Incidence, Characteristics, and Outcomes of Macular Neovascularization in Extensive Macular Atrophy with Pseudodrusen-like Appearance

Abstract

Purpose: To report the incidence, features, and clinical outcomes of macular neovascularization (MNV) in a large Italian cohort of patients with extensive macular atrophy with pseudodrusen-like appearance (EMAP).

Methods: Retrospective, longitudinal study including 79 EMAP patients (158 eyes) with ≥6 months of follow-up at three retina clinics. Medical records and imaging were reviewed for demographic and clinical data, including age, best-corrected visual acuity (BCVA), MNV features, and retinal pigment epithelium (RPE) atrophy size, measured by short-wavelength autofluorescence and refined with near-infrared and OCT imaging. Main outcomes included cumulative MNV incidence, MNV risk factors, and BCVA and RPE atrophy changes in eyes with and without MNV.

Results: Over a mean follow-up of 40.4 months, MNV developed in 14 eyes (10 patients), with a 4-year cumulative incidence of 15.2%. Most MNVs were type 2 (86%) and subfoveal (64%). Cox regression identified younger age, fellow eye involvement, smaller RPE atrophy size, and greater central subfield thickness (all p<0.01) as significant risk factors for MNV. While eyes with MNV had lower baseline BCVA (58.4 vs. 71.4 letters, approximately 20/63 vs. 20/40 Snellen; p=0.005), BCVA decline over time was similar between the two groups (-3.9 vs. -4.1 letters/year, p=0.69). However, RPE atrophy progressed faster in MNV eyes (3.4 vs. 2.8 mm 2 /year, p=0.02).

Conclusions: In this EMAP cohort, MNV had a cumulative incidence of 15.2% at 4 years. Although BCVA outcomes were comparable, MNV was associated with faster atrophy progression, potentially due to a more aggressive disease phenotype or fibro-atrophic changes.

Keywords: CNV; EMAP; MNV; extensive macular atrophy with pseudodrusen-like appearance; incidence; macular neovascularization; outcomes; risk factors.

Fig. 1.

Multimodal imaging of Type 2 MNV in a patient with EMAP. A. A true-color fundus photograph shows large areas of RPE atrophy with a predominant vertical axis, surrounded by several pseudodrusen-like lesions. A loss of the foveal reflex is visible just inferior to the RPE atrophy. B. SW-AF reveals moderately hypoautofluorescent areas corresponding to RPE atrophy, with a pronounced hypoautofluorescent signal superonasal to the fovea. C. Early-phase fluorescein angiography displays a hyperfluorescent neovascular membrane, with pronounced leakage in the late phase (E). Indocyanine green angiography (ICGA) highlights the neovascular network (D), with characteristic late-phase staining (F). Swept-source optical coherence tomography angiography, superimposed on the near-infrared image, provides further visualization of the neovascular network (G). The optical coherence tomography scan through the fovea shows significant neuroretinal thickening because of intraretinal cysts, subretinal hyperreflective material, subretinal neovascularization (H).

Fig. 2.

RPE atrophy size, lesion activity, and intravitreal injection frequency in eyes with EMAP complicated by MNV. A. A bimodal distribution of MNV development can be observed in relation to RPE atrophy size, with the first peak occurring between 6.1 mm² and 9.0 mm² and the second peak between 21.1 mm² and 24.0 mm². B. Mean lesion activity was highest in the first year (41.7%) and progressively declined over time, ranging from 7.5% to 16.8% in subsequent years. C. Similarly, the average number of intravitreal anti-VEGF injections decreased from 2.7 in the first year to 0.3 to 0.9 per year during years 2 to 4. Error bars in panels (B and C) represent the standard error of the mean.

Fig. 3.

Cumulative incidence and risk factors for MNV development in EMAP. A. The Kaplan–Meier curve demonstrates a steady increase in MNV cases over time, with a cumulative incidence of 6.1% (95% CI: 4.1%–8.0%) at 1 year, rising to 15.2% (95% CI: 11.6%–18.7%) by year 4 (red curve). Excluding left-censored data, cumulative incidence was estimated at 0.9% (95% CI: 0.0–2.7) at 1 year and 4.2% (95% CI: 0.0–8.9) at 4 years (blue curve). The 95% confidence intervals are shaded in gray. B. Forest plots depict HR and 95% confidence intervals from a multivariable mixed-effects Cox regression analysis, identifying significant risk factors for MNV development. Notably, greater central subfield thickness (CST; P < 0.001), smaller RPE atrophy size (P = 0.001), fellow eye involvement (P = 0.01), and younger age (P < 0.001) were all significantly associated with an increased risk of MNV. CI, confidence interval.

Fig. 4.

Estimated changes in RPE atrophy size and BCVA between eyes with and without MNV over time. A. After adjusting for age and baseline areas, eyes with MNV demonstrated a significantly higher rate of RPE atrophy progression compared with eyes without MNV (3.42 vs. 2.76 mm²/year, P = 0.02). B. After adjusting for age and baseline BCVA, no difference in the rate of BCVA decline was observed between eyes with and without MNV (−3.9 vs. −4.1 ETDRS letters/year, P = 0.69). C. This panel illustrates the predicted trend of BCVA changes around the onset of MNV, showing a marked decline in BCVA followed by visual improvement after anti-VEGF treatment.

Fig. 5.

Longitudinal measurement of RPE atrophy in an eye with EMAP, complicated by MNV, from 12 months before its onset (A–C) to the final follow-up (P–R). The first row shows near-infrared images used to improve the delineation of RPE atrophy. The second row illustrates the progression of atrophy over time, as captured by SW-AF. The third row depicts the measured RPE atrophy area using the expert modus of the RegionFinder software (version 2.6.4; Heidelberg Engineering GmbH), with atrophic areas highlighted in yellow and constraints marked in red. SW-AF, short-wavelength autofluorescence.