Abstract
Purpose: :
To evaluate the sensitivity and specificity of surrogate markers for AMD such as the retinal thickness map or subretinal hemorrhage, for detecting the presence of OCT features of exudation commonly used to guide retreatment decisions.
Methods: :
We retrospectively collected clinical and Cirrus spectral domain (SD) OCT data from patients with active neovascular AMD undergoing anti-VEGF therapy who presented to our institute between May and November 2010. Volume SDOCT macular cube (512x128) B-scans were reviewed and the presence of intraretinal, subretinal, or sub-RPE fluid was recorded. The presence of any fluid on visual inspection of all OCT B-scans was deemed to be the gold-standard indicator for treatment. Separately, in a masked fashion, OCT retinal thickness maps were reviewed and qualitative or quantitative evidence of persistent disease activity was noted (e.g. increased or persistent thickening). High definition 5-line raster scans were also reviewed for presence of fluid. In addition, the presence of a drop in visual acuity of 5 more letters and the presence of new subretinal hemorrhage were both recorded from review of clinical records. The sensitivities and specificities of the thickness map, 5-line raster scan, decrease in vision, and subretinal hemorrhage for predicting the need for re-treatment were computed.
Results: :
A total of 74 visits (35 patients) were reviewed, among which 57.2 % were deemed to have fluid on volume OCT suggesting need for re-treatment. When compared to this gold-standard of visual inspection of all 128 B-scans for the presence of any fluid, the sensitivities (SN) and specificities (SP) of the various surrogate markers were: vision loss of greater than 5 letters (SN 27.8%, SP 81.5%), central subfield (CSF) thickness increase of greater than 100 microns (SN 2%, SP 100%), presence of new subretinal hemorrhage on clinical examination (SN 12.8%, SP 100%), worsening of disease based on a qualitative comparison of the OCT Thickness Map (SN 29.8%, SP 82.8%), any abnormality on the OCT Thickness Map (SN 98%, SP 20.1%), presence of fluid on OCT 5-line raster scans (SN 91.5%, SP 65.5%).
Conclusions: :
Surrogate markers such as subretinal hemorrhage and increased CSF thickness are specific but not sensitive for identifying patients requiring anti-VEGF re-treatment. Considering any qualitative abnormality on the thickness map improves sensitivity but at the expense of specificity. Review of these surrogate markers may not be an adequate replacement for comprehensive review of all OCT B-scans when treating patients with neovascular AMD with anti-VEGF therapy.
Keywords: age-related macular degeneration • imaging/image analysis: clinical • retina