• endophenotype;
  • genetics;
  • primary open-angle glaucoma;
  • quantitative trait loci


  1. Top of page
  2. A
  3. Introduction
  4. Conclusion
  5. References

Genome-wide association studies are a powerful tool for the identification of genetic risk factors for complex disease. This methodology has been successfully applied to primary open-angle glaucoma through the analysis of primary open-angle glaucoma (POAG) as well as specific subgroups of patients including those with normal tension glaucoma and advanced glaucoma. In addition, the analysis of quantitative traits important in POAG, including optic disc area and vertical cup-to-disc ratio has also identified genes important in POAG development. This review explores findings of genome-wide association studies for POAG and related traits.


  1. Top of page
  2. A
  3. Introduction
  4. Conclusion
  5. References

Primary open-angle glaucoma (POAG) has long been known to have a strong genetic component. Around 60% of patients have a family history of the disease1 and family-based studies have demonstrated a 10-fold increased risk of POAG for first-degree relatives of affected individuals.2 The heritability of POAG itself is calculated to be around 0.81 and the heritability of related quantitative traits is also high.3 Several genes have been identified that appear to increase the risk of POAG. Most notably, the myocilin gene, the first gene identified as a risk factor for POAG in 1997,4 accounts for 3–4% of POAG. Genes identified through family-based and candidate gene studies have recently been reviewed by Fingert.5

Genome-wide association studies (GWAS) have recently been used for the investigation of many complex diseases including POAG. GWAS are generally applied to large case-control cohorts and are designed to assess the entire genome for association without bias to known protein coding genes. Single-nucleotide polymorphisms (SNPs) throughout the genome are genotyped in cases and controls and the allele frequencies between the two groups compared. The genotyping strategy is based on the premise of linkage disequilibrium. This is the non-random assortment of alleles on the chromosome because of their proximity to each other and a tendency not to recombine. This phenomenon means that much of the common variation in the genome can be assayed indirectly, by typing ‘tag’ SNPs that can provide proxy information about genotypes at nearby SNPs. Modern SNP genotyping arrays are designed with the linkage disequilibrium structure of the human genome taken into account. Although useful for limiting the depth of genotyping required, linkage disequilibrium can also hinder attempts to determine which SNP is actually the causative variant as all SNPs in linkage disequilibrium with each other will give similar association signals. Thus GWAS can identify genomic regions associated with a disease or trait of interest, but rarely will identify the functional variant and further in depth genotyping and functional testing is usually required. A further consideration in GWAS study design is the threshold required for statistical significance. Because of the large number of SNPs tested, a strict multiple testing correction should be applied. A P-value of less than 5 × 10−8 is generally considered acceptable, accounting for around 1 million independent tests.6 Replication of the findings in independent cohorts is also considered a standard requirement for claiming a successful gene identification from a GWAS study. Even with these limitations, however, the methodology has been successful at identifying loci associated with POAG and its related quantitative traits. This review will describe GWAS findings in POAG.

The first GWAS in glaucoma

The first GWAS to investigate glaucoma was a study of 195 patients with open-angle glaucoma from Iceland, compared with 14 474 controls.7 This study identified two coding variants at the LOXL1 locus, rs1048661 and rs3825942. The patient cohort consisted of 90 POAG patients and 75 patients with pseudoexfoliation glaucoma. On subanalysis, only the pseudoexfoliation patients were contributing to the association signal. The association has since been confirmed in multiple cohorts with pseudoexfoliation, with and without glaucoma (reviewed in8) but this locus is not associated with POAG in any cohort examined.9,10 This study highlighted the importance of careful phenotyping in patient cohorts to ensure homogeneity of the sample. The effect size of the LOXL1 locus in pseudoexfoliation was large enough that the signal was still observed in the mixed phenotype population. However, this is unlikely to be the case for most genetic associations.

Normal tension glaucoma

The first GWAS for POAG was published in 2009 by Nakano et al.11 This was a two stage GWAS where a proportion of the available samples were genotyped on genome-wide SNP arrays and the top ranked SNPs were then typed in the remaining samples. This two stage design is a common method used to maintain statistical power while minimizing the costs of genome-wide genotyping.12 The study evaluated 827 cases and 748 controls from Kyoto, Japan. The patients predominantly had normal tension glaucoma, which is the most common presentation of POAG in Japan.13 This study identified three genomic regions associated in both stages of the GWAS, with a P-values of the order of 10−5 in the whole cohort. Although support for association was observed in both stage 1 and stage 2, none of the SNPs reached genome-wide significance (P < 5 × 10−8) even in the combined analysis and therefore these loci have not yet demonstrated statistical robustness and await further evaluation in additional cohorts. The loci reported are on chromosomes 1, 10 and 12 with the nearest annotated genes ZP4, PLXDC2 and DKFZp762A217, respectively.

In 2010 the Normal Tension Glaucoma Genetic Study Group of the Japan Glaucoma Society published a GWAS limited to patients with normal tension glaucoma (NTG).14 This small GWAS of 305 patients compared with 355 examined controls identified one locus at genome-wide significance. SNP rs3213787 in an intron of the SRBD1 gene gave an odds ratio of 2.8 and P-value of 2.5 × 10−9. Other SNPs in the vicinity were also associated with P-values in the order of 10−6. The A allele at the most associated SNP, rs3213787, is the most common allele, being present on over 80% of control chromosomes, but is even further enriched in NTG patients. This scenario of the most common allele being associated with disease risk has been observed before, notably at the LOXL1 locus for pseudoexfoliation syndrome.15 As with the report by Nakano et al., the findings from this NTG GWAS are awaiting replication in an independent cohort. The authors also report that the SRBD1 gene is expressed in the retina of neonatal mice14 and that the expression level of the SRBD1 mRNA in peripheral whole blood of human participants was elevated in those carrying the NTG risk allele at this SNP, compared with patients without this allele.14 This suggests that the SNP or one in linkage disequilibrium with it is involved in regulating expression of this gene. It was not clear if these patients had NTG or if the levels were higher in NTG patients compared with controls and thus a direct link between transcript levels and risk of NTG is awaiting further evidence.

Caveolin locus

The Caveolin locus was the first genomic region to be identified at genome-wide significance and replicated in independent cohorts. The initial discovery was described by Thorleifsson and colleagues who reported on a GWAS utilizing 1263 POAG patients from Iceland compared with 34 877 controls with no reported glaucoma.16 This study identified the A allele of SNP rs4236601 as associated with POAG in Iceland with an odds ratio of 1.36 and P-value of 5.0 × 10−10. The study included several independent replication cohorts; A European cohort from Sweden, the UK and Australia and an Asian cohort of Hong Kong and Shantou Chinese. In the European replication cohort (2064 controls and 2175 cases), the overall odds ratio at this SNP was 1.18 with P = 0.0015. The Asian cohort (1607 controls and 299 cases) also showed strong replication with odds ratio of 3.33 and P-value of 0.003. In addition, the risk of POAG attributable to this locus appears to be stronger in the subset of cases with normal tension glaucoma (IOP ≤ 21 mmHg) although this only reached statistical significance when all European cohorts with available data (including the discovery cohort) were combined.

The Asian replication cohort demonstrated much higher odds ratios than the European discovery and replication cohorts. This SNP, rs4236601, is relatively rare in this Chinese cohort with the A allele frequency of less than 1% in the controls, compared with 20–26% in European controls. The authors postulate that this variant may tag the functional rare variant with stronger linkage disequilibrium in the Asian population than in Europeans. This is supported by a larger block of linkage disequilibrium in Asians than Europeans, although it may also be that the locus contributes different risk in the two populations.

The associated SNPs at this 7q34 locus lie in a block of linkage disequilibrium encompassing the CAV1 and CAV2 genes. These genes encode caveolin 1 and 2, proteins involved in the generation and function of caveolae, which are small invaginations of the cell membrane rich in proteins and specific lipids. Caveloae are involved in cell signalling as well as endocytosis. The mechanism through which these genes might contribute to POAG are not yet known; however, they may influence TGF-β or nitric oxide signalling.16 It is also not yet known which is the functional variant at this locus that actually contributes to POAG risk.

Several other studies have followed up these findings in European cohorts. Kuehn et al. reported a failure to replicate association at this locus.17 They genotyped SNP rs4236601 in 545 POAG patients and 297 controls of European descent from Iowa, USA. IOP was an inclusion criteria for both cases (IOP > 21 mmHg) and controls (IOP ≤ 21 mmHg). The study reported no association at this SNP (P = 0.22). From the data provided, the observed odds ratio was around 1.08. For comparison the reported odds ratio in the similar Caucasian populations in the original replication cohorts published by Thorleifsson et al. was 1.18. The Kuehn et al. study reported sufficient power to detect a significant association at this SNP with an odds ratio of 1.4 or better, which is similar to the odds ratio reported in the Icelandic discovery cohort. It is well recognized that the initial report of an association tends to overestimate the effect size18 and therefore, it would have been more appropriate to power this follow-up study to detect the effect size observed in the replication cohorts. In addition, the cohort in Kuehn et al. was limited to patients with IOP >21 mmHg, whereas in the original report, the association was strongest in patients without elevated IOP.16 The data presented by Kuehn et al. does actually support a role for this locus in POAG, at similar effect size to other reported Caucasian/European populations, although does not reach statistical significance.

A second independent study has also investigated this locus in an American cohort. Wiggs et al. report an odds ratio of 1.28 in a cohort of 1000 POAG cases and 1183 controls with P = 0.0014, consistent with the original report.19 This study also identified two risk haplotypes in the region associated with POAG with similar odds ratios to the individual SNP, although not containing either of the SNPs reported by Thorleifsson et al. There was no evidence however that these haplotypes were associated independently of the individual SNPs and likely tag the same causative variant as the individual SNPs. In addition, this study identified a sex specific association at this locus, with rs4236601 and rs1052990 showing association in women but not men, as well as replicating the observation of stronger association in normal tension POAG than high tension POAG.

Thus, the CAV1/CAV2 locus on 7q34 has been repeatedly associated with POAG in Caucasian (European) populations, although with small effect sizes with odds ratios in the range 1.1–1.3 in most cohorts. The association appears to be stronger in females and in patients with normal tension glaucoma. The locus is also associated in Asia with larger effect size than observed in Caucasians, although this finding is yet to be replicated in independent cohorts.

Advanced glaucoma

A second GWAS has also been conducted in a Caucasian population. This study utilized 615 patients with advanced glaucoma, requiring severe visual field defects for study entry.20 The patients were compared with 3946 historic controls who had not undergone ocular examination. A proportion of this GWAS discovery cohort was included in the Australian replication cohort in the study of Thorleiffson et al. that identified the CAV1/CAV2 locus. Interestingly, this locus did not rank highly in the Australian GWAS of advanced glaucoma, although a similar effect size was observed as reported previously in this population, as would be expected. The main findings from the Australian GWAS were two loci (1q24 and 9p21), associated with advanced glaucoma, but also replicated in three independent Australian cohorts, one with advanced POAG and the other two with less severe disease.

The top ranked SNP at 1q24 was rs4656461, giving an odds ratio of 1.68 with P-value of 6.1 × 10−10 in the discovery cohort. This SNP is located downstream of the TMCO1 gene. A second SNP reaching genome-wide significance and also showing replication in all three cohorts was situated in an intron of the same gene. The function of the TMCO1 gene is not yet known, although recessive mutations cause a rare syndrome consisting of skeletal and craniofacial dysmorphism and mental retardation.21 The top ranked SNP at 9p21 was rs4977756 with odds ratio of 1.50 and P-value of 4.7 × 10−9. Again, the second ranked SNP at this locus also reached genome-wide significance and both SNPs were replicated in all three additional cohorts. The associated SNPs at 9p21 are situated within a long non-coding RNA called CDKN2B-AS1 (also known as ANRIL). This functional RNA is involved in the regulation of other genes at this locus, notably the tumour suppressor genes CDKN2A and CDKN2B.22 These genes are involved in the cell cycle, promoting senescence and an apoptotic response to stress.22 All four genes at these loci (TMCO1, CDKN2B-AS1, CDKN2A and CDKN2B) are expressed in retinal ganglion cells.20 In addition, CDKN2A and CDKN2B are upregulated in response to elevated intraocular pressure (IOP) in an animal model of glaucoma20 suggesting that an altered expression of these genes in response to IOP (or other stressors) may promote apoptosis of retinal ganglion cells, leading to glaucomatous visual field loss.


In the evaluation of complex diseases, it is often useful to evaluate endophenotypes, or quantitative traits that contribute to the overall disease phenotype.3 Endophenotypes may have overlapping genetic risk factors with the disease of interest. In the case of POAG, proposed endophenotypes include vertical cup-to-disc ratio (VCDR),3 IOP,3 central corneal thickness (CCT)3 and optic disc area (ODA)23 and GWAS in normal populations have been published for VCDR,24 CCT25–28 and ODA23,24,29 (Table 1). Several populations have been used to evaluate these traits, often through meta-analysis. The Australian and UK Twins cohorts, the Rotterdam Study and Croatian study are predominantly Caucasian samples whereas the SINDI, SiMES, SCES and BES studies are in Asian populations.

Table 1.  Loci reaching genome-wide significance in GWAS studies of glaucoma related endophenotypes
StudyTraitPopulationLocusGenePOAG association?
  1. References are given as superscript numbers. The locus and nearest gene for each result is shown along with the result for association with POAG. ‘–’ indicates not evaluated in relation to POAG to date. AusTwins, Brisbane Adolescent Twin Study and Twins Eye Study in Tasmania; BES, Beijing Eye Study; CCT, central corneal thickness; CROATIA, From the Croatian islands of Vis and Korcula; GWAS, genome-wide association studies; POAG, primary open-angle glaucoma; ODA, optic disc area; Rotterdam, The Rotterdam Study; SCES, Singapore Chinese Eye Study; SINDI, Singapore Indian Eye Study; SiMES, Singapore Malay Eye Study; TwinsUK, St Thomas' Hospital Adult Twin Registry, London; VCDR, vertical cup : disc ratio.

Lu25CCTAusTwins + TwinsUK16q24.2ZNF469
 CROATIA + AusTwins + TwinsUK15q25.3AKAP13
Vithana27CCTSINDI + SiMES16q24.2ZNF469
Cornes28CCTSINDI + SiMES + SCES + BES16q24.2ZNF469
Macgregor23ODAAusTwins + TwinsUK10q21.3-q22.1ATOH7Yes31
Ramdas24ODARotterdam Study1p22CDC7/TGFBR3Borderline31
 Rotterdam + TwinsUK16q12.1SALL1Borderline31
Ramdas24VCDRRotterdam Study9p21CDKN2A/BYes20,31,32
 Rotterdam + TwinsUK10q21.3-q22.1ATOH7Yes31

Three independent GWAS for CCT identified the ZNF469 locus.25–27 The associated SNPs are in what appears to be a regulatory region upstream of the gene. Mutations in ZNF469 cause brittle cornea syndrome, characterized in part by a thin cornea.30 The COL5A1 gene has also been implicated in GWAS of both Caucasian26 and Asian27 cohorts whereas several other loci have been reported in GWAS and meta-analysis of GWAS and replicated to some level in independent cohorts. FOXO1 was reported at genome-wide significance by Lu et al.25 The data of Vitart et al.26 demonstrated replication, but the Asian study of Vithana et al.27 did not. Vitart et al.26 also reported AKAP13 and AVGR8. AKAP13 was replicated in Asians by Vithana et al.27 but AVGR8 was not. In addition, Vithana et al.27 identified association at COL8A2 in Singapore Indians and Malays, and later replicated this association in Chinese from Singapore and Beijing.28 This later study also identified four additional loci through meta-analysis of all four Asian cohorts including SNPs in or near IBTK, LRRK1/CHYS1, a gene desert on 9p23 and C7orf42. These loci are also yet to be replicated in independent cohorts. Thus strong evidence exists for the association of SNPs upstream of ZNF469 and within COL5A1 and COL8A2 with CCT and some evidence exists for an association of FOXO1 and AKAP13. All these genes are awaiting analysis in relation to POAG.

ODA has been evaluated in three independent GWAS.23,24,29 All three identified the ATOH7 gene on chromosome 10. Macgregor et al.23 also demonstrated that mutations in the gene were associated with optic nerve hypoplasia, a common form of childhood blindness. Ramdas et al.31 were also able to demonstrate an association of this gene with POAG itself. The CDC7/TGFBR3 region was also associated in two studies24,29 but was of borderline significance for association with POAG.31

To date, only one GWAS reaching genome-wide significance has been published for VCDR.24 This study identified the CDKN2B region on chromosome 9p21, the same region as implicated directly in POAG by GWAS20 discussed earlier. This locus has also been associated with POAG in two further independent replication studies.31,32 Similarly, the SIX1/SIX6 region identified in the same study has also been associated with POAG.31 Interestingly, the ATOH7 locus associated with ODA is also associated with VCDR in this GWAS.24 Other loci associated with VCDR on meta-analysis of the Rotterdam study with the TwinsUK study (Table 1) including SNPs between FERM8 and SCYL1 and around CHEK2 and DCKL1 are awaiting replication for VCDR association, but so far show no association with POAG.31

Thus, the analysis of endophenotypes has identified a range of candidate genes and loci for POAG. Several of these genes, particularly those associated with VCDR have already proven to be important risk factors for POAG. In all cases, the specific functional variant and its mechanism of action in relation to these quantitative traits is yet to be identified.


  1. Top of page
  2. A
  3. Introduction
  4. Conclusion
  5. References

Significant inroads into understanding the genetic basis of POAG have occurred in recent years, due in large part to the application of GWAS methodology. The CDKN2B-AS1 locus on chromosome 9 is without doubt a major genetic risk factor for glaucoma. This locus, as well as the SIX1/SIX6 locus were identified through analysis of VCDR in normal populations and have since shown reproducible association with POAG itself. The CAV1/CAV2 locus is also reproducibly associated, although the effect size is small in most Caucasian populations investigated to date. In conclusion, GWAS has been a valuable tool in the hunt for genetic risk factors for POAG, although the identified loci, including high penetrance genes identified in family studies, still account for less than 10% of POAG. Thus further loci remain to be identified and this will likely be achieved through a variety of methodologies including GWAS and the more traditional family-based studies for POAG and its endophenotypes.


  1. Top of page
  2. A
  3. Introduction
  4. Conclusion
  5. References
  • 1
    Green CM, Kearns LS, Wu J et al. How significant is a family history of glaucoma? Experience from the Glaucoma Inheritance Study in Tasmania. Clin Experiment Ophthalmol 2007; 35: 7939.
  • 2
    Wolfs RC, Klaver CC, Ramrattan RS et al. Genetic risk of primary open-angle glaucoma. Population-based familial aggregation study. Arch Ophthalmol 1998; 116: 16405.
  • 3
    Charlesworth J, Kramer PL, Dyer T et al. The path to open-angle glaucoma gene discovery: endophenotypic status of intraocular pressure, cup-to-disc ratio, and central corneal thickness. Invest Ophthalmol Vis Sci 2010; 51: 350914.
  • 4
    Stone EM, Fingert JH, Alward WL et al. Identification of a gene causing primary open angle glaucoma. Science 1997; 275: 66870.
  • 5
    Fingert JH. Primary open-angle glaucoma genes. Eye 2011; 25: 58795.
  • 6
    Dudbridge F, Gusnanto A. Estimation of significance thresholds for genomewide association scans. Genet Epidemiol 2008; 32: 22734.
  • 7
    Thorleifsson G, Magnusson KP, Sulem P et al. Common sequence variants in the LOXL1 gene confer susceptibility to exfoliation glaucoma. Science 2007; 317: 1397400.
  • 8
    Schlotzer-Schrehardt U. Genetics and genomics of pseudoexfoliation syndrome/glaucoma. Middle East Afr J Ophthalmol 2011; 18: 306.
  • 9
    Abu-Amero KK, Osman EA, Azad MT et al. Lack of association between LOXL1 gene polymorphisms and primary open angle glaucoma in the Saudi Arabian population. Ophthalmic Genet 2011; doi: 10.3109/13816810.2011.575430 [Epub ahead of print].
  • 10
    Chen H, Chen LJ, Zhang M et al. Ethnicity-based subgroup meta-analysis of the association of LOXL1 polymorphisms with glaucoma. Mol Vis 2010; 16: 16777.
  • 11
    Nakano M, Ikeda Y, Taniguchi T et al. Three susceptible loci associated with primary open-angle glaucoma identified by genome-wide association study in a Japanese population. Proc Natl Acad Sci U S A 2009; 106: 1283842.
  • 12
    Satagopan JM, Venkatraman ES, Begg CB. Two-stage designs for gene-disease association studies with sample size constraints. Biometrics 2004; 60: 58997.
  • 13
    Iwase A, Suzuki Y, Araie M et al. The prevalence of primary open-angle glaucoma in Japanese: the Tajimi Study. Ophthalmology 2004; 111: 16418.
  • 14
    Meguro A, Inoko H, Ota M, Mizuki N, Bahram S. Genome-wide association study of normal tension glaucoma: common variants in SRBD1 and ELOVL5 contribute to disease susceptibility. Ophthalmology 2010; 117: 13318 e5.
  • 15
    Hewitt AW, Sharma S, Burdon KP et al. Ancestral LOXL1 variants are associated with pseudoexfoliation in Caucasian Australians but with markedly lower penetrance than in Nordic people. Hum Mol Genet 2008; 17: 71016.
  • 16
    Thorleifsson G, Walters GB, Hewitt AW et al. Common variants near CAV1 and CAV2 are associated with primary open-angle glaucoma. Nat Genet 2010; 42: 9069.
  • 17
    Kuehn MH, Wang K, Roos B et al. Chromosome 7q31 POAG locus: ocular expression of caveolins and lack of association with POAG in a US cohort. Mol Vis 2011; 17: 4305.
  • 18
    Beavis W. The power and deceit of {QTL} experiments: lessons from comparative {QTL} studies. In: Wilkinson DB, ed. Proceedings of the Corn and Sorghum Industry Research Conference1994. Washington D.C.: American Seed Trade Association, 1994; 25066.
  • 19
    Wiggs JL, Hee Kang J, Yaspan BL et al. Common variants near CAV1 and CAV2 are associated with primary open-angle glaucoma in Caucasians from the USA. Hum Mol Genet 2011; 20: 470713.
  • 20
    Burdon KP, Macgregor S, Hewitt AW et al. Genome-wide association study identifies susceptibility loci for open angle glaucoma at TMCO1 and CDKN2B-AS1. Nat Genet 2011; 43: 5748.
  • 21
    Xin B, Puffenberger EG, Turben S et al. Homozygous frameshift mutation in TMCO1 causes a syndrome with craniofacial dysmorphism, skeletal anomalies, and mental retardation. Proc Natl Acad Sci U S A 2010; 107: 25863.
  • 22
    Aguilo F, Zhou MM, Walsh MJ. Long noncoding RNA, polycomb, and the ghosts haunting INK4b-ARF-INK4a expression. Cancer Res 2011; 71: 53659.
  • 23
    Macgregor S, Hewitt AW, Hysi PG et al. Genome-wide association identifies ATOH7 as a major gene determining human optic disc size. Hum Mol Genet 2010; 19: 271624.
  • 24
    Ramdas WD, van Koolwijk LM, Ikram MK et al. A genome-wide association study of optic disc parameters. PLoS Genet 2010; 6: e1000978.
  • 25
    Lu Y, Dimasi D, Hysi P et al. Common genetic variants near the Brittle Cornea Syndrome Locus ZNF469 influence the blinding disease risk factor central corneal thickness. PLoS Genet 2010; 13: e1000947.
  • 26
    Vitart V, Bencic G, Hayward C et al. New loci associated with central cornea thickness include COL5A1, AKAP13 and AVGR8. Hum Mol Genet 2010; 19: 430411.
  • 27
    Vithana EN, Aung T, Khor CC et al. Collagen-related genes influence the glaucoma risk factor, central corneal thickness. Hum Mol Genet 2011; 20: 64958.
  • 28
    Cornes BK, Khor CC, Nongpiur ME et al. Identification of four novel variants that influence central corneal thickness in multi-ethnic Asian populations. Hum Mol Genet 2011; 21: 43745.
  • 29
    Khor CC, Ramdas WD, Vithana EN et al. Genome-wide association studies in Asians confirm the involvement of ATOH7 and TGFBR3, and further identify CARD10 as a novel locus influencing optic disc area. Hum Mol Genet 2011; 20: 186472.
  • 30
    Christensen AE, Knappskog PM, Midtbo M et al. Brittle cornea syndrome associated with a missense mutation in the zinc-finger 469 gene. Invest Ophthalmol Vis Sci 2010; 51: 4752.
  • 31
    Ramdas WD, van Koolwijk LM, Lemij HG et al. Common genetic variants associated with open-angle glaucoma. Hum Mol Genet 2011; 20: 246471.
  • 32
    Fan BJ, Wang DY, Pasquale LR, Haines JL, Wiggs JL. Genetic variants associated with optic nerve vertical cup-to-disc ratio are risk factors for primary open angle glaucoma in a US Caucasian population. Invest Ophthalmol Vis Sci 2011; 52: 178892.