Patient organizations have stopped asking pharma for a seat at the table. In rare disease, a growing number are building their own table, and increasingly they are the ones sponsoring the trials. That was the throughline of a Clinical Leader conversation between Craig Lipset, co-founder and co-chair of the Decentralized Trials & Research Alliance and founder of Clinical Innovation Partners, and Dr. Derk Arts, founder and CEO of Castor, that aired on 17 September 2026. The math explains why it is happening now. In a strong year the FDA’s drug center approves roughly 50 new medicines. In that same year, researchers characterize somewhere between 150 and 250 new rare diseases. With more than 10,000 rare diseases on the books, and industry able to pursue only the commercially returnable ones, someone has to move the rest. The first hands up have been patient-led nonprofits.
What the session covered
For fifteen years the industry line was “include patients in design.” Lipset’s argument is that the ground has already shifted past that. These groups are no longer advisory voices.
“This isn’t partnering with patients. This is the patients who are driving the car.”Craig Lipset
Some patient-led nonprofits now sponsor molecules in Phase 1 and beyond with no pharma partner at all. The FOXG1 Research Foundation and the TESS Research Foundation are both advancing gene therapies that the FDA cleared to begin trials in the past year. When these groups invite industry to their table, Lipset said, they often do not get the same turnout they were promised for years, and his challenge to industry was blunt: answer the call.
The hard part is evidence, and it has a specific shape. Lipset relayed a challenge from a workshop led by Janet Woodcock, the former FDA principal deputy commissioner and longtime director of its drug center: when a population is too small for a randomized trial, the answer is not to run an underpowered one. Underpowered trials waste patients’ time and can bury good medicines as failures. The law, Arts noted, calls for substantial evidence from adequate and well-controlled studies, not specifically a randomized trial, which leaves room for self-controlled designs, natural history and real-world data, and model-informed external controls.
Arts pushed hardest on who should control the data these studies run on. He described helping an ultra-rare group stand up its own natural history study, with families pulling together their own medical records (a patient-mediated, direct-to-patient route) that then feed AI-assisted data extraction with human review in Castor Catalyst, work his team has taken on pro bono so that parents stay in charge of the data rather than handing it off.
“It’s quite elegant to have patients actually govern their own data sets and then decide who gets access and for what it gets used.”Derk Arts
Lipset’s analogy was the credit report. You can see who pulled your financial data, lock it, and audit the trail, so why not your health data? “Data sharing for research shouldn’t happen in the dark,” he said, pointing to a bipartisan run of US data-access policy and the European Health Data Space as parallel signs the direction is set, if slow.
What’s in the full recording
The recording goes further than a recap can. Lipset walks through three ways a patient group can actually recover money from a derisked asset, including a pediatric priority review voucher that has sold on the open market for roughly $100 to $200 million (19:54). He and Arts get specific on how to cluster related rare conditions, such as FOXG1 and CACNA1A, into a single master protocol modeled on oncology’s Lung-MAP and I-SPY 2 (40:56). There is a candid exchange on why fifteen years of data-portability policy still has not given patients a usable “credit report” for their health records (31:16). And Lipset closes on obsolescence, with a pair of subway-token cufflinks, in a way that reframes what AI should mean for everyone working in trials (49:05).
Watch the full conversation
It is built for clinical-development leaders deciding how to work with the patient organizations already running ahead of them, and for patient-organization leaders mapping their own data and evidence strategy.
Then go deeper: the insight brief on patient-led drug development in rare disease is the practical companion to this conversation.
That brief traces the path from a registry to a trial, and shows how patient-reported outcomes captured through ePRO and a lean, patient-owned clinical data management approach make that path runnable for a small group.
Frequently asked questions
What is patient-led drug development?
Patient-led drug development is when a patient organization, usually a nonprofit founded by affected families, takes the sponsor role in developing a therapy rather than only advising a pharma company. In rare disease these groups now fund natural history studies, select targets, and sponsor early-phase trials themselves, often because no commercial sponsor will take on a population that small.
Why are patient organizations sponsoring their own clinical trials?
The number of newly characterized rare diseases each year far outpaces the number of new medicines the industry approves, and most rare conditions are too small to be commercially attractive. As Lipset put it in the session, industry will chip away at the commercially returnable diseases first, so patient-led nonprofits are stepping in to move the rest, in several cases as the sole sponsor of a molecule in early clinical development.
How can a small population generate evidence without a large randomized trial?
FDA approval requires substantial evidence from adequate and well-controlled studies, not specifically a randomized controlled trial. For very small populations that opens the door to self-controlled study designs, natural history and real-world data, and model-informed approaches such as external control arms and digital twins. The goal discussed in the session is a shared taxonomy so regulators can set a proportionate level of evidence for each type of rare indication.
Editorial note
During the live session, the scarcity of new approvals was described as roughly 60 to 80 new medicines a year. For accuracy: the FDA’s Center for Drug Evaluation and Research (CDER) approved 50 novel drugs in 2024 and 55 in 2023, averaging about 50 over the past decade (per FDA Novel Drug Approvals). The contrast with the pace of new rare disease discovery holds, and this post uses the verified figure. On the priority review voucher: the pediatric program lapsed at the end of 2024 and was reauthorized in 2026 through 30 September 2029. A move to make it permanent has been proposed but is not yet law.
Sources
- FDA, Novel Drug Approvals 2024 and 2023, Center for Drug Evaluation and Research. Available at: https://www.fda.gov/drugs/novel-drug-approvals-fda
- Consolidated Appropriations Act, 2026 (P.L. 119-75), Rare Pediatric Disease Priority Review Voucher reauthorization through 30 September 2029. Available at: https://www.congress.gov