Moving Beyond Traditional Roles With AI in the Era of Precision Medicine

A paradigm shift is underway in cancer treatment development, combining regulatory innovation with technological advancements in artificial intelligence (AI). In this edition of The Way Ahead: The Convergence of Technology and Cancer Care, Waqas Haque, MD, MPH, Hematology/Oncology Fellow at the University of Chicago, explores emerging trends incorporating AI applications into precision medicine approaches in oncology.

Regulatory Shift From Animal Testing Toward In Silico Models

The FDA recently announced its plans to phase out animal testing requirements for drug development, particularly for monoclonal antibodies and other therapies. This shift marks a departure from decades of standard practice, with FDA Commissioner Martin Makary emphasizing that the new approach will “get safer treatments to patients faster and more reliably, while also reducing research & development (R&D) costs and drug prices. It is a win-win for public health and ethics.”

The FDA’s new framework will replace traditional animal testing with more effective, human-relevant methods including:

• AI-based computational models of toxicity
• Human organ model-based lab testing (organoids and organ-on-a-chip)
• Real-world human data analysis

This adjustment builds upon the FDA Modernization Act 2.0 that was passed by Congress in 2022, which authorized non-animal alternatives in preclinical testing. Implementation begins immediately for investigational new drug applications, with a roadmap including a pilot program for monoclonal antibody developers using primarily non-animal testing strategies. Similar efforts are underway through the National Institute of Health’s new Office of Research Integrity for Validation and Algorithm.

The Limitations of RCTs in the Precision Medicine Era

While randomized clinical trials (RCTs) have long been considered the gold standard for drug approval in oncology, they present significant challenges for precision oncology, particularly for patients with rare mutations for their particular cancer:

1. Patient heterogeneity: Traditional RCTs rely on large, homogeneous patient populations, whereas precision oncology targets increasingly specific molecular subsets.
2. Small patient pools: For rare mutations, identifying sufficient patients for adequately powered trials is often impossible within reasonable timeframes. For example, among patients with non–small cell lung cancer (NSCLC), EGFR mutations are present in up to 50% of Asian patients and around 14% of patients overall, but ALK rearrangements are found in only 2% to 7% percent of NSCLC cases.
3. Statistical challenges: Low patient numbers “would in most settings preclude meaningful empirical inferences about treatment benefits or risks,” as acknowledged by the FDA itself.
4. Inefficient resource allocation: The current paradigm means that patients with rare mutations are often underserved with limited therapeutic options.

In response, the FDA has been developing more flexible evidence standards for low-frequency mutations, potentially favoring internal validity over traditional external validity requirements. This creates opportunities to further clinical research on rare mutations in both rare and more common diseases.

The AI Revolution in Drug Discovery

The integration of AI into drug discovery has already demonstrated remarkable potential to accelerate development timelines. Ren et al (2023) describes how a team of researchers used AlphaFold, an AI-powered protein structure database, to design and synthesize a potential treatment for hepatocellular carcinoma targeting CDK20 in just 30 days. The study demonstrates how AI can dramatically compress development timelines that traditionally take years. Other companies have reported similar breakthroughs, with some AI platforms capable of delivering a phase 1 clinical candidate in under 30 months from initial target discovery.

Looking Ahead

In our future blog posts, we’ll explore other topics related to AI-assisted drug discovery, including:

• The integration of multimodal data (clinical, molecular, and imaging)
• How AI foundation models are transforming candidate selection and clinical trial design
• Regulatory considerations for AI-discovered therapeutics
• The potential for AI to democratize drug discovery for ultra-rare conditions

The convergence of regulatory flexibility for rare mutations with AI-powered discovery platforms promises to fundamentally transform how we develop treatments for cancer patients most in need of innovation.

Stay tuned for future monthly posts exploring the intersection of precision medicine, regulatory science, and AI in oncology!

About Dr. Haque

Waqas Haque, MD, MPH, is a Hematology/Oncology Fellow at the University of Chicago. He recently completed his Internal Medicine Residency at New York University (NYU) in a Clinical Investigator Track. Dr. Haque has balanced his patient care work with a variety of research projects. During his fellowship training at the University of Chicago, he plans to further his work in innovative clinical trial design, value-based care delivery to cancer patients, and clinical investigation.