Purchase this article with an account.
Himeesh Kumar, Liubov Robman, Galina Makeyeva, Xavier Hadoux, Maxime Jannaud, Lauren A.B. Hodgson, Kai Lyn Goh, Tony Pham, Robyn H Guymer, Zhichao Wu; The influence of Quantity and Location of Hyperpigmentary Abnormalities on the Progression of Age-related Macular Degeneration. Invest. Ophthalmol. Vis. Sci. 2021;62(8):84.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To evaluate the potential added predictive value of the quantity and location of hyperpigmentary abnormalities (HP) for progression of age-related macular degeneration (AMD).
140 participants with bilateral large drusen were followed longitudinally at six-monthly intervals for up to 36 months. They underwent multimodal retinal imaging to determine late AMD development. Baseline colour fundus photographs were graded for the presence of any pigmentary abnormalities. All pixels containing HP were then manually labelled on these baseline images, and the quantity of HP within and outside the central 3mm diameter region centred on the fovea was determined. Two predictive models that included age and any pigmentary abnormalities, one with and one without the quantitative parameters of the HP extent, were developed and evaluated.
HP were detected in 81 (29%) eyes, all of which (100%) had HP inside the central 3mm region, and 41 (51%) eyes had HP outside the central 3mm region. The quantity of HP within the central 3mm region (P = 0.036), but not outside (P = 0.361), was independently associated with progression to late AMD. The addition of the quantity of HP within the central 3mm to a predictive model including presence of any pigmentary abnormalities and age did not significantly improve its performance (area under the receiver operator characteristic curve [AUC] = 0.81), compared to the model without it (AUC = 0.80; P = 0.23). The sensitivity of these two predictive models were 54% and 46% respectively at 90% specificity (P = 0.32).
The extent of HP within the central 3mm was associated with AMD progression at the population level. However, its addition to a model based on presence of any pigmentary abnormalities and age did not result in an improved predictive performance at the individual level in our cohort. Further analysis of spatial patterns of HP, would provide additional parameters to improve the model.
This is a 2021 ARVO Annual Meeting abstract.
This PDF is available to Subscribers Only