Revisiting the Structure-Function Relationship in Glaucoma: Implications from
Recent Clinical Trials.
Chair: John G. Flanagan PhD, MCOptom, FAAO
Professor, School of Optometry, University of Waterloo and
Dept. of Ophthalmology and Vision Sciences, University of Toronto, Ontario, Canada.
Pamela Sample PhD
Professor, Hamilton Glaucoma Centre
Department of Ophthalmology
University of California, San Diego
Linda Zangwill PhD
Professor, Hamilton Glaucoma Centre
Department of Ophthalmology
University of California, San Diego
Chris Johnson PhD,
Professor, Discoveries in Sight,
Devers Eye Institute, Portland
Abstract
This course will review the current understanding of the structure-function relationship as
it relates to glaucoma, and reviews important information from recent and ongoing
clinical trials that may force us to modify current perceptions. Included will be a
discussion of the OHTs ancillary studies involving both Short-Wavelength Automated
Perimetry and the Heidelberg Retina Tomograph. In addition there will be discussion of
the African Americans with Glaucoma (AAG) study.
Learning Objectives
1. Review current understanding of the structure-function relationship in glaucoma
2. Understand the implications of the OHTs ancillary study incorporating SWAP
3. Understand the implications of the OHTs ancillary study incorporating HRT
4. Review the results of the AAG study
Outline
1. Short Wavelength Automated Perimetry (SWAP) in the Ocular Hypertension Treatment
Study (OHTS)
1.1 Purpose:
i. To determine the percentage of SWAP deficits for participants enrolled in the
OHTS
ii. To evaluate the relationship between SWAP results and other baseline clinical
characteristics and
iii. To assess the predictive value of initial SWAP deficits.
1.2 Methods: All participants had normal and reliable standard Humphrey visual fields at
baseline. SWAP testing was performed in 397 eyes of 199 participants enrolled in the OHTS
SWAP ancillary study at nine participating clinical centers, using the previously-determined
1
optimal testing conditions for SWAP. A standardized protocol was used for all clinical centers
and all technicians were required to complete a certification process prior to testing participants.
1.3 Results:
i. The prevalence of SWAP abnormalities at the initial test, for the Glaucoma
Hemifield Test (GHT) analysis was 20.9% when there was no detectable visual field deficit for
standard (white-on-white) automated perimetry.
ii. There was a higher prevalence of SWAP abnormalities in African Americans (32.6%) as
compared to white, not Hispanics (19.2%) (p<.026).
ii. There was a higher prevalence of SWAP abnormalities for participants with a high school
education or less (31.6%) as compared to participants with 1 or more years of college (19.1%)
(p<.016).
iv. No other significant associations were found for the prevalence of SWAP deficits as a function
of age, gender, IOP, diabetes, high blood pressure, low blood pressure, family history of
glaucoma or migraine.
v. SWAP abnormalities at the initial test were not significantly associated with prediction of a
POAG endpoint.
1.4 Conclusions: i. Approximately 20% of OHTS participants enrolled in the SWAP ancillary
study demonstrated abnormalities on the SWAP GHT analysis, despite having normal and
reliable visual fields on standard automated perimetry.
ii. Our results suggest that the prevalence of SWAP deficits is higher in African Americans and in
participants with a lower education level.
ii. Initial SWAP deficits were not significantly associated with the development of a Primary
Open Angle Glaucoma (POAG) endpoint, although further follow-up may provide further
information in this regard.
2. The Confocal Scanning Laser Ophthalmoscopy (CSLO) Ancillary Study to the OHTS
2.1 Study Design and Baseline Factors.
2.1.1 Purpose: To describe the study design of the CSLO Ancillary Study to the OHTS
and to examine the associations between optic disk topography, and baseline demographic,
clinical, and ocular factors at study entry.
2.1.2 Methods: Associations between Heidleberg Retina Tomograph (HRT) topographic
optic disk measurements and IOP, baseline photographic estimates of horizontal and vertical cupto-disk diameter ratios by the OHTS Optic Disk Reading Center, baseline visual field indices, and
demographic and clinical factors were assessed using linear mixed effects models.
2.1.3 Results: 439 participants with good quality images were included. No associations
between HRT topographic optic disk measurements and diabetes, systemic hypertension,
cardiovascular disease, IOP, or visual function were found. The HRT topographic optic disk
measurements were associated with baseline stereophotographic estimates of horizontal and
vertical cup-to-disk diameter ratios. The strongest associations were found between
stereophotographic assessment of horizontal and vertical cup-to-disk diameter ratios, and HRT
cup-to-disk area ratio (r = .85 and .84, respectively), rim-to-disk area ratio (r = -.85 and -84,
respectively), mean cup depth (r = .84 and .83, respectively), and cup area (r = .83 and .80,
respectively). After adjusting for optic disk area, all HRT topographic optic disk measurements
remained associated with stereophotographic assessment of horizontal and vertical cup-to-disk
diameter ratios.
2.1.4 Conclusions: The CSLO ancillary study to the OHTS is the first multicenter clinical
trial to use CSLO imaging to monitor changes in the optic disk. At study entry, HRT topographic
measurements corresponded well with both horizontal and vertical stereophotographic-based
estimates of cup-to-disk diameter ratio in OHTS participants.
2
2.2 Racial Differences in Optic Disc Topography
2.2.1 Purpose: To examine racial differences in optic disc topography among ocular
hypertensive participants in the Ocular Hypertension Treatment Study.
2.2.2 Methods: 439 participants with good-quality baseline CSLO images obtained using
a HRT were included. Differences in HRT topographic optic disc parameter measurements by
self-identified race were assessed using a mixed-effects linear model to control for confounders
and for the use of both eyes in the model.
2.2.3 Results: By self-attribution, 74 (17%) of the 439 participants were of African
origin, 329 (75%) were white, 24 (5%) were Hispanic, and 12 (3%) were Native American,
Native Alaskan, Asian, Pacific Islander, or unknown. The African American participants had
statistically significantly (P<.001) larger mean (SD) optic disc areas than the other participants,
2.17 (0.41) mm2 vs. 1.87 (0.38) mm2, respectively. African American participants had a larger
cup area, cup volume, cup depth, neuroretinal rim area, rim volume, and smaller rim–to-disc area
ratios than the other participants. No difference between African American and the other
participants was found for cup shape and retinal nerve fiber layer thickness. After controlling for
optic disc area, none of the differences between African American and the other participants
found in the univariate analysis remained statistically significant (P>.10).
2.2.4 Conclusions: This study found that differences in topographic optic disc parameters
between African Americans with ocular hypertension and other racial groups are largely
explained by the larger optic disc area in the African Americans. These results highlight the need
to consider race and optic disc size when evaluating the appearance of the optic disc in glaucoma.
2.3 Baseline Topographic Optic Disc Measurements are Associated with the Development of
POAG
2.3.1 Purpose: To determine whether baseline CSLO optic disc topographic
measurements are associated the development of POAG in ocular hypertension.
2.3.2 Methods: 865 eyes from 438 participants with good quality baseline CSLO images
were included. Each baseline CSLO parameter was assessed in univariate and multivariate
proportional hazards models to determine its association with the development of POAG (a
confirmed visual field abnormality or a confirmed clinically significant confirmed optic disc
deterioration attributed to glaucoma.)
2.3.3 Results: 41 eyes from 36 CSLO Ancillary Study participants developed POAG. 31
(77%) eyes reached an optic disc endpoint only, 9 (22%) eyes reached a visual field endpoint
only, and 1 eye reached both types of endpoints concurrently. Several baseline CSLO
measurements were significantly associated with the development of POAG in both univariate
and multivariate analysis including larger cup-to-disc area ratio, mean cup depth, mean height
contour, cup volume, reference plane height, and smaller rim area, rim area to disc area, and rim
volume. In addition, classification as “outside normal limits” by the HRT “Classification” and the
Moorfields Regression Analysis (MRA) (overall, global, temporal inferior, nasal inferior and
superior temporal regions) classifications were significantly associated with the development of
POAG. However, within the follow-up period of this analysis, the predictive value of a positive
test was low for many CSLO indices, ranging from 14% (HRT Classification and MRA Overall)
to 27% for MRA Global.
2.3.4 Conclusion: Several baseline CSLO measurements alone or when combined with
baseline clinical and demographic factors were significantly associated with the development of
POAG among OHTS participants. Longer follow-up is required to evaluate the true predictive
accuracy of CSLO measures.
3
3. Comparison of New Diagnostic Techniques in Black and White Participants.
3.1 Background: Glaucoma is 5 times more likely to occur in Blacks than Whites, and is the
leading cause of irreversible blindness in Blacks. Little is known about whether new
diagnostic tests including vision function specific tests and optic disc and retinal nerve fiber
layer imaging can improve our ability to detect and monitor glaucoma in the Black
population.
3.2 Purpose: To compare differences in visual function tests and measures of optic nerve
structure in normal, ocular hypertensive, and glaucomatous Black and White participants, to
determine whether function specific perimetry tests show improved diagnosis in Blacks as
they do in Whites, and to compare the rate and patterns of progression of glaucomatous
damage in the two racial groups.
3.3 Methods: 125 normal and 275 Black participants with glaucoma. Black participants will
be recruited primarily from New York Eye and Ear Infirmary and the University of Alabama
Glaucoma Service. White participants are recruited under separate funding at UC San Diego.
UC San Diego serves as a Reading Center for data processing, quality control and analysis
for this study. The glaucomatous eyes will be followed with annual testing for 5 years.
Visual fields will include standard clinical perimetry, and two function-specific tests, SWAP
and Frequency Doubling Perimetry. Optic nerve structure will be qualitatively evaluated with
stereoscopic fundus photography and quantitatively using the HRT and Optical Coherence
Tomograph. Comparisons between Blacks and Whites will be made on each test and between
the measures of function and structure.
3.4 Results: Preliminary results will be presented.
3.5 Significance: A greater understanding of the importance of race for development and
progression of glaucoma will result leading to more precise endpoints for future clinical trials
with this at-risk minority population.
4