Repeat kidney transplant recipients with active rejection have elevated donor‐derived cell‐free DNA

In 2015, 13% of recipients of kidney transplants in the United States were recipients of repeat transplants for prior allograft failure.1 Compared to single kidney transplant recipients (SKTRs), these repeat kidney transplant recipients (RKTRs) have inferior graft survival and a higher risk of rejection.1, 2 Despite these risks, transplant patients benefit from a repeat transplant compared to dialysis support.3 The use of noninvasive biomarkers of allograft injury may optimize the care of RKTRs.

Donor‐derived cell‐free DNA (dd‐cfDNA) has established utility in diagnosing the probability of active rejection in SKTRs,4 but its value in RKTRs remains undetermined. In this study, we evaluated the dd‐cfDNA in patients with more than one kidney transplant. We compared the levels of dd‐cfDNA in 12 RKTRs to levels in 202 SKTRs. These cohorts were drawn from the Circulating Donor‐Derived Cell‐Free DNA in Blood for Diagnosing Acute Rejection in Kidney Transplant Recipients (DART) study (ClinicalTrials.gov Identifier: NCT02424227),4 where surveillance for rejection began less than 2 months posttransplant and there was no clinically indicated biopsy at the first visit and no rejection while on the study. Of the 12 RKTRs, 11 patients had 2 and one patient had 3 kidney allografts in situ. Median dd‐cfDNA in the RKTR surveillance cohort (n = 12) (0.29%) was higher than in the SKTR surveillance cohort (0.19 %, P < .001) (Figure 1A). However, both were significantly lower than the established 1% dd‐cfDNA rejection threshold.4

Figure 1.

Donor‐derived cell‐free DNA measured in kidney transplant recipients. (A) Percent of dd‐cfDNA in recipients with repeat transplants (RKTRs) is significantly higher than in recipients with single kidney transplant (SKTRs). (B) In RKTRs, at the time of clinically indicated biopsies, dd‐cfDNA is significantly higher in patients with biopsy‐proven rejection than in patients without rejection. The 6 rejection cases (and respective dd‐cfDNA) are one acute Banff 1A T cell–mediated rejection (TCMR, 0.94%); one Banff IB TCMR (0.88%); one acute/active antibody‐mediated rejection (ABMR, 0.54%), and 3 chronic, active ABMR (1.78%, 2.33%, and 3.44%). The biopsy findings in the 5 nonrejection graft dysfunction episodes consisted of acute tubular injury (ATI) and grade I interstitial fibrosis/tubular atrophy (IF/TA) (0.12%); glomerulonephritis, ATI, and grade I IF/TA (0.36%); BK polyomavirus, ATI, and grade I IF/TA (0.41%); glomerulonephritis and grade I IF/TA (0.48%); and one had no major findings (0.59%). The TCMR median (0.91%) was not significantly higher than the 5 nonrejection median (0.41%) (P = .095), whereas the ABMR median (2.06%) was significantly elevated (P = .032) in this small sample

We also examined dd‐cfDNA levels in all 11 RKTR patients (9 patients with 2, one patient with 3, and one patient with 4 kidney allografts in situ) at the time of a clinically indicated biopsy of their current allograft. We compared the dd‐cfDNA levels in the subset of 5 RKTRs with no rejection findings to the 6 RKTRs with biopsy diagnosis of rejection (2 RKTRs with T cell–mediated rejection and 4 with antibody‐mediated rejection). dd‐cfDNA in RKTRs with rejection was higher (median 1.36%) than in RKTRs without rejection (median 0.41%, P = .009) (Figure 1B). The dd‐cfDNA levels in RKTRs with rejection are similar (median 1.36%) to the levels in SKTR with rejection4 (n = 27, median 1.62%, P = .85).

The AlloSure method used in this study to determine the dd‐cfDNA level does not require genotyping of the donor(s) and does not distinguish which allograft(s) may contribute to the total dd‐cfDNA measured.5 However, our results demonstrate higher “combined” dd‐cfDNA levels in RKTRs at the time of surveillance testing than the levels observed in SKTRs. Of practical consequence, these “combined” dd‐cfDNA levels in RKTRs at the time of surveillance testing are significantly below the dd‐cfDNA levels seen in RKTRs at the time of rejection. Although patients with rejection have significantly higher levels of dd‐cfDNA compared to those without rejection, the small number of cases precludes defining test sensitivity or specificity. Moreover, ongoing inflammation in prior allograft(s) could be a confounding feature and limit test utility. Our findings warrant further study and highlight the potential utility of dd‐cfDNA monitoring regardless of prior kidney transplantation.

AUTHOR CONTRIBUTIONS

DJH, MG, and JPY participated in research design; SGM, JHC, TA, JSB, DJH, MG, JPY, and RBM participated in the writing of the paper; SGM, JHC, TA, JSB, RBM, DJH, MG, and JPY participated in carrying out the research; and DJH participated in data analysis.

DISCLOSURE

The authors of this manuscript have conflicts of interest to disclose as described by the American Journal of Transplantation. TA, SGM, and RBM have received grant support from CareDx. DJH, MG and JPY are employees of CareDx. JSB has received grant support and honoraria from CareDx. JHC has no conflicts of interest to disclose.