ESMO 2026: HERIZON-GEA-01 and the evolving HER2-positive GE A treatment landscape banner image.
HER2-Positive GI Cancer / Resource Centers 11/03/2026

ESMO 2026: HERIZON-GEA-01 and the Evolving HER2-Positive GEA Treatment Landscape

Key Points

  • Zanidatamab plus chemotherapy significantly improved overall survival (OS) versus trastuzumab plus chemotherapy in first-line HER2-positive gastroesophageal adenocarcinoma (GEA), with the benefit concentrated in IHC 3+ disease.
  • The zanidatamab/tislelizumab/chemotherapy triplet achieved a median OS of 27 months, with durable benefit observed with longer follow-up.
  • OS benefits with the triplet were observed across PD-L1 subgroups, although the comparison between the triplet and zanidatamab doublet did not reach statistical significance.
  • Diarrhea was a common toxicity with zanidatamab-based regimens, but mandatory antidiarrheal prophylaxis reduced the incidence of grade 3 diarrhea.

The 2026 European Society for Medical Oncology (ESMO) Congress in Madrid, Spain, featured the updated OS analysis of the phase 3 HERIZON-GEA-01 trial (NCT05152147), presented by Kohei Shitara, MD, of National Cancer Center Hospital East, Kashiwa, Japan, as a late-breaking abstract. The results of this study came just months after the FDA approved two regimens containing zanidatamab-hrii for first-line HER2-positive advanced GEA.

GEA Background

Approximately 20% of GEAs are HER2-positive, and trastuzumab plus fluoropyrimidine/platinum chemotherapy has been the first-line backbone since the phase 3 ToGA trial. Zanidatamab is a bispecific (biparatopic) HER2-directed antibody that binds two distinct extracellular HER2 domains, driving receptor internalization and downregulation along with complement- and cell-mediated cytotoxicity.

At the first interim analysis, both zanidatamab arms significantly improved progression-free survival (PFS), and the zanidatamab/tislelizumab/chemotherapy triplet significantly improved OS (median 26.4 vs 19.2 months; HR, 0.72). The OS comparison for zanidatamab plus chemotherapy alone had not yet reached statistical significance.

In August 2026, the FDA approved zanidatamab plus tislelizumab and chemotherapy for HER2-positive (IHC 3+ or IHC 2+/ISH+) unresectable, locally advanced or metastatic GEA, as well as zanidatamab plus chemotherapy for HER2-positive IHC 3+ disease.

HERIZON-GEA-01 Trial Design

  • Population: 914 patients with previously untreated, unresectable, locally advanced or metastatic HER2-positive GEA (IHC 3+ or IHC 2+/ISH+ by central testing).
  • Study arms: randomized 1:1:1 to:
  • Arm A: trastuzumab plus chemotherapy (control)
  • Arm B: zanidatamab plus chemotherapy
  • Arm C: zanidatamab plus tislelizumab and chemotherapy
  • Dosing: zanidatamab 1800 mg (<70 kg) or 2400 mg (≥70 kg) IV every 3 weeks; tislelizumab 200 mg IV every 3 weeks; chemotherapy was CAPOX (capecitabine and oxaliplatin) or 5-FU (fluorouracil)/cisplatin.
  • Primary end points: PFS by blinded independent central review and OS.
  • Key note: Arm C versus Arm B was a secondary comparison with limited power to show superiority.

Key Takeaways From Arm B: Updated Overall Survival

  • Median OS: Zanidatamab plus chemotherapy improved median OS versus trastuzumab plus chemotherapy (22.9 months [95% CI, 19.9–27.6] versus 19.6 months [95% CI, 16.9–22.2]).
  • Statistical significance: The improvement was statistically significant (HR, 0.76; 95% CI, 0.62–0.93; P = .0088).
  • Clinical significance: This represents a 24% reduction in the risk of death and converts the numerical OS trend observed at the first interim analysis into a statistically significant result.

Landmark OS

  • At 2 years, OS was 49.6% with zanidatamab plus chemotherapy versus 40.7% with trastuzumab plus chemotherapy. 
  • OS at 3 years was 31.5% versus 22.9%, respectively.

PFS remained consistent with the primary analysis: median 12.4 months versus 8.3 months (HR, 0.65; 95% CI, 0.53–0.80). Moreover, the confirmed objective response rate was 69.6% with zanidatamab plus chemotherapy, with a median duration of response of 14.3 months.

Clinical Significance

Arm B isolates the HER2-directed agent. This is the first phase 3 trial in which a HER2-targeted antibody, without an added checkpoint inhibitor, improved OS over trastuzumab in first-line GEA. The National Comprehensive Cancer Network (NCCN) lists zanidatamab-containing regimens as preferred options alongside trastuzumab-based regimens. Clinically, the doublet matters most for patients for where immunotherapy is contraindicated, such as those with organ transplants or severe autoimmune disorders requiring systemic therapy. Of note, the FDA label for the doublet is restricted to HER2 IHC 3+ disease, and the subgroup data explain why: the OS benefit was concentrated in IHC 3+ tumors (HR, 0.71; 95% CI, 0.57–0.88), with no apparent benefit in IHC 2+/ISH+ tumors (HR, 1.20; 95% CI, 0.77–1.87). HER2 testing results should be reviewed closely before selecting this regimen.

Key Takeaways From Arm C: Longer-Term Follow-Up

  • Median OS: Zanidatamab plus tislelizumab and chemotherapy achieved an OS of 27 months (95% CI, 23.7–32.2) versus 19.6 months with trastuzumab plus chemotherapy (HR, 0.65; 95% CI, 0.52–0.80).
  • The OS benefit deepened with longer follow-up, with the HR improving from 0.72 at the first interim analysis to 0.65. This updated comparison was descriptive, as statistical significance had already been established at the first interim analysis.

Landmark OS

  • 2-year OS: 55.8% with zanidatamab plus tislelizumab and chemotherapy versus 40.7% with trastuzumab plus chemotherapy.
  • 3-year OS: 39.8% with zanidatamab plus tislelizumab and chemotherapy, 31.5% with zanidatamab plus chemotherapy, and 22.9% with trastuzumab plus chemotherapy.

PFS

  • Median PFS was 12.7 months versus 8.3 months with trastuzumab plus chemotherapy (HR, 0.61; 95% CI, 0.50–0.76).

Arm C Versus Arm B

  • Median OS was approximately 4.1 months longer with the triplet than the zanidatamab doublet (27 months vs 22.9 months; HR, 0.81; 95% CI, 0.65–1.00; P = .0502), which did not reach statistical significance.
  • Mean duration of treatment was longest with the triplet (64.1 weeks) compared with the doublet (52.6 weeks) and trastuzumab plus chemotherapy (42.3 weeks).

PD-L1 Subgroups (Second Interim Analysis)

  • Roughly 30% to 36% of patients in each arm had tumors with PD-L1 tumor area positivity (TAP) below 1%.
  • The triplet improved OS in both PD-L1 TAP below 1% (HR, 0.50; 95% CI, 0.35–0.73) and TAP of 1% or greater (HR, 0.68; 95% CI, 0.52–0.89) disease.
  • Adding tislelizumab showed a consistent OS trend regardless of PD-L1 status: TAP below 1% (HR, 0.76; 95% CI, 0.52–1.10) and TAP of 1% or greater (HR, 0.77; 95% CI, 0.58–1.02).

Implications for First-Line Treatment Selection

The HERIZON-GEA-01 triplet showed benefit regardless of PD-L1 status and is FDA-approved for both IHC 3+ and IHC 2+/ISH+ disease. The subgroup data should be interpreted cautiously, as they were not powered for formal comparison. Currently, the NCCN recommends the triplet regimen for combined positive score (CPS) of 1 or greater, while awaiting further OS analyses. 

Safety and Practical Management

Treatment exposure was longest with the triplet regimen (median 46.6 weeks vs 31.1 weeks with the zanidatamab doublet and 30 weeks with trastuzumab plus chemotherapy), which partly accounts for the higher cumulative rates of adverse events (AEs).

Noninfectious pulmonary toxicity was more frequent with the triplet (7.5% vs 1.6% and 1.3%), consistent with the addition of a PD-1 inhibitor. Prophylaxis for infusion-related reactions and diarrhea was mandatory in both zanidatamab arms.

Health-related quality of life (QoL) analyses (ESMO GI 2026) showed that physical functioning was maintained without compromising  overall QoL.

Figure 1. Safety Profile at the Second Interim Analysis

Arm C: Zani + Tisle + CTx (n = 294)Arm B: Zani + CTx (n = 305)Arm A: Trastuzumab + CTx (n = 302)
Grade ≥3 TRAEs72.4%59.0%59.9%
Any-grade diarrhea83.0%79.0%53.3%
Grade ≥3 diarrhea25.2%20.0%12.9%
Zanidatamab/trastuzumab discontinuation due to TRAEs11.9%9.2%2.3%
Treatment-related deaths2.4%0.3%1.3%
Infusion-related reactions25.2%25.2%13.2%
Left ventricular dysfunction9.2%6.6%4.3%
Immune-mediated AEs40.1%14.4%10.6%

Abbreviations: AE, adverse event; CTx, chemotherapy; TRAE, treatment-related adverse event; Zani, zanidatamab; Tisle, tislelizumab.

Updated safety findings presented at ESMO 2026 showed that with an additional 9.4 months of follow-up, no new safety signals emerged and the safety profiles of both zanidatamab regimens were consistent with the first interim analysis.

Diarrhea: What to Expect

The zanidatamab label carries a Boxed Warning for diarrhea (and embryo-fetal toxicity). At interim analysis 2, treatment-related diarrhea occurred in 81.6% of patients receiving the triplet and 76.7% of those receiving the doublet (grade ≥3: 24.8% and 20%) compared with 48.3% (grade ≥3: 12.9%) with trastuzumab plus chemotherapy. Diarrhea occurred mainly in cycles 1 and 2, with a median time to first onset of 7 days with the triplet and 6 days with the doublet. Grade 3 or greater diarrhea had a median time to onset of16 and 11 days, respectively. 

Prophylaxis was beneficial, as grade 3 diarrhea fell from 52% to 24% after mandatory antidiarrheal prophylaxis was introduced.

Conclusion

The updated HERIZON-GEA-01 results from ESMO 2026 strengthen the case for zanidatamab as the first-line HER2-directed backbone in HER2-positive advanced GEA. Arm B shows that changing the HER2 agent alone improves survival: zanidatamab plus chemotherapy significantly improved OS over trastuzumab plus chemotherapy (HR, 0.76), with the benefit concentrated in IHC 3+ disease. With longer follow-up, Arm C showed durable benefit with a median OS of 27 months. Adding tislelizumab extended median OS by 4.1 months over the zanidatamab doublet, although this comparison did not reach statistical significance. The OS benefit was consistent across PD-L1 subgroups, so it remains to be seen whether these mature data will broaden NCCN recommendations beyond PD-L1 CPS of 1 or greater. Diarrhea remains the main toxicity, but mandatory antidiarrheal prophylaxis showed improvement in grade 3 diarrhea. No new safety signals emerged with longer follow-up. Remaining questions include how much tislelizumab adds over the zanidatamab doublet and how best to sequence therapy after progression on a zanidatamab-based regimen. 

References

  1. Shitara K, Elimova E, Liu T, et al. Zanidatamab with and without tislelizumab in HER2-positive gastroesophageal cancer. N Engl J Med. Published online May 28, 2026. PMID: 42202319.
  2. Shitara K, et al. Zanidatamab plus chemotherapy with or without tislelizumab for first-line HER2-positive advanced/metastatic gastroesophageal adenocarcinoma: updated overall survival from HERIZON-GEA-01. Presented at: ESMO Congress 2026; Madrid, Spain. Abstract LBA79. 
  3. Rha S, Shitara K, Shen L, et al. Zanidatamab + chemotherapy (CT) ± tislelizumab for first-line (1L) HER2-positive (HER2+) locally advanced or metastatic gastroesophageal adenocarcinoma (mGEA): PD-L1 subgroup analysis from HERIZON-GEA-01. J Clin Oncol. 2026;44(16 suppl):4010. Presented at: ASCO Annual Meeting; Chicago, Illinois. 2026.