Background  

The aim of this meta-analysis was to summarize and to discuss the results of the four main submacular surgical procedures for age-related macular degeneration (AMD) as reported in the literature through 2004 and to compare them to the Submacular Surgery Trials (SST) data.

Methods  

The existing data in the literature on submacular surgery for AMD from 1992 to 2004 were evaluated. The main outcomes were proportion of patients with two or more lines of improvement in visual acuity (VA) and proportion with two or more lines of deterioration in VA after surgery.

Results  

Eighty-eight studies including 1,915 cases met the inclusion criteria. Estimates for the treatment outcome within the four groups of treatment based on a logistic regression model gave comparable results for removal of choroidal neovascularization (CNV) (improvement of VA 28%, deterioration of VA 25%), macular translocation (improvement of VA 31%, deteriora-tion of VA 27%), and for transplantation of pigment epithelium (improvement of VA 22%, deterioration of VA 21%). Estimates for removal of subretinal hemorrhage were significantly different (improvement of VA 62%, deterioration of VA 13%).

Conclusions  

Selected case series showed superior results of VA compared to the SST. The question of whether this is due to selection bias that seems inevitable when dealing with medium-sized nonrandomized case series or due to better results in single centers cannot be answered. In our opinion there still seem to be indications for submacular surgery such as in patients with AMD with low preoperative VA due to large hemorrhagic or fibrotic membranes or nonresponders to photodynamic therapy (PDT).