First Wave of AI-Designed Drugs Faces Major Clinical Setbacks

• Facebook
• Twitter
• Reddit
• Tumblr
• LinkedIn
• Pinterest
• RSS Feed
• Link

First Wave of AI-Designed Drugs Faces Major Clinical Setbacks

• AstraZeneca Admits for the First Time that its COVID Vaccine Has Blood Clot Side Effects
• Was COVID virus leaked from the Chinese WIV lab?
• HIV Cure Research: New Study Links Viral DNA Levels to Spontaneous Control
• FDA has mandated a top-level black box warning for all marketed CAR-T therapies
• Can people with high blood pressure eat peanuts?
• What is the difference between dopamine and dobutamine?
• How long can the patient live after heart stent surgery?

First Wave of AI-Designed Drugs Faces Major Clinical Setbacks

The first batch of drugs designed by artificial intelligence has largely ended in failure.

In early October, the UK-based AI pharmaceutical company Exscientia updated its pipeline, discontinuing the development of its cancer candidate drug, EXS-21546, in its Phase I/II studies.

Officially, the reason given for this A2A antagonist was that “achieving appropriate therapeutic efficacy for the drug would be challenging based on modeling of clinical and preclinical data.”

Last year, Exscientia, along with its partner Sumitomo Dainippon Pharma, decided to abandon another AI-designed drug. Similarly, AI pharmaceutical companies BenevolentAI and Relay had pipeline setbacks.

“We are now also aware that if we want to change the likelihood of clinical success, it’s not just about better molecules,” said Exscientia founder Hopkins, “we also need better translational models.”

The price is too high

Over the past year, the first batch of molecules created by artificial intelligence has suffered significant setbacks, with some put on hold in development and others having their clinical trial priorities reduced.

The AI companies behind these drugs enthusiastically brought them into clinical trials, heralding a new era in drug discovery, only to quietly shelve them after paying the price.

The table above only lists clinical drugs designed with AI, as some AI pharmaceutical companies acquired the rights to discontinued drugs.

For example, in October 2022, Recursion stopped the development of REC-3599 for treating GM2 gangliosidosis, but this drug was brought into Phase III more than two decades ago by Eli Lilly for treating diabetic retinopathy and later changed hands to Recursion.

Another prominent example is Exscientia.

In 2021, Exscientia went public with three clinical-stage drugs, with a market value of $2.6 billion at the time. Two of these drugs, DSP-1181 and DSP-0038, were developed in collaboration with Sumitomo Dainippon Pharma, while EXS-21546 was jointly developed with Evotec SE. DSP-1181 was promoted as the world’s first AI-designed molecule to enter clinical trials.

However, Exscientia’s pipeline has undergone a major overhaul, replacing the A2A antagonist (EXS-21546), 5-HT1A agonist, and 5-HT2A antagonist (DSP-0038 and DSP-1181) with the PKC-θ inhibitor EXS-4318 and the CDK7 inhibitor GTAEXS617.

Although Exscientia can rapidly “replenish” its pipeline and advance new drugs into clinical trials, it’s difficult to avoid speculation from outsiders about the quality of its molecules, relying on speed and quantity to sustain company growth.

Some companies once proudly declared that AI had reduced the speed and cost of drug development when announcing drugs entering clinical trials, but it seems that AI hasn’t reduced the risks of clinical drug development.

“Speed and cost are the easiest things to demonstrate, especially in preclinical stages. However, increasing the success of drugs in clinical trials will inevitably require the longest time to verify,” said Ivan Griffin, Co-Founder and COO of BenevolentAI.

This statement also highlights why AI pharmaceutical companies’ promotional materials all tout speed and time, as proving effectiveness indeed takes a considerable amount of time, making it challenging for the outside world to buy into the AI pharmaceutical story.

People once held unrealistic fantasies about AI, expecting it to elevate the success rate of drug development from 5% to 90%. However, the reality is that AI can increase the rate from 5% to 6% or 7% at most.

Meanwhile, the success rate of drugs in the clinical stage is inherently low, with estimates of only 5% to 10% of clinical-stage drugs gaining approval.

Today, these AI pharmaceutical companies are beginning to focus on hiring top drug development experts, optimizing their business models and research projects, and no longer harboring illusions that AI can revolutionize drug development overnight.

“We are now also aware that if we want to change the likelihood of clinical success, it’s not just about better molecules,” said Exscientia founder Hopkins. “We also need better translational models.”

A Polarized Market

Apart from Schrodinger and Recursion, the first wave of AI pharmaceutical companies’ market values have dropped below $1 billion.

Exscientia has a market value of $640 million, and BenevolentAI is only valued at $117 million. The stock prices of listed AI pharmaceutical companies have fallen by at least 75%, underperforming the biotech market.

Many companies have had to resort to layoffs due to pipeline failures. For instance, BenevolentAI laid off 50% of its workforce, while Atomwise also cut one-third of its employees due to business adjustments.

Schrodinger, a standout, actively removed the AI label, openly stating that it’s not an AI company but rather a pharmaceutical company with proprietary software.

CEO Ramy Farid once expressed concern about the hype around AI in the market, warning of a possible steep fall when the hype reaches its peak.

However, for supporters of AI in drug development, the success of projects like Moderna’s Covid-19 vaccine or Nimbus Therapeutics’ TYK2 inhibitor to some extent relied on artificial intelligence. These pharmaceutical companies simply didn’t label these drugs as AI-designed, unlike Exscientia, BenevolentAI, and Recursion, which prominently feature their AI-driven or AI-designed approach.

The recent coldness in the secondary market hasn’t deterred significant funding in the primary market over the past month or two.

Investment in AI pharmaceuticals mainly comes from technology funds like Nvidia, internet funds, and large capital groups, with fewer appearances from industry capital and biopharmaceutical funds.

The divided attitude has made it challenging for the market to understand these companies.

Dylan Reid, a partner at Zetta Venture Partners, said, “The market has never known how to assess these companies. They have been both excited and disappointed. They once reached valuations of X billion dollars for their platform, but today, this could weigh down valuations.”

The only way to create significant value using artificial intelligence in drug discovery is to help obtain specific approved drugs.

People have also realized that faster AI in pharmaceuticals does not necessarily mean better. Faster failure is not success. In the face of scrutiny in clinical trials, AI drug molecules must take the step of proving their quality and moving beyond the hype.

First Wave of AI-Designed Drugs Faces Major Clinical Setbacks

(source:internet, reference only)

Disclaimer of medicaltrend.org

Important Note: The information provided is for informational purposes only and should not be considered as medical advice.