Introduction: Detection of subclinical kidney transplant(KTx) rejection by surveillance biopsies and early management improves graft survival. However, these are cumbersome and invasive. Donor derived cell free DNA(dd-cfDNA) has emerged as a tool to predict KTx rejection in adult patients when >1% is present. Limited data is available in pediatrics. We aimed to determine whether surveillance with dd-cfDNA detects subclinical KTx rejection in children using a threshold >1%.

Methods: Retrospective chart review of all patients(n=30) followed at a single pediatric KTx program from 1/2021-12/2021. Patients with multiorgan transplant(n=2) and without dd-cfDNA(n=1) were excluded. dd-cfDNA was performed as surveillance, starting on 12/2020(every 3-6months). dd-cfDNA, demographics, eGFR, BK status, donor specific antibodies(DSA) and biopsy(Bx) findings(Banff 2019 criteria) were collected. Descriptive statistics and diagnostic test validity assessments were performed. Rejection outcome(for diagnostic validity testing) was defined as any of these: borderline T-cell mediated rejection(bTCMR), TCMR, antibody mediated rejection(ABMR), or mixed.

Results: During study period, 27 patients had dd-cfDNA performed(ESKD cause: 65%CAKUT, 31%GN; sex: 27%female; age at KTx: 12yrs; baseline egfr: 87 ml/min/1.73m2 [IQR 78-96]). Post KTx time at dd-cfDNA evaluations: 30%(n=8) < 1yr, 33%(n=9) 1-3yrs, 37%(n=10) > 3yrs. 11 patients had dd-cfDNA >1%; two of them were not biopsied(1 had BK viremia, dd-cfDNA decreased < 1% after lowering immunosuppression(IST); 1 had DSAs and abnormal coagulations, dd-cfDNA decreased < 1% after increasing maintenance IST). 44% of patients(12/27) had a biopsy: 9 with dd-cfDNA >1% (median 1.8% [IQR 1.6-3.2%]; Bx findings: 3 TCMR, 4 bTCMR, 2 mixed rejection) and 3 with dd-cfDNA < 1%, performed due to another indication(AKI=2 and BK viremia=1; median 0.78%[IQR 0.7-0.8%]; Bx findings: 1 bTCMR, 1 TCMR and 1 normal). dd-cfDNA >1% had a 100% specificity and 81.8% sensitivity(AUC: 0.909) to diagnose rejection. In patients with dd-cfDNA < 1% and no Bx, median dd-cfDNA was 0.23%[IQR 0.16-0.42%].

Conclusions: Subclinical rejection was detected using dd-cfDNA >1% in this cohort with high specificity/sensitivity, supporting its potential use as a graft rejection surveillance tool. Two patients had dd-cfDNA < 1% and rejection suggesting that lower cut-offs may be needed to enhance detection. Given the small sample size and low ABMR prevalence in this cohort, we were unable to assess whether dd-cfDNA discriminates between rejection types. Validation of dd-cfDNA is needed in larger scale pediatric studies.