AstraZeneca joins Tempus, Pathos in cancer AI project

AstraZeneca has started work with tech companies Tempus and Pathos AI on a project that aims to generate a wide-ranging and comprehensive generative artificial intelligence model for oncology.

The three partners are hoping to come up with a multimodal foundation model – trained on massive datasets spanning everything from text and omics data to medical imaging – that will help researchers look for biological and clinical patterns in cancer, find new drug targets, and develop new treatments.

At the heart of the project is a massive database operated by Tempus that includes more than 7.3 million de-identified patient records – including 1.4 million with imaging data, 1.3 million with genomic information, and 260,000 with full transcriptomics profiles - and over 1.1 billion pages of accompanying clinical text.

Shares in the Chicago-based company, which stands to receive up to $200 million under the terms of the project, shot up nearly 15% after the partnership was announced.

Pathos AI is tasked with using Tempus' platform to build the foundation model which, after it is completed, will be shared by all three companies so they can advance their own individual programmes, according to a joint statement.

Like all big pharma companies, AZ is already starting to deploy generative AI in its R&D operations and has said it is already seeing significant benefits in identifying novel targets, achieving better design of small and large molecules, informing clinical trial design, and improving the efficiency of its regulatory submissions.

"Cancer drug discovery and clinical development are being transformed by the ability to analyse vast amounts of rich data using artificial intelligence," said Jorge Reis-Filho, CAZ's chief AI and data scientist for oncology R&D.

The alliance with Tempus and Pathos AI "will accelerate and increase the probability of clinical success across our diverse pipeline," he added.

Pathos AI starts first clinical trials

Also this week, Pathos AI revealed that it has started patient dosing in a phase 1b/2a clinical trial of pocenbrodib, a CBP/p300 inhibitor that is the lead drug identified using its AI platform. CBP/p300 are proteins that activate genes that promote cancer cell growth and proliferation and represent a novel target in oncology.

The trial is testing the drug in patients with metastatic castration-resistant prostate cancer (mCRPC) as a monotherapy and in combination with other prostate cancer therapies, including hormonal therapy abiraterone acetate, AZ's PARP inhibitor Lynparza (olaparib), and Novartis' radioligand therapy Pluvicto (lutetium Lu 177 vipivotide tetraxetan).