GSK and Hansoh press on despite Pfizer’s exit

B7-H4 has proven a tough nut to crack, but GSK and Hansoh still seem to have confidence in the target, with a new Chinese phase 3 study of the partners’ ADC, GSK5733584, recently listed on clinicaltrials.gov.

The move comes amid a backdrop of disappointing data with B7-H4 projects, including on a long delayed study of AstraZeneca’s puxitatug samrotecan, and the exit of Pfizer, which last month ditched its own ADC, felmetatug vedotin.

It’s becoming increasingly clear that B7-H4-targeting ADCs work best – and perhaps only – in patients with high levels of B7-H4 expression. It’s therefore curious that the new phase 3, which is sponsored by Hansoh, makes no mention of patients’ B7-H4 status as part of its enrolment criteria.

Meanwhile, a GSK-sponsored global phase 1, which started last July, does require testing, but only retrospectively.

Mersana warning

Pfizer’s aforementioned felmeta-v, AstraZeneca’s puxita-s, and Mersana’s emiltatug ledadotin have all yielded early-stage data showing responses driven by B7-H4-expressers. Mersana, most notably, focused on patients with B7-H4 levels of 70% or higher.

Pfizer then walked away from its asset, saying felmeta-v was “unlikely to achieve a meaningful improvement over standard-of-care chemotherapy”.

GSK might take comfort from the fact that Pfizer and Mersana’s projects use auristatin payloads, while GSK5733584 employs a more fashionable topoisomerase 1 inhibitor. However, Astra’s contender also has a topo1 payload, and has still disappointed.

Pfizer also has a topo1-based project in play, the preclinical SGN-B7H4C, while BeiGene’s DualityBio-originated BG-C9074 uses a topo1 inhibitor. The latter is due to generate proof-of-concept data later this year, which could give more clues about B7-H4’s relevance.

Astra doesn’t appear to have given up on puxita-s, either, promising more phase 1/2 results in 2026. As well as evaluating monotherapy, this trial has also begun combining the ADC with Astra’s anti-PD-1 x TIGIT MAb rilvegostomig.

If GSK5733584 does turn out to be a dud at least it won’t be an expensive one: GSK licensed the asset for just $85m up front in 2023. However, after various recent oncology disappointments, the group could do without another cancer misstep.

Clinical-stage B7-H4-targeting projects

Source: OncologyPipeline.