M García-Fiñana1, S Murjaneh, S Mahmood, S P Harding. 1. Centre for Medical Statistics and Health Evaluation, Faculty of Health and Life Sciences, University of Liverpool, and St Paul's Eye Unit, Royal Liverpool University Hospital, Liverpool, UK.
Abstract
PURPOSE: To analyse the influence of baseline clinical characteristics on the outcome of verteporfin photodynamic therapy (VPDT) for neovascular age-related macular degeneration (nAMD). METHODS: A total of 1008 patients commencing VPDT for nAMD in a single UK centre entered a prospective observational study between 1999 and 2006 and were followed for 2 years. Longitudinal linear mixed-effects modelling was applied to assess the influence of baseline covariates, such as best corrected visual acuity (BCVA), contrast sensitivity (CS), age, lesion size, and lesion type, on changes of BCVA over time in patients after VPDT. A logistic regression analysis was used to analyse clinical features significantly associated with treatment failure. RESULTS: Study eye BCVA was significantly better on average throughout the course of treatment in patients with better baseline BCVA and CS in the study eye (P<0.001 and P<0.01, respectively) and lower age (P=0.01). Mean BCVA showed a significant reduction over time with a significant quadratic relationship between 0 and 6 months and with stabilisation between 6 and 9 months. Patients with better BCVA and worse CS at baseline, and those in whom BCVA dropped during the first 3 months of follow-up, were more likely to lose >/=15 letters after 12 months. CONCLUSIONS: Findings from our large longitudinal data set provide estimates of likely outcome based on baseline features and response at 3 months in patients commencing a course of VPDT for nAMD. Statistical modelling built up for this large data set can be applicable to other studies in ophthalmology research.
PURPOSE: To analyse the influence of baseline clinical characteristics on the outcome of verteporfin photodynamic therapy (VPDT) for neovascular age-related macular degeneration (nAMD). METHODS: A total of 1008 patients commencing VPDT for nAMD in a single UK centre entered a prospective observational study between 1999 and 2006 and were followed for 2 years. Longitudinal linear mixed-effects modelling was applied to assess the influence of baseline covariates, such as best corrected visual acuity (BCVA), contrast sensitivity (CS), age, lesion size, and lesion type, on changes of BCVA over time in patients after VPDT. A logistic regression analysis was used to analyse clinical features significantly associated with treatment failure. RESULTS: Study eye BCVA was significantly better on average throughout the course of treatment in patients with better baseline BCVA and CS in the study eye (P<0.001 and P<0.01, respectively) and lower age (P=0.01). Mean BCVA showed a significant reduction over time with a significant quadratic relationship between 0 and 6 months and with stabilisation between 6 and 9 months. Patients with better BCVA and worse CS at baseline, and those in whom BCVA dropped during the first 3 months of follow-up, were more likely to lose >/=15 letters after 12 months. CONCLUSIONS: Findings from our large longitudinal data set provide estimates of likely outcome based on baseline features and response at 3 months in patients commencing a course of VPDT for nAMD. Statistical modelling built up for this large data set can be applicable to other studies in ophthalmology research.