Are AI-Predicted Clinical Trials the Future of Drug Development?

Let’s talk about AI for a second.

There are problems that AI can solve, problems that AI can’t solve and problems that AI shouldn’t solve.

But every industry is looking at AI and trying to figure out how it fits in.

And healthcare is no different.

Medical imaging is setting the pace for the healthcare industry with cardiologists and radiologists already using FDA approved AI products to support disease detection.

But where else AI can fit into the industry?

AI in drug discovery…

AI for personalised medical care…

(let’s be honest, we’ve all used ChatGPT to check our symptoms at one point)

But where I think the most interesting place for AI in healthcare is if AI can speed up the clinical trial pathway.

It takes 10 years on average for a drug to go from pre-clinical studies through to a commercial product.

The prize can be big for those drugs that are lucky enough to make it through the entire process.

But it does take time and capital.

The gold standard is the double-blind, placebo-controlled clinical study…

But what if AI could be used to support or predict clinical outcome trials BEFORE they happen?

In this week’s episode of the GlobalData Healthcare podcast, this exact idea was discussed at a panel at the Jefferies Healthcare Conference last month in London (around the 15 minute mark).

Every panellist agreed that the FDA is way more likely to accept AI predictions of drug effects as supplementary to actual drug effects.

This would be a big change for the industry, as AI clinical prediction tools may become the difference between a drug making it to market or not.

(scary thought)

While the double blind placebo controlled study is unlikely to be changed any time soon…

The idea of AI predictions of drug safety and efficacy is no longer an absurd pipe dream.

With this backdrop, the FDA announced on Monday that it would “expand artificial intelligence capabilities”

The organisation plans to use AI for various tasks including pre-market reviews.

This is just the beginning.

So will we one day see a world where drugs are entirely approved just based on an AI algorithm?

Who knows… 

But I think that anything that can speed up the 10-year timeframe to bring life saving drugs to market is a good thing.

The Daily Check-up

DorsaVi (ASX: DVL | MC: $43M) continues testwork on its RRAM microchip at 22 nanometers, optimising performance while shrinking the size. DVL acquired this technology in June for use in various robotics and medical wearable device applications. (DVL)

Algorae Pharmaceuticals (ASX:1AI | MC: $23M) AI prediction tool demonstrated predictive power in identifying synergistic and non-synergistic combinations of 21 CBD-drugs when tested against cancer cells. Results from the test tube study were independently validated by the Peter MacCallum Cancer Centre. (1AI)

Island Pharmaceuticals (ASX:ILA | MC: $130M) secured a US patent for its drug Galadisivir for the COVID-19 virus.

🪑 Hey ILA, 2021 is calling and they want their announcement back. 

Althea Group (ASX:AGH | MC: $21M) has changed its name to Peak Processing Limited, stock code change to PKP next Tuesday. (AGH)

The UK has agreed to pay 25% more for new medicines by 2035 as part of a US-UK drug pricing deal. (The Guardian) 

🪑It will be interesting to see if Australia gets caught up in this as well. 

There is a really good article by the AFR that discusses what this US-UK deal may mean for the Australian Pharmaceutical Benefits Scheme and why Australia’s largest healthcare company CSL has the most at stake. (AFR)

US-based Crescent Biopharma and Chinese-based Kelun‑Biotech strikes a global oncology partnership in a deal that could deliver up to US $1.25 billion in milestone payments. (Biospace)

AlphaFold’s John Jumper reflects on five years at the company and how the technology is evolving from predicting proteins to driving a broader new era of AI-powered biological discovery. (Technology Review)

See you tomorrow.

The Armchair Analyst

Jason Segal