KEYNOTE-671 Trial: Perioperative Pembrolizumab Plus Chemotherapy Improves Outcomes in Stage II and III NSCLC

Get Permission

Neoadjuvant therapy with pembrolizumab plus platinum-based chemotherapy followed by surgery and then adjuvant pembrolizumab led to significantly improved event-free survival in patients with resectable stage II and III non–small cell lung cancer (NSCLC) compared with platinum-based chemotherapy plus placebo in the randomized phase III KEYNOTE-671 trial.1 Overall survival trended higher in the pembrolizumab-containing arm, although longer-follow-up is needed. The results of KEYNOTE-671 were presented at the 2023 ASCO Annual Meeting1 and published simultaneously in The New England Journal of Medicine.2

Event-free survival was improved by 42% with pembrolizumab plus chemotherapy compared with placebo plus chemotherapy (P < .00001). At a median follow-up of 25.2 months, median event-free survival was not reached with pembrolizumab plus chemotherapy vs 17 months with placebo plus chemotherapy. The 2-year event-free survival rates were 62.4% and 40.6%, respectively. Major pathologic response and pathologic complete response rates were significantly higher with the addition of pembrolizumab.

“In KEYNOTE-671, neoadjuvant pembrolizumab plus chemotherapy followed by surgery and adjuvant pembrolizumab achieved statistically significant and clinically meaningful event-free survival in patients with resectable stage II and III NSCLC,” stated lead author Heather A. Wakelee, MD, FASCO, of Stanford Cancer Institute. “The adverse event profile was as expected. These results argue for the importance of screening patients. The data provide strong support for perioperative pembrolizumab as a promising treatment option for stage II and III NSCLC.”

“The data provide strong support for perioperative pembrolizumab as a promising treatment option for stage II and III NSCLC.”
— Heather A. Wakelee, MD, FASCO

Tweet this quote

Immune checkpoint inhibitors blocking PD-1 or PD-L1 are standard approaches for later-stage NSCLC. Studies have also shown a significant benefit for checkpoint inhibitors used before or after resection in NSCLC. The perioperative approach—using checkpoint inhibitor therapy both before and after surgery—was assessed in the AEGEAN and Neotorch studies.3,4 In AEGEAN, the addition of durvalumab, and in Neotorch, the addition of toripalimab, to neoadjuvant platinum-based chemotherapy followed by surgery and adjuvant PD-1/PD-L1 inhibition showed improved event-free survival compared with neoadjuvant platinum-based chemotherapy and surgery alone. KEYNOTE-671 evaluated pembrolizumab in this setting.

KEYNOTE-671 Details

The double-blind, placebo-controlled phase III KEYNOTE-671 trial randomly assigned 797 patients with resectable stage II, IIIA, or IIIB NSCLC to receive pembrolizumab at 200 mg intravenously every 3 weeks or placebo for four cycles of cisplatin-based chemotherapy. This was followed by surgery and continuation of pembrolizumab or placebo for up to 13 cycles. Stratification factors included stage, tumor proportion score (TPS) for PD-L1, histology, and geographic region of enrollment. The dual primary endpoint was event-free survival and overall survival.

All patients received at least one dose of treatment. The first interim analysis was prespecified after 326 events—about 5 months. The final analysis was for pathologic complete response rate.

Among all 797 randomly assigned patients, about 75% on each arm completed all four cycles of treatment. Patients could go on to surgery if they had at least one treatment. At a median follow-up of 25.2 months, 40% of the pembrolizumab arm and 35% of the placebo arm completed all four treatments.

At baseline, almost two-thirds (62%) had a smoking history, and about 12% had never smoked. Approximately 30% had stage II disease, and 70% had stage III disease; one-third had a TPS of at least 50% PD-L1 expression; about 30% had a TPS between 1% and 49% PD-L1 expression; and about 35% had a TPS less than 1% PD-L1 expression. About 4% had EGFR mutations, and about 2.7% had ALK mutations. Most surgeries were lobectomies: 79% and 75%, respectively. Pneumonectomy was performed in 11% and 12%, respectively.

At 30 days, six patients on the pembrolizumab arm and two patients on the placebo arm died of any cause. A total of 35% of patients given pembrolizumab and 51% of those assigned to placebo sustained an event.

Key Results and Toxicity

Event-free survival was statistically significant favoring the pembrolizumab arm (P < .00001).

“All subgroups uniformly benefited from pembrolizumab vs placebo. Histology made no difference, and a benefit was observed across all stages. The magnitude of benefit was greater in those with higher PD-L1 levels, but a benefit was observed across all PD-L1 levels of expression,” Dr. Wakelee told the audience.

There appeared to be a benefit of pembrolizumab in a small cohort of 33 patients with EGFR mutations. However, she noted, it is difficult to draw conclusions from this observation, given the superiority of osimertinib in EGFR-mutated NSCLC.

At follow-up, the survival curves were starting to separate for the two arms, with a 27% improvement in overall survival for perioperative pembrolizumab at this analysis.

Patients in the pembrolizumab group were more likely than those in the placebo group to attain major pathologic response: 30.2% vs 11%, respectively (P < .00001) and pathologic complete response: 18.1% vs 4%, respectively (P < .00001). In an exploratory analysis looking at event-free survival according to pathologic complete response rates, patients with a major pathologic response and a pathologic complete response did better than those who did not reach these milestones; pembrolizumab was superior to placebo plus chemotherapy in both subgroups—regardless of the depth of response achieved.

More Information

For more on the KEYNOTE-671 trial of neoadjuvant pembrolizumab and platinum-based chemotherapy followed by surgery and adjuvant pembrolizumab in stage II, IIIA, or IIIB non–small cell lung cancer, see a conversation with Narjust Florez, MD, and Heather A. Wakelee, MD, FASCO, on The ASCO Post Newsreels at

The toxicity was as expected in a patient population treated with immunotherapy, chemotherapy, and surgery. More immune-related adverse events were reported on the pembrolizumab arm. The most frequent immune-related adverse events in the pembrolizumab and placebo arms, respectively, included hypothyroidism (11.1% vs 1.8%), hyperthyroidism (5.6% vs 3.3%), and pneumonitis (5.6% and 1.8%). Grades 3 to 5 adverse events were reported in 5.8% of patients given pembrolizumab vs 1.5% of those given placebo. Treatment-related adverse events led to treatment discontinuation in 12.6% of pembrolizumab-treated patients and 5.3% on the placebo arm. One death occurred in the pembrolizumab arm because of pneumonitis. 

DISCLOSURE: Dr. Wakelee has served as a consultant or advisor to Mirati Therapeutics; has received institutional research funding from ACEA Biosciences, Arrys Therapeutics, AstraZeneca/MedImmune, Bristol Myers Squibb, Clovis Oncology, Genentech/Roche, Helsinn Therapeutics, Merck, Novartis, Seagen, and Xcovery.


1. Wakelee HA, Liberman M, Kato T, et al: KEYNOTE-671: Randomized, double-blind, phase 3 study of pembrolizumab or placebo plus platinum-based chemotherapy followed by resection and pembrolizumab or placebo for early-stage NSCLC. 2023 ASCO Annual Meeting. Abstract LBA100. Presented June 3, 2023.

2. Wakelee H, Liberman M, Kato T, et al: Perioperative pembrolizumab for early-stage non–small-cell lung cancer. N Engl J Med. June 3, 2023 (early release online).

3. Heymach JV, Harpole D, Mitsudomi T, et al: AEGEAN: A phase 3 trial of neoadjuvant durvalumab + chemotherapy followed by adjuvant durvalumab in patients with resectable NSCLC. AACR Annual Meeting 2023. Abstract CT005. Presented April 16, 2023.

4. Lu S, Wu L, Zhang W, et al: Perioperative toripalimab + platinum-doublet chemotherapy vs chemotherapy in resectable stage II/III NSCLC: Interim event-free survival analysis of the phase III Neotorch study. ASCO Plenary Series. Abstract 425126. Presented April 20, 2023.

Related Articles

Expert Point of View: Mark Awad, MD, PhD

Formal discussant of KEYNOTE-671, Mark Awad, MD, PhD, of the Dana-Farber Cancer Institute, Boston, was optimistic about these data. “A hazard ratio of 0.58 for event-free survival and signals favoring overall survival in stage II to III non–small cell lung cancer [NSCLC] are truly impressive....