TRIAL SUCCESS

Astellas builds on IZERVAY's GATHER1 win with new driving and biomarker data

The original GATHER1 result behind IZERVAY's approval is now dated, but Astellas is layering pooled follow-up data on driving eligibility and photoreceptor biomarkers to extend the case at ASRS 2026.

By AppliedXLJuly 9, 2026Trial NCT02686658

Executive Summary

Astellas restated that avacincaptad pegol met the primary endpoint of its pivotal GATHER1 trial and is presenting new follow-on analyses at a major retina meeting, extending evidence for a drug already on the market.
The new material centers on driving-eligibility maintenance and a photoreceptor-health biomarker, giving the approved drug's benefit a functional, patient-relevant frame beyond the anatomical endpoint that supported approval.
IZERVAY sits in a sparse geographic atrophy field for its target and mechanism, with a single late-stage rival advancing the same complement pathway through a different modality.
Because the drug is already approved, the stake is less about proving efficacy and more about whether the accumulating biomarker and real-world data broaden how physicians and payers view its durability and functional benefit.

The disclosure

Astellas said in a July 9, 2026 release that IZERVAY met its primary endpoint in GATHER1 (NCT02686658), the change in geographic atrophy area measured by fundus autofluorescence at Month 12. The trial enrolled 286 patients across sites in the United States, Hungary, Israel, Czechia, Estonia, Croatia and Latvia, comparing avacincaptad pegol against a sham injection. The registered result, first met and reported publicly in November 2021, is not new: this release restates an outcome already reflected in the drug's approved label. IZERVAY is approved for GA in the United States, Australia and Macau, and conditionally in Japan. Astellas+1

What's actually new

The substance of the July 9 disclosure is forward-looking about presentation content, not new trial data: Astellas will give a podium presentation on the relationship between ellipsoid zone integrity, a biomarker for photoreceptor health, and low-luminance visual deficit, plus additional analyses on hyper-transmission progression and real-world treatment patterns in patients receiving concurrent anti-VEGF therapy. A pooled post hoc analysis of GATHER1 and GATHER2, presented earlier in 2026, found IZERVAY reduced the 24-month risk of losing driving eligibility to 12.6% versus 20.1% with sham, a 41% relative risk reduction with a nominal p-value of 0.0594. The trial's open-label extension also reported growth-rate reductions of 40.5% and 37% versus projected sham, with p<0.001, and no new safety signals. Astellas

Safety and design context

The trial's own posted safety data show serious adverse events occurred in 36 of 176 participants across pooled Zimura dose groups (any dose) versus 28 of 110 on sham, with three deaths on active treatment versus one on sham across the combined Part 1 and Part 2 populations. Astellas characterizes IZERVAY's current safety profile as consistent with its known label. The trial's enrollment target rose from 200 to 286 participants at the time enrollment completed, tracked as a registry change rather than a mid-study cut, and the trial reached Completed status cleanly after enrollment closed. NCT02686658+1

Where it sits competitively

Geographic atrophy targeting the C5 complement protein remains a narrow field: two active industry trials study C5 in this indication, and no distinct sponsor has recorded a failure in that specific pairing. Regeneron Pharmaceuticals' cemdisiran, also an RNA therapy against C5, is in Phase 3 testing in geographic atrophy (NCT06541704), making it the nearest direct comparator on both target and modality. Other C5-targeting assets, including Roche's crovalimab and Regeneron's pozelimab, are in development for paroxysmal nocturnal hemoglobinuria rather than geographic atrophy, so they inform the target's broader safety and mechanism record without competing directly in this indication. NCT02686658

This analysis was produced using AI-assisted reporting systems, AppliedXL data, and official public records. These systems undergo editorial review, quality checks, and regular audits by human experts. Errors may still occur, as with any automated system. Always consult the linked primary sources. Read our AI Editorial Policy.