SD-OCT: What Can It Offer Glaucoma Management?

Continued evolution may allow for earlier diagnoses.

By Scott D. Smith, MD, MPH

As a glaucoma specialist, the most difficult decision that must be made when managing patients is to decide when to advance treatment from medications alone to some type of surgical intervention. In taking the step to surgery, I know, based on my past experience with patients, that there is a small but real possibility of developing a complication that may have short or long-term effects on the vision. On the other hand, a delay in performing necessary surgery could also result in serious and permanent loss of vision for the patient.

A decision to move from meds to surgery should ideally be based upon strong evidence — and the more compelling the evidence, the better. In an attempt to achieve this level of compelling evidence, ophthalmologists perform a number of tests that measure both functional and structural change, seeking strong confirmation that disease progression is taking place. In addition, we would like to know the rate of this progression, as small changes in the vision of a very elderly patient with early glaucoma are not nearly as concerning as more rapid changes in the vision of a younger patient.

Until recently, most ophthalmologists would agree that confirmation of the occurrence of disease worsening (event analysis) and determining the rate of change from exam to exam (trend analysis) have been more art than science, with decisions often based upon largely subjective test results and the experience of seeing thousands of glaucoma patients over a period of many years. We have made decisions somewhat like Sherlock Holmes would, matching this piece of evidence with that piece of evidence and using deductive reasoning to make our treatment choices.

Figure 1. Characteristic superotemporal and inferotemporal RNFL defects are seen on the RNFL thickness (left) map and correspond to regions of RNFL loss on fundus photography (right). ALL OCT IMAGES COURTESY OF NATHAN RADCLIFFE, MD

What healthcare providers have been seeking for years is an objective test that can take some of the uncertainty out of measuring two key standards for treatment decisions: (1) confirmation that glaucoma progression is taking place and (2) a more refined apprehension of the rate of progression.

Here, I will discuss the increasing value of spectraldomain optical coherence tomography (SD-OCT) and meaningful progression analysis as useful tools in diagnosing glaucoma progression. For me, the availability of an SDOCT instrument with a normative database that enables me to track progression over time represents a major step forward in monitoring glaucoma patients. Having such an instrument that offers objective, numerical evaluation of data has definitely increased my confidence level in the interpretation of the widely used tests for glaucoma progression.

Although my commentary here is based on my experience with the Cirrus SD-OCT (Carl Zeiss Meditec), other SD-OCT devices also provide data analysis tools to improve glaucoma diagnosis. Robust competition in the marketplace will surely yield continual improvements to all platforms in the coming years.

The Limits of Subjective Tests

Tests designed to evaluate glaucoma progression are useful but, unfortunately, they have inherent limitations. The problem with these glaucoma tests can be summed up with a single key question. Are changes in our measurements reflecting random variation or true change due to disease progression?

With a functional test such as a visual field test, for example, many factors can influence patient performance aside from the disease status. The patient could be tired or distracted for any number of reasons. As a result, reliable identification of visual field progression requires repeated confirmatory tests. Even objective measures of medical conditions can suffer from this problem. To use an everyday analogy, we have all had our blood pressure taken many times and I'm sure we have all seen variations in our blood-pressure readings. Our blood pressure reading taken during a relaxing vacation may be quite different than that taken after a long drive in traffic to a crowded doctor's office following a stressful day at work. Which of the readings is the more meaningful, or are either of the readings meaningful at all? And in addition to variation in the value being measured, measurement error itself contributes to the problem.

The detection of structural change through optic nerve and nerve fiber layer imaging brings a greater level of objectivity to glaucoma progression testing. Earlier-generation time-domain OCT was primarily used to measure retinal nerve fiber layer (RNFL) thickness. In this test, a circumpapillary scan of the retina is performed that doesn't allow precise alignment on repeat testing, which leads to variability in the measured nerve fiber layer thickness due to difference in placement of the scan circle. Sources of measurement variability such as this lead to difficulty in discerning measurement error from true change, in spite of the objective nature of the test.

Figure 2. An RNFL thickness map with neuroretinal rim segmentation (gray) in a glaucoma suspect with a normal visual field. The image demonstrates that although the vertical CD ratio is enlarged at about 0.6, the retinal fiber thickness is normal.

The Advantages of SD-OCT

The higher scan speed of SD-OCT allows acquisition of data from the entire peripapillary retina and optic nerve head. The technological advancements incorporated into SD-OCT that allow this higher scan speed, as well as the higher image resolution, have been shown to reduce test variability. With SD-OCT, superimposition of sequential images also reduces measurement error, and improves the ability to detect and track glaucoma progression or change over time. Essentially, SD-OCT takes a great deal of measurement error out of the equation and provides ophthalmologists with the opportunity to identify real change with a higher level of confidence.

Even with the reduced test variability of SD-OCT, however, the question of how much change represents a real change in the patient's disease status still remains. The incorporation of statistical software into SD-OCT that can measure disease progression or change over time provides direction to the clinician as to whether a meaningful change in the patient's status has occurred, along with information about the rate of change of different SD-OCT parameters. I feel that this represents another important step forward in the value of optic nerve and nerve fiber layer imaging for glaucoma diagnosis, as it allows us to make more rational use of the information provided by SD-OCT.

For example, studies of SD-OCT variability have shown that parameters such as the mean nerve fiber layer thickness and quadrant nerve fiber layer thickness values can vary by as much as 6% to 10% in people with stable glaucoma. Rather than using these estimates as a sort of off-the-cuff estimate of whether glaucoma progression has occurred, software that performs serial comparisons between exams allows the determination of statistically significant change within an individual patient based upon the whole series of measurements over time. It also provides an estimate of the rate of change of these parameters, as well as the presence of a statistically significant trend over time. This can give a more solid foundation to determining whether — given the rate of change, the age and the severity of glaucoma damage — a patient is at risk for visual impairment in his or her lifetime. While a degree of uncertainty always remains in diagnosing glaucoma progression in any individual case, SD-OCT testing can be very helpful in minimizing the level of that uncertainty.

Is SD-OCT the Game Changer?

I have been asked whether the value of SD-OCT in accurately detecting glaucoma progression is such that a glaucoma fellow using SD-OCT can make a more accurate diagnosis of progression than a glaucoma specialist with 20 years of experience and having access to all tests except SDOCT. My answer to that question is that having SD-OCT provides significant advantages to the clinician but that the data derived from SD-OCT testing alone cannot override all of the data from the other common glaucoma tests such as tonometry, visual field testing and clinical examination of the optic nerve.

Figure 3. HD-OCT of the RNFL and neuroretinal rim in a patient with optic disc drusen and an inferior visual field defect. Note the superior RNFL loss seen on the RNFL thickness and deviation maps in the presence of a greater than average neuroretinal rim thickness.

We still need to look at the overall picture in making treatment decisions. However, if I get a strong indication of disease worsening from SD-OCT testing, it makes me look very carefully at the other measures to try to identify correlation between the SD-OCT and the other tests. It is, in fact, discovery of these correlations that allows the identification of subtle glaucoma progression that could otherwise go undetected.

At this point, SD-OCT cannot yet be called a true revolution in glaucoma-progression testing, but it certainly qualifies as an important step forward in this area. SD-OCT may become an even more useful tool with further improvement in image resolution and more robust normative data, such as, for example, a refractive database that could account for retinal nerve fiber layer changes in patients with myopia. OCT devices also offer imaging of the anterior segment angle, can capture pachymetry data and can image the ganglion cell complex, among other capabilities. As with RNFL, these measures can also be analyzed to identify change over time.

With continued advances on numerous fronts, SDOCT imaging will play an ever-increasing role in the diagnosis of glaucoma. OM