Ophthalmologists confidently use visual fields and stereo disc photographs to assess disease progression in glaucoma patients, but would love to remove the subjectivity from the equation and inject some objectivity. To this end, researchers have been hard at work refining glaucoma imaging technology to try to increase its speed, resolution and reproducibility. Here’s a look at the currently available technologies’ role in tracking progression.
Why Imaging?
Many hope that if imaging’s resolution and reproducibility were advanced enough it might be akin to seeing a flashing red arrow that pointed unerringly toward progression in glaucoma.
“The major strength of imaging for tracking progression is objectivity and reproducibility,” says Christopher Leung, MD, associate professor in the Department of Ophthalmology and Visual Sciences at the Chinese University of Hong Kong. “Relying on optic disc photos to evaluate structural change over time in glaucoma is challenging because photographic assessment is subjective and, more important, it’s not reproducible. There’s a large inter-observer variability in optic disc assessment in terms of diagnosis and tracking progression.
“There are three commercially available imaging modalities for optic disc and retinal nerve fiber layer assessment: confocal scanning laser ophthalmoscopy; scanning laser polarimetry and optical coherence tomography,” Dr. Leung continues. “These methods provide reliable measurements of optic disc and retinal nerve fiber layer parameters. Using them, we’ll be in a better position to judge if these structures are changing over time.”
The other goal with imaging is to find a correlation between it and functional changes on visual fields. “The dilemma right now is tracking functional vs. structural progression,” says Shan Lin, MD, associate professor of clinical ophthalmology at the University of California, San Francisco. “There are papers that show the two may correlate, but for just some of the time.1Often, one will suggest progression but the other one won’t. The hope is that the newer devices, mainly the higher-resolution Fourier-domain OCT or spectral-domain OCT, will be able to detect progression a little better than older devices such as time-domain OCT or confocal scanning lasers. Whether they actually correlate with functional progression remains to be seen.”
Confocal Scanning Ophthalmoscopy
The Heidelberg Retinal Tomo-graph is the imaging modality that’s been around the longest. The current version is the HRT-3.
According to Heidelberg, in the HRT-3, a laser scans the retina in 24 ms sequential scans, starting above the retinal surface, then capturing parallel images at increasing depths. The stacks of images can then be combined to create a 3D topographic image of the retina. The system aligns the images and compares them using software called TruTrack for both individual examinations and for detecting progression over time. TruTrack is designed to get the scans to be oriented and registered as closely as possible to each other in order to eliminate differences that could result from eye movement or patient positioning.
The HRT-3 uses a module called Topographic Change Analysis that employs TruTrack to look for differences between sequential scans from different visits.
“The HRT is designed to measure the optic disc, basically measuring the neuroretinal rim and the optic disc area,” says Dr. Leung. “The HRT’s Topographic Change Analysis is the most widely known module for tracking change from glaucoma. It provides a topographic analysis of the change in surface topology of the optic disc. It’s basically an event-based analysis: The algorithm collects a baseline and then follow-up exams to compute the baseline variability, since we’re aware that each instrument has some variability in its measurements. If the pooled variability of all the baseline and follow-up examinations is significantly bigger than the baseline variability, then the analysis will show you where the change is on the optic disc surface topology map.”
Last year, a study compared the TCA to optic disc photography in tracking progression in 44 pairs of eyes over a mean follow-up of nine years, with 18 HRT images for each pair.
Ultimately, the study found that TCA performed at least as well as either an individual observer or the best combination of observer classifications of disc photographs.2
Scanning Laser Polarimetry
Where the HRT focuses on the optic disc, the GDx from Carl Zeiss Meditec focuses on the retinal nerve fiber layer, using a laser to measure its thickness.
In an attempt to track the progression of glaucoma over time, the GDx has Guided Progression Analysis, which is similar to the GPA used in the company’s Humphrey Visual Field Analyzer. GPA allows the user to compare each patient’s exam-to-exam variability with the nerve fiber layer variability of a widespread population. It then identifies statistically significant, reproducible changes in focal, regional and diffuse global defects, and presents the results in a single-page report. It also has a corneal compensator to overcome issues regarding the cornea interfering with RNFL readings. There is some data that show this corneal compensator helps give better results.3
A recent study evaluated the GDx’s GPA software’s ability to detect progression, comparing it to the GPA of standard automated perimetry and masked assessment of optic disc stereophotos by expert graders. The researchers found that the GDx detected glaucoma progression in a significant number of cases (50 percent) that showed progression by the conventional methods and had high specificity and positive likelihood ratios for detection of progression. They concluded that the instrument might be useful to complement clinical evaluation in the detection of longitudinal change in glaucoma.4
Optical Coherence Tomography
Like their colleagues in retina and the anterior segment, glaucoma specialists are excited about the potential of optical coherence tomography, both in terms of diagnosis of the disease and tracking its progression.
The first OCT devices were known as time-domain, and provided good images that many feel helped evaluate the condition of ocular structures. The latest generation of OCT devices acquire their images through spectral-domain principles, which users and companies say result in better image quality and acquisition ability. Spectral domain actually gives the user more data from a larger region in a shorter period of time.
The latest spectral-domain OCT devices usually have features to aid physicians in tracking progression. Here are some of these devices and their progression features:
• Heidelberg Spectralis TruTrack.Taking a cue from the HRT, the Spectralis TruTrack uses anatomical features, such as blood vessels, to align the images, and discards any scans that are of low quality. Physicians also say the speed of the device—it takes 40,000 scans per second—helps avoid the negative impact of movement on the scan. It also allows the user to center future scans on the same spot as previous ones, to help track progression.
• Carl Zeiss Meditec Cirrus Guided Progression Analysis. The Cirrus HD-OCT’s GPA compares the measurements of retinal nerve fiber layer thickness over time to try to determine if statistically significant change has occurred. GPA’s features include the ability to view eight maps at once as well as a color-coded progression analysis.
With color-coded analysis, the user can compare up to six maps with two baseline maps. Areas in yellow denote the first instance of statistically significant change in that location, and they become red when the change persists over several visits. It can also plot the retinal nerve fiber layer thickness over time and indicate when a rate of RNFL change reaches statistical significance.
• Optovue RTVue FD-OCT’s Asymmetry Analysis. In addition to a high speed (26,000 scans per second), the RTVue can also provide users with a printout of RNFL thickness that shows asymmetry or change over a central area. It also uses vessel registration to help ensure that subsequent scans are registered to previous ones, so any change will stand out to an observer.
• Topcon 3D OCT-2000’s Trend Analysis. This feature compares disc topography and RNFL thickness from a minimum of two visits. For the disc, it provides cup/disc ratio, cup volume and cup area. For the RNFL, it gives the average thickness of inferior RNFL, superior RNFL and the average of the total RNFL.
Proponents of spectral domain OCT say its improvements over the previous generation of OCT technology in the areas of resolution, reproducibility of measurements and speed of image acquisition align well with the goal of tracking progression.
“We’re deriving a benefit from the greater resolution of these [spectral domain OCT] devices so that we can potentially detect a little bit of loss earlier on. And with a tighter reproducibility of the test, we may be able to detect these changes a little bit better than we did in the past with imaging,” says Dr. Lin.
“The OCT test/retest variability has been shown to be very tight,” says associate medical director of the Bascom Palmer Eye Institute in Miami, Don Budenz, MD, MPH.
“For Stratus OCT, we’ve done several studies that show the test/retest variability of the machine is about 8 µm.5,6 Meaning that if you have thinning of the nerve fiber layer beyond 8 µm, that’s beyond the test/retest variability in 95 percent of patients, and we can say with a 95-percent certainty that this looks like worsening glaucoma. With Cirrus, or spectral-domain OCT, the test/retest variability is even less, around 4 or 5 µm.7 So, since these instruments accurately measure an important aspect of the disease, and they do it reproducibly, we can start to use them to follow glaucoma progression.”
Blanca Monsalve, MD, of Gregorio Marańon University Hospital in Madrid, says that, in a study of 95 consecutive glaucoma patients imaged with the Spectralis, her group found that the intraclass correlation coefficient—a measure of the reliability of the data—was 98 percent in most cases. “This is the best we’ve gotten,” she says. (Monsalve B, et al. IOVS 2010;51:ARVO E-Abstract 242)
Dr. Leung says the area encompassed by the scans is helpful, as well. “For earlier, time-domain OCT, the retinal nerve fiber layer is measured based on a 3.4-mm circle centered on the optic disc,” he explains. “But with spectral-domain OCT, since it’s faster and has a higher resolution, it can generate a map of the retinal nerve fiber layer. In a sense, we’re able to track not only the measurements from a circle scan, but the whole peripapillary region.”
However, the physicians who are excited about its potential say they’ll also have to be patient, since SD-OCT has only been on the market for about a year and more time will be needed to gauge its ability to track progression, given that glaucoma is a slow-moving disease.
“Right now, there’s really not a good study on progression, because that takes time,” says Dr. Lin. “One year isn’t really enough. At one year, your test could just be showing artifact, so there are issues of reproducibility to keep in mind. The good thing is that reproducibility seems to be very good with spectral-domain OCT.”
Challenges to Imaging
Though speed, resolution and reproducibility are improving in the latest devices, the traditional challenges to getting good, sharp images and accurate diagnoses remain.
Dr. Monsalve says that you have to have an idea of the size of the optic disc when using OCT scanning on a patient. “If the disc is too big, you have to be careful because the machine will magnify the numbers,” she warns. “If it’s too small, it makes the numbers smaller. It doesn’t analyze how big the disc is. You have to keep in mind the size of the disc when you determine if the OCT data are reliable or not.”
In a similar vein, Dr. Leung says a scan’s signal strength can affect the results. “In a past journal article, we showed that RNFL thickness measurement on OCT is closely related to signal strength,”8 says Dr. Leung. “If signal strength is low, image quality is poor, and you actually get thinner RNFL thickness readings. When it’s high, even in the same eye, you tend to get thicker RNFL thicknesses. So, if you don’t take this into account in your evaluation of progression, you may end up with artifacts that aren’t related to progression at all.”
As alluded to earlier, proper registration of scans is key if a clinician wants to compare them over time. “With spectral-domain OCT that uses three-dimensional images, you have an OCT fundus image that’s registered from visit to visit that accounts for the acquisition aspect of the scan,” says Joel Schuman, MD, professor and chair of ophthalmology at the University of Pittsburgh Medical Center, and co-inventor of OCT technology. “There can still be artifact due to eye movement that’s not compensated for, though. So, as a clinician, you need to be careful about how you interpret the image. However, the active tracking in the Spectralis helps reduce the variability.”
Also, the presence of a cataract can affect the results. In a study from Dr. Budenz’s group at Bascom Palmer Eye Institute, researchers used Stratus time-domain OCT to image 45 patients before and after their cataract surgery. Twenty-three of them had glaucoma. The researchers found that the presence of a cataract may decrease the peripapillary RNFL thickness measurements as an artifact of decreased signal strength on OCT. They warn that OCT measurements should be interpreted with caution in glaucomatous eyes with significant cataract, particularly if the signal strength is reduced (less than 6/10).9
A concept known as the floor effect can also come into play in cases of advanced disease. “We know that OCT measurements of the average RNFL thickness rarely go below 40 or 30 µm,” says Dr. Leung. “The algorithms in all OCT software also measure the thickness of retinal blood vessels. This means that you’ll never get a zero measurement because of the presence of the vessels. So, in a patient with advanced glaucoma where the NFL is already 45 µm, it’s very hard to track further changes with OCT due to the floor effect, so I believe the role of imaging tests will be more for detecting progression in the early stages.”
Finally, glaucoma specialists still agree that it’s a good idea to get stereo disc photos at baseline and periodically if change is suspected, since imaging technology changes, but photos, whether digital or film, stay constant. “While cameras change, a photograph is a photograph and one can still get a sense of whether there’s a change in the amount of cupping five, 10 or 20 years later,” says Dr. Budenz. “I’ve dragged stereo photos out from 20 years ago and have reassured people that they haven’t changed. You can’t do that with an imaging device because the hardware changes every five years or so.
“Also, clinicians generally agree that structure changes before measurements on visual fields,” continues Dr. Budenz. “But, in clinical practice, studies have shown that most clinicians are following visual fields. I think we’re missing an opportunity to aid diagnosis if we don’t pay attention to our optic disc photos or get them frequently enough to decide whether people are progressing.”
Ultimately, the final arbiter of imaging’s value will be time. Says Dr. Schuman, “Imaging technology and its modules for tracking progression are a piece of the puzzle that needs to be correlated with other findings. You don’t want to take the measurements of imaging technology as gospel—you want to use them as a tool to help with management.”
1. Chauhan BC, Nicolela MT, Artes PH. Incidence and rates of visual field progression after longitudinally measured optic disc change in glaucoma. Ophthalmology. 2009;116:11:2110-8.
2. Chauhan BC, Hutchison DM, Artes PH, et al. Optic disc progression in glaucoma: Comparison of confocal scanning laser tomography to optic disc photographs in a prospective study. Invest Ophthalmol Vis Sci 2009;50:4:1682-91.
3. Medeiros FA, Zangwill LM, Alencar LM, Sample PA, Weinreb RN. Rates of Progressive Retinal Nerve Fiber Layer Loss in Glaucoma Measured by Scanning Laser Polarimetry. Am J Ophthalmol 2010 Apr 8. [Epub ahead of print]
4. Alencar LM, Zangwill LM, Weinreb RN, Bowd C, Vizzeri G, Sample PA, Susanna R Jr, Medeiros FA. Agreement for detecting glaucoma progression with the GDx guided progression analysis, automated perimetry, and optic disc photography. Ophthalmology 2010;117:3:462-70.
5. Budenz DL, Chang RT, Huang X, et al. Repro-ducibility of retinal nerve fiber thickness measurements using the Stratus OCT in normal and glaucomatous eyes. Invest Ophthalmol Vis Sci 2005;46:2440.
6. Budenz DL, Fredette M-J, Feuer WJ, Anderson DR. Reproducibility of peripapillary retinal nerve fiber thickness measurements with the Stratus OCT in glaucomatous eyes. Ophthalmology 2008;115:661-6.
7. Leung CK, Cheung CY, Weinreb RN, et al. Retinal nerve fiber layer imaging with spectral-domain optical coherence tomography: A variability and diagnostic performance study. Ophthalmology 2009;116:1257-1263.
8. Cheung CY, Leung CK, Lin D, et al. Relationship between retinal nerve fiber layer measurement and signal strength in optical coherence tomography. Ophthalmology 2008;115:8:1347-51.
9. Mwanza JC, Bhorade AM, Sekhon N, et al. Effect of cataract and its removal on signal strength and peripapillary retinal nerve fiber layer OCT Measurements. J Glaucoma 2010 Mar 19. [Epub ahead of print]
|
