KEYNOTE-598: No Improvement With Addition of Ipilimumab to Pembrolizumab in NSCLC

Get Permission

In the phase III KEYNOTE-598 study, the addition of ipilimumab to pembrolizumab increased toxicity without boosting efficacy as first-line therapy for metastatic non–small cell lung cancer (NSCLC) in patients with high expression of PD-L1. The findings were presented at the International Association for the Study of Lung Cancer 2020 World Conference on Lung Cancer Singapore, held in virtual format in January 2021,1 and published simultaneously in the Journal of Clinical Oncology.2

“The addition of ipilimumab to pembrolizumab did not improve efficacy ... for NSCLC with high PD-L1 and no targetable mutations.”
— Michael Boyer, MBBS, PhD

Tweet this quote

The research, which was conducted on patients with metastatic NSCLC who had high expression of PD-L1, was presented by Michael Boyer, MBBS, PhD, Clinical Professor of Medicine at the Chris O’Brien Lifehouse and the Central Clinical School of the University of Sydney, Australia.

“Nonbinding futility criteria were met, and ipilimumab and placebo were discontinued per the external data monitoring committee’s recommendation,” Dr. Boyer announced.

“The addition of ipilimumab to pembrolizumab did not improve efficacy, and it did increase toxicity for first-line therapy for NSCLC with high PD-L1 and no targetable mutations. This means that pembrolizumab monotherapy remains a standard-of-care first-line treatment for this group of patients…. Future research will focus on other ways to improve outcomes,” he commented.

KEYNOTE-598 Details

The previous KEYNOTE-024 study showed that pembrolizumab monotherapy significantly improved survival vs platinum-doublet chemotherapy for patients with metastatic NSCLC and a tumor proportion score (TPS) of at least 50% without targetable EGFR or ALK aberrations.3 The current randomized, double-blind phase III KEYNOTE-598 study aimed to determine whether adding ipilimumab to pembrolizumab improved efficacy over pembrolizumab alone in such patients.

“The idea was to take a group of patients who already responded well to pembrolizumab (ie, high–PD-L1 expressors) and see whether it was possible to further enhance response and survival outcomes,” Dr. Boyer said.

The trial enrolled 568 high–PD-L1 expressors from Europe, Asia, and North America; they were assigned 1:1 to the combination of pembrolizumab at 200 mg every 3 weeks for up to 35 cycles plus ipilimumab at 1 mg/kg every 6 weeks or pembrolizumab and placebo. Subjects were stratified by European Cooperative Oncology Group score (0 vs 1), region (East Asia vs not East Asia), and histology (squamous vs nonsquamous). The primary endpoints were overall survival and progression-free survival in the intent-to-treat population.


  • The phase III KEYNOTE-598 trial found no value in adding ipilimumab to pembrolizumab for the first-line treatment of metastatic NSCLC.
  • Median overall survival was 21.4 months with ipilimumab plus pembrolizumab compared with 21.9 months with pembrolizumab alone (hazard ratio [HR] = 1.08; 95% confidence interval [CI] = 0.85–1.37; P = .74), which met the futility criteria.
  • Median progression-free survival was 8.2 months vs 8.4 months (HR = 1.06; 95% CI = 0.86–1.30; P = .72), and 45% of patients in each group responded, for a median duration of approximately 17 months.
  • Toxicity was greater with ipilimumab plus pembrolizumab.
  • Single-agent pembrolizumab remains a standard of care for these patients.

The protocol-specified first interim analysis was to occur upon approximately 255 deaths and approximately 12 months after the last participant was randomly assigned. Nonbinding futility criteria at first interim analysis were differences in the restricted mean survival time between ipilimumab plus pembrolizumab and placebo plus pembrolizumab of ≤ 0.2 at the maximum observation time and ≤ 0.1 at 24 months of follow-up.

No Differences in Endpoints

With 272 deaths, median overall survival was 21.4 months for ipilimumab plus pembrolizumab compared with 21.9 months for pembrolizumab alone (hazard ratio [HR] = 1.08; 95% confidence interval [CI] = 0.85–1.37; P = .74). Restricted mean survival time differences were –0.56 at the maximum observation time and –0.52 at 24 months, which met the futility criteria, according to Dr. Boyer.

With 372 events, median progression-free survival was 8.2 months for ipilimumab plus pembrolizumab compared with 8.4 months for placebo plus pembrolizumab (HR = 1.06; 95% CI = 0.86–1.30; P = .72). The objective response rate was 45.4% in both arms; the median duration of response was 16.1 months with the combination vs 17.3 months with pembrolizumab alone.

No subgroups derived a benefit from adding ipilimumab, Dr. Boyer reported. Aside from tumor mutational burden, which the investigators have not examined, “there were no other subsets where the results were any different from the overall results of the study … none where the combination works better or worse,” he commented.

More Toxicity With the Combination

“Safety in the pembrolizumab-plus-ipilimumab and pembrolizumab-plus-placebo groups was as expected,” Dr. Boyer said. The addition of ipilimumab led to increased rates of treatment-related adverse events vs placebo:

  • Any-grade adverse events: 76% vs 68% (immune-related: 45% vs 32%)
  • Grade ≥ 3 adverse events: 35% vs 20% (immune-related: 20% vs 8%)
  • Serious adverse events: 28% vs 14% (immune-related: 19% vs 7%)
  • Deaths: 2% vs 0 (immune-related: 2% vs 0).

Pruritus, rash, and hypothyroidism were the most common adverse events and were seen in more patients on the combination therapy. Patients on the two drugs also had higher rates of treatment discontinuation due to adverse events. In the combination arm, 6% discontinued ipilimumab and 3% discontinued placebo; 19% and 7%, respectively, discontinued both drugs. Immune-mediated adverse events and infusion reactions led to discontinuation of ipilimumab or placebo in 2% and 1%, respectively, and to both drugs in 12% and 4%, respectively, he reported.

The increased toxicity with the combination may have been a factor in the negative efficacy outcome, Dr. Boyer said, though, he added, “it was not unexpectedly bad.” 

DISCLOSURE: Dr. Boyer has received honoraria from AstraZeneca; has served as an institutional consultant or advisor to AstraZeneca, Bristol Myers Squibb, Janssen, and Merck Sharp & Dohme; has received institutional research funding from Amgen, Ascentage Pharma, AstraZeneca, Boehringer Ingelheim, Bristol Myers Squibb, Genentech/Roche, Janssen, Lilly, Merck Serono, Merck Sharp & Dohme, Novartis, and Pfizer; and has been reimbursed for travel, accommodations, or other expenses by Genentech/Roche and Merck Sharp & Dohme.


1. Boyer M, Sendur MAN, Rodríguez-Abreu D, et al: Pembrolizumab plus ipilimumab vs pembrolizumab plus placebo as 1L therapy for metastatic NSCLC of PD-L1 TPS ≥ 50%: KEYNOTE-598. 2021 World Lung Cancer Conference. Abstract PS01.09. Presented January 30, 2021.

2. Boyer M, Sendur MAN, Rodriguez-Abreu D, et al: Pembrolizumab plus ipilimumab or placebo for metastatic non-small-cell lung cancer with PD-L1 tumor proportion score ≥ 50%: Randomized, double-blind phase III KEYNOTE-598 study. J Clin Oncol. January 29, 2021 (early release online).

3. Reck M, Rodriguez-Abreu D, Robinson AG, et al: Pembrolizumab versus chemotherapy for PD-L1-positive non-small-cell lung cancer. N Engl J Med 375:1823-1833, 2016.

Related Articles

Expert Point of View: Yun Fan, MD, and Deborah Doroshow, MD, PhD

Yun Fan, MD

Yun Fan, MD

Invited discussant of KEYNOTE-598,1Yun Fan, MD, Director of Thoracic Tumor Center at Zhejiang Cancer Hospital in Hangzhou, China, suggested that patients with non–small cell lung cancer (NSCLC) with low PD-L1 expression and those with high tumor mutational burden may, in fact,...