September 28 2023
The integration of AI into drug discovery has piqued the interest of many players within the pharmaceutical industry, with regulatory changes expected in due course.
Drug discovery processes are known to be costly, time-consuming, and at times inaccurate.
Traditional drug discovery methods are known to be ‘hit-and-miss’ in nature. They are associated with high costs alongside low accuracy and low success rates, with only 10% of candidates making it past pre-clinical development. Morgan Stanley estimated the cost of R&D to be “as high as $2.5 billion per marketed therapy when factoring in abandoned trials and clinical failures".
Technological advancements have provided the opportunity for streamlining drug development processes with potentially improved outcomes.
The current challenges have led to an increased demand for novel applications that could at the both accelerate and de-risk the drug development process. This would not take the place of skilled scientists, but rather lend a helping hand in groundbreaking research. Already, AI platforms, such as AlphaFold by Deepmind, offer the possibility to predict protein structures and drug candidate interactions to streamline the candidate selection process.
Artificial Intelligence (AI) and machine learning (ML) are not new to drugmakers. The Head of Information Management at Pfizer previously highlighted that “using data to make faster decisions on a medicine’s potential, would allow to re-allocate resources, dollars, and expertise to the next promising candidate faster”. Indeed, AI had the potential to cut down costs and lower the barriers to start-up discovery by reducing exploratory research timelines, aid biomarker identification, define target patient populations and by optimizing clinical trial design.
Adoption has so far been driven by a new wave of AI-based biotechs with R&D workflows built around AI tools. Bigger pharmaceutical players have been more careful after notable failures (e.g., Sensyne Health), instead easing in via partnerships.
From in silico approaches to actual patients, how are pharma companies leveraging AI?
To successfully integrate AI into drug discovery processes, the limitations must first be acknowledged.
Data collection is a crucial hurdle for AI platforms as they need vast datasets to generate precise results through ML algorithms. Having access to these past datasets is critical for AI-assisted drug development, but this may not be feasible in certain instances, such as with rare diseases. Moreover, the dependability of AI platform outcomes hinges on the quality of data used for machine learning since inadequate data can lead to prejudiced predictions. This needs to be heavily monitored and regulated.
The regulatory landscape across markets is already adapting to the increased AI use
The FDA’s Center for Drug Evaluation and Research (CDER) released an initial discussion paper in May 2023. This recognized AI and ML playing a significant role in drug development and personalized treatment approaches, noting plans to be flexible and collaborative with regard to its regulatory framework.
In Europe, the EMA followed suit by publishing a draft reflection paper on the use of AI in drug development in July 2023. This highlights that pharmaceutical companies will be responsible for ensuring the AI they use is “fit for purpose and in line with ethical, technical, scientific, and regulatory standards as described in GxP standards and current EMA scientific guidelines”.
What to expect going forward
With hundreds of AI-driven drug-discovery companies, AI has promise to revolutionize drug discovery. In the coming years, bigger pharmaceutical players stand to benefit by increasing the use of AI across all stages of a product’s lifecycle. This will be accompanied by clearer regulatory frameworks set out across markets.
To stay ahead of the rapidly advancing companies in the space, it will be important to be prepared and proactive. The Eradigm team can provide comprehensive support to pharmaceutical companies that are willing to take the plunge, accelerate the adoption of AI-led discovery techniques and create compelling value propositions.