BRIDGING THE GAP

A multimodal approach, combining spectral domain optical coherence tomography (SD-OCT) and fundus autofluorescence, can provide very useful means of quantifying the extent and morphology of atrophic lesions in eyes with geographic atrophy, said Ramzi G Sayegh MD, Medical University of Vienna, Vienna, Austria, at the 13th EURETINA Congress in Hamburg. “Until recently imaging of geographic atrophy consisted in two dimensional en-face imaging with fundus photography and fundus autofluorescence, and OCT’s role in geographic atrophy had yet to be established. But our findings show it is now possible to reliably delineate the area of atrophic lesions with spectral domain OCT,” Dr Sayegh said.

He noted that fundus autofluorescence enables an exact assessment of the atrophic lesion area and that the autofluorescence patterns have a prognostic value, in the lesions' progression. Lesions with a banded or diffuse pattern grow more rapidly than those with a focal pattern or those with no fundus autofluorescence abnormalities. However, fundus autofluorescence does not provide insights into the changes in retinal morphology. On the other hand, research conducted at the University of Vienna shows that modern spectral domain optical coherence tomography enables the visualisation of ultrastructural alterations of retinal and subretinal layers which is highly diagnostic and which correlates closely with vision loss, Dr Sayegh said. But unlike fundus autofluorescence, SD-OCT cannot as yet distinguish any features that can predict the rapidity of the disease’s progression.

SD-OCT vs. fundus autofluorescence

Dr Sayegh outlined a series of studies he and his associates conducted comparing the findings obtained through SD-OCT with those obtained through fundus autofluorescence, time domain OCT and other technologies. In the first study they compared complete SD-OCT datasets of 81eyes of 42 patients with geographic atrophy and looked at specific morphological changes in the lesion area and compared them with fundus autofluorescence images from the same eyes. Two readers of the Vienna Reading Centre graded all of the imaging obtained with the two techniques.

The researchers found that the OCT parameters which correlated most closely with FAF were choroidal signal enhancement, with a 96 per cent correlation, the area of outer plexiform layer shifting from 84 per cent, and external limiting membrane loss with a 93 per cent correlation. A follow-up of 48 eyes produced very similar results, Dr Sayegh noted. “We found a very good agreement between OCT and fundus autofluorescence measurement results at baseline and at follow-up. However, we also saw that the area of complete RPE loss correlated to only 40 per cent of the lesion area detected by fundus autofluorescence at baseline and follow-up,” he said.

Further comparisons

To determine which alterations in retinal morphology corresponded to loss of visual function, Dr Sayegh and his associates conducted another study in which they combined SD-OCT imaging with microperimetry fundus images overlaid with an infrared image (MP1, Nidek). To perform their analysis they used the OCT-tool-kit, a suite of software designed by the Vienna Reading Centre that enables the delineation of the lesion area in SD-OCT volume scans and which locates the sensitivity points obtained through microperimetry in their precise relationship to the retina’s anatomy.

The study involved 43 eyes of patients with geographic atrophy. The researchers graded 1,005 sensitivity points in each eye. They found that in regions where the external limiting membrane was lost, the sensitivity was reduced by 1.3 decibels, and where both the the external limiting membrane and the retinal pigment epithelium were lost there was a complete loss of sensitivity. “From this we concluded that grading outer retinal layers in addition to the retinal pigment epithelium is important,” Dr Sayegh said.

In yet another study, involving 81 eyes of patients with geographic atrophy, the Vienna researchers compared spectral domain OCT with the CirrusR (Carl-Zeiss Meditec) and the SpectralisR (Heidelberg- Engineering), to time domain OCT StratusR OCT imaging (Carl-Zeiss Meditec) and to fundus autofluorescence and fundus autofluorescence combined with infrared imaging. For external parameters they also measured visual acuity and performed microperimetry and fixation testing (MP1, Nidek). That study showed that visual acuity correlated significantly more closely with grading of the fovea in OCT images than with FAF imaging (p<0.0001), and that the SD-OCT performed significantly better than time domain OCT. In addition, central fixation correlated closely with foveal sparing as detected by OCT.

Summarising, Dr Sayegh said that the studies show that SD-OCT can precisely determine the size of geographic lesions and that the best parameters are choroidal signal enhancement, outer plexiform layer shifting and external limiting membrane loss. In addition, their research also showed that external limiting membrane loss corresponds with sensitivity reduction and that foveal sparing is best detected by SD-OCT. “However, since we still haven’t found any OCT parameter that corresponds to the lesion growth rate, multimodal imaging in geographic atrophy with fundus autofluorescence and OCT is required,” he added.