Real‐world safety and efficacy data of ipilimumab in Japanese radically unresectable malignant melanoma patients: A postmarketing surveillance

Abstract Treatment with immune checkpoint inhibitors has improved prognosis among patients with cutaneous melanoma, but there are still unmet medical needs in Japan, especially for mucosal melanoma and acral lentiginous melanoma (ALM) subtypes. Ipilimumab, a fully human monoclonal antibody that specifically blocks cytotoxic T‐lymphocyte‐associated antigen 4 and potentiates antitumor T‐cell response, was approved in Japan in 2015 for the treatment of radically unresectable malignant melanoma. This postmarketing surveillance (prospective, non‐interventional, multicenter, observational study) evaluated the safety (occurrence of adverse drug reactions [ADR]) and efficacy (overall survival [OS]) of ipilimumab in a real‐world setting in Japan. All patients with radically unresectable malignant melanoma undergoing treatment with ipilimumab in Japan during the registration period between August 2015 and February 2017 were enrolled. In total, 547 patients were analyzed; 67.5% were 60 years old or more, 85.7% had an Eastern Cooperative Oncology Group performance status of 0–1, 50.3% had melanoma of the skin (mainly of the ALM subtype) and 73.5% had negative BRAF mutation status. Most patients had experienced recurrence and received multiple treatments. The overall incidence of ADR and serious ADR was 69.5% and 40.8%, respectively. The most common ADR and serious ADR were liver disorder, colitis and diarrhea. The most common ADR of special interest were liver‐related ADR (22.5%), skin‐related ADR (22.1%), gastrointestinal‐related ADR (20.3%) and endocrine system‐related ADR (16.3%). Most of these events had recovered or were in remission by the last evaluation. The median OS was 7.52 months (95% confidence interval, 6.47–8.74). Median OS was 6.31 and 8.44 months in patients with mucosal melanoma and melanoma of the skin; 9.43 and 3.75 months in patients with and without ADR; and 10.32 and 6.11 months in patients with and without serious ADR, respectively. Ipilimumab was tolerable and showed efficacy in improving OS for these patients.


INTRODUCTION
While malignant melanoma is a common skin cancer among Caucasians, it is considered a rare type of skin cancer for non-Caucasians. 1 The incidence of malignant melanoma in Japan in 2011-2013 was 1.75/100 000 persons per year. 2 The number of Japanese patients with skin cancer reportedly increases rapidly between the ages of 40 and 49 years and peaks at the age of 60 years. 3 Consequently, with the continuous growth of the aging population in Japan, the mortality rate associated with skin cancer is also increasing. 2 Notably, approximately 40% of the deaths caused by skin cancer in Japan are attributable to malignant melanoma. 4 There are remarkable differences between malignant melanoma in Japanese populations and malignant melanoma affecting Caucasians. Some major differences are the overall frequency of cases (more frequent in Caucasian than Japanese populations), 2 proportions of tumor subtypes (higher frequency of acral lentiginous melanoma [ALM] and mucosal melanoma compared with Caucasians), types of mutations (lower proportions of BRAF mutations compared with Caucasians) and tumor site (sole of the foot in Japanese patients compared with the trunk in Caucasian patients). [4][5][6] Prior to the introduction of immune checkpoint inhibitors, the prognosis of patients with melanoma was poor. 2,3 Although advances in treatment with immune checkpoint inhibitors have resulted in improved prognosis among patients with cutaneous melanoma, there are still unmet treatment needs in Japan, particularly for mucosal melanoma and ALM subtypes. 7 Ipilimumab is a fully human monoclonal antibody of the immunoglobulin (Ig)G1 isotype that specifically binds to anticytotoxic T-lymphocyte-associated antigen 4 (CTLA-4) and augments the antitumor response. 8 Improvements in overall survival (OS) were observed in the global phase III study among previously treated patients with metastatic melanoma treated with a total of four doses of ipilimumab 3 mg/kg, every 3 weeks. 9 Thus, ipilimumab was approved for melanoma as monotherapy (3 mg/kg, every 3 weeks for four doses) by the US Food and Drug Administration and the European Medicines Agency in 2011. In Japan, ipilimumab was approved in 2015 for the treatment of radically unresectable melanoma patients based on the results of the global phase III study 9 and a Japanese phase II study. 10 In the Japanese phase II study, 10 the best overall response rate (ORR) was 10% (95% confidence interval [CI], 1.2-31.7), median OS was 8.71 months (95% CI, 3.71-not reached) and median progression-free survival (PFS) was 2.74 months (95% CI, 1.25-2.83). Twelve patients (60%) had at least one drug-related adverse event (AE), and 12 patients (60%) reported immune-related adverse events (irAE).
As there were limited data on the safety and efficacy of ipilimumab among Japanese patients with radically unresectable melanoma, the Japan Ministry of Health requested the marketing authorization holder (Bristol-Myers Squibb, Tokyo, Japan) to conduct a postmarketing surveillance (PMS) to provide data on ipilimumab use for the approved indication in a real-world setting. The primary objectives of this postmarketing surveillance were to evaluate safety in terms of the occurrence of adverse drug reactions (ADR) and ADR of special interest (ADRI), assess the efficacy of ipilimumab based on OS, and identify factors that may affect the safety and efficacy of ipilimumab for Japanese patients with radically unresectable malignant melanoma in a real-world setting based on the conditions of its approval.

Study design, patients and treatment
This was a prospective, non-interventional, non-controlled, multicenter (146 institutions), observational study (all-case postmarketing surveillance; ClinicalTrials.gov Identifier, NCT02717364). The registration period of all Japanese patients with radically unresectable malignant melanoma was from August 2015 to February 2017 and the survey implementation period was from August 2015 to January 2019.
The study was conducted in accordance with Japanese regulatory requirements stipulated in Good Post-marketing Study Practice, 11 and approval from an ethics committee and written informed consent from the patients were not mandated as per the ministerial ordinance. Patients who had received at least one dose of ipilimumab were enrolled in the study by their treating physician, and each patient was followed up for 12 months. All patients with radically unresectable malignant melanoma treated with ipilimumab during the registration period were included in this postmarketing surveillance. There were no prespecified exclusion criteria.
This was a non-interventional study; thus, ipilimumab treatment was prescribed by the treating physicians under routine, daily practice, in compliance with the recommendations in the Japanese prescribing information. 12 The approved ipilimumab dose was 3 mg/kg of bodyweight administrated i.v. every 3 weeks for a total of four doses as a monotherapy. Treating physicians made treatment-related decisions such as initiation, duration and discontinuation of treatment. If the treatment was not completed with four doses of ipilimumab, the patient was considered to have discontinued treatment.

Data collection
The data collected in the case report forms (CRF) included patient demographics, clinical characteristics such as Eastern Cooperative Oncology Group performance status (ECOG PS), onset and recurrence, date of diagnosis, primary site, subtype, disease stage at baseline, M category, location of metastases, BRAF mutation status, prior therapy, complications, medical history, history of allergies, previous history of treatment for primary disease, hospitalization status, health-care data related to ipilimumab treatment (date of administration, dose administrated, reasons for discontinuation and concomitant medications), laboratory test results (if available) and other examinations or procedures that patients underwent during the observation period.

Safety
The safety of ipilimumab was evaluated based on the occurrence of ADR (AE) during 12 months after the initiation of ipilimumab treatment. An AE is any untoward medical occurrence in a patient or clinical investigation subject administrated a pharmaceutical product that does not necessarily have to have a causal relationship with this treatment. A serious AE is any untoward medical occurrence that, at any dose, results in death, is life-threatening, requires inpatient hospitalization or causes prolongation of existing hospitalization, results in persistent or significant disability/incapacity, is a congenital anomaly/birth defect or is an important medical event. Non-serious AE data were collected and confirmed for 6 months from the first dose of ipilimumab while only serious AE data were collected for 12 months of ipilimumab administration.
For AE and ADR, Preferred Terms (PT) and System Organ Class (SOC) terminology from the Medical Dictionary for Regulatory Activities, Japanese version (MedDRA/J version 21.1) were used. The seriousness of AE and ADR was determined by treating physicians based on the evaluation as per ICH E2A and E2D guidelines. The ADRI were gastrointestinal-related ADR (diarrhea, colitis and gastrointestinal perforation), liver-related ADR, skin-related ADR, endocrine system-related ADR (hypophysitis, hypopituitarism, hypothyroidism and adrenal Excluded from the safety analysis set n = 7 Received ipilimumab treatment before launch of ipilimumab n = 4 No information before the patient was transferred from another institution where ipilimumab was used No treatment with ipilimumab n = 1 No information after the patient was transferred to another institution n = 2 Exclusion from the efficacy analysis set n = 0

Efficacy
Survival at 12 months after the start of ipilimumab was regarded as effective, and the overall survival was estimated using the Kaplan-Meier method.

Statistical analysis
The planned survey sample size was 400 patients, based on an estimated number of patients who would receive ipilimumab within 2 years of product launch. The 400-patient target was considered to allow for the detection, with a probability of 95% or higher, of at least one event of the rarest AE, which were AE of special interest. For safety, patients who received at least one dose of ipilimumab were included in the analysis. For efficacy, patients who received at least one dose of ipilimumab for the approved indication were included in the analysis (off-label use would be excluded).
All statistical analyses were conducted using SAS version 9.4 (SAS Institute, Cary, NC, USA). Descriptive statistics including frequency distributions, means and standard deviations were used to analyze the data. The OS of patients who received treatment with ipilimumab was estimated using the Kaplan-Meier method. The statistical analysis method used

Patient disposition and characteristics
Of the 578 registered patients, 22 had invalid registration forms. Of the 556 eligible patients, CRF could not be collected for two patients; thus, 554 CRF were collected. Of these, seven patients were excluded from the safety and efficacy analysis sets; thus, 547 patients were analyzed for safety and efficacy ( Fig. 1). Table 1 summarizes the patient demographics and clinical characteristics at baseline. In the safety analysis set, approximately half of the patients were male (51.9%). Most patients were between 60 and 80 years of age (58.7%). The median (range) age was 66 (17-94) years. Most patients either had an ECOG PS of 0 (55.6%) or 1 (30.2%). The proportions of patients with ECOG PS 2, 3 and 4 were 7.9%, 4.4% and 1.8%, respectively. Nearly two-thirds (64.7%) had experienced disease recurrence, and 50.3% had melanoma of the skin, followed by mucosal melanoma (32.9%) as the primary tumor site. Among those with melanoma of the skin, the most common tumor subtypes were ALM (19.6%), nodular melanoma (12.6%) and superficial spreading melanoma (7.5%). 3 (1.0) † When more than one reason for discontinuation of the administration of ipilimumab was provided, the data were tabulated for each reason for discontinuation.

Dosing status
The mean (AE standard deviation) number of total doses was 2.9 AE 1.2 and the median (range) was 3.0 doses (1-5) (Fig. 2a). The percentage of patients who completed four doses of ipilimumab was 44.6% (244/547 patients). One patient received five doses of ipilimumab. This patient experienced AE after the first dose and discontinued the administration of ipilimumab. Subsequently, the patient was transferred to another hospital and received four doses at that hospital, without presenting new ADR after the fifth dose. Figure 2(b) shows the number of ipilimumab doses received with respect to previous treatments (i.e. first-, second-, third-line or further lines of treatment). Among patients with no prior treatment, 66.7% received four doses of ipilimumab, while among those with one and two or more previous treatments, 42.9% and 39.4% received four doses of ipilimumab, respectively.
The main reasons for treatment discontinuation by the number of ipilimumab doses received at the time of discontinuation are summarized in Table 2. With the first dose (n = 100), the most common reason for discontinuation was AE (40.0%), followed by disease progression (31.0%) and death due to primary disease (27.0%). Similarly, with the second (n = 115) and third doses (n = 87), the most common reasons for discontinuation were AE (53.0% and 66.7%, respectively) and disease progression (36.5% and 29.9%, respectively).
Of the 152 liver-related ADR, 88 were recovered, 40 were in remission and 23 were not recovered at the time of assessment. One death due to liver disorder was reported (Table 5b). Of the 137 skin-related ADR, 76 were recovered, 42 were in remission and 15 were not recovered at the time of assessment (of these, five cases were leukoderma and two with rash). No deaths related to skin-related ADR occurred (Table 6b) Endocrine systemrelated ADRs (Hypophysitis, hypopituitarism, hypothyroidism, and adrenal insufficiency) Figure 4. Incidence of adverse drug reaction at each dose. Each priority survey item has a duplicate count. When the same event occurred multiple times in the same case, it was counted using the first event. ADR, adverse drug reaction.
the 129 gastrointestinal-related ADR, 82 were recovered, 40 were in remission and three were not recovered at the time of assessment (two were diarrhea and one was colitis). One death due to intestinal perforation was reported (Table 7b). Of the 124 endocrine system-related ADR, 37 were recovered, 43 were in remission and 39 were not recovered at the time of assessment (of these, 12, eight, nine and six were hypothyroidism, hypopituitarism, adrenal insufficiency and hypophysitis, respectively) (Table 8b).

Efficacy
Among the 547 patients included in the efficacy analysis, treatment outcomes were death in 299 patients (54.7%), survival in 201 patients (36.7%) and outcomes unknown in 47 patients (8.6%) at 12 months after the start of ipilimumab.
The OS estimated by the Kaplan-Meier method is shown in Figure 5(a). The median OS was 7.52 months (95% CI, 6.4-8.74). Patients with a PS of 0 and 1 had longer OS than those with a PS of 2, 3 or 4 (Fig. 5b). Patients without prior drug treatment had longer OS than those who had received prior treatment (Fig. 5c). By primary tumor site, patients with tumors located in the mucosa had a shorter OS than those with skin or ocular tumors (Fig. 5d). The median OS was 7.85 months (95% CI, 5.39-not reached [NR]) in nodular melanoma, 7.16 months (95% CI, 4.99-10.32) in acral lentiginous melanoma, and NR in lentigo maligna melanoma (95% CI, 5.72-NR) and superficial spreading melanoma (95% CI, 6.24-NR) (Fig. 5e). There were no statistically significant differences among melanoma subtypes. OS was longer with the presence of ADR (Fig. 5f) and serious ADR (Fig. 5g).

DISCUSSION
As there were limited data on the safety and efficacy of ipilimumab in Japanese patients, the regulatory entity requested the conduct of this postmarketing surveillance of ipilimumab for Japanese patients with radically unresectable malignant melanoma in the real-world clinical setting. In the studied population, most patients were more than 60 years of age, had an ECOG PS of 0 or 1, half had melanoma of the skin (mainly of the ALM subtype) and approximately three-quarters had negative BRAF mutation status. Most patients had received multiple treatments as well as several rounds of pharmacotherapy with nivolumab, dacarbazine or vemurafenib, including third-line ipilimumab in nearly half of the patients, and most had experienced recurrence. The characteristics of the Japanese - † Hepatitis, hepatobiliary disease, blood bilirubin increased (one for each). Multiple events in the same case were tabulated for each event. ADR, adverse drug reaction; ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; c-GT, c-glutamyl transferase.

842
© Regarding the safety results, the overall incidence rate of ADR was 69.5%, and that of serious ADR was 40.8%. The most common ADR and serious ADR were liver disorder, colitis and diarrhea. The most common ADRI were liver-related ADR (22.5%), skin-related ADR (22.1%), gastrointestinal-related ADR (20.3%) and endocrine system-related ADR (16.3%). The present incidence rates seemed slightly higher than the rates observed in the Japanese phase II study in which the incidence of ADR was 60% (12/20 patients). 10 Meanwhile, the present incidence rates were lower than the ADR incidence rate observed in the global phase III study in which the ADR incidence rate was 80.2% (105/131). 9 As the Japanese phase II study included only 20 patients, 10 we consider that the ADR incidences observed in the present postmarketing surveillance results are within the expected range for a study of this nature with a considerably larger sample size. Accordingly, no major changes were noted in the risk-benefit profile of ipilimumab and no new safety concerns were raised; thus, the safety profile of ipilimumab was further validated. Further, the ADR -- † Alopecia areata, dermatitis exfoliative generalized, erythema nodosum, papule, pemphigoid, prurigo, rash macular, skin discoloration, achromotrichia acquired, papuloerythroderma of Ofuji (one for each). Multiple events in the same case were tabulated for each event. ADR, adverse drug reaction.

843
© 2020 The Authors. The Journal of Dermatology published by John Wiley & Sons Australia, Ltd on behalf of Japanese Dermatological Association observed were mostly irAE; this was expected based on the mechanism of action of ipilimumab.
The most frequently reported ADRI occurred within 8 weeks, but endocrine system-related ADR tended to occur at 8 weeks or later. Moreover, many of these endocrine system-related ADR occurred after the fourth dose, while most of the other ADR were found to have occurred before the fourth dose when analyzing the incidence of ADR at each dose (Fig. 4). This relatively later occurrence of endocrine system-related ADR may be a unique characteristic of ipilimumab compared with other programmed cell death 1/programmed death ligand 1 inhibitors, such as nivolumab. 13 In the present study, the most common ADRI with more than 2%, which also tended to be serious ADR, were liver disorder, hepatic function abnormal, and aspartate aminotransferase and alanine aminotransferase increased; rash, pruritus and skin disorder; diarrhea, colitis and enterocolitis; and hypothyroidism, hypopituitarism, adrenal insufficiency and hypophysitis. Notably, the vast majority of these events had recovered or were in remission by the time of the last evaluation.
Regarding ipilimumab efficacy, among the 547 patients analyzed, 201 survived (36.7%) at 12 months after the start of the treatment. The median OS estimated with the Kaplan-Meier method was 7.52 months. Mucosal melanoma seems to be less sensitive to ipilimumab compared with melanoma of the skin. This trend differed from that observed in the nivolumab postmarketing surveillance, 14 in which mucosal melanoma and melanoma of the skin had a similar sensitivity to treatment. These differences could be explained by differences in mutation patterns and these patients could perhaps benefit from nivolumab and ipilimumab combination therapy. 15,16 Overall, the OS observed in the present postmarketing surveillance (7.52 months) was similar to that reported in the Japanese phase II study evaluating ipilimumab monotherapy (median OS, 8.71 months) 10 and other overseas real-world data (median OS, 6.1-7.2 months); [17][18][19] however, it was slightly shorter than that in the global phase III study (median OS, 10.1 months). 9 The clinical studies enrolled patients with PS 0-1. In contrast, patients with PS 0-1 accounted for 85.7% of the patients in this postmarketing surveillance, while 14.1% of patients had a PS of 2 or higher.
Patients in later lines of treatment seem to be less sensitive to ipilimumab than first-line or treatment-na€ ıve patients. A retrospective study analyzed OS in treatment-na€ ıve and previously treated patients who were treated with ipilimumab 10 mg/kg every 3 weeks for four doses. Notably, treatment-na€ ıve patients achieved better survival rates (69.4%, 62.9% and 56.9% at 12, 18 and 24 months, respectively) than previously treated patients (50.0%, 37.7% and 28.5% at 12, 18 and 24 months, respectively). 20 Patients with ADR and serious ADR showed higher efficacy than patients without ADR. Based on recent reports on nivolumab, another immune checkpoint inhibitor, it has been suggested that the onset of immune-related ADR of immune checkpoint inhibitors is associated with clinical benefit for these patients. [21][22][23] Our findings support this hypothesis.
In this postmarketing surveillance, 44.6% of patients received four doses of ipilimumab, and patients received a median of three doses overall. The number of ipilimumab Multiple events in the same case were tabulated for each event. ADR, adverse drug reaction.

844
© 2020 The Authors. The Journal of Dermatology published by John Wiley & Sons Australia, Ltd on behalf of Japanese Dermatological Association doses seems to be different in patients receiving ipilimumab as first-and later-line treatment. In the present study, 66.7% of patients received four doses and had no prior treatments while the proportion of those with one and two or more previous treatments who received four doses of ipilimumab was 42.9% and 39.4%. The main reason for discontinuation was ADR (52.6%), followed by disease progression or onset of new lesion (32.8%), and death due to the primary disease (13.6%). Ipilimumab is used to treat patients at later stages of the disease, at which point, most patients will have received several lines of treatment; this could affect the efficacy of ipilimumab. In this postmarketing surveillance, 78.2% of patients had been treated with nivolumab before ipilimumab. Ipilimumab seems to be used mainly as second-or later-line treatment in patients at later stages of the disease. The fact that nivolumab was approved in Japan earlier than ipilimumab may be one reason. Tsutsumida et al. 24 reported the actual situation of sequential treatment patterns of ipilimumab and nivolumab in Japan. They reported on 61 of 68 patients treated with nivolumab before ipilimumab, and seven of 68 patients treated with ipilimumab before nivolumab. Among patients who were switched from nivolumab to ipilimumab, the most common reason for switching treatment was progressive disease.
In a global phase III study (CheckMate 067), 25 treatment-na€ ıve patients with advanced melanoma were randomly assigned to receive nivolumab plus ipilimumab, nivolumab alone or ipilimumab alone. At 5 years, the OS in the nivolumab plus ipilimumab group was 52%; in the nivolumab group, OS was 44%; and in the ipilimumab group, OS was 26%. 26 These monotherapy groups showed improved results compared with previous studies, and this may have been caused by the treatment cross-over observed among the groups. In a multicenter, single-arm Japanese study, nivolumab plus ipilimumab also showed increasing survival of treatment-na€ ıve Japanese patients with unresectable stage III/IV or recurrent melanoma. 27 Based on these studies, the combination of nivolumab and ipilimumab has been approved in Japan since 2018. 25 In Japan, further studies are needed not only to establish the optimal treatment of advanced malignant melanoma in order to improve patient survival and prognosis but also to clarify the optimal treatment pattern of ipilimumab, nivolumab and ipilimumab plus nivolumab to yield the most benefits for these patients with fewer ADR.

Limitations
This study was a non-interventional observational study with no control group and data were collected by physicians in usual clinical practice. Therefore, physicians' implementation of data collection might have introduced bias, and the types of safety and effectiveness data were limited compared with interventional clinical trials. Possible misclassifications of events cannot be ruled out because events were assessed by treating physicians and were not confirmed by an independent adjudication committee.

Conclusions
Based on the incidence of ADR and serious ADR, ipilimumab was found to be tolerable and no new safety concerns were raised. Ipilimumab showed efficacy in improving the median OS of patients according to its condition of approval. Notably, patients receiving ipilimumab as later lines of treatment were less sensitive to ipilimumab. Additionally, those with mucosal melanoma seemed to be less sensitive than those with skin type melanoma, but relatively higher efficacy was observed in patients with ADR and serious ADR.