Adjuvant Pembrolizumab Sustains DFS Benefit in Resected NSCLC

Adjuvant pembrolizumab (Keytruda) continued to extend disease-free survival (DFS) compared with placebo in patients with completely resected stage IB to IIIA non–small cell lung cancer (NSCLC), but it did not improve overall survival (OS), according to the final analysis of the phase 3 PEARLS/KEYNOTE-091 trial (NCT02504372) presented at the 2026 ESMO Congress.¹

Findings showed that at a median time of 87.7 months (range, 68.6-120.1) from randomization to data cutoff, median DFS in the intention-to-treat (ITT) population was 58.7 months (95% CI, 47.0-70.4) in the pembrolizumab arm (n = 590) vs 43.3 months (95% CI, 35.0-55.6) in the placebo arm (n = 587; HR, 0.85; 95% CI, 0.73-0.98). The 6-year DFS rates were 44.0% (95% CI, 39.7%-48.2%) vs 39.2% (95% CI, 35.0%-43.3%), respectively.

Median OS was not reached in either arm (HR, 0.98; 95% CI, 0.81-1.18; one-sided P = .41826).

"At final analysis with long-term follow-up, adjuvant pembrolizumab continues to demonstrate a DFS benefit compared [with] placebo," Mary O'Brien, MD, FRCP, a consultant medical oncologist at The Royal Marsden NHS Foundation Trust in the United Kingdom, said during the presentation. "There was no OS benefit with adjuvant pembrolizumab, nor was there overall detriment."

Prior data from KEYNOTE-091 supported the January 2023 FDA approval of pembrolizumab as an adjuvant treatment following resection and platinum-based chemotherapy for patients with stage IB (T2a ≥4 cm), II, or IIIA NSCLC.²

How was the trial designed?

In the randomized, triple-blind PEARLS/KEYNOTE-091 study, patients who had undergone complete resection, with or without subsequent adjuvant chemotherapy, received pembrolizumab or placebo.^1,3 In total, 1177 patients were randomly assigned 1:1 to pembrolizumab (n = 590) or placebo (n = 587).¹

Patients in the pembrolizumab arm received 200 mg intravenously every 3 weeks for a maximum of 18 doses (approximately 1 year); placebo was given on an identical schedule. Stratification factors included disease stage, PD-L1 tumor proportion score (TPS; <1% vs 1%-49%, vs ≥50%), receipt of adjuvant chemotherapy, and geographic region.

Up to 4 cycles of adjuvant chemotherapy were allowed. Chemotherapy was to be considered for patients with stage IB disease and a tumor of at least 4 cm and was strongly recommended for those with stage II or IIIA disease. Chemotherapy had to start within 12 weeks of surgery, and study treatment then began 3 to 12 weeks after the last chemotherapy dose. Patients who did not receive chemotherapy started study treatment within 12 weeks of surgery.

Per protocol, the final OS analysis could be triggered either by a prespecified number of events or by reaching 10 years from the first patient's randomization. Because the event threshold had not been met, the time-triggered analysis was presented; data cutoff was January 21, 2026.

Investigator-assessed DFS in the ITT population and in patients with a PD-L1 TPS of at least 50% served as dual primary end points. DFS events included recurrence, a second primary cancer, or death from any cause, and a second extrapulmonary malignancy also counted as an event.

Secondary end points comprised DFS in patients with a PD-L1 TPS of at least 1%; OS in the ITT, TPS of at least 50%, and TPS of at least 1% populations; lung cancer–specific survival (LCSS) in the ITT population; and safety. Health-related quality of life was an exploratory end point.

What were the patient characteristics?

Patients in both arms had a median age of 65 years (pembrolizumab, range, 31-87; placebo, range, 37-85). Men made up 68.0% and 68.7% of the pembrolizumab and placebo arms, respectively. Most patients were White (76.3% vs 77.5%). The majority had stage II disease (55.9% vs 57.6%); 29.7% vs 27.3% had stage IIIA disease, and 14.4% vs 14.8% had stage IB disease. Nonsquamous histology was reported in 67.5% vs 61.8% of patients, an ECOG performance status of 1 in 35.6% vs 41.6%, and current or former smoking in 85.3% vs 88.8%.

Adjuvant chemotherapy had been given to 85.8% vs 85.9% of patients. PD-L1 TPS was below 1% in 39.5% of both arms, 1% to 49% in 32.0% vs 32.4%, and 50% or higher in 28.5% vs 28.1%.

EGFR mutations were present in 6.6% vs 5.8% of patients, and ALK translocations occurred in 1.2% of patients in each arm, although EGFR and ALK status was unknown for most patients.

What additional DFS and OS findings were reported?

In the subgroup that received adjuvant chemotherapy, median DFS reached 58.7 months (95% CI, 47.0-70.5) with pembrolizumab compared with 39.1 months (95% CI, 31.3-51.7) with placebo (HR, 0.80; 95% CI, 0.68-0.94); 6-year DFS rates were 43.6% and 38.1%, respectively.

Among patients with a PD-L1 TPS of at least 50%, the updated DFS HR was 0.87 (95% CI, 0.65-1.17).

Six-year OS rates in the ITT population were 68.5% in the pembrolizumab arm and 66.4% in the placebo arm. In the adjuvant chemotherapy subgroup, the OS HR was 0.91 (95% CI, 0.74-1.12), with 6-year OS rates of 68.9% vs 65.4%; median OS was not reached in either arm. Among patients with disease progression, 12.8% in the pembrolizumab arm and 29.3% in the placebo arm went on to receive subsequent immunotherapy.

LCSS numerically favored pembrolizumab in the ITT population (HR, 0.81; 95% CI, 0.63-1.03), with 6-year rates of 81.3% vs 76.9%, respectively. In patients who had received adjuvant chemotherapy, the LCSS HR was 0.69 (95% CI, 0.53-0.91), and respective 6-year rates were 81.6% vs 75.0%.

The investigators reported numerical imbalances in certain causes of death unrelated to lung cancer, such as cardiovascular disease and COVID-19, and cautioned that the small number of these events limits interpretation. Health-related quality of life did not worsen with pembrolizumab.

What long-term safety findings were reported?

Adverse effects (AEs) of any grade occurred in 95.9% vs 91.0% of patients in the experimental and placebo arms, respectively; grade 3 to 5 AEs were reported at respective rates of 34.0% vs 25.6%, and serious AEs rates were 24.5% vs 15.5%, respectively.

Treatment-related AEs (TRAEs) were reported at respective rates of 75.2% vs 52.5%, with grade 3 to 5 TRAEs at 15.3% vs 4.3%. TRAEs led to treatment discontinuation in 17.1% of patients treated with pembrolizumab vs 3.3% given placebo. Four patients (0.7%) in the pembrolizumab arm died of TRAEs (myocarditis with cardiogenic shock, myocarditis with septic shock, sudden death, and pneumonia) vs none in the placebo arm.

Immune-mediated AEs and infusion reactions affected 40.9% of patients in the pembrolizumab arm vs 13.8% of patients in the placebo group, including grade 3 to 5 AEs at respective rates of 7.9% vs 1.9%. These AEs led to discontinuation rates of 10.3% vs 1.5%, respectviely. Longer follow-up did not reveal any new safety concerns.

References

1. O'Brien ME, Paz-Ares L, Marreaud S, et al. Final analysis of adjuvant pembrolizumab in resected early-stage non–small-cell lung cancer: phase 3 PEARLS/KEYNOTE-091 study. Paper presented at: ESMO Congress 2026; October 23-27, 2026; Madrid, Spain. Abstract 2511O. Accessed October 26, 2026. https://www.eortc.org/blog/2026/10/22/eortc-research-presented-at-esmo-2026/
2. FDA approves pembrolizumab as adjuvant treatment for non-small cell lung cancer. News release. FDA. January 26, 2023. Accessed October 26, 2023. https://www.fda.gov/drugs/resources-information-approved-drugs/fda-approves-pembrolizumab-adjuvant-treatment-non-small-cell-lung-cancer
3. Study of pembrolizumab (MK-3475) vs placebo for participants with non-small cell lung cancer after resection with or without standard adjuvant therapy (MK-3475-091/KEYNOTE-091). ClinicalTrials.gov. Updated March 17, 2026. Accessed October 26, 2026. https://clinicaltrials.gov/study/NCT02504372