Key points from article :
Insilico Medicine has announced a new preclinical drug candidate, ISM3830, designed to boost the immune system’s ability to fight advanced cancers. The work, led by researchers at Insilico, builds on findings recently published in the Journal of Medicinal Chemistry, where the team described the discovery and optimisation of a series of CBLB inhibitors. CBLB is an intracellular protein that acts as a brake on T cells and natural killer (NK) cells. High levels of CBLB are seen across many tumour types, and blocking it has been shown to restore exhausted immune cells and improve anti-tumour responses.
ISM3830 is a highly selective, orally available inhibitor of CBLB generated using Chemistry42, Insilico’s generative AI design platform. The system uses more than 40 AI models to explore chemical space, predict ADMET properties, and optimise drug-like behaviour. This enabled the discovery of a novel molecular scaffold—only 42% similar to known structures—with improved absorption, metabolism, and safety characteristics. Early safety screens show low risk of off-target toxicity, strong selectivity, and favourable pharmacokinetics across multiple preclinical species.
In animal studies, ISM3830 produced strong tumour-shrinking effects in several mouse models and triggered long-lasting immune protection, confirmed by tumour rechallenge tests. It also showed potential to work alongside existing therapies, including immune checkpoint inhibitors, chemotherapies, and targeted drugs—important for patients who do not respond to current immunotherapies. IND-enabling studies are now underway to support first-in-human trials in people with advanced solid tumours.
This nomination adds to Insilico’s rapidly expanding AI-generated pipeline, which includes programs published in Nature Biotechnology, Nature Communications, and Nature Medicine. With ISM3830, the company aims to push AI-designed cancer immunotherapies closer to the clinic, addressing major unmet needs in treatment-resistant cancers.
