, 2-chamber, and long-axis images by 2D speckle tracking echocardiography. Results A total of 244 dd-cfDNA samples were analyzed from 46 patients in fifteen runs performed monthly between October 2022 and December 2023. The dd-cfDNA fraction remained below the injury cut-off in most patients (81%). There were no missed rejection episodes. 88% of routine EMBs that would have otherwise been performed were safely avoided. Eighteen for-cause EMBs were performed based on dd-cfDNA levels suggesting injury. One EMB verified moderate cellular rejection (ISHLT grade 2R) that required steroid pulse therapy. Two EMBs proved mild cellular rejection, one EMB confirmed mild antibody-mediated rejection, while one EMB proved mixed rejection. Thirteen EMBs found no rejection. Mean LVEF was preserved (57.6%±7.6%). Mean LV GLS was -15.1%±2.5% in the whole cohort, while -15.6%±4.7% in recipients with HTx rejection on indication EMB. There was no significant correlation between dd-cfDNA and LV GLS (n=159). Conclusion The noninvasive dd-cfDNA assay is effective in ruling out acute rejection and safely decreases the necessity of invasive surveillance EMBs after HTx. LV GLS is less sensitive in comparison with the dd-cfDNA assay for detection of myocardial injury in the early stages of graft rejection in HTx recipients at low risk for rejection.