Purpose
To identify characteristics associated with measures of suboptimal response to anti-VEGF therapy in patients with wAMD.
Methods
Post-hoc exploratory analyses of integrated data from the VIEW 1 and 2 trials of intravitreal aflibercept or ranibizumab for wAMD were performed. Response variables included change in visual acuity (VA) and fluid status at weeks 52 (W52) and 96 (W96); descriptive variables included demographic and disease characteristics at baseline (BL) and during treatment. Stepwise logistic regression was used for categorical target variables (e.g., ≥3 line loss, fluid status wet), generalized linear modeling was used for continuous target variables (e.g., VA change from BL to W52).
Results
Analyses were performed on data from 2412 patients (ITT population) using 28 response and 19 descriptive variables. When suboptimal response was defined as a ≥3 line loss from BL to W52, associated descriptive variables included: BL lesion size (p<0.001), lesion size change from BL to W52 (p<0.001], age (p<0.01), and VA (ETDRS letters) at BL (p<0.05). Similarly, when defined as a ≥3 line loss from BL to W96, associated variables included: BL lesion size and lesion size change from BL to W96 (both p<0.001), age and VA at BL (both p<0.01). For suboptimal response defined as VA worse than 20/200 at W96, associated variables included: age, VA at BL, BL lesion size, lesion size change from BL to W96 (all p<0.001), as well as W52 fluid status (dry vs wet, p<0.05) and CRT change from BL to W52 (p<0.05). Additional statistical modeling showed that, along with other variables, W52 fluid status was associated with VA change from BL to W52 (estimated LS mean letter gain: 6.9 [95% CI: 5.2-8.6] vs 9.2 [7.6-10.8] for wet vs dry, p<0.001). When used as a suboptimal response variable, W52 fluid status (wet) was one with which treatment was associated (p<0.05), as well as W12 fluid status, lesion size change from BL to W52, CRT at BL and change from BL to W52 (all p<0.001), and BL lesion size (p<0.01).
Conclusions
Variables consistently shown as significantly associated with suboptimal VA response in patients with wAMD included age, VA at BL, lesion size at BL, and lesion size change from BL to endpoint. Variables significantly associated with suboptimal fluid response at W52 included W12 fluid status, treatment, BL and change from BL in lesion size, and CRT.
Keywords: 412 age-related macular degeneration •
453 choroid: neovascularization •
748 vascular endothelial growth factor