Givastomig

Givastomig is a bispecific antibody (an engineered protein designed to grab two different molecular targets at once) from I-Mab Biopharma - originally licensed from Korea's ABL Bio under the code ABL111/TJ033721 and now co-developed 50/50 ex-Greater China and South Korea - that combines two distinct ideas: hit a tumor-specific antigen called CLDN18.2 on gastric cancer cells, and only then flip on a T cell switch called 4-1BB. The first-in-human Phase 1 monotherapy data, published in Clinical Cancer Research in 2025 [1], established a recommended Phase 2 dose range of 5-12 mg/kg and a modest but real signal (confirmed ORR 11% across 73 evaluable patients, 16% in the 43-patient gastroesophageal adenocarcinoma subset, DCR 37%) with manageable safety. Phase 1b combination data with nivolumab and mFOLFOX in first-line metastatic gastroesophageal carcinoma were disclosed at ESMO GI 2025 [10]. The randomized Phase 2 GIVA-2 trial (NCT07432295) [2] now tests givastomig added to nivolumab plus mFOLFOX as first-line treatment for CLDN18.2-positive, PD-L1-positive, HER2-negative metastatic gastric and gastroesophageal junction adenocarcinoma, against nivolumab + chemo alone - squarely in the territory Astellas's zolbetuximab (Vyloy) opened up with its FDA approval in 2024 [5]. The commercial bet is that conditional 4-1BB agonism, delivered only when the antibody is also engaged on a tumor cell, adds enough immune lift on top of chemo and a PD-1 blocker to beat the standard of care that CheckMate 649 [8] established and that zolbetuximab is now extending.

Givastomig is a novel compound, never approved anywhere. The Phase 1 monotherapy data in CLDN18.2-positive advanced solid tumors were published in June 2025 in Clinical Cancer Research by Ku et al. [1]; that parent study (NCT04900818) [3] continued into Phase 1b dose-escalation with nivolumab and mFOLFOX, with combination results disclosed at ESMO GI 2025 (abstract 388MO) [10] - establishing that the combination catalyst calendar is real and not speculative. The randomized Phase 2 GIVA-2 trial (NCT07432295) [2] opened in 2025 with a target enrollment of 180 patients and progression-free survival by blinded independent central review as the primary endpoint. (Note on the NCT identifier: ClinicalTrials.gov rolled forward to the NCT07 prefix in 2025; this is the correct, current registration ID.) No publicly disclosed FDA breakthrough therapy, fast track, or orphan drug designation for givastomig is known as of mid-2025 - none appear in I-Mab's recent investor materials [4]. The compound has no PDUFA date; with first-line gastric SoC median PFS ~7-8 months and a 180-patient design, a primary PFS readout from GIVA-2 is plausible in 2027, with any registrational Phase 3 follow-on pushing approval into the late-2020s. The near-term catalyst calendar is dominated by Phase 1b combination updates at GI ASCO (January) and ESMO GI (mid-year), where I-Mab has already disclosed combination data once [10].

Claudin 18.2 is a sealing protein that lives in the tight junctions between stomach lining cells - think of it as the caulk between bathroom tiles. In healthy adults, it is buried inside the gastric epithelium where antibodies cannot reach it. When stomach cells become cancerous, they lose their architecture and expose CLDN18.2 on the cell surface, where a drug can bind. About a third of gastric and GEJ adenocarcinomas express it at meaningful levels. Astellas's zolbetuximab (Vyloy) [5] proved this target works as an anchor: its 2024 FDA approval rests on two Phase 3 trials, SPOTLIGHT [6] and GLOW, showing survival benefit when added to first-line chemo. 4-1BB (CD137) is a costimulatory receptor on T cells, the gas pedal that amplifies an immune response once a T cell has already recognized something. Drugs that mash this pedal systemically (urelumab, utomilumab) [7] either fried the liver or did nothing useful. Givastomig's design is the workaround: the 4-1BB arm only crosslinks and fires when the antibody is also docked onto CLDN18.2 on a tumor cell. Activation is supposed to happen in the tumor and nowhere else. Importantly, Ku et al. [1] observed responses across CLDN18.2 expression levels from 11% to 100% of tumor cells, suggesting givastomig may engage patients below zolbetuximab's ≥75% 2+/3+ IHC bar - a potentially broader addressable population than the validated CLDN18.2 monoclonal. The genetic and target-validation case for CLDN18.2 is now strong because of zolbetuximab. The case for 4-1BB as a productive immune lever is still entirely a bet on conditional engagement.

GIVA-2 (NCT07432295) [2] is a randomized, open-label, multicenter Phase 2 study testing givastomig at two dose levels combined with nivolumab and mFOLFOX versus nivolumab and mFOLFOX alone as the comparator, in first-line CLDN18.2-positive, PD-L1-positive, HER2-negative metastatic gastric and GEJ adenocarcinoma. Enrollment target is 180 patients, currently recruiting, sponsor I-Mab Biopharma US. The primary endpoint is PFS by blinded independent central review, which is sensible for a Phase 2 designed to set up a registrational Phase 3. The chemo backbone choice (mFOLFOX rather than CAPOX) aligns with SPOTLIGHT [6] and simplifies cross-trial comparison, though it also concentrates GI toxicity risk that will need to be parsed cleanly against any 4-1BB-driven hepatic signal. The competitive context that matters here: nivolumab plus chemo (the CheckMate 649 regimen [8]) is now standard of care in PD-L1-positive gastric cancer in the US, while zolbetuximab plus chemo is the new option for CLDN18.2-positive patients. Beyond zolbetuximab, AstraZeneca's AZD0901 (CLDN18.2 ADC, Phase 3 in gastric) and Innovent's IBI343 (CLDN18.2 ADC, Phase 3 in gastric) are advancing payload-based approaches against the same target - a different therapeutic class (ADC vs. immune-cell-engaging bispecific) but the same anchor. Givastomig is being slotted into a triplet, which raises questions about attribution if the combination works and about tolerability if it does not. The CLDN18.2 cutoff threshold GIVA-2 uses for eligibility is consequential - Ku et al. saw responses well below zolbetuximab's ≥75% 2+/3+ bar - and will shape both who enrolls and what comparison to standard-of-care data is fair. The trial does not appear to be powered for OS at n=180, so any positive PFS signal will need to be re-litigated in a larger study.

Our model puts this drug's approval odds at 5%. That starts from the historical approval rate for Phase 2 drugs in this area, which is about 13%, then adjusts based on ten facts about the trial and sponsor. The estimate rises because the trial has more secondary endpoints than usual, but falls because the sponsor has a thin approval record and earlier-phase results were weak. Most other factors are close to average, so they don't shift the number much either way.

Safety: 4-1BB has a body count. Urelumab's grade 4 liver enzyme elevations stopped its monotherapy program over a decade ago [7]. Givastomig's bispecific architecture aims to localize activation, and the Ku et al. [1] Phase 1 monotherapy paper reported manageable safety with dose-proportional exposure across the 5-12 mg/kg range chosen for combination - but AST/ALT trends and immune-mediated hepatitis at the active doses in the triplet setting will be the single most important safety read at the next major GI conference. Efficacy: CLDN18.2 expression is heterogeneous, and the cutoff defining 'positive' is not standardized across programs. Ku et al. saw responses from 11% to 100% expression, which is encouraging for breadth but complicates head-to-head durability comparisons against zolbetuximab's narrower ≥75% population. The monotherapy ORR of 11% (16% in GEA) sets a modest bar - the combination must deliver well beyond additive activity for the conditional-4-1BB thesis to look real. Execution: I-Mab Biopharma US ended 2024 with $173.4M in cash, cash equivalents, and short-term investments [4], which the company guides as sufficient to fund the givastomig program into 2027 - enough to reach the Phase 2 readout but not to self-fund a registrational Phase 3 against a Big Pharma incumbent. The 50/50 ex-Greater China/Korea co-development with ABL Bio [4] means any out-licensing, partnership, or buyout economics get split, which both reduces I-Mab's upside per dollar of trial spend and complicates clean deal structures. Commercial: zolbetuximab launched in late 2024 and is establishing the CLDN18.2 first-line position [5]; AZD0901 and IBI343 are advancing ADCs against the same target. A bispecific entrant has to show better PFS, better OS, or a clearly better safety profile to displace any of them.

Worth watching, but not yet a portfolio thesis. The signal that would matter: Phase 1b combination expansion data showing an ORR meaningfully above the ~58% rate achieved by nivolumab+chemo in CheckMate 649 [8], in a CLDN18.2-selected population, without grade 3+ hepatotoxicity above a few percent. Initial Phase 1b combination data [10] are already in the public record and the next update is most likely to surface at GI ASCO (January) or ESMO GI (mid-year). The corporate question is whether I-Mab can run a registrational Phase 3 on its own - $173.4M and a co-development partner [4] suggest no - or whether a strategic buyer or licensee steps in after GIVA-2. Both are plausible given the target validation zolbetuximab provided and ABL Bio's incentive to monetize. Check back after the next major GI conference for updated Phase 1b combination data, and watch I-Mab's next annual filing [4] on capital position and partnership intent. Until there is a clean efficacy and safety readout in a triplet, the conditional 4-1BB-activation thesis remains a hypothesis with one approved drug (zolbetuximab) doing all the target-validation work for the entire category.