Fight Retinal Blindness SPAIN. Report 3: clinical outcomes of vascular endothelial growth factor inhibitors in low vision eyes with neovascular age-related macular degeneration. A national database study

Abstract

Objectives: To compare visual outcomes for low vision eyes (LVE) (<35 letters LogMAR or <20/200 Snellen) versus non-low vision eyes (NLVE) (>35 letters LogMAR or >20/200 Snellen) at the time of the first injection in a clinical practice setting.

Methods: Subgroup analysis of a multicenter national database of treatment- naïve eyes neovascular age related macular degeneration (nAMD) treated with anti-VEGF intravitreal injections divided into LVE and NLVE. Demographics, visual acuity (VA) at baseline and subsequent timepoints (12, 24, and 36 months), number of injections and visits data were collected using a validated web-based tool (Fight Retinal Blindness!).

Results: 3138 eyes were included, 705 LVE and 2433 NLVE. The LVE group had the greatest VA gain (p < 0.001), at 12, 24, and 36 months (+15, +15, and +13 letters respectively). The proportion of patients with VA loss (-5 letters) differed between groups at 12, 24, and 36 and was significantly greater (p < 0.001) in NLVE. The proportion of patients with VA gain (+5 letters) was significantly higher (p < 0.001) in LVE in all timeframes. The proportions of LVE that still had VA ≤ 35 letters at 12, 24, and 36 months were 54%, 54%, and 57%. Conversely, 8%, 9%, and 8% of LVE achieved VA ≥ 70 letters at 12, 24, and 36 months. LVE received fewer intravitreal injections than NLVE throughout follow-up (6, 9, 12 vs 7, 11, 15).

Conclusion: Findings of this study support the need for ongoing therapy in patients with initial visual acuity less than 35 letters since sustained visual improvements can be achieved and maintained for the first 3 years of treatment.

Fig. 1. Visual outcomes at 12 months.

Treatment-only visits were excluded from these models. Mean visual acuity (VA, top left) shown in locally weighted scatterplot smoothing (LOESS) regression curve. Distribution of eyes by VA levels and subgroup analysis (top right) shown in a hybrid boxplot. The whiskers represent the 25th and 75th quartiles plus or minus the interquartile range. Predicted VA (bottom left) and VA change (bottom right) displayed in longitudinal generalized additive mixed models. These models included longitudinal data from completers and non-completers. The predicted mean (95% CI) VA change at 12 months when data from completers and non-completers were included was +2.9 (2.4, 3.3) letters for Overall, +12.5 (11.1, 13.9) letters for Baseline 0–35 Letters and +0.3 (−0.2, 0.8) letters for Baseline 36–100 Letters. Significant predicted change comparison at 12 months: Baseline 0–35 Letters vs Baseline 36–100 Letters (p < 0.001).

Fig. 2. Visual outcomes at 24 months.

Treatment-only visits were excluded from these models. Mean visual acuity (VA, top left) shown in locally weighted scatterplot smoothing (LOESS) regression curve. Distribution of eyes by VA levels and subgroup analysis (top right) shown in a hybrid boxplot. The whiskers represent the 25th and 75th quartiles plus or minus the interquartile range. Predicted VA (bottom left) and VA change (bottom right) displayed in longitudinal generalized additive mixed models. These models included longitudinal data from completers and non-completers. The predicted mean (95% CI) VA change at 24 months when data from completers and non-completers were included was +0.3 (−0.4, 1.0) letters for Overall, +9.4 (7.4, 11.4) letters for Baseline 0−35 Letters and -2.5 (−3.3, −1.8) letters for Baseline 36–100 Letters. Significant predicted change comparison at 24 months: Baseline 0–35 Letters vs Baseline 36–100 Letters (p < 0.001).

Fig. 3. Visual outcomes at 36 months.

Treatment-only visits were excluded from these models. Mean visual acuity (VA, top left) shown in locally weighted scatterplot smoothing (LOESS) regression curve. Distribution of eyes by VA levels and subgroup analysis (top right) shown in a hybrid boxplot. The whiskers represent the 25th and 75th quartiles plus or minus the interquartile range. Predicted VA (bottom left) and VA change (bottom right) displayed in longitudinal generalized additive mixed models. These models included longitudinal data from completers and non-completers. The predicted mean (95% CI) change in visual acuity at 36 months when data from completers and non-completers were included was −3.2 (−4.2, −2.2) letters for Overall, +3.9 (1.1, 6.8) letters for Baseline 0–35 Letters and −6.0 (−7.0, −4.9) letters for Baseline 36–100 Letters. Significant predicted change comparison at 36 months: Baseline 0–35 Letters vs Baseline 36–100 Letters (p < 0.001).