Delaying pegcetacoplan by two years costs about two-thirds of the tissue saved

Design

analysis of a phase 3 open-label extension (GALE) of two randomised phase 3 trials (OAKS and DERBY), with crossover of sham-observed eyes at 24 months

Population

eyes with non-subfoveal or subfoveal geographic atrophy secondary to age-related macular degeneration, up to 48 months of continuous treatment

Primary outcome

geographic atrophy area growth rate, retinal tissue preserved, risk of progression to absolute scotoma, and safety

Effect

growth rate reduced by up to 24% vs projected sham (1.88 mm2 preserved); non-subfoveal early treatment preserved up to 3.16 mm2 vs 1.11 mm2 delayed; scotoma progression risk down 32% and 43% at central 4 and 16 loci

The GALE open-label extension carried patients from the phase 3 OAKS and DERBY trials forward to as much as 48 months of continuous pegcetacoplan, while sham-observed eyes crossed over to active treatment at 24 months. That design creates a natural early-versus-delayed comparison in geographic atrophy secondary to age-related macular degeneration.

Across the overall population, 48 months of treatment reduced the atrophy growth rate by up to 24% against projected sham, corresponding to 1.88 mm2 of retinal tissue preserved. The early-versus-delayed contrast is sharper in non-subfoveal disease: up to 3.16 mm2 preserved with monthly treatment from the start, against 1.11 mm2 in eyes whose treatment began two years later - roughly three times as much. Risk of progression to absolute scotoma fell by 32% at the central 4 loci and 43% at the central 16. Safety over 48 months was consistent with the parent trials.

The finding is about timing rather than about whether the drug works, and that is where the difficulty lies. Geographic atrophy is slow, the benefit is anatomical, and a patient asked to attend monthly for injections wants to know what they get. What this provides is a way to answer honestly: the tissue not lost is greater the earlier treatment starts, and the scotoma data give the first functional handle on what that means. It remains a treatment that slows loss rather than restoring vision, and the comparator for the growth rate is a projected rather than a concurrent sham.

• Frame the discussion around timing: the tissue preserved is roughly three times greater when treatment starts two years earlier in non-subfoveal disease.
• Use the scotoma endpoints - 32% and 43% reduction at the central 4 and 16 loci - as the closest thing here to a functional outcome.
• Be explicit that this slows progression and does not recover vision.
• The comparator beyond 24 months is a projected sham, not a concurrent one, which weakens the growth-rate estimate.
• Weigh injection burden honestly against an anatomical benefit; the monthly regimen produced the largest tissue preservation.

The statistics, in plain English

The phrase 'up to 24%' signals that this is the largest of several dosing and subgroup estimates rather than a single effect, so it should be read as the ceiling. More importantly, once the sham arm crossed over at 24 months there was no untreated comparison group left, and the 48-month growth rate is compared against a projected sham - a modelled counterfactual, not observed data. Open-label extensions also retain the patients who tolerated and responded to treatment, which biases long-term estimates favourably. The early-versus-delayed comparison is the more robust part, because both groups were randomised at the outset and differ only in when treatment started.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.