Subgroup analysis of the MARINA study of ranibizumab in neovascular age-related macular degeneration

Objective: An examination of clinically relevant subgroups of patients in the MARINA study of ranibizumab in treatment of minimally classic or occult with no classic choroidal neovascularization (CNV) secondary to age-related macular degeneration (AMD) was done. Objectives were to determine the effectiveness of ranibizumab across subgroups, compare the effectiveness of ranibizumab with that of sham injection within subgroups, and evaluate the relationship between selected baseline characteristics and visual acuity (VA) outcomes. Design: Retrospective subgroup analyses of 24-month data from the MARINA study. Participants and Controls: Seven hundred sixteen patients were randomly assigned to 0.3 mg ranibizumab (n = 238), 0.5 mg ranibizumab (n = 240), or sham treatment (n = 238). Methods: Efficacy outcomes were compared across subgroups based on patients' gender, age, baseline VA score, baseline CNV lesion size, CNV lesion type, and duration of neovascular AMD using univariate analyses. Multivariate analyses were performed on the change from baseline to 24 months in VA score-to assess further the correlation between baseline characteristics and VA outcome. Main Outcome Measures: Proportion of patients losing < 15 letters from baseline, proportion gaining >= 15 letters from baseline, and mean VA score change from baseline. Results: For each of the 3 VA end points, all subgroups of ranibizumab-treated patients did better on average than the sham-treated patients. Increasing age, larger CNV lesion size at baseline, and a higher baseline VA score were all associated with greater loss of letters in the sham group or less gain of letters in the ranibizumab groups. However, the net benefit of ranibizumab versus sham treatment was greater in patients who scored higher than in those who scored lower in baseline VA. Conclusions: This subgroup analysis of 24-month data from the MARINA study indicates that ranibizumab treatment was associated with an average increase from baseline VA in all subgroups evaluated, and that ranibizumab treatment was superior to sham treatment across all subgroups. The most important predictors of VA outcomes were, in decreasing order of importance, baseline VA score, CNV lesion size, and age.