Purchase this article with an account.
S. Joeres, P. G. Updike, S. Ongchin, S. Msutta, A. C. Walsh, S. R. Sadda; Quantitative Correlation Between Fluorescein Angiography and OCT in Neovascular Age-Related Macular Degeneration. Invest. Ophthalmol. Vis. Sci. 2008;49(13):3210. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To analyze the correlation between fluorescein angiographic features and optical coherence tomography (OCT) findings in 65 patients with neovascular age-related macular degeneration (AMD).
Fluorescein angiograms (FA) and radial line StratusOCT B-scans (set of six each) from 65 consecutive eyes with active neovascular AMD were collected. Cases were graded independently by certified reading center graders using a standard protocol and analysis software tools (OCTOR and GRADOR), which allow the user to manually draw the boundaries of various features on angiogram frames and OCT B-scans. Differences between graders were adjudicated to achieve a consensus reading. For FA, area and GLD of the total lesion as well as the individual lesion components (classic choroidal neovascularization (CNV), occult CNV, staining scar, serous pigment epithelial detachment, blood and blocked fluorescence) were calculated. For OCT, the volume (mm3) of the retina, subretinal fluid, subretinal tissue, pigment epithelial detachments (PEDs) and the combination of all 4 spaces (termed "height from choroid") were computed for the whole macula as well as for the foveal subfield alone. Results of FA and OCT analysis were correlated using Pearson Correlation.
There were correlations between the total area of the lesion on FA and the total volume of the "height from choroid" on OCT (p<0.0001, R=0.6), between classic CNV on FA and subretinal tissue on OCT (p<0.0001, R=0.6), and between occult CNV on FA and PED on OCT (p<0.0001, R=0.6). Retinal thickness on OCT correlated best with classic CNV on FA (p<0.01, R=0.4), but not with the total lesion size or occult CNV. The volume of subretinal fluid on OCT correlated with occult CNV on FA (p<0.05, R=0.4), but not with classic CNV or total lesion size. Overall, area measurements on FA showed slightly stronger correlations with OCT features than GLD measurements.
Significant correlations were observed between angiographic features and OCT findings in neovascular AMD. The findings were consistent with the supposition that classic CNV is frequently akin to Type II histologic CNV and occult CNV is often analogous to Type I histologic CNV. Further studies are needed to evaluate if new OCT technologies may allow for even stronger correlations and if OCT may be as sensitive as angiography for detecting and classifying CNV.
This PDF is available to Subscribers Only