NEW TECHNOLOGY: GDx VCC

ONE OF THE EARLIEST clinically identifiable changes occurring in glaucoma is damage to the nerve fiber layer (NFL). Nerve fiber layer damage typically precedes visual field loss, often by several years. Thus, a careful examination of the NFL is essential to identify early patients who have developed glaucoma. Focal NFL defects, particularly when they are juxtaposed against healthy NFL, are easily detected ophthalmoscopically. However, diffuse or subtle NFL damage is more difficult to appreciate. The need to objectively and accurately quantify the NFL and differentiate normal patients from those with glaucoma led to the development of the GDx Nerve Fiber Layer Analyzer.

This device utilizes scanning laser polarimetry (SLP) to measure the thickness of the NFL for comparison against a normative database. The principles of SLP utilize birefringence and retardation. Briefly, the NFL is birefringent due to the parallel arrangement of microtubules within NFL axons. Birefringence changes the polarization of the incident light into orthogonal planes. The time delay between the return of these fast and slow axes is called retardation. Polarized light passing through a birefringent medium (NFL) is split into two rays. Rays of laser light that travel perpendicular to fibers passes through at a speed different from those rays passing through in parallel. The phase shift results in retardation and provides the basis for which to measure the thickness of the NFL.

Unfortunately, the cornea and, to a lesser extent, the lens are also birefringent structures that essentially can contribute to signal noise that can artifactually alter the analysis, much the same way a cataract can affect the results of a visual field. Early iterations of SLP utilized a fixed corneal compensator that assumed a set corneal polarization magnitude and axis. While this fixed corneal compensator corrected for anterior segment aberration in the vast majority of patients, there were patients whose corneal polarization magnitude and axis did not conform to that of the general population, and their anterior segment birefringence was not eliminated.¹ This led to artifacts in the analysis with the true retrolenticular birefringence not being accurately measured.

To address this problem, variable corneal compensation (VCC) has been incorporated into the scanning laser polarimeter (GDx VCC). This device now measures the corneal polarization axis and magnitude in all eyes, then sets the device to cancel individual corneal birefringence so that the true retrolenticular (retinal NFL) birefringence is accurately measured. The method that the GDx VCC employs is quite simple. For the initial exam on all patients, a macular scan is first obtained. Birefringence around the fovea is known to be uniform due to the absence of ganglion cells, and hence NFL in this region. A non-uniform pattern at the fovea must be caused by the cornea, and an hourglass pattern indicates the axis and magnitude of the corneal birefringence. When the variable compensator is set to cancel the corneal birefringence, the hourglass disappears, indicating successful corneal compensation.²

The goal of scanning laser polarimetry is to objectively and accurately discriminate between normal patients and those with glaucoma. Previously reported poor diagnostic precision of SLP with a fixed corneal compensator was likely attributable to inflated intersubject variability resulting from ineffective corneal birefringence compensation. New research using variable corneal compensation has demonstrated success at differentiating normal patients from those with glaucoma. The variable corneal birefringence compensation has significantly narrowed the range of normals and much more accurately allows for differentiation between those with and without glaucoma. The sensitivity of this discrimination has increased greatly due to the removal of artifact stemming from improperly compensated corneal polarization axis and magnitude.^3-5 It appears that the GDx VCC accurately measures the true retrolenticular birefringence and provides a reproducible measurement of the thickness of the NFL. The VCC now allows for accurate measurement of the NFL, which correlates well with both structure as measured by optical coherence tomography⁶ and function as measured by SITA perimetry.⁷ A normative database, which also represents patients of African descent as well as other races prone to glaucoma, allows for a statistical comparison of the NFL.

NEW TECHNOLOGY: CONFOCAL SCANNING LASER TOMOGRAPHY

WHILE CONFOCAL SCANNING LASER TOMOGRAPHY is not exactly a "new" technology, its widespread use in clinical practice today merits special mentioning. The Heidelberg Retinal Tomograph II (HRT II) offers a combined approach for detection of both glaucoma and macular edema.

For glaucoma detection, the HRT II obtains a series of optical section images of the optic disc and peripapillary retina at several depths. This series of images is combined layer-by-layer into a three-dimensional image remarkably similar to a CT scan. The HRT II gives both a pictorial image of the disc and cup as well as numerical values of several disc parameters. Data obtained from a patient is compared against a normative database in the Moorfields Regression Analysis. The Moorfields Regression Analysis classifies eyes based upon the relationship between the cup and rim area, among other parameters. The fact that the HRT II linear regression analysis takes into account the overall disc area makes the device more sensitive than clinical assessment of stereoscopic optic disc photographs in distinguishing between healthy patients and those with early glaucoma.¹

Confidence intervals are set at 95% and 99.9%. That is, if the data obtained on a patient shows that the area of the neuroretinal rim is larger than or equal to that found in 95% of the patients in the normative database, the patient is classified as "within normal limits." If the area of the neuroretinal rim for the patient falls between 95% and 99.9% of the area for the neuroretinal rim of patients in the normative database, the patient is classified as "borderline." If the patient's obtained data indicates that 99.9% of the patients in the normative database have more area to the neuroretinal rim, the patient is labeled as "outside normal limits." These analyses are performed for several areas of the disc. Following a patient for change over time can be done either with the stereometric parameters or by using the Progression and Change Probability program, which is the preferred method.

After an image is obtained, the operator must manually plot the edge of the disc with a "contour line." This can be seen as a source of introduced error, as the analysis is highly dependent upon the correct determination of the disc contour. Further, the delineation between the cup and rim is accomplished through a reference plane. The average thickness of the papillo-macular bundle located at 350º to 356º has been found to be, on average, 50µm. This is the bundle that remains intact through disease progression. The separation of rim from cup is set 50µm below the average thickness of this area. Essentially, everything above the reference plan is rim, and everything below is cup. This has been argued to be another source of introduced error.

However, importing the reference plane and contour line into successive analyses compensates for any introduced error. The Progression and Change Probability program negates any sources of introduced error and remains a powerful feature of the HRT II. Further, the Progression and Change Probability program can operate without the operator ever defining a contour line. Built-in software automatically, objectively compares subsequent patient exams against the original baseline exam and identifies significant changes. Measurement of disc stereometric parameters is highly reproducible.² There are sensitive parameters to indicate the quality of captured images and follow-up exams that are unfocused by more than 1.00D are easily identified. Thus, the HRT II remains an outstanding device to measure changes in disc parameters over time.³ Further, the HRT II can meet or exceed the disc topographic change detection performance of even the most experienced clinicians.⁴ The ability to detect subtle change in the disc topography over time is critical because this pathological alteration frequently occurs prior to the onset of visual field progression and may be one of the first indications of glaucomatous deterioration.^5,6 The ability to detect change over time remains the greatest strength of the HRT II.

The HRT II also provides an objective way to measure retinal edema. The macular edema module of the HRT II analyzes signal width. It then takes into account signal width and retinal reflectance at each retinal location obtained in the optic slices to generate an index that is sensitive to the presence of retinal edema.

NEW INSIGHT ON TREATING GLAUCOMA

WHILE GLAUCOMA HAS LONG BEEN considered a pressure- sensitive optic neuropathy, only recently has the benefit of intraocular pressure reduction been scientifically proven.^1-5 Recently, a study has been published that may be the only study to include an untreated study arm of patients known to have the disease. The Early Manifest Glaucoma Trial (EMGT) was a controlled clinical trial that evaluated the effectiveness of reducing IOP in patients with newly detected, previously untreated, early glaucoma.⁶ Results from this landmark study have shed light on managing this condition.

The EMGT randomized 255 patients, aged 50 to 80 years with early stage glaucoma in at least one eye, into either a treated arm or an observational arm of the study. Patients were detected through population-based screenings at two cities in Sweden and were included if they had a new diagnosis of early manifest primary open-angle glaucoma (POAG), normal-tension glaucoma, or pseudo-exfoliative glaucoma. Patients were excluded if they had advanced visual field defects, poor acuity in either eye, mean IOP greater than 30mm Hg or any IOP greater than 35mm Hg, or any condition that would interfere with serial observation for progression. Thus, the patients in the study truly had early glaucoma only.

Patients in the treated arm received standard therapy consisting of betaxolol and argon laser trabeculoplasty. While there was not a target pressure or desired IOP reduction, the standard treatment lowered IOP by 25%, which was maintained throughout the study. The patients in the observational arm were followed closely with very sensitive indicators designed to detect progression. While some wondered whether it was ethical to not treat patients with glaucomatous damage, the study was designed carefully so that any untreated patient who demonstrated progression was immediately offered treatment. Thus, no patient was allowed to lose a significant amount of visual field while undergoing observation.

The results were somewhat surprising. After fours years of follow-up, 30% of patients in the treated arm demonstrated progression, while 49% of the untreated patients progressed. After six years of follow-up, 45% of the patients in the treated group had progressed, compared to 62% of the untreated patients. While this clearly demonstrated a benefit of treatment, it was apparent that glaucoma progressed in a large percentage of treated patients. The time that it took for glaucoma to progress varied greatly among patients and was sometimes rather short, even in the treated group. However, many patients remained stable throughout the course of the study, even in the untreated group. Results of this study showed that each mm Hg of pressure reduction conferred approximately a 10% reduction of risk for glaucoma progression. This may force doctors to reevaluate what constitutes a "clinically significant" pressure reduction, as small reductions in IOP may be critical.

The EMGT clearly shows that treatment for newly diagnosed early glaucoma should be individualized and tailored to the needs of the patient. One option could include no initial treatment, but observation with treatment commencing if progression is seen. The decision not to treat patients in the early stage may postpone or obviate the need for medications.

The results of this study must be viewed in context with the limitations of this study. Virtually all patients were Caucasian. Thus, it may not be prudent to apply the results of this study to patients of African descent, who may have an entirely different form of the disease. Over half of the patients had normal-tension glaucoma, which has been shown to be largely a non-progressive disease.⁵ Also, there is no data on patients with very high IOP or advanced visual field loss, and there is no long-term follow-up of patients beyond EMGT progression.