Highlights
– Dual TIGIT (domvanalimab) and PD‑1 (zimberelimab) blockade plus FOLFOX showed an objective response rate (ORR) of 59% (90% CI 44.5–71.6%) in 41 patients with previously untreated, HER2‑negative advanced gastric, gastroesophageal junction or esophageal adenocarcinoma (EDGE‑Gastric cohort A1).
– Median progression‑free survival (PFS) was 12.9 months (90% CI 9.8–14.6) and median overall survival (OS) was 26.7 months (90% CI 18.4–NE). PD‑L1 positive (TAP ≥1%) and high (TAP ≥5%) subgroups showed numerically higher ORR and longer PFS/OS.
– Immune‑related adverse events (irAEs) occurred in 27% of patients; overall safety aligned with expectations for anti‑PD‑1 therapy combined with platinum‑based chemotherapy. The regimen is now being tested in the phase 3 STAR‑221 trial (NCT05329766).
Background and Disease Burden
Advanced gastric and gastroesophageal junction (GEJ) adenocarcinomas carry a high global morbidity and mortality burden. Despite incremental advances, median overall survival in the first‑line metastatic setting historically ranged from ~9–12 months with chemotherapy alone. The addition of anti‑PD‑1 therapy to first‑line chemotherapy has become standard for many patients, improving survival for PD‑L1 positive tumors. Nonetheless, response rates and durability remain suboptimal for a large portion of patients, and new strategies to augment antitumor immunity are needed.
TIGIT (T cell immunoreceptor with Ig and ITIM domains) is an inhibitory receptor expressed on T cells and natural killer (NK) cells that can dampen antitumor immune responses. Preclinical and translational data suggest TIGIT blockade can synergize with PD‑1/PD‑L1 inhibition by restoring effector T cell and NK cell function, providing a strong rationale for dual checkpoint blockade in cancers where T cell dysfunction contributes to immune escape.
Study Design
The EDGE‑Gastric study is a multicenter, international phase 2 trial comprising three cohorts across first‑line and later‑line settings. Cohort A is a first‑line cohort containing four arms (A1–A4); arm A1 was a single‑arm evaluation of the combination of domvanalimab (an Fc‑silent anti‑TIGIT monoclonal antibody) and zimberelimab (anti‑PD‑1) added to standard FOLFOX chemotherapy in patients with previously untreated, advanced HER2‑negative gastric, GEJ, or esophageal adenocarcinoma.
Key eligibility elements included measurable disease and no prior systemic therapy for advanced disease. Forty‑one patients were treated in arm A1. Primary and secondary endpoints included objective response rate (ORR), progression‑free survival (PFS), overall survival (OS), and safety. PD‑L1 expression was assessed using tumor area positivity (TAP) with predefined cutoffs of ≥1% (PD‑L1 positive) and ≥5% (PD‑L1 high) to explore biomarker associations.
Key Findings and Clinical Results
Population: Forty‑one treated patients comprised the evaluable cohort for efficacy and safety analyses reported for arm A1.
Antitumor activity
– Confirmed ORR (primary efficacy readout): 59% (90% CI 44.5–71.6%). Responses included complete and partial responses per RECIST‑type criteria (detailed breakdown not provided in the report abstract).
– Median PFS: 12.9 months (90% CI 9.8–14.6 months). This indicates a durable disease control signal compared with historical controls of chemotherapy alone and aligns favorably with prior first‑line PD‑1 plus chemotherapy studies in selected populations.
– Median OS: 26.7 months (90% CI 18.4 months–not estimable). Median OS exceeded typical first‑line chemotherapy outcomes and is notable given the small sample size and single‑arm design; however, follow‑up duration and confidence interval censoring (NE) indicate immaturity of the OS estimate.
PD‑L1 subgroup analyses
The authors reported efficacy stratified by PD‑L1 tumor area positivity (TAP):
– TAP ≥1% (PD‑L1 positive): ORR 62% (90% CI 45.1–77.1%); median PFS 13.2 months (90% CI 11.3–15.2); median OS 26.7 months (90% CI 19.5–NE).
– TAP ≥5% (PD‑L1 high): ORR 69% (90% CI 45.2–86.8%); median PFS 14.5 months (90% CI 11.3 months–NE); median OS not reached (90% CI 17.4 months–NE).
These numerically higher response rates and time‑to‑event outcomes in PD‑L1 enriched tumors are biologically plausible and consistent with PD‑1–directed therapy experience; however, small subgroup sizes limit inference and formal statistical comparison.
Safety
Immune‑related adverse events were reported in 27% of patients. The authors state the safety profile was consistent with known toxicities for anti‑PD‑1 agents combined with platinum‑based chemotherapy. No unexpected or novel safety signals attributed specifically to TIGIT blockade were highlighted in the summary. Details on grade ≥3 toxicities, treatment discontinuations, immune‑related serious adverse events, and laboratory abnormalities are necessary for comprehensive risk appraisal and will be informative in the phase 3 program.
Interpretation and Mechanistic Rationale
The observed ORR of 59% and median PFS of 12.9 months with domvanalimab plus zimberelimab and FOLFOX are encouraging in this phase 2, single‑arm cohort. The addition of TIGIT blockade to PD‑1 inhibition aims to further reverse T cell exhaustion and potentially engage NK cells, broadening immune effector mechanisms beyond PD‑1 alone. The numerical enhancement of responses in PD‑L1 TAP ≥1% and ≥5% subgroups suggests that PD‑L1 expression remains an important biomarker of benefit, while TIGIT blockade might augment responses across PD‑L1 strata; confirmatory data are required.
Notably, the TIGIT antibody used (domvanalimab) is Fc‑silent, designed to avoid FcγR‑mediated depletion of TIGIT‑expressing effector cells, a mechanistic design choice intended to preserve antitumor T cell and NK cell populations while relieving inhibitory signaling.
Expert Commentary and Limitations
Strengths of the report include a clear biological rationale for dual checkpoint blockade, prospectively collected PD‑L1 biomarker data, and clinically meaningful outcome estimates that support further testing. However, important limitations temper enthusiasm and must be emphasized:
- The cohort is small (n=41) and single‑arm, which precludes definitive comparisons with established first‑line regimens such as PD‑1 plus chemotherapy in randomized trials.
- The reported confidence intervals use 90% bounds rather than the conventional 95% limits, which provides less conservative uncertainty estimates.
- Follow‑up duration appears immature for OS; median OS confidence intervals extend to “not estimable,” and longer observation is needed to fully characterize survival benefit and late toxicities.
- Detailed safety data (frequency of grade ≥3 AEs, immune‑related serious events, dose modifications, and discontinuations) were not presented in the summary and are essential for clinical risk–benefit assessment.
- Generalizability may be limited by patient selection and geographic heterogeneity; phase 3 randomized data are required to establish comparative effectiveness and identify subgroups that derive the greatest benefit.
Given these caveats, the ongoing randomized phase 3 STAR‑221 trial will be pivotal to determine whether TIGIT plus PD‑1 blockade adds clinically meaningful benefit over standard PD‑1 plus chemotherapy regimens and to clarify safety in a larger, controlled population.
Clinical Implications and Future Directions
If the phase 3 program confirms improvement in PFS and OS with acceptable toxicity, dual TIGIT/PD‑1 blockade could become an important addition to first‑line systemic therapy for selected patients with advanced gastric and GEJ adenocarcinoma. Key questions for future work include:
- Which biomarker(s) — PD‑L1 TAP, TIGIT expression, tumor mutational burden, immune gene signatures — best predict benefit from TIGIT co‑blockade?
- What is the incremental toxicity attributable to TIGIT inhibition when combined with PD‑1 blockade and chemotherapy, particularly for immune‑related organ toxicities?
- Can TIGIT blockade improve long‑term durable responses and the proportion of patients achieving deep, sustained remissions?
- What is the role of TIGIT inhibition in later lines of therapy or as part of novel combinations (e.g., with anti‑angiogenic agents, targeted therapies, or other immune modulators)?
Conclusion
The EDGE‑Gastric cohort A1 results indicate that domvanalimab (anti‑TIGIT) plus zimberelimab (anti‑PD‑1) with FOLFOX demonstrates promising efficacy and a manageable safety profile in first‑line HER2‑negative advanced gastroesophageal adenocarcinoma. These phase 2 data justify further investigation in randomized phase 3 trials to determine whether dual TIGIT/PD‑1 blockade delivers clinically meaningful improvements over current standards of care and to define predictive biomarkers and safety parameters for routine clinical use.
Funding and ClinicalTrials.gov
The trial is registered at ClinicalTrials.gov (NCT05329766). Detailed funding sources are listed in the original publication (Janjigian et al., Nat Med 2025).
References
1. 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 Oct 18. doi:10.1038/s41591-025-04022-w. Epub ahead of print. PMID: 41109921.
2. ClinicalTrials.gov. NCT05329766. EDGE‑Gastric study (STAR‑221 phase 3 program). Available at: https://clinicaltrials.gov/ (accessed via identifier NCT05329766).
Thumbnail image prompt (AI‑friendly)
A multidisciplinary oncology team (oncologist, medical researcher, pathologist) gathered around a lightbox and a laptop showing CT scans and immunohistochemistry slides; subtle overlays of schematic antibodies (Y shapes) and molecular markers (PD‑L1/TIGIT icons) in a modern clinical research setting, bright natural light, professional and hopeful tone.