Main topic: Functional Assessment in Glaucomaglaucomanow.com/media/15319/glaucomanow_issue1_2012.pdf · Glaucoma Now Issue No 1 2012. 1 issue 1 2012 Main topic: Functional Assessment

Glaucoma Now – Issue No 1, 2012. www.glaucomanow.com 1

i s s u e 1

2 0 1 2

Main topic:

Functional Assessment in Glaucoma

TARGET AUDIENCEThis educational activity is aimed at general ophthalmologists, glaucoma specialists and ophthalmo logy residents.

Glaucoma Now is published and adminis-tered by the editorial board and supported by an unrestricted educational grant from Allergan, inc. Copyright 2010 Editorial Board. All rights reserved. No responsibility as-sumed for injury or damage to persons or property arising from the use of information or ideas contained in this publication.

LEARNING OBJECTIVES• Special Focus: An overview of currently available testing technologies with focus on optimal test methods and technologies to detect glaucoma damage and progression.

• What’s New: A comprehensive summary discussing the possible approaches such as pattern electroretinogram (PeRG), visually evoked potential (VeP), short wavelength automated perimetry (sWAP) and frequency doubling technology (FDT).

• Clinical Issues: A summary highlighting the best approach to decide on what tests and strategies to use on which patient at a particular point in time.

• Practical Tips: An in-depth discussion on how to handle or minimize artifacts. Tips on how to avoid typical pitfalls in interpretation of data available.

Special Focus:

Optimal testing, techno-logies to detect glaucoma damage, technologies to detect progression

PAGe 2

Clinical Issues: What tests and strategies to use for whom and when?

PAGe 9

What’s New:

Where are we going with approaches such as PeRG, VeP, sWAP, FDT?

PAGe 6

Practical Tips:

Minimizing artifacts and avoiding pitfalls in interpretation

PAGe 10

inside this issue:

EDITORIAL BOARDIvan Goldberg MD, FRANZCO, FRACs sydney eye Hospital, sydney, NsW, Australia.

Remo Susanna MD, Professor and Head of Department of Ophthalmology, university of são Paulo, Brazil.

Glaucoma Now is a continuing medical education publication.Distributed worldwide, our goal is to educate and update general ophthalmologists, glaucoma specialists and ophthalmology resi-dents.

international leaders in the field of glaucoma are invited to contribute to this journal, sharing their most recent insights.

supported by an unrestricted educational grant, the publication is non-promotional and has a fully independent editorial Board.Glaucoma Now is in the process of gaining accreditation for Continuing Medical education (CMe) by the european Accredi-tation Council for Continuing Medical Accreditation.

Executive officer: Patricia Buchholz RPh, PhD Karlsruhe, Germany [email protected] Production by Phosworks www.phosworks.com

2 Glaucoma Now – Issue No 1, 2012. www.glaucomanow.com

special Focus:

Optimal testing, technologies to detect glaucoma damage and progressionLinda Zangwill PhD, Christopher Bowd, PhDuniversity of California san Diego, Hamilton Glaucoma Center and the shirley eye Center, California usA

Technologies to Detect Glauco-matous Visual Field DamageThere are several ways to identify glauco-matous defects from standard automated perimetry (SAP) print outs (Figure 1). Firstly, inspect sensitivity thresholds at each test point. It is not particularly useful to identify defects, because these values are not age-corrected. In older patients low thresholds could be attrib-

depression that could be caused by other pathologies (e.g. cataract). So in the PD plot, localized defects most likely from glaucoma are highlighted. Localized patterns of abnormality, identified on the PD probability plot, can be assessed to determine if they are what is consid-ered characteristic for glaucoma (e.g. na-sal step, superior or inferior arcuate pat-terns of defect). Odd-looking patterns

Figure 1 shows all of the information available on a standard automated perimetry print out.

utable entirely to age. More meaning-ful information is available in the total deviation (TD) and pattern deviation (PD) probability plots (Figure 2). The TD probability plot shows the point-by-point probability of abnormality (in grey scale) compared with an age corrected normative database. The PD probabil-ity plot provides similar information adjusted for both age and generalized

Core Concepts • There are 2 global indices of vi-sual field sensitivity: mean deviation (MD) and pattern standard deviation (PsD). Once glaucoma has been di-agnosed MD and PsD provide data on disease severity.• To be useful, tests must be reli-able with few fixation losses and few false positive/false negative re-sponses.• Data on disease severity withmeasured visual field deterioration, along with patient age and health status should be used to estimate likelihood of visual impairment.• Progression detection algorithmsavailable are GPA i and ii.• Frequency of follow-up testingshould be dictated by the relative risk of progression. The higher the risk, the more frequently testing should take place.• TheWorldGlaucomaAssociationConsensus Group on Glaucoma-tous Progression suggests at least 2 reliable fields in the first 6 months after diagnosis and at least 2 more fields within the next 18 months.• Increased testing from baselineup to 2 years identifies more rapidly progressing patients with a lower false positive rate.• When central vision is compro-mised it may be useful to obtain 10-2 visual fields that use a2-degree spacing.


might result from a test-taking artifact or patient inattention. Two global indices of visual field sen-sitivity are available. Mean deviation (MD) is the weighted average of the point-by-point decibel deviations from normal that are shown in the TD plot. Pattern standard deviation (PSD) repre-sents the relative point-by-point decibel deviations from normal across the visual field. PSD increases as the point-by-point deviations from normal across the visual field become more variable. PSD increases in the presence of localized de-fects. Although these global indices are not used for diagnosis, a PSD outside normal limits is likely to be more infor-mative than an MD outside of normal limits because it is more likely to detect local defects earlier. Once glaucoma has been diagnosed, MD and PSD provide a general idea of the severity of the disease. The Glaucoma Hemifield Test1 is a summary parameter that describes the relative difference in sensitivity in five zones of the superior hemifield com-pared with five corresponding zones in the inferior hemifield. Results are either “outside normal limits”, “borderline” or “within normal limits”. This parameter is particularly useful to detect early to moderate disease because glaucoma-tous visual field defects usually manifest asymmetrically across the visual field horizontal midline. The Visual Field Index (VFI) is another summary pa-rameter; it is based on geographically weighted PD probability map values with emphasis on the central test points; it has been reported to represent the percentage of remaining useful vision.2 Theoretically the VFI ranges from 100 in a healthy eye to zero in a seriously damaged eye. There is no suggested ab-normality cut-off for this parameter. Consider other important items to identify glaucomatous defects. Firstly, tests must be reliable, with few fixation losses and few false positive and false negative responses. The latter reliability criterion is less important in individuals with advanced disease because a reported false negative may be a truly unseen test point that is the result of a severe defect. Secondly, observed abnormalities should be repeatable and consecutive abnormal tests should have the same general pat-terns of defect. In the Ocular Hyperten-sion Treatment Study, 86% of 702 ini-tial abnormal SAP tests did not show a

Figure 2 shows the Total Deviation and Pattern Deviation probability plots from an eye with a moderate to severe inferior visual field defect. Global indices are also shown.

Figure 3 shows the Total Deviation and Pattern Deviation probability plots from an eye with a small pupil. This pattern of defect generally is not characteristic of glaucoma.

Figure 4 shows the first incidence of possible and likely progression, based on GPA analysis, in an eye with a developing inferior temporal visual field defect. Half filled triangles represent test points progressed, relative to baseline variability, in two consecutive tests (possible progression). Filled triangles represent progression in three consecutive tests (likely progression).

subsequent abnormal test. In fact, 66% of these eyes had all indices within nor-mal limits on follow-up testing3, thus the importance of repeatable abnormal

results to identify true glaucomatous defects. Finally, visual field results are quite variable and learning effects, ar-tifact caused by drooping eyelids, small


pupils and by inattention can result in different patterns of defects that are not disease related: hence the need to check for similar and recognizable patterns of defect in consecutive tests (Figure 3).

Technologies to Detect Glauco-matous ProgressionGlaucoma management aims to pre-serve visual function for an individual’s lifetime. To this end, clinicians need to monitor closely the magnitude and rate of visual fields changes so that appropri-ate treatment can be initiated to preserve visual function. Information on the rate of visual field deterioration along with the severity of the disease at diagnosis and the patient’s age and general health status (i.e. likely longevity) should be used to estimate the likelihood of visual impairment. Similar to detection of disease using SAP, disease progression also can be assessed subjectively based on informa-tion available in TD and PD probability plots. To do this, a longitudinal series of TD or PD plots can be compared for increases in scotoma size and/or depth or the development of new scotomas. Because of the large variability in SAP results, this information is not always meaningful. Two useful progression detection algorithms currently are avail-able in the Humphrey Statpac software: Guided Progression Analysis (GPA)4 and linear regression of Visual Field In-dex (VFI).2

GPA is a PD plot-based event analy-sis calculation that relies on the com-parison of change at each test location to the variability observed between two baseline tests at the same location (Fig-ure 4). According to the Early Manifest Glaucoma Trial criteria4, “Likely Pro-

gression” is assigned to a visual field if at least three test points are flagged as significantly progressing at the same lo-cation over three consecutive tests (“Pos-sible Progression” is assigned if the same three points have progressed over two consecutive tests). GPA is attractive be-cause it is objective, but it relies on the variability between baseline tests. Be-cause of this, its sensitivity to change can be variable. Also, it does not require the progressing points to be contiguous, so the observed change may not represent the predicted increase in scotoma size that is associated with disease progres-sion.5 Once a GPA result of “Likely Pro-gression” has been assigned, new base-line examinations should be obtained to allow identification of a subsequent pro-gression event. As suggested by the 2011 World Glaucoma Association Consen-sus Group,6 new baseline examinations also should be obtained after any signifi-cant therapeutic intervention. Unlike the GPA, linear regression of the VFI can provide the rate of change over time (i.e. change in VFI value per year) and, because it is derived from PD and not MD, it likely is fairly resistant to the effects of developing/worsening cat-aracts. Knowing the rate of progression is more important for long-term disease management than simply knowing if a patient has progressed or not (Figure 5). This is in part because faster progression is a predictor for future progression and increases the likelihood of visual impair-ment in one’s lifetime.7 Like GPA, VFI is not perfect. For instance, using linear regression to suggest a predicted rate of change over time assumes that change over time is linear; this may not be true.8 It also assumes a stable treatment regi-men and it does not consider intervening

surgical procedures. Finally, assuming change over time is linear estimates of change of VFI over time (i.e. estimates of the negative slope of VFI) will be more accurate when more tests are in-cluded in the regression equation. This is because adding more test points will result in a better (“tighter”) linear fit. Another available progression de-tection technique uses ordinary least squares linear regression and is imple-mented using Progressor® software.9 Based on five or more consecutive tests, Progressor software performs linear regression (like regression of VFI) of threshold sensitivity on a point-by-point basis using all test points in the 24-2 test pattern. Individual points are defined as progressed if the negative slope of the point-wise change in sensitivity meets or exceeds a user-selected cut-off. A global progression event can be assigned (similar to GPA) based on a user-defined global progression criterion (e.g. ≥ 3 of the same test points with a slope ≥ 1.0 dB/year for inner points and ≥ 2.0 dB for edge points, confirmed in three consecu-tive exams10). No matter the technique used for progression detection, as sug-gested by the 2011 World Glaucoma As-sociation Consensus Group statements,6 frequency of follow-up testing should be dictated by the relative risk of significant progression (based on extent of damage, life expectancy and any other relevant clinical observations, such as structural change and IOP control11), with fre-quency increasing with increasing risk.

Optimal TestingSeveral critical clinical questions must be answered to determine how best to “monitor closely” visual function in glaucoma patients, using any or all of the techniques described above. How many visual fields are really needed? How of-ten should visual field tests be obtained? Given the variability of visual field re-sults, a sufficient number of visual field tests must be obtained in order confi-dently to identify changes that are great-er than the variability of the measure-ment. More visual fields, offer a better estimate of the rate of visual field pro-gression. But how many visual field tests are needed over what period of time? According to Chauhan et al,13 to detect relatively slow progression (-0.50 dB/yr) in 2 years, 7 visual fields are needed each year; clearly more than is practical. On Figure 5 shows the VFI regression analysis results from the same eye shown in Figure 4.


the other hand, to detect fast progress-ing glaucoma (-0.20 dB/yr), Chauhan et al 13 recommend that newly diagnosed patients complete visual field testing three times per year during the first two years after glaucoma diagnosis.13 These recommendations were adopted by The European Glaucoma Society in 2008.14 The 2011 World Glaucoma Association Consensus Group6 on Glaucomatous Progression recently developed simi-lar guidelines on the optimal number of visual field tests required with newly diagnosed glaucoma patients. Specifi-cally, during the first two years after a diagnosis of glaucoma “at least two reli-able visual fields is optimal in the first six months… and at least two further visual fields should be performed within the next 18 months. Where the lifetime risk of visual disability is high, such as those who already have advanced dam-age, three baseline visual fields may be necessary.” After the initial two years following a diagnosis of glaucoma, the frequency of visual field testing should be based on the risk of visual impairment during the patient’s lifetime. The 2011 World Glau-coma Association Consensus Group6 recommends that, “In low and moder-ate risk patients, subsequent visual field frequency should be one visual field per year … and, as a rule, repeated sooner …if other clinical observations are sugges-tive of possible progression or increased risk of progression.” However, “In high risk patients, subsequent visual field fre-quency should be two visual fields per year and repeated sooner if possible pro-gression is identified on the basis of an ‘event’ analysis, or if other clinical obser-vations are suggestive of progression or increased risk of progression.” Recently, computer simulation has been used to determine the optimal spacing for the intervals between test-ing for detection of rapidly progressing glaucoma (rate of loss: -2dB/yr). Crabb and Garway-Heath15 suggest that com-pared with evenly-spaced follow-up ev-ery six months, visual field testing 2 or 3 times at baseline and at the end of a 2 year period identifies more patients with rapidly progressing visual fields (58% and 62%, respectively) with a lower false positive rate (5.9% and 0.4%, respec-tively). By aggregating the visual field testing earlier and later in the follow-up period, there is increased efficiency

to reduce the false positive rate, and to estimate the rate of visual field loss. This “wait and see” approach, developed us-ing “moderate” variability of the visual fields in the computer simulation to de-tect fast progression, is being tested in longitudinal studies. The number of fields and intervals be-tween them can also influence the meth-ods used to measure the rate of change. For example, Medeiros et al. 12 recently reported that a Bayesian-based slope es-timate technique was more predictive of future impairment than ordinary least square regression, particularly in eyes with moderate to fast rates of change. Most patients can be monitored using the 6-degree by 6-degree grid spacing of 24-2 visual field test pattern. However, when central vision is compromised or threatened, it may be useful to obtain 10-2 visual fields that use 2-degree by 2-degree spacing to better document how glaucoma is affecting central vi-sion.16, 17 Using a relatively small number of eyes, Hood et al.17 reported evidence of visual field damage in the central vi-sual field using 10-2 test pattern that is not detected in the peripheral field using 24-2 testing. Zhang et al.16 documented that the 10-2 test pattern results are use-ful to make clinical management deci-sions; in about one third of visual fields considered to threaten fixation based on 30-2 or 24-2 test patters, the threat, based on 10-2 testing was not imminent. In some patients, 10-2 testing can be im-portant to determine the level of visual impairment expected from glaucoma-tous visual field loss in the central and peripheral visual field.

References1. Asman P, Heijl A. Glaucoma Hemifield Test. Automated visual field evaluation. Arch Ophthal-mol 1992;110:812-819.

2. Bengtsson B, Heijl A. A visual field index for calculation of glaucoma rate of progression. Am J Ophthalmol 2008;145:343-353.

3. Keltner JL, Johnson CA, Cello Ke, et al. Clas-sification of visual field abnormalities in the ocular hypertension treatment study. Arch Ophthalmol 2003;121:643-650.

4. Leske MC, Heijl A, Hyman L, Bengtsson B. early Manifest Glaucoma Trial: design and base-line data. Ophthalmology 1999;106:2144-2153.

5. Boden C, Blumenthal eZ, Pascual J, et al. Patterns of glaucomatous visual field progression identified by three progression criteria. Am J Oph-thalmol 2004;138:1029-1036.

6. World Glaucoma Association Consensus series 8. Progression of Glaucoma. eds. R.N. Weinreb & D.F. Garway-Heath & C. Leung & J.G.

Crowston & F.A. Medeiros.. Amsterdam: Kugler; 2011.

7. Nouri-Mahdavi K, Hoffman D, Gaasterland D, Caprioli J. Prediction of visual field progres-sion in glaucoma. invest Ophthalmol Vis sci 2004;45:4346-4351.

8. Caprioli J, Mock D, Bitrian e, et al. A method to measure and predict rates of regional visual field decay in glaucoma. invest Ophthalmol Vis sci 2011;52:4765-4773.

9. Fitzke FW, Hitchings RA, Poinoosawmy D, McNaught Ai, Crabb DP. Analysis of visual field progression in glaucoma. Br J Ophthalmol 1996;80:40-48.

10. Krupin T, Liebmann JM, Greenfield Ds, Ritch R, Gardiner s, Group L-PGs. A randomized trial of Brimonidine versus timolol in preserving visual function: Results from the Low-pressure Glau-coma Treatment study. American Journal of Oph-thalmology 2011;151:671-681.

11. De Moraes CG, Juthani VJ, Liebmann JM, et al. Risk factors for visual field progression in treat-ed glaucoma. Arch Ophthalmol 2011;129:562-568.

12. Medeiros FA, Zangwill LM, Weinreb RN. im-proved prediction of rates of visual field loss in glaucoma using empirical bayes estimates of slopes of change. J Glaucoma 2012;21:147-154.

13. Chauhan BC, Garway-Heath DF, Goni FJ, et al. Practical recommendations for measuring rates of visual field change in glaucoma. Br J Ophthalmol 2008;92:569-573.

14. society eG. Terminology and Guidelines for Glaucoma. savona, italy; 2008. Available at:

http://www.eugs.org/eng/eGs_guidelines.asp. Accessed March 19, 2012

15. Crabb DP, Garway-Heath DF. intervals be-tween visual field tests when monitoring the glau-comatous patient: Wait and see. invest Ophthal-mol Vis sci 2012.

16. Zhang L, Drance sM, Douglas GR. Auto-mated perimetry in detecting threats to fixation. Ophthalmology 1997;104:1918-1920.

17. Hood DC, Raza As, de Moraes CG, et al. initial arcuate defects within the central 10 de-grees in glaucoma. invest Ophthalmol Vis sci 2011;52:940-946.


What’s New:

Where are we going with approaches such as PeRG, VeP, sWAP and FDT?Stuart L Graham, MD Professor of Ophthalmology and Vision Science, Australianschool Advanced Medicine, Macquarie university, Australia

Is anything new in functional testing?There have been many attempts to refine psychophysical and electrophysiologi-cal techniques over the last 20 years to enable earlier detection and/or possibly provide an objective means for monitor-ing function. While a few techniques survive in clinical practice, there has generally not been any major break-throughs in recent times in the func-tional testing area. Standard automated perimetry (SAP) with white on white stimulus still remains the gold standard despite all its well known limitations.

PeRG/ sTR/ PhNRThe pattern electroretinogram (PERG) consists of an averaged response to a uni-formly reversing checkerboard pattern, and is derived from the central visual field. It has been described as abnormal in glaucoma in many studies dating back as far as the 1980’s. Reduction of the two main components P50 (positive wave at

50 ms) and N95 (negative wave at 95ms) has been reported, but in glaucoma there may be more significant effects on the N95 1. The Freiburg group use a PERG ratio comparing responses at 0.8 degree check size (which are more affected in early glaucoma) to 16 degree checks2. Latency changes are also reported, al-though not as substantial as that seen in demyelinating diseases (optic neuritis). The PERGLA (Lace Elettronica, Pisa, Italy) is a system designed to pro-vide more rapid investigation using a short protocol steady state PERG. While many animal and human stud-ies agree that the PERG is reduced by ganglion cell loss, and source analysis confirms it to be generated in the in-ner retina, clinical correlations have not been accurate enough for it to become widely used for diagnostic or monitor-ing purposes. There is a wide range of variability in normal and disease with considerable overlap. Also it is subject to reductions from any other pathology

Figure 1. Binocular simultaneous mfVEP (dichoptic). Shows reduction of signal in scotoma areas in both eyes compared to corresponding SITA field

Core Concepts • Standard automated perimetry(sAP) with white on white stimulus re-mains the gold standard for functional testing.• Patternelectroretinogram(PERG)can be used where perimetry is not possible; it is valuable to investige unexplained visual loss.• Scotopic threshold response(sTR) has not been found useful for clinical detection of glaucoma.• Photopic negative response(PhNR) has not been shown to be a good discriminator to detect dis-ease.• Multifocal visually evoked poten-tial (mfVeP) has been found useful to investigate patients with visual field loss that does not match the clinical picture.• Short wavelength automated pe-rimetry (sWAP) has been shown use-ful in younger patients who are good observers, without nuclear sclerosis, where there is high suspicion of early damage.• Frequency doubling technology(FDT) matrix is at least as sensitive as sAP for detection of damage.• Flicker defined form (FDF) maydetect glaucomatous defects simi-larly to sAP. More data is needed to determine sensitivity/specificity and ability to detect early disease.• Alternative strategy subjectivetests sWAP, FDT matrix and per-haps HeP can be used in “pre-peri-metric glaucoma” patients , as they may detect changes earlier.


affecting visual acuity and retinal func-tion, so like so many of our functional tests, it is non-specific for glaucoma. It can be of use in cases where perimetry is not possible and is of value in investi-gating unexplained visual loss. The fact that there appears some generalised loss which does not correspond topographi-cally, has confounded the possibility of adapting a multifocal PERG. The scotopic threshold response (STR) is a signal recorded to a dim stim-ulus at the depths of dark adaptation. It appears to be generated in inner retina and is sensitive to optic nerve damage in animal models, and regularly used in lab based research. In humans it has not been found to be useful for clinical de-tection of glaucoma. The photopic negative response (PhNR) is a slow negative response fol-lowing the b wave peak. It is easier to record than the STR, and seems to be dependent on retinal ganglion cell func-tion. However while there are reports of its reduction in human glaucoma, it also has not been shown to be a good dis-criminator for detecting disease.

mfVePThe conventional pattern VEP (visu-ally evoked potential) represents a single summed response that is mainly derived from the macular region, and due to cortical anatomy, it is dominated by the inferior hemifield. In fact, a dense su-perior field defect can go undetected on conventional VEP testing3. To evaluate local VEP responses from the visual cortex and to map them topographically the multifocal VEP (mfVEP) has been developed. Using a multichannel mfVEP technique it is possible to objectively detect visual field defects4. Multifocal stimulation is now available commercially in several dif-ferent electrophysiological systems. The visual stimulus is usually generated on a CRT screen (eg, 22-inch high-resolution display), but with faster refresh rates, LCD flat screens can now be used. We have also used virtual reality goggles and twin LCD screens to present the stimu-lus dichoptically5 (see Figure 1). A blue-on-yellow mfVEP has also been de-scribed using a sparse blue pattern-onset stimulus (instead of pattern reversal) on a yellow-adapting background. The goal was to target the koniocellular pathway. This approach showed good sensitiv-

ity and displayed more extensive scotomata than the conventional black-white mfVEP (92.2% sensitivity), and it correlated well with SAP6. The advantage of the blue on yellow stimulus over standard mfVEP, however, was probably more likely related to the fact that the pattern onset stimulus was spatially sparse 7 (likely due to less lateral inhibition) and was of low luminance, both of which we have demonstrated may increase sensitivity 8. The mf VEP is particularly useful for investigating patients with field loss that does not match the clinical picture, either because they are poor performers on subjective tests or there is a suspicion of other pathology. It supplements but does not replace the findings of subjective SAP and does not have the suitable follow up capability to compare with SITA or Octopus progression analysis software. Pupil perimetry has also been re-ported9, as a form of objective perimetry and uses a multifocal type stimulus with sophisticated pupil tracking to capture responses. It has the issues associated with pupil variability to deal with and requires an intact efferent pathway (in at least one eye) to give meaningful results, but does seem to work well as a screen-ing tool and is completely non-invasive.

sWAP (short wavelength automat-ed perimetry), and FDT (frequency doubling technology)The SWAP technique for perimetry presents a 440nm blue size V target on a bright yellow background to attempt to selectively test the koniocellular pathway. Initial studies had reported

favourable early detection of glaucoma, despite some of the limitations including greater fluctuation of responses, lens ab-sorption of the blue light stimulus, and generally lower patient acceptance of the test. Some more recent reports suggest no clinical advantage in detection over SITA white-on-white10. It may still be useful in younger patients who are good observers where there is a high index of suspicion of early damage, but despite the introduction of SITA SWAP it has generally not been as widely adopted as once anticipated. The FDT technique was also intro-duced as a means of attempting to tar-get a subpopulation of ganglion cells, in this case the magnocellular pathay, to

Figure 2. Comparison of standard perimetry (Humphrey SITA) (fig 2a) with Heidelberg Edge Perimeter (HEP) which uses flicker defined form (fig 2b) for the same subject.


Figure 3. Structure function correlation map produced by Heidelberg Retinal Tomograph and HEP. Inner coloured ring represents Moorfield’s classification, while outer ring represents HEP visual field result for corresponding sector of disc.

achieve early detection of ganglion cell loss. The initial FDT perimeter which had 19 test zones, has been expanded in the Matrix Perimeter (Zeiss) to include a test grid similar to the standard 24-2 and 30-2 Humphrey fields. The test strategy is relatively fast, being around 5-6 min-utes. Studies suggest FDT Matrix is at least as sensitive as standard automated perimetry (SAP)11.

Flicker defined form (FDF)Flicker defined form is a recently adapt-ed technique in the Heidelberg Edge Pe-rimeter (HEP, Heidelberg Engineering) which presents as a stimulus counter-phase dots to discrete areas within the field. In my early experience it seems to be well received by patients and detects glaucomatous defects with similar dis-tribution to SAP. Further studies are needed to test its sensitivity/specificity and ability to detect early disease. The addition of a combined structure-function map by

earlier in some cases. However SWAP is quite variable and better reserved for younger pa-tients who are good observers. I have found the Matrix FDT sometimes works well for pa-tients who fatigue on SAP, and it can be easier for those with back problems due to easier po-sitioning, however none of the alternate strategies have the same follow up software capa-bilities for progression analysis.

References1. Holder, G.e., et al., electrodiag-nostic assessment in optic nerve dis-ease. Current Opinion in Neurology, 2009. 22(1): p. 3-10.

2. Bach, M., et al., Pattern eRG as an early Glaucoma indicator in Ocular Hypertension: A Long-Term, Pro-spective study. invest Ophthalmol Vis sci, 2006. 47: p. 4881-7.

3. Graham, s.L. and B. Fortune, elec-trophysiology in glaucoma, in Glau-coma, T. shaaraway, et al., editors. 2009, elsevier,, 2009,: London, uK. p. 151-171.

4. Graham, s.L., A. Klistorner, and i. Goldberg, Clinical application of the multifocal VeP in glaucoma. Arch Ophthalmol, 2005. 123: p. 729-73.

5. Arvind, H., et al., Dichop-tic stimulation improves detection of glaucoma with multifocal visual evoked potentials. investigative Oph-thalmology & Visual science, 2007. 48(10): p. 4590-4596.

6. Arvind, H., et al., identifying pre-perimetric Functional loss in Glaucoma: A Blue-on-yellow Multifocal Visual evoked Potentials study. Oph-thalmology, 2009. 116: p. 1134-1141.

7. Fortune, B., et al., Multifocal visual evoked potential responses to pattern-reversal, pattern-onset, pattern-offset, and sparse pulse stimuli. Visual Neuroscience, 2009. 26(2): p. 227-35.

8. Arvind, H., et al., Low-luminance contrast stim-ulation is optimal for early detection of glaucoma using multifocal visual evoked potentials invest Ophthalmol Vis sci 2011. 52(6): p. 3744-50.

9. Carle, C.F., et al., High-Resolution Multifocal Pupillographic Objective Perimetry in Glaucoma. invest Ophthalmol Vis sci, 2011. 52: p. 604–610.

10. van der shoot, J., et al., The ability of short wavelength automated perimetry to predict con-version to glaucoma. Ophthalmology, 2010. 117: p. 30-34.

11. Racette, L., et al., Diagnostic accuracy of the Matrix 24-2 and original N-30 frequency dou-bling technology tests compared with standard automated perimetry. invest Ophthalmol Vis sci, 2008. 49: p. 954-960.

combining the test results of the HEP with the HRT (Heidelberg Retina To-mograph) scan structural data has been a nice addition for clinicians. However, like all structure-function correlations, there appears to be many cases where the two simply do not match. Figure 2 shows an example of HEP where there is quite reasonable correlation with SAP findings. Figure 3 shows an example of a structure-function correlation map.

Clinical applicationsElectrophysiology testing using either PERG or mfVEP can be helpful in pa-tients who are poor perimetrists, in cases where there is disproportionate field loss compared to the clinical picture, and in other cases of unexplained vision loss in-cluding functional overlay. The alternative strategy subjective tests SWAP, FDT Matrix and perhaps now HEP also, can be used in investigat-ing patients with suspect early disease, as they may be able to detect changes


Clinical issues:

What tests and strategies to use for whom and when?Andrew Toren, MD, Marcelo T. Nicolela, MDDepartment of Ophthalmology and Visual sciences, Dalhousie university, Halifax, Nova scotia, Canada

Functional Assessment in GlaucomaAll glaucoma patients and suspects should have periodic visual field examinations if they are able to perform this test. This functional assessment in glaucoma is cru-cial to provide the information necessary to make the best decisions in a patient’s management. The clinician must choose thoughtfully the best visual field test (or combination of tests) as well as the interval between examinations for each patient. As testing technologies and software analysis for visual fields continue to evolve, standard automated perimetry (SAP) re-mains the gold standard for diagnosis and monitoring of progression in glaucoma. While other testing modalities are avail-able, e.g. frequency doubling technology (FDT) and short-wavelength automated perimetry (SWAP), these should not be used at the expense of SAP1. As the most widely utilized hardware, the Humphrey field analyzer (HFA) is will

be the example in this article, although the same principles apply to other SAP modali-ties such as Octopus perimetry. To diagnose and monitor glaucoma, several testing strategies can be employed: Full Threshold, SITA (Swedish Interac-tive Thresholding Algorithm) standard, FAST PAC, screening strategies and SITA FAST. Full threshold requires a long pe-riod of concentration and does not appear to offer additional benefits to SITA stan-dard2, 3. A SITA FAST test takes less time but has a greater degree of variability2,4. It is less useful to monitor glaucoma, but can be used as a first screening test for diagnosis. Using the same test pattern over time improves the chance to detect visual field changes. As the 24-2 test pattern saves time compared with the 30-2 pattern, without compromising the ability to diagnose or to monitor glaucoma progression, it is the pre-ferred test pattern for most patients6. In patients with advanced glaucoma and/or visual field loss approaching fixation, a HFA 10-2 test is useful to monitor for pro-gression in this critical central region of vision. 10-2 tests can be alternated or com-bined with 24-2 tests in patients with re-maining peripheral vision, such as patients with one hemifield threatening fixation but with the other hemifield relatively normal. Although there is currently no statistical

progression analysis available within the HFA for 10-2 fields, thresholds from these important points close to fixation can be compared subjectively. (Figure 1). Alter-natively, pointwise linear analysis can be utilized with third party software packages such as PROGRESSOR or PeriData. In patients with severe visual field dam-age, particularly with decreased central visual acuity, where a traditional Size III stimulus does not provide a dynamic range to observe progression, a Size V stimulus can be useful to assess the remaining visual function. (Figure 2). The number and frequency of visual field tests need to be considered carefully in the context of the burden of repeated tests on patients, the number of visits required, the reliability of test results and financial implications. The optimal frequency for visual field tests depends on the course and severity of the disease and patient factors such as age, test tolerability and perfor-mance. As there is considerable variation between individuals for rate of progres-sion, more frequent initial tests provide an earlier and more accurate analysis of this rate for the individual. Ideally, 6 visual field tests in the first two years following a glaucoma diagnosis should be done, and then the number can be adjusted (1). In summary, a correct selection of vi-sual fields, both for diagnosis and moni-toring will depend on the individual patient’s characteristics. Most frequently a SITA 24-2 test will offer the best com-bination of test speed, reliability and

Figure 1. 24-2 Matrix of the 24-2 pattern (A) vs the 10-2 pattern (B) on retina.

Figure 2. Comparison of Size III (A) vs Size V (B) testing stimulus in a patient with ad-vanced visual field damage.

Core Concepts • Foreachpatient,choosethoughtfullythebest visual field test or combination of tests as well as intervals between tests. This may change with time.• Frequency doubling technology (FDT)and short-wavelength automated perim-etry (sWAP) should not be used routinely instead of standard automated perimetry (sAP).• Todiagnoseandmonitorglaucomasev-eral testing strategies may be useful such as Full Threshold, siTA standard, FAsT PAC and siTA FAsT.• Fullthresholdtestingdoesnotofferad-ditional benefit to siTA and requires more time; siTA FAsT takes less time again, but with greater variability

• SITA24-2savestimecomparedwiththe30-2 pattern without compromising ability to diagnose or monitor glaucoma.• Inpatientswithadvancedglaucomaand/or visual field loss approaching fixation 10-2 testing is useful to monitor for progres-sion in critical central vision• In patientswith severe damage (espe-cially with decreased VA) a size V is pre-ferred to a size iii stimulus.• Morefrequentinitialvisualfieldtestsaremore accurate to assess the rate of any progression.• TheSITA24-2 testmost frequentlyof-fers the best combination of speed, reli-ability and availability of analysis software.


Practical Tips:

Minimizing artifacts and avoiding pitfalls in interpretationRavi Thomas MD, FRANZCOQueensland eye institute & university of Queensland. Australia

Core Concepts • Successful communication be-tween perimetrist and patient are of utmost importance.• Thepatientneedstobecomfort-ably and properly positioned. Acti-vate the forehead rest alarm. • Comments by the operator onpatient performance and accompa-nying circumstances are useful for interpretation. • Potential operator errors are list-ed in the field analyser manual and should be reviewed by the techni-cian from time to time.• Thegreyscaleisoflimitedusebutdistinctive patterns such as the Clo-

ver leaf, the Maltese Cross or swiss Cheese suggest that test data is not reliable. • The foveal threshold measure-ment option should be switched on. • Use of the diamond fixation tar-gets can help obtain reliable fields in patients with central vision de-fects (e.g. from macular degenera-tion)• “Baselinefields”shouldbeestab-lished as soon as possible and up-dated as needed.• Always consider the completeclinical picture including the role of coexisting conditions. Never inter-pret the visual fields in isolation.

Given the objectives (and space con-straints), I will mention a few well-known artifacts but concentrate more on selected issues that are easily overlooked in a busy clinic. Instilling positive attitudes in patients and staff is a neglected but important component for effective perimetry.1 The person conducting the test does not re-quire special qualifications but is as im-portant as the machine. Their explana-tion, demonstration and interaction with the patient help obtain valid information.Though the grey scale is the least useful part of the printout, the occasional, typi-cal patterns produced by some artifacts do make it worth just a passing glance. Distinctive appearances on the grey scale include the “Clover Leaf ” pattern sug-gestive of patient fatigue, the “Maltese

Cross” that occurs when the patient is a “slow starter” and the “white” scotomas (or “Swiss Cheese” ap-pearance) produced by the false positives of “trigger happy” pa-tients. The presence of more abnor-mal points on the pattern deviation as compared to the total deviation plot is a subtle indicator of false positives even when they have not been identified or labelled as exces-sive. Such issues usually improve with proper instruction and as the pa-tient gets over the learning curve. It usually takes two or three fields to get over the learning curve but a first field is not entirely useless. If an unreliable first field correlates with clinical findings, it still pro-

Figure 1

Figure 2

availability of analysis software to be utilized to monitor glaucoma. Other test strategies such as SITA Fast, 10-2 pat-terns and size V stimulus can be used in special circumstances.

References:1 Chauhan BC, Garway-Heath DF, Goni FJ, Rossetti L, Bengtsson B, Viswanathan AC, et al. Practical recommendations for measuring rates of

visual field change in glaucoma. Br J Ophthalmol. 2008 Apr;92(4):569-73.

2 Artes PH, iwase A, Ohno Y, Kitazawa Y, Chauhan BC. Properties of perimetric threshold estimates from full threshold, siTA standard, and siTA fast strategies. invest Ophthalmol Vis sci. 2002 Aug;43(8):2654-9.

3 sekhar GC, Naduvilath TJ, Lakkai M, Jayaku-mar AJ, Pandi GT, Mandal AK, et al. sensitivity of swedish interactive threshold algorithm compared with standard full threshold algorithm in hum-

phrey visual field testing. Ophthalmology. 2000 Jul;107(7):1303-8.

4 Budenz DL, Rhee P, Feuer WJ, Mcsoley J, Johnson CA, Anderson DR. sensitivity and speci-ficity of the swedish interactive threshold algo-rithm for glaucomatous visual field defects. Oph-thalmology. 2002 Jun;109(6):1052-8.

5 Khoury JM, Donahue sP, Lavin PJ, Tsai JC. Comparison of 24-2 and 30-2 perimetry in glau-comatous and nonglaucomatous optic neuropa-thies. J Neuroophthalmol. 1999 Jun;19(2):100-8.


vides valuable information required to initiate manage-ment. Conversely, it is dif-ficult to produce a reliable normal field; so a reliable, normal field, even if it is the first the patient has done, is likely to be normal. 1. Elderly patients with “poor vision” from macular degeneration may not see the usual central fixation target. Use of the diamond fixation targets can help obtain reli-able fields.2

2. Patient positioning is not just about comfort. If the forehead drifts away from the forehead-rest, the rim of the correcting lens pro-duces an artifact that can mimic a glaucomatous field defect. A patient I managed was advised urgent surgery for such a defect. A repeat field with adequate po-sitioning was the safer intervention, and unlike surgery, even made the “defect” disappear. Switching on and heeding the alarm that sounds when the patient’s forehead drifts away from the forehead-rest helps avoid this artifact. 3. The operator should be encouraged to provide comments on patient perfor-mance. The defect depicted in Figure 1 was compatible with the advanced op-tic disc changes but no comments were added. The clinician noted the ptosis in

portant to establish a good baseline as soon as possible. The baseline should ex-clude fields obtained during the learning curve and should be changed as needed. Following surgery, the field in Figure 4 provides a “likely” progression message that disappeared with the new baseline (Figure 5). Remember that the fields “triggering” the event can be used for the new baseline. 6. The foveal threshold measurement option should be turned on. Foveal threshold serves as internal validity for vision and provides a feel for the “drop

off” of the hill of vision. With the usual statistical comparisons to “normal”, it is easy to forget that a scotoma is actually a de-pression compared to the surround. The pattern deviation plot in Fig-ure 6 suggests a central scotoma. As the foveal threshold is higher than the surround this is not a central scotoma but an artifact produced by (otherwise very helpful) sophisticated analyses. Measuring the foveal threshold helps identify a true scotoma: if the fovea has lost its physiological superiority, its sensitiv-ity will be less than that of the surround.3 The su-

this, the patients “good” eye. A repeat field with the eyelid taped up (as shown in Figure 2) changed the prognosis (and management). Think beyond glaucoma. 4. Some operator errors escape casual inspection. Careful inspection of Figure 3 shows that the baseline for the visual field index has merged the R and L eyes. The field analyser manual details, how this operator error can occur.3

5. Detection of progression is the most important decision in glaucoma manage-ment and while there are logistical and reimbursement issues involved, it is im-

Figure 3 Figure 4

Figure 6Figure 5


STATEMENT OF NEED AND PROGRAM DESCRIPTIONRecent months and years have seen significant advances in our under-standing of glaucoma. Much has been learned, not only about damage mechanisms and pathogenesis, but also about diagnosis and management. Treatment options – both medical and surgical – continue to expand. This program will review this new knowledge with an emphasis on incorporating recent insights into day-to-day practice.

DATE OF ORIGINAL RELEASEMay 2012. Approved for a period of 12 months.

DISCLAIMERParticipants have an implied responsibil-ity to use newly acquired information to enhance patient outcomes and profes-sional development. The information presented in this activity is not meant to serve as a guideline for patient care. Any procedures, medications, or other cours-es of diagnosis or treatment discussed or suggested in this activity should not be used by clinicians without evalua-tion of their patient’s conditions and possible contraindications or dangers in use, applicable manufacturer’s product information, and comparison with recom-mendations of other authorities.

CONTRIBUTORS• Linda Zangwill, PhD is an Associate

Professor in epidemiology and Bio statistics at the university of Cali fornia san Diego, Hamilton Glaucoma Center and at the shirley eye Center, san Diego, California, usA. she has no commercial

relationships to disclose

• Christopher Bowd, PhD is a research scientist at the university of

California san Diego, Hamilton Glaucoma Center and at the shirley eye Center, san Diego, California, usA. He has no commercial relationships to disclose.

• Stuart L Graham, MD is Professor of Ophthalmology and Vision science, Australian school Advanced Medicine, Macquarie university, Australia. He has disclosed to be developer designated on a patent to mutifocal techniques.

• Andrew Toren, MD is Professor of Ophthalmology at the Department of Ophthalmology and Visual sciences, Dalhousie university, Halifax, Nova scotia, Canada. He has no commer-cial relationships to disclose.

• Marcelo T. Nicolela, MD is Associate Professor of Ophthalmol-ogy at the Department of Ophthal-mology and Visual sciences, Dalhousie university, Halifax, Nova scotia, Canada. He has no commer-cial relationships to disclose.

• Ravi Thomas MD, FRANZCO is Professor at the Queensland eye institute & university of Queensland, Australia. He has received honoraria from Allergan inc, and Zeiss.

DISCLOSURE STATEMENT EDITORIAL BOARDClive Migdal serves on the Faculty and Advisory Boards of the following com-panies: Alcon, Allergan, Merck, Pfizer and santen.

Ivan Goldberg serves on the Faculty and Advisory Boards of the following com panies: Alcon, Allergan, Merck and Pfizer.

Remo Susanna serves on the Faculty and Advisory Boards of the following companies: Alcon, Allergan, Merck and Pfizer.

perior resolution of the 10-2 program is useful to detect such suspected scotomas (Figure 7). 7. A well defined posterior sub capsu-lar cataract is a rare cause for a field de-fect that fulfils the criteria for a glauco-matous defect.4 Cataract surgery rather than glaucoma treatment is needed to address such a defect. I’ll conclude with some common sense perimetrisms from Patel Bhai (personal communication)• Ontheirown,eventhebestautomat-ed perimeters are just dumb machines. • The number of ways in which theperimetrist and / or patient can mess up the results is unlimited.• The number of ways a clinician canmess up interpretation is also seemingly unlimited.• Never, never ever interpret a visualfield in isolation.

References 1 Patella M. Heijl A. effective Perimetry. Chap-ter in Humphrey Field Analyzer users Manual. in press.

2 Anderson D, Patella M,. Automated static Pe-rimetry 2nd edition. st Louis: Mosby 1999.

3 Humphrey Field Analyzer ii-i series system software Version 5.1 user Manual. 2010. Carl Zeiss Meditec.

4 Thomas R, Kuriakose T, George T. Glaucoma like defect in automated perimetry caused by cataract. indian Journal of Ophthalmology 1996; 44:167-168

Figure 7

CME credits can be obtained via the questions on the website very soon. The process of obtaining CME accreditation for the journal is currently ongoing.

Documents

Main topic: Functional Assessment in Glaucomaglaucomanow.com/media/15319/glaucomanow_issue1_2012.pdf · Glaucoma Now Issue No 1 2012. 1 issue 1 2012 Main topic: Functional Assessment