New JCI Viewpoint: FDA Must Rethink How It Evaluates Precision Medicines for Rare Diseases
Authors argue that rare disease innovation requires learning more from fewer patients—not enrolling more patients
NEW YORK, NY — August 17, 2026 — Precision medicine is dividing once-common diseases into increasingly smaller, genetically defined patient populations. The traditional clinical trial model is thus reaching its practical limits say the authors of an important Viewpoint published in the August 16th issue of the Journal of Clinical Investigation. They argue that the future of regulatory science will no rely solely on RCTs but instead increasingly on extracting stronger causal evidence from subjects that are only available in small numbers.
The article, "Do the Math: Modernizing the FDA's Evidence Paradigm for Precision Medicine," is authored by Peter J. Pitts, Executive Chairman, Brainstorm Therapeutics, former FDA Associate Commissioner and President of the Center for Medicine in the Public Interest; Turing Award laureate Judea Pearl of UCLA; bioethicist Arthur Caplan of NYU Grossman School of Medicine; and Robert Goldberg of the Center for Medicine in the Public Interest.
The authors develop the concept of "denominator collapse" -- the reality that advances in molecular biology are dividing diseases into ever-smaller patient populations. While this precision improves scientific understanding, it also makes conventional large randomized clinical trials increasingly difficult, and in some cases impossible, to conduct.
"For most of the modern era, uncertainty in drug development has been addressed by enrolling more patients," said Peter J. Pitts. "That approach transformed medicine, but it is no longer sufficient for many rare and genetically defined diseases. The question is no longer how to study larger populations, but how to learn more from smaller ones."
The article argues that the FDA's recently proposed Plausible Mechanism Framework represents an important step toward recognizing that mechanistic biology itself carries evidentiary value. The authors contend, however, that this framework should be strengthened by incorporating modern methods of causal inference—analytical approaches that integrate randomized trials, real-world evidence, natural-history studies, and external controls to determine not simply whether a treatment appears effective on average, but whether it causes meaningful benefit for individual patients.
"Precision medicine demands precision evidence," said Pitts. "Causal inference provides the mathematical framework needed to connect biological understanding with reliable regulatory decision-making when conventional statistical approaches reach their limits."
Rather than advocating lower evidentiary standards, the authors argue that causal inference formalizes scientific rigor by making assumptions explicit, testing competing explanations, and quantifying uncertainty in ways particularly suited to rare diseases and individualized therapies.
The Viewpoint also connects advances in regulatory science to economic innovation. Scientific uncertainty becomes regulatory uncertainty, the authors note, increasing development costs and discouraging investment in therapies for small patient populations. Better methods for evaluating causal evidence could reduce those uncertainties, lower the cost of capital, and encourage investment in treatments for devastating rare diseases.
Looking ahead to negotiations over the next Prescription Drug User Fee Act (PDUFA VIII), the authors recommend that Congress and the FDA invest not only in faster review technologies, but also in the scientific expertise needed to evaluate causal evidence consistently across emerging therapeutic platforms.
"The future of precision medicine will not be determined by our ability to enroll more patients," the authors conclude. "In many diseases, those patients simply do not exist. Success will depend on our ability to learn more from the patients we have."
About the Authors
Peter J. Pitts ([email protected] ) is President of the Center for Medicine in the Public Interest and former Associate Commissioner of the U.S. Food and Drug Administration. Judea Pearl is Chancellor's Professor of Computer Science and Statistics at UCLA and recipient of the ACM A.M. Turing Award. Arthur L. Caplan ([email protected] ) is Professor Emeritus at NYU Grossman School of Medicine. Robert Goldberg ([email protected] ) is Vice President of the Center for Medicine in the Public Interest.
Article Information
Pitts PJ, Pearl J, Caplan AL, Goldberg R. Do the Math: Modernizing the FDA's Evidence Paradigm for Precision Medicine. Journal of Clinical Investigation. 2026;136(16):e211143.
