Diagnosing open-angle glaucoma may be particularly challenging when damage is mild or early. Although most clinicians will not have difficulty detecting glaucomatous damage when both functional and structural defects are evident, patients presenting with a normal visual field combined with optic nerve and/or retinal nerve fiber layer (RNFL) damage may make practitioners stop and think. The physician must decide if the optic nerve's or RNFL's appearance is a result of glaucoma or represents a normal variation. Because this scenario arises so often, supplemental testing such as optic nerve/RNFL imaging or selective perimetry can play a useful role in everyday practice.
The definition of glaucoma has evolved such that optic nerve findings alone now may be considered a sufficient indication for a positive diagnosis.1-3 Explanations for frequently seen disparities between standardized automated perimetry (SAP) findings and optic nerve/RNFL evaluations are complex. Often, these disparities are associated with factors ranging from redundancy in the visual system to the use of log scales in one automated measurement versus linear scales in another.
A series of instruments has been developed to help clinicians recognize glaucomatous damage when obvious loss is not apparent. Automated imaging devices evaluate the optic nerve, RNFL, and macula. They either compare results to age-corrected normative databases or plot results over time to detect threatening rates of change. Specialized perimetric tests have been developed to selectively assess visual function, with the goal of detecting visual field loss not found by SAP. This article describes these selective perimetric tests and their role in the diagnosis of glaucoma.
SHORT-WAVELENGTH
AUTOMATED PERIMETRY
Short-wavelength automated perimetry (SWAP) or
blue-yellow perimetry was commercially introduced
more than 15 years ago. SWAP is available on the
Humphrey perimeter (Carl Zeiss Meditec, Inc., Dublin,
CA) and on Octopus perimeters (Haag-Streit USA Inc.,
Mason, OH). Both instruments also perform SAP, which
can be useful when the clinician suspects glaucoma and
the standard field is within normal limits.
With SWAP, a large Goldmann size V blue target is projected against a bright yellow background. The background reduces the sensitivities of the green and red cones, thus isolating the short-wavelength–sensitive blue cones and their associated small, bistratified retinal ganglion cells. Because only about 10% of retinal ganglion cells are of the bistratified variety, SWAP tests only a small fraction of the visual system. Early theories suggesting that SWAP works because bistratified cells are the first ones damaged in glaucoma are no longer widely accepted. Instead, the most popular theory is that SWAP owes its success to the fact that fewer cells are tested, and a reduced amount of overlap in tested ganglion cell-receptive fields exposes defects early. Defects found with SWAP often appear larger and deeper than with SAP, and many studies have suggested that SWAP can detect visual field loss earlier than SAP (Figure 1).4-8
Of course, all techniques have attendant difficulties. The original SWAP test was normally performed using a tedious full-threshold algorithm. Several years ago, Swedish interactive testing algorithm (SITA) SWAP was introduced, which reduced testing time by half while retaining similar sensitivity and reproducibility. SITA SWAP also possesses an improved dynamic range. Nonetheless, many patients still find the test difficult, perhaps because the visual system tested by SWAP has low resolution and responds slowly. Even with optimal refractive correction, the stimulus usually seems blurry and is not seen to turn on and off crisply. Because patients are unaccustomed to seeing under these conditions, it is hardly surprising that there is often a learning effect when they are introduced to SWAP testing, even among individuals who have long experience with SAP testing. Moreover, testing the blue cone system leads to higher intra- and intertest variability compared with white testing.9,10
SWAP tests require a somewhat different interpretative strategy compared with standard white testing. SWAP's blue stimulus is highly attenuated by yellowed crystalline lenses. In addition, because yellowing of the crystalline lens varies greatly among patients, simple age corrections are not sufficient. Thus, the analysis of SWAP results relies mostly on metrics that self-correct for media effects (the familiar pattern deviation probability plots, the glaucoma hemifield test, and the pattern standard deviation index) and ignores raw threshold sensitivities and metrics that simply apply an age correction (eg, the grayscale, mean deviation, and the total deviation probability plots). In patients who have excessively yellow lenses, lenticular attenuation of the stimulus may be so strong that there is not enough brightness range to fully determine the depth of very deep scotomata, further complicating the evaluation of tests.
Given patients' frequently disquieting experience, increased testing variability, reduced testing range, and the need for a somewhat different strategy for interpreting the test, most practitioners seldom use SWAP. Overall, these problems do not appear to be correctable, and it is doubtful that, after all these years, SWAP will become a widely used clinical test.
FREQUENCY DOUBLING TECHNOLOGY
Frequency doubling technology (FDT) perimetry was
commercially introduced in 1997, a few years after
SWAP. FDT has been reported to selectively test the
sensitivity of the magnocellular portion of the visual
system, which serves a different subset of visual functions
than SWAP. One difference between FDT and
SWAP is that the former can only be performed on a
stand-alone instrument. A significant paradigm shift
would be needed for it to be replaced by FDT, because
SAP is the primary perimetric test for most clinicians.
FDT has been marketed primarily to the optometric
community, where it is broadly used in clinical case
detection.
FDT perimetry uses a low-spatial-frequency sinusoidal grating target that undergoes high-temporalfrequency counter-phase flicker. FDT is a flicker-type test in which patients are asked to respond when they notice a shimmering or flickering stimulus. The test measures the central 30° of the field of vision, with the original instrument presenting large, 10° X 10° peripheral stimuli and a 5° X 5° macular stimulus. Only 17 to 19 test point locations are evaluated with the original FDT perimeter, depending upon the testing pattern used.
The Humphrey Matrix FDT perimeter (Carl Zeiss Meditec, Inc.) is the most recent FDT tester and presents 50 grating stimuli throughout the central field. In addition to the original FDT suprathreshold tests, the Humphrey Matrix also offers 10-2, 24-2, and 30-2 threshold testing. Thus, the instrument is intended for use both in clinical case detection and in glaucoma diagnosis.
Like SWAP, SAP, and all other testing modalities, Humphrey Matrix testing is not without its challenges. Patients frequently report that the testing screen seems to fade during the test (Troxler phenomenon), requiring them to blink to bring it back. Also, the threshold testing points often seem to be sporadically flagged as being outside normal limits. The latter problem may relate to the limited number of steps in stimulus strength that are used during threshold testing. On the Humphrey perimeter, stimulus strength is varied in 1-dB steps over the full range of vision from 0 dB to about 40 dB. Thus, each decibel level is possible such that threshold scores may be, for example, 32, 31, 30, 29, 28, etc. With the Humphrey Matrix, 13 available stimulus contrasts are arranged in uneven intervals across the full testing range. Because steps are generally larger in the normal sensitivity range and smaller in scotomata, it is not uncommon for normal or nearly normal testing locations to be variably flagged as outside normal limits simply because of normal testing variability. The algorithm used is called zest, and it is conceptually similar to SITA in that it is quick and uses forecasting principles. As with SWAP, a number of articles have suggested that FDT field defects often precede SAP defects (Figure 1).11,12
THE HEIDELBERG EDGE PERIMETER
The new Heidelberg Edge Perimeter (HEP; Heidelberg
Engineering GmbH, Heidelberg, Germany) also selectively
tests retinal ganglion cells, in this case, the magnocellular
system. The HEP is available in many countries,
and US regulatory clearance is pending. The technology
is based on the concept of flicker-defined form,
in which a high-temporal-frequency stimulus undergoes
counter-phase flicker leading to a phantom contour
illusion. The objective is to recognize early glaucomatous
damage, and the instrument is similar to
perimeters with SWAP in that it can also perform SAP
testing (Figure 2). With the HEP, a flickering black-andwhite
patch creates an illusory edge due to differences
in flicker phase between the stimulus and the background;
patients perceive a circular stimulus. An adaptive
staircase thresholding algorithm makes testing
times comparable to those with other algorithms.
Further studies are needed to understand the HEP's
ability to recognize early loss, but the introduction of a
new perimeter is exciting.
CONCLUSION
When a patient presents with findings that may indicate
early glaucomatous damage but full SAP fields
(and not a complete burden of proof), confirming
damage with a selective perimetric test may be useful.
SWAP has been clinically disappointing, and FDT is
often confusing. HEP is the latest test to emerge and shows potential. It will take time, however, to understand
if this test will be a useful addition to clinicians'
armamentaria.
Murray Fingeret, OD, is a clinical professor at the SUNY College of Optometry and chief of the Optometry Section, Department of Veterans Affairs, New York Harbor Health Care System, Brooklyn Campus. He is a consultant to and sits on the advisory board of Carl Zeiss Meditec, Inc., and he has received research support from Heidelberg Engineering, Inc. Dr. Fingeret may be reached at (718) 298- 8498; murrayf@optonline.net.
