Artikel
Type-1 choroidal neovascularization reduces the progression of atrophy in eyes with AMD
Suche in Medline nach
Autoren
-
Maximilian Pfau
- Bonn/D
-
L. von der Emde
- Bonn/D
-
C. Dysli
- Inselspital, Universitätsspital Bern/CH
-
P.T. Möller
- Bonn/D
-
S. Thiele
- Bonn/D
-
M. Lindner
- Nuffield Department of Clinical Neurosciences, University of Oxford/GB
-
M. Schmid
- Institut für Medizinische Biometrie, Informatik und Epidemiologie, Bonn/D
-
S. Schmitz-Valckenberg
- Bonn/D
-
F.G. Holz
- Bonn/D
-
M. Fleckenstein
- Bonn/D
| Veröffentlicht: | 5. Februar 2020 |
|---|
Gliederung
Text
Background: To investigate the impact of co-existent quiescent and exudative type-1 choroidal neovascularization (CNV) on the progression of retinal pigment epithelium (RPE) atrophy in age-related macular degeneration (AMD).
Methods: A total of 114 eyes with AMD of 57 patients (40 female, 17 male) with a mean (± SD) baseline age of 76.54±6.66 years and a median [IQR] review period of 1.24 years [1.02, 1.55] from the prospective, non interventional natural history study DSGA (Directional-Spread-in-Geographic-Atrophy [NCT02051998]) were included. Eyes were subdivided in three categories: geographic atrophy (GA) without evidence of CNV (n=78), GA with quiescent CNV (n=5) and RPE-atrophy with exudative CNV (n=10). Longitudinal fundus-autofluorescence and infrared-reflectance images were semi-automatically annotated for RPE-atrophy using the RegionFinder software. CNV lesions were spatially mapped to these annotations based on optical coherence tomography angiography (OCTA). Both overall and localized RPE-atrophy progressions in topographic relation to the presence of CNV were analyzed using cross-validated mixed-effects models.
Results: Baseline GA size ([estimate ± SE] 9.06±1.06 mm2]) and age did not differ significantly among the subgroups. The model to predict overall progression rates, considering the baseline lesion size and further shape-descriptive factors, achieved a high accuracy (cross-validated R2=0.428). Hereby, the observed progression rates were significantly slower than the predicted progression rates for eyes with quiescent CNV (discrepancy of 0.35±0.16 mm2/year) and exudative CNV (0.32±0.14 mm2/year). The localized prediction model achieved excellent Results with a cross-validated Dice coefficient [95 % CI] of 0.87 [0.85, 0.89]. The localized presence of CNV significantly (P<0.001) reduced the odds for future atrophy involvement by a factor of 0.45 [0.42, 0.48].
Conclusion: The Results indicate that the presence of quiescent and exudative type 1 CNV is associated with a reduced overall and localized RPE-atrophy progression. This observation highlights the potential protective effect of CNV on the RPE and overlying neurosensory retina.
