First Author: S.M.Waldstein AUSTRIA
Co Author(s): J. Wright J. Warburton P. Margaron C. Simader U. Schmidt-Erfurth 0 0 0 0 0 0 0 0 0
Back to previous
To establish predictive factors for 12-month visual outcomes of different injection frequencies of ranibizumab for neovascular age-related macular degeneration (AMD).
Post-hoc analysis of prospective, multicenter clinical trial data.
Data of 353 patients randomized in the EXCITE trial, receiving fixed monthly 0.3 mg (n=115) or quarterly (n=238) 0.3 or 0.5 mg ranibizumab treatment following three initial monthly loading doses, were analyzed. Best-corrected visual acuity (BCVA) was measured using ETDRS charts at each monthly visit. Standardized grading of retinal morphology including intra-retinal cysts (IRC), sub-retinal fluid (SRF), pigment-epithelial detachment (PED), classification of the vitreo-macular interface (VMI) configuration, including vitreomacular adhesion (VMA) and posterior vitreous detachment (PVD) was performed and central retinal thickness (CRT) was measured monthly using optical coherence tomography (OCT) scans at the Vienna Reading Center. Analysis of covariance was used to assess the impact of OCT-based retinal morphology on change in BCVA at month 12. Potential predictor variables included treatment frequency, VMI and the baseline values of BCVA and CRT. Separate models added the baseline and month 3 values of PED, SRF and IRC as well as lesion type determined by fluorescein angiography. Data were available for 319 patients (monthly: n= 102; quarterly: n= 217) for baseline and 277 patients (monthly: n= 92; quarterly: n= 185) for the month 3 time point.
The impact of baseline parameters on functional outcome was first assessed: Overall, adjusted BCVA change at month 12 was +7.12 letters for monthly and +3.08 for quarterly treatment. Significant (p<0.05) predictive factors for BCVA change at month 12 were treatment frequency (p=0.01), VMI (p<0.01), baseline SRF (p=0.05) and baseline BCVA (p<0.01). The unadjusted mean differences in BCVA gains at month 12 between monthly and quarterly treatments were +0.92 for patients with SRF versus +12.31 letters without SRF, whereas they were similar whether IRC and PED were present or not. VMA was associated with a higher BCVA gain in monthly compared to quarterly treatment (mean estimate of gain difference: +10.9, p=0.03), whereas PVD was associated with similar BCVA gains without significant difference at the 5% level. Using the month 3 parameters, significant predictive factors for BCVA change at month 12 were: PED (p<0.01), lesion type (p=0.04), treatment frequency (p<0.01), VMI (p<0.01) and baseline BCVA (p<0.01). The unadjusted mean differences in BCVA gains from month 3 to month 12 between monthly and quarterly treatments were +2.36 in patients with PED at month 3 versus +4.80 letters without PED at month 3.
Characterization of patients by OCT allows the prediction of functional outcomes in ranibizumab therapy of neovascular AMD. In patients with SRF at baseline or with PVD, monthly and quarterly treatment performed similarly. In contrast, in patients without SRF at baseline or with VMA, monthly treatment was substantially superior to quarterly treatment. Our results may facilitate a reduction of treatment frequency without jeopardizing functional outcomes.