Pembrolizumab Plus SBRT Fails to Improve EFS in Unresected Early-Stage NSCLC

Adding the PD-1 inhibitor pembrolizumab (Keytruda) to stereotactic body radiotherapy (SBRT) did not improve event-free survival (EFS) and increased treatment-related toxicity in patients with unresected stage I or II non–small cell lung cancer (NSCLC), according to results from a randomized phase 3 study published in the Journal of Clinical Oncology.¹ Merck halted the trial for futility in August 2024 on the recommendation of an independent data monitoring committee after its first interim analysis, and all study treatment was discontinued.^1,2

In the double-blind KEYNOTE-867 trial (NCT03924869), median EFS by blinded independent central review was 31.2 months (95% CI, 22.5-39.1) with SBRT plus pembrolizumab vs 28.3 months (95% CI, 19.8-33.1) with SBRT plus placebo (HR, 0.92; 95% CI, 0.69-1.24; P = .29). Grade 3 to 5 treatment-related adverse events (TRAEs) occurred in 20.4% vs 3.7% of patients, respectively, and all 5 treatment-related deaths occurred in the pembrolizumab arm.¹

Study Design

KEYNOTE-867 enrolled adults at 193 sites worldwide with untreated stage I or II (T1 to limited T3, N0, M0) NSCLC that was medically inoperable or for which the patient declined surgery. Patients were randomly assigned 1:1 to pembrolizumab 200 mg or placebo every 3 weeks for up to 17 cycles, with SBRT to the primary tumor delivered over no more than 2 weeks starting on day 1 of cycle 1. Fractionation depended on tumor location, and the most common regimen in both arms was 50 Gy in 5 fractions. Randomization was stratified by stage, ECOG performance status, region, and reason for not undergoing surgery.

EFS, defined as time to local, regional, or distant recurrence or death from any cause, was the primary end point. Secondary end points included overall survival (OS), time to distant metastasis or death, safety, and patient-reported outcomes.

Between September 2019 and January 2024, 448 patients were randomly assigned to pembrolizumab (n = 227) or placebo (n = 221). The median age was 73.0 years; 14.7% of patients were younger than 65 years. The majority, 82.4%, were deemed medically inoperable. Tumors 2 cm or smaller (stage IA1 or IA2) accounted for 39.3% of the population. At the June 11, 2024, data cutoff, the median time from randomization was 20.6 months (range, 1.3-56.8).

Survival Outcomes

OS was not formally tested because EFS did not reach statistical significance. The OS HR was 1.33 (95% CI, 0.93-1.90), with OS events occurring in 30.4% of the pembrolizumab arm vs 24.4% of the placebo arm. The median OS was 54.7 months (95% CI, 33.5-not reached [NR]) with pembrolizumab and not reached (95% CI, 44.4-NR) with placebo.

Time to distant metastasis or death did not differ between arms (median, 39.1 vs 41.3 months; HR, 1.02; 95% CI, 0.74-1.42). At the end-of-study analysis (data cutoff, January 20, 2025), disease-specific survival was also not improved (HR, 1.44; 95% CI, 0.75-2.76).

In subgroup analyses, EFS and OS results were generally consistent with the overall population. Point estimates favored pembrolizumab in patients younger than 65 years (EFS HR, 0.48; 95% CI, 0.18-1.24; OS HR, 0.53; 95% CI, 0.18-1.56), but this group included only 66 patients, the confidence intervals were wide, and the trial was not designed to assess outcomes by age. Among patients 65 years or older, the OS HR was 1.53 (95% CI, 1.04-2.24).

Safety Findings

TRAEs occurred in 73.5% of patients who received pembrolizumab and 50.0% of those who received placebo. Serious TRAEs occurred in 14.2% vs 2.3%, and TRAEs led to discontinuation of pembrolizumab or placebo in 15.9% vs 2.3%; no patients discontinued radiation because of TRAEs. The most frequent TRAEs with pembrolizumab vs placebo were hypothyroidism (13.7% vs 2.8%), rash (12.4% vs 5.0%), pruritus (11.9% vs 5.5%), and fatigue (11.5% vs 8.7%).

The 5 treatment-related deaths in the pembrolizumab arm were due to pneumonia (n = 3), bronchopneumopathy (n = 1), and pneumonitis (n = 1). All-cause pneumonitis was more common with pembrolizumab (20.8% vs 10.6%), whereas radiation pneumonitis rates were similar (9.7% vs 9.2%). Seven patients (3.1%), all in the pembrolizumab arm, had grade 3 to 5 pneumonitis or radiation pneumonitis. Global health status scores declined from baseline to week 24 in both arms without a difference between them (difference in least-squares mean of −1.21 [95% CI, −4.90 to 2.48]; P = .52).

Clinical Context and Limitations

The findings contrast with those of a randomized phase 2 study (NCT03110978) in which adding nivolumab (Opdivo) to stereotactic ablative radiotherapy significantly improved 4-year EFS in early-stage or isolated parenchymal recurrent node-negative NSCLC.^1,3 The authors noted that the phase 2 study used a shorter course of nivolumab (4 to 7 cycles every 4 weeks), included patients with recurrent disease, and enrolled a higher proportion of patients with tumors 2 cm or smaller (60.9% vs 39.3%). In that trial, a subgroup analysis also suggested EFS benefit in patients at or below the median age of 72 years but not in older patients.¹

The authors also pointed to a preliminary report from the phase 3 SWOG/NRG S1914 trial (NCT04214262) of SBRT with or without atezolizumab (Tecentriq) in high-risk early-stage NSCLC, which showed no difference in OS or progression-free survival and more grade 3 to 5 adverse events with atezolizumab.⁴ Full results from S1914 and from the phase 3 PACIFIC-4 trial (NCT03833154) of SBRT with durvalumab (Imfinzi) or placebo are expected to further define the role of immunotherapy with SBRT in this setting.

Results were reported at a prespecified interim analysis that became the final analysis, and follow-up was relatively short for an early-stage population.

“Further investigations are needed to evaluate the potential utility of adding immunotherapy to SBRT in younger patients,” wrote Salma K. Jabbour, MD, of the Department of Radiation Oncology at Rutgers Cancer Institute, Rutgers University, and colleagues, in the study.¹

REFERENCES

1. Jabbour SK, Sotér S, Eaton M, et al. KEYNOTE-867: a phase III, randomized, double-blind study of stereotactic body radiotherapy with pembrolizumab for unresected stage I or II non–small cell lung cancer. J Clin Oncol. Published online October 28, 2026. doi:10.1200/JCO-26-00404

2. Merck provides update on phase 3 KEYNOTE-867 and KEYNOTE-630 trials. News release. Merck. August 29, 2024. Accessed November 6, 2026. https://tinyurl.com/9f2tp94p

3. Chang JY, Lin SH, Dong W, et al. Stereotactic ablative radiotherapy with or without immunotherapy for early-stage or isolated lung parenchymal recurrent node-negative non-small-cell lung cancer: an open-label, randomised, phase 2 trial. Lancet. 2023;402(10405):871-881. doi:10.1016/S0140-6736(23)01384-3

4. Daly ME, Simone CB 2nd, Redman MW, et al. Induction and consolidation atezolizumab with stereotactic body radiation therapy versus radiation alone in high-risk, early-stage non-small-cell lung cancer (SWOG/NRG S1914): a multicentre, open-label, superiority, phase 3, randomised controlled trial. Lancet. 2026;408(10562):1313-1325. doi:10.1016/S0140-6736(26)01655-7