![The sights of Avalon Air Show 2025: Day Three [PHOTOS]](https://breakingdefense.com/wp-content/uploads/sites/3/2025/03/f-35-avalon-final-day-scaled-e1743079275404.jpg?#)
STAT+: This Canadian company’s AI tool could help drugmakers cut risk in late-stage cancer trials
Toronto-based Altis Labs uses AI tool to review lung cancer CT scans to predict future patient survival.
When AstraZeneca announced the results of its failed lung cancer immunotherapy trial in 2017, its stock plunged 16%. But that data was good for something — today, the drugmaker presented a re-evaluation of data from that trial using an AI tool, which outperformed the conventional method of predicting overall survival in lung cancer patients.
The Toronto-based company, Altis Labs, claims its AI model is similarly superior at analyzing breast cancer and colorectal cancer scans, potentially offering drugmakers a better way to design their clinical trials.
Much of the fanfare for AI in drug development has gone to companies at the pre-clinical stage of the pipeline: those churning up new targets, figuring out how to re-purpose drugs, or designing new molecules and proteins that may make good therapeutics. But the real bottleneck for drug development, and where AI could really make a difference, is in clinical trials, particularly in the Phase 3 trials that take years and millions of dollars to conduct. According to a 2016 study of 640 drugs that entered trials between 1998 and 2008, 54% of Phase 3 trials fail.
