Slowing Stargardt: What a 36% reduction in lesion growth could mean for adolescent patients

Stargardt disease type 1 (STGD1), the most common inherited macular dystrophy, has no approved treatment, but that may change soon. Belite Bio's phase 3 DRAGON trial (NCT05244304) randomly assigned 104 patients aged 12 to 20 years 2:1 to oral tinlarebant or placebo for 24 months. Tinlarebant is a retinol-binding protein 4 (RBP4) antagonist. In the trial it reduced the growth of definitely decreased autofluorescence (DDAF) lesions by about 36% vs placebo (P = .0033).¹

The US Food and Drug Administration (FDA) has since accepted the new drug application for priority review and set a target action date of February 12, 2027.² Visual acuity changed little in either group over 2 years. This is expected from natural history data, but it leaves open the question of how well an anatomical end point shows clinical benefit.

Michel Michaelides, BSc, MBBS, MD(Res), FRCOphth, FACS, of Moorfields Eye Hospital and the UCL Institute of Ophthalmology in London, England, presented the DRAGON topline results at the ESCRS 2026 Annual Meeting. In this interview, he explains what slower lesion growth means for adolescent patients and why functional end points are hard to measure in a 2-year trial. He also covers tinlarebant's safety profile, which follows from its mechanism, and how an oral systemic therapy might work alongside ABCA4 gene therapy. Finally, he discusses the gaps in genetic testing that still limit who can be diagnosed and treated.

Q&A with Michel Michaelides, BSc, MBBS, MD(Res), FRCOphth, FACS

Transcript edited lightly for clarity and length.

DRAGON hit its primary endpoint—roughly a 36% reduction in lesion growth on DDAF. In plain terms, what does that mean for a teenager with Stargardt?

Michaelides: Lesion size on DDAF represents tissue that is irreversibly lost, and these regions correspond to decreased vision. Reduction in lesion growth observed with tinlarebant is expected to preserve retinal tissue cumulatively over many years with disease. Whether this ultimately delays clinically important milestones of visual disability requires longer-term follow-up beyond this trial.

The imaging signal is strong, but does slowing lesion growth actually preserve vision? What do the functional data show so far?

Michaelides: Lesion size on DDAF represents tissue that is irreversibly lost, and these regions correspond to decreased vision. Reduction in lesion growth observed with tinlarebant is expected to preserve retinal tissue cumulatively over many years with disease. Differences in functional vision are difficult to measure over a 2-year trial. For instance, in the ProgStar natural history study, BCVA declined by only approximately 0.55 ETDRS letters per year over two years. There is also substantial test-retest variability in Stargardt. As such, BCVA and microperimetry were secondary endpoints for the DRAGON trial.

Tinlarebant works by cutting retinol delivery to the retina. How was that tolerated—any effect on night vision or dark adaptation?

Michaelides: Tinlarebant was generally well tolerated in DRAGON; most ocular and non-ocular treatment-emergent adverse events (TEAEs) were mild and the majority resolved while patients remained on study; there were no serious ocular adverse events.

• The most frequently reported ocular TEAEs were xanthopsia, delayed dark adaptation, and night vision impairment, which are consistent with tinlarebant’s mechanism of action. These TEAEs were generally mild, and most resolved while the patients continued the study.
• There were 2 severe and no serious ocular TEAEs.
• The 2 severe TEAEs were delayed dark adaptation and night vision impairment.
• In a total of 69 subjects treated with tinlarebant, 4 TEAEs resulted in study drug discontinuation.
• (delayed dark adaptation [1], xanthopsia [1], and night vision impairment [2])
• 2 of these TEAEs led to study discontinuation.

ABCA4 gene therapy has been hard because the gene is so large. How does a systemic oral drug sit alongside gene-therapy efforts—competitor or complement?

Michaelides: Complementary. While ABCA4 dysfunction is the cause for Stargardt disease and required for DRAGON trial inclusion, it does not target or modify the ABCA4 gene or protein. Therefore, this approach may possibly be complementary to potential treatments that target ABCA4 DNA or RNA, such as gene therapy or gene or RNA editing. Further studies are necessary to show how a systemic drug therapy and gene therapy can work alongside each other.

Regulators have been cautious about anatomical endpoints like atrophy growth. Is the field turning a corner on accepting them?

Michaelides: There is precedence for anatomical endpoints in clinical trials of geographic atrophy secondary to age-related macular degeneration. This was the primary efficacy endpoint for complement inhibitors, and this reduction in lesion growth formed the principal efficacy evidence supporting regulatory approval.³

RBP4 antagonism is also being explored in geographic atrophy. Could this mechanism reach beyond Stargardt?

Michaelides: Uncoupling of retinol delivery to the retina is expected to reduce accumulation of bisretinoids and formation of lipofuscin. The PHOENIX trial (NCT05949593) is underway to explore whether tinlarebant reduces lesion growth in geographic atrophy.

Genetic testing is the gateway to trials like this. Is diagnosis reaching young Stargardt patients early enough to act on?

Michaelides: In the US, it is estimated that 20,000 individuals are diagnosed with Stargardt disease following genetic testing. Based on disease and molecular prevalence estimates, there are likely 50,000-60,000 individuals living in the US with Stargardt disease. There are many people who remain undiagnosed or [who are] diagnosed later in disease. Expansion of genetic testing access is of critical importance for all trials and medications where genetic information is an inclusion criteria.⁴

The biggest misconception about treating inherited retinal disease today?

Michaelides: That we can only diagnose disease. There are many clinical trials and potential treatments on the horizon for patients. The hope is that there will be options for patients in the very near future.

REFERENCES:

1. Belite Bio releases topline results from phase 3 DRAGON trial of tinlarebant for STGD1. Ophthalmology Times. December 2025. https://www.ophthalmologytimes.com/view/belite-bio-releases-topline-results-from-phase-3-dragon-trial-of-tinlarebant-for-stgd1

2. Belite Bio announces US Food and Drug Administration acceptance and priority review of new drug application for tinlarebant for the treatment of Stargardt disease type 1. News release. Belite Bio; August 11, 2026. https://investors.belitebio.com/news-releases/news-release-details/belite-bio-announces-us-food-and-drug-administration-acceptance

3. Lad EM, et al. Annu Rev Vis Sci. 2024;10(1):455-476.

4. Mata NL, et al. Ophthalmic Res. 2025;68(1):555-572.