ESMO Congress 2026

CLDN18.2 is a hotly contested target in gastric/gastroesophageal junction adenocarcinoma (G/GEJA): zolbetuximab is the first-generation anti-CLDN18.2 mAb, while next-generation agents — including arcotatug tavatecan (IBI343) — spanning ADCs, bispecific antibodies (emaretamig, givastomig) and CAR-T cells (IMC002), have shown single-agent objective response rates higher than that of zolbetuximab ^[2]. Within the ADC track, arcotatug tavatecan, garetatug rezetecan, sonesitatug vedotin and tecotabart vedotin are all in phase III development for CLDN18.2-positive G/GEJA ^[1]. Gastrointestinal toxicity, nausea and vomiting are common on-target/off-tumor adverse effects of this class ^[2]; tavatecan itself is an exatecan-based topoisomerase I inhibitor payload also used in other ADCs under development ^[3][4].

At ESMO26, the reported readout is the first interim analysis of the phase 3 G-HOPE-001 study (presented by Kohei Shitara, with the Innovent and Takeda teams involved in development): arcotatug tavatecan versus TPC in CLDN18.2-positive, ≥3L G/GEJA. Conducted in China and Japan, 2600 patients were prescreened; enrollment was recorded as 430 in one post and 460 in another (to be confirmed). Key numbers: ORR 37.4% vs 7.7%; mPFS 7.1 vs 3.1 months (HR 0.45, updated; another post cited 0.47); mOS 10.2 vs 6.9 months (HR 0.57, p<0.0001). For safety, grade ≥3 treatment-related AEs were 47.4% vs 55.9% and dose reductions 17.8% vs 35.1%, summarized by posters as a strong signal of efficacy with acceptable safety.

Whereas the background literature previously only noted its phase III status ^[1] and early single-agent ORR exceeding zolbetuximab ^[2], this readout delivers randomized, controlled dual-endpoint benefits in PFS and OS: in the 3L+ population, mPFS extended from 3.1 to 7.1 months and mOS from 6.9 to 10.2 months crossing the significance threshold, with ORR of 37.4% vs 7.7%, while grade ≥3 treatment-related AEs and dose reductions were both lower than with control chemotherapy. This provides phase 3 efficacy and tolerability evidence for upgrading CLDN18.2 targeting from mAbs ^[2] to ADCs; as for the gastrointestinal on-target/off-tumor toxicity common to this class ^[2], the on-site posts only summarized it as 'acceptable safety,' with the specific AE profile undisclosed.

Limitations and items to confirm: all figures come from on-site social media posts, with discrepancies between posts on the same metric (enrollment 430 vs 460, PFS HR 0.45 vs 0.47, ORR 37% vs 37.4%, and CLDN18.2-positive vs CLDN18.2-high wording), which must be verified against the official abstract and full publication; this is a first interim analysis and OS data will mature further; the composition of the TPC arm, the p value for PFS, subgroup analyses and the CLDN18.2 assay cutoff were not disclosed in the posts.