Canadian cohort expanded-access program of nivolumab plus ipilimumab in advanced melanoma

wild-type advanced melanoma 5,6 . In the randomized, phase iii CheckMate 067 study, nivolumab plus ipilimumab or nivolumab alone, compared with ipilimumab alone, was associated with a significant improvement in the ABSTRACT Background The combination of nivolumab and ipilimumab is approved in several jurisdictions (United States, European Union, Canada) for the first-line treatment of patients with advanced melanoma. CheckMate 218 is a North American expanded-access program ( eap ) of nivolumab plus ipilimumab in patients with advanced melanoma. Here, we report safety and survival outcomes for the Canadian cohort in the eap . Methods Eligible patients were those 18 years of age or older with unresectable stage iii or iv melanoma, an Eastern Cooperative Oncology Group performance status of 0 or 1, and no prior anti–PD-1 or anti– ctla -4 therapy. Patients were treated with nivolumab 1 mg/kg and ipilimumab 3 mg/kg every 3 weeks for 4 cycles (induction phase); they then continued with nivolumab 3 mg/kg every 2 weeks (maintenance phase) until progression, unacceptable toxicity, or a maximum of 48 weeks, whichever occurred first. Safety and overall survival ( os ) data were collected. Of 194 patients enrolled, 174 were treated, and 51% continued on nivolumab maintenance. Median follow-up was 12.9 months. All-grade and grades 3–4 treatment-related adverse events were reported in 98% and 60% of patients respectively and led to treatment discontinuation in 40% and 28% of patients. Two treatment-related deaths were reported. The 12- and 18-month os rates were 80% [95% confidence interval ( ci ): 73% to 86%] and 76% (95% ci : 67% to 82%) respectively. In this Canadian population, nivolumab plus ipilimumab demonstrated a safety profile and survival outcomes consistent with phase ii and iii clinical trial data.


INTRODUCTION
Immune checkpoint inhibitors such as nivolumab, an anti-PD-1 antibody, and ipilimumab, an anti-ctla-4 antibody, have changed the treatment paradigm for advanced melanoma. Before the era of immune checkpoint inhibition, median overall survival (os) for patients with advanced melanoma ranged from 6 months to 12 months 1 ; today, it is 4 years or longer 2 . Nivolumab with or without ipilimumab is used for the first-line treatment of patients with advanced melanoma 3,4 .
Results from randomized clinical trials (rcts) indicate that, compared with ipilimumab alone, the combination of nivolumab and ipilimumab is associated with improved os in patients with treatment-naïve advanced melanoma, but with a higher frequency of treatment-related adverse events (traes) than occur with monotherapy 2,5-7 . In the randomized, phase ii CheckMate 069 study, treatment with nivolumab plus ipilimumab was compared with ipilimumab alone and was associated with a significant improvement in the objective response rate and median progression-free survival in patients with treatment-naïve BRAF wild-type advanced melanoma 5,6 . In the randomized, phase iii CheckMate 067 study, nivolumab plus ipilimumab or nivolumab alone, compared with ipilimumab alone, was associated with a significant improvement in the objective response rate, median progression-free survival, and median os in patients with treatment-naïve advanced melanoma 2,7 . The combination of nivolumab and ipilimumab was approved for advanced melanoma in 2015 in the United States and in 2016 in Canada.
CheckMate 218 (see NCT02186249 at https://ClinicalTrials. gov/) is a North American expanded-access program (eap) for nivolumab plus ipilimumab in patients with unresectable stage iii or iv melanoma, including cutaneous, ocular or uveal, mucosal, and acral melanoma. The eap provided the combination to patients with treatment-naïve disease or disease that progressed with other therapies, excluding anti-ctla-4 or anti-PD-1 therapies, until market authorization was granted for the combination. The overall patient population for the eap included 754 patients: 580 treated in the United States and 174 treated in Canada. We previously reported earlier os data for the combined U.S. (1-year follow-up) and Canadian (6-month follow-up) cohorts 8 . Here, we report updated safety and os data for the Canadian cohort, with a median follow-up of 12.9 months. Results from the Canadian cohort support reimbursement decisions in Canada related to this approved treatment. Results from the total North American population (United States and Canada combined) will be published separately.

Patients
Eligible patients were 18 years of age or older with unresectable stage iii or iv metastatic melanoma per the American Joint Committee on Cancer staging system (7th edition) 9 . Patients were required to have an Eastern Cooperative Oncology Group performance status (ecog ps) of 0 or 1 and to be treatment-naïve to anti-ctla-4 and anti-PD-1 agents. Patients could have received other systemic treatments for localized or metastatic disease, including braf or mek inhibitors. (Initially, patients with BRAF mutation-positive tumours who had received prior treatment with targeted therapy were excluded, but the protocol was amended within a few weeks to remove that exclusion.) Patients were excluded if they had active (symptomatic) or untreated brain metastases or leptomeningeal metastases, a life expectancy of less than 6 weeks, autoimmune disease, or conditions requiring systemic corticosteroids or other immunosuppressive medications within 14 days of drug administration. Patients were also excluded if they required other systemic antineoplastic therapy while receiving nivolumab.

EAP Design and Treatment
An eap protocol was used for administration of sequential doses of nivolumab plus ipilimumab. Investigators had previous experience with the administration of nivolumab and ipilimumab, either as monotherapy or in combination. Patients received nivolumab (intravenously over 60 minutes at 1 mg/kg) and ipilimumab (intravenously over 90 minutes at 3 mg/kg every 3 weeks) for 4 doses during the induction phase. Subsequently, they continued with single-agent nivolumab (intravenously over 60 minutes at 3 mg/kg every 2 weeks) during the maintenance phase, until disease progression or unacceptable toxicity, or until a maximum of 48 weeks from the first nivolumab monotherapy dose, whichever occurred first (supplemental Figure 1).
In Canada, patients who stopped combination therapy because of toxicity were allowed to resume nivolumab monotherapy if toxicities had resolved and upon discussion with the medical monitor. Patients who discontinued treatment were followed for adverse events (aes). After U.S. approval of the combination treatment, the eap was closed in the United States because patients were transitioned to the commercial supply of nivolumab, and data collection was continued only for the Canadian cohort. The eap was subsequently closed in Canada, and some patients experiencing clinical benefit at eap completion were provided nivolumab through a post-eap drug access program funded by the Bristol Myers Squibb Company.

EAP Endpoints
Safety parameters-collected according to health authority regulations starting at cycle 1 and recommended for monitoring until 100 days after discontinuation of therapy-were aes, physical examination, ecog ps, and laboratory results. Evaluation of aes commenced with the first dose and ended 30 days after the last dose of therapy. Severity was assessed according to the U.S. National Cancer Institute's Common Terminology Criteria for Adverse Events, version 4.0. Serious aes were those that resulted in death, that were life-threatening, that resulted in persistent or significant disability or incapacity, or that required intervention or hospitalization.
Overall survival was defined as the time from treatment start to death from any cause. Collection of survival data was recommended for up to 5 years from the first nivolumab monotherapy dose, but the eap was closed early when a safety analysis determined that patient safety was consistent with observations across the nivolumab program. Response data were not collected.

Statistical Analysis
A descriptive statistical analysis was performed on the collected safety and os data; no formal hypothesis-testing was conducted. The overall study sample size for the descriptive analysis was based on the projected number of patients meeting the enrolment criteria during a specific period, because one of the objectives was to provide access to therapy. In the overall study population (United States and Canada), it was estimated that, with a screening failure rate of 15%, 1000 patients would have to enrol, provide informed consent, and be screened to allow for 850 patients to be treated for a maximum of 48 weeks in the maintenance phase or until the first of disease progression, loss to follow-up, death, withdrawal of consent, or unacceptable toxicity. 12.9 months (range: 0.3-21.2 months). All patients discontinued eap therapy throughout the on-eap treatment period, with the most common reasons for discontinuation being drug toxicity (n = 75, 43%) and disease progression (n = 46, 26%). Discontinuations during the induction phase totalled 86 [including 21 (24%) because of disease progression and 58 (67%) because of toxicity] and 88 during the maintenance phase [including 25 (28%) because of disease progression and 17 (19%) because of toxicity]. At database lock, 123 patients (71%) continued to be followed for aes. At the end of the eap, 21 of the 174 patients who discontinued were experiencing clinical benefit and were provided nivolumab through a separate post-eap drug access program, without interruption in treatment.
Median treatment duration of nivolumab and ipilimumab in the induction phase was 1.7 months (range: 0.03-3.9 months) and 1.6 months (range: 0.03-3.9 months) respectively. The median number of nivolumab and ipilimumab doses received during the induction phase was 3.0 each (range: 1-4), and the median number of nivolumab doses received during the maintenance phase was 1.0 (range: 0-35). From the induction phase, 85 patients (51%) went on to receive maintenance nivolumab monotherapy. Median duration of nivolumab treatment in the maintenance phase was 6.6 months (range: 0.03-16.6 months). In 86 patients (49%), no nivolumab doses were administered during the maintenance phase; 48 patients (28%) received more than 10 doses. In the overall eap, 57 patients (33%) received more than 10 doses of nivolumab. Overall, a relative dose intensity of 90% or greater with nivolumab and ipilimumab was achieved by 47 patients (27%) and 133 patients (76%) respectively. A dose delay with nivolumab and ipilimumab occurred in 102 patients (59%) and 47 patients (27%) respectively, with the most common reason being occurrence of an ae.
Select traes of any grade (those with a potential immunologic cause) occurred most frequently in skin in 118 patients (68%) and in the gastrointestinal system in 84 patients (48%, Table iii). Of the select traes of any grade, the most common were diarrhea in 80 patients (46%), maculopapular rash in 57 (33%), and pruritus in 43 (25%). Of the select traes of grades 3-4, the most common were diarrhea in 18 patients (10%), increased alanine aminotransferase in 16 (9%), and increased aspartate aminotransferase in 15 (9%).

Efficacy
With a median follow-up of 12.9 months, median os was 20.5 months [95% confidence interval (ci): 20.5 months to not reached], and 12-month and 18-month survival rates were 80% (95% ci: 73% to 86%) and 76% (95% ci: 67% to 82%) respectively ( Figure 2). The drop in os after 18 months is an artefact and reflects the end of follow-up, as patients transitioned to post-eap drug access.

DISCUSSION
This large Canadian eap allowed patients to access lifeprolonging medications while physicians were provided with experience in managing the toxicities associated with the novel agents. Moreover, real-world safety and efficacy data were collected to ensure that the results accorded with those from registrational rcts 2,5-7 . Compared with patient populations in the Check-Mate 069 and CheckMate 067 trials 5,7 , the eap population was younger (median age: 56 years vs. 59-64 years), consisted of proportionally fewer patients with an ecog ps of 0 (58% vs. 73%-83%), and had a higher incidence of BRAF mutation-positive tumours (51% vs. 24%-32%). The relatively high proportion of patients with BRAF mutationpositive tumours is likely explained by the inclusion criteria, which allowed enrolment of patients who had progressed on braf-or mek-targeted therapy. However, the incidence of elevated baseline serum ldh was similar in the eap and rct populations (36% vs. 25% -36%). In contrast to the CheckMate 069 and CheckMate 067 trials, CheckMate 218 enrolled patients who could previously have received other prior systemic treatment. Overall, the eap population had some prognostic features that were more negative than those in the rct populations, but outcomes for the eap population were comparable.
Despite differences between the e ap and the rct populations, survival outcomes with nivolumab plus ipilimumab for the patients in the eap compared favourably. The 12-month os rate was 80% in the eap population and 73% for those who received nivolumab plus ipilimumab in the CheckMate 069 and CheckMate 067 trials 6,10 . Further underscoring the survival benefit, the 12-month os rate was greater than 70% in groups with poor prognostic characteristics (elevated ldh and ocular or uveal melanoma, for instance). Patients with ocular or uveal melanoma were not eligible to enrol in the rcts, and although such patients constituted a small proportion (9%) of the eap, survival in that subgroup was better than expected. Ocular or uveal melanoma typically demonstrates an aggressive course 11 , with a 5-year relative survival rate of 19% in the metastatic setting 12 . Our observations accord with those in a retrospective study of 8 patients in whom ipilimumab plus nivolumab was active in uveal melanoma 13 . In addition, compared with female patients, male patients experienced increased survival, consistent with a recent systematic review of immune checkpoint inhibitors 14 . Survival rates were higher for patients 65 years of age and older than for patients less than 65 years of age, suggesting that the combination might be of benefit in elderly patients. A similar trend of an increased response to anti-PD-1 therapy in elderly patients has also been observed in another study 15 . Furthermore, survival rates were similar for patients with BRAF mutation-negative and BRAF mutation-positive tumours, suggesting that BRAF mutation status is not predictive of response. However, because of the observational nature of the present eap and its short follow-up, os data from this eap should be interpreted with caution. Notably, the os differences observed between patients with BRAF mutation-negative and BRAF mutation-positive tumours in CheckMate 067 were not readily evident at 12 and 18 months, beginning to emerge only at 2 years of follow-up.
Because of the nature of the study design, response data were not collected in the present eap. The absence of response and progression-free survival data restricts further comparisons with clinical studies and represents a limitation.
Nivolumab plus ipilimumab was well tolerated. Safety results were consistent with those in the rcts 5,7 . Grades 3-4 traes were reported in 60% of patients and led to treatment discontinuation in 28% of the population. In Check-Mate 069 and CheckMate 067 5,7 , grades 3-4 traes were reported in 54%-55% of patients and led to treatment discontinuation in 29%-38%. The most common grades 3-4 traes included diarrhea (10%) and increased alanine aminotransferase and aspartate aminotransferase (9% each), which were also among the most common aes, occurring at similar rates in CheckMate 069 and CheckMate 067 5,7 . Two treatment-related deaths were reported, but no new safety signals were identified.

CONCLUSIONS
CheckMate 218 was an observational eap evaluating the safety and efficacy of the nivolumab plus ipilimumab combination in real-world patients with advanced melanoma. Its results provide additional insights into the use of the combination in patients with advanced melanoma in the real-world setting, including subpopulations not studied in rcts. Clinical benefit was noted in patients with various melanoma subtypes and in those who had received prior treatments. The unexpected benefit in uveal melanomaalbeit in a limited number of patients-suggests that the combination should be explored further in that melanoma subtype. The results of the eap are consistent with rct data, further supporting the use of the nivolumab plus ipilimumab combination for the treatment of advanced melanoma.

DATA AVAILABILITY
Bristol Myers Squibb policy on data-sharing can be found at https://www.bms.com/researchers-and-partners/independentresearch/data-sharing-request-process.html.