- Design
- Prospective multicentre cohort study with life-table analysis
- Population
- 81 treatment-naive eyes with polypoidal choroidal vasculopathy at four Thai university hospitals; mean age 64 years
- Primary outcome
- Fluid-free interval between first regression and first exudative recurrence on OCT
- Effect
- 13 of 26 eyes (50%, 95% CI 30-70%) recurred within 2 years; fluid-free probability 57% (37-77%) at week 24 and 53% (33-73%) at week 84; median acuity change on recurrence -3.5 letters
Complete polypoidal regression is the stated goal of treating polypoidal choroidal vasculopathy, and what to do once it is achieved has never been settled — keep injecting on a fixed schedule, extend, or stop and watch.
This prospective cohort at four Thai university hospitals took 81 treatment-naive eyes and gave three monthly aflibercept 2.0 mg injections. Eyes with regression at week 12 moved to an as-needed regimen with monthly follow-up in year one and two-monthly in year two; eyes without regression got three more injections and, if regressed at week 24, joined the same pathway. Thirty-two eyes (40%) regressed.
Of the 26 followed on the as-needed pathway, 13 (50%, 95% CI 30-70%) had an exudative recurrence within two years. Fluid-free probability was 57% (95% CI 37-77%) at week 24 and 53% (33-73%) at week 84 — so most recurrences happened early and the curve then flattened. Among 14 eyes that recurred, median acuity change from the previous visit was -3.5 letters (IQR -5.0 to -0.25); three (21%) lost ten letters or more and five (36%) had new haemorrhage.
Post hoc, baseline massive subretinal haemorrhage predicted lower recurrence (HR 0.18, 95% CI 0.04-0.80, P=.01) while leakage without subretinal haemorrhage suggested higher (HR 2.65, 95% CI 0.89-7.94, P=.07). That is counterintuitive and exploratory, and should be read as a question rather than an answer.
- Deferral after regression is defensible, but the patient must be monitored monthly in the first year
- Most recurrences came early — the flat part of the curve is after week 24
- A third of recurrences brought new haemorrhage; that is the reason monitoring cannot be loose
- Only 40% achieved regression on three or six injections in the first place
- The baseline haemorrhage signal is post hoc — do not use it to select patients yet
Why it matters
It puts a number on the risk a patient accepts when treatment is deferred after regression.
Don't overread it
The OCT biomarker findings are post hoc and exploratory in a cohort of 26 eyes.
The statistics, in plain English
Twenty-six eyes is a small denominator, which is why the confidence interval around that 50% recurrence rate runs from 30% to 70% — the true rate could be either. The hazard ratio of 0.18 for baseline massive subretinal haemorrhage comes from a post hoc analysis of a subgroup of a small cohort, and the competing finding (HR 2.65) does not reach significance at all. Treat both as observations that need testing.
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.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for ophthalmology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free