Malignant melanoma is an aggressive tumor originating from melanocytes. Recent developments in antibodies targeting the checkpoint molecules cytotoxic T-lymphocyte-associated protein 4, programmed cell death protein 1 (PD1), and programmed death ligand 1 (PD-L1) have notably improved the overall survival (OS) of patients with melanoma (1). The 5-year OS of patients with melanoma treated with nivolumab was reported to be 39%, and the median survival time was longer for patients with 5% or more PD-L1-positive tumor cells than for those who did not (2). The 5-year OS was 52%, 44%, and 26% in patients treated with immune checkpoint inhibitors (ICIs) nivolumab plus ipilimumab, nivolumab, and ipilimumab, respectively (3). Thus, ICIs were approved as a standard treatment for patients with metastatic melanoma, and they have shown remarkable curative effects in some patients; however, around 70% of patients show resistance to ICI therapy. Other than PD-L1 expression, several biomarkers, such as major histocompatibility complex (MHC) class I/II, tumor-infiltrating lymphocytes (TILs), tumor neoantigens, and the microbiota, have been suggest to be factors that might predict response to ICI therapy (4).

Herein, we report an autopsy case of metastatic melanoma that showed a slight beneficial response to ICI treatment. We evaluated several molecules related to melanoma antigens, human leukocyte antigen (HLA)/MHC, and the microenvironment in both primary and post-treatment (autopsy) specimens by means of immunohistochemistry.

A 77-year-old man with unresectable malignant melanoma of the left nasal cavity was referred to our hospital (Figure 1A). His medical history included hypertension, angina pectoris, and lacuna infarction. ¹⁸F-Fluorodeoxyglucose positron-emission tomography integrated with computed tomography revealed no metastatic lesion. Nivolumab (3 mg/kg every 2 weeks) was given as first-line therapy. After the 10^th infusion of nivolumab, the nasal tumor had increased in size in what turned out to be progressive disease. We administered ipilimumab (3 mg/kg every 3 weeks) as second-line therapy. After the third infusion of ipilimumab, the patient developed severe diarrhea and was diagnosed as having immune-related diarrhea and colitis, and cytomegalovirus enterocolitis. The detailed information on the cytomegalovirus enterocolitis of this patient has been reported previously (5). Stable disease was achieved with ipilimumab treatment but ipilimumab had to be discontinued due to severe adverse events. Thereafter, the patient refused any active treatment, and 1 year after discontinuation of treatment, he died from tumor metastasis at 23 months after the diagnosis of melanoma.

Pathological findings. Histologically, the biopsy sample showed proliferative melanoma cells along with sporadic melanin deposition (Figure 1B). Melanin deposition was confirmed by means of H₂O₂ treatment. The autopsy was performed 13 hours after death. Macroscopically, multiple metastatic nodules were seen in the liver and lungs, and a metastatic nodule was also seen in the spleen. The metastatic tumor nodules were elastic, hard, and whitish in color (Figure 2A). A small-sized old myocardial infarction and mild atherosclerosis were also detected; however, no other particular changes were detected. Microscopically, the recurrent tumors comprised melanoma cells with nuclear enlargement and a high nuclear/cytoplasm ratio, and notably, no melanin deposition was observed (Figure 2B).

Next, Immunohistochemistry was performed using primary (pre-treatment) and autopsy (post-ICI therapy/recurrent) specimens. The tumor cells of both the primary and recurrent tissue samples were positive for melanoma antigen recognized by T-cells (MART1) and protein S100; however, the expression of these melanoma-related antigens was slightly down-regulated in the post-ICI tumor (Figure 3). Notably, the level of HMB-45 antigen (Figure 3), which is a melanoma antigen that was strongly positive in the primary tumor, was significantly reduced in the post-ICI tumor. The levels of human leukocyte antigen (HLA) class I and class II (HLA-DR), which were strongly and weakly positive in the primary tumor, respectively, were significantly reduced in the post-ICI tumor (Figure 4). As shown in Figure 5, the primary tissue was partially and weakly positive for PD-L1, whereas the recurrent tumor tissue showed an even slighter level of PD-L1. PD-L1 expression was restricted to macrophages. Many cluster of differentiation (CD) 8-positive T-cells were seen in the primary tissue sample, but their numbers were significantly reduced in the post-ICI tissue sample. CD204-positive macrophages were detected in both the primary tissue and post-ICI tissue samples, although the density was lower in the post-ICI tissue sample. CD4-positive T-cells and FOXP3-positive regulatory T-cells were rarely seen in both the primary tissue and post-ICI tissue samples (data not shown).

Although ICI therapy augments antitumor immune responses in several kinds of solid tumors it can cause multiple adverse effects (immune-related adverse events), including neurological syndromes, endocrine disorders, dermatological symptoms, hepatitis, pneumonitis, myocarditis, and mucositis (6). The patient of the present case had polymyalgia rheumatica during ICI treatment. We previously reported the clonal expansion of cytotoxic T-cells in muscles that is potentially involved in myositis (7). Patients with immune-related adverse events were found to have a better beneficial effect from ICI therapy (8,9). Clonal expansion after ICI therapy was suggested to influence the partial response to ICI therapy in the present case.

In the present case, melanin deposition and HMB-45 clearly disappeared after ICI treatment, and the expression of MART1 and S100 was down-regulated after the ICI treatment. The loss of neoantigens is one mechanism of resistance to ICI therapy. In addition, the expression of MHC class I and II was also significantly reduced after ICI treatment. These indicated that the loss of neoantigens and MHC proteins contributed to immune escape and resistance to ICI therapy in the present case. Recently, the significance of MHC class II expression on tumor cells in ICI therapy was reported (10), and the helper and killer functions of CD4-positive T-lymphocytes in tumor tissues were suggested to influence antitumor immune responses (11,12). The expression of MHC class II on tumor cells might have been involved in the beneficial partial response to ICI therapy in the present case.

Based on the PD-L1 expression on tumor cells and the presence of TILs, it was suggested that is possible to classify melanoma tumors as PD-L1⁺/TILs^high (‘hot’ tumor), PD-L1⁻/TILs^low (‘cold’ tumor), PD-L1⁺/TILs^low, or PD-L1⁻/TILs^high, and that being ‘hot’ or ‘cold’ potentially predicted a response or no response, respectively, to ICI therapy (13). In the present case, the primary tumor was classified as PD-L1⁺/TILs^high (‘hot’ tumor), whereas the recurrent tumor was classified as PD-L1^–/TILs^low (‘cold’ tumor). This indicates that a phenotypical change between ‘hot’ and ‘cold’ tumors is a critical mechanism in ICI resistance.

In conclusion, we present an autopsy case of a patient with metastatic melanoma treated with ICI. A partial response was seen but the patient died 2 years after the diagnosis. Pathological analysis of the primary lesion and a recurrent (post-ICI therapy) lesion indicated that the loss of immunogenic antigens and the down-regulation of MHC class I and II were critical factors causing resistance to ICI.

The Authors have no conflicts of interest to declare.

YK and TM gathered the patient’s data and wrote the article. YK and SF were responsible for pathological diagnosis of this case. YK, HS, and TN carried out the immunohistochemistry. YK, TM, AM, and SF discussed the data and helped write the article. All Authors approved the final article.

The Authors thank Ms. Chiemi Shiotsu for their technical assistance. All procedures were performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. Informed consent was obtained from the patient included in this study.