Kidney retransplantation after anti-programmed cell death-1 (PD-1)-related allograft rejection

Abstract

In this report, we describe the first kidney retransplantation performed after anti-programmed cell death-1 (PD-1)-related allograft rejection. In 2014, we administered pembrolizumab (anti-PD-1) for ~9 months to a 57-year-old kidney transplant recipient with metastatic cutaneous squamous cell carcinoma (CSCC). The patient experienced both a complete anti-tumor response and T cell-mediated allograft rejection requiring reinitiation of hemodialysis. Four-and-a-half years after initiating pembrolizumab, the patient remained without evidence of CSCC relapse and received a kidney transplant from a living unrelated donor. Ten-and-a-half months after kidney retransplantation, the allograft is functioning well and the patient’s CSCC remains in remission. This case illustrates the potential for PD-1 blockade to bring about durable immune-mediated tumor control in chronically-immunosuppressed patients, and begins to address the feasibility of kidney retransplantation in patients who have previously received immune checkpoint inhibitor therapy for cancer. Results from this and future cases may help elucidate mechanisms of anti-tumor immunity and allograft tolerance, and inform updates to transplant decision models. Our report also underscores the need for clinical trials testing novel immunotherapy combinations in solid organ transplant recipients designed to uncouple anti-tumor and anti-allograft immunity.

Introduction

Over the past several years, reports describing the administration of anti-programmed cell death-1 (PD-1) or other immune checkpoint inhibitors (ICIs) to solid organ transplant recipients (SOTRs) with advanced cancers have emerged with increasing frequency (1). However, there is a paucity of data about the durability of anti-tumor and anti-allograft immunity in this chronically-immunosuppressed patient population, which makes assessment of suitability for retransplantation challenging. Here, we describe the first reported kidney retransplantation performed after anti-PD-1-related allograft rejection. Our observations illustrate the potential for PD-1 blockade to trigger long-term cancer remissions in chronically-immunosuppressed patients and begin to inform ongoing discussions about suitability for retransplantation in SOTRs who have previously received immune checkpoint inhibitor therapy for cancer.

Case Report

In 1989, a 33-year-old woman with autosomal dominant polycystic kidney disease underwent kidney transplantation (Figure 1). Her allograft was from a deceased donor, though additional details such as donor human leucocyte antigen (HLA) type are no longer available. Her maintenance immunosuppression regimen administered over the decades that followed included cyclosporine, prednisone, and azathioprine or mycophenolic acid. About 7 years after transplantation, the patient began to develop cutaneous squamous cell carcinomas (CSCCs) involving the head and neck, trunk and extremities that were successfully addressed with various therapies, including excisions, liquid nitrogen applications, and topical imiquimod or 5-fluorouracil.

Figure 1:

Timeline illustrating the course of a 57-year-old kidney transplant recipient with metastatic cutaneous squamous cell carcinoma (CSCC) who received pembrolizumab (anti-PD-1) and experienced allograft rejection and a complete anti-tumor response to therapy. Four-and-a-half years after initiating pembrolizumab, the patient remained without evidence of CSCC relapse and underwent kidney retransplantation. Ten-and-a-half months later, the allograft is functioning well and the patient’s CSCC remains in remission.

In 2012 (22.5 years post-transplant; age 55), the patient underwent Mohs micrographic surgery for an invasive CSCC of the scalp and left posterior auricular area. Several months later, her disease recurred in the left posterior auricular and neck lymph nodes. After left neck lymphadenectomy revealed CSCC in 9 of 24 lymph nodes, she received adjuvant irradiation and cetuximab. She was maintained on prednisone 5 mg daily but cyclosporine was discontinued and sirolimus was added.

In 2014 (~25 years post-transplant, age 57), a PET/CT scan demonstrated bilateral pulmonary nodules, one of which was biopsied, demonstrating metastatic CSCC. Sirolimus was discontinued and the patient remained on prednisone monotherapy at 5mg po daily. Her kidney allograft function was unaffected.

She was treated first with cetuximab (epidermal growth factor receptor (EGFR) inhibitor), but experienced progressive disease. She then received trametinib (MEK inhibitor), which was administered because her tumor contained a loss-of-function mutation in NF1. Her disease progressed through that therapy as well.

Administration of anti-PD-1 was then considered based on her Eastern Cooperative Oncology Group (ECOG) performance status=2 (Ambulatory and capable of all selfcare but unable to carry out any work activities) and poor candidacy for cytotoxic chemotherapy. At that time, clinical data about the use of anti-PD-1 in SOTRs had not yet been published, but ex vivo analyses of CSCC in SOTRs demonstrated expression of PD-1 and its ligands, suggesting that PD-1 pathway blockade may be effective in this patient population (2). Indeed, subsequent immunohistochemical evaluation of our patient’s CSCC demonstrated focal PD-L1 expression on tumor cells and infiltrating immune cells, which was geographically associated with an activated CD8+ cytotoxic T cell response, including lymphocytes expressing PD-L2 and PD-1 (3). In September 2014, after a clear discussion of the risks of immune-mediated toxicities associated with anti-PD-1, including kidney allograft rejection, the patient began therapy with intravenous pembrolizumab (anti–PD-1), 2mg/kg of body weight administered every 3 weeks.

Two months after initiation of anti–PD-1, the patient experienced acute allograft rejection. Despite administration of high-dose corticosteroids, the patient’s transplanted kidney did not recover. She underwent surgical removal of the allograft due to hematuria. Histologic and immunohistochemical evaluation of the explanted kidney revealed severe acute and chronic T cell-mediated rejection with arteritis (Banff II) with chronic transplant glomerulopathy (3). Anti–PD-1 was continued for another ~7 months for a total duration of therapy of ~9 months. She experienced a complete response to therapy per RECIST v1.1 (4), including regression of lung and subcutaneous metastases.

Although the patient tolerated hemodialysis well for a few years, she began to feel increasingly “drained,” both physically and emotionally. She developed severe hyperparathyroidism with a serum intact parathyroid hormone level >1000 pg/mL (normal range: 10–65), requiring a subtotal parathyroidectomy. She began having difficulty maintaining her dry weight, at one point becoming so fluid overloaded that her left ventricular ejection fraction dropped to 25%. Aggressive ultrafiltration with dialysis resulted in recovery of her cardiac function. The thrice-weekly dialysis schedule made travelling difficult and began to have an increasingly negative impact on her day-to-day quality of life. About 4 years after initiation of pembrolizumab, she remained without evidence of CSCC relapse, and requested evaluation for retransplantation.

Pre-transplant evaluation included assessment of medical, surgical and psychosocial suitability. To avoid sensitization, the patient had been maintained on prednisone 5 mg daily since the prior transplant. She tested negative for HLA antibodies. An infectious disease workup was unremarkable, including hepatitis B and C, HIV, and tuberculosis. An exercise stress test and transthoracic echocardiogram were only remarkable for mild diastolic dysfunction. Pap smear, mammogram and colonoscopy showed no evidence of pre-cancerous or cancerous lesions. Routine laboratory studies revealed a low serum immunoglobulin G (IgG) level (382 mg/dL, range 700–1600) and a low absolute lymphocyte count (ALC, 680 cells/mL, range 1100–4800). Risk assessment for relapse of CSCC in the post-transplant period was challenging given the limited published experience describing SOTRs treated with anti-PD-1, but given her ongoing complete response to ICI ~3 years after discontinuation of pembrolizumab, she was deemed an acceptable candidate for retransplantation. We advised the patient that we planned to administer relatively low-dose immunosuppression post-transplant, and that close monitoring for cancer relapse would continue. The patient underwent a one-week tacrolimus challenge trial to assess drug tolerability, which she completed without incident.

After careful consideration, it was determined that low-dose lymphocyte depletion therapy for induction, coupled with low-dose maintenance immunosuppression would be safest and most effective. To allow for this, we recommended a zero, or at most, one DR mismatched donor kidney. In this setting, we anticipated utilizing donor-derived cell-free DNA (dd-cfDNA) (Allosure, CareDx, Brisbane, CA) testing for close monitoring of allograft function in the post-transplant period.

Four-and-a-half years after initiation of pembrolizumab, the patient remained without evidence of relapsed CSCC and underwent kidney transplantation from a living unrelated donor as part of a kidney-paired donation chain. No donor specific antibodies (DSA) were detected and the patient’s Panel Reactive Antibodies (cPRA) score was zero. The HLA match was a 2A, 2B, 1DR match and 1DR mismatch. The donor was seropositive for cytomegalovirus (CMV) and Epstein–Barr virus (EBV). The recipient was seronegative for CMV but seropositive for EBV. Because of her history of cancer, lymphopenia and hypogammaglobulinemia, she received reduced immunosuppression: anti-thymocyte globulin (rabbit) 3 mg/kg total body weight, tacrolimus (target trough level 6–9 ng/mL for 3 months, 3–7 ng/mL thereafter), mycophenolate mofetil (MMF, 250 mg twice daily, initially), and a rapid prednisone taper to 5 mg daily by post-operative day 7. She experienced immediate graft function with a nadir serum creatinine level of 0.8 mg/dL. Valganciclovir prophylaxis for CMV was prescribed.

At ~7.5 months post kidney retransplantation she required intravenous ceftriaxone for E. coli urosepsis and pyelonephritis. She was again found to be hypogammaglobulinemic (serum IgG = 331 mg/dL) and received 5 mg/kg of intravenous IgG. Valganciclovir was discontinued, as was MMF because of persistent lymphopenia (ALC <500).

At 10.5 months after retransplantation, she remains without evidence of allograft rejection or relapsed CSCC.

Dd-cfDNA levels, measured every 3 months, peaked at 0.36% in the first month post-transplantation, and dropped to below the limit of detection (0.15%) thereafter. Of note, a level <1% has an 85% negative predictive value for rejection. DSA, measured monthly, have been undetectable (5). Her ECOG performance status is 0 (fully active, able to carry on all pre-disease performance without restriction), and her weight has returned to her pre-dialysis weight. Her immunosuppressive regimen includes tacrolimus and prednisone. She continues to undergo surveillance imaging and clinical medical oncology and dermatology evaluations every ~3 months.

Discussion

This report describes the first kidney retransplantation performed after anti-PD-1-related allograft rejection. The clinical significance of this “proof-of-principle” case report rests primarily with the long-term impacts of ICIs, which are distinct from those associated with small molecule inhibitors or cytotoxic chemotherapy. Tumor regressions and immune-mediated adverse reactions (i.e., side effects) associated with ICIs can continue for years after cessation of therapy, suggesting an enduringly disinhibited immune system (6). Although these pharmacodynamic properties provide clinical benefit (i.e., durable anti-tumor responses), we were unsure whether our patient’s immune system might still have been primed to attack foreign kidney antigens as it had during the first rejection episode. Similarly, it was not clear whether reinitiation of transplant immunosuppression would negatively impact established anti-tumor immunity. Although longer-term follow-up from this and future patients will be needed to better understand mechanisms of anti-tumor and anti-allograft immunity, our group’s decision to pursue retransplantation in this patient was based on several factors.

First, despite improvements in the past decade, the risk of death for patients on dialysis remains higher than for kidney transplant recipients. Based on United States Renal Data System 2016 data, the adjusted mortality rate for patients on dialysis was 164/1000 patient-years compared to 29/1000 patient-years for transplant recipients (7).

Second, maturing data from clinical trials testing ICIs in the general population demonstrate that, for a considerable percentage of patients, durable (i.e., ≥5 years) anti-tumor immunity is triggered by blockade of PD-1 and its ligands (8–10). Although data specifically from patients with CSCC is not yet mature, survival curves from the aforementioned studies that included patients with other cancer types (e.g., melanoma, lung cancer) begin to plateau at ~3 years, which provided reassurance about risk of relapse in our patient, who was 4.5 years out from initiation of pembrolizumab. Although ongoing analyses of the long-term impact of ICIs on a patient’s immune landscape may someday reveal a salutary effect on the risk of retransplantation (e.g., non-depletion of regulatory T cell function (11)), optimism about ongoing tumor control and, therefore, retransplantation in this case was largely based on data from previous studies demonstrating maintenance of an anti-tumor response after discontinuation of ICI. As data from this case and others like it mature, we will learn more about two clinically-relevant issues: 1) whether anti-tumor immunity generated after administration of ICI can persist after re-initiation of immunosuppression in the setting of retransplantation, and 2) whether previous receipt of anti-PD-1 impacts the longevity of a kidney allograft.

There is a dearth of data regarding the optimal timing of retransplantation among SOTRs with previous malignancies. Published guidelines generally advise minimum waiting periods (e.g., 2–5 years) before listing a patient for transplantation, which vary by tumor type and differ among expert panels (12). Some recent studies have shown lower risks of disease recurrence post-transplant than previously described (13). For example, Bavinck and colleagues reported that among 46 patients with non-melanoma skin cancers (NMSC) pre-transplant, there was no increased incidence of NMSC-related mortality post-transplant (14). Furthermore, requirements for longer waiting times between cancer remission and transplantation may not necessarily improve survival. Indeed, longer times on dialysis pre-transplant have been associated with poor outcomes such as increased cardiovascular mortality (15).

Although length of time from cancer treatment can be used to estimate risk of disease recurrence, it is not a direct measure of the presence or absence of cancer. Looking forward, just as dd-cfDNA levels are beginning to demonstrate utility as indicators of allograft rejection in SOTRs undergoing treatment with ICIs (16), so too might blood-based markers of neoplasia provide objective evidence of tumor activity – or lack thereof – in patients with treated cancers who are undergoing pre-transplant evaluation (17).

In summary, our case illustrates the potential for PD-1 blockade to bring about durable immune-mediated tumor control in chronically-immunosuppressed patients, and begins to inform transplant decision models and ongoing discussions about suitability for retransplantation in SOTRs who have previously received immune checkpoint inhibitor therapy for cancer (12,18). In the years to come, emerging survival data from this patient and others like her may reveal the impact of reinitiating transplant immunosuppression on established anti-tumor immunity, and the long-term effects of anti-PD-1 on a kidney allograft. Our findings also underscore the need for clinical trials testing novel immunotherapy combinations in SOTRs designed to uncouple – and elucidate mechanisms behind – anti-tumor and anti-allograft immunity [NCT03816332 and Ref. (19)].

Abbreviations

ALC

absolute lymphocyte count

CMV

cytomegalovirus

cPRA

Panel Reactive Antibodies

CSCC

cutaneous squamous cell carcinoma

dd-cfDNA

donor-derived cell-free DNA

DSA

Donor specific antibodies

EBV

Epstein - Barr virus

ECOG

Eastern Cooperative Oncology Group

EGFR

epidermal growth factor receptor

HLA

human leucocyte antigen

ICI

immune checkpoint inhibitors

IgG

immunoglobulin G

MMF

mycophenolate mofetil

NMSC

non-melanoma skin cancer

PD-1

programmed cell death-1

SOTR

solid organ transplant recipient