Emerging technologies open new horizons in glaucoma structural evaluation, management

Technological evolution, with optical coherence tomography in the forefront, is gradually but radically changing the diagnostic and management approach to glaucoma. Earlier detection of structural changes allows prediction of functional consequences and quantification of progression.

“The physiological structure-function relationship is beyond question and the basis for glaucoma management; however, the available technologies to assess either aspect are far from perfect. Current developments suggest that the threshold for detecting damage might be lower in structural imaging than in standard automated perimetry,” Herbert Reitsamer, MD, said. “The impact of structural documentation on therapy decisions has not been so great in the past but will become increasingly important with the high specificity and sensitivity of emerging technologies.”

“OCT will be able to detect glaucomatous abnormalities prior to our current functional measurement technology being able to detect them, and the earlier in the disease that you can detect damage or progression, the more likely it is you are able to prevent further progression with less intensive intervention,” OSN Glaucoma Board Member and co-inventor of OCT Joel S. Schuman, MD, said.

Wolfgang Drexler, PhD, who played a leading role as a biomedical engineer in the development of OCT, added that “quantifying function is significantly harder than quantifying morphology. Visual field is not a very precise technique, and progress in this area is much more challenging.”

According to Schuman, an OCT system is a better method of detecting and following glaucoma once it hits a tipping point of 75 µm of retinal nerve fiber layer (RNFL) thickness because it shows a good correspondence between structural and functional loss.

“At about 50 µm to 55 µm, OCT is no longer able to measure change in the RNFL thinning, but visual field can be used to measure the progression of glaucoma at that point,” Schuman said.

Swept-source OCT

Swept-source OCT has brought significant improvement in imaging speed and depth range, the latter due to less sensitivity loss with scanning depth.

“An ophthalmic OCT system that is commercially available (Topcon Medical Systems) enables 100 kHz A-scan rate. In the near future, we’ll see up to 200 kHz, and in the long term, 400 kHz. Spectral-domain systems at the moment go up to 70 kHz,” Drexler said.

Faster speed allows properly sampled three-dimensional scanning of a larger area of the retina. In addition, a longer scanning depth is provided thanks to less depth dependent sensitivity reduction of the system, which has overcome the sensitivity losses of spectral-domain OCT.

“Swept source has a uniform high sensitivity that allows, from the nerve fiber layer to the choroid and lamina cribrosa, very good sensitivity in the same scan,” Drexler said.

The higher speed and better sensitivity combined with the 1050 nm wavelength of current systems also allow better OCT measurement of the retina in eyes with cataract or corneal opacities. Access and visualization of the choroid are possible because the longer wavelength allows for reduced scattering, irrespective of the fundus pigmentation.

“For glaucoma, that means you can have all the retinal layers properly segmented. You can have access to the choroid. You can visualize the optic disc better, especially in myopic long eyes. You can see the lamina cribrosa properly. You might use this also as a biomarker of early glaucoma changes,” Drexler said.

The role of choroidal changes and choroidal blood flow in the pathophysiology of glaucoma is currently the focus of much interest among investigators, and OCT technology could play a key role.

Using OCT technology as a screening tool for glaucomatous changes has good potential because of its high sensitivity and specificity as long as the quality of the image is good, Schuman said.

“If you have a good image and the segmentation had worked properly, the sensitivity and specificity for OCT are quite high and may have utility in screening for the disease,” he said.

Finding a place for visual fields

Although OCT is more beneficial than visual fields when it comes to diagnosing and managing glaucoma, visual field testing has not lost its place.

“The [OCT] test is much less unpleasant for the patient and has significantly lower variability than visual fields, but that does not mean that visual fields do not have a place [in glaucoma diagnosis and management],” Schuman said.

There is good correspondence between visual fields and OCT, and there is a point at which glaucoma becomes more advanced and OCT, as it currently stands, is no longer helpful. At that point, visual fields can be used to detect glaucoma progression, according to Schuman.

In a study published in British Journal of Ophthalmology, Schuman and colleagues investigated the “tipping point,” the point at which functional loss measured by visual fields becomes detectable in relation to structural changes in RNFL thickness, and found that visual fields differ significantly from one another.

“Pre-perimetric glaucoma can be identified by OCT RNFL thickness with the detectable visual field loss first appearing with RNFL approximately 17% below that expected for healthy eyes,” Schuman and colleagues said.

Research shows that OCT detects abnormalities and change before measurable visual field perimetry, Schuman said.

“You need to lose about 17% to 20% of nerve tissue before an abnormality is likely to be present on a standard achromatic visual field,” Schuman said. “The tissue after that shows pretty good correspondence between progression by visual field and progression by OCT. The floor effect appears to be an artifact of the RNFL segmentation algorithms that are used commercially and the effect makes it appear as if the RNFL is stable, but the visual fields can continue to show progression.”

Measuring blood flow

The retina is one of the most oxygen- and blood flow-demanding tissues in the body and is simultaneously served by two distinct circulatory systems.

“One comes from the choroid, which has high autonomic nervous input and strong capacity for local blood flow regulation, and the other one comes from retinal circulation, which is primarily metabolically regulated. And this all occurs within 200 µm, which is a very unique situation in the body,” Reitsamer said.

Impaired blood flow regulation is likely to play a role in glaucoma, as shown in several studies. However, the reasons are still unclear, and it is difficult, Reitsamer said, to get a grip on blood flow, first in terms of measurement and even more in terms of treatment.

Blood flow is still an overlooked concept in clinical practice and research. There is therefore a lack of large-scale studies on blood flow-related factors, Reitsamer said.

“The main reason might be that the methods so far are not good enough. If we could measure volumetric blood flow on a capillary level in a reproducible way, this technique would find its way into the clinics very quickly,” he said.

The old method of measuring choroidal blood flow with scanning laser Doppler flowmetry with the Heidelberg Retina Flowmeter (HRF) lacked reproducibility. Because measurements were made on small spots, the signal-to-noise ratio depended on what was backscattered from the surface, and the variability of measurements was high.

“A dark surface gives a different signal than a bright surface. Therefore, the studies performed with the HRF had different results. In absence of volumetric blood flow, interindividual comparison is not possible because people have very different optic situations in terms of reflectance and how far the signal goes into the tissue,” Reitsamer said.

New approaches of measuring choroidal blood flow are under investigation. Doppler OCT enables measurement of erythrocyte velocity, which, correlated with blood vessel diameter, potentially gives the volumetric blood flow. However, the OCT resolution currently allows for measurement of diameter only in larger vessels.

“It is an evolving technique. Swept-source OCT works at a faster speed and has therefore enhanced capability for looking at dynamic processes like blood flow. The bidirectional Doppler OCT also looks particularly promising,” Reitsamer said.

OCT-based oximetry may also be a possibility in the future.

“Looking at the speed of development in OCT research, there is a place for that. Blood flow, after all, is only a surrogate for whatever function blood flow serves in the tissues, and the delivery of oxygen is one of the most important functions,” Reitsamer said.

A view into the cells

The application of adaptive optics (AO) to retinal image acquisition techniques enables in vivo visualization of cellular structures by real-time measurement and correction of ocular aberrations. AO has been applied to fundus cameras, scanning laser ophthalmoscopy (SLO) and, more recently, OCT.

The only commercially available system is the Adaptive Optics Retinal Camera produced by Imagine Eyes, “a super-easy-to-use flood illumination AO camera which provides 2-D cellular resolution imaging of the retina on a large field of view,” Drexler said.

An AO-SLO system is at the prototype stage as a collaborative project among Canon and a number of universities and medical research institutes in Japan and abroad.

“AO-SLO allows the surgeon to evaluate integrity of individual photoreceptor cells, primarily cones, particularly in eyes with macular diseases and retinal degenerative diseases. It shows circulation of leukocytes and red blood cells in the retinal capillaries, which enables detection of non-perfusion in eyes with diabetic retinopathy. In eyes with glaucoma, it allows us to detect pathologic changes in retinal nerve fiber bundles,” Masanori Hangai, MD, PhD, said.

In a study, Hangai and co-authors at Kyoto University were able to demonstrate by AO-SLO that nerve fiber bundles were narrower in glaucomatous eyes than in normal eyes.

“These abnormalities were associated with visual field defects, suggesting that AO-SLO may be useful for detecting early nerve fiber bundle abnormalities associated with loss of visual function,” he said.

AO combined with OCT adds a third dimension and might allow 3-D visualization of not only photoreceptors but also retinal pigment epithelial cells, choriocapillaris and nerve fiber bundles.

“From the resolution point of view, AO-OCT has also the capability to visualize ganglion cells. The problem is that the contrast does not yet allow any AO imaging system to resolve cells without any kind of contrast enhancement by labeling,” Drexler said.

Research is ongoing, and the first experimental studies showed potential. However, stepping from prototype to production and clinical practice requires a compact, easy-to-use AO-OCT system that is cost-effective.

“We need to demonstrate that the additional cost is justified because you get much more cellular-based diagnostically relevant information. I received funding for a European project where together with Imagine Eyes we are looking into a compact AO-OCT system that probably will be used in a multicenter trial in the future,” Drexler said.

DARC: Measuring ganglion cell apoptosis

A new, noninvasive real-time imaging technology called DARC (detection of apoptosing retinal cells) utilizes fluorescently labeled annexin 5 and SLO to visualize retinal ganglion cells undergoing apoptosis.

“We see three scenarios for its use. First is to allow glaucoma to be detected much earlier than it is currently possible. The gold standard for glaucoma now is when a [visual field] defect develops, and we know that this could be as much as after 50% of retinal ganglion cells have died. There is a delay of about 10 years between when the disease starts and when it is detected. Hopefully DARC will allow us to pick up the disease 10 years earlier than it is currently possible,” Francesca Cordeiro, MD, PhD, said.

The second scenario, she said, is to monitor the effects of treatment through the response of retinal ganglion cells.

“If treatment stops or slows down cell apoptosis, then we know it is working. Rather than unnecessarily wait 6 weeks or 3 months or 6 months, we think we should have an answer in weeks,” she said.

The third application might be to use DARC as a surrogate marker in clinical trials. With drugs that normally require long-term use to show efficacy, such as neuroprotective drugs, DARC could shorten the duration of clinical trials to much less than is currently expected.

Cordeiro expects that toward the end of the year authorization should be obtained to start phase 1 clinical trials. Preclinical trials were performed initially with intravenous administration of the DARC agent, but a noninvasive eye drop version has been developed.

“The protein we are looking at, which is annexin, has been used in about 30 clinical trials around the world. Phase 2 clinical trials are currently ongoing, though they are testing with a radio-labeled annexin as opposed to a fluorescence-labeled annexin, which is the basis of DARC,” Cordeiro said. – by Michela Cimberle and Nhu Te

• References:
• Bizrah M, et al. BMC Bioinformatics. 2014;doi:10.1186/1471-2105-15-169.
• Cordeiro MF, et al. Eye (Lond). 2011;doi:10.1038/eye.2011.64.
• Cordeiro MF, et al. Ophthalmic Surg Lasers Imaging. 2009;doi:10.3928/15428877-20090901-07.
• Drexler W, et al. J Biomed Opt. 2014;doi:10.1117/1.JBO.19.7.071412.
• Galvao J, et al. Curr Opin Pharmacol. 2013;doi:10.1016/j.coph.2012.08.007.
• Guo L, et al. Prog Brain Res. 2008;doi:10.1016/S0079-6123(08)01130-8.
• Kersey T, et al. Indian J Med Res. 2013;137(4):659-668.
• Normando EM, et al. Cell Tissue Res. 2013;doi:10.1007/s00441-013-1554-5.
• Sommer A, et al. Arch Ophthalmol. 1991;doi:10.1001/archopht.1991.01080010079037.
• Stalmans I, et al. Acta Ophthalmol. 2011;doi:10.1111/j.1755-3768.2011.02178.x.
• Takayama K, et al. Am J Ophthalmol. 2013;doi:10.1016/j.ajo.2012.11.016.
• Torti C, et al. Opt Express. 2009;doi:10.1364/OE.17.019382.
• Werkmeister RM, et al. Cell Tissue Res. 2013;doi:10.1007/s00441-013-1600-3.
• Wollstein G, et al. Br J Ophthalmol. 2012;doi:10.1136/bjo.2010.196907.

• For more information:
• Francesca Cordeiro, MD, PhD, is professor of retinal neurodegeneration and glaucoma studies at University College London. She can be reached at UCL Institute of Ophthalmology, Bath Street, London EC1V9EL, England; 44-207-608-6938; email: m.cordeiro@ucl.ac.uk.
• Wolfgang Drexler, PhD, is professor of medical physics at the University of Vienna. He can be reached at Medical University Vienna, Center for Medical Physics and Biomedical Engineering, General Hospital Vienna, 4L, Waehringer Guertel 18-20, A-1090 Vienna, Austria; 43-1-40400-1986; email: wolfgang.drexler@meduniwien.ac.at.
• Masanori Hangai, MD, PhD, is professor and chairman at the department of ophthalmology of Saitama Medical University. He can be reached at 38 Morohongo, Moroyama, Iruma, Saitama 350-0495, Japan; 81-49-2761250; email: hangai@saitama-med.ac.jp.
• Herbert Reitsamer, MD, is senior physician and professor at the Paracelsus University of Salzburg, Austria. He can be reached at Salzburger Landeskliniken – Landeskrankenhaus, Müllner Hauptstr. 48 , A-5020 Salzburg, Austria; 43-662-44823701; email: ha.reitsamer@gmail.com.
• Joel S. Schuman, MD, is Distinguished Professor and Chairman of Ophthalmology at University of Pittsburgh and the director of UPMC Eye Center. He can be reached at Department of Ophthalmology, UPMC Eye Center, University of Pittsburgh School of Medicine. 203 Lothrop St., Suite 816, Pittsburgh, PA 15213; email: schumanjs@upmc.edu.

Disclosures: Cordeiro is employed by UCL, which holds the patent of DARC technology. Drexler is a consultant to Carl Zeiss Meditec and Imagine Eyes. Hangai is on the advisory board of Nidek, received consulting fees from Topcon and received research funding from Canon. Reitsamer has no relevant financial disclosures. Schuman receives royalties licensed by Massachusetts Institute of Technology and Massachusetts Eye and Ear Infirmary to Zeiss.

What are the current possibilities to correlate structure and function in glaucoma progression?

The view of the clinician

Correlating structure and function in glaucoma progression is currently a challenge. At the onset of glaucoma, progression is more likely to be detected with structure analysis by Heidelberg Retina Tomograph, GDx and optical coherence tomography. With conventional perimetry, visual field shows changes relatively late. On a later stage, when there is advanced damage of the optic nerve and retinal nerve fiber layer, imaging systems for structural changes do not provide us with much useful information because a lot of the tissue is already lost and progression is better detected by perimetry. If changes occur, perimetry should be performed frequently, but it does not make a lot of sense, in my opinion, to perform structural assessment more often than once a year, or twice in a few cases. There are patients who need visual field assessment three or four times a year. Once visual field progression has been detected, you have to confirm it and then maybe reconfirm a third time. Measurements should not be affected by learning curve effects, fluctuations or other factors.

In some of my patients, mainly young ones, I use the Heidelberg Edge Perimeter (HEP), especially in early glaucoma or in the presence of a normal visual field. Special software allows us to combine the Heidelberg Spectralis OCT and HEP, correlating structural and functional changes. One machine is speaking with the other. Hopefully, further technological advances will allow us, in a not too distant future, to correlate structural deterioration in a certain area with visual field analysis in the same area and to find specifically located visual field defects. Heidelberg and Zeiss are researching this possibility, and other manufacturers will follow.

Anton Hommer, MD, is from Sanatorium Hera, Vienna, Austria. Disclosure: Hommer is a consultant to Allergan and Santen.

The view of the physicist

The nature of the structure-function correlation is of great interest to both clinicians and scientists. While it seems natural that this correlation exists, proving it in practical terms turns out to be a rather difficult problem, limiting its current clinical application.

First, the correlation may not be simply linear; it depends on how function and structure are represented, for example using a common unit of measurement, such as the number of retinal ganglion cells, to express both structure and function results in an inherently linear correlation. Even then, other factors that vary between individuals are involved, such as the retinal location, the patient’s age and the density of structural tissue. New imaging technologies, such as polarization-sensitive optical coherence tomography, may enable the estimation of these factors for individual patients in the clinic.

The second issue is spatial correspondence. Structure and function are measured at different locations, and we need to define their relation. This becomes a problem when correlating, for example, the retinal nerve fiber layer near the optic disc with perimetric data of the central visual field. Such spatial correlation has been investigated for groups of patients, but the considerable interindividual variation severely limits its use in individual patients. High-resolution adaptive optics imaging could be used to measure the spatial correspondence in an eye by accurately tracing fiber bundles.

Finally, structure and function likely do not progress simultaneously. While this makes analysis difficult, it is precisely this aspect that holds the greatest promise for clinical applications. If, as is suggested by multiple studies, structural change precedes functional change, imaging provides an opportunity to identify progression before functional damage results.

In conclusion, correlating structure and function in clinical practice currently seems largely unfeasible. But with new imaging technology, such as polarization-sensitive OCT and adaptive optics, we can make great progress to make this possible in the future.

Koen Vermeer, PhD, is a senior scientist at the Rotterdam Ophthalmic Institute, The Netherlands. Disclosure: Vermeer has no relevant financial disclosures.