Dr. Peter Marks On Why Gene Therapy May Need A New Regulatory Playbook

By Erin Harris, Editor-In-Chief, Cell & Gene
Follow Me On Twitter @ErinHarris_1

I attended the 2026 @Philly Cell and Gene Therapy Annual Conference last week, and Dr. Peter Marks, SVP of Molecule Discovery and Head of Infectious Disease at Eli Lilly (and former CBER Director), gave a keynote that was equal parts history lesson, policy challenge, and call to action.

Speaking to a standing-room-only crowd, Dr. Marks celebrated how far the science has come. He shared that we’re at an exciting inflection point; gene therapy is advancing faster than ever, creating an opportunity for regulatory and commercialization systems to evolve and bring life-changing treatments to thousands more rare disease patients.

Setbacks Do Not Stop the Science

Dr. Marks opened by tracing gene therapy’s uneven history. The concept first appeared in scientific literature in 1972. The first U.S. gene therapy trial launched in 1990. Then came major setbacks, including the death of Jesse Gelsinger in a University of Pennsylvania adenoviral vector trial that reshaped the field’s safety culture and slowed progress for years.

The completion of the Human Genome Project, advances in immunology, and the emergence of viral vector engineering helped unlock entirely new therapeutic possibilities. Dr. Marks noted how Philadelphia has been central to many of those breakthroughs. CAR-T therapies pioneered at the University of Pennsylvania led to the first approvals in 2017. That same year, Spark Therapeutics won approval for Luxturna, the first in vivo AAV gene therapy approved in the U.S.

Dr. Peter Marks

Since then, progress has been substantial. Dr. Marks pointed out the roughly 28 approved gene therapies now available worldwide as evidence that the field has matured. He stated that pace is nowhere near fast enough. “We have 28 [approved therapies],” he said. “But there are more than 10,000 rare diseases.”

Throughout the keynote, Dr. Marks returned repeatedly to one idea. Gene therapies are fundamentally different from traditional drugs, and regulators may need to treat them differently.

Indeed, he highlighted several examples that illustrate the transformative potential of the modality. He described the long-term remission seen in pediatric leukemia patient Emily Whitehead following CAR-T treatment. He also referenced one of the field’s most closely watched recent stories. Baby KJ, treated at Children’s Hospital of Philadelphia with an individualized CRISPR therapy for CPS1 deficiency, is now thriving at two years old. “These are the things we can aspire to with gene therapy,” he said.

The Challenge is Not Scientific Possibility. It’s Scalability.

Many rare diseases affect only dozens, or sometimes only single digit numbers, of patients worldwide. Traditional clinical development frameworks were not designed for populations that small. Natural history data are often incomplete. Recruiting enough patients for randomized studies can be impossible.

At the same time, manufacturing remains expensive and operationally complex. Dr. Marks warned that unless industry finds ways to reduce production costs and simplify delivery, genetic medicine will remain inaccessible in many parts of the world.

He argued that this issue is especially urgent in low- and middle-income countries, where one-time therapies could potentially provide enormous public health value. “Gene therapy would be best adapted in some of those countries, and yet it’s out of reach because of the cost.”

He also spent considerable time discussing regulatory friction. Today, developers often must repeat costly submissions and reviews across multiple jurisdictions even for therapies intended for extremely small patient populations. He noted that redundancy discourages global expansion and limits access.

His proposed solution suggests that regulators should begin thinking more seriously about platform-based or process-based approvals for certain classes of genetic medicines.

Under such a framework, developers could establish a validated manufacturing and delivery platform, such as an AAV vector system or a CRISPR editing backbone and then reuse that platform across multiple disease targets without restarting the entire regulatory process each time.

Instead of treating every new insert or guide RNA as a separate product, regulators could focus on the novel components while relying on previously established manufacturing, toxicology, and quality data. “It sounds crazy,” Dr. Marks admitted. “But it’s not as crazy as you think.”

He compared the concept to personalized cancer vaccines, where regulators effectively approve the manufacturing process even though each individualized product differs from patient to patient. He argued that a similar framework could dramatically accelerate development for ultra rare diseases while reducing the cost burden associated with repeated toxicology studies, manufacturing validation, and regulatory filings.

He also emphasized the role advanced manufacturing and AI could play in supporting that transition. Automation, analytics, and AI enabled regulatory workflows could all help streamline production and reduce operational overhead.

Perhaps the most provocative part of the keynote centered on benefit risk tolerance.

Dr. Marks suggested that regulators, developers, and patient communities may need to rethink how uncertainty is handled in devastating rare diseases where no treatment options exist. “We don’t want the bar to be set so high that we have 95 percent of children not getting potentially effective therapies to avoid 5 percent ineffective therapies,” he said.

Questions from the Audience on Single Arm Trials and Randomized Controlled Studies

As is usually the case with questions from the audience, the brief Q&A at the end of Dr. Marks’ address proved to be beneficial to everyone in the room.

Asked about recent FDA guidance related to single trial approvals and plausible mechanism pathways, Dr. Marks acknowledged growing concerns around consistency within the agency following staffing disruptions and organizational changes. “The number one thing I heard from companies was consistency,” he said. He noted that sponsors can often avoid randomized studies when therapies demonstrate dramatic clinical effects or when highly reliable biomarkers clearly demonstrate biological benefit.

“The problem is trying to do a randomized clinical trial for some of the rare diseases where there are 20 people in the United States,” he said. “That’s just not going to work.”

Another audience member asked how the industry can balance increasingly sophisticated capsid engineering with the need for scalable manufacturing and platform standardization.

Dr. Marks acknowledged the tension. Better capsids may improve potency and tolerability, but under current regulatory frameworks, changing capsids can effectively create an entirely new product development program.

He suggested future regulatory models may need to allow more flexibility for iterative improvements without forcing sponsors to restart from scratch each time.

By the end of the session, Dr. Marks had laid out a vision that extended well beyond gene therapy itself. His broader argument was that biomedical regulation was largely designed around mass market drugs for large patient populations. Genetic medicine challenges those assumptions at nearly every level.

He warned that the current system simply cannot keep pace with the number of rare diseases waiting for treatment. “A calculation was done, and if we keep the current pace up for rare disease treatments, it will take something like 600 years to get through another couple hundred diseases.” For a field built around the idea of rewriting biology itself, Dr. Marks made clear that rewriting regulatory frameworks may now be just as important.