Donor-derived Cell-free DNA and Acute Kidney Rejection: A Golden Opportunity or a False Hope?

Acute rejection is a major impediment to long-term allograft function,1 which is a critically important outcome for kidney transplant recipients, caregivers, and health professionals.2 Early detection through screening in asymptomatic individuals may allow initiation of appropriate interventions to prevent significant injury to the allograft before it reaches an irreversible state. In clinical practice, a robust and reliable screening test must conform to the following criteria: it must be reliable and accurate, safe to administer, cost-effective, acceptable, and available to patients who are at risk of developing the condition.3 Serum creatinine is the oldest and the most frequently used biomarker of kidney injury. Although serum creatinine is economical and widely available to patients with kidney transplants, its ability to discriminate acute rejection from other pathologies such as acute tubular necrosis, ischemia, disease recurrence, and infections (polyomavirus nephropathy) is limited.4 Kidney biopsy has been the gold standard for diagnosis of acute rejection in kidney allograft. Histopathology can uncover and reliably detect a wide range of disease in the allograft with some certainty, but protocol biopsy is not routinely performed in many transplant centers worldwide because of its invasiveness, and potential complications including bleeding and pain. In addition, the optimal frequency and the cost–benefit ratios of conducting surveillance biopsy are highly heterogenous across different countries, regions, and clinical settings. Some also argued that the transplant biopsy is not always the perfect reference standard as there are significant interobserver variations between histopathologists and the definition of acute rejection, based on the Banff criteria, is evolving over time. Therefore, a noninvasive and accurate tool for early detection of kidney injury because of acute rejection remains an unmet need. Over the past decades, a plethora of biomarkers using urine and blood samples has been proposed as alternative index tests for screening and monitoring for acute rejection in the kidney allograft. Donor-derived cell-free DNA (dd-cfDNA) is fragments of DNA that are released into the bloodstream from injured donor allograft cells and are therefore a potential indicator of cell injury and death resulting from acute rejection within the allograft.5 There are many methods of quantifying dd-cfDNA that include next-generation sequencing or targeted approaches using preselected single-nucleotide polymorphisms in droplet digital PCR (ddPCR). DD-cfDNA can be measured in absolute values, or relative to the total cell-free DNA in the recipient's blood.6 A "dose-dependent relationship" between levels of dd-cfDNA and severity of acute rejection has been observed in kidney transplant cohorts, with rising relative dd-cfDNA levels corresponding to higher grades of acute rejection.7 The diagnostic value of dd-cfDNA has also been extensively reviewed in the literature. A systematic review and meta-analysis of diagnostic test accuracy studies of dd-cfDNA in kidney transplant recipients reported reasonable test performance for detecting acute rejection. Using a dd-cfDNA cut point of 1%, the pooled test sensitivity and specificity estimates were 70% and 78%, respectively.8 The predictive values for these tests are generally acceptable but, of course, are dependent on the disease prevalence. Any "sensitive" tests applied to a low disease prevalence population will undoubtedly perform poorly. This is an important concept because acute rejection risk differs between individuals and populations, based on their immunological risk profiles as well as the choice of immunosuppression. Real-world data to assess such variability are therefore welcome. In this issue of Transplantation, Bromberg et al9 presented the findings of a longitudinal cohort study that assessed the association between the trajectory of dd-cfDNA and acute rejection. The authors also performed a similar analysis using repeated measures of serum creatinine values as exposures. Using data from the ProActive Registry, a multicenter study that enrolled 4997 kidney transplant recipients from 54 sites in the United States. This substudy included 424 transplant recipients, who were representative of the original cohort of kidney transplant recipients within the registry. Using multilevel regression modeling that accounted for the clustering effects of repeated measures of dd-cfDNA and serum creatinine values of the individual, the authors found an association between rising dd-cfDNA measures and biopsy-proven acute rejection. This relationship was observed for both acute T cell–mediated rejection and antibody-mediated rejection. On the contrary, a similar trend was not observed for the longitudinal assessments of serum creatinine and acute rejection. A separate analysis was conducted to evaluate the performance of the dd-cfDNA assay and its ability to correctly classify patients with and without acute rejection. In this subset of 249 recipients with matched biopsy, at the dd-cfDNA threshold of 1%, the positive and negative predictive values for acute rejection were 32.6% and 97.9%, respectively. Therefore, assuming the annual prevalence of acute rejection in this cohort is around 8%, a "negative" test could potentially rule "out" a rejection event with some certainty (translating to approximately 1 episode of missed rejection per patient per year), but a "positive" test cannot accurately rule "in" a diagnosis of acute rejection. Among those with a positive test, only 32% had acute rejection. Hence, approximately two-thirds of positive tests (around 68%) would still be wrong. This study offers some novel insights and enticing clinical implications. Serial measurements of dd-cfDNA may provide some important prognostic information regarding the potential risk of acute rejection above and beyond what standard and routine measures of serum creatinine can offer. A rising dd-cfDNA may predict those at risk of acute rejection so that appropriate early intervention such as augmentation of immunosuppression can be instigated to prevent the detrimental effects of acute rejection. However, like many excellent studies, it has limitations and therefore the findings must be interpreted with caution. Given the observational and retrospective nature of the current design, this study is subjected to several biases including selection, ascertainment, spectrum, incorporation, and verification biases. Also, it does not permit a direct comparison between serum creatinine alone and the added value of dd-cfDNA for the prediction and diagnosis of acute rejection in kidney transplant recipients. To do so, a well-powered and well-designed randomized controlled trial is needed to compare the efficacy and costs of using dd-cfDNA as a prognostic assessment of acute rejection (for both antibody-mediated rejection and acute T cell–mediated rejection) with early treatment guided by the findings of dd-cfDNA versus standard posttransplant care using serum creatinine measures. In this ideal trial, subgroups such as recipients of different immunological risks, sexes, and ethnic populations, should be predefined and assessed. An economic evaluation alongside this trial to estimate the incremental cost–benefits ratio is essential to inform reimbursement and policy decisions and advise regulatory authorities about the cost-effectiveness of this test if this is implemented in the wider transplant populations, particularly in low-resource settings. As with all clinical trials, consumer engagement is invaluable throughout the entire research cycle and should involve operational, organizational, intellectual and leadership roles across all stages from priority setting and design to conduct, dissemination, and implementation.10 Although the transplant community eagerly waits for more evidence to address the many unresolved questions about dd-cfDNA, the findings of this study affirm the desire to adopt a standardized and evidence-based approach to posttransplant evaluation and monitoring of graft health.