Session Title: Quick Fire Free Paper 4
Session Date/Time: Sunday 14/09/2014 | 08:00-10:00
Paper Time: 08:05
Venue: Boulevard B
First Author: : G.Lambrou FRANCE
Co Author(s): : G. Liew P. Keane D. Sim C.A. Egan J. Warburton A. Tufail
Intravitreal injections of ranibizumab are an effective treatment for neovascular age-related macular degeneration (nAMD), with the majority of patients retaining baseline vision and a proportion experiencing improved vision. Baseline visual acuity (VA) is known to be an important predictor of treatment response, but how this influences changes in vision compared to sham injections or natural history is unknown. We therefore performed a post hoc subgroup analysis of data from the pivotal ANCHOR and MARINA trials to evaluate response to ranibizumab, stratified by presenting baseline visual acuity (bVA).
Pooled post-hoc subgroup analysis of 24-month visual outcomes in ranibizumab-treated and sham-treated patients from the randomized clinical trials ANCHOR and MARINA.
ANCHOR and MARINA were the pivotal trials that demonstrated clear superiority of monthly intravitreal ranibizumab over photodynamic therapy or sham injections in the treatment of classic and occult nAMD. These trials followed 423 and 716 participants respectively over 2 years. We conducted an individual patient-level pooled analysis of these trials and evaluated VA gains (VG) measured in ETDRS letters gained or lost, stratified by bVA in 0.3 logMAR (15 letter) interval groups. We also calculated visual benefits (VB) defined as the difference in VA between treated and control arms at each time point, and plotted these by bVA groups.
The overall mean VG and VB in the pooled analysis (all patients) were 8.2 and 23.3 letters respectively at 2 years. When stratified by baseline VA, mean (m) VGs and VBs were as follows: Group A (bVA 70-85 ltrs): mVG = -1.1 and mVB = 20.9; group B (bVA 55-70 ltrs): mVG = 4.8 and mVB = 23.8; group C (bVA 40-55 ltrs): mVG = 8.4 and mVB = 20.2
Visual gain under treatment – the change of VA from baseline – is strongly dependent on bVA: eyes with good to best bVA have little room for improvement and therefore show limited visual gains (ceiling effect) which may be misinterpreted as lack of response to treatment. On the other hand, visual benefit from treatment – the difference in VA outcomes between treated and untreated patients – was relatively consistent across baseline VAs, suggesting a similar magnitude of response in the three bVA groups. These results highlight the importance of treating patients while they still retain fairly good vision, and have implications on interpretation of visual gain as response to treatment as well as in future trial design