Abstract
Purpose :
Retinal fluid status preceding therapy is of significance for visual acuity in neovascular age-related macular degeneration (nAMD). However, the implications of persistent fluid after initiation of treatment with intravitreal anti-VEGF therapy are unclear. This analysis investigates prevalence and functional impact of fully automatically quantified persistent fluid after the loading dose in 2-year data of the HARBOR clinical trial.
Methods :
Monthly volume scan SD-OCT data of n=904 eyes of 904 patients with nAMD, enrolled in the HARBOR trial were included into the analysis. Intraretinal fluid (IRF) and subretinal fluid (SRF) were quantified using validated fully automated per-voxel segmentations based on deep learning. The study population was grouped according to persistent IRF and SRF in the fovea and parafovea at month 3, following the loading dose. A longitudinal mixed effects model was applied to compare the mean BCVA trajectories between the two groups. Two-sided Mann-Whitney-U test was used to compare BCVA change by month 24.
Results :
Frequency distributions of fluid at baseline and month 3 are displayed in Table 1. Figure 1 shows BCVA development depending on persistent fluid status. Patients with persistent foveal and parafoveal IRF had reduced baseline BCVA by an average of 5.6 (p<0.01) and 2.5 (p=0.025) letters and showed a significant lower BCVA change after the first injection by 2.3 (p=0.035) and 1.9 (p=0.026) letters, respectively. In fully quantitative terms, BCVA change after the first injection was lower by 1.2 letters per every 10nl foveal and lower by 0.6 letters per every 10nl parafoveal persistent IRF (both p<0.01). Patients with persistent foveal and parafoveal SRF, had lower BCVA change after the first injection by 2.6 (p=0.028) and 1.9 (p=0.026) letters, respectively. After the first injection, no significant changes in BCVA development occurred between groups. Overall change in BCVA by year 2 was significantly lower in patients with persistent vs. no persistent parafoveal IRF (6.7 vs. 9.3, p=0.047).
Conclusions :
Persistent SRF following the loading dose correlated with less initial BCVA gain. Furthermore, patients with persistent IRF might constitute a defined subgroup, as they present with lower baseline BCVA and less BCVA gain following the first injection. Optimal treatment regimens for patients presenting with persistent fluid remain to be established.
This is a 2020 ARVO Annual Meeting abstract.