Real-world safety and effectiveness of nivolumab for advanced renal cell carcinoma in Japan: a post-marketing surveillance

Background This all-case post-marketing surveillance (PMS) evaluated the real-world safety and effectiveness of nivolumab monotherapy in Japanese patients with un-resectable or metastatic renal cell carcinoma (RCC). Methods This multicenter, open-label, non-interventional, observational PMS study (registered from August 2016 to January 2017) was conducted in patients who were newly initiated on nivolumab monotherapy. Assessments included treatment-related adverse events (TRAEs) of special interest, patient characteristics affecting safety, and effectiveness over 12 months. Results Overall, 580 patients were enrolled; 555 and 554 patients comprised the safety and effectiveness analysis sets, respectively. The median (range) age of the population was 66 (14–90) years. Nivolumab was initiated as 1st-, 2nd-, and ≥ 3rd-line treatment in 0.2%, 42.0%, and 57.8% of patients, respectively. TRAEs were reported in 275 (49.5%) patients. The most common TRAEs of special interest included thyroid dysfunction (9.5%), hepatic dysfunction (8.6%), and interstitial lung disease (6.7%). The incidence of TRAEs was significantly higher in elderly patients (≥ 65 vs < 65 years; ≥ 75 vs < 75 years); patients with lower C-reactive protein levels (< 5 vs ≥ 5 mg/dL); and patients with vs without a past medical history, including hepatic, thyroid, and autoimmune diseases. The 6- and 12-month survival rates were 71.8% and 57.9%, respectively. Conclusion The safety profile of nivolumab monotherapy in Japanese patients with advanced RCC was similar to that in the phase 3 CheckMate 025 trial. No new safety signals were observed in this study. Supplementary Information The online version contains supplementary material available at 10.1007/s10147-022-02155-3.


Introduction
Nivolumab is the world's first human anti-programmed cell death protein 1 monoclonal antibody that was approved in Japan in 2014 for patients with malignant melanoma and in 2016 for patients with renal cell carcinoma. Currently, nivolumab is approved for numerous cancers in more than 65 countries worldwide.
A multinational phase 3 study (CheckMate 025 trial) conducted in patients with un-resectable or metastatic renal cell carcinoma (RCC) in Japan and other countries demonstrated the superior efficacy of nivolumab over everolimus. Median overall survival (OS) as the primary endpoint was 25.0 months with nivolumab and 19.6 months with everolimus. Fewer grade 3 or 4 treatment-related adverse events (TRAEs) were reported with nivolumab than with everolimus [1]. A subsequent subgroup analysis showed that in Japanese patients from the CheckMate 025 trial, the efficacy 1 3 and safety of nivolumab versus everolimus were consistent with that reported in the global population [2]. Consequently, in August 2016, the manufacturing and marketing approval for nivolumab was updated to include the indications of un-resectable or metastatic RCC in Japan.
However, since the number of patients enrolled in the clinical trial was limited in Japan (37 patients with nivolumab) [2], an all-case post-marketing surveillance (PMS) study was mandated as a condition for approval. This is the first report of an all-case PMS for nivolumab in RCC conducted to evaluate the real-world safety, including TRAEs of special interest and risk factors of patient characteristics affecting safety, and effectiveness of nivolumab monotherapy in patients with un-resectable or metastatic RCC.

Study design
This PMS was a multicenter, open-label, non-interventional, observational study conducted at 237 centers across Japan using a central registration system. Each participating center signed a contract with the sponsor to undertake this surveillance. The study complies with the ministerial ordinance of Good Post-Marketing Study Practice in Japan [3]. Written informed consent of patients and Institutional Review Board approval were waived as these were not required for the PMS study. Japanese patients with un-resectable or metastatic RCC who were to be treated with nivolumab between August 26, 2016, and January 31, 2017, were registered and observed for 12 months after initiating treatment with nivolumab. Patients who discontinued treatment before completion of the 12-month observation period were monitored up to the end of the observation period as frequently as possible.

Patients
All patients with un-resectable or metastatic RCC who were newly initiated on nivolumab monotherapy for the prescribed indication (per prescribing information) were registered for this PMS study. Patients who had not been treated with systemic therapy before initiating nivolumab were excluded from the effectiveness analysis set.

Assessments
Adverse events (AEs) were recorded 12 months after the start of treatment, and their relation to nivolumab was judged by each attending physician. Incidence of TRAEs in patients with RCC, particularly that of events described as important identified risks in the drug risk management plan, was captured by this PMS study. We used the Japanese version of the Medical Dictionary for Regulatory Activities version 22.1 and the National Cancer Institute Common Terminology Criteria for Adverse Events version 4.0 for classifying and grading each TRAE, respectively. Effectiveness was evaluated as the OS rate at 6 and 12 months after the first nivolumab dose.

Statistical analysis
The incidence of TRAEs was characterized by patient background factors using the Fisher's exact test, Wilcoxon ranksum test, and chi-square test. Multivariate regression analyses evaluated the association of the occurrence of hepatic dysfunction, thyroid dysfunction, and interstitial lung disease (ILD) with various risk factors (multiple independent variables) using the Fine and Gray model [4].
For hepatic dysfunction, thyroid dysfunction, and ILD, we used the sub-distribution hazards model by Fine and Gray, in which the onset of AEs was defined as events and any death before the onset of adverse effects as competing risks. Cessation/termination was defined as discontinuation or the end of observation period other than the occurrence of the relevant adverse effect or death. The starting point of the survival period was defined as the initial date of nivolumab use. Statistical analyses of risk factors associated with AEs are presented in Online Resource 1. Survival rate could not be calculated by the Kaplan-Meier method, since data on the exact date of patients' death were not collected.

Patients
In total, 580 patients were enrolled between August 26, 2016, and January 31, 2017, from 237 facilities. After excluding patients not treated with nivolumab, 555 patients were included in the safety analysis set. The effectiveness analysis set comprised 554 patients after excluding one patient who was not treated with systemic therapy before initiating nivolumab (Fig. 1).
In the safety analysis set, median (range) age was 66 (14-90) years, with 14.8% of patients being ≥ 75 years; 23.1% of patients had a Karnofsky performance status (KPS) score of ≤ 70. The most common histological type of RCC was reported to be clear cell in 84.5% of patients, followed by papillary in 6.5% of patients. On the other hand, in the phase 3 CheckMate 025 trial, the median (range) age was 62 (23-88) years, with 8.3% of patients being ≥ 75 years; 5.9% of patients had a KPS score of ≤ 70, and all patients had clear cell RCC [1].
Incidences of TRAE of special interest in this PMS study and the CheckMate 025 trial are shown in Online Resource 3. Among TRAEs of special interest, the incidence of allgrade ILD, adrenal disorder, and cardiac disorder in this PMS population (6.7%, 2.5%, and 2.3%, respectively) was higher to an extent (1.0% or more) than that in the overall population of the CheckMate 025 trial (4.9%, 1.5%, and 1.2%, respectively). Of the 37 patients who developed ILD in this PMS, the outcome was recovered/improved in 29, recovered with sequelae in 1, not recovered in 5, and death in 2. Of the 14 patients who developed adrenal disorders (adrenal insufficiency, n = 13; secondary adrenocortical insufficiency, n = 1) in this PMS, the outcome was recovered/improved in 11, not recovered in 1, and unknown in 2. Of the 13 patients who developed cardiac disorders in this PMS, the outcome was recovered/improved in 9, death in 3, and unknown in 1.
In addition, 13.5% (5/37) of the patients who developed ILD had a history of ILD. Of the 14 patients who developed adrenal disorders, none had a history of adrenal disorders or adrenal metastases. Of the patients who developed cardiac disorders, 23.1% (3/13) had a history of cardiac disease (congestive heart failure, n = 2; heart failure, n = 1).

TRAEs categorized by patient background factors
The incidences of TRAEs in patients categorized by background factors are shown in Table 3 Table 3).
Multivariate regression analyses showed that the risk of hepatic dysfunction was significantly higher among patients with versus without a past medical history of hepatic disease (Online Resource 4). Further, these analyses showed Fig. 1 Patient disposition. CRF case report form that the risk of thyroid dysfunction was significantly higher among patients with versus without a past medical history of thyroid disease and among those aged ≥ 75 years versus < 75 years (Online Resource 5). Similarly, multivariate analyses showed that the risk of ILD was significantly higher among patients with versus without a past medical history of ILD and autoimmune diseases (Online Resource 6).

Discussion
This large PMS study clarified the real-world safety and effectiveness of nivolumab in Japanese patients with advanced RCC with results similar to those observed in the phase 3 CheckMate 025 trial and another retrospective realworld study [1,2,5].
Among TRAEs of special interest, the incidence of allgrade ILD, adrenal disorder, and cardiac disorder in this PMS population was higher to an extent than that in the overall population of the CheckMate 025 trial. In contrast, the incidence of all-grade ILD in the Japanese population of the CheckMate 025 trial (8.1%) was comparable to that in this PMS population (6.7%). The prevalence of druginduced pneumonia has been reported to be relatively high in the Japanese population [6], suggesting that the higher incidence of ILD in this PMS compared with that in the global population could be attributed to the influence of differences in patient background factors.
No patients who developed adrenal disorders had a history of adrenal disorders or adrenal metastases. The relationship between history of adrenal disorders and the occurrence of AEs has not been analyzed in this PMS study or in the phase 3 CheckMate 025 trial and the aforementioned other retrospective real-world study [1,2,5]. However, in most patients, the adrenal disorders recovered/improved during the study period, suggesting that these AEs could be managed appropriately according to existing AE management algorithms [7]. Nevertheless, since adrenal disorders are often exacerbated, subjective symptoms, such as malaise, consciousness disturbed, and nausea/vomiting, should be carefully monitored for early detection.
It is known that in general, cardiac disorders are more likely to occur with molecular targeted drugs. The incidence of cardiac disorders was 2.3% in this PMS study and 1.2% in the CheckMate 025 trial [1]. The eligibility criteria for the CheckMate 025 trial stipulated the inclusion of "Patients with one or two angiogenesis inhibitors (including, but not limited to, sunitinib, pazopanib, and axitinib) as treatment for advanced or metastatic RCC, but not more than three prior regimens, with or without prior treatment with cytokine therapy (e.g., interleukin-2 and interferonalpha), vaccine therapy, or cytotoxic anticancer drugs." As prior therapies in the CheckMate 025 trial, 72.0% of patients received one angiogenesis inhibitor and 28.0% of patients received two angiogenesis inhibitors [1]. On the other hand, in this PMS, where 57.8% of patients received nivolumab as ≥ 3rd-line treatment, prior therapies included mammalian target of rapamycin inhibitors or cytokine therapies in addition to angiogenesis inhibitors. Although direct comparison is difficult, it is likely that more patients in this PMS used more numbers of molecular targeted drugs compared with the patients in the CheckMate 025 trial [1]. In addition, 23.1% of patients who developed cardiac disorders had a history of cardiac disease. Based on the above, the higher incidence of cardiac disorders in this PMS may be attributed to differences in patient background factors, such as prior treatment, past medical history, and comorbidities. We observed that the incidence of TRAEs was significantly lower in the younger age group (< 65 or < 75 years) versus the older age group (≥ 65 or ≥ 75 years). Cross-tabulation with the number of nivolumab doses showed that the younger age group received a lower number of doses (data not shown). Similarly, the incidence of TRAEs was lower in the CRP ≥ 5 mg/dL versus CRP < 5 mg/dL group, and cross-tabulation with the number of nivolumab doses showed that patients in the CRP ≥ 5 mg/dL group received a lower number of nivolumab doses (data not shown). In the younger group and the CRP ≥ 5 mg/dL group, many patients were considered to have died or stopped the therapy before the onset of adverse reactions, which may have affected the incidence of adverse reactions.
Furthermore, patients with a "past medical history" and/or "present comorbidities" related to an organ are likely to experience more TRAEs due to organ dysfunction. Multivariate analyses confirmed that patients with a past medical history of hepatic/thyroid dysfunction may experience significantly more TRAEs. Similarly, patients with a past medical history related to ILD may experience significantly more TRAEs.
Overall, no new safety signals were identified. Although some TRAEs occurred more frequently in this PMS study than in the CheckMate 025 trial, no specific concerns were raised.
The 6-and 12-month survival rates (71.8% and 57.9%) in this study were lower than those in the CheckMate 025 trial (89.2% and 76.0%), respectively. Although direct comparison may not be appropriate partly due to the different evaluation methods, these differences in the survival rates may be attributed to different patient characteristics, such as KPS score, histological types, and age, that are known to be related to poor prognosis; this PMS study included more patients with poor (≤ 70) KPS score (23.1% vs 5.9%), histological types other than clear cell type (15.5% vs none), and age ≥ 75 years (14.8% vs 8.3%) compared with the CheckMate 025 trial [1]. The effectiveness in a real-world setting may help us understand the net benefit of nivolumab for un-resectable or metastatic RCC patients. Effectiveness in this study was evaluated by the 1-year OS rate since the follow-up period of the study was up to 1 year, which made it difficult to evaluate long-term

Limitations
This was a PMS study, and treatment was based on the physician's discretion. We did not conduct a central review of the TRAEs in case report forms. Thus, TRAE incidences

Conclusion
The results of this PMS study showed that the safety of nivolumab monotherapy in Japanese patients with un-resectable or metastatic RCC was similar to that observed in the phase 3 CheckMate 025 trial and another retrospective realworld study. No new safety concerns were identified over the 12-month observation period in patients with RCC. These results support nivolumab as a treatment option for patients with advanced RCC in the real world.
Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.