AI Is Pharma's Newest Power Tool For Enhancing Clinical Trials

It’s not exactly news that Big Pharma warmed up to the possibilities brought by artificial intelligence with major drug makers using this tool to speed up their clinical trials, potentially saving millions along the way. A scientist from Roche Holding AG RHHBY subsidiary Flatiron Health, Blythe Adamson, was among the many who confirmed that AI empowers scientists to examine real-world patient data quickly, and at a scale. Amgen Inc AMGN, Bayer Aktiengesellschaft BAYRY and Novartis AG NVS are training AI to scan public health records, prescription data, medical insurance claims, as well as their internal data. Novartis confirmed that the AI tool it uses has made enrolling patients in trials faster, cheaper and more efficient.

Johnson & Johnson Is Using AI To Ensure All Groups Benefit From Scientific Advances

Besides speeding up the process, AI is also being used to increase diversity in clinical trials. Embracing diversity is a very important aspect for medical breakthroughs considering that the effectiveness of a drug can vary depending on the race, ethnicity, age, gender and sex. Yet, about 75% of participants in clinical trials for new drug that approved in 2020 were White, leaving the Hispanic representatives with a 11% share and the Black representatives with mere 8%. This issue is particularly concerning in the case of multiple myeloma. Black Americans are twice as likely to develop this disease compared to their White counterparts yet their participation cancer in clinical trials for bone marrow cancer treatments is only 4.8%. Johnson & Johnson JNJ is among the first to take this bold step to change these statistics. By using AI, J&J pinpointed centers where this neglected race group might seek treatment. By using algorithms, J&J managed to lift the Black enrollment rate in five ongoing studies to about 10%. Moreover, J&J is using AI to increase diversity in 50 trials while planning to increase that number to 100 next year. 

Mainz Biomed Is Also Using AI For Its Product Development

Mainz Biomed NV MYNZ, a molecular genetics diagnostic company specializing in the early detection of cancer, has also boarded the AI wagon. It recently announced it has entered a strategic partnership with Liquid Biosciences, a bio-analytics company, to expand the use of its proprietary AI analysis technology platform, EMERGE.The announced collaboration is in fact a deepening of the company’s utilization of Liquid Bioscences’ EMERGE platform, or more precisely, a formal acknowledgment of the strategic relationship. With EMERGE, Mainz Biomed analyzed its ColoFuture international study and recently reported groundbreaking results from its test for colorectal cancer that integrates a portfolio of mRNA biomarkers. The study found that the company’s flagship product, ColoAlert®, has sensitivity for colorectal cancer of 94% with a specificity of 97%, along with a sensitivity for advanced adenoma of 80% that ‘opens’ the door to colorectal cancer.

With the partnership, Mainz Biomed gained the right to extend its use of EMERGE for its its product development, as well as to analyze its eAArly DETECT study, which is the U.S. arm of the ColoFuture clinical trial whose results are expected to be included in the undergoing quarter, as well as ReconAAsense, the forthcoming U.S. pivotal FDA PMA trial. If successful, Mainz Biomed will be able to advance the existing capabilities of its tests and commercialize a next-generation of its self-administered CRC screening tool. The goal of using this evolutionary platform is to develop a single fixed machine learning AI-based algorithm.

AI Saved Bayer Millions

Bayer used AI to lower the number of participants needed by several thousand for a late-stage trial for its experimental drug designed to reduce the long-term risk of strokes in adults. Buyer used AI to link the mid-stage trial results to real-world data from millions of patients in Finland and the United States to predict the long-term risks in a similar population. After it got equipped with data, Bayer kicked off the late-stage trial with fewer participants. Buyer stated that AI saved it millions while reducing the recruitment time by nine months. Strengthened by these results, Bayer now wants to test that drug, asundexia, in children with the same condition, but by using real-world patient data to generate an external control arm that could eliminate the need for the placebo patients. Due to the fact that the condition is so rare in the chosen age group, recruiting patients would be difficult while also raising an ethical dilemma of giving patients a placebo. Therefore, Buyer plans to mine anonymised real-world data of children with similar vulnerabilities, hoping it will be enough to help discern how effective the drug is. Amgen already adopted this approach for its drug Blincyto that treats a rare form of leukaemia and consequently received U.S. approval. But once the drug went on sale, Amgen also had to conduct a follow-up study to confirm its beneficial effect.

The Main Risk With AI Use Is In Getting The Wrong Answer

As an associate director from FDA's Center for Drug Evaluation and Research, in charge of real-world evidence analytics in the Office of Medical Policy, John Concato, put it, using AI increases the risk of not getting the right answer as to whether a drug is efficient. Although the head of global development operations at Novartis, Badhri Srinivasan, openly stated the benefits harnessing the power of AI, even a Novartis representative acknowledged that the tool as only as good as the data it gets. Undoubtedly, AI is also bringing a new set of challenges as this developing technology also brings the risk of making things worse than they are through algorithmic bias. This is why J&J formed an AI ethics council to monitor trials and remove data bias while increasing representation but there will be an increasing need for regulatory support as this groundbreaking technology continues to disrupt every aspect of healthcare.

For the good or the bad, AI is here to stay.

DISCLAIMER: This content is for informational purposes only. It is not intended as investing advice.