Catherine Batsford: Welcome back to Research Ethics Reimagined.
Dan McLean: I’m your host, Dan McLean.
Catherine Batsford: And I'm your cohost, Catherine Batsford. Today we are joined by a leading expert in the intersection of law, medicine, and research ethics who's working to shape the national standards for how we conduct medical science. Holly Fernandez Lynch is the Associate Professor of Medical Ethics at the University of Pennsylvania's Perelman School of Medicine. Holly's work focuses on clinical research ethics, institutional review board quality, and access to investigational medicines and FDA pharmaceutical policy, especially related to early approval pathways and post market study requirements. She is a former board member of PRIM&R and figure in the research oversight community.
Her influence extends deep into federal policy. She served as a member of SACHRP, the secretary's advisory committee of human research protections, which provide guidance to the U. S. Department of Health and Human Services. She is also co founder and co chair of ARIO, which is an ongoing collaborative effort to understand and evaluate and improve IRB quality and effectiveness.
Holly has testified before Congress and authored a wide range of reports, including those which explore the promise and perils of FDA's new plausible mechanism pathway, which is what we'll focus on during our episode. Holly brings a vital perspective to some of the most pressing questions in scientific research today. Holly, thanks again for being here.
Holly Fernandez Lynch: I’m so glad to be with you.
Dan McLean: We’d like to start with how you became interested in bioethics and how that led to your focus on this area of FDA policy.
Holly Fernandez Lynch: I often tell the PRIM&R community that I found bioethics by accident. As a freshman in college, I enrolled in a math class that was more advanced than I was prepared for. I quickly realized I was in over my head and asked my adviser to move me out of it. She suggested “Introduction to Biomedical Ethics.” I agreed, largely because it was not math, and I immediately fell in love with the field.
I had long been interested in science, medicine, and health, but bioethics was still emerging at that time. I had not encountered it in high school. The class exposed me to pressing ethical issues—everything from reproductive technologies to research policy—and I realized I was particularly interested in the policy dimensions. That led me to law school.
After law school, I joined a law firm’s FDA practice group, where I learned how clinical trials are designed and how pharmaceutical and biotechnology companies approach regulation. I later worked at the Division of AIDS, which provided a different perspective—how a government sponsor thinks about its ethical and regulatory responsibilities, including in global research contexts. Today, much of my work centers on FDA drug and biotechnology policy, especially questions about evidentiary standards and regulatory flexibility.
The plausible mechanism pathway sits squarely within that work because it asks: How much evidence is enough before we allow a product onto the market, particularly for rare and ultra-rare diseases?
Dan McLean: The phrase “plausible mechanism” is not intuitive. What does it mean, and what problem is the FDA trying to solve?
Holly Fernandez Lynch: A plausible mechanism refers to a strong biologic rationale for why an intervention should work. For example, if we know that a disorder is caused by a particular gene or biological process, and we can intervene directly on that process, we might reasonably expect improved outcomes. That expectation is the “plausible mechanism.”
However, plausibility does not guarantee success. Sometimes an intervention appears mechanistically sound but fails to produce the expected clinical benefit.
For common diseases, we would test a plausible mechanism in a well-designed randomized controlled trial. But for diseases affecting only one or a handful of individuals—true “N-of-1” or “N-of-few” situations—traditional clinical trials are not feasible. In those cases, requiring full-scale clinical trials for each genetic variant becomes impractical.
Historically, every molecular change to a product could be treated as a new drug, requiring new investigational applications and supporting data. For individualized therapies built on platform technologies, that approach becomes unworkable.
Dan McLean: Can you explain what you mean by “platform technologies”?
Holly Fernandez Lynch: Platform technologies share a common backbone, while only a specific component is individualized.
One example is antisense oligonucleotides (ASOs), which influence how messenger RNA reads genetic code. The backbone of the ASO remains the same, but the nucleotide sequence can be tailored to a particular patient’s mutation. The first widely discussed individualized therapeutic of this type was developed for a patient named Mila in 2018.
A more recent example involves gene editing. An infant referred to publicly as “KJ” was treated at Children’s Hospital of Philadelphia for a urea cycle disorder caused by a specific genetic variant. His team developed a base editor delivered through a lipid nanoparticle. The delivery system and core editing machinery were consistent, while the guide RNA was customized to target his specific mutation.
Under a traditional regulatory framework, each change to the guide RNA might be considered a new product, requiring new submissions and supporting data. That would make individualized treatment effectively impossible.
Catherine Batsford: How expensive would that be?
Holly Fernandez Lynch: It would be extraordinarily expensive and time-consuming. You could not realistically repeat the entire development process for every individual variant. FDA has recognized this challenge for several years and issued guidance directed at clinician-investigators treating individual patients, including recommendations on consent and manufacturing. But those guidances were not designed to support scalable commercialization.
Dan McLean: How was KJ able to be treated?
Holly Fernandez Lynch: KJ was treated under a single-patient expanded access investigational new drug application. Expanded access is an established pathway for patients who cannot enroll in clinical trials and lack satisfactory treatment alternatives. In his case, FDA reviewed the application quickly.
What was unique was the individualized gene-editing approach and the close collaboration between the scientific team and FDA. In describing that experience, FDA later articulated factors supporting a plausible mechanism pathway: clear molecular cause, targeted intervention, known natural history of the disease, confirmation of target engagement in preclinical models, and observable clinical improvement beyond what would be expected from natural history alone.
Those elements now serve as reference points for the plausible mechanism framework.
Catherine Batsford: There is also an equity concern. How do we move from one patient at the right institution to broader access?
Holly Fernandez Lynch: That is a central ethical issue. If each case requires a bespoke regulatory solution, access will be limited to those in well-resourced settings. The goal of the plausible mechanism pathway and umbrella trial approaches is to allow shared data across variants, enabling more scalable development and eventual commercialization. Commercialization matters for coverage, because insurers typically do not pay for investigational products.
However, even with commercialization, pricing and equitable access remain open questions.
Catherine Batsford: Gene editing also raises unique consent challenges. Once administered, it cannot simply be withdrawn.
Holly Fernandez Lynch: Correct. Gene editing may involve permanent or long-lasting changes. There is also the possibility of off-target effects—unintended genetic changes. Preclinical models help, but they are not perfect. In cases like KJ’s, where the alternative may be death or transplant, families may reasonably accept substantial uncertainty. That context heightens the importance of careful IRB review and clear communication in informed consent.
Dan McLean: This seems like FDA being appropriately flexible. Yet you have expressed concerns. What are they?
Holly Fernandez Lynch: I view the pathway as promising for true N-of-1 and N-of-few situations. My concern is whether it might be extended to contexts where randomized trials are feasible. Plausible mechanisms can fail, and robust clinical evidence remains important where it can reasonably be generated. FDA has described the pathway in a way that leaves some ambiguity about its scope.
I also have process concerns. The pathway was introduced through a journal article rather than through draft guidance or regulation. Guidance documents typically undergo public comment, which supports transparency and stakeholder engagement. Regulations provide even greater clarity and durability. If this pathway is intended to shape long-term policy, formalization would provide greater stability.
Dan McLean: You have also referenced changes in administrative law.
Holly Fernandez Lynch: Yes. Recent Supreme Court decisions have altered how courts evaluate agency interpretations of statutes. Although FDA retains regulatory authority, its interpretations may be more vulnerable to judicial challenge. Congressional action affirming FDA’s authority or codifying a framework for this pathway would provide the greatest durability, though legislative action is never simple.
Dan McLean: Looking ahead, where might this lead?
Holly Fernandez Lynch: For rare and ultra-rare genetic disorders, particularly those affecting children, this could be transformative. Sequencing costs have fallen dramatically. If regulatory flexibility, scientific innovation, and commercial incentives align, we may see significant advances in personalized genetic medicine.
Dan McLean: For IRBs, what should be top of mind?
Holly Fernandez Lynch: These cases require early and sustained collaboration between investigators and IRBs. IRBs should focus on mechanism, preclinical evidence, potential off-target effects, monitoring plans, and clarity in consent. This is not a routine submission; it is a partnership to protect participants while enabling responsible innovation.
Dan McLean: Holly, thank you for this thoughtful discussion.
Holly Fernandez Lynch: Thank you for the opportunity. It is remarkable to consider what these technologies may make possible.
)