Stargardt disease is a rare inherited retinal disorder that gradually affects central vision, often starting in childhood or adolescence. The phase 3 DRAGON study (NCT05244304) is evaluating tinlarebant, an oral therapy designed to slow the growth of retinal atrophy in adolescents with Stargardt disease.

Quan Dong Nguyen, MD, MSc, FAAO, FARVO, FASRS, presented interim results from the trial at the FLORetina 2025–International Congress on OCT and OCT Angiography (ICOOR), held from 4 to 7 December in Florence, Italy.¹ Nguyen is professor of ophthalmology at the Byers Eye Institute at Stanford, and professor of medicine and paediatrics at Stanford University School of Medicine in Palo Alto, California, USA.

In this Q&A conversation, the Eye Care Network caught up with Nguyen to learn more about the design, end points, and interim results of the DRAGON study, as well as how tinlarebant may affect disease progression and patient outcomes in adolescents with Stargardt disease.

Note: Transcript edited lightly for clarity and length.

What is the primary objective of the DRAGON study, and how does it differ from earlier trials?

Quan Dong Nguyen, MD, MSc, FAAO, FARVO, FASRS: The primary objective of the DRAGON study is to assess the efficacy of tinlarebant in slowing the growth rate of atrophic lesion(s) in adolescent Stargardt (STGD1) subjects. This objective aligns with the efficacy objective of the previous phase 2 study in STGD1 subjects, although the primary objectives of that phase 2 study were safety and dose determination.

What end points are most critical for evaluating efficacy in this phase 3 study?

Nguyen: The most critical assessment for the phase 3 study is the annualized rate of lesion growth in the aggregate area of atrophy (definitely decreased autofluorescence, DDAF) from baseline. In addition, the annualized growth rate in the total area of decreased autofluorescence (DAF; the sum of DDAF and questionably decreased autofluorescence, QDAF) were listed as a key secondary end point in the study.

Why is slowing the progression such an important goal for patients?

Nguyen: Stargardt disease is a hereditary retinal condition that leads to gradual loss of central vision, often starting in childhood or adolescence. Patients may still see reasonably well for some time, but underlying damage in the retina and specifically in the macula continues to expand. Slowing that damage means delaying the point at which it becomes difficult to read, recognize faces, and see at school or work. So even if the vision chart number does not change over 1 or 2 years, slowing the disease gives patients more years of useful vision.

Belite Bio recently released a news update showing that the DRAGON study met its primary end point with statistical significance.² How clinically meaningful is the improvement observed in the primary end point?

Nguyen: The improvement observed in the primary end point is a significant and exciting milestone for the medical community and people living with Stargardt disease as there are no approved therapies for Stargardt. Tinlarebant [is] the first therapeutic candidate to demonstrate statistically significant clinical benefit in a phase 3 trial for Stargardt disease. The DRAGON trial achieved the primary end point: statistically significant reduction in lesion growth rate (definitely decreased autofluorescence, DDAF) of 36% vs placebo (P = .0033) as measured by autofluorescence imaging. Such finding means that, on average, the area of retinal damage grew about one-third more slowly in treated patients than in those on placebo. Slowing lesion progression by 36% will likely lead to a significantly longer period of preserved vision.

What safety signals have emerged so far, and how are they being monitored?

Nguyen: Tinlarebant (5 mg orally, daily) has been well tolerated in adolescent STGD1 patients. There was no drug- or trial-discontinuations due to nonocular adverse effects [AEs], while 4 discontinuations were related to treatment. The most common drug-related ocular AEs were xanthopsia and delayed dark adaptation. Most cases of xanthopsia, delayed dark adaptation, and night vision impairment were mild and resolved during the trial. Headaches were the most commonly reported treatment-related nonocular AE. All events were being closely monitored through scheduled safety assessments and regular safety reporting throughout the study.

What role do imaging biomarkers play in assessing outcomes in this trial?

Nguyen: The Ocular Imaging Research and Reading Center (OIRRC, Sunnyvale, California, USA) serves as the centralized, global reading center for the DRAGON study, assessing and grading various images. DDAF, QDAF, and DAF are key imaging biomarkers in the DRAGON trial and are part of the study objectives to evaluate lesion growth. Specifically, DDAF represents areas of definite atrophy and serves as the primary end point, while QDAF identifies regions of potential early atrophy. DAF, the sum of DDAF and QDAF, provides a broader assessment of retinal changes and is evaluated as a secondary end point. Tinlarebant modulates the visual cycle and reduces the accumulation of toxic vitamin A byproducts in the retina. By targeting this mechanism, the drug helps to protect photoreceptors and slows the progression of atrophic lesions. Together, these biomarkers provide an objective, quantitative measure of lesion progression over time, reflecting treatment effect of tinlarebant.

Are there any adaptive design elements incorporated into the study?

Nguyen: The study design for the DRAGON trial included an adaptive sample size re-estimation that would determine the need for an increase in sample size in order to enhance power, based on a treatment effect observed at the interim analysis. According to the protocol, if interim analysis showed efficacy within the “promising efficacy zone,” additional subjects could be enrolled to enhance conditional power. No subjects would be added if efficacy was below or beyond this zone. The DSMB concluded that no study modifications were required, the trial continued with the planned sample size, and submission for regulatory review was recommended, indicating that efficacy exceeded the promising zone.

Quan Dong Nguyen, MD, MSc, FAAO, FARVO, FASRS
E: [email protected]
Nguyen is professor of ophthalmology at the Byers Eye Institute at Stanford, and professor of medicine and pediatrics at Stanford University School of Medicine in Palo Alto, California. He did not indicate any relevant financial disclosures.

REFERENCES

1. Nguyen QD. Tinlarebant for adolescent Stargardt disease: interim results of phase 3 DRAGON study. Presented at: FLORetina 2025–International Congress on OCT and OCT Angiography (ICOOR); December 4-7, 2025; Florence, Italy.

2. New hope for people living with a disease once deemed untreatable: Belite Bio announces positive topline results from the pivotal global, phase 3 DRAGON trial of tinlarebant in adolescents with Stargardt disease. News release. December 1, 2025. Accessed December 11, 2025. https://investors.belitebio.com/news-releases/news-release-details/new-hope-people-living-disease-once-deemed-untreatable-belite