Purpose
To assess the effect of anti-VEGF treatment on visual acuity outcome in patients with neovascular AMD presenting with very low vision.
Methods
Retrospective analysis of electronic care record entries between March 2010 and June 2013 assessing a final dataset comprised of 420 eyes diagnosed and treated with an intravitreal anti-VEGF for neovascular AMD. We included all eyes (n = 329) with BCVA ≤35 letters and sampled approximately 10 % of eyes within VA range groups 36 to 85 letters. The selected sample (N = 494) was then scrutinized and we subsequently excluded 44 other non-AMD exudative maculopathies,11 cases of nAMD with no follow up and second eyes in 19 cases on bilateral treatment.<br /> The database was populated with demographic parameters, BCVA in the treated eye at baseline and follow up, baseline BCVA in fellow eye, the number of intravitreal treatments, duration of follow up. The presence or absence of haemorrhage, RPE tear, geographic atrophy (GA) was noted and a graded categorical scale was used to record extent of fibrosis within the lesion. We scrutinised the accompanying database of colour, OCT, autofluorescence, FFA and ICG images to measure and classify the type of lesions in the selected eyes. SPSS 21 was used for statistical analysis.
Results
Two thirds of eyes in our selected database with initial BCVA < 23 ETDRS letters improved to ≥ 23 on year 1 BCVA. Considering the entire group, the change in BCVA from baseline to best BCVA achieved during year 1 was highly statistically significant with a mean improvement of 9.95 letters, 95% CI= 8.81, 11.0 p = 0.000 (paired samples t test).<br /> Patients with worst presenting VA were more likely to be older, receive fewer treatments, have larger lesion areas and shorter follow up. Presence of RPE rips resulted in statistically significantly worse baseline BCVA. Regression analysis identified GA and fibrosis as predictors of best and average BCVA with the model having an R2 of 0.7; Presence of moderate to severe fibrosis and/or atrophy at baseline reduced the likelihood of improving VA during follow up.
Conclusions
Our study supports the use of anti VEGF agents in eyes with low VA particularly when fibrosis and atrophy are absent and suggests algorithms to predict outcome for combinations of visual acuity and lesion characteristics across the full visual acuity range.