Purpose: To describe the natural history and progression of visual loss in eyes with untreated neovascular age-related macular degeneration (AMD).

Design: Systematic review and meta-analysis.

Participants: Four thousand three hundred sixty-two untreated neovascular AMD patients from published interventional studies.

Methods: A systematic review of the literature from 1980 to August 2005 was performed. Studies reporting disease progression outcomes for untreated patients with neovascular AMD were included. Outcome measures were summarized using simple counts and means. Random effects meta-analyses were conducted and tests of heterogeneity were performed where appropriate.

Main outcome measures: Changes in visual acuity (VA) loss, development of comorbidities, and fellow eye involvement.

Results: Fifty-three primary studies were included. Nearly half of the studies (28) were randomized clinical trials. The quality of the studies was high, with over 80% providing level I or II evidence. Mean baseline VA among study patients was 0.64 logarithm of the minimum angle of resolution (logMAR) (approximately 20/87 Snellen). The mean VA change in logMAR progressed from 0.1 (1 line lost) at 3 months to 0.3 (2.7 lines lost) after 12 months and 0.4 (4 lines lost) after 24 months. The proportion of patients who developed severe vision loss (>6 lines) from baseline increased from 21.3% at 6 months to 41.9% by 3 years. The proportion of patients with VA worse than logMAR 1.0 (20/200 Snellen) increased from 19.7% at baseline to 75.7% by 3 years. Neovascular AMD developed in the fellow eye in 12.2% of patients by 12 months and in 26.8% by 4 years. Meta-analyses of vision outcome by subtype of neovascular AMD were not possible.

Conclusions: A doubling of the visual angle of presenting VA may be expected to occur in the year after initial presentation in eyes with untreated neovascular AMD. No conclusions can be drawn as to the differences in rates of disease progression by neovascular AMD subtype. The diversity of reporting formats, paucity of long-term natural history data, and heterogeneity among the reported clinical studies impose limits to the clear understanding of long-term prognosis for visual function in neovascular AMD.