Vitreomacular Adhesion and the Risk of Neovascular Age-Related Macular Degeneration

To assess the prevalence of vitreomacular adhesion (VMA) in consecutive naïve eyes diagnosed with exudative age-related macular degeneration (AMD) in comparison with eyes with nonexudative AMD and age-matched controls, and to evaluate prospectively the incidence of vitreomacular interface changes over time and their influence on choroidal neovascularization (CNV) development.

Retrospective cross-sectional analysis and longitudinal cohort study conducted at Sacrocuore Hospital, Negrar, Verona, Italy.

A total of 1067 eyes examined at Sacrocuore Hospital between August 2008 and June 2015 met the inclusion criteria and were evaluated in this study.

Eyes were classified into 3 groups: 403 eyes of 364 patients (mean [standard deviation; SD] age 77.8 [8.0] years) affected by exudative AMD; 350 eyes of 298 subjects (mean [SD] age 78.1 [8.2] years) with nonexudative AMD; and 314 eyes of 214 subjects (mean [SD] age 74.2 [8.2] years) with no signs of AMD enrolled as the control group. The vitreomacular interface status was evaluated by spectral-domain optical coherence tomography (OCT) and was graded according to the OCT-based International Classification System developed by the International Vitreomacular Traction Study Group by 2 independent masked observers.

VMA was present in 101 (25.1%) eyes with exudative AMD, 84 (24.0%) eyes with nonexudative AMD, and 84 (26.8%) eyes with no signs of AMD (no statistical difference was found; P = 0.3384). Spontaneous release of VMA (RVMA) was found in 15 (15.3%) eyes with exudative AMD, 21 (28.0%) eyes with nonexudative AMD, and 10 (24.4%) eyes with no signs of AMD over a mean follow-up of 25.5, 25.9, and 24.1 months, respectively. The incidence of RVMA in exudative AMD eyes was significantly lower compared with nonexudative (P = 0.0207) and lower, but not statistically significant, with respect to eyes with no signs of AMD (P = 0.1013). In eyes with nonexudative AMD, de novo development of CNV occurred in 91 eyes (30.6%). There was no significant difference regarding the rate of CNV development in the presence or absence of VMA (P = 0.0966).

The present study found no significant difference in the prevalence of VMA in eyes affected by AMD compared with age-matched controls and no difference in the rate of de novo CNV development in eyes with or without VMA. Conversely, a lower incidence of RVMA over time was found in eyes affected by exudative AMD. The results of this study suggest that VMA might be a consequence rather than a causative factor in the development of CNV.