Dr. Leen Kawas, a biotechnology innovator with decades of pertinent experience, delves into the potential benefits of AI in drug development.
The integration of artificial intelligence (or AI) into global companies’ operations continues to accelerate. The biotechnology industry’s highly complex operations are especially well suited to AI applications. For perspective, since 2023 the drug discovery and treatment repurposing sectors have seen a steady flow of AI-based innovations.
Leen Kawas, Ph. D. is well acquainted with the drug discovery and development cycle. As Propel Bio Partners’ Managing General Partner, Dr. Kawas oversees the biotech venture capital firm’s partnerships with start-up and early-stage biotech companies. Dr. Kawas welcomes pitches from innovative biotech entrepreneurs, especially female and minority candidates.
Prior to her current role, Dr. Kawas served as Athira’s Chief Executive Officer (or CEO). While there, she managed several successful drug development cycles. Dr. Kawas also led Athira’s initial public offering (or IPO) in September 2020. She offers well-informed insights on AI tools’ integration into drug discovery and development operations.
Reassessing AI’s Role in the Biotech and Biopharma Industries
With AI’s unexpectedly rapid infusion into diverse industries, many companies were caught without a concrete evaluation and implementation plan. In January 2024, International Society of Pharmaceutical Engineering president and CEO Thomas Hartman noted that some businesses are taking a more objective look at AI’s potential applications.
“Now, as AI fatigue begins to set in and there’s less of a race to implement AI in various processes, companies will have more time to take a step back and properly evaluate their technology and resources,” Thomas Hartman remarked. To gain expert insights, businesses may turn to AI consultants and special interest groups. These entities will ideally inform companies on generative AI integration and regulatory requirements best practices.
AI Tools Can Streamline Two Key Drug Development Steps
The extended drug development process has historically contained significant bottlenecks. Today, Dr. Leen Kawas highlighted two AI-driven advancements that together will decrease these two key phases’ duration and lower costs accordingly.
Drug Target Selection and Verification
Selecting and verifying drug development targets is an inexact and laborious undertaking. Today, AI’s deep learning algorithms enable rapid analysis of huge datasets including clinical, genomic, and proteomic data. The results lead to faster, more accurate potential target identification.
During the near future, AI algorithms will incorporate increasingly diverse datasets that include real-world patient information. In turn, this will enable more accurate drug target identification. Over time, the algorithms’ predictive accuracy will likely show additional improvements.
Dr. Leen Kawas noted that this approach shaves several months off the projected target identification timeline. A faster drug development process, and better chances of clinical trial success, will together deliver substantial research cost reductions.
The Drug Formulation Sequence
During the drug formulation sequence, researchers identify a new drug’s distinctive chemical structure and composition. An AI algorithm enables increasingly efficient formulation by predicting relevant compounds’ solubility and stability. Over time, AI will likely commence integration of quantum computing and other sophisticated simulation techniques.
Collectively, these methodologies will enable more accurate molecular behavior predictions, thus further decreasing the required drug formulation time. Looking at the bigger picture, improving drug formulation is mostly geared to fine-tuning existing processes rather than adopting new ones. That said, anything that decreases drug development timeframes is regarded as a positive occurrence.
Projected AI and Machine Learning Benefits
The drug development pipeline is a multi-year proposition, with successful candidates typically taking an average of 12 years to complete the cycle. That said, AI and machine learning (or ML) appear poised to positively impact every aspect of this time-consuming undertaking. Dr. Leen Kawas detailed four potentially game-changing applications.
- Determining Potentially Viable Drug Targets and Creating Innovative Drug Compounds
- Pinpointing Existing Drugs that are Repurposing Candidates, Decreasing Time and Development Costs
- Optimizing Clinical Trial Patient Recruitment and Real-Time Data Monitoring
- Helping Clinicians More Accurately Interpret Medical Imaging and Diagnose Diseases
Emerging Challenges in AI Integration
Besides AI tools’ potential benefits, biotech firms must overcome challenges during the technology’s use. Dr. Roger Palframan oversees U.S. research at Boston-based UCB. In a Technology Networks interview, Dr. Palframan delivered a big-picture assessment of AI’s challenges in drug discovery.
“I think having a vision for the future is great, but I’m not entirely sure how we get there. Therefore, as this evolves quickly, how do we set ourselves up for success to be agile and not constrain the power of AI, especially in early research? How do we free the AI to be able to impact discovery?
“How do we apply the appropriate risk and framework to some areas, such as patient-facing AI where it naturally needs to be more regulated? But also, how do we free it in other areas to be able to maximize its potential? The danger is applying blanket restrictions on everything that will limit innovation,” Dr. Palframan concluded.
The FDA Readies AI-Enhanced Drug Development Guidelines
Artificial intelligence’s (or AI’s) rapid integration into the biotechnology arena caught Food and Drug Administration (or FDA) regulators off guard. The FDA feels a particular sense of urgency over the lack of regulations regarding AI in the drug development process.
To address this deficiency, FDA regulators plan to introduce a 2024 draft document providing guidance on using AI/ML (or machine learning) in drug development cycles. Concurrently, the FDA will develop a database of consistent AI-related terminologies and best practices.
The FDA’s guidance will partially result from the agency’s regulatory science research. The FDA also continues to review over 300 submissions integrating AI components. These documents relate to drug discovery through post-market safety monitoring protocols. In addition, feedback on two 2023 discussion papers will help inform the FDA’s evaluation process.
Collectively, biotech leaders have asked for clarification on the projected scope of the FDA’s AI oversight. Biotech professionals also sought definitive criteria regarding the documentation (and level of detail) to support AI/ML uses in FDA drug applications.
AI Tools Can Help Predict Regulatory Obstacles
While biotech leaders await the FDA’s guidance, AI tools are currently able to analyze regulatory decisions’ historical data. Equipped with the results, AI can identify possible regulatory obstacles along with optimal approval pathways. Data-supported submissions may help streamline the regulatory review cycle. This increased efficiency contributes to faster drug introduction into the marketplace.
The Future of AI-Enhanced Drug Development
Over time, AI tools’ integration will reduce pharmaceutical firms’ high drug development costs. That frees the companies to enlist AI applications to accurately develop patient-centric medications.
Dr. Leen Kawas emphasized that the AI tools’ use will greatly streamline the drug discovery and development process. Similar advancements would take place in the clinical trials and drug approval sequences. These innovations should together enable faster, more efficient delivery of potentially useful medications to patients who can benefit from them.