Abstract
Purpose:
To assess the life time treatment outcomes of anti-VEGF treatment for neovascular age-related macular degeneration (nv AMD) in routine medical practice by modeling over ten years as this is the average remaining life expectancy of nv AMD patients starting treatment.
Methods:
200 patients undergoing long-term anti-VEGF treatment for nv AMD were included. As half of patients were treated in both eyes, visual acuity (VA) in both eyes was categorized in different vision states to reflect changes in visual acuity in each eye over time. Each vision state encompasses a combination of VA in both eyes, i.e. 1st eye no visual impairment (>6/12), 2nd eye moderate impairment (6/12 - >6/18) or severe visual impairment (<6/18), etc., with six possible combinations). Multi-stage modeling (MSM) was used to extrapolate our real life data over 10 years (software: msm package for R). Bootstrapping was used to adjust for age and gender, and to generate confidence intervals for obtained transition probabilities between vision states.
Results:
For our real life cohort, average treatment duration was 37 (±13) months and mean number of injections was 21(±11). On average, VA in the treated eye improved from baseline to last follow-up (+7 letters; 49 to 56 ltrs, p<0.001) and remained stable in the better eye (66 to 63 ltrs, p=0.001). Forty percent of patients were treated in both eyes. Based on the MSM, persons with good vision in both eyes (VA>20/40) at baseline had a 16% chance at 5 years and an 11% chance at 10 years to remain at that level of vision. On the contrary, they had a 29% chance of losing vision in one and a 21% chance of losing vision in both eyes at 5 years and a 20% and 14% chance at 10 years, respectively. Irrespective of baseline vision, patients had a 30-35% chance at 5 years and a 50-55% chance at 10 years of treatment cessation.
Conclusions:
Based on our model results, most patients will cease treatment, and - beforehand - experience a deterioration of vision in both eyes despite treatment over ten years. However, a proportion of patients will remain with good vision (VA >20/40) in both eyes, and a number of patients whose vision loss will be either limited to one eye only or limited as to the extent of vision loss, will retain vision above 20/60. Further modeling will provide outcomes for a hypothetical treatment cohort.
Keywords: 412 age-related macular degeneration •
459 clinical (human) or epidemiologic studies: biostatistics/epidemiology methodology •
473 computational modeling