Abstract
In rare diseases, choosing the right measure of success can matter just as much as developing the treatment itself. My capstone project, Clinical Trial Data Analysis: Finding Endpoints that Matter to Choroideremia Patients, studies how existing natural history data can be used to identify better clinical trial endpoints for choroideremia(CHM). I was motivated by how current trial measures often miss younger patients with the disease. My STS research paper, Rethinking “Success” in Rare-Disease Trials: Sociotechnical Pathways for Stabilization-Focused Endpoints in CHM, examines why outcomes that reflect preserved vision are often harder to accept as evidence of benefit. I pursued this research because patients with CHM often value keeping the vision they still have, while trial systems often reward short-term improvement instead. At their core, both projects ask what success should look like in a rare-disease trial. The capstone approaches that question through data analysis, while the STS research explains how clinical, regulatory, and social systems shape which outcomes are treated as meaningful.
My capstone project addresses a major problem in CHM research: common FDA-desired endpoints, such as visual acuity boosts, may not fully capture disease progression or treatment benefit. CHM progresses slowly and unevenly, so trials can miss important changes if they rely too heavily on one measure. To address this, our team used data from the NIGHT natural history study to compare functional and structural endpoints such as visual acuity, microperimetry, contrast sensitivity, ellipsoid zone area, and fundus autofluorescence. We also used age-stratified analysis to compare patients aged 40 and younger with patients over 40, since disease progression is not uniform across the lifespan. This approach aimed to identify more sensitive endpoints for future trials.
The capstone project shows that endpoint performance depends heavily on patient age and disease stage. In the data we analyzed, younger patients generally showed higher and more stable microperimetry and contrast sensitivity values over time, while older patients had lower baseline values and greater variability. These findings may improve group selection and should guide age-specific trial design. Overall, our project supports better endpoint selection, better patient selection, and less reliance on visual acuity improvement alone.
My STS paper asks why stabilization-focused endpoints and patient-reported outcomes struggle to count as primary evidence in CHM trials. Patients often define success as keeping vision stable, especially night vision, mobility, and daily independence, while trial systems often favor outcomes that show obvious short-term improvement. To study this issue, I used a document-based and case-based approach. I analyzed FDA rare-disease endpoint materials, published CHM trials, and payer and health technology assessment perspectives to understand how different groups decide what counts as credible evidence. I framed the paper around the idea that endpoints are not just technical measurements. They are also shaped by institutions, standards, and decision-making systems.
My STS research found that stabilization often fails to count as strong evidence not because it lacks value, but because it is harder to interpret within current regulatory and payer systems. Improvement-centered endpoints, especially visual acuity gains, remain dominant because they are familiar, standardized, and easier to defend. In contrast, stabilization needs stronger support through natural history data, repeatability benchmarks, and clear links to patient function. The paper argues that stabilization can become more credible if researchers anchor it to expected decline, combine functional and structural evidence, and frame results in ways that are useful for both regulators and payers. Taken together, my capstone and STS projects show that improving rare-disease trials is not only a technical problem. It is also a problem of how evidence is defined, interpreted, and accepted.
