Gene Expression Profiling in Kidney Transplant Recipients on Immune Checkpoint Inhibitors

In the decade since their development, immune checkpoint inhibitors have quickly become part of standard treatment for an expanding list of solid-organ and hematologic malignancies. Immune checkpoint inhibitors are mAbs that stimulate a T cell–mediated antitumor response by inhibiting “immune checkpoints”—receptor-mediated signaling pathways that provide negative regulatory control over T-cell immunity. Prime targets for immune checkpoint blockade are cytotoxic T lymphocyte–associated antigen 4 and programmed cell death protein-1, both of which are T-cell surface receptors. Binding of these receptors to cognate ligands found on antigen-presenting cells and parenchymal cells, including tumor cells, leads to inactivation of antigen-specific T cells. Thus, blockade of the cytotoxic T lymphocyte–associated antigen 4 and programmed cell death protein-1 axis by immune checkpoint inhibitors is an attractive strategy in cancer treatment, because tumor cells upregulate some of the ligands (such as programmed death ligand-1) to evade the immune system. Indeed, use of immune checkpoint inhibitors stimulates an inflammatory antitumor response that improves outcomes for advanced malignancies that have progressed on prior lines of treatment (1).

Despite improved outcomes, use of immune checkpoint inhibitors is limited by immune-related adverse events, which may affect multiple organs, including the kidneys. A variety of native kidney lesions have been reported, including, but not limited to, acute interstitial nephritis, minimal change disease, and FSGS (2). The incidence of AKI from immune checkpoint inhibitor use has been reported to be between 1% and 5%, with acute interstitial nephritis comprising >90% of cases (3). Similar data from kidney transplant recipients are lacking because immune checkpoint inhibitor trials have historically excluded these patients, despite their higher incidence of malignancy. The rationale for this exclusion has been the putative high risk of allograft rejection with immune checkpoint inhibitors that would take place with reactivation of allograft-specific T cells.

Some of these concerns have been borne out in reports detailing use of immune checkpoint inhibitors in kidney and other solid organ transplant recipients. A recent retrospective, multicenter study by Murakami et al. (4) describes an allograft rejection rate of 42%, with a median time to rejection of 24 days since first immune checkpoint inhibitor dose in 69 kidney transplant recipients. Importantly, 65% of patients with acute rejection experienced graft loss, despite treatment with corticosteroids and/or intravenous Ig, based on histologic features of T cell–mediated rejection versus antibody-mediated rejection. At the same time, compared with non–immune checkpoint inhibitor therapy, use of immune checkpoint inhibitors was associated with lower risk of disease progression in patients with squamous cell cancer and melanoma, highlighting the thoughtful deliberation that initiation of immune checkpoint inhibitor therapy in kidney transplant recipients requires.

A clinical dilemma that arises in the use of these drugs in the treatment of kidney transplant recipients is the difficulty in differentiating between immune checkpoint inhibitor–induced acute interstitial nephritis and acute T cell–mediated rejection on the kidney biopsy sample. Unlike acute interstitial nephritis associated with non–immune checkpoint inhibitor drugs, where an eosinophilic infiltrate is pathognomonic, immune checkpoint inhibitor–induced acute interstitial nephritis and T cell–mediated rejection Banff 1A both display predominantly lymphocytic infiltrates, making a definitive diagnosis challenging to establish. Establishing the diagnosis is important, however, because the treatment and prognosis may differ, with immune checkpoint inhibitor–associated acute interstitial nephritis having a more robust response to corticosteroids, and T cell–mediated rejection potentially requiring more potent immunosuppression.

In this issue of CJASN, Adam et al. (5) address this diagnostic conundrum by gene expression profiling to find unique molecular signatures for acute interstitial nephritis versus T cell–mediated rejection in kidney transplant recipients receiving immune checkpoint inhibitors. Specifically, they used the NanoString nCounter platform to analyze gene expression in archival kidney biopsy specimens. A variation on DNA microarray, this technology can identify RNA targets, including those pathways known to be significant in innate and adaptive immune responses. In this study, 725 immune-related genes were queried in discovery and validation cohorts. Of the 75 total biopsy samples, 25 formed the discovery cohort and included samples with non–immune checkpoint inhibitor T cell–mediated rejection and immune checkpoint inhibitor–associated acute interstitial nephritis. The remaining 50 samples formed the validation cohort and comprised immune checkpoint inhibitor–associated acute interstitial nephritis, T cell–mediated rejection, immune checkpoint inhibitor–associated crescentic GN, drug-induced acute interstitial nephritis, BK virus nephropathy, and normal kidneys. The initial analysis of the 39 genes known to be associated with T cell–mediated rejection noted no differential expression between the non–immune checkpoint inhibitor T cell–mediated rejection and immune checkpoint inhibitor–induced acute interstitial nephritis groups. After an expanded query of all 725 genes, one gene—interferon alpha-inducible protein 27—demonstrated significantly higher expression in the T cell–mediated rejection biopsy specimens in both the discovery and validation cohorts. Expanding the same analysis to the key clinical question of whether there is a different signature in immune checkpoint inhibitor–associated acute interstitial nephritis versus immune checkpoint inhibitor–associated T cell–mediated rejection, IFI27 again discriminated between the two by higher expression in the latter.

Lastly, the authors focused on 30 genes with differential expression in T cell–mediated rejection versus drug-associated acute interstitial nephritis, because these would potentially reveal a predominant hypersensitivity versus alloimmune fingerprint. Interestingly, there was significant overlap in gene signatures across all biopsy cohorts (drug-associated acute interstitial nephritis, immune checkpoint inhibitor–associated acute interstitial nephritis, T cell–mediated rejection, immune checkpoint inhibitor–associated T-cell rejection, and immune checkpoint inhibitor–associated GN) compared with normal kidney biopsy specimens, with the most overlap seen among the immune checkpoint inhibitor–associated acute interstitial nephritis, drug-associated acute interstitial nephritis, and immune checkpoint inhibitor–associated T cell–mediated rejection cohorts.

This study contains several points for discussion. It is the first description of an association between T cell–mediated rejection of kidney allografts and increased gene expression of IFI27, an IFN-α–dependent protein implicated in apoptosis and antiviral immunity and in several cancer models (6 –9). IFN-γ gene expression has previously been noted in association with cardiac antibody–mediated rejection (10), and recombinant IFN-α has been associated with acute kidney allograft rejection (11), indicating the IFN family of genes plays a role in alloimmunity as well. Further work needs to be done at the gene and protein level to support a significant role of IFI27 in T cell–mediated rejection.

The suggestion that IFI27 may be a biologic marker to differentiate immune checkpoint inhibitor–associated acute interstitial nephritis from immune checkpoint inhibitor–associated T cell–mediated rejection is intriguing, but the biopsy specimen sample size of five in the latter group is insufficient to draw any conclusions. For that, these initial data need to be validated in a larger, prospective cohort of kidney transplant recipients on immune checkpoint inhibitors. Furthermore, the clinical utility of checking IFI27 or other potential biomarkers remains unclear in patients who are at a complex juncture of both under- and over-immunosuppression. Both early T cell–mediated rejection and acute interstitial nephritis are treated with steroids with differing intensity and duration, but immune checkpoint inhibitors may have reduced efficacy with any increase in immunosuppression, making the clinical implications of any distinction less clear. At the same time, it may avoid over-treatment of late immune checkpoint inhibitor–associated T cell–mediated rejection, and the potential adverse effects of such treatment, if a distinction can be made beyond histology. It is also important to remember that patients on immune checkpoint inhibitors have often had tumor progression on prior therapy. Ultimately, the decision to use immune checkpoint inhibitors in transplant patients is complex and involves multidisciplinary discussions with the patient that include understanding that immune checkpoint inhibitor therapy may prolong survival even if it leads to allograft loss.

Clinically, we know that acute interstitial nephritis and T cell–mediated rejection due to immune checkpoint inhibitors have histopathologic overlap that make differentiation difficult. The novelty of this study is that it largely confirms the same on a molecular level as well. Apart from IFI27, it is remarkable that gene profiles were similar across all disease-affected biopsy specimens and that immune checkpoint inhibitor–associated T cell–mediated rejection had an expression profile closer to acute interstitial nephritis from immune checkpoint inhibitor or other drugs versus non–immune checkpoint inhibitor–associated T cell–mediated rejection. This may be due to the limited genes for analysis in the NanoScript platform or it may point to an under-appreciation for the role of T-cell hypersensitivity responses in T cell–mediated rejection. In a sense, that itself is a validation of the efforts in kidney disease to move beyond simple histologic diagnosis and seek other tools to assay pathophysiologic mechanisms to ask whether there is more than meets the eye when it comes to kidney disease.

Adam et al. (5) have used novel technology to decipher unique molecular signatures in kidney adverse events related to immune checkpoint inhibitor use in kidney transplant recipients. This population is expected to grow in the next several years as immune checkpoint inhibitors are established in clinical use. It will hopefully pave a roadmap for future studies evaluating the role of immune pathways in T cell–mediated rejection in all kidney transplant recipients, including those on immune checkpoint inhibitors.

Disclosures

A.C. Shirali reports serving as a member of the American Society of Nephrology and early program faculty for 2021 and having consultancy agreements with, and receiving honoraria from, OnViv. The remaining author has nothing to disclose.

Funding

None.