Randomized Phase II Trial of Nivolumab Versus Nivolumab and Ipilimumab for Recurrent or Persistent Ovarian Cancer: An NRG Oncology Study

Abstract

Purpose: Single-agent PD-1 blockade exhibits limited efficacy in epithelial ovarian cancer (EOC). We evaluated ipilimumab plus nivolumab compared with nivolumab alone in women with persistent or recurrent EOC.

Methods: Eligibility criteria included measurable disease, 1-3 prior regimens, and platinum-free interval (PFI) < 12 months. Participants were randomly allocated to intravenous nivolumab (every 2 weeks) or induction with nivolumab plus ipilimumab for 4 doses (every 3 weeks), followed by every-2-week maintenance nivolumab for a maximum of 42 doses. The primary null hypothesis was equal probability of objective response within 6 months of random allocation in each arm.

Results: One hundred patients were allocated to receive either nivolumab (n = 49), or nivolumab plus ipilimumab (n = 51), with PFI of < 6 months in 62%. Six (12.2%) responses occurred within 6 months in the nivolumab group and 16 (31.4%) in the nivolumab plus ipilimumab group (odds ratio, 3.28; 85% CI, 1.54 to infinity; P = .034). The median progression-free survival (PFS) was 2 and 3.9 months in the nivolumab and nivolumab plus ipilimumab groups, respectively, with a PFI-stratified hazard ratio of 0.53 (95% CI, 0.34 to 0.82); the respective hazard ratio for death was 0.79 (95% CI, 0.44 to 1.42). Grade ≥ 3 related adverse events occurred in 33% of patients in the nivolumab group and 49% in the combination group, with no treatment-related deaths. PD-L1 expression was not significantly associated with response in either treatment group.

Conclusion: Compared with nivolumab alone, the combination of nivolumab and ipilimumab in EOC resulted in superior response rate and longer, albeit limited, PFS, with toxicity of the combination regimen comparable to prior reports. Additional combination studies to enhance durability of the dual regimen are warranted.

FIG 1.

Eligibility, random allocation, and follow-up of the study patients. All patients who enrolled were included in the intention-to-treat analysis of efficacy end points. Patients were stratified by last platinum-free interval (< 6 months v 6-12 months), then randomly allocated in a 1:1 ratio to four intravenous infusions of the nivolumab or the nivolumab plus ipilimumab induction regimen: nivolumab 3 mg/kg every 2 weeks (nivolumab) or nivolumab 3 mg/kg plus ipilimumab 1 mg/kg every 3 weeks (nivolumab plus ipilimumab). Each induction regimen was followed by a common maintenance regimen: nivolumab 3 mg/kg every 2 weeks for a maximum of 42 doses. The analysis of safety included all patients receiving at least one dose of study therapy. The date of data cutoff was September 3, 2018. (*) Pathology report from original diagnosis indicated coexisting superficially invasive endometrioid adenocarcinoma of the endometrium along with stage IIIb high-grade serous fallopian tube cancer.

FIG 2.

RECIST sum by time since enrollment, according to treatment group. The comparison of percentage change in RECIST sum across treatment groups over time is shown in these spider plots, where nodes below the red lines define objective response. Response durations of at least 6 months without evidence of new disease occurred in 4 (8.2%) and 8 (15.7%) patients in the (A) nivolumab and (B) nivolumab plus ipilimumab groups, respectively. (*) RECIST increase beyond 100%.

FIG 3.

Analyses of progression-free survival (PFS) and overall survival (OS), according to treatment group. Analysis of (A) PFS and (B) OS, respectively, for all 100 patients randomly assigned to receive nivolumab induction followed by nivolumab maintenance therapy or nivolumab plus ipilimumab (Ipi) induction followed by nivolumab maintenance therapy, after stratification for the most recent platinum-free interval. Summary PFS and OS as well as hazard ratio (HR) and P value are presented in the respective tables. There was a significant, time-dependent decrease in the hazard of progression in the nivolumab plus ipilimumab group as compared with the nivolumab group (HR, 0.528; 95% CI, 0.339 to 0.821; P = .0041). As compared with the nivolumab group, the hazard of death was 0.789 for the nivolumab plus ipilimumab group (95% CI, 0.439 to 1.418; P = .43).

FIG 4.

Forest plot for selected baseline characteristics. Associations are determined with asymptotic methods. Because the number of events in some sets is small, the associated P value may be inaccurate. Control, nivolumab arm; Experimental, nivolumab plus ipilimumab arm; PFS, progression-free survival.