Purchase this article with an account.
Tunde Peto, Irene Leung, Ferenc Sallo, Daniela Florea, Alan Bird, MacTel Study Group; Comparison of main OCT features of Type 2 macular telangiectasia in time- and spectral domain OCTs. Invest. Ophthalmol. Vis. Sci. 2013;54(15):3604.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Type 2 macular telangiectasia (MacTel) has characteristic optical coherence tomographic (OCT) signs, including inner and outer retinal cavities and a break in the inner segment/outer segment (IS/OS) line. The location of these relative to the foveal center and the presence of temporal hyper-reflectivity can be identified in time-domain (TD) OCTs. During the MacTel Study, all sites upgraded their OCT equipment to spectral domain (SD) units. All OCT characteristics have been graded by the Moorfields Reading Centre (RC). Our aim was to compare the abnormalities in TD- and SD-OCTs taken at time of upgrading.
All MacTel Study patients have OCT scans done annually. At the time of upgrade, patients were asked to have both types of scans carried out consecutively, using the standard image acquisition method for the given OCT. All images were graded by a trained grader at the RC, sequentially, in order of arrival, but with an interval of at least two weeks between OCT types. At the time of grading of the first OCT set, the grader was masked to the availability of a matching alternate set of images. OCT scans were graded according to the MacTel Study-specific protocol and gradings were analyzed statistically.
A total of 339 eyes had gradable image sets for comparing Stratus TD scans with Spectralis (S-Sp), 233 for Cirrus (S-C) and 170 for Topcon (S-T) SD scans. All SD-OCT images took at least twice as long to grade as TD-OCTs. Agreement on the presence of pigment and neovascular changes was near perfect (κ 0.908-1.0). IS/OS break gradings showed good agreement (κ 0.81-0.86) between all 3 comparison groups. Agreement was lowest for the presence of inner and outer empty spaces on all 3 sets of comparisons (from κ=0.919 for S-T to κ =0.686 for S-Sp). Both features were detected most frequently in Spectralis scans, but differences relative to TD did not reach statistical significance.
Higher detection rates of retinal cysts are probably attributable to the higher resolution of the SD scans, small cavities may not be appreciable in low resolution TD scans. While en-face imaging of the IS/OS from SD-OCT data is a more sensitive method for following progression in MacTel, our study shows that detecting change in the longitudinal follow-up of the MacTel study may not be compromised by the upgrade from TD to SD OCT.
This PDF is available to Subscribers Only