The Optic Nerve Head in Myocilin Glaucoma

Abstract

purpose. Approximately 1 in 30 unselected patients with open-angle glaucoma (OAG) have a mutation in the myocilin gene. The purpose of this study was to describe the morphologic features of the optic nerve head (ONH) in myocilin glaucoma.

methods. A case-control design was adopted. Sixty-six patients heterozygous for a range of myocilin mutation (cases) were matched in disease severity to 105 patients with OAG known not to have a myocilin mutation (controls), using visual field findings. Quantifiable analysis of the ONH was undertaken of stereoscopic photographs, by using custom software with a z-screen. Subjective grading of the cup depth, lamina cribrosa pore shape and orientation, and the slope of the neuroretinal rim was performed by an examiner masked to the subject’s mutation status. Mutation screening was conducted using either direct sequencing or single-stranded conformation polymorphism analysis.

results. Patients with a myocilin mutation had glaucoma diagnosed earlier (P < 0.001) and had higher maximum recorded intraocular pressures (P < 0.001) than did the control OAG subjects. There was no significant (P > 0.05) difference in global disc area, global neuroretinal rim area, α-parapapillary atrophy, β-parapapillary atrophy, slope of neuroretinal rim, or visible lamina cribrosa morphology between myocilin mutation carriers and patients with nonmyocilin glaucoma. Disc hemorrhages were identified more frequently in those without mutations (14/209 vs. 1/129), though this was not significant after correction for multiple hypothesis testing.

conclusions. No major structural or morphologic difference of the ONH was detected in pooled data from subjects who had myocilin mutations compared with data from individuals with nonmyocilin glaucoma.

Primary open-angle glaucoma (POAG) is the leading cause of optic neurodegeneration worldwide. ¹ Progressive retinal ganglion cell apoptosis, the hallmark of this complex heterogeneous disease, principally manifests through changes in the optic nerve head (ONH). ^{2 3} Approximately 1 in 30 unselected cases of POAG have a myocilin (MYOC) gene mutation. ⁴ More than 50 MYOC mutations have been described, predominantly in the gene’s olfactomedin domain of exon 3. ⁵ Although MYOC is expressed at the ONH, evidence suggests that pathogenic mutations cause dysfunctional secretion of the MYOC protein in trabecular meshwork cells, thereby inhibiting homeostatic aqueous filtration. ^{6 7 8 9} Although other purported mechanisms include abnormal interaction between MYOC and extracellular matrix or cell surface proteins. ¹⁰  

MYOC-related glaucoma is predominantly associated with elevated intraocular pressure (IOP), ¹¹ and there are strong phenotype–genotype correlations with the different MYOC mutations. ² Worldwide, the most prevalent MYOC mutation is Gln368Stop, which appears to have arisen from a common founder. ^{4 5 12} Patients with this specific mutation typically have glaucoma diagnosed in the early fifth decade and have a mean peak IOP of 31 mm Hg, and approximately 30% of carriers undergo trabeculectomy. ^{11 13 14} The other MYOC mutations generally manifest as more severe disease. For example, individuals with the Pro370Leu mutation have a mean age at diagnosis of 10 years, and a mean peak IOP of 45 mm Hg, with trabeculectomy being performed in almost all mutation carriers. ^{2 15 16} Additional mutations, such as Thr377Met, confer disease of intermediate severity. Patients with the Thr377Met MYOC mutation typically have a mean age at diagnosis of 37 years and a mean peak IOP of 35 mm Hg, and approximately 56% undergo trabeculectomy. ^{11 16 17 18 19}  

The ability to anticipate accurately the likely natural history or clinical progression in a patient would significantly improve therapeutic algorithms and thus result in enhanced visual preservation. The usefulness of molecular diagnosis may be lessened if discerning features are detectable at the slit lamp. The purpose of this study was to describe the morphologic features of the ONH in MYOC glaucoma, compared with those in severity-matched patients with POAG known not to have MYOC mutations. In particular, the question of whether MYOC expression in the ONH has any functional role, has not been addressed, and discernible differences in the disc appearance would be relevant for future investigation. 

Methods

A case–control design was adopted, whereby case subjects represented patients with an MYOC mutation and control subjects were patients with POAG who did not have an MYOC mutation. An MYOC disease-causing mutation was defined as one that altered the predicted amino acid sequence and had previously been found to be consistently more common in glaucomatous cases than in age-matched normal individuals. ¹¹ A total of 66 case subjects from 25 genealogically separate pedigrees were identified. These cases were matched by visual field severity to 105 control patients who had POAG without MYOC mutations. This study was approved by the relevant ethics committees of the Royal Victorian Eye and Ear Hospital and the Royal Hobart Hospital. Written informed consent was obtained from each subject, and the study was conducted in accordance with the Declaration of Helsinki and subsequent revisions. 

Glaucoma was defined by the presence of, in at least one eye, visual field loss, with corresponding optic disc cupping (cup–disc ratio ≥ 0.7); by a 0.2 intereye disparity in cup–disc ratio; or by focal rim notching. For inclusion, subjects had to have an abnormal visual field as graded by the Glaucoma Hemifield Test (Humphrey Field Analyzer II; Carl Zeiss Meditec, Inc., Dublin, CA). Mean deviation scores in the most recent, reliable visual field test were used to match patients with glaucoma without MYOC mutations (control subjects) to mutation-carrying subjects. 

All recruited subjects underwent a comprehensive clinical examination that included anterior segment examination, gonioscopy, IOP measurement by Goldmann applanation tonometry, visual field assessment, refraction, and mydriatic optic disc assessment. Highly myopic eyes (refraction exceeding −7 D) were excluded because of confounding myopia-related ONH appearance, including optic disc tilting and tessellation of vessels. Color 35-mm slides of the ONH were taken with a nontelecentric fundus camera (3-Dx/F; Nidek, Gamagori, Japan). The resultant simultaneous stereoscopic images were digitalized at a high resolution (2102 × 1435 pixels, 2900 ppi, 8-bit color) with a slide scanner (CoolScan IV ED; Nikon Corp., Tokyo, Japan). 

Preliminary ONH quantification was performed stereoscopically with custom software with a z-screen (Real D Corp. Beverly Hills, CA) by a grader masked to the subject’s mutation status. ^{20 21 22} This stereoscopic system has been described in detail elsewhere. ^{20 21 22} In brief, it consists of a normal CRT monitor and an overlying high speed modulating panel. Flicker-free stereoscopy, achieved by alternatively displaying the component images of the stereo pair on the monitor at 60 Hz, can be viewed with polarized glasses. Cursor depth may then be adjusted to coincide with Elschnig’s rim, such that the neuroretinal rim as well as the inner margin of the disc can be outlined at the depth of the scleral plane. We corrected for image magnification by using keratometry readings, refraction, and camera specifications, by using established methods, to provide scaled estimates of disc parameters. ^{20 21 22}  

Using this custom software, the optic disc size, neuroretinal rim area, maximum disc and cup diameters, length of the central retinal vessel trunk along the floor of the cup, and the size of any disc hemorrhages, as well as parapapillary atrophy, were quantified. Parapapillary atrophy was differentiated into a central β zone demarcated by visible sclera and large choroidal vessels close to the optic disc border and a peripheral α zone with irregular pigmentation. ²³ A neuroretinal rim notch was defined as a 60° arc of disc, in the center of which the neuroretinal rim was thinner than two thirds of the rim width at both peripheral borders of this disc sector. Disc and cup ovality was determined by the ratios of their respective maximum vertical and horizontal diameters. 

Using a stereo viewer (Stereo Viewer-II; Pentax Stereo Viewer-II Pentax Imaging Company, Golden, CO) the color 35-mm slides, with all patient identifying information removed, were subjectively graded. The central retinal artery entry site was categorized as being: 1, far nasal; 2, midnasal; 3, central; 4, midtemporal; or 5, far temporal. Cup depth was graded in a Likert range between 0 for no cupping and 5 for very deep cupping. ²⁴ The slope in most (>50%) of the neuroretinal rim was scaled as being 1 for very shelved; 2 for shelved; 3 if vertical; and 4 when undermined. When undermined, the slope of the remaining neuroretinal rim was graded 1 to 3. The presence of nerve fiber layer defects and baring of vessels were also noted. When visible, the lamina cribrosa appearance was recorded. According to the description of Miller and Quigley, lamina pore shapes were described as being: 1, round or dotlike; 2, polygonal; 3, oval; and 4, striate or slitlike. ²⁵ The configuration of the laminar pores were also characterized as being: 1, circumferential; 2, radial or spiral; 3, hourglass; and 4, random or disorganized. ²⁵  

The stereo disc photographs of the ONH of each subject were reviewed by an observer (SLB) masked to mutation status and clinical parameters. To determine the reproducibility and internal validity of ONH grading, 43 (12.7%) randomly selected optic disc photographs were analyzed twice. The κ values for the central retinal artery entry site, slope of the neuroretinal rim, and cup depth were all greater than 0.85; however, they were 0.69 and 0.62 for lamina cribrosa pore shape and orientation, respectively. 

Laboratory Techniques

Mutation screening was conducted by either direct sequencing or single-stranded conformation polymorphism (SSCP) analysis. Genomic DNA was isolated from peripheral blood samples, and the coding regions of MYOC were amplified with previously published oligonucleotide primers. ¹¹ In preparation for SSCP, PCR products were denatured for 3 minutes at 94°C and, after electrophoresis, were stained with silver nitrate for review. Mutations detected by SSCP were subsequently confirmed by sequencing. Sequencing reactions were performed by using dye termination chemistry (Big Dye Terminator kit; Applied Biosystems, Scoresby, Australia), with 25 cycles of 10 seconds at 95°C, 5 seconds at 50°C, and 4 minutes at 60°C, as specified by the manufacturer. Analysis was performed with a genetic analyzer (Prism 310; Applied Biosystems) the resultant outputs were reviewed (Sequencher; Gene Codes Corp., Ann Arbor, MI). 

Data Analysis

The presence of MYOC-specific ONH features were investigated through comparing both the “better” and “worse” eyes of case and control subjects. The worse eye was determined by mean deviation in the most recent reliable visual field test. Case subjects were analyzed on a pooled and mutation-specific basis to assess genotype-phenotype correlations. 

Statistical analysis was performed (Intercooled Stata 7.0 for Windows; Stata Corp., College Station, TX), with the Student’s t-test used for parametric data and the Kruskal-Wallis and Mann-Whitney tests used to determine significant differences in nonparametric data. Differences in categorical proportions were tested with the χ² test. The Bonferroni correction was used to account for multiple hypothesis testing. Power calculations were performed with the PS program version 1.0.17 for Windows. ²⁶ Unless otherwise indicated, data are presented as the mean ± SD. 

Results

A total of 66 patients with an MYOC mutation (Gln368Stop, n= 38; Thr377Met, n = 17; Gly252Arg, n = 6; Pro370Leu, n = 4; Asp380Gly, n = 1) were matched by visual field findings to 105 patients known not to have an MYOC mutation. Four eyes of four subjects (three cases and one control) were excluded from analysis because of poor fundus imaging. 

Patients with an MYOC mutation were diagnosed earlier (P < 0.001) and had higher maximum recorded IOP (P < 0.001) than did control patients without MYOC mutations (Table 1) . Case subjects with the Gln368Stop mutation had a later diagnosis (53.8 ± 12.9 years) than those with any other MYOC mutations (P < 0.001). Case subjects with Pro370Leu had the lowest age at diagnosis (15.7 ± 9.8 years). There was a stepwise increase in mean maximum recorded IOP among cases with the Gln368Stop, Thr377Met, Gly252Arg, and Pro370Leu MYOC mutations (lowest to highest, respectively; Kruskal-Wallis P < 0.001). The subject with the Asp380Gly mutation was aged 28 years at diagnosis and had a maximum recorded IOP of 44 mm Hg. 

The MYOC mutation-carrying group and the MYOC mutation free group did not vary significantly in optic disc area, neuroretinal rim area, steepness and depth of disc cupping, and optic cup shape, as determined by the ratio of vertical-to-horizontal cup diameters, and the size of β- and α-zones of parapapillary atrophy (Tables 2 3) . There was no significant difference in visible lamina cribrosa morphology between mutation carriers and non-MYOC patients. Disc hemorrhages were identified more frequently in the control group (P = 0.01), but this finding was not significant after correction for multiple hypothesis testing (Table 4) . Our cohort afforded 90% power to detect a 10-fold increased prevalence of disc hemorrhages at the 0.05 significance level. In all eyes examined, MYOC mutation carriers had more neuroretinal rim notches than did control subjects; once again, however, the difference was not significant after Bonferroni correction (Table 4) . 

There was a significant stepwise increase in neuroretinal rim area to optic disc area ratios between cases with the Gly252Arg, Thr377Met, and Gln368Stop MYOC mutations (Kruskal-Wallis, P = 0.003). However, this trait did not differ significantly between Gln368Stop MYOC-mutation carriers and patients with glaucoma without MYOC mutations (P = 0.43). Subanalysis revealed no other significant MYOC mutation–specific morphologic characteristics (Fig. 1) . 

Discussion

No major morphologic differences in the ONH of glaucomatous MYOC mutation carriers compared with non-MYOC patients, matched for visual field severity, were identified. This work supports that of Alward et al., ¹¹ that maximum recorded IOP and age at diagnosis may be the best discerning features of MYOC glaucoma. ONH damage is the final common endpoint for glaucoma, and our study underscores the fact that cases with different primary etiologies may have indistinguishable endpoints. Molecular diagnosis remains the best means of anticipating the likely natural history and disease progression in approximately 4% of unselected patients with glaucoma and their family members. ⁴  

MYOC protein of patients with the Gly252Arg, Gln368Stop, Pro370Leu, and Thr377Met MYOC mutations are known to be Triton assay insoluble, ^{16 27} and firm genotype–phenotype correlations for age at diagnosis, maximum recorded IOP, and ratio of neuroretinal rim area to optic disc area were identified. The age of diagnosis and peak IOP in case subjects with the Gln368Stop, Pro370Leu, or Thr377Met MYOC mutations did not differ from those previously reported. ^{11 13 15 16 17 18} The mean age of onset of POAG in our Gly252Arg cases is greater than that previously presented. ^{16 28} The Asp380Gly MYOC mutation has been identified in a single patient with juvenile-onset open-angle glaucoma (JOAG), and the amino acid substitution has a Blosum62 matrix score of −1, indicating a low tolerance for this particular exchange during natural selection. ^{29 30} A different nonsynonymous change to alanine at codon 380 is known to render the MYOC protein insoluble. ³¹  

Analysis of the ONH slides suggested a lower rate of disc hemorrhages in MYOC carriers, and although this finding was not significant after correction for multiple hypothesis testing, it should be noted that the power of our study to detect a significant change is limited. Of note, earlier work has suggested that optic disc hemorrhages are less common in patients with JOAG; however, the analysis of this work may have been biased through not adjusting for age. ³² Nevertheless, a similar large review of patients with either high or normal-pressure POAG has also found optic disc hemorrhages to be more frequently associated with the latter group. ³³  

The finding that neuroretinal rim notches were more common in MYOC mutation carriers than in non-MYOC patients was unexpected. Once again, however, this finding was not significant after correction for multiple testing. Jonas and Budde ³² observed that neuroretinal rim notching was more common in patients with JOAG than in those with normal-pressure POAG. Given that a subset of MYOC mutations are known to cause JOAG, we expected to reach a similar conclusion. Using red-free fundus photographs, Jonas and Budde found that, when present, localized retinal nerve fiber layer defects were narrower in subjects with JOAG than in patients with normal-pressure POAG. The number of nerve fiber layer defects identified in our cohort is relatively low and may reflect the inherent difficulty in identifying them from color slides. ³⁴  

Extracellular matrix remodeling at the ONH occurs due to elevated IOP. ³⁵ Fibroblast activation and expression of matrix metalloproteinases may alter the laminar pore shape and orientation. Miller and Quigley ²⁵ have found that patients with high IOP-related POAG are more likely to have an hourglass appearance of connective tissue bundles at the lamina cribrosa. Although grading of many of the qualitative ONH traits had excellent reproducibility, lamina cribrosa pore shape and orientation were found to have substantial retest agreement (κ > 0.60). This finding reflects the inherent difficulty in the subjective categorization of lamina characteristics. Despite our study’s being similar in size to that conducted by Miller and Quigley, no significant difference in the clinically visible laminar cribrosa architecture between MYOC cases and non-MYOC control subjects with glaucoma was noted. We corroborate the finding of Healey and Mitchell ³⁶ that lamina cribrosa pore visibility is greater in eyes with larger cup–disc ratios. The findings in our study do not fully preclude other fundamental differences in the composition of extracellular matrix remodeling between patients with MYOC or non-MYOC glaucoma, and the pattern of preglaucoma IOP spiking in those with MYOC mutation requires further investigation. The increased density of mitochondria in the prelaminar regions of the optic nerve has been found primarily to facilitate the higher energy requirements for electrical conduction in this unmyelinated region. ³⁷ Hence, additional work is warranted to investigate the impact that various degrees of oxidative stress and mitochondria dysfunction in patients with POAG may play in contributing to the ONH phenotype. ³⁸  

Caprioli and Spaeth ³⁹ found that, despite having similar mean total visual field loss, patients with normal-pressure POAG tend to have smaller neuroretinal rim areas, particularly in the inferotemporal regions, than do patients with high-pressure POAG. Given this finding they suggested that the ONH appearance may be useful in differentiating subgroups of POAG. We found that neuroretinal rim tissue was not preferentially lost in any particular region in MYOC cases compared with non-MYOC cases (data not shown). In addition, the site of central artery entry and length of the vessel trunk along the floor of the cup, a direct surrogate for vessel bayoneting, did not differ between case and control subjects. After adjusting for disease severity, Tezel et al. ³³ concluded that the clinical appearance of the ONH did not differ between patients with normal or high-pressure POAG. 

Many of the mutation-carrying case subjects were from the same pedigree. A full description of the general clinical features and the overall pedigree structure for some of the Gln368Stop and Thr377Met MYOC mutation cases analyzed as part of this study’s cohort have been presented previously. ^{14 19} Randomly selecting only one affected case from each pedigree may be the most rigorous means for investigation; however, the small number available would significantly limit the ability to detect significance in such analysis. The power to detect signature morphologic features in specific MYOC mutations uncommon in our population, such as Pro370Leu, was limited. Any potential bias introduced through adherence to therapy was minimized by matching the cases to control subjects by disease severity. 

In summary, no major structural or morphologic difference of the ONH could be clearly detected in pooled subjects who had an MYOC mutation when compared to subjects in non-MYOC glaucoma cases. Nonetheless, quantitative trait analysis of specific ONH traits may prove useful in the identification of novel POAG-related genes. ⁴⁰ Longitudinal analysis of the ONH in MYOC cases may reveal specific, though subtle, characteristics that are relevant to the natural history of MYOC-related optic cup excavation and glaucomatous damage. 

 Supported by National Health and Medical Research Council (NHMRC) Project Grant 229960, the Clifford Craig Medical Research Trust, the Ophthalmic Research Institute of Australia, and Glaucoma Australia. AWH is the recipient of an NHMRC Medical Postgraduate Scholarship, JEC is supported in part by an NHMRC Practitioner Fellowship, and DAM is supported by a Pfizer Australia Research Fellowship.

 Submitted for publication June 6, 2006; revised August 17, 2006; accepted November 10, 2006.

 Disclosure: A.W. Hewitt, None; S.L. Bennett, None; J. H. Fingert, None; R.L. Cooper, None; E.M. Stone, None; J.E. Craig, None; D.A. Mackey, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Composition of the Study Groups

	MYOC Glaucoma	Non-MYOC Glaucoma	P
n	66	105
Gender (F/M)	36/30	67/38	0.228 (NS)
Age at diagnosis (y)	46.2 ± 15.1; 9–83	60.0 ± 11.3; 30–82	<0.0001
Age at review (y)	60.6 ± 16.5; 16–96	68.4 ± 10.5; 41–93	<0.0001
Maximum recorded IOP (mm Hg)	30.8 ± 10.1; 17–60	23.2 ± 6.4; 15–44	<0.0001
Refractive error (D)	−0.12 ± 1.4; −6.19–+3.5	0.04 ± 1.6; −5.25–+6.13	0.522 (NS)

 Data are the mean ± SD; range. MD, mean deviation.

Table 2.

 

Optic Nerve Head Characteristics of Subject’s Worse Eye, as Determined by Visual Field Mean Deviation

	MYOC Glaucoma	Non-MYOC Glaucoma	P *
n	66	105
Visual field severity (MD in dB)	−10.73 ± 10.19; −31.65–−1.35	−8.90 ± 8.45; −30.80–−1.10	0.217
Optic disc size (mm2)	2.13 ± 0.47; 1.25–13.33	2.14 ± 0.45; 1.02–13.17	0.892
Neuroretinal rim area (mm2)	1.04 ± 0.42; 0.25–12.21	1.14 ± 0.37; 0.28–12.59	0.088
Neuroretinal rim/optic disc area	0.42 ± 0.21; 0.06–10.91	0.46 ± 0.17; 0.10–10.86	0.236
ONH Shape
Disc ovality	1.09 ± 0.07; 0.94–11.32	1.10 ± 0.08; 0.89–11.31	0.243
Cup ovality	1.16 ± 0.18; 0.81–11.71	1.18 ± 0.19; 0.81–12.24	0.497
Optic disc cupping
Steepness	3.01 ± 1.00; 1–14	2.97 ± 0.99; 1–14	0.779
Depth	4.00 ± 0.97; 1–15	3.95 ± 1.14; 1–15	0.922
Lamina cribrosa architecture
n gradable (%)	28 (42.4)	38 (36.2)	0.415
Lamina pore shape	2.18 ± 1.12; 1–14	2.14 ± 1.12; 1–14	0.904
Lamina pore orientation	3.00 ± 1.32; 1–14	3.13 ± 1.24; 1–14	0.713
Parapapillary atrophy
β zone (mm2)	0.20 ± 0.66; 0–15.09	0.27 ± 0.61; 0–14.74	0.487
α zone (mm2)	0.88 ± 0.72; 0–13.72	1.03 ± 0.71; 0–14.15	0.180
Site of central retinal artery entry	2.41 ± 0.63; 1–14	2.47 ± 0.62; 1–14	0.496
Length of vessel trunk along floor of cup (mm)	0.26 ± 0.21; 0–10.78	0.24 ± 0.20; 0–10.96	0.548

 Data are the mean ± SD; range.

*  All are nonsignificant.

Table 3.

 

Optic Nerve Head Characteristics of Subject’s Better Eye, as Determined by Visual Field Mean Deviation

	MYOC Glaucoma	Non-MYOC Glaucoma	P *
n	63	104
Visual field severity (MD in dB)	−6.45 ± 8.81; −30.98–2.02	−4.70 ± 6.34; −27.82–2.62	0.144
Optic disc size (mm2)	2.19 ± 0.53; 1.39–3.74	2.15 ± 0.53; 1.15–3.57	0.640
Neuroretinal rim area (mm2)	1.15 ± 0.42; 0.38–2.34	1.21 ± 0.39; 0.42–2.51	0.413
Neuroretinal rim/optic disc area	0.46 ± 0.18; 0.12–0.82	0.49 ± 0.18; 0.13–0.90	0.339
ONH Shape
Disc ovality	1.12 ± 0.09; 0.86–1.36	1.10 ± 0.07; 0.91–1.26	0.301
Cup ovality	1.19 ± 0.21; 0.82–1.79	1.19 ± 0.22; 0.71–1.85	0.786
Optic disc cupping
Steepness	2.94 ± 0.92; 1–4	2.79 ± 0.98; 1–4	0.341
Depth	3.90 ± 0.98; 0–5	3.79 ± 1.09; 0–5	0.564
Lamina cribrosa architecture
n gradable (%)	25 (39.7)	30 (28.9)	0.149
Lamina pore shape	1.76 ± 1.20; 1–4	1.93 ± 1.01; 1–4	0.564
Lamina pore orientation	3.59 ± 0.85; 1–4	3.18 ± 1.17; 1–4	0.183
Parapapillary atrophy
β zone (mm2)	0.19 ± 0.35; 0–1.68	0.25 ± 0.62; 0–3.28	0.439
α zone (mm2)	0.92 ± 0.72; 0–3.29	1.09 ± 0.75; 0–4.84	0.161
Site of central retinal artery entry	2.40 ± 0.59; 1–4	2.34 ± 0.54; 1–4	0.522
Length of vessel trunk along floor of cup (mm)	0.22 ± 0.17; 0–0.73	0.20 ± 0.22; 0–1.00	0.581

 Data are the mean ± SD; range.

*  All are nonsignificant.

Table 4.

 

Disc Hemorrhages, Neuroretinal Rim Notching, Nerve Fiber Layer Defects, and Vessel Baring in All Eyes Examined

	MYOC Glaucoma	Non-MYOC Glaucoma	P
n of eyes	129	209
Disc hemorrhages
- Frequency (%)	1 (0.8)	14 (6.7)	0.010
- Size (mm2)	0.02	0.10 ± 0.07; 0.06–0.14	0.165 (NS)
Frequency of vessel baring (%)	70 (45.2)	115 (45.0)	0.952 (NS)
Frequency of neuroretinal rim notching (%)	38 (29.5)	39 (18.7)	0.022
Frequency of nerve fiber layer defects (%)	4 (3.1)	3 (1.4)	0.307 (NS)

 Data are the mean ± SD; 95% confidence interval.

Figure 1.

 

Representative ONH photographs in patients with the Gly252Arg (ACT02-26, aged 52; ACT02-2, aged 64; and ACT02-7, aged 71); Gln368Stop (GTas2-74, aged 50; GVic117-1, aged 62; and GVic309-6, aged 75); Thr377Met (GVic1-1, aged 34, GVic1-8, aged 48; and GVic1-53, aged 78); or Pro370Leu (GNSW23-2, aged 18; GNSW23-1, aged 41; and GNSW23-7, aged 62) MYOC mutations.

The authors thank Lisa Kearns and David Dimasi for providing technical assistance, Adrian Esterman and Catherine McCarty for statistical advice, and the research participants and their ophthalmologists.