Utility of non-invasive rejection biomarkers to guide immunomodulation in kidney transplant patients with active COVID-19 infection

Purpose: Treatment of Coronavirus Disease 2019 (COVID-19) infection in organ transplant recipients has involved reduction of immunosuppression (IS). It is plausible that the cytokine release syndrome associated with COVID-19 coupled with reduction in IS may predispose transplant patients to rejection. Methods: Since March 2020 we initiated a protocol of measuring donor derived cell free DNA (ddcfDNA) (Allosure, CareDx) and HLA DSA at the time of diagnosis of COVID-19 infection (prior to reduction or cessation of antimetabolite). ddcfDNA, HLA DSA and additional clinical markers of allograft function were serially monitored until the point of nasopharyngeal (NP) swab clearance of COVID-19. Results: Thirty three transplant recipients were included, the majority were kidney only (KT) (31/33;94%) and the other two were simultaneous liver kidney transplant (SLK) recipients. Most were African American (23/33, 70%) with deceased donor transplants (25/33;76%) and underwent induction with rabbit anti-thymocyte globulin (31/33, 94%). Baseline IS included mycophenolate in 32/33 (97%) patients. Patients presented with COVID-19 at a median of 57 months (Range: 1-182) post transplant. Antimetabolite was stopped in 20/33 (61%) patients while it was reduced in the others (13/33;39%). Mean ddcfDNA at diagnosis was 0.45±0.39% in the KT recipients. In the two SLK recipients the values were 4.8% and 6.4% (below the cut off of 10% to detect rejection in liver transplants). Nineteen (58%) patients underwent repeat ddcfDNA testing at a median follow-up of 20 days (Range 14-127). In these recipients the ddcfDNA did not change from 0.44±0.37% to 0.30±0.36% (p=0.172). At a median NP swab clearance of 41 days (Range: 20-128), the mean eGFR at diagnosis was 57.9±30.5 ml/min/1.73m2 and remained unchanged at 57.4±29 ml/ min/1.73m2. Three KT recipients had an initial ddcfDNA of greater than 1%, two of these had pre-existing chronic active ABMR. One patient had a value greater than 1%, who then underwent a biopsy for new low grade class II DSA after his NP swab cleared and had early evidence of transplant glomerulopathy without active microvascular inflammation. Twenty-nine patients (88%) did not have any HLA DSA at the time of diagnosis and only one developed de-novo HLA DSA during infection. Four (12%) sensitized patients had pre-formed HLA DSA at the time of diagnosis and during the infection with reduced IS there were no new DSA. Conclusions: In the setting of active Covid-19 infection, we report the utility of ddcfDNA, a validated biomarker of immunological graft injury as a non-invasive tool to monitor allograft function while allowing for reduction in IS. In the majority of patients, ddcfDNA was low initially and remained low on subsequent testing arguing against allograft injury. In addition despite reduced IS, COVID-19 infection did not stimulate alloimmune responses in the short-term as evidenced by a very low rate of de-novo DSA.