Senior Editor, OncLive®
Hayley Virgil heads OncLive's feature article efforts and specializes in social issues and equality in oncology. Prior to joining the company in early 2020, she worked as an editor in numerous industries, including media, marketing, hospitality, and computer science, and freelanced in subjects such as history, culture, and the natural sciences.
The addition of ipilimumab to nivolumab as an adjuvant treatment failed to improve relapse-free survival in the intent-to-treat and PD-L1 of less than 1% populations compared with nivolumab alone in patients with resected stage IIIB to D/IV melanoma.
The addition of ipilimumab (Yervoy) to nivolumab (Opdivo) as an adjuvant treatment failed to improve relapse-free survival (RFS) in the intent-to-treat (ITT) and PD-L1 of less than 1% populations compared with nivolumab alone in patients with resected stage IIIB to D/IV melanoma, missing the dual primary endpoints of the phase 3 CheckMate-915 trial (NCT03068455) that were presented during the virtual AACR Annual Meeting 2021.1
Findings showed that, in the ITT population, the 2-year RFS rate was 64.6% vs 63.2% with nivolumab alone. The median RFS was not reached (NR) in either arm (HR, 0.92; 97.295 CI, 0.77-1.09; P = .269).
In those who had PD-L1 expression of less than 1%, the 2-year RFS rate was 53.6% vs 52.4% in the doublet and monotherapy arms, respectively. Here, the median RFS was 33.2 months and 25.3 months in each arm, respectively (HR, 0.91; 95% CI, 0.73-1.14).
“The dual primary end points of RFS in the [ITT population] and PD-L1 of less than 1% population were not met,” lead study author Georgina V. Long, BSc, PhD, MBBS, FRACP, FAHMS, co-medical director of the Melanoma Institute Australia (MIA) and chair of melanoma medical oncology and translational research at MIA and Royal North Shore Hospital, The University of Sydney, said during a virtual presentation on the findings.
“Nivolumab combined with ipilimumab vs nivolumab in resected stage IIIB to D and IV melanoma did not improve the RFS or distant metastasis-free survival; these results reaffirm nivolumab as an adjuvant standard of care.”
The rationale for the research was based on several prior studies, one of which was the phase 3 EORTC 18071 trial of adjuvant ipilimumab vs placebo in patients with stage III melanoma who had undergone complete resection.2 Results showed that a 10-mg/kg dose of ipilimumab every 3 weeks elicited an improved 5-year RFS rate vs placebo (65% vs 54%, respectively; HR, 0.72; 95% CI, 0.58-0.88).
The phase 3 CheckMate-238 trial, which examined the role of nivolumab vs ipilimumab in patients who had experienced recurrent melanoma after complete resection of stage IIIB/C or stage IV melanoma, helped to establish nivolumab as a standard of care in this patient population.3 The 3-mg dose of nivolumab every 2 weeks yielded a 4-year RFS rate of 52% vs 41% in patients who received a 10-mg/kg dose of ipilimumab (HR, 0.71; 95% CI, 0.60-0.86).
Early data from phase 3 CheckMate-067 trial (NCT01844505), which examined nivolumab or nivolumab/ipilimumab vs ipilimumab alone in patients with previously untreated advanced melanoma, indicated that patients regardless of PD-L1 expression may experience benefit from the combination.4
In the CheckMate-915 trial, which was conducted at 122 sites across 19 countries, randomized patients with completely resected stage IIIB to D or IV no evidence of disease melanoma to receive either 240 mg of nivolumab every 2 weeks and 1 mg/kg of ipilimumab every 6 weeks (n = 920) or 480 mg of nivolumab every 4 weeks (n = 924). Treatment was given for 1 year.
The trial also included an ipilimumab-monotherapy arm, wherein patients received the CTLA-4 inhibitor at 10 mg/kg (n = 99), though this arm was discontinued on July 20, 2017.
Stratification factors were by tumor PD-L1 expression (<1% vs 1% to <5% vs ≥5%), as well as by disease stage (IIIB vs IIIC-D vs IV). Complete lymph node dissection was not required.
The median follow-up was 28 months. Dual primary end points were RFS both in the ITT population and the PD-L1 less than 1% population. Key secondary end points included OS, correlation between PD-L1 and RFS, and outcomes on subsequent therapies. Exploratory end points consisted of distant metastasis-free survival and quality of life (QoL).
The median age was 55 years and the majority of patients (57%) were male. Thirty-eight percent of patients on both arms had a PD-L1 of less than 1% and 31% of patients had a BRAF mutation. Overall, 31% of patients had stage IIIB disease, 52.5% had stage IIIC disease, 3% had stage IIID disease, and 13% had stage IV disease. Fifty-three percent of patients on the nivolumab/ipilimumab arm and 52% of patients on the nivolumab arm had stage IIIC disease.
Notably, patients who received the combination regimen had a median duration of therapy of 7.6 months vs 11.1 months in the nivolumab-alone arm.
“The cumulative dose of nivolumab in the [combination] arm was 3840 mg, which was considerably lower than the cumulative dose in the nivolumab monotherapy [cohort], which was 6240 mg,” Long explained. “If we break down duration by time periods, we can see that nearly one-third of patients in the combination arm received less than 3 months of drug therapy, whereas only 13% in the nivolumab monotherapy arm received less than 3 months of therapy.”
Additional data showed that when examining RFS by stage, patients with stage III disease who were treated with nivolumab/ipilimumab and single-agent nivolumab experienced a 2-year RFS rate of 64.7% and 63.6%, respectively. The median RFS was NR in either arm (HR, 0.94; 95% CI, 0.80-1.11).
In patients with stage IV disease, the 2-year RFS rate in the doublet arm was 63.6% compared with 61.1% in the single-agent arm. Similarly, the median RFS was NR in either arm (HR, 0.88; 95% CI, 0.58-1.32).
Additionally, patients with stage III disease who were treated with the combination had a 2-year dMFS rate of 75.4% with the combination compared with 77.4% with nivolumab alone. The median dMFS was NR in either arm (HR, 1.01; 95% CI, 0.83-1.23).
In those with stage IV disease, the 2-year dMFS rates were 68.4% and 67.9% in the nivolumab/ipilimumab and nivolumab-alone arms, respectively. Similarly, the median dMFS was not reached in this subgroup in either arm (HR, 0.94; 95% CI, 0.70-1.25).
A post hoc analysis of the combination arm evaluated whether duration of therapy impacted RFS by comparing patients who discontinued the regimen after 6 months or less with those who did not discontinue after 6 months or less, according to Long.
“Although we do see a 5% difference in the landmark RFS for the combination in these 2 groups, it is not sufficient to explain the overall negative results of this trial,” Long said.
Use of subsequent therapies was comparable in both arms, she noted, at 32% with the combination and 36% with the monotherapy.
Regarding safety, 94% of patients on nivolumab/ipilimumab experienced all-grade treatment-related adverse effects (TRAEs) and 33% experienced grade 3/4 TRAEs; these rates were 86% and 13% in the nivolumab arm, respectively. Thirty-two percent and 19% of patients in the combination arm experienced any-grade and grade 3/4 TRAEs, respectively, that led to treatment discontinuation compared with 10% and 6% in the monotherapy arm, respectively. No treatment-related deaths occurred in the monotherapy arm vs 4 in the combination arm.
The most common any-grade AEs with nivolumab/ipilimumab were pruritis (33%), rash (24%), fatigue (30%), diarrhea (27%), and hypothyroidism; the most common grade 3/4 AEs were increased lipase (5%), increased alanine amino transferase (3%), diarrhea (2%), hypophysitis (2%), and increased aspartate amino transferase (2%).
In the nivolumab-alone arm, the most common any-grade AEs were fatigue (30%), rash (21%), pruritis (21%), and diarrhea (20%), with common grade 3/4 AEs consisting of increased lipase (2%), rash (1%), and diarrhea (1%).
QoL was comparable between the 2 arms.
“A persistent, stable quality of life [was experienced] by both arms through the duration of the study. There were no improvement or deterioration in [clinically meaningful differences] during the course of the study for both arms. Similar trends were observed for both EQ-5D utility index and EQ-5D visual-analogue scale,” Long concluded.