EyeNet Magazine

Clinical Update: Glaucoma
Posterior Segment Imaging: Correlating Structure and Function
By Annie Stuart, Contributing Writer
Interviewing Sanjay G. Asrani, MD, Jeffrey M. Liebmann, MD, and Joel S. Schuman, MD
Academy members: login to read or make comments on this article.

(PDF 382 KB)

Given its insidious nature, glaucoma has long presented ophthalmologists with diagnostic challenges. The advent of optic nerve imaging some 15 years ago opened an important new window on glaucomatous damage and disease progression. With continuing advancements in technology, the structural assessment made possible by posterior segment imaging now serves as a valuable adjunct to the functional appraisal provided by perimetry. Such structure-function correlations are moving the clinical evaluation of open-angle glaucoma increasingly from the subjective to the objective.

“In terms of checking change in the optic nerve, imaging devices today can bring the general ophthalmologist up to the level of the glaucoma specialist,” said Jeffrey M. Liebmann, MD, director of glaucoma services at the Manhattan Eye, Ear and Throat Hospital and New York University in New York. “These devices add information that we have trouble getting otherwise.”


Types of Imaging Technologies

For the posterior segment, three main technologies are in use today: confocal scanning laser ophthalmoscopy (CSLO), scanning laser polarimetry (SLP) and optical coherence tomography (OCT). In addition to aiding diagnosis, these devices provide quantifiable data that can be used to help predict which patients might progress, said Dr. Liebmann.

Confocal scanning laser ophthalmoscopy. CSLO provides topographic mapping of the optic nerve head and peripapillary retina, as well as an estimate of the retinal nerve fiber layer (RNFL) around the optic nerve head.

The most commonly used device is the Heidelberg Retinal Tomograph (HRT), said Dr. Liebmann, and one of its strengths is that the acquisition of data has remained consistent since the device’s introduction. This permits the use of patient data across time to watch for changes in optic disc topography.

“The absence of change over time using this device can confirm the stability of the optic disc,” said Dr. Liebmann. “For many patients, confirming stability is just as important as detecting change, which of course suggests that treatment should be advanced.”

A drawback of the HRT, said Sanjay G. Asrani, MD, associate professor of ophthalmology at Duke University, is that it evaluates a global parameter: the optic nerve, where all the axons converge. “If there is a localized loss somewhere,” he said, “it may not be easy to pick up at the convergence point.” Dr. Asrani added that because there is great variability in what’s normal, it’s difficult to obtain a baseline standard against which to compare early abnormalities. However, once an abnormality is found, HRT is useful for following patients.

Scanning laser polarimetry. SLP measures RNFL birefringence to estimate tissue thickness. “It looks at how the nerve fiber layer reflects light back,” said Dr. Asrani, noting that the Carl Zeiss Meditec GDx is currently the most commonly used device.

Polarimetry assesses RNFL thickness around the optic nerve. Because the technology is based on reflectivity, measurement is hampered by polarization of the ocular media, said Dr. Asrani, and this can lead to confounding by non-RNFL birefringence. A newer model, the GDx-ECC with enhanced corneal compensation, produces more reproducible results and more accurate discrimination of glaucoma than in the past.1

As with other technologies, GDx can detect progression, said Joel S. Schuman, MD, professor of ophthalmology at the University of Pittsburgh. “There is some variability in nerve fiber index,” he said, “so there’s a need for more data to validate the utility of the guided progression analysis software provided by the manufacturer, but this parameter appears to be promising.”

Optical coherence tomography. Based on low-coherence interferometry, OCT examines backscattered light, creating high-resolution tomographic images. “OCT provides cross-sectional data, allowing measurement of RNFL thickness, retinal thickness as a surrogate for ganglion cell thickness, and the optic nerve structure,” said Dr. Asrani. He said that the ability to obtain information about localized and general RNFL loss, ganglion cell thickness and retinal anatomy has contributed to the growing use of OCT.

“Spectral-domain OCT (SD-OCT) scans tissue at a much faster rate with higher resolution than the older time-domain (TD-OCT) versions, providing more detailed images,” said Dr. Asrani. SD-OCT works by collecting backscattered light frequencies simultaneously, with each frequency representing a different tissue depth. This allows for nearly instantaneous acquisition of each scan and fewer motion artifacts.1

“With SD-OCT, you can create 3-D analysis of tissue with the ability to register images from visit to visit, which decreases variability and increases sensitivity in detecting progression,” said Dr. Schuman. SD-OCT now allows ophthalmologists to image tissue that could not be accessed before, he said. This new ability to evaluate tissues such as the lamina cribrosa has provided new insights into glaucoma and the likelihood of its progression. SD-OCT is also being used in research to measure retinal blood flow, which appears to correspond well with visual function and the presence of glaucoma, he said.

OCT is particularly beneficial for comprehensive practices, said Dr. Asrani, because it can be used to evaluate both glaucoma and retinal disease. However, just as with other technologies, OCT has its drawbacks, such as confounding from media opacities, said Dr. Asrani. “If the vitreous has too many floaters, even the OCT cannot get a good measure.” He added that if the eye moves during OCT, you won’t have a good baseline image, making it more difficult to monitor the patient over time. Another limitation is that data from the older TD-OCT cannot be transferred to the newer platforms, making it more difficult to determine progression.


Spotting Structural Changes Early

“These technologies allow us to detect disease at an earlier stage or to decide that disease is probably not present in a glaucoma suspect,” said Dr. Schuman.

This is most useful in patients with ocular hypertension when the optic nerve looks healthy, and the clinician is in a quandary about whether to treat, said Dr. Asrani. Imaging can help ensure early treatment in those ocular hypertension patients who have structural loss.

Pediatric patients, who often cannot give reliable visual fields, also stand to benefit. “Even when completely normal, the optic nerves in some young patients may look quite suspicious—appearing hollowed out as a result of myopia or racial differences,” said Dr. Asrani. “But if we can get a normal RNFL scan, that gives us a lot of reassurance.” However, cautioned Dr. Liebmann, there are no approved normative databases that can be applied to pediatric populations.

Optic nerve imaging can also add an extra degree of confidence in other cases, said Dr. Schuman. “For example, you might see an early visual field defect that’s tempting to write off as variability or noise.” But an abnormality on the patient’s imaging tests suggests a structure-function correspondence, allowing for earlier diagnosis. Dr. Liebmann agreed. “Given the variability in visual field testing for many patients, the presence of a structural abnormality obtained by computerized imaging devices will affect the physician’s decision making.”


Watching for Progression

“In terms of detection of glaucoma progression,” said Dr. Schuman, “we are just scratching the surface.” One challenge is that there is very little congruency among the technologies: Each identifies different groups of patients as progressing. “This diminishes our ability to detect progression earlier than we can with visual field testing,” said Dr. Schuman. “Overall, the data for tracking progression, especially in comparison to visual fields, are still lacking. And, in many patients, functional loss can precede evidence of structural changes in the optic disc.”2

Once again, however, structure-function correlation can be of great value, he said. “If you have an area of RNFL or optic disc or neuroretinal rim that’s getting thinner, and it corresponds to an area of the visual field that appears to be changing, you have pretty clear evidence—even without doing repeat visual fields—that something is happening and the patient needs an intervention.”

Dr. Schuman relies on statistically significant structural change derived from imaging as the sole indicator of needed treatment when the next step is a low-intensity intervention, such as adding or switching medications. “But if the next intervention is the operating room, I still require progression on the visual field.”

To improve detection of progression, imaging companies are all working diligently on software, said Dr. Liebmann. “Software development is the next major hurdle.”


Other Pearls for Clinical Practice

Drs. Asrani, Liebmann and Schuman offered tips for making effective use of these technologies in practice.

Don’t rush to judgment. Dr. Liebmann reminded colleagues that although a high-risk patient needs to be evaluated more carefully than a low-risk individual, glaucoma usually progresses slowly. “If a person has a 30-year disease, you don’t have to make all the decisions in the first month,” he said. Further, even with advanced technologies available, “The value of gathering information about the range of IOP, baseline optic nerve anatomy via traditional stereophotography and/or imaging, and reliable visual field information cannot be overestimated.”

Ensure image quality up front. To improve your chances of getting a good image, confirm that the patient has good fixation, and encourage frequent blinking before the scan.3 “Also assess the quality of the scan before trusting the results,” said Dr. Schuman. That means ensuring good signal strength or other quality parameters.

“Check the thumbnail images to make sure that the quality is good and that the algorithm segmenting the tissue has done so properly.” If image quality is poor or the software has failed to accurately identify the RNFL margins, for example, don’t trust those numbers, he said.

Be consistent. “There’s little value in switching from one technology to the next,” said Dr. Liebmann. “It’s better to acquire good information many times from one type of device than to use a variety of devices and collect conflicting information.”

Share results with the patient. Show your patients what a normal result looks like compared with their own. This helps to improve adherence to treatment, said Dr. Asrani, because it allows patients to perceive an objective measure of loss despite an absence of symptoms. Many patients discount visual field results because they are subjective, but pay better attention to imaging printouts.

Consider imaging in context. “Interpret results based on what you are seeing clinically, not the other way around,” said Dr. Asrani. Taking machine measurements at face value is not foolproof. For example, highly myopic patients may appear abnormal on OCT because they aren’t included in the normative database, he said.

It’s also important to remember that other diseases can have a confounding effect. For example, said Dr. Schuman, evaluating the macula for glaucoma progression may be less beneficial in a patient with macular edema or diabetes.

Look at all available information. “Technology can aid in decision making,” said Dr. Liebmann. “Don’t let it make all the decisions.”

Dr. Asrani has received lecture honoraria from Heidelberg Engineering. Dr. Liebmann is a consultant for or has received research support from Carl Zeiss Meditec, Heidelberg Engineering, Optovue and Topcon. Dr. Schuman receives royalties for intellectual property licensed by Massachusetts Institute of Technology to Carl Zeiss Meditec.

1 Townsend, K. A. et al. Br J Ophthalmol 2009;93(2):139–143.
2 De Moraes, C. G. et al. Expert Rev Ophthalmol 2010;5(4):451–462.
3 Asrani, S. et al. Glaucoma 2010;19(4):237–242


Learn More in Orlando

You’ll have opportunities aplenty to expand your knowledge of the latest in glaucoma imaging technologies in Orlando.

Whether you want to gain hands-on experience with these devices or have an informal, interactive discussion over coffee, consider the following events:

  • Computerized Scanning Imaging of the Optic Nerve and Retinal Nerve Fiber Layer (Sunday, Oct. 23, 10:15 a.m. to 12:30 p.m.). You can take this as an informative, stand-alone instruction course (168). Or, if you want to go deeper into the topic, you can use it as the prerequisite (Lec168) for the Skills Transfer lab later that afternoon (1 to 3 p.m.; Lab168a) or on Monday (noon to 2 p.m.; Lab168b).
  • Ocular Imaging: Can We Improve Glaucoma Diagnosis and Management With These Devices? (Monday, Oct. 24, 7:30 to 8:30 a.m.; Breakfast With the Experts B272).
  • Identification of the Preperimetric Glaucoma Patient (Monday, Oct. 24, 2 to 3 p.m.; course 387).

And if you’re registered for any of the Subspecialty Day meetings, be sure to stop in at the Glaucoma meeting, too.


Academy members: login to read or make comments on this article.
About Us Academy Jobs Privacy Policy Contact Us Terms of Service Medical Disclaimer Site Index