Session Title: Free Paper Session 25: Imaging IV
Session Date/Time: Sunday 10/09/2017 | 10:00-11:30
Paper Time: 10:18
Venue: Room 113
First Author: : A.Bagchi UK
Co Author(s): : R. Schwartz S. Sivaprasad
Fundus fluorescein angiography (FFA) is the gold standard for detecting myopic choroidal neovascularisation (mCNV). New imaging modalities are routinely used in the management of mCNV. However, the diagnostic accuracy of these modalities remains unclear. The current study will compare the diagnostic performance of spectral domain optical coherence tomography (SD-OCT) and optical coherence tomography angiography (OCT-A) either as a single modality or in combination in eyes with mCNV compared to FFA. We will also address whether any additional information obtained on OCT-A adds value to the disease mechanism.
Medical Retina clinics at Moorfields Eye Hospital between October 2016 and February 2017.
Consecutive patients with high myopia who presented with recent onset of visual disturbances and underwent FFA, SD-OCT and OCT-A to confirm the diagnosis of mCNV were included in this study. Exclusion criteria included poor quality of images due to large haemorrhages, lack of one or other imaging modality, presence of any other ocular condition other than CNV, ocular inflammation, vitreo-retinal disorders and glaucoma. The standard imaging modality used in this study is FFA and the other 2 imaging modalities were compared to FFA. Presence of CNV on FFA was identified by early hyperfluorescence with increasing intensity in late frames or hypofluorescence due to sub-retinal haemorrhage. On SD-OCT, hyper-reflectivity; sub/intra-retinal fluid were used to detect CNV. OCT-A imaging was performed on AngioPlex CIRRUS HD-OCT model with a scanning area of 3x3 mm centred on the fovea. A bright interlacing or tangled group of vessels was used to detect the CNV.
Thirteen eyes met the eligibility criteria. The images quality was good. 12 eyes clearly demonstrated CNV on FFA (92.3%). CNV was not identified in one patient. OCT-A alone accurately detected CNV in 91.6% (n=11; true positives). One eye had features of CNV on OCT-A with no CNV on FFA (n=1; false positive). Detection rate of CNV on SD-OCT was 83.3% (n=10). The one eye in which both OCT-A and SD-OCT did not detect a CNV complex was 0.07 mm2 in size on FFA. Presence of haemorrhage made detection of the CNV doubtful on SD-OCT in one patient. On OCT-A, 10 eyes had a foveal CNV and 2 eyes and extra-foveal CNV. The margins of the CNV were well defined in 9 eyes and ill - defined in 3. The appearance of the CNV had a few characteristic features. Most frequently, the CNV appeared as a “tight net” (n=6, 50 %) with a few eyes having a filamentous appearance (n=2, 17 %), glomerulus like appearance (n=2, 17 %) and a loose interlacing pattern seen in 2 eyes (n= 2, 17 %). The core of the CNV was only visible in 3 of 12 eyes (n= 3, 25 %).
The diagnostic performance of both OCT-A and OCT are excellent but FFA remains the gold standard. If the CNV is detectable on OCT-A and OCT, FFA can be avoided. However, an absence of CNV on OCT-A and OCT warrants an FFA in symptomatic patients with mCNV.