Delayed immune-mediated hepatitis after three cycles of pembrolizumab for the treatment of sinonasal melanoma

Abstract

Immune checkpoint inhibitors (ICIs) are monoclonal antibodies that induce the anti-tumor effects of T cells by targeting co-inhibitory immune checkpoints. The development of ICIs has revolutionized the clinical practice of oncology, leading to significant improvements in outcomes; therefore, ICIs are now standard care for various types of solid cancers. Immune-related adverse events, the unique toxicity profiles of ICIs, usually develop 4–12 weeks after initiation of ICI treatment; however, some cases can occur >3 months after cessation of ICI treatment. To date, there have been limited reports about delayed immune-mediated hepatitis (IMH) and histopathologic findings. Herein, we present a case of delayed IMH that occurred 3 months after the last dose of pembrolizumab, including histopathologic findings of the liver. This case suggests that ongoing surveillance for immune-related adverse events is required, even after cessation of ICI treatment.

Introduction

Immune checkpoint inhibitors (ICIs) are monoclonal antibodies that induce the anti-tumor effects of T cells by targeting co-inhibitory immune checkpoints, such as programmed cell death protein 1 (PD-1) or cytotoxic T-lymphocyte antigen-4 (CTLA-4).[1] Immunotherapy with ICIs have significantly improved the outcomes of patients with solid tumors; thus, their use in (neo) adjuvant and metastatic settings of cancers has dramatically increased. Pembrolizumab, first approved in 2014, is a monoclonal antibody targeting PD-1. Nowadays, it is one of the most widely used ICIs for patients with various types of solid cancers, including non-small-cell lung cancer and malignant melanoma.

However, due to the unique mechanism of ICIs, which induce upregulation of immunity, unique adverse events, so-called immune-related adverse events (IRAEs), are key challenges for physicians.[1] In general, the median time to onset of IRAEs from the commencement of ICI treatment is known to be 8–12 weeks, and most cases occur during immunotherapy or <3 months from the last dose.[2]

However, a few cases occur >3 months after the last administration of immunotherapy.[3,4] In the literature, few cases of delayed immune-mediated hepatitis (IMH) and their histologic findings have been reported. Herein, we report a case of delayed IMH that occurred 3 months after cessation of pembrolizumab treatment in a patient with malignant melanoma and the associated histopathological findings in the liver.

Case Report

A male patient in his 40s visited the hospital because of blurred vision and periorbital pain in the right eye for two weeks on October 7, 2021. The patient had no past medical history. His visual acuity in the right eye was decreased to 20/40, and that in the left eye was 20/20. Formal visual field testing demonstrated a right-sided superior nasal and inferior field defect. Brain magnetic resonance imaging (MRI) revealed an enhancing, 2.6 cm soft tissue lesion on the right posterior ethmoid sinus that involved the medial extraconal space of the right orbit and optic nerve [Figure 1a]. The mass was confirmed to be malignant melanoma by biopsy. After evaluation, the patient was diagnosed with unresectable sinonasal malignant melanoma, and 200 mg of pembrolizumab was initiated on November 8, 2021. Unfortunately, disease progression was confirmed after three cycles of pembrolizumab [Figure 1b]. Hence, the treatment was switched to dacarbazine. After three cycles of dacarbazine, the primary lesion showed a partial response [Figure 1c].

Figure 1.

MRI scan of the neck: (a) at diagnosis (October 2021), T1-weighted axial selective partial inversion recovery image; (b) at observation of disease progression after three cycles of pembrolizumab (January 2022), T1 contrast enhanced fat-suppression axial image; and (c) at observation of a partial response after three cycles of dacarbazine (March 2022), T1-weighted contrast enhanced fat-suppression axial image

On April 15 (158 days after the initiation of pembrolizumab and 105 days after the last administration of pembrolizumab), the patient's laboratory findings demonstrated severe liver injury. There had been no recent medication changes, and the patient was asymptomatic; however, his total bilirubin (TB), aspartate aminotransferase (AST), and alanine aminotransferase (ALT) levels were 5.85 mg/dL [grade 3 toxicity by Common Terminology Criteria for Adverse Events (CTCAE), version 5.0; normal, 0.2–1.2 mg/dL], 107 U/L (grade 1 toxicity; normal, 13–33 U/L), and 192 U/L (grade 2 toxicity; normal, 8–42 U/L), respectively. Markers of coagulation function were normal.

We admitted the patient and stopped chemotherapy and other medication including pregabalin, rebamipide, naproxen, esomeprazole, and mosapride. However, on April 25, liver injury markers further increased with TB of 10.15 mg/dL (grade 3 toxicity) [Figure 2, Supplementary Table 1]. An abdominal CT scan revealed no structural abnormalities of the liver or biliary system. Cytomegalovirus immunoglobulin M, hepatitis A and C viral antibodies (Ab), hepatitis B surface antigen, anti-nuclear Ab, anti-mitochondrial Ab, anti-smooth muscle Ab, liver–kidney microsomal Ab, C-anti-neutrophil cytoplasmic antibodies (ANCA), and P-ANCA were all negative. For differential diagnosis, a liver biopsy was performed. It showed portal cellular infiltration, mostly by lymphocytes, and lobular inflammation with multiple spotty hepatocyte necrosis. In the portal tract, the bile duct showed reactive ductular proliferation, and the parenchyma showed mild hepatocellular and sinusoidal cholestasis [Figure 3].

Figure 2.

Graphical representation of the treatment course and laboratory findings. Abbreviations: AST, aspartate aminotransferase; ALT, alanine aminotransferase; TB, total bilirubin; PD, disease progression; PR, partial response

Figure 3.

Histopathologic findings of the liver: (a) inflammatory cell infiltration in portal tract with expansion (inside circle); (b) mixed inflammatory cell infiltration (mostly lymphocytes [white arrow] and some eosinophils [black arrow] and neutrophils [blue arrow]); (c) lobular inflammation (asterix) with intrahepatic cholestasis (blue arrow) and sinusoidal cholestasis (black arrow) (a: hematoxylin and eosin staining 100×, b: 200×, and c: 400×); (d) lobular injury with multiple areas of spotty necrosis of hepatocytes (black arrow, D-PAS staining, 200×)

Supplementary Table 1.

Laboratory findings of the patient

Dates	Events	AST (U/L)	ALT (U/L)	ALP (U/L)	TB (mg/dl)
2021-10-07		40	67	99.2	0.91
2021-10-28	Biopsy of para-nasal mass	41	47	96	0.66
2021-11-08	Pembrolizumab Cycle 1	25	28	95	0.35
2021-11-16		32	46	102	0.69
2021-11-30		84	140	96	0.93
2021-12-03		113	129	89	0.8
2021-12-10	Pembrolizumab Cycle 2	76	35	82	0.73
2021-12-31	Pembrolizumab Cycle 3	44	16	90	0.53
2022-01-21	DTIC Cycle 1	102	38	96	0.33
2022-01-28		31	23	101	0.39
2022-02-11	DTIC Cycle 2	43	20	97	0.33
2022-03-04	DTIC Cycle 3	36	28	89	0.57
2022-03-23	DTIC Cycle 4	33	25	83	0.98
2022-04-15		107	192	225	5.85
2022-04-16		148	243	239	7.78
2022-04-18	Admission	153	260	199	7.36
2022-04-20		135	211	173	6.95
2022-04-21		134	188	174	7.34
2022-04-22		120	151	165	7.34
2022-04-24		131	121	197	9.32
2022-04-25	Biopsy of liver	170	114	206	10.15
2022-04-26		199	113	177	8.75
2022-04-27		166	117	177	9.49
2022-04-28		145	112	183	10.76
2022-04-29	Initiation of steroid	92	88	172	9.64
2022-04-30		70	73	172	8.72
2022-05-01		49	56	159	7.94
2022-05-02		48	43	158	6.87
2022-05-10		58	29	168	5.75
2022-05-17		71	97	144	3.87
2022-05-24		47	54	116	2.43
2022-05-31		41	30	96	1.5
2022-06-07	DTIC Cycle 5	35	22	98	1.23
2022-06-14		29	38	86	0.9
2022-06-21		31	51	85	0.73

AST, aspartate aminotransferase; ALT, alanine aminotransferase; ALP, alkaline phosphatase; TB, total bilirubin; DTIC, Dacarbazine.

Based on clinical findings and biopsy results, a diagnosis of IMH was made. For treatment, 70 mg of intravenous methylprednisolone (1 mg/kg/day) was administered, and the patient's laboratory findings gradually improved with daily decreases in AST, ALT, and TB levels [Figure 2, Supplementary Table 1]. After 5 days, the patient was discharged with oral prednisolone at a dose of 60 mg per day. The corticosteroids were slowly tapered in an outpatient clinic, and the patient's laboratory findings normalized after 40 days of corticosteroid administration. The 5^th cycle of dacarbazine was administered on June 7 (40 days after initiation of corticosteroids), and the patient is currently being followed at an outpatient clinic.

Discussion

In this case, we reported delayed IMH that occurred 3 months after cessation of ICI treatment. Two previous reports presented cases similar to ours.[3,5] In these reports, IMH occurred after four doses and 5 months of pembrolizumab administration, respectively, whereas it occurred after three doses of pembrolizumab in our case. Our case suggests that relatively short courses of ICIs could induce delayed IRAEs. Although the incidence of delayed IRAEs is low,[2] it could significantly impact patient clinical outcomes. Thus, prompt recognition and treatment of IRAEs are essential, and clinicians should be aware that delayed IRAEs are possible in patients with histories of ICI treatment, regardless of the timeframe and ICI doses administered, even after ICI cessation.

Although the precise mechanisms of IRAEs have not yet been fully elucidated, T-cell-mediated toxicity is thought to be the main factor causing IRAEs. Inhibition of CTLA-4 or PD-1 by ICIs increases T-cell activation and suppresses regulatory T-cell functions. This results in the loss of peripheral tolerance for the patient's own tissues and results in a condition similar to an autoimmune disease.[1,6] However, the mechanisms of delayed IRAEs that occur after ICI cessation remain unclear. The aforementioned two case reports showed liver histopathologic findings of “delayed” IMH,[3,5] similar to ours, and those of IMH (non-delayed) due to pembrolizumab, which showed cellular infiltration, chiefly by lymphocytes.[5,7] These findings suggest that delayed IMH and IMH (non-delayed) are caused by similar mechanisms.

One possible pathophysiology of delayed IRAEs is consistent with ICI occupancy of PD-1 receptors on T cells.[8] The serum half-life of pembrolizumab is 12–20 days; however, pharmacodynamic testing showed that PD-1 receptor occupancy was sustained at >70% up to 2 months after a single dose and remained at 40% more than 8 months after three doses.[8] Thus, the remaining activated circulating T cells with occupied PD-L1 receptors could be related to delayed IRAEs, even several months after the cessation of ICIs. The similar histopathologic findings of delayed IMH and IMH (non-delayed) support this theory.

Still, little is known about the delayed and long-term toxicities of immunotherapy. Clinical trials of immunotherapy generally focus on the anti-tumor efficacy and acute side effects of those drugs, and the median follow-up duration of trials was relatively short, at 12 months.[9] Thus, late-onset toxicities might be under-reported.[9] Considering that immunotherapy has been started to be used in (neo) adjuvant settings, the clinical importance of late-onset or chronic toxicities of immunotherapy has become bigger. Real-world data, such as our case, will be helpful to understand the clinical manifestations of delayed IRAEs. Long-term follow-up data from clinical trials and real-world data will be needed for a better understanding of late-onset IRAEs.

Indeed, the diagnosis of IMH is a diagnosis of exclusion,[7] and routine use of liver biopsy for the diagnosis of IMH is controversial.[7,10] The liver biopsy could be helpful to exclude other differential diagnosis and be used to evaluate the severity of the liver injury; however, it is an invasive procedure, and a consensus regarding pathognomonic findings for IMH is not yet established.[7] Thus, liver biopsy is generally indicated (1) when the diagnosis is ambiguous with noninvasive methods and (2) when patients do not respond to initial empirical corticosteroids.[10]

Conclusion

As highlighted in our case, IRAEs can occur after cessation of immunotherapy, even after >3 months. Thus, close monitoring of patients who have received ICIs should be required, and physicians should be aware of the possibility of IRAEs. Furthermore, long-term follow-up of clinical trials will be warranted.

Declaration of patient consent

The authors certify that appropriate patient consent was obtained.

Financial support and sponsorship

This work was in part supported by the National Research Foundation of Korea (NRF) grant funded by the Korean Government (MSIT) (No. 2017R1A5A2015385).

Conflicts of interest

There are no conflicts of interest.