Funduscopic versus hrt iii confocal scanner vertical cup disc ratio assessment in normal tension and by Sāng Dé Luó CR

Short Communication Received: April 20, 2016 Accepted: May 7, 2016 Published online: August 4, 2016

Ophthalmic Res DOI: 10.1159/000446659

Funduscopic versus HRT III Confocal Scanner Vertical Cup-Disc Ratio Assessment in Normal Tension and Primary Open Angle Glaucoma (The Leuven Eye Study) Koen Willekens a Sophie Bataillie a Inge Sarens a Sofie Odent a Luìs Abegão Pinto b Evelien Vandewalle a Karel Van Keer a Ingeborg Stalmans a a

Department of Ophthalmology, University Hospitals Leuven, Leuven, Belgium; b Department of Ophthalmology, Centro Hospitalar Lisboa Norte, Lisbon, Portugal

Abstract Purpose: To compare funduscopic and confocal scanning vertical cup-disc ratio (VCDR) assessments and their respective predictive value for estimating functional glaucomatous damage. Methods: Data from a single eye of open angle glaucoma patients from the Leuven Eye Study were included: age, gender, intra-ocular pressure, visual acuity, refractive error, visual field mean deviation and pattern standard deviation, funduscopic and HRT III VCDRs as well as mean retinal nerve fibre layer thickness. Non-parametric tests to compare differences within and between diagnostic groups were used, and receiver-operating characteristic curves as well as Bland-Altman plots constructed. Results: Three hundred and one eyes of 301 subjects with primary open angle glaucoma (POAG) and normal tension glaucoma (NTG) were included. The average VCDR assessed with HRT III was significantly smaller than the funduscopic measurement (0.69 ± 0.16 vs. 0.81 ± 0.14, respectively; p < 0.001). The predictive value of both measurement techniques did not differ in NTG patients, but the funduscopic estimate yielded a significantly larger predictive power in patients with severe POAG.

Conclusion: Funduscopic and confocal scanner estimates of VCDR differ significantly and should not be used interchangeably. In POAG patients with severe glaucoma, a subjective VCDR predicts functional glaucomatous damage significantly better. © 2016 S. Karger AG, Basel

Introduction

Glaucoma is a neurodegenerative disorder characterized by irreversible progressive loss of retinal ganglion cells and one of the leading causes of blindness in developed countries [1, 2]. It is defined as having characteristic glaucomatous visual field and optic disc changes, i.e. thinning of the inferior and/or superior rim, vertical cup-disc ratio (VCDR) asymmetry of >0.2, not due to optic disc size asymmetry, together with matching visual field defects [3]. VCDR can be assessed rapidly by direct or indirect funduscopy and is reliably estimated with a good agreement between clinicians [4, 5]. Researchers have shown that structural damage might precede functional damage and VCDR might be altered in preperimetric glaucoma [6–8]. Therefore it is of paramount importance to investigate and measure changes in the optic disc and Ingeborg Stalmans, MD, PhD Department of Ophthalmology, University Hospitals Leuven, Campus St-Raphaël Kapucijnenvoer 33 BE–3000 Leuven (Belgium) E-Mail ingeborg.stalmans @ uzleuven.be

Downloaded by: Univ. of California San Diego 132.239.1.231 - 1/2/2017 7:58:37 PM

Key Words Vertical cup-disc ratio · Glaucoma · Visual field

Table 1. Demographics

Number Women Age, years Visual acuity, logMAR IOP, mm Hg Spherical equivalent, dpt MD, dB Pattern standard deviation Funduscopic VCDR HRT III VCDR Mean RNFL, μm

Total

NTG

POAG

301 161 (53.5) 70.69 ± 11.68 0.21 ± 0.40 13.05 ± 3.62 –0.29 ± 2.55 –9.13 ± 7.80 6.79 ± 3.32 0.81 ± 0.14* 0.69 ± 0.16* 160 ± 90

140 (46.5) 81 (57.9) 70.57 ± 12.08 0.22 ± 0.47 12.17 ± 3.17 –0.36 ± 2.43 –8.25 ± 7.40 6.76 ± 3.29 0.80 ± 0.12* 0.71 ± 0.13* 170 ± 90

161 (53.5) 80 (49.7) 70.80 ± 11.36 0.21 ± 0.33 13.81 ± 3.82 –0.22 ± 2.67 –9.91 ± 8.09 6.81 ± 3.36 0.82 ± 0.15* 0.68 ± 0.17* 150 ± 90

Difference (p value) 0.157 0.942 0.221 <0.0001 0.677 0.081 0.91 0.051 0.238 0.151

Demographic data are depicted as means with standard deviation for continuous data and numbers with percentages in parentheses for discrete variables. p values for non-parametric comparison between groups depicted on the right. The asterisk indicates a significant difference between funduscopic and HRT VCDR measurements within groups. IOP = Intra-ocular pressure.

MD > –3

–12 < MD ≤ –3 MD ≤ –12

NTG Funduscopic VCDR HRT VCDR

0.71 ± 0.13 0.67 ± 0.10

0.77 ± 0.12 0.68 ± 0.14

0.90 ± 0.09 0.81 ± 0.12

POAG Funduscopic VCDR HRT VCDR

0.66 ± 0.15 0.57 ± 0.15

0.82 ± 0.12 0.69 ± 0.18

0.91 ± 0.07 0.73 ± 0.14

Funduscopic and HRT-derived VCDR for early, moderate and advanced glaucoma. In all 3 groups, both measures differ significantly (p = 0.024 mild NTG, p < 0.001 other).

retinal nerve fibre layer (RNFL) at the time of diagnosis and during the follow-up of glaucoma patients. Clinically, without the help of technical investigations, this morphology assessment is merely limited to an estimate of the VCDR. Confocal and optical coherence tomography scanners can assess various parameters in optic disc morphology more specifically [9]. Furthermore, stereo photographs can be of help to assess disc morphology and stored for future reference [10]. These photos can be assessed by the clinician or analysed by specifically designed software [11]. The relationship and agreement between a clinical subjective (funduscopic) and technological objective assessment were investigated in multiple studies [5, 12–16]. Most focused on glaucoma as a whole but few 2

Ophthalmic Res DOI: 10.1159/000446659

have looked at the possible difference in the relationship between the VCDR and functional glaucomatous damage in low tension versus high tension open angle glaucoma patients [17–19]. Some have found significant differences in optic nerve head morphology, whereas others selected subjects for matching disc morphology and found differences in visual field damage with more extensive visual field defects in primary open angle glaucoma (POAG) patients. Other studies investigated the typical disc excavation in normal tension glaucoma (NTG) and POAG and concluded that NTG patients tend to have more often inferior notching, whereas others found no difference [20, 21]. These variations in morphologic disc changes could have an influence on VCDR measurements. Therefore, this study aims at investigating the difference between a funduscopic and confocal scanner VCDR assessment and their respective prognostic power to estimate the extent of glaucomatous visual field loss in low and high tension open angle glaucoma.

Methods Patient Selection A total of 301 open angle glaucoma patients (161 POAG and 140 NTG) from the Leuven Eye Study were included. The Leuven Eye Study is a large cross-sectional monocentre study investigating the vascular aspects of glaucoma. This study was approved by the Institutional Review Board of the University Hospitals Leuven and adhered to the tenets of the Declaration of Helsinki. All eligible

Willekens et al.

Downloaded by: Univ. of California San Diego 132.239.1.231 - 1/2/2017 7:58:37 PM

Table 2. VCDR as a function of MD

POAG 100

100 p = 0.233

80 p = 0.014

p = 0.003

40 20

Sensitivity

Mild

40 60 80 100 – specificity

CD CDHRT 0

p = 0.245

p = 0.060

Sensitivity

p = 0.895

40 60 80 100 – specificity

100

0 0

40 60 80 100 – specificity

60 40

100

80 p = 0.372

p = 0.732

p = 0.146

40 60 80 100 – specificity

100

40 60 80 100 – specificity

p = 0.0004

CD CDHRT

100

p = 0.558

CD CDHRT

Sensitivity

100

Sensitivity

100

Severe

p = 0.052

CD CDHRT

Moderate

Color version available online

NTG

CD CDHRT 0

40 60 80 100 – specificity

100

Fig. 1. ROC curves. For both diagnostic groups, ROC curves were made for the 3 different levels of glaucomatous

patients who agreed to participate in the study signed an informed consent prior to enrolment. Methodology of inclusion and data gathering are published elsewhere [22]. Demographic data on gender and age as well as ophthalmologic data on intra-ocular pressure, visual acuity, spherical equivalent and funduscopic VCDR assessment of the worst eye were retained for analysis. The VCDR was estimated by a senior glaucoma specialist (I.St./E.V.). Mean deviation (MD) and pattern standard deviation, displayed via the

visual field analysis program peridata (Peridata Software GmbH, Huerth, Germany, http://www.peridata.org), were recorded. Visual fields were tested using a Humphrey field analyser, program 24-2, Sita standard strategy (Carl Zeiss, Oberkochen, Germany) or an Octopus 301, program G1 (30°), dynamic strategy (Interzeag AG, Schlieren, Switzerland). Optic nerve head tomography was performed in all eyes using a Heidelberg retinal tomograph III confocal scanner (HRT III, Heidelberg retina tomography; Heidel-

Vertical Cup-Disc Ratio Assessment

Ophthalmic Res DOI: 10.1159/000446659

Downloaded by: Univ. of California San Diego 132.239.1.231 - 1/2/2017 7:58:37 PM

damage. p values comparing the curves (funduscopic vs. HRT-derived VCDR) are within the graph. p values in between the graphs compare the same measurement technique between the two groups (colour coded according to ROC). CD = Clinical diagnosis; CDHRT = clinical diagnosis after HRT.

Table 3. LRs for different funduscopic VCDRs to predict severe NTG or POAG (MD <–12 dB)

NTG

POAG

criterion

sensitivity

specificity

positive LR

criterion

sensitivity

specificity

positive LR

>0.7 >0.75 >0.8 >0.85 >0.9 >0.95 >1

91.18 91.18 79.41 79.41 41.18 26.47 0

40.78 42.72 74.76 78.64 96.12 100 100

1.54 1.59 3.15 3.72 10.6

>0.7 >0.75 >0.8 >0.85 >0.9 >0.95 >1

96.3 96.3 88.89 85.19 38.89 20.37 0

40 43 64 66 93 98 100

1.6 1.69 2.47 2.51 5.56 10.19

Positive LRs for different cut-offs of funduscopic VCDR in NTG and POAG patients.

Statistical Analysis SPSS 22.0 (IBM SPSS statistics for Windows, Armonk, N.Y., USA) was used to perform normality tests according to Kolmogorov-Smirnov and non-parametric Mann-Whitney U tests for independent samples to compare baseline characteristics between the diagnostic groups. Within groups, the Wilcoxon signed rank test for related samples was performed. Correlations were investigated using the non-parametric Spearman test. Receiver-operating characteristic (ROC) curves according to Hanley and McNeil as well as Bland-Altman plots were constructed using Medcalc (Medcalc Software, Ostend, Belgium), and areas under the curve (AUC) where compared between and within groups [24]. Likelihood ratios (LRs) were calculated. Graphs were made with Graphpad Prism version 5.0 (Graphpad Software Inc., La Jolla, Calif., USA).

Results

A total of 301 eyes from the same number of patients were included. In both groups, data was not distributed normally except for mean RNFL thickness in NTG and visual field pattern standard deviation in POAG. Table 1 depicts the demographic and ocular characteristics. Intraocular pressure was, on average, significantly higher in POAG patients (p < 0.001). The average VCDR assessed by funduscopy was significantly larger than the measurement derived from the HRT software, in both NTG and POAG patients [0.80 vs. 0.71 (p < 0.001) and 0.83 vs. 0.67 (p < 0.001), respectively]. This statistically significant dif4

Ophthalmic Res DOI: 10.1159/000446659

ference remained when analysing per diagnostic group for mild (MD >–3 dB), moderate (–12 dB < MD ≤ –3 dB) and severe (MD ≤–12 dB) glaucomatous damage (table 2). Although significantly different, both measurements are correlated with increased visual field damage in both diagnostic groups (Spearman ρ = 0.48 with p < 0.001 and 0.6 with p < 0.001, for NTG and POAG, respectively). ROC curves were constructed comparing funduscopic VCDR measurements with HRT-derived values for all 3 levels of glaucomatous damage (fig. 1). In NTG patients, no significant difference in AUC was observed between both curves. This finding was confirmed in the POAG patients, except for the advanced glaucoma subgroup (p = 0.0004). When comparing between diagnostic groups, in mild glaucoma, the AUC was significantly lower in NTG compared to POAG, both for funduscopic and HRT-derived VCDR measurements (p = 0.003 and p = 0.01, respectively). The opposite is true for the moderate and advanced glaucoma subgroups where there was no significant difference between NTG and POAG estimates of the AUC for both measurement techniques. LRs were calculated to determine the likelihood of having severe glaucomatous damage depending on various levels of funduscopic and HRT VCDR (tables 3, 4). For both groups the positive LR became higher than 10 for both measurement options, except for the HRT VCDR estimates in POAG patients. Indeed, starting from a VCDR of 0.9 funduscopically or 0.86 on HRT, the posttest probability of having severe glaucomatous damage rises significantly for NTG patients. For POAG patients, however, funduscopic VCDR assessment yielded an increased post-test probability starting only from a VCDR of 0.95, whereas the HRT-derived values were not associated with significant changes in post-test probability.

Willekens et al.

Downloaded by: Univ. of California San Diego 132.239.1.231 - 1/2/2017 7:58:37 PM

berg Engineering, Dossenheim, Germany). The images were analysed using the HRT version 3.0 software. Only images with a standard deviation below 40 μm were included. The analysis technique has previously been described [23]. In brief, the optic disc margin was manually marked by a trained technician at the inner edge of Elschnig’s ring, with the use of stereoscopic optic nerve head photographs. The standard reference plane was used for calculations of optic disc topography. HRT III-derived measurements for VCDR and mean RNFL were included for analysis.

Table 4. LRs for different HRT-derived VCDRs to predict severe NTG or POAG (MD <–12 dB)

NTG

POAG

criterion

sensitivity

specificity

positive LR

criterion

sensitivity

specificity

positive LR

>0.7 >0.71 >0.72 >0.73 >0.74 >0.75 >0.77 >0.78 >0.79 >0.8 >0.83 >0.84 >0.85 >0.86 >0.88 >0.92 >0.93

82.35 82.35 76.47 76.47 73.53 70.59 70.59 64.71 61.76 55.88 55.88 52.94 50 38.24 29.41 17.65 11.76

50.49 54.37 59.22 65.05 70.87 77.67 82.52 84.47 85.44 86.41 92.23 93.2 93.2 97.09 99.03 99.03 100

1.66 1.8 1.88 2.19 2.52 3.16 4.04 4.17 4.24 4.11 7.19 7.79 7.36 13.13 30.29 18.18

>0.7 >0.71 >0.72 >0.73 >0.74 >0.75 >0.76 >0.77 >0.78 >0.8 >0.81 >0.82 >0.83 >0.84 >0.85 >0.86 >0.87 >0.88 >0.89 >0.9 >0.93

72.22 66.67 59.26 53.7 50 46.3 42.59 38.89 31.48 31.48 25.93 24.07 20.37 12.96 11.11 11.11 9.26 5.56 3.7 3.7 1.85

62 64 66 68 70 71 74 76 80 83 85 87 88 89 89 93 94 94 95 97 97

1.9 1.85 1.74 1.68 1.67 1.6 1.64 1.62 1.57 1.85 1.73 1.85 1.7 1.18 1.01 1.59 1.54 0.93 0.74 1.23 0.62

Positive LRs for different cut-offs of HRT III VCDR in NTG and POAG patients.

Discussion

VCDR assessment is one of the most important cornerstones in glaucoma diagnosis and follow-up, since it serves as an indicator for structural and functional glaucomatous damage [25, 26]. Several studies examined the agreement between subjective and objective VCDR assessment [5, 11–15]. In general, the value obtained from objective measurement is smaller than the clinically assessed VCDR, but some found the opposite to be true [13]. Our results confirm the larger VCDR as assessed by funduscopy compared to HRT for every stage of open angle glaucoma, regardless of the initial intra-ocular pressure at diagnosis. Since both are estimates of the same ratio and were shown to be significantly correlated, the question arises which of both is more powerful in predictVertical Cup-Disc Ratio Assessment

ing glaucomatous damage. Consequently, ROCs were constructed for 3 levels of disease severity in both diagnostic groups. These showed that, within diagnostic groups, there was either no difference in diagnostic power (NTG) or the AUC was in favour of the clinical VCDR estimate (POAG). In POAG rim loss is generally more diffuse than in NTG where inferior and superior notching is more prevalent [27]. Therefore, an objective measurement of the cup-disc ratio along the vertical axis might underestimate the visual field damage expressed as the mean deviation in POAG. During subjective VCDR assessment, glaucoma specialists will also take into account focal rim loss, nasal thinning, disc pallor and even RNFL defects which may contribute to a larger subjective VCDR estimate that is more congruent with the functional damage. This resulted in a significantly larger positive predictive power for severe glaucomatous damage in POAG, but not in NTG. Apparently, the functional damage in NTG corresponds better to rim loss along the vertical axis than in POAG. In moderate glaucoma, both measurements seem to lose diagnostic power, and more so in POAG. Apparently both the clinician and the confocal scanner have trouble estimating glaucomatous damage in Ophthalmic Res DOI: 10.1159/000446659

Downloaded by: Univ. of California San Diego 132.239.1.231 - 1/2/2017 7:58:37 PM

Bland-Altman plots for NTG and POAG showed significantly lower values for HRT-derived VCDR than the funduscopic measurements in both groups, with a progressively larger difference in patients with advanced glaucomatous damage (p < 0.0001) (fig. 2).

0.6

0.6 +1.96 SD

0.4

0.35

0.2

Mean 0.09

+1.96 SD

–1.96 SD

–0.2

0.44

0.2

Mean 0.14

–1.96 SD

–0.2

–0.17

–0.16

–0.4

0.4

Color version available online

0.8

CD – CDHRT

0.8

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

Fig. 2. Bland-Altman plots to compare both VCDR measurement techniques within groups. Bland-Altman plots

displaying the difference between the HRT VCDR and funduscopic VCDR with the latter as reference measurement (p < 0.0001 for both). CD = Cup/disc fundoscopic; CDHRT = cup/disc HRT. a NTG. b POAG.

Ophthalmic Res DOI: 10.1159/000446659

in mild glaucoma, both the subjective and the objective measurement possessed significantly more predictive power in POAG than in NTG patients. With more advanced glaucoma, this difference disappeared. These results are reflected in the LRs, which increase substantially in both groups with severe glaucoma, resulting in a posttest probability of having severe (MD ≤–12 dB) visual field damage of at least 45%, according to LR conversion following McGee [33]. Surprisingly, where the HRT VCDR estimate derives the highest LR in the NTG group, it is, in contrast, significantly outperformed by the clinical estimate in POAG patients. In this diagnostic subgroup, and for patients with severe glaucomatous damage, one should therefore rely more on a clinical VCDR determination than on an HRT III-derived measurement. This study is limited because of the comparison of only 2 means of estimating VCDR.

Conclusion

Funduscopic and confocal scanner (HRT)-derived measurements for VCDR in glaucoma patients differ significantly within patients and should not be used interchangeably in clinical practice or studies. They perform equally well in predicting functional glaucomatous damage, except in POAG patients with severe glaucoma, where the confocal scanner was outperformed by the clinical assessment. Willekens et al.

Downloaded by: Univ. of California San Diego 132.239.1.231 - 1/2/2017 7:58:37 PM

between early and severe glaucoma. It seems important that, particularly for a patient in this stage of the disease, a comprehensive approach and close follow-up with functional testing are mandatory. Bland-Altman plots confirmed this difference between the measurements and revealed that the greater the glaucomatous damage, the greater the difference in VCDR estimation. Although this did not result in any significant difference in predicting power in NTG patients with severe glaucomatous damage, the opposite is true in the severe glaucoma POAG group where the clinical estimate seems to be more reliable than the objective method. This finding is also reflected in the high LRs achieved with subjective VCDR assessment in contrast to low values for HRT VCDR. Besides comparing the performance in predicting glaucomatous damage for both VCDR estimates, it is also important to investigate whether there is a difference when assessing patients with NTG or POAG. Since the pathophysiology of both subtypes of open angle glaucoma might differ, one might expect a difference in rim loss pattern and typical visual field deterioration [28]. For instance, the appearance of optic nerve head haemorrhages has been suggested as a significant risk factor for NTG progression [29]. There is conflicting evidence on possible differences in the pattern of visual field change accompanying glaucomatous progression [30]. Researchers have speculated on a possible difference in lamina cribrosa thickness and pore configuration in NTG versus POAG [31, 32]. When comparing between both groups,

Acknowledgements

Disclosure Statement

The authors would like to thank Ellen Rossou and Isabel Van den Abeele for their contribution in data gathering.

The authors have no conflict of interest or any relevant sources of funding.

References

Vertical Cup-Disc Ratio Assessment

12 Schuman JS, Wollstein G, Farra T, Hertzmark E, Aydin A, Fujimoto JG, Paunescu LA: Comparison of optic nerve head measurements obtained by optical coherence tomography and confocal scanning laser ophthalmoscopy. Am J Ophthalmol 2003;135:504–512. 13 Hatch WV, Trope GE, Buys YM, Macken P, Etchells EE, Flanagan JG: Agreement in assessing glaucomatous discs in a clinical teaching setting with stereoscopic disc photographs, planimetry, and laser scanning tomography. J Glaucoma 1999;8:99–104. 14 Varma R, Spaeth GL, Steinmann WC, Katz LJ: Agreement between clinicians and an image analyzer in estimating cup-to-disc ratios. Arch Ophthalmol 1989;107:526–529. 15 Arthur SN, Aldridge AJ, De León-Ortega J, McGwin G, Xie A, Girkin CA: Agreement in assessing cup-to-disc ratio measurement among stereoscopic optic nerve head photographs, HRT II, and Stratus OCT. J Glaucoma 2006;15:183–189. 16 Durmus M, Karadag R, Erdurmus M, Totan Y, Feyzi HI: Assessment of cup-to-disc ratio with slit-lamp funduscopy, Heidelberg Retina Tomography II, and stereoscopic photos. Eur J Ophthalmol 2008;19:55–60. 17 Jonas JB, Gründler AE: Correlation between mean visual field loss and morphometric optic disk variables in the open-angle glaucomas. Am J Ophthalmol 1997;124:488–497. 18 Mikelberg FS, Parfitt CM, Swindale NV, Graham SL, Drance SM, Gosine R: Ability of the Heidelberg retina tomograph to detect early glaucomatous visual field loss. J Glaucoma 1995;4:242–247. 19 Jayasundera T, Danesh-Meyer HV, Donaldson M, Gamble G: Agreement between stereoscopic photographs, clinical assessment, Heidelberg retina tomograph and digital stereoscopic optic disc camera in estimating vertical cup:disc ratio. Clin Exp Ophthalmol 2005;33:259–263. 20 Caprioli J, Spaeth GL: Comparison of the optic nerve head in high- and low-tension glaucoma. Arch Ophthalmol 1985; 103: 1145– 1149. 21 Iester M, Mikelberg FS: Optic nerve head morphologic characteristics in high-tension and normal-tension glaucoma. Arch Ophthalmol 1999;117:1010–1013.

Ophthalmic Res DOI: 10.1159/000446659

22 Abegão Pinto L, Willekens K, Van Keer K, Shibesh A, Molenberghs G, Vandewalle E, Stalmans I: Ocular blood flow in glaucoma – the Leuven Eye Study. Acta Ophthalmol 2016, Epub ahead of print. 23 Weinreb R, Dreher A, Bille J: Quantitative assessment of the optic nerve head with the laser tomographic scanner. Int Ophthalmol 1989; 13:25–29. 24 Hanley JA, McNeil BJ: The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 1982;143:29– 36. 25 Gordon MO, Beiser JA, Brandt JD, Heuer DK, Higginbotham EJ, Johnson CA, Kass MA: The Ocular Hypertension Treatment Study: baseline factors that predict the onset of primary open-angle glaucoma. Arch Ophthalmol 2002;120:714–720. 26 Jonas JB, Budde WM, Panda-Jonas S: Ophthalmoscopic evaluation of the optic nerve head. Surv Ophthalmol 1999;43:293–320. 27 Eid TE, Spaeth GL, Moster MR, Augsburger JJ: Quantitative differences between the optic nerve head and peripapillary retina in lowtension and high-tension primary open-angle glaucoma. Am J Ophthalmol 1997; 124: 805– 813. 28 Shields MB: Normal-tension glaucoma: is it different from primary open-angle glaucoma? Curr Opin Ophthalmol 2008; 19: 85– 88. 29 Martus P, Stroux A, Budde WM, Mardin CY, Korth M, Jonas JB: Predictive factors for progressive optic nerve damage in various types of chronic open-angle glaucoma. Am J Ophthalmol 2005;139:999–1009. 30 Araie M: Pattern of visual field defects in normal-tension and high-tension glaucoma. Curr Opin Ophthalmol 1995;6:36–45. 31 Park HYL, Jeon LSH, Park CK: Enhanced depth imaging detects lamina cribrosa thickness differences in normal tension glaucoma and primary open-angle glaucoma. Ophthalmology 2012;119:10–20. 32 Miller KM, Quigley HA: Comparison of optic disc features in low-tension and typical openangle glaucoma. Ophthalmic Surg 1987; 18: 882–889. 33 McGee S: Simplifying likelihood ratios. J Gen Intern Med 2002;17:647–650.

Downloaded by: Univ. of California San Diego 132.239.1.231 - 1/2/2017 7:58:37 PM

1 Weinreb RN, Aung T, Medeiros FA: The pathophysiology and treatment of glaucoma: a review. JAMA 2014;311:1901–1911. 2 Bourne RR, Jonas JB, et al: Prevalence and causes of vision loss in high-income countries and in Eastern and Central Europe: 1990– 2010. Br J Ophthalmol 2014;98:629–638. 3 Jampel HD: Target pressure in glaucoma therapy. J Glaucoma 1997;6:133–138. 4 Tielsch JM, Katz J, Quigley HA, Miller NR, Sommer A: Intraobserver and interobserver agreement in measurement of optic disc characteristics. Ophthalmology 1988;95:350–356. 5 Zangwill L, Shakiba S, Caprioli J, Weinreb RN: Agreement between clinicians and a confocal scanning laser ophthalmoscope in estimating cup/disk ratios. Am J Ophthalmol 1995;119:415–421. 6 Springelkamp H, Lee K, Wolfs RC, Buitendijk GH, Ramdas WD, Hofman A, Jansonius NM: Population-based evaluation of retinal nerve fiber layer, retinal ganglion cell layer, and inner plexiform layer as a diagnostic tool for glaucoma. Invest Ophthalmol Vis Sci 2014; 55:8428–8438. 7 Lalezary M, Medeiros FA, Weinreb RN, Bowd C, Sample PA, Tavares IM, Zangwill LM: Baseline optical coherence tomography predicts the development of glaucomatous change in glaucoma suspects. Am J Ophthalmol 2006;142:576–582. 8 Mohammadi K, Bowd C, Weinreb RN, Medeiros FA, Sample PA, Zangwill LM: Retinal nerve fiber layer thickness measurements with scanning laser polarimetry predict glaucomatous visual field loss. Am J Ophthalmol 2004;138:592–601. 9 Varma R, Steinmann WC, Scott IU: Expert agreement in evaluating the optic disc for glaucoma. Ophthalmology 1992;99:215–221. 10 Medeiros FA, Zangwill LM, Bowd C, Weinreb RN: Comparison of the GDx VCC scanning laser polarimeter, HRT II confocal scanning laser ophthalmoscope, and Stratus OCT optical coherence tomograph for the detection of glaucoma. Arch Ophthalmol 2004; 122: 827– 837. 11 Pena-Betancor C, Gonzalez-Hernandez M, Fumero-Batista F, Sigut J, Medina-Mesa E, Alayon S, de la Rosa MG: Estimation of the relative amount of hemoglobin in the cup and neuroretinal rim using stereoscopic color fundus images. Invest Ophthalmol Vis Sci 2015;56:1562–1568.