AstraZeneca's NCT06764875 pits a rilvegostomig–Enhertu regimen against the pembrolizumab comparator in first-line HER2+ gastric cancer

AstraZeneca builds a first-line gastric regimen it controls end-to-end

AstraZeneca is recruiting NCT06764875, a Phase 3 randomized, open-label, sponsor-blinded, three-arm global trial testing a combination it largely controls — rilvegostomig plus a fluoropyrimidine plus trastuzumab deruxtecan (T-DXd, Enhertu) — as first-line treatment for HER2-positive, PD-L1 CPS ≥ 1 gastric and gastroesophageal junction (GEJ) adenocarcinoma. ^[1] The study opened March 1, 2025 and carries an estimated primary completion of April 27, 2029; it plans to enroll 840 patients and has posted no results. ^[1] The launch is documented in the ClinicalTrials.gov registry — Prognyx found no corresponding AstraZeneca press release or SEC filing in the sources reviewed. ^[1]

The arm structure is the experiment. Arm A is rilvegostomig + fluoropyrimidine + T-DXd. Arm B, the comparator, is trastuzumab + chemotherapy + pembrolizumab. Arm C is rilvegostomig + trastuzumab + chemotherapy, added explicitly to assess the contribution of components. ^[1] The two primary endpoints are progression-free survival and overall survival, and protocol chemotherapy spans 5-fluorouracil, capecitabine, cisplatin and oxaliplatin. ^[1]

Lay the arms side by side and the design reads almost factorial. Arm A versus Arm C both hold rilvegostomig constant, so the contrast broadly isolates the antibody-drug conjugate's contribution — though the per-arm chemo backbones are not fully detailed in the registry (Arm A pairs a fluoropyrimidine with T-DXd; Arm C pairs trastuzumab with chemotherapy), so it is not a guaranteed single-variable read. Arm C versus Arm B holds the trastuzumab-plus-chemotherapy backbone constant and pits rilvegostomig against pembrolizumab — a clean rilvegostomig-versus-pembrolizumab read. Arm A versus Arm B is the headline: the full new regimen against the comparator. ^[1] AstraZeneca has engineered the trial to win the top-line comparison and to show that each proprietary swap pulls its own weight.

Eligibility is restricted to PD-L1 CPS ≥ 1 tumors — a PD-L1-selected population. ^[1] Two primary endpoints — PFS and OS — across 840 patients mark this as a registration-grade bet, not a signal-finding study. ^[1]

The asset — and the gap

T-DXd is the approved anchor: ChEMBL lists trastuzumab deruxtecan (CHEMBL4297844) as a HER2/ERBB2-binding agent at maximum phase 4.0. ^[3] Rilvegostomig is the newer piece — an AstraZeneca asset registered in ChEMBL (CHEMBL5314706) at maximum phase 3.0. ^[1][2] Here is the honest gap: Prognyx could not confirm rilvegostomig's molecular target or mechanism from the primary sources reviewed for this briefing — ChEMBL records no mechanism, and no label or peer-reviewed mechanism description was available. ^[2] The combination's clinical rationale is at least on the record: rilvegostomig plus T-DXd appears in a 2026 review of bispecific antibodies and ADCs in advanced gastric/GEJ adenocarcinoma. ^[7]

Why it matters

The strategic logic is ownership. Enhertu is co-developed by AstraZeneca and Daiichi Sankyo ^[5]; rilvegostomig is an AstraZeneca asset. ^[1][2] The comparator AstraZeneca chose for this trial includes pembrolizumab. ^[1] Pembrolizumab is marketed by Merck as Keytruda — general background, not established in the primary sources reviewed here. If Arm A prevails, AstraZeneca would displace a comparator regimen containing a competitor's product with one built on assets it and its partner control — a control play as much as a clinical one. Whether that comparator represents today's approved first-line standard of care is not established in the sources reviewed.

Precision matters on the regulatory backdrop. Around May 22, 2026 — per secondary reporting (Google News RSS redirects, with Business Wire and Stock Titan also citing AstraZeneca and Daiichi Sankyo) — the EMA's CHMP recommended Enhertu for EU approval in previously treated HER2-positive metastatic solid tumours, a later-line, tumour-agnostic setting. ^[4][5] That is a positive opinion, not an approval, and it is not a first-line gastric indication. The 1L gastric/GEJ use of T-DXd in NCT06764875 remains investigational. ^[1]

Who is exposed, who gained

Pembrolizumab is the named comparator agent in this first-line HER2-positive gastric/GEJ trial — directly in the crosshairs if Arm A shows superiority. ^[1] The trastuzumab-only HER2 backbone is the second target: a positive Arm A-versus-Arm C result would argue that swapping naked antibody for the ADC adds first-line value. ^[1]

The broader HER2 field would feel the bar move. Open Targets catalogs 47 ERBB2 drugs and candidates, including Phase 3 assets disitamab vedotin, trastuzumab botidotin, dacomitinib, poziotinib, pyrotinib, KN-026 and HEMAY-022. ^[6] A successful first-line T-DXd combination raises the threshold every one of them must clear in gastric disease.

Gained: AstraZeneca and Daiichi Sankyo ^[5], who would extend Enhertu into a large first-line gastric population, and rilvegostomig, which gets a pivotal anchor indication. ^[1] Separately, AstraZeneca is running a Phase 2 perioperative study in locally advanced resectable gastroesophageal adenocarcinoma (NCT07069712); the registry describes it as a study of novel agents/combinations and does not confirm whether it evaluates these same assets. ^[8]

What to watch — with dates

• Primary completion: estimated April 27, 2029. ^[1] This is the dated catalyst. ClinicalTrials.gov posts no interim-analysis dates, so the definitive read sits years out. ^[1]
• Enrollment. The trial is recruiting toward 840 patients now; accrual pace is the near-term gating item. ^[1]
• Rilvegostomig disclosure. Its mechanism and any earlier-phase data are absent from the primary sources reviewed — the first credible read on what this molecule is and does is a watch item in itself. ^[2]
• EMA action on the CHMP opinion — separate from this trial, and in a later-line, tumour-agnostic setting, not first-line gastric. ^[4][5]

The now-what

Three options sit on the table for an operator in gastric IO or HER2:

1. Re-price your comparator. AstraZeneca has effectively dated the first-line HER2-positive gastric question to ~2029. If you are designing a competing first-line trial, pembrolizumab + trastuzumab + chemo is the comparator AstraZeneca put in play, and a T-DXd combination could become the bar to beat. Build for the bar that will exist in 2029, not 2026.

1. Stress-test your HER2 ADC thesis. The Arm A-versus-Arm C comparison will help quantify ADC-over-antibody value in first line — a data point that reprices the entire HER2 ADC field. If you hold an ADC, sharpen the differentiation case (payload, bystander effect, CNS activity, tolerability) before that number lands. ^[1][6]

1. Don't pre-judge the rilvegostomig swap. Arm C versus Arm B is a clean rilvegostomig-versus-pembrolizumab test on a shared backbone — the most direct read on the new agent in the trial. But bank nothing on it until rilvegostomig's mechanism and clinical data are public. ^[2]

Verdict — strategic threat, not an imminent one: AstraZeneca has staked a largely owned first-line HER2-positive gastric regimen against a pembrolizumab-anchored comparator, but with primary completion estimated in April 2029 the data window is years away. Act now on positioning; the data won't reshape first-line practice until late this decade.