TL;DR:
- FDA has observed a significant increase in drug applications incorporating AI and ML elements over the past five years.
- In 2022, the FDA received 170 submissions integrating AI and ML components, compared to just three in 2018.
- AI and ML tools are used to enhance various aspects of drug development, including study design, patient selection, dosing optimization, and toxicity prediction.
- Anakinra, a drug used for COVID-19 treatment, received emergency use authorization with the help of a machine learning model for patient selection.
- The FDA supports the appropriate use of AI/ML in drug development and aims to provide regulatory clarity through workshops, white papers, and guidance documents.
- Analysis reveals that the majority of AI/ML-driven submissions were in oncology (27%), followed by psychiatry (15%) and gastroenterology (12%).
- These AI/ML elements are most frequently utilized during the clinical phase of drug development.
- The FDA seeks public feedback on the use of AI/ML to inform the regulatory landscape and ensures safety and efficacy standards.
Main AI News:
The utilization of artificial intelligence (AI) and machine learning (ML) technologies in drug development has witnessed a remarkable upswing, as indicated by the US Food and Drug Administration (FDA). Hao Zhu, the Director of the Division of Pharmacometrics in FDA’s Center for Drug Evaluation and Research (CDER), disclosed that over the past five years, the agency has observed a significant surge in the number of drug applications incorporating AI and ML components. In fact, in 2022 alone, the FDA received a staggering 170 submissions that integrated these groundbreaking elements into their core framework.
This notable shift in the industry landscape becomes even more apparent when we examine the statistics provided by Zhu. Back in 2018, a mere three submissions encompassed AI/ML components. This exponential growth was highlighted by Zhu during his insightful discourse at the DIA Global Annual Meeting on June 27th.
The integration of AI and ML tools has proven instrumental in enhancing various facets of the drug development process. These cutting-edge technologies aid in designing studies, identifying suitable patients for clinical trials, assessing patient risks, optimizing dosing protocols, evaluating endpoints and biomarkers, predicting drug toxicity, and even facilitating drug discovery and repurposing initiatives.
To illustrate the tangible impact of AI in a real-world scenario, Zhu cited the emergency use authorization (EUA) granted for Anakinra, a drug employed in the treatment of hospitalized COVID-19 patients. In this particular case, a machine learning model was utilized to meticulously select patients for inclusion in clinical trials. Zhu further emphasized that two distinct machine learning models were employed to derive the score-based rule governing patient selection.
In light of these advancements, Zhu expressed the FDA’s unwavering support for and encouragement of the judicious implementation of innovative tools such as AI and ML to streamline new drug development. In fact, an enlightening article published in the esteemed journal Clinical Pharmacology & Therapeutics in June 2022 delved into the extensive application of AI and ML in regulatory submissions for drug development between 2016 and 2021. The article specifically focused on scouring submissions for key terms like “machine learning” or “artificial intelligence.”
The comprehensive analysis conducted revealed that in 2021, the FDA received a staggering 132 applications incorporating AI and ML elements for novel drugs and biologics. Among these submissions, a significant majority—27%—were primarily in the domain of oncology, followed by 15% in psychiatry and 12% in gastroenterology.
Considering the various stages of drug development, these AI and ML elements were predominantly employed during the clinical phase, followed by the nonclinical, postmarketing, and drug discovery and repurposing stages.
To ensure that the regulatory landscape remains well-informed and adaptable in this rapidly evolving realm, the FDA seeks to engage with the public and gather valuable feedback on the appropriate utilization of AI and ML technologies. In pursuit of this goal, the agency previously solicited input through two discussion papers focusing on the use of AI and ML in drug development and manufacturing. The FDA aims to carefully review and analyze the comments received, with the intent of developing public workshops, white papers, and guidance documents that offer regulatory clarity based on the invaluable feedback received.
As the pharmaceutical industry continues to embrace the power of AI and ML, the FDA remains committed to fostering an environment conducive to innovation while ensuring the highest standards of safety and efficacy. With the potential to revolutionize drug development, these technologies hold the key to accelerating the delivery of novel therapies to patients in need, ultimately transforming healthcare as we know it.
Conclusion:
The rapid rise in drug and biologic submissions incorporating AI and ML technologies signals a transformative shift in the market. This surge highlights the industry’s growing recognition of the potential of these innovative tools in expediting drug development processes, enhancing patient selection, and optimizing treatment outcomes. The FDA’s support and encouragement of AI/ML adoption reflect the agency’s commitment to fostering an environment conducive to innovation while ensuring regulatory clarity. As AI and ML continue to revolutionize the pharmaceutical landscape, companies operating in this market need to stay ahead of the curve by leveraging these transformative technologies to drive efficiency, improve patient outcomes, and deliver novel therapies in a timely manner.
