Introduction

Immune checkpoints are critical modulators in the immune system that either enhance a signal (co-stimulatory molecules) or attenuate a signal (co-inhibitory molecules). It is now clear that tumors modulate immune checkpoints as one of the mechanisms employed to escape immune surveillance [1]. The US Food and Drug Administration (FDA)-approved the Programmed Death 1 (PD-1) immune checkpoint pathway inhibitor, nivolumab, a fully human IgG4 monoclonal antibody indicated for the treatment of patients with unresectable or metastatic melanoma and squamous cell lung carcinoma [2]. However, immune checkpoint blockade has recently emerged as a promising therapeutic approach for various other malignancies including Hepatocellular Carcinoma (HCC) [3]. Interim results of an ongoing Phase I dose-escalation study involving nivolumab treatment of advanced HCC patients with or without chronic viral hepatitis demonstrated an estimated 62% survival rate in evaluable patients (n=47) at 12 months [3].

The clinical benefit of PD-1 inhibitors can be hampered by immune-related adverse events including immune-mediated hepatitis caused by dysregulation of the immune system. Hepatic toxicity often manifests as asymptomatic elevated levels of hepatic transaminases [4]. Moreover, PD-1 inhibitors were not tested within clinical trials in patients treated with long-term immunosuppressive medications. A recent publication by Lipson et al., [5] demonstrated that anti-PD-1 antibody (pembrolizumab) administration to a kidney transplant recipient with metastatic cutaneous squamous-cell carcinoma was associated with a robust antitumor response, but was accompanied by kidney allograft rejection [5]. Here, we report for the first time a case in which an anti-PD-1 antibody (nivulomab) was administered to a liver transplant recipient with recurrent HCC in lung.

Case

A 65-year-old man presenting primary sclerosing cholangitis and hepatocellular carcinoma (HCC) which was outside the Milan criteria (solitary tumor ≥8 cm), received a liver transplant from a deceased donor in December 2012 in China by patient selection. Thereafter, he received standard long-term immunosuppression therapy, including tacrolimus and certican. His graft function remains stable and normal. In November 2014, following detection of increased serum Alphafetoprotein (AFP) levels (from normal to 78.2micg/l), a lung metastasis, detected in the left upper lung by chest x-ray and CT, was diagnosed (Figure 1A). The lung biopsy collected under CT guidance was compatible with metastatic HCC. AFP staining was weakly positive (data not shown). PDL1 and PD1 were not expressed on tumor cells and infiltrating immune cells. There was no evidence of HCC recurrence in the grafted liver. Immunosuppression therapy dosage was decreased (tacrolimus) and treatment with sorafenib (a multikinase inhibitor) 800mg was initiated in January 2015. However, the lung metastasis progressed, as observed in repeated chest CTs, and mediastinal bilateral hilar adenopathy developed. In addition, serum AFP levels continued to increase, reaching 225micg/l. Due to disease progression, sorafenib treatment was discontinued.

The possibility and availability of administrating immune checkpoint-blocking drugs was discussed, in light of the published results of the Phase I dose-escalation study of nivolumab in advanced HCC patients [3]. The risk of immune-related toxic effects associated with anti-PD-1 therapy, including liver allograft rejection [4,5], was clearly explained to the patient. Although there was no endorsement for the administration of anti PD-1 by our local tumor board in the Sheba Medical Center, treatment with nivulomab was initiated in December 2015 in a private clinic, at a regimen of 80mg every 2 weeks. Following four cycles of treatment with nivulomab, serum AFP levels further increased to 1513micg/l, and a repeated chest CT clearly confirmed further disease progression, with new diffused metastatic bilateral lung nodules (Figure 1B). At the same time, that patient’s serum liver enzymes and bilirubin levels increased considerably from their baseline normal levels (Figure 2) (alanine aminotransferase (ALT) 428U/ l, aspartate aminotransferase (AST) 210U/ l, bilirubin 7.3mg/dl, alkaline phosphatase (ALP) 702U/ l , GGT 1419U/ l , INR 1.4, albumin 4.0g/dl). Serology and PCR testing for HCV, HBV, HAV, CMV, EBV and herpes viruses were all negative. In addition, immunoserology for antinuclear a, anti-smooth-muscle, and anti-mitochondrial antibodies proved negative. Repeated abdominal CT, MRI, and Doppler ultrasound examinations excluded hepatic artery thrombosis, CBD and bile ducts dilatation and there was no evidence of HCC recurrence in the grafted liver. A liver biopsy (Figure 3A) demonstrated severe panlobular hepatitis with foci of confluent necrosis and prominent perivenular infiltrate, hepatocanalicunar cholestasis, diffused sinusoidal and portal neutrophilic and lymphocytic infiltrates. Many foamy macrophages were noted. There was evidence neither of acute cellular rejection nor of humoral rejection. Immunohistochemistry staining for C4d was negative. Very few PD-1-positive T cells were noted (Figure 3B) however, membranous PD-L1 positivity was noted in hepatocytes (Figure 3C), endothelial cells and macrophages (Figure 3D). The infiltrating immune cells were predominantly CD8-positive findings consistent with an activated cytotoxic T-cells (Figure 3E). Evaluation of the patient’s serum for antibodies to HLA class I and II antigens was negative. Liver histology findings were compatible with reported immune-related toxic effects associated with anti-PD-1. Nivulomab administration was discontinued and prompt treatment with three cycles of Intravenous (IV) methyl prednisolone 500mg, followed by the administration of prednisolone 60 mg daily was initiated, but no improvement was achieved. At the beginning of March 2016, the patient was admitted to our hospital due to severe weakness and fatigue, dyspnea, and jaundice. Laboratory findings peaked at WBC 17K, Hb 13.2 g/dl, PLT 106K, AST 968U/ l , ALT1618U/ l , total bilirubin 41mg/dl (direct 29.8mg/dl) (Figure 2), ALP 877U/ l and GGT 806U/ l albumin 2.9g/dl, INR 3.46, creatinine 3.6 mg/ dl and urea 78 mg/dl. Multiple blood cultures tested sterile. Fluid resuscitation as well as the administration of IV methylprednisolone 125 mg daily and Cellcept 1000mg daily was not associated with any improvement, the patient continued to deteriorate rapidly and died on March 2016.

Discussion

Immune checkpoint blockade has recently emerged as a promising therapeutic approach for various malignancies, including HCC [1-3]. In a Phase I dose-escalation study of the efficacy of nivolumab in advanced HCC patients [3], an interim analysis demonstrated an estimated survival rate of 62% in evaluable patients (n=47) at 12 months. Durable partial responses and complete responses were detected in one out of five nivolumab-treated patients. It is of note that PD-1 and PD-L1were not expressed in the HCC lung metastatic tissue, however, focal PDL1 expression in some tumors may be missed in small biopsy specimens, such as needle biopsies [6]. Nivulomab administration in our patient was associated with further disease progression. Moreover, the patient developed grade 4 immune-mediated hepatitis. Checkpoint drug-related immune-mediated hepatitis, which manifest as asymptomatic increased liver function tests, mainly AST, ALT and GGT, rarely bilirubin, must be distinguished from other etiologies of hepatic injury, such as recurrent HCC in the grafted liver, viral infections and effects of other medications [4], which were all ruled out in our patient. Biopsy of the liver showed severe panlobular hepatitis with foci of confluent necrosis and no evidence of acute cellular or humoral rejection. The hepatic PDL1 expression in endothelial cells, hepatocytes and macrophages might indicate that the previous depression of T-cells caused by PD-1/PD-L1 coupling was disrupted by the administration of nivolumab, thereby releasing T cells and thus inducing immunemediated hepatitis. The median time to onset of hepatic drug-related adverse effects is highly variable, ranging from 25 weeks (range 4-31 weeks) in lung cancer patients, and 19 weeks (range 0.3-93 weeks) in patients treated with pembrolizumab, to 4 weeks (range 0.1-23 weeks) in melanoma patients treated with nivolumab [4]. The immune-mediated hepatitis in our patient developed within 7 weeks of nivulomab treatment. The frequency and severity of “adverse events of specific interest” with PD-1 inhibitors are 1-6% and mainly grade 1-2, respectively [4]. Severe liver adverse events (grade ≥3) occur in 0.2% of patients.

Histological findings of hepatitis related to ipilimumab, an immune checkpoint blocker targeting cytotoxic T-lymphocyte antigen 4 (CTLA-4), were recently described [7]. Samples often indicated panlobular active hepatitis and zone 3 hepatitis that resembles autoimmune hepatitis however, findings were not specific in several cases.

The standard algorithm for the management of hepatic checkpoint-related adverse effects involving grades 3-4 hepatotoxicity, includes the intravenous administration of highdose glucocorticoids for 24 to 48 h, followed by an oral steroid with prednisone 1-2 mg/kg, tapered over no less than 30 days. If serum transaminase levels do not decrease within 48 h after initiation of systemic steroids, oral mycophenolate mofetil (Cellcept) 500 mg every 12 h is considered. Severe cases require permanent discontinuation of PD1 inhibitor treatment due to hepatic drug-related adverse events and might require prolonged immunosuppressive therapy [4]. Hospitalization is considered, as cases of death from liver failure have been reported in patients with this level of hepatotoxicity [8]. Our patient did no recover from the grade 4 immune-mediated hepatitis.Lipson et al., reported, for the first time, on two kidney transplanted patients with metastatic melanoma receiving immune checkpoint blockade therapy. Renal allografts from both patients appeared to have been unaffected by administration of ipilimumab (anti-CTLA-4) [9]. Both patients experienced an anti-tumor response to therapy. Similarly, allograft tolerance remained intact in two liver-transplanted patients who received ipilimumab for metastatic melanoma [10,11]. One [11] experienced increased AST and ALT levels (grade 3) and no hyperbilirubinemia in week 16 of treatment which resolved later. There was no evidence of graft rejection. In a recent report by Alhamad et al., [12] the use of pembrolizumab (anti-PD-1 antibody) in kidney transplant recipient with metastatic melanoma was associated with biopsy-proven rejection. Although treatment with methylprednisolone was initiated, kidney function was not recovered and the patient ended up on hemodialysis. Lipson et al., [5] has also demonstrated that anti-PD-1 antibody (pembrolizumab) administration to a kidney transplant recipient with metastatic cutaneous squamous-cell carcinoma was associated with a robust antitumor response along with kidney allograft rejection. Despite administration of high-dose glucocorticoids, the patient’s transplanted kidney did not recover [5].

The administration of immune checkpoint inhibitors in organ transplant recipients might be associated with some adverse clinical impact. Immune checkpoint inhibitors require a competent T cell population to carry out their antitumor function, and coadministration of immunosuppressive agents to prevent graft rejection may impact the antitumoral efficacy of the immune checkpoint inhibitors. Our patient clearly experienced tumor progression while on nivolumab therapy. In addition, the ability of nivolumab to activate T cells specific for non-self antigens expressed by the allograft, might precipitate organ rejection. However, liver grafts are considered to be the least immunogenic organs for transplant and thus can sustain less aggressive immunosuppressive regimens; evidently, our patient did not develop acute cellular rejection. It is important to identify the patients who may tolerate a reduction of immunosuppression as well as the use of immunomodulatory agents; for instance, increased length of time from transplantation may serve as a predictive factor for tolerance of immunosuppression withdrawal. Due to the conflicting reports highlighted above, further studies in a large patient cohort are warranted to study the potential effects of immune checkpoint inhibitors in immunocompromised patients. The risks surrounding graft rejection in the setting of immune activation with these agents must be carefully considered when making treatment decisions in this population. Our case shows that the antitumoral effect of PD-1 inhibitors might be compromised in the context of long-term immunosuppression in a patient with HCC and that it can be associated with severe and fatal hepatitis. Similarly, in kidney-transplanted patients with metastatic melanoma, the administration of anti-PD-1 was associated with a robust antitumor response, yet, the kidney allografts were rejected and the patients ended up on hemodialysis [5,12]. In contrast, administration of anti-CTLA-4 to kidney-transplanted patients was associated with an anti-tumor response without an effect on the allografts [9-11], suggesting that the CTLA-4 pathway might have an advantage over PD-1 pathway in the transplant setting, although the number of cases reported is very small.

The transplant community must be aware of the potential risk of immune-mediated hepatitis in liver transplant recipients considering use of anti-checkpoint inhibitor antibodies. Close monitoring of liver function and moderate reduction of immunosuppression are warranted. The benefits of tumor regression versus the potential risk of allograft rejection and allograft loss should be weighed carefully for each individual patient. Additionally, the utility of the reduction of immunosuppressive therapy and its relative contribution to the overall antitumor effect should be investigated.

References

1. Pardoll DM (2012) The blockade of immune checkpoints in cancer immunotherapy. Nat Rev Cancer 12: 252-264.
2. http://www.accessdata.fda.gov/drugsatfda_docs/label/2015/125527s000lbl.
3. l-Khoueiry AB, et al. (2015) Phase I/II Opdivo (nivolumab) trial demonstrate anti-tumor activity in patients with hepatocellular carcinoma; ASCO 2015: LBA101.
4. Eigentler TK, Hassel JC, Berking C, Aberle J, Bachmann O, et al. (2016) Diagnosis, monitoring and management of immune-related adverse drug reactions of anti-PD-1 antibody therapy. Cancer Treat Rev 45: 7-18.
5. Lipson EJ, Bagnasco SM, Moore J Jr, Jang S, Patel MJ, et al. (2016) Tumor Regression and Allograft Rejection after Administration of Anti-PD-1. N Engl J Med 374: 896-898.
6. Topalian SL, Taube JM, Anders RA, Pardoll DM (2016) Mechanism-driven biomarkers to guide immune checkpoint blockade in cancer therapy. Nat Rev Cancer 16: 275-287.
7. Johncilla M, Misdraji J, Pratt DS, Agoston AT, Lauwers GY, et al. (2015) Ipilimumab-associated Hepatitis: Clinicopathologic Characterization in a Series of 11 Cases. Am J Surg Pathol 39: 1075-1084.
8. O’Day SJ, Maio M, Chiarion-Sileni V, et al. (2010) Efficacy and safety of ipilimumab monotherapy in patients with pretreated advanced melanoma: a multicenter single-arm phase II study. Ann Oncol 21: 1712-738.
9. Lipson EJ, Bodell MA, Kraus ES, Sharfman WH (2014) Successful administration of ipilimumab to two kidney transplantation patients with metastatic melanoma. Journal of Clinical Oncology 32: e69-71.
10. Ranganath HA, Panella TJ (2015) Administration of ipilimumab to a liver transplant recipient with unresectable metastatic melanoma. J Immunother 38: 211.
11. Morales RE, Shoushtari AN, Walsh MM, Grewal P, Lipson EJ, et al. (2015) Safety and efficacy of ipilimumab to treat advanced melanoma in the setting of liver transplantation. J Immunother Cancer 3: 22.
12. Alhamad T, Venkatachalam K, Linette GP, Brennan DC (2016) Checkpoint Inhibitors in Kidney Transplant Recipients and the Potential Risk of Rejection. Am J Transplant 16: 1332-1333.