GENE THERAPY

Advances in imaging technologies and the development of increasingly sophisticated testing methods are helping to drive forward efforts to quantify the therapeutic effect of clinical trials for hereditary retinal diseases, according to an expert in the field.

“While phenotyping was very important in the past to find genes, now we increasingly have the opportunity to enter clinical trials for genetic eye disease. And in this respect it is critically important to have the most sensitive methods possible in order to monitor any possible treatment effect,” Birgit Lorenz MD, PhD, FEBO told delegates attending the 2013 Congress of the Society of European Ophthalmology (SOE).

Current research in the field of therapeutic monitoring and management is focusing on several distinct goals, said Prof Lorenz, chairman and head of the Department of Ophthalmology and professor of neuroophthalmology, paediatric ophthalmology and ophthalmic genetics at Justus-Liebig-University, Giessen, Germany.

“The aim now is to quantify layer-specific changes in retinal dystrophies longitudinally, as at present the gene therapies that are available or will become available are not able to reverse the disease but just stop progression. We also want to be able to quantify genotype-specific changes, to correlate morphology with psychophysics in a multimodal manner, and to provide objective measures for treatment trials in retinal dystrophies,” she said.

Many inherited retinal pathologies such as retinitis pigmentosa typically result in reordering of the intraretinal layers, with some layers becoming even thicker over time as the disease progresses, said Prof Lorenz.

“This is important with respect to analysing OCT data. Analysing the full retinal thickness alone, however, may lead to errors, because if the inner retina is thickened but there is atrophy of the outer retinal layer it may look relatively normal. So layer thickness is important in current experimental treatment trials for gene therapy. We need to be aware that the photoreceptor layer may be reduced while the whole retina thickness is still normal, so further analysis is warranted,” she said.

To assist in that process, Prof Lorenz’s group at the University of Giessen have developed device-independent OCT analysis software (DIOCTA) which allows for the automated segmentation of intraretinal layers as well as thickness measurement of individual layers (Pilch M, Stieger K, Wenner Y, Preising MN, Friedburg C, Meyer Zu Bexten E, Lorenz B.

Invest Ophthalmol Vis Sci. 2013 Jun 27;54(6):4385-93. doi: 10.1167/iovs.12-11396. [PubMed - in process].

Concerning retinal layer analysis, there is currently no gold standard for such measurements, Prof Lorenz told EuroTimes.

“Three major points have to be discussed in this respect. The first is that layer segmentation as well as total retinal thickness values are generally device dependent. This is due to the fact that different reference lines and hallmarks are used. We and others have published on this effect (e.g. Chopovska Y, Jaeger M, Rambow R, Lorenz B., Ophthalmologica. 2011;225(1):27-36. doi: 10.1159/000316693. Epub 2010 Aug 7.) While DIOCTA was developed to overcome this problem, the system needs to get the primary data which not all companies are willing to give. Presently, we have access to the original data of Spectralis, OPTOVue and Stratus III,” she said.

The second issue, said Prof Lorenz, relates to the particular challenge of developing effective algorithms for diseased retina. “If experienced graders segment diseased retinae, variations among them will surely arise. Because manual segmentation is time consuming and depends on the experience of the grader, several groups are working towards automated layer segmentation. When analysing healthy retinae, most algorithms work fine. However, when it comes to pathological alterations of the retina, most algorithms fail, because layer structures disappear or locally defined structures appear and the entire retina is modified,” she said.

Finally, Prof Lorenz highlighted the issue of identifying which structures are really important in a given disease and therefore worth segmenting.

“Are functional and morphological correlations possible? Currently, the so-called ellipsoids seem to be an indicator of visual function. They are anatomically defined as the part of the inner segments containing the mitochondria and on the OCT scan present as a strong reflecting line (dark or bright or red according to the mode of display). If this line is visible on an OCT, the corresponding retinal zone seems to be functional or potentially capable to maintain function and thus could be treated by gene therapy (gene addition of the respective gene). But this is only theory and needs to be confirmed in large studies. In this regard, it is too early to talk about any standardised method in analysing morphologically gene therapy treatments,” she said.

For a retinal disease such as achromatopsia, scheduled to be one of the next disorders to be treated by gene therapy, defined local alterations at the level of outer segments can be observed on OCT in the region of the foveola, potentially reflecting the pathologic alterations of cone outer segments, said Prof Lorenz.

“This alteration can be quantified by DIOCTA. It is possible that these morphological alterations reverse to a physiological state following gene therapy, and the change in volume or thickness of these alterations could be quantified by DIOCTA, thus representing a biomarker,” she said.

Although significant progress has been made in imaging technologies and software, some obstacles remain to be overcome to take testing to the next level, said Prof Lorenz.

“Resolution of clinical OCT devices is still limited. Also, in patients with nystagmus, such as Leber’s congenital amaurosis and achromatopsia, the quality of the scans may not be optimal. Faster scans help to overcome this problem. Also, unsteady fixation or no fixation is a serious problem. Much higher resolution comes from adaptive optics, but nystagmus will still be a problem with this method that with good fixation allows us to visualise individual cones. Rods cannot yet be imaged individually due to their much smaller size. Also, imaging of the foveola remains difficult,” she said.

In terms of future developments, Prof Lorenz cited multicolour imaging, ultrawide imaging and adaptive optics as offering exciting possibilities for progress in retinal imaging.

“Multicolour imaging is a technique that uses light at different wavelengths that consequently enters the retina with varying depth thus providing morphological information from different layers all in one image. It therefore might be an improved funduscopy technique enhancing visibility of certain structures. The value of this multicolour imaging remains to be proven,” she said.

Ultrawide imaging enables the ophthalmologist to obtain an image of almost the entire retina with one shot, in contrast to the necessity to take multiple images around the macula and papilla and then merging them afterwards, said Prof Lorenz.

Adaptive optics also holds rich potential for retinal imaging, said Prof Lorenz, providing a means to display the retina en face with such a high resolution that even single cones can be visualised in the parafoveal zone.

“Consequently, especially cones can be analysed with this technique, generating data about their viability in the natural course of diseases or following treatment,” she concluded.