KEYNOTE-859 Data Show 1.4-Month OS Benefit With Frontline Pembrolizumab Plus Chemo in Advanced HER2– Gastric or GEJ Cancer


The addition of pembrolizumab to chemotherapy resulted in a significant and clinically meaningful improvement in overall survival, progression-free survival, and objective response rate vs chemotherapy alone when used as first-line treatment in patients with advanced HER2-negative gastric or gastroesophageal junction adenocarcinoma.

gastric cancer

The addition of pembrolizumab (Keytruda) to chemotherapy resulted in a significant and clinically meaningful improvement in overall survival (OS), progression-free survival (PFS), and objective response rate (ORR) vs chemotherapy alone when used as first-line treatment in patients with advanced HER2-negative gastric or gastroesophageal junction (GEJ) adenocarcinoma, according to data from the phase 3 KEYNOTE-859 trial (NCT03675737).

In the intention-to-treat (ITT) population, the median OS in the investigative arm was 12.9 months (95% CI, 11.9-14.0) vs 11.5 months (95% CI, 10.6-12.1) in the control arm, translating to a 22% reduction in the risk of death (HR, 0.78; 95% CI, 0.70-0.87; P < .0001). The median PFS was 6.9 months (95% CI, 6.3-7.2) with pembrolizumab plus chemotherapy and 5.6 months (95% CI, 5.5-5.7) with chemotherapy alone, equating to a 24% reduction in the risk of disease progression or death (HR, 0.76; 95% CI, 0.67-0.85; P < .0001).

The ORR achieved with pembrolizumab plus chemotherapy was 51.3% (95% CI, 47.7%-54.8%) vs 42.0% (95% CI, 38.5%-45.5%) with chemotherapy alone, which translated to a 9.3%-increase (95% CI, 4.4%-14.1%; P = .00009). The median duration of response (DOR) in the investigative arm was 8.0 months (range, 1.2+ to 41.5+) vs 5.7 months (range, 1.3+ to 34.7+) in the control arm.

“[These] data support pembrolizumab plus chemotherapy as a new treatment option for this patient population,” Sun Young Rha, MD, PhD, of Yonsei Cancer Center, Yonsei University Health System, Seoul, Republic of Korea, said in a presentation of the data during the 2023 ESMO Plenary session.

Gastric cancer still has a large morbidity and mortality burden worldwide, according to Rha, who added that the 5-year survival rate for those with advanced or metastatic disease is less than 10% despite available options. She added that chemotherapy is the main component in the frontline setting, irrespective of tumor histology or HER2 status. The standard regimen is comprised of a fluoropyrimidine and a platinum.

Phase 3 trials have evaluated the combination of PD-1 inhibitors like nivolumab (Opdivo), sintilimab (Tyvyt), and pembrolizumab and chemotherapy in patients with gastric cancer, in both the HER2-negative and -positive settings. “Although regulatory approvals and guideline recommendations are different globally, immunotherapy plus chemotherapy is becoming the standard of care for the first-line treatment of metastatic gastric cancer,” Rha said.

The randomized, double-blind, phase 3 KEYNOTE-859 trial enrolled patients with histologically or cytologically confirmed adenocarcinoma of the stomach or GEJ who had locally advanced unresectable or metastatic disease. They needed to have an ECOG performance status of 0 or 1, known PD-L1 status, and HER2-negative status. They could not have previously received treatment.

Study participants were randomly assigned 1:1 to receive pembrolizumab or placebo intravenously (IV) at 200 mg every 3 weeks for up to 35 cycles plus chemotherapy. In terms of chemotherapy, they could have received IV 5-fluororacil at 800 mg/m2 given continuously on days 1 to 5 every 3 weeks plus IV cisplatin at 80 mg/m2 every 3 weeks (FP) or oral capecitabine at 1000 mg/m2 twice daily on days 1 to 14 every 3 weeks plus IV oxaliplatin at 130 mg/m2 every 3 weeks (CAPOX).

Patients were stratified based on geographic region (Europe/Israel/North America/Australia vs Asia vs rest of world), PD-L1 combined positive score (CPS; <1 vs ≥1), and choice of chemotherapy (FP vs CAPOX).

OS served as the primary end point of the trial, and secondary end points comprised PFS, ORR, DOR, and safety. Response was assessed using RECIST v1.1 criteria and blinded independent central review.

The planned enrollment for the trial is approximately 1542 patients. Efficacy was evaluated in the ITT population, and safety was evaluated in all-randomized patients who received at least 1 dose of study treatment. The family-wise error rate was strongly controlled at 1-sided 2.5% across all hypotheses.

“There are 9 hypotheses total, which included OS, PFS, and ORR,” Rha said. “Each was assessed in the overall population, in patients with PD-L1 CPS ≥1, and in patients with PD-L1 CPS ≥10.” The alpha was initially split between testing OS in the ITT population (α = 0.8%) and the PD-L1 CPS ≥10 population (α = 1.7%).

The data cutoff date for the interim analysis was October 3, 2022. The P value boundaries for demonstrating statistical significance in the overall ITT population were P = .006079 for OS, P = .025 for PFS, and P = .025 for ORR.

At the meeting, Rha only shared findings for the ITT population, and noted that data in the PD-L1 subgroups will be presented at a later date. At a median follow-up of 31.0 months (range, 15.3-46.3), a total of 1579 patients were randomly allocated and included in the ITT population.

A total of 790 patients were allocated to the investigative arm and 789 patients were allocated to the control arm. In the pembrolizumab arm, 785 patients received treatment, 60 completed treatment, and 685 discontinued study treatment. The most common reason for discontinuation was clinical or radiographic progression (n = 515), followed by toxicity (n = 103), withdrawn consent (n = 39), physician decision (n = 13), and another unspecified reason (n = 12). Forty patients are still receiving treatment.

“The baseline demographics and disease characteristics were as expected for an advanced gastric cancer population, and were balanced between the arms,” Rha said.

In the ITT population, the median age was 61.5 years (range, 21-86), with 38.9% of patients aged 65 years and older. More than half of patients were male (67.8%) and had an ECOG performance status of 1 (63.2%). Approximately 33% of patients were from Asia; ~25% from Europe, Israel, North America, or Australia; and ~41% from the rest of the world. “Approximately 5% of patients were known to have microsatellite instability–high [MSI-H] tumors,” Rha said. Regarding PD-L1 status, 78.2% had a CPS of 1 or higher and 34.9% had a CPS of 10 or higher. Moreover, 86.3% of patients received CAPOX at randomization and 13.7% had FP.

Additional data showed that the 24-month OS rate in the pembrolizumab arm was 28.2% vs 18.9% in the placebo arm. The addition of pembrolizumab to chemotherapy was observed across prespecified subgroups. “Adding pembrolizumab provided benefit regardless of region of enrollment, choice of chemotherapy, primary tumor location, or histologic subgroup,” Rha said. “Not surprisingly, based on previous studies of immune checkpoint inhibitors in gastric cancer, patients with MSI-high tumors appear to have greater relative benefit with the addition of pembrolizumab [HR, 0.34].”

The 24-month PFS rates in the pembrolizumab and placebo arms were 17.8% and 9.4%, respectively. Again, the benefit of adding pembrolizumab to chemotherapy was observed across the prespecified subgroups. “The 95% CIs for all subgroups overlap [in] the total population, with the exception of patients with MSI[-H] tumors for whom there appears to be a larger benefit with the addition of pembrolizumab,” Rha added.

Regarding safety, 95.7% of patients in the pembrolizumab/chemotherapy arm experienced any-grade treatment-related adverse effects (TRAEs) vs 93.5% of those in the chemotherapy-alone arm; these effects were grade 3 to 5 in 59.4% and 51.1% of patients, respectively. Moreover, 23.4% of those in the pembrolizumab arm experienced serious TRAEs vs 18.6% of those in the placebo arm. Eight patients on the investigative arm died, and 16 patients on the control arm died. TRAEs resulted in treatment discontinuation in 26.4% and 20.1% of patients, respectively.

The most common TRAEs to be reported in 15% or more of patients in the investigative and control arms were nausea (41.4% vs 41.4%, respectively), diarrhea (32.1% vs 27.2%), anemia (31.0% vs 26.9%), vomiting (27.4% vs 22.2%), decreased platelet count (25.0% vs 22.5%), decreased neutrophil count (24.6% vs 21.6%), palmar-plantar erythrodysesthesia (24.1% vs 21.1%), decreased appetite (21.4% vs 21.3%), fatigue (20.0% vs 20.5%), peripheral neuropathy (19.1% vs 20.8%), neutropenia (18.1% vs 17.2%), increased aspartate aminotransferase (17.7% vs 13.0%), and peripheral sensory neuropathy (17.5% vs 16.6%).

Any-grade immune-mediated toxicities were experienced by 27.1% of those in the investigative arm vs 9.3% of those in the control arm; these effects were grade 3 to 5 in 7.9% vs 1.7%, respectively. The most common immune-mediated AEs reported in 2 or more of patients in the pembrolizumab and placebo arms were hypothyroidism (15.3% vs 4.3%, respectively), hyperthyroidism (5.6% vs 1.7%), colitis (3.3% vs 1.8%), pneumonitis (3.2% vs 0.9%), severe skin reactions (2.0% vs 0.1%), adrenal insufficiency (1.3% vs 0.1%), hepatitis (1.1% vs 0.5%), thyroiditis (0.4% vs 0%), type 1 diabetes (0.6% vs 0.1%), nephritis (0.5% vs 0%), hypophysitis (0.4% vs 0%), pancreatitis (0.4% vs 0.4%), and vasculitis (0.3% vs 0.1%). One patient in each arm died.

“Safety profile was as expected for a regimen containing pembrolizumab and chemotherapy,” Rha concluded.

Editor’s Note: Dr Rha received study funding from Arcus, Amgen, AstraZeneca, Beigene, BMS, Daiichi Sankyo, Eisai, Incyte, Indivumed, Lilly, Merck, MSD, and Zymeworks. She is a steering committee member for Amgen. She also receives consulting fees from Amgen, Astellas, Daiichi Sankyo, Eisai, Indivumed, LG Biochemical, and MSD. The KEYNOTE-859 study and medical writing and editorial assistance in support of the presentation were funded by Merck Sharp & Dohme LLC, which is a subsidiary of Merck & Co., Inc.


Rha SY, Wyrwicz LS, Weber Y, et al. Pembrolizumab plus chemotherapy as first-line therapy for advanced HER2-negative gastric or gastroesophageal junction cancer: phase 3 KEYNOTE-859 study. Ann Oncol. Published online February 16, 2023. doi:10.1016/j.annonc.2023.01.006

Related Videos
Dae Won Kim, MD, Gastrointestinal Oncology Program, Moffitt Cancer Center
Michael J. Overman, MD, The University of Texas MD Anderson Cancer Center,
John Michael Bryant, MD,
Jacob Shreve, MD, MS, hematology/oncology fellow, Mayo Clinic
Efrat Dotan, MD, Fox Chase Cancer Center
A panel of 4 experts on gastrointestinal cancers
A panel of 4 experts on gastrointestinal cancers
Zev A. Wainberg, MD
A panel of 4 experts on gastrointestinal cancers
A panel of 4 experts on gastrointestinal cancers