Bristol seeks novelty in TIGIT

If Bristol Myers Squibb’s discontinuation of the anti-TIGIT MAb renvistobart didn’t come as much of a surprise then doubling down on the bispecific TIGIT asset BMS-986442 – and revealing its second target – did. BMS-986442 is in fact an Fc-selected dual anti-TIGIT/anti-CD96 MAb, previously coded AGEN 1777.

Bristol had licensed it from Agenus for $200m up front in 2021, at which point its second target was undisclosed. Bristol highlighting this project during yesterday’s R&D day is thus a strong endorsement of Agenus and its technology, and perhaps the most important feature of CD96 is its role in the so-called CD226 axis, a theme familiar to ardent watchers of the TIGIT space.

CD96 is a protein also known as “Tactile”, expressed in particular on T and NK cells, and both it and TIGIT share CD155 as a ligand. Bristol yesterday said CD96 and TIGIT were complementary targets in the same pathway that negatively regulate T and NK cell function in the tumour microenvironment, and that BMS-986442 could enhance the “quality and magnitude” of T-cell responses through dual inhibition in antigen-presenting as well as tumour cells.

Mention of CD96 will naturally make biotech investors curious about what else is going in with this approach. OncologyPipieline reveals just two other industry assets that hit CD96: GSK’s GSK6097608 and Samyang Biopharmaceuticals’ SYB-020.

Industry projects targeting CD96

Note: CD96 is also known as Tactile. Source: OncologyPipeline.

A phase 1 study of single-agent BMS-986442 has been completed, and a phase 1/2 trial, looking at combining it with PD-(L)1 blockade and or chemotherapy, is in dose escalation and should yield data next year. Bristol also cited gastric cancer – a relatively new departure for TIGIT – as well as NSCLC as an initial tumour of interest.

Meanwhile, the monospecific anti-TIGIT MAb renvistobart (BMS-986207) has been scrapped, Bristol confirmed yesterday. Renvistobart remains in three active studies, according to clinicaltrials.gov, but notably its phase 2 test in first-line NSCLC was withdrawn in January after treating just one patient, owing to an “adverse change in risk/benefit”.

The CD226 axis

Disclosure that Bristol is doubling down on a combinatorial TIGIT approach is also highly relevant for the micro-cap player Compugen, which has long argued that TIGIT blockade has to be combined with action on other members of the so-called CD226 axis. To this end Compugen is testing its Fc-silent anti-TIGIT project COM902 with the anti-PVRIG MAb COM701.

But why is this relevant? Because the third member of the CD226 axis is none other than CD96. 

It might come as a surprise that only GSK boasts MAbs against all three CD226 axis members; as well as the anti-CD96 GSK6097608 the UK company’s pipeline includes the iTeos-derived anti-TIGIT MAb belrestotug, and the anti-PVRIG molecule GSK4381562/SRF813, licensed in 2020 from Surface Oncology. If the CD226 axis theory has legs then it’s likely that GSK will pursue double and triple combos to achieve maximal efficacy.

It will also be noted that AstraZeneca has apparently eschewed the TIGIT monospecific approach, betting instead on rilvegostomig/AZD2936, which it has promised to take into phase 3 this year. However, rilvegostomig doesn’t hit other members of the CD226 axis beyond TIGIT, its second target antigen simply being PD-1.

There’s a final twist, however: rilvegostomig is derived from a deal with Compugen, and is based on COM902. If continuing developments on Roche’s tiragolumab suggest that the TIGIT approach is far from dead, then earlier work shows that the science behind it won’t get simpler any time soon.

Industry projects targeting PVRIG

Note: PVRIG is also known as CD112R. Source: OncologyPipeline.

This story has been updated to add the D3 Bio project.