Pembrolizumab Plus Chemo Improves PFS in PD-L1+ HR+/HER2– Breast Cancer

At the European Society for Medical Oncology (ESMO) Congress 2026, Hope S. Rugo, MD, of City of Hope Comprehensive Cancer Center in Duarte, California, presented findings from the phase 3 KEYNOTE-B49 trial (NCT04895358), which showed that adding pembrolizumab (Keytruda) to physician’s choice of chemotherapy significantly improved progression-free survival (PFS) vs placebo plus chemotherapy in patients with hormone receptor (HR)–positive, HER2-negative, PD-L1–positive advanced breast cancer that progressed on prior endocrine therapy.¹ The combination did not, however, significantly improve overall survival (OS).

At the interim analysis, with a median follow-up of 29.1 months, the median progression-free survival (PFS) was 6.7 months (95% CI, 6.2-8.3) with pembrolizumab plus chemotherapy vs 6.0 months (95% CI, 4.2-8.1) with placebo plus chemotherapy (HR, 0.74; 95% CI, 0.58-0.95; P = .0089), meeting the prespecified statistical criterion. The 12-month PFS rates were 26.3% vs 20.3%, respectively, and the 24-month rates were 15.0% vs 4.3%. A descriptive updated analysis at a median follow-up of 34.9 months yielded the same HR of 0.74, with 24-month PFS rates of 14.1% vs 3.5%.

In patients with a combined positive score (CPS) of at least 10, a secondary end point, the median PFS was 7.5 months (95% CI, 4.2-10.3) vs 4.4 months (95% CI, 4.0-8.3; HR, 0.59; 95% CI, 0.35-0.98), and 12-month PFS rates were 23.6% vs 9.3%. Investigators reported a generally consistent PFS benefit across other prespecified subgroups, with point estimates that numerically favored pembrolizumab among patients with 1 prior line of endocrine therapy for metastatic disease (HR, 0.53) and those with prior neoadjuvant or adjuvant taxane exposure (HR, 0.55).

At the final analysis, the median OS in the CPS of at least 1 population was 21.0 months (95% CI, 18.9-22.9) with pembrolizumab vs 22.2 months (95% CI, 17.6-26.3) with placebo (HR, 1.04; 95% CI, 0.81-1.33; P = .6185), which did not meet the prespecified criterion. In patients with a CPS of at least 10, the median OS was 22.9 months vs 22.5 months (HR, 0.82; 95% CI, 0.47-1.43). More patients in the placebo arm received subsequent therapy (86.0% vs 73.1%), including antibody-drug conjugates (38.0% vs 29.2%). In a post hoc exploratory analysis that used a 2-stage model to adjust for this imbalance, the OS HR was 0.94 (95% CI, 0.74-1.21), suggesting that subsequent therapy may have confounded the observed OS effect.

The objective response rate in patients with a CPS of at least 1 was 31.6% with pembrolizumab vs 28.1% with placebo, with a median duration of response of 8.3 vs 8.4 months. Among patients with a CPS of at least 10, the response rates were 44.4% vs 30.6%, and the median duration of response was 8.3 vs 6.2 months.

KEYNOTE-B49 randomly assigned 342 patients 1:1 to receive pembrolizumab at 200 mg every 3 weeks or placebo for up to 35 cycles, each in combination with physician’s choice of paclitaxel, nab-paclitaxel, liposomal doxorubicin, or capecitabine. Eligible patients had locally recurrent inoperable or metastatic HR-positive/HER2-negative disease with a PD-L1 CPS of at least 1, had received no prior chemotherapy for metastatic disease, and had progressed on at least 1 prior line of endocrine therapy, typically with a CDK4/6 inhibitor. Randomization was stratified by PD-L1 expression (CPS 1–9 vs CPS ≥10), presence of visceral disease, and planned chemotherapy. The primary end point was PFS by blinded independent central review (BICR) in patients with a CPS of at least 1, and OS in that population was the key secondary end point.

A protocol amendment reduced the planned sample size from 800 to 340 patients because enrollment was slower than expected owing to a lower-than-anticipated prevalence of PD-L1–positive tumors. Of 1767 patients screened, 1425 did not meet eligibility criteria, 59% of them because of negative PD-L1 expression. Following the amendment, statistical power remained above 80% for PFS but fell to approximately 65% for OS.

Baseline characteristics were balanced between the arms. The median age was 58 years in the pembrolizumab arm and 57 years in the placebo arm, 78.9% of patients in each arm had a CPS of 1 to 9, roughly 85% had visceral metastases, and nearly all (99.4% and 98.8%, respectively) had received a prior CDK4/6 inhibitor. Capecitabine was the most commonly used on-study chemotherapy, given to 53.2% of patients in each arm.

Treatment-related adverse events (AEs) occurred in 91.2% of patients in the pembrolizumab arm and 90.5% of those in the placebo arm, including grade 3 or higher events in 37.6% and 32.7%, respectively. Serious treatment-related AEs occurred in 13.5% vs 7.7% of patients, and treatment-related AEs led to discontinuation in 11.2% vs 5.4%. Three treatment-related deaths (1.8%) occurred in the pembrolizumab arm, attributed to autoimmune hepatitis, gastrointestinal inflammation, and pneumonitis; none occurred in the placebo arm. The most common treatment-related AEs with pembrolizumab plus chemotherapy were palmar-plantar erythrodysesthesia (30.6%), anemia (29.4%), and nausea (25.9%).

Immune-mediated AEs and infusion reactions occurred in 37.1% of patients receiving pembrolizumab vs 14.9% of those receiving placebo, with grade 3 or higher events in 4.7% vs 1.8%. The most frequent were hypothyroidism (20.0% vs 3.0%), hyperthyroidism (9.4% vs 1.2%), and pneumonitis (4.7% vs 2.4%).

The investigators concluded that KEYNOTE-B49, the largest study to date of immunotherapy plus chemotherapy in this endocrine-resistant, PD-L1–positive population, met its primary end point, and that the safety profile was consistent with the known profiles of each regimen with no new signals. They noted that interpretation of the OS findings is limited by the reduced statistical power. The ongoing phase 3 TroFuse-010 trial (NCT06312176) is further evaluating pembrolizumab in combination with sacituzumab tirumotecan in both the overall population and a PD-L1–defined subgroup.

Reference

Rugo HS, Kok M, Goncalves A, et al. KEYNOTE-B49: a phase 3, randomized, double-blind study of pembrolizumab vs placebo plus chemotherapy in participants with HR+/HER2− PD-L1–positive advanced breast cancer. Presented at: European Society for Medical Oncology Congress 2026; October 23-27, 2026; Madrid, Spain.