Domvanalimab/Zimberelimab Plus Chemotherapy Fails to Improve OS in HER2-Negative Gastroesophageal Adenocarcinoma

The addition of the Fc-silent anti-TIGIT antibody domvanalimab to the anti–PD-1 antibody zimberelimab plus chemotherapy did not improve overall survival (OS) vs nivolumab (Opdivo) plus chemotherapy as first-line treatment for patients with advanced HER2-negative gastric, gastroesophageal junction (GEJ), or esophageal adenocarcinoma, missing the primary end point of the phase 3 STAR-221 trial (NCT05568095) presented at the 2026 ESMO Gastrointestinal (GI) Cancers Congress.¹

In the intention-to-treat (ITT) population, the median OS was 14.0 months (95% CI, 13.1-15.7) with domvanalimab plus zimberelimab and chemotherapy (n = 521) vs 13.4 months (95% CI, 12.3-15.4) with nivolumab plus chemotherapy (n = 519; HR, 1.01; 95% CI, 0.86-1.17; P = .526). No OS benefit was observed in the PD-L1–high population (tumor area positivity [TAP] ≥ 5%), where the median OS was 15.2 months (95% CI, 13.4-18.3) vs 17.2 months (95% CI, 13.3-19.9), respectively (HR, 1.10; 95% CI, 0.87-1.39; P = .795). In the PD-L1–positive population (TAP ≥ 1%), the median OS was 14.7 months in both arms (HR, 0.99; 95% CI, 0.83-1.18); per the prespecified testing procedure, statistical significance for OS was not formally assessed in this population.

“In this study, there was no clear analysis population with particular benefit from the domvanalimab plus zimberelimab and chemotherapy combination for patients with advanced gastroesophageal adenocarcinoma,” presenting study author Sun Young Rha, MD, PhD, said in the presentation. Rha is a professor of medical oncology in the Department of Internal Medicine and director of Songdang Institute for Cancer Research, Yonsei University College of Medicine, Yonsei University Health System in Seoul, Korea.

Why do the STAR-221 results matter for first-line gastroesophageal cancer?

The addition of anti–PD-1 antibodies to chemotherapy has improved survival in advanced gastroesophageal adenocarcinoma and established nivolumab plus chemotherapy and pembrolizumab (Keytruda) plus chemotherapy as first-line standards of care based on the phase 3 CheckMate 649 (NCT02872116) and KEYNOTE-859 (NCT03675737) trials, respectively.^2,3 Despite these advances, median OS remains approximately 13 to 15 months, leaving a need for regimens that build on the PD-1 backbone.¹

TIGIT is an inhibitory immune checkpoint expressed on activated T cells and natural killer cells that limits antitumor immunity, providing a rationale for combining TIGIT and PD-1 blockade. That rationale was supported by data from the phase 2 EDGE-Gastric A1 trial (NCT05329766), in which first-line domvanalimab plus zimberelimab and FOLFOX (leucovorin, fluorouracil, and oxaliplatin) generated a confirmed objective response rate (ORR) of 59% (90% CI, 44.5%-71.6%) and a median OS of 26.7 months (90% CI, 18.4-not estimable) among 41 treated patients.⁴ STAR-221 was designed to test whether that signal would translate into a survival advantage over an active PD-1 inhibitor–based comparator. In December 2025, the trial's sponsors announced that STAR-221 would be discontinued for futility after a prespecified interim OS analysis showed no improvement over nivolumab plus chemotherapy; the ESMO GI presentation reported the full dataset behind that decision.⁵

How was the STAR-221 trial designed?

STAR-221 was a global, randomized, open-label, phase 3 trial that enrolled patients at least 18 years of age with histologically confirmed, locally advanced unresectable or metastatic HER2-negative gastric, GEJ, or esophageal adenocarcinoma; an ECOG performance status of 0 or 1; measurable disease per RECIST 1.1; and no prior systemic therapy for advanced disease.¹ A total of 1040 patients were randomly assigned 1:1 to domvanalimab plus zimberelimab and chemotherapy or to nivolumab plus chemotherapy, with stratification by PD-L1 expression (TAP ≥ 5% vs < 5%), ECOG performance status (0 vs 1), and region (Asia vs Canada/EU5/US vs rest of world).

In the investigational arm, domvanalimab and zimberelimab were administered every 4 weeks with FOLFOX every 2 weeks or with CAPOX (capecitabine and oxaliplatin) every 3 weeks. The primary end point was OS, assessed in the ITT, PD-L1–high (TAP ≥ 5%), and PD-L1–positive (TAP ≥ 1%) populations. Key secondary end points were progression-free survival (PFS) and time to first deterioration (TTD) by FACT-Ga; other secondary end points included ORR, duration of response (DOR), and safety. Baseline characteristics were balanced between arms; most patients had gastric cancer and metastatic disease, and approximately 80% had PD-L1–positive tumors.

What did the efficacy analyses show across PD-L1 populations?

PFS also did not favor the domvanalimab combination. In the ITT population, the median PFS was 8.4 months (95% CI, 7.3-9.6) with domvanalimab plus zimberelimab and chemotherapy vs 8.3 months (95% CI, 7.2-9.0) with nivolumab plus chemotherapy (HR, 1.00; 95% CI, 0.87-1.16). Corresponding HRs were 1.00 (95% CI, 0.80-1.25) in the PD-L1–high and 0.98 (95% CI, 0.83-1.16) in the PD-L1–positive populations. Because the OS testing gate was not met, statistical significance was not assessed for PFS.

Response rates were comparable between arms. In the ITT population, the ORR was 55% (95% CI, 50.1%-58.8%) with the domvanalimab combination vs 56% (95% CI, 51.7%-60.4%) with nivolumab plus chemotherapy, with a median DOR of 10.5 months (95% CI, 9.5-12.4) vs 9.5 months (95% CI, 8.1-11.2), respectively. The one measure that numerically favored the domvanalimab combination was TTD by FACT-Ga: the median TTD was 8.4 months (95% CI, 6.8-11.7) vs 6.5 months (95% CI, 5.1-8.9) in the ITT population (HR, 0.78; 95% CI, 0.63-0.97), with similar HRs of 0.70 in both the PD-L1–high (95% CI, 0.52-0.96) and PD-L1–positive (95% CI, 0.55-0.89) populations. Significance was not assessed for TTD per the testing procedure.

What did the safety analysis show?

The safety profile of domvanalimab plus zimberelimab and chemotherapy was consistent with that of anti–PD-(L)1 plus chemotherapy and was similar between arms. Among 522 and 507 patients in the respective safety populations, any-grade adverse effects (AEs) occurred in 98% and 97% of patients, and grade 3/4 AEs occurred in 63% of patients in both arms. AEs led to death in 6% vs 8% of patients and to discontinuation of any study drug in 48% vs 47%, respectively. Infusion-related reactions were less frequent with the domvanalimab combination (15% vs 27%), and immune-mediated AEs occurred in 22% vs 24% of patients.

“These findings suggest that safety did not contribute to the lack of any additional clinical benefit with domvanalimab plus zimberelimab and chemotherapy compared with nivolumab plus chemotherapy,” Rha concluded.

Disclosures: Rha reported receiving grants or contracts for other studies from ABL Bio, Amgen, Arcus Biosciences, Astellas Pharma, AstraZeneca, BeOne, Bold Therapeutics, Celltrion, Daiichi Sankyo, DongA, Eisai, Henlius, Ipsen, Jazz, Merck KGaA, Merck Sharp and Dohme, Ono Pharmaceutical, Pfizer, and Taiho Pharmaceutical. She reported serving in consulting, steering committee, and advisory board roles for Amgen, AstraZeneca, Astellas Pharma, Arcus Biosciences, BeOne, Celltrion, Daiichi Sankyo, Gilead, Henlius, Indivumed, LG Chem, Eisai, Merck, MSD, Takeda, and Toray.

References

1. Janjigian YY, Rha SY, Wainberg ZA, et al. STAR-221: a phase 3 study of first-line domvanalimab, zimberelimab, and chemotherapy vs nivolumab + chemotherapy in advanced HER2-negative gastric, gastroesophageal junction, or esophageal adenocarcinoma. Presented at: 2026 ESMO Gastrointestinal Cancers Congress; July 1-4, 2026; Munich, Germany. Abstract 493O.
2. Janjigian YY, Shitara K, Moehler M, et al. First-line nivolumab plus chemotherapy versus chemotherapy alone for advanced gastric, gastro-oesophageal junction, and oesophageal adenocarcinoma (CheckMate 649): a randomised, open-label, phase 3 trial. Lancet. 2021;398(10294):27-40. doi:10.1016/S0140-6736(21)00797-2
3. Rha SY, Oh DY, Yañez P, et al. Pembrolizumab plus chemotherapy versus placebo plus chemotherapy for HER2-negative advanced gastric cancer (KEYNOTE-859): a multicentre, randomised, double-blind, phase 3 trial. Lancet Oncol. 2023;24(11):1181-1195. doi:10.1016/S1470-2045(23)00515-6
4. Janjigian YY, Oh DY, Pelster M, et al. Domvanalimab and zimberelimab in advanced gastric, gastroesophageal junction or esophageal cancer: a phase 2 trial. Nat Med. 2025;31(12):4274-4280. doi:10.1038/s41591-025-04022-w
5. Flaherty C. Phase 3 STAR-221 study of first-line domvanalimab/zimberelimab plus chemo in upper GI cancers to be discontinued. OncLive. Published December 12, 2025. Accessed July 7, 2026. https://www.onclive.com/view/phase-3-star-221-study-of-first-line-domvanalimab-zimberelimab-plus-chemo-in-upper-gi-cancers-to-be-discontinued