February 2018
Volume 59, Issue 2
Open Access
Glaucoma  |   February 2018
Testosterone Pathway Genetic Polymorphisms in Relation to Primary Open-Angle Glaucoma: An Analysis in Two Large Datasets
Author Affiliations & Notes
  • Jessica N. Cooke Bailey
    Department of Population and Quantitative Health Sciences, Case Western Reserve University School of Medicine, Cleveland, Ohio, United States
    Institute for Computational Biology, Case Western Reserve University School of Medicine, Cleveland, Ohio, United States
  • Puya Gharahkhani
    Statistical Genetics, QIMR Berghofer Medical Research Institute, Royal Brisbane Hospital, Brisbane, Australia
  • Jae H. Kang
    Channing Division of Network Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts, United States
  • Mariusz Butkiewicz
    Department of Population and Quantitative Health Sciences, Case Western Reserve University School of Medicine, Cleveland, Ohio, United States
    Institute for Computational Biology, Case Western Reserve University School of Medicine, Cleveland, Ohio, United States
  • David A. Sullivan
    Schepens Eye Research Institute, Massachusetts Eye and Ear Infirmary, Harvard Medical School, Boston, Massachusetts, United States
  • Robert N. Weinreb
    Department of Ophthalmology, Hamilton Glaucoma Center and Shiley Eye Institute, University of California at San Diego, La Jolla, California, United States
  • Hugues Aschard
    Department of Epidemiology, Harvard T. H. Chan School of Public Health, Harvard Medical School, Boston, Massachusetts, United States
  • R. Rand Allingham
    Department of Ophthalmology, Duke University Medical Center, Durham, North Carolina, United States
  • Allison Ashley-Koch
    Department of Medicine, Duke University Medical Center, Durham, North Carolina, United States
  • Richard K. Lee
    Bascom Palmer Eye Institute, University of Miami Miller School of Medicine, Miami, Florida, United States
  • Sayoko E. Moroi
    Department of Ophthalmology and Visual Sciences, University of Michigan, Ann Arbor, Michigan, United States
  • Murray H. Brilliant
    Center for Human Genetics, Marshfield Clinic Research Institute, Marshfield, Wisconsin, United States
  • Gadi Wollstein
    Department of Ophthalmology, NYU Langone Medical Center, NYU School of Medicine, New York, New York, United States
  • Joel S. Schuman
    Department of Ophthalmology, NYU Langone Medical Center, NYU School of Medicine, New York, New York, United States
  • John H. Fingert
    Departments of Ophthalmology and Anatomy/Cell Biology, University of Iowa, College of Medicine, Iowa City, Iowa, United States
  • Donald L. Budenz
    Department of Ophthalmology, University of North Carolina, Chapel Hill, North Carolina, United States
  • Tony Realini
    Department of Ophthalmology, WVU Eye Institute, Morgantown, West Virginia, United States
  • Terry Gaasterland
    Scripps Genome Center, University of California at San Diego, San Diego, California, United States
  • William K. Scott
    Institute for Human Genomics, University of Miami Miller School of Medicine, Miami, Florida, United States
  • Kuldev Singh
    Department of Ophthalmology, Stanford University, Palo Alto, California, United States
  • Arthur J. Sit
    Department of Ophthalmology, Mayo Clinic, Rochester, Minnesota, United States
  • Robert P. Igo, Jr
    Department of Population and Quantitative Health Sciences, Case Western Reserve University School of Medicine, Cleveland, Ohio, United States
  • Yeunjoo E. Song
    Department of Population and Quantitative Health Sciences, Case Western Reserve University School of Medicine, Cleveland, Ohio, United States
    Institute for Computational Biology, Case Western Reserve University School of Medicine, Cleveland, Ohio, United States
  • Lisa Hark
    Wills Eye Hospital, Glaucoma Research Center, Philadelphia, Pennsylvania, United States
  • Robert Ritch
    Einhorn Clinical Research Center, New York Eye and Ear Infirmary of Mount Sinai, New York, New York, United States
  • Douglas J. Rhee
    Department of Ophthalmology, Case Western Reserve University School of Medicine, Cleveland, Ohio, United States
  • Douglas Vollrath
    Department of Genetics, Stanford University, Palo Alto, California, United States
  • Donald J. Zack
    Wilmer Eye Institute, Johns Hopkins University Hospital, Baltimore, Maryland, United States
  • Felipe Medeiros
    Department of Ophthalmology, Hamilton Glaucoma Center and Shiley Eye Institute, University of California at San Diego, La Jolla, California, United States
  • Thasarat S. Vajaranant
    Department of Ophthalmology, University of Illinois College of Medicine at Chicago, Chicago, Illinois, United States
  • Daniel I. Chasman
    Division of Preventive Medicine, Brigham and Women's Hospital, Boston, Massachusetts, United States
  • William G. Christen
    Division of Preventive Medicine, Brigham and Women's Hospital, Boston, Massachusetts, United States
  • Margaret A. Pericak-Vance
    Institute for Human Genomics, University of Miami Miller School of Medicine, Miami, Florida, United States
  • Yutao Liu
    Department of Cellular Biology and Anatomy, Augusta University, Augusta, Georgia, United States
  • Peter Kraft
    Department of Epidemiology, Harvard T. H. Chan School of Public Health, Harvard Medical School, Boston, Massachusetts, United States
    Department of Biostatistics, Harvard T. H. Chan School of Public Health, Harvard Medical School, Boston, Massachusetts, United States
  • Julia E. Richards
    Department of Ophthalmology and Visual Sciences, University of Michigan, Ann Arbor, Michigan, United States
  • Bernard A. Rosner
    Channing Division of Network Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts, United States
    Department of Biostatistics, Harvard T. H. Chan School of Public Health, Harvard Medical School, Boston, Massachusetts, United States
  • Michael A. Hauser
    Department of Ophthalmology, Duke University Medical Center, Durham, North Carolina, United States
    Department of Medicine, Duke University Medical Center, Durham, North Carolina, United States
  • Jamie E. Craig
    Department of Ophthalmology, Flinders University, Adelaide, SA, Australia
  • Kathryn P. Burdon
    School of Medicine, Menzies Research Institute of Tasmania, Hobart, Australia
  • Alex W. Hewitt
    Centre for Ophthalmology and Visual Science, Lions Eye Institute, University of Western Australia, Perth, Australia
  • David A. Mackey
    School of Medicine, Menzies Research Institute of Tasmania, Hobart, Australia
    Centre for Ophthalmology and Visual Science, Lions Eye Institute, University of Western Australia, Perth, Australia
  • Jonathan L. Haines
    Department of Population and Quantitative Health Sciences, Case Western Reserve University School of Medicine, Cleveland, Ohio, United States
    Institute for Computational Biology, Case Western Reserve University School of Medicine, Cleveland, Ohio, United States
  • Stuart MacGregor
    Statistical Genetics, QIMR Berghofer Medical Research Institute, Royal Brisbane Hospital, Brisbane, Australia
  • Janey L. Wiggs
    Department of Ophthalmology, Mass Eye and Ear Infirmary, Harvard Medical School, Boston, Massachusetts, United States
  • Louis R. Pasquale
    Channing Division of Network Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts, United States
    Department of Ophthalmology, Mass Eye and Ear Infirmary, Harvard Medical School, Boston, Massachusetts, United States
  • Correspondence: Louis R. Pasquale, Mass Eye and Ear Infirmary, 243 Charles Street, Boston, MA 02114, USA; Louis_Pasquale@meei.harvard.edu
  • Footnotes
     JEC, KPB, AWH, DAM, and SM are members of the ANZRAG Consortium.
Investigative Ophthalmology & Visual Science February 2018, Vol.59, 629-636. doi:10.1167/iovs.17-22708
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Jessica N. Cooke Bailey, Puya Gharahkhani, Jae H. Kang, Mariusz Butkiewicz, David A. Sullivan, Robert N. Weinreb, Hugues Aschard, R. Rand Allingham, Allison Ashley-Koch, Richard K. Lee, Sayoko E. Moroi, Murray H. Brilliant, Gadi Wollstein, Joel S. Schuman, John H. Fingert, Donald L. Budenz, Tony Realini, Terry Gaasterland, William K. Scott, Kuldev Singh, Arthur J. Sit, Robert P. Igo, Yeunjoo E. Song, Lisa Hark, Robert Ritch, Douglas J. Rhee, Douglas Vollrath, Donald J. Zack, Felipe Medeiros, Thasarat S. Vajaranant, Daniel I. Chasman, William G. Christen, Margaret A. Pericak-Vance, Yutao Liu, Peter Kraft, Julia E. Richards, Bernard A. Rosner, Michael A. Hauser, Jamie E. Craig, Kathryn P. Burdon, Alex W. Hewitt, David A. Mackey, Jonathan L. Haines, Stuart MacGregor, Janey L. Wiggs, Louis R. Pasquale, for the Australian and New Zealand Registry of Advanced Glaucoma (ANZRAG) Consortium; Testosterone Pathway Genetic Polymorphisms in Relation to Primary Open-Angle Glaucoma: An Analysis in Two Large Datasets. Invest. Ophthalmol. Vis. Sci. 2018;59(2):629-636. doi: 10.1167/iovs.17-22708.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose: Sex hormones may be associated with primary open-angle glaucoma (POAG), although the mechanisms are unclear. We previously observed that gene variants involved with estrogen metabolism were collectively associated with POAG in women but not men; here we assessed gene variants related to testosterone metabolism collectively and POAG risk.

Methods: We used two datasets: one from the United States (3853 cases and 33,480 controls) and another from Australia (1155 cases and 1992 controls). Both datasets contained densely called genotypes imputed to the 1000 Genomes reference panel. We used pathway- and gene-based approaches with Pathway Analysis by Randomization Incorporating Structure (PARIS) software to assess the overall association between a panel of single nucleotide polymorphisms (SNPs) in testosterone metabolism genes and POAG. In sex-stratified analyses, we evaluated POAG overall and POAG subtypes defined by maximum IOP (high-tension [HTG] or normal tension glaucoma [NTG]).

Results: In the US dataset, the SNP panel was not associated with POAG (permuted P = 0.77), although there was an association in the Australian sample (permuted P = 0.018). In both datasets, the SNP panel was associated with POAG in men (permuted P ≤ 0.033) and not women (permuted P ≥ 0.42), but in gene-based analyses, there was no consistency on the main genes responsible for these findings. In both datasets, the testosterone pathway association with HTG was significant (permuted P ≤ 0.011), but again, gene-based analyses showed no consistent driver gene associations.

Conclusions: Collectively, testosterone metabolism pathway SNPs were consistently associated with the high-tension subtype of POAG in two datasets.

Primary open-angle glaucoma (POAG), a leading cause of chronic progressive optic nerve degeneration worldwide,1 is a strongly age-related disease.2,3 Testosterone and estradiol production decline with age,4,5 and accumulating evidence suggests that the retina and optic nerve are sex hormone–sensitive tissues69; thus, these sex hormones may be implicated in the glaucomatous process. Postmenopausal therapy with estrogen alone was associated with modest reductions in IOP in a post hoc analysis from a randomized clinical trial10 and with lower POAG risk in a large observational study11; yet, there is a gap in our understanding of the role testosterone plays, if any, in the glaucomatous process. 
Nongonadal sources of sex hormone precursors from the adrenal gland provide dehydroepiandrosterone (DHEA), which is converted to sex hormones in peripheral tissues (Fig.).12 Intracrinology refers to the synthesis of sex steroids in these peripheral tissues from adrenal precursors.13 In fact, intracrine metabolism accounts for essentially all androgen (and estrogen) synthesis in peripheral tissues for postmenopausal women and up to approximately 40% of androgen synthesis in aging men.13 Thus, biochemical factors that lead to differences in intracrine metabolism may impact the availability of sex hormones and influence disease processes related to declining hormones. For example, various isoforms of 17-beta hydroxysteroid dehydrogenase (17β HSD) are essential to the local intracellular generation of testosterone and estradiol from DHEA (Fig.).12 Interestingly, Coca-Prados and colleagues14 documented that the neuroendocrine secretory ciliary epithelium expresses 17β HSD isoforms 2, 5, and 7, and these cells actively metabolize androgen and other sex hormones. However, genetic variants of 17β HSD have been little studied specifically in relation to POAG. 
Figure
 
The testosterone pathway is depicted. Gene names are as follows: AKR1C3-ALDO-KETO reductase family 1, member C3, also known as HSD17B5; HSD17B 1, 2, 3, 7, 8, 12, 14–17 beta hydroxysteroid dehydrogenase isoforms; CYP19A1: cytochrome P450, family 19, subfamily A, polypeptide 1; SULT1E1: sulfotransferase family 1E, estrogen-preferring, member 1; SULT1A1: sulfotransferase family 1A, cytosolic, phenol-preferring, member 1; SRD5A1, 2, 3: steroid 5-alpha-reductase 1, 2, 3; HSD3B 1, 2: 3-beta-hydroxysteroid dehydrogenase 1, 2. AR, androgen receptor; ER, estrogen receptor; DHT, dehydroxytestosterone.
Figure
 
The testosterone pathway is depicted. Gene names are as follows: AKR1C3-ALDO-KETO reductase family 1, member C3, also known as HSD17B5; HSD17B 1, 2, 3, 7, 8, 12, 14–17 beta hydroxysteroid dehydrogenase isoforms; CYP19A1: cytochrome P450, family 19, subfamily A, polypeptide 1; SULT1E1: sulfotransferase family 1E, estrogen-preferring, member 1; SULT1A1: sulfotransferase family 1A, cytosolic, phenol-preferring, member 1; SRD5A1, 2, 3: steroid 5-alpha-reductase 1, 2, 3; HSD3B 1, 2: 3-beta-hydroxysteroid dehydrogenase 1, 2. AR, androgen receptor; ER, estrogen receptor; DHT, dehydroxytestosterone.
Using a pathway analysis, we previously showed that, collectively, genetic variants in estrogen metabolism enzymes were associated with POAG in women and not men.15 Here, we formed a custom testosterone metabolism genetic variant panel in case-control datasets from the United States and Australia/New Zealand to further evaluate if there is an association overall, or stratified by sex, with POAG or POAG subtypes defined by IOP. This panel focused on genes related to the intracrine generation of testosterone, as local production of sex steroid may be most relevant to the glaucomatous process. 
Methods
Description of the Study Populations
The US data are derived from the National Eye Institute Glaucoma Human Genetics Collaboration Heritable Overall Operational Database (NEIGHBORHOOD), a genetic consortium that includes the following eight independent datasets: Massachusetts Eye and Ear Infirmary; National Eye Institute Glaucoma Human Genetics Collaboration; Iowa; Marshfield; the Ocular Hypertension Treatment Study; the Women's Genome Health Study; and two datasets from the Glaucoma Genes and Environment Study: one genotyped on the Affymetrix platform and the other genotyped on the Illumina HapMap Series.16 The NEIGHBORHOOD dataset has a total of 3853 POAG cases and 33,480 controls. The Australian and New Zealand data are derived from the Australian and New Zealand Registry of Advanced Glaucoma (ANZRAG) and consist of 1115 advanced POAG cases and 1992 controls genotyped on the Illumina Omni 1M or the OmniExpress array. Cases and controls were drawn from Southern Adelaide Health Service/Flinders University, University of Tasmania, Queensland Institute of Medical Research, and the Royal Victorian Eye and Ear Hospital.17 All participants in both datasets were of European ancestry. The institutional review boards of all participating institutions approved this study. 
Ophthalmic Characteristics of Cases and Controls
Across the eight datasets in NEIGHBORHOOD, cases and controls lacked evidence of secondary IOP elevation on slit lamp biomicroscopy, such as exfoliation syndrome, pigment dispersion syndrome, or trauma. For cases and controls, slit lamp examination or gonioscopy did not reveal evidence of significant irido-trabecular meshwork apposition suggestive of angle closure in either eye. For cases, fundus examination revealed a cup-disc ratio (CDR) of at least 0.7 or an intereye difference in CDR of at least 0.2. Each case had at least one eye with visual field (VF) loss consistent with nerve fiber layer dropout on a reliable test. In the absence of VF loss, the CDR was 0.8 or higher in both eyes. Elevated IOP was not a criterion for inclusion as a case or a control. Controls had a CDR of 0.6 or less in both eyes and a CDR intereye difference of 0.1 or less. IOP at diagnosis was collected and used to categorize POAG cases into high-tension glaucoma (HTG with IOP ≥22 mm Hg) and normal tension glaucoma (NTG with IOP <22 mm Hg) subtypes when available. The exact definition of POAG across the eight NEIGHBORHOOD sites can be found in Supplementary Table 2 of Cooke Bailey et al.16 
Advanced POAG cases in ANZRAG had best-corrected visual acuity worse than 6/60 due to POAG or a reliable 24-2 VF with a mean deviation worse than −22 db or at least two of four central fixation squares affected with a pattern SD of <0.5%. The less severely affected eye was also required to have signs of glaucomatous disc damage with care taken to exclude secondary glaucomas of all types. Unscreened participants from the Australian Cancer Study (225 esophageal cancer cases, 317 Barrett's esophagus cases, and 552 controls) and from a study of inflammatory bowel diseases (303 cases and 595 controls) were chosen as controls. 
Genotyping Data and Imputation
Details regarding the genotyping of the US and Australian datasets, including information about the genotyping platforms and quality control measures, can be found in the Supplemental Note of Cooke Bailey et al.16 Estimated genotypic probabilities for the loci in the US and Australian dataset imputed to the 1000 Genomes Project reference panel (March 2012)18 were analyzed. 
Generation of Genetic Data for the Testosterone Pathway Analysis
The genome-wide associations between single nucleotide polymorphism (SNP) allele dosage in relation to POAG were analyzed with ProbABEL (GenABEL project developers; http://www.genabel.org/)19 for NEIGHBORHOOD and SNPTEST (University of Oxford; http://mathgen.stats.ox.ac.uk) in ANZRAG.20,21 Logistic regression models adjusting for age, study-specific eigenvectors, and study-specific covariates for each dataset were evaluated. Using METAL (Center for Statistical Genetics, University of Michigan; http://csg.sph.umich.edu/abecasis/metal/index.html) we performed a meta-analysis to assess SNP dosages in relation to POAG across the US datasets.22 SNPs with imputation quality score >0.7 and minor allele frequency >0.05 were carried forward for the US and Australian datasets.16,17 For this work, we used the P values for association with POAG from the 2974 gene variants in NEIGHBORHOOD and the 2617 gene variants in ANZRAG (with 2609 consensus SNPs between datasets) that were attributable to testosterone metabolism for pathway analyses (see Fig. for genes). The number of gene variants differed slightly for each analysis. 
Pathway Analysis by Randomization Incorporating Structure (PARIS) Analysis
As part of the testosterone pathway, we chose to include genes involved in the formation of androstenediol and testosterone, because although they are made in the testes, they are also formed from DHEA produced by the adrenal glands and then undergo intracrine conversion to both androgens and estrogens in local tissues.12 We generated a custom SNP panel derived from 16 genes across 12 chromosomes comprising the testosterone metabolic pathway, as depicted in the Figure. We submitted the P values from SNPs within 50 kB of the start and end sites of these genes to PARIS (v2.4).23 We have previously described PARIS23,24 and used a prior version of this software to assess the estrogen metabolism pathway gene variants in relation to POAG.15 PARIS derives a P value for association between a given gene variant set and outcome of interest using a permutation procedure. Specifically, it first creates a random collection of SNPs with genomic features that mimic features of the user-defined pathway (in this case, testosterone metabolism), then compares the number of statistically significant (P < 0.05) features within the user-defined pathway to the random pathway. We chose to permute 10,000 times to determine an overall likelihood of the random pathway containing more significant features than the user-defined one. For example, for the testosterone pathway SNP set association with POAG among men in NEIGHBORHOOD, PARIS reported 44 significant features; specifically, 27 of 238 “simple features” (SNPs not in any linkage disequilibrium block [LD block]) and 17 of 45 “complex features” (an LD block with two or more SNPs) had P value less than 0.05 for association with POAG. PARIS calculated a permuted P = 0.0001, indicating that only 1 of 10,000 random pathways with genetic architectures similar to the testosterone pathway had a higher significant feature count (>44 significant features with P < 0.05). Initially, these analyses were carried out in men and women together for the outcomes of overall POAG as well as of the HTG and NTG subtypes. Subsequently, associations between testosterone metabolism SNPs and POAG were repeated in men and women separately. We also used the “- -paris-details” flag to investigate specifically which of the genes in the testosterone metabolism pathway were contributing to the significant signal in the overall pathway. Analyses in NEIGHBORHOOD were repeated in ANZRAG using a dataset-specific testosterone SNP set, that is, SNPs in the ANZRAG dataset located within 50 kB of the start and end sites of the 16 genes comprising the testosterone metabolism pathway, because various platforms were used across studies, and different sets of SNPs passed the quality control filters. However, in secondary analysis, we did use the 2609 consensus SNPs between both datasets as the exposure of interest in relation to the various glaucoma phenotypes. Finally, we explored whether our outcomes differed if we excluded SNPs from five genes that coded for overlapping estrogen-metabolizing enzymes (CYP19A1, SULT1E1, HSD17B1, HSD3B1, and SRD5A1). Although unlikely, we also excluded the four small genes with only one feature (HSD17B8, SULT1A1, HSD3B1, and HSD3B2) to minimize any bias they might introduce, as we were primarily interested in collections of genes that worked in biochemical pathways in relation to glaucoma outcomes. 
Results
The mean ages of cases and controls stratified by sex and POAG subtype (HTG versus NTG) are provided in Tables 1 and 2 for the US and Australian datasets, respectively. There is a preponderance of female controls in the US dataset due to the large size of the Women's Genome Health Study,25 which has a case-cohort design. 
Table 1
 
The Mean Age Distribution of POAG Cases and Controls in the NEIHBORHOOD, Stratified by Sex and by IOP (HTG or NTG)
Table 1
 
The Mean Age Distribution of POAG Cases and Controls in the NEIHBORHOOD, Stratified by Sex and by IOP (HTG or NTG)
Table 2
 
The Mean Age and Distribution of POAG Cases and Controls in the ANZRAG, Stratified by Sex and HTG or NTG
Table 2
 
The Mean Age and Distribution of POAG Cases and Controls in the ANZRAG, Stratified by Sex and HTG or NTG
In NEIGHBORHOOD, the testosterone pathway was not associated with POAG overall (permuted P = 0.77; Table 3) but a significant association was noted in ANZRAG (permuted P = 0.018; Table 3). In both datasets, the testosterone pathway was associated with POAG overall among men (permuted P ≤ 0.033) but not among women (permuted P ≥ 0.42). In both datasets, the testosterone pathway was significantly associated with HTG (permuted P ≤ 0.011), but there were inconsistent results with respect to NTG (Table 3). Although the testosterone pathway was associated with NTG in ANZRAG (permuted P < 0.0001), it was not associated with NTG in NEIGHBORHOOD (permuted P = 1.00). These results were essentially identical when evaluating only the overlapping SNPs (as opposed to the dataset-specific SNP sets) between the US and Australian datasets (see Supplementary Material). Further stratification by sex for the POAG subtypes of HTG and NTG as outcomes was not performed due to the smaller sample sizes in both datasets. In both datasets, the relationship between the testosterone pathway and POAG stratified by sex was similar if the five genes that overlap between the estrogen metabolism pathway15 and the testosterone pathway (CYP19A1, SULT1E1, HSD17B1, HSD3B1, and SRD5A1) were excluded from analysis (data not shown). Furthermore, we also performed an analysis deleting four small genes with only one feature (HSD3B1, HSD3B2, HSD17B8, and SULT1A1), as such genes will yield a P < 0.0001 if there is only one SNP in the feature block with P < 0.05; the results of this analysis were the same as the main results (see Supplementary Material). 
Table 3
 
Relation Between the Testosterone Pathway Genetic Variants and POAG HTG and NTG With Sex-Stratified Results
Table 3
 
Relation Between the Testosterone Pathway Genetic Variants and POAG HTG and NTG With Sex-Stratified Results
Similar to the pathway approach, we tested the association between the 16 testosterone pathway genes and our various outcomes to determine the genes that were driving observed associations. When comparing the US and Australian datasets, there was no overlap in the significant driver genes with >1 feature (genes with permuted P < 0.05) responsible for the associations between testosterone pathway and POAG in men and women (Table 4). Furthermore, there were no overlapping significant driver genes with >1 feature responsible for the association between testosterone metabolism SNPs and HTG in both the US and Australian datasets (Table 5). Several genes were responsible for the relationship between the testosterone SNP panel and NTG in ANZRAG, including AKR1C3, HSD17B2, and HSD17B14 (permuted P for gene ≤ 0.036; Table 5). 
Table 4
 
Gene Significance Within the Testosterone Pathway for POAG Patients Stratified by Sex in NEIGHBORHOOD (United States) and ANZRAG (Australia)
Table 4
 
Gene Significance Within the Testosterone Pathway for POAG Patients Stratified by Sex in NEIGHBORHOOD (United States) and ANZRAG (Australia)
Table 5
 
Gene Significance Within the Testosterone Pathway for POAG Patients Stratified by IOP in NEIGHBORHOOD (United States) and ANZRAG (Australia)
Table 5
 
Gene Significance Within the Testosterone Pathway for POAG Patients Stratified by IOP in NEIGHBORHOOD (United States) and ANZRAG (Australia)
We also analyzed individual SNPs in the testosterone metabolic pathway and the various outcomes. As expected, no SNP achieved a P value that passed Bonferroni-corrected significance level (2609 consensus SNPs evaluated for five outcomes; thus, the new significance level is 3.8 × 10–6). The complete results can be found in the Supplementary Material
Discussion
Very little is known about the role of testosterone metabolism in POAG. This work assessed the relationship between gene variants related to the intracrine testosterone metabolism and POAG using two large datasets. In both datasets, we observed that the assembled testosterone pathway SNP set was consistently associated with HTG in two datasets. Furthermore, the pathway was consistently associated with POAG in men but not in women. 
Some of the testosterone pathway genetic associations across the US and Australian datasets were not consistent. Specifically, although the testosterone SNPs were not associated with POAG overall in the US dataset (permuted P = 0.77), a significant association was found in the Australian dataset (permuted P = 0.018). Also, the relationship between the testosterone SNP set and NTG was null in the US dataset, whereas it was significant in the Australian dataset. Various sensitivity analyses using only overlapping SNPs or excluding genes predominately involved in estrogen metabolism did not change the results that were consistent between the US and Australian datasets. We suspect that the inconsistencies between the datasets are due to differing sample size, as the individual genes have very modest effects and no common gene sets emerged as driving the pathway results replicating across the datasets; however, we cannot rule out different disease definitions and environmental influences as the source of the differences we report. 
For the associations between the testosterone SNP set and HTG in the US and Australian datasets, there was no common genetic driver of the relationship between testosterone metabolism SNPs and HTG. Different 17β HSD isoforms play critical roles in the testosterone pathway and are involved in the intracellular generation of markers that bind both androgen and estrogen receptors (Fig.) via the interconversion of androstenedione and testosterone as well as the interconversion of estrone and estradiol. The literature would suggest that the trabecular meshwork cells26 and retinal cells, including retinal ganglion cells,9,27 are under estrogenic control, whereas ocular adnexal tissues are predominately under androgenic control.28 However, little is known of the relationship between the various 17β HSD isoforms and glaucoma, particularly the product of HSD17B14, which was first discovered in the human retina29 and predicts a favorable response to tamoxifen in estrogen receptor–positive breast cancer tissue.30 As for 17β HSD isoforms 7 and 12, which were associated with HTG only in the US dataset, they are involved in local production of estradiol and feature prominently in the endoplasmic reticulum,31 whose subcellular organelles are abundant in the trabecular meshwork.32 More work is needed to understand if and how testosterone metabolism and functional polymorphisms in the various 17β HSD enzymes contribute to endoplasmic reticulum stress found in POAG.33 
The association between the testosterone gene variant set and POAG in men but not women in two datasets was a notable finding. There is some evidence that declining estrogen levels are linked to higher IOP in postmenopausal women34 and that in women, postmenopausal hormones might lower IOP10,35 and be associated with lower glaucoma risk.11,36 In contrast, there are scarce data of the impact of sex hormones on health outcomes in aging males.37 There are, however, sporadic case reports of syndromic male hypogonadism associated with elevated IOP.38,39 Although specific genes in the testosterone pathway that accounted for this apparent sexual dimorphism were identified in both datasets (for example, HSD17B7 in the US dataset and HSD17B1 in the Australian dataset), no common gene drivers of this relation were found. There are a myriad of reasons why gene drivers common to the US and Australian datasets could not be found, including the dissimilar sample sizes, the slight variation in case-control ascertainment, and differences in genotyping platforms that were used. Furthermore, differences in the quality control and the imputation process generated different SNP sets that conformed to our definition of testosterone gene variants in the US and Australian datasets. Nonetheless, the lack of consistent cross-study driver genes raises questions about whether the relationship between testosterone metabolism SNPs and POAG is truly sex specific. This is in contrast to the relationship between the estrogen pathway SNPs and POAG, where we found COMT to be a driver gene for the sexual dimorphic association with glaucoma (associated with HTG in women but not men) in two separate datasets.15 
Study limitations include the fact that none of the individual testosterone pathway SNPs were significantly associated with POAG stratified by sex or with POAG subtypes of HTG/NTG after correcting for multiple comparisons. Nonetheless, it is well known that existing genome-wide association studies of POAG are underpowered to find biologically meaningful gene variants of modest effects due to the need to minimize false discovery rates. In addition, the testosterone metabolic pathway could be construed to extend beyond biochemical pathways focused on intracrine production of sex steroids, such as cholesterol biosynthesis.40 Finally, there is a lack of evidence that a genetic signature associated with sex steroid metabolism is related to varying concentrations of estradiol or testosterone in cells relevant to POAG. Nonetheless, there is accumulating evidence that exposures altering estrogen levels modify the risk of POAG.11,36,41,42 
Our study does have strengths, including the use of two large datasets, the use of common imputed SNPs across the genome, and the use of updated PARIS software with its enhanced ability to refine gene margins. By including a second dataset to compare findings, we were able to provide a more careful interpretation of the relationship between the testosterone metabolic pathway and POAG and POAG subtypes stratified by sex. By jointly analyzing association signals across a large number of genetic variants, pathway analysis allowed for identification of modest cumulative effects, which could have been missed in standard analyses of individual variants. 
In conclusion, in this study involving 40,440 participants from two continents, we observed a significant relationship between the testosterone metabolism SNPs collectively and POAG among men but not among women. We also found that these SNPs were associated with the high-tension subtype of POAG in both the US and Australian dataset, although there was no consensus on driver genes for these pathway associations across the two datasets. 
Acknowledgments
The authors thank Bronwyn Usher-Ridge and Emmanuelle Souzeau for patient recruitment and data collection, Patrick Danoy and Johanna Hadler for genotyping, and Rhys Fogarty for data cleaning. Controls for the ANZRAG discovery cohort were drawn from the Australian Cancer Study, the Study of Digestive Health, and a study of inflammatory bowel diseases. The authors thank David Whiteman and Graham Radford-Smith for collecting the samples used as controls. 
Presented in part at the annual meeting of the Association for Research in Vision and Ophthalmology, Seattle, Washington, United States, May 2016. 
Supported by the National Institutes of Health Grant EY015473, the Harvard Glaucoma Center of Excellence and Margolis fund (Boston, MA, USA) (LRP, JLW), Research to Prevent Blindness, Inc. (New York, NY, USA) (LRP, JER, JLW), the Arthur Ashley Foundation (LRP), the Glaucoma Research Foundation (San Francisco, CA, USA) (YL), American Health Assistance Foundation (Clarksburg, MD, USA) (YL), and the Glaucoma Foundation (New York, NY, USA) (YL). A Horizon Grant to the Massachusetts Eye and Ear Infirmary from Allergan (Irvine, CA, USA) supported the collection of some glaucoma feature data. 
The following infrastructure grants supported portions of this work. UM1 CA186107 provided infrastructure support for the Nurses Health Study (NHS). UM1 CA167552 provided infrastructure support for the Health Professional Follow-up Study (HPFS). NHS program project grant P01 CA87969 supported cancer research, cheek cell collection, and cancer endpoints, as controls from these studies were used in this project. R01 CA49449 supported blood draws in the NHS. R01 HL034594 supported documentation of fatal and nonfatal coronary heart disease in NHS. R01 HL35464 supported documentation of fatal and nonfatal coronary heart disease in HPFS. Patients with cardiovascular endpoints were a subset of the NEIGHBORHOOD study. R01 HL088521 supported documentation of stroke, as controls with this endpoint were included in this study. 
The following grants from the National Human Genome Research Institute (Bethesda, MD, USA) supported the Glaucoma Gene Environment Initiative: HG004728 (LRP), HG004424 (Broad Institute to support genotyping), and HG004446 (C. Laurie, University of Washington, for genotype data cleaning and analysis). 
Genotyping services for the NEIGHBOR study were provided by the Center for Inherited Disease Research (CIDR) and were supported by the National Eye Institute through grant HG005259-01 (JLW). Additionally, CIDR is funded through a federal contract from the National Institutes of Health to The Johns Hopkins University, contract number HHSN268200782096C. The National Eye Institute (Bethesda, MD, USA) through American Recovery and Reinvestment Act grants 3R01EY015872-05S1 (JLW) and 3R01EY019126–02S1 (MAH) supported the collection and processing of samples for the NEIGHBOR dataset. 
Funding for the collection of cases and controls was provided by National Institutes of Health (Bethesda, MD, USA) Grants EY015543 (RRA), HG8701 and UL1TR000427 (MHB), EY006827, HL073389, EY13315 (MAH), EY09611, and EY015473 (LRP), EY009149, HG004608, EY008208 (F. Medeiros), EY015473 (LRP), EY012118 (MAP-V), EY015682 (TR), EY011671 (JER), EY09580 (JER), EY013178 (JSS), EY015872 (JLW), EY010886 (JLW), EY009847 (JLW), EY011008 (L. Zangwill), EY144428, EY144448, and EY18660. 
Support for recruitment of the Australian and New Zealand Registry of Advanced Glaucoma (ANZRAG) was provided by the Royal Australian and New Zealand College of Ophthalmology (RANZCO) Eye Foundation. Funding was provided by the National Health and Medical Research Council (NHMRC) of Australia (No. 535074, 1031362, 1023911, and 1021105). 
The Australian Cancer Study was supported by the Queensland Cancer Fund and the NHMRC of Australia (program number 199600, awarded to David C. Whiteman, Adele C. Green, Nicholas K. Hayward, Peter G. Parsons, David M. Purdie, and Penelope M. Webb and program number 552429, awarded to David C. Whiteman). The Study of Digestive Health was supported by Grant 5 R01 CA 001833 from the US National Cancer Institute (awarded to David C. Whiteman). The Barrett's and Esophageal Adenocarcinoma Genetic Susceptibility Study (BEAGESS) sponsored the genotyping of cases with esophageal cancer and Barrett's esophagus, which were used as unscreened controls in the ANZRAG discovery cohort. BEAGESS was funded by grant R01 CA136725 from the U.S. National Cancer Institute. SM is supported by a Future Fellowship from the Australian Research Council. The authors alone are responsible for the content and writing of the paper. 
Disclosure: J.N. Cooke Bailey, None; P. Gharahkhani, None; J.H. Kang, None; M. Butkiewicz, None; D.A. Sullivan, None; R.N. Weinreb, Eyenovia (C), Bausch&Lomb (C); H. Aschard, None; R.R. Allingham, None; A. Ashley-Koch, None; R.K. Lee, None; S.E. Moroi, None; M.H. Brilliant, None; G. Wollstein, None; J.S. Schuman, None; J.H. Fingert, None; D.L. Budenz, None; T. Realini, None; T. Gaasterland, None; W.K. Scott, None; K. Singh, None; A.J. Sit, None; R.P. Igo Jr, None; Y.E. Song, None; L. Hark, None; R. Ritch, None; D.J. Rhee, None; D. Vollrath, None; D.J. Zack, None; F. Medeiros, None; T.S. Vajaranant, None; D.I. Chasman, None; W.G. Christen, None; M.A. Pericak-Vance, None; Y. Liu, None; P. Kraft, None; J.E. Richards, None; B.A. Rosner, None; M.A. Hauser, None; J.E. Craig, None; K.P. Burdon, None; A.W. Hewitt, None; D.A. Mackey, None; J.L. Haines, None; S. MacGregor, None; J.L. Wiggs, None; L.R. Pasquale, Eyenovia (C), Bausch&Lomb, Inc. (C), The Glaucoma Foundation (S) 
References
Tham YC, Li X, Wong TY, Quigley HA, Aung T, Cheng CY. Global prevalence of glaucoma and projections of glaucoma burden through 2040: a systematic review and meta-analysis. Ophthalmology. 2014; 121: 2081–2090.
Leske MC, Connell AMS, Wu SY, et al. Incidence of open-angle glaucoma: the Barbados Eye Studies. The Barbados Eye Studies Group. Arch Ophthalmol. 2001; 119: 89–95.
Mukesh BN, McCarty CA, Rait JL, Taylor HR. Five-year incidence of open-angle glaucoma: the visual impairment project. Ophthalmology. 2002; 109: 1047–1051.
Walther A, Philipp M, Lozza N, Ehlert U. The rate of change in declining steroid hormones: a new parameter of healthy aging in men? Oncotarget. 2016; 7: 60844–60857.
Noth RH, Mazzaferri EL. Age and the endocrine system. Clin Geriatr Med. 1985; 1: 223–250.
Deschenes MC, Descovich D, Moreau M, et al. Postmenopausal hormone therapy increases retinal blood flow and protects the retinal nerve fiber layer. Invest Ophthalmol Vis Sci. 2010; 51: 2587–2600.
Akar ME, Taskin O, Yucel I, Akar Y. The effect of the menstrual cycle on optic nerve head analysis in healthy women. Acta Ophthalmol Scand. 2004; 82: 741–745.
Agapova OA, Kaufman PL, Hernandez MR. Androgen receptor and NFkB expression in human normal and glaucomatous optic nerve head astrocytes in vitro and in experimental glaucoma. Exp Eye Res. 2006; 82: 1053–1059.
Munaut C, Lambert V, Noel A, et al. Presence of oestrogen receptor type beta in human retina. Br J Ophthalmol. 2001; 85: 877–882.
Vajaranant TS, Maki PM, Pasquale LR, Lee A, Kim H, Haan MN. Effects of hormone therapy on intraocular pressure: The Women's Health Initiative-Sight Exam Study. Am J Ophthalmol. 2016; 165: 115–124.
Newman-Casey PA, Talwar N, Nan B, Musch DC, Pasquale LR, Stein JD. The potential association between postmenopausal hormone use and primary open-angle glaucoma. JAMA Ophthalmol. 2014; 132: 298–303.
Luu-The V, Labrie F. The intracrine sex steroid biosynthesis pathways. Prog Brain Res. 2010; 181: 177–192.
Labrie F. Adrenal androgens and intracrinology. Semin Reprod Med. 2004; 22: 299–309.
Coca-Prados M, Ghosh S, Wang Y, Escribano J, Herrala A, Vihko P. Sex steroid hormone metabolism takes place in human ocular cells. J Steroid Biochem Mol Biol. 2003; 86: 207–216.
Pasquale LR, Loomis SJ, Weinreb RN, et al. Estrogen pathway polymorphisms in relation to primary open angle glaucoma: an analysis accounting for gender from the United States. Mol Vis. 2013; 19: 1471–1481.
Bailey JN, Loomis SJ, Kang JH, et al. Genome-wide association analysis identifies TXNRD2, ATXN2 and FOXC1 as susceptibility loci for primary open-angle glaucoma. Nat Genet. 2016; 48: 189–194.
Gharahkhani P, Burdon KP, Fogarty R, et al. Common variants near ABCA1, AFAP1 and GMDS confer risk of primary open-angle glaucoma. Nat Genet. 2014; 46: 1120–1125.
Abecasis GR, Auton A, Brooks LD, et al. An integrated map of genetic variation from 1,092 human genomes. Nature. 2012; 491: 56–65.
Aulchenko YS, Struchalin MV, van Duijn CM. ProbABEL package for genome-wide association analysis of imputed data. BMC Bioinformatics. 2010; 11: 134.
Wellcome Trust Case Control Consortium. Genome-wide association study of 14,000 cases of seven common diseases and 3,000 shared controls. Nature. 2007; 447: 661–678.
Marchini J, Howie B. Genotype imputation for genome-wide association studies. Nat Rev Genet. 2010; 11: 499–511.
Willer CJ, Li Y, Abecasis GR. METAL: fast and efficeint meta-analysis of genomewide association scans. Bioinformatics. 2010; 26: 2190–2191.
Butkiewicz M, Cooke Bailey JN, Frase A, et al. Pathway Analysis by Randomization Incorporating Structure-PARIS: an update. Bioinformatics. 2016; 32: 2361–2363.
Yaspan BL, Bush WS, Torstenson ES, et al. Genetic analysis of biological pathway data through genomic randomization. Hum Genet. 2011; 129: 563–571.
Ridker PM, Chasman DI, Zee RY, et al. Rationale, design, and methodology of the Women's Genome Health Study: a genome-wide association study of more than 25,000 initially healthy American women. Clin Chem. 2008; 54: 249–255.
Zhou L, Cheng EL, Rege P, Yue BY. Signal transduction mediated by adhesion of human trabecular meshwork cells to extracellular matrix. Exp Eye Res. 2000; 70: 457–465.
Cascio C, Deidda I, Russo D, Guarneri P. The estrogenic retina: the potential contribution to healthy aging and age-related neurodegenerative diseases of the retina. Steroids. 2015; 103: 31–41.
Rocha EM, Wickham LA, da Silveira LA, et al. Identification of androgen receptor protein and 5alpha-reductase mRNA in human ocular tissues. Br J Ophthalmol. 2000; 84: 76–84.
Haeseleer F, Palczewski K. Short-chain dehydrogenases/reductases in retina. Methods Enzymol. 2000; 316: 372–383.
Sivik T, Gunnarsson C, Fornander T, et al. 17 beta-Hydroxysteroid dehydrogenase type 14 is a predictive marker for tamoxifen response in oestrogen receptor positive breast cancer. PLoS One. 2012; 7: e40568.
Tsachaki M, Birk J, Egert A, Odermatt A. Determination of the topology of endoplasmic reticulum membrane proteins using redox-sensitive green-fluorescence protein fusions. Biochim Biophys Acta. 2015; 1853: 1672–1682.
Grierson I, Marshall J, Robins E. Human trabecular meshwork in primary culture: a morphological and autoradiographic study. Exp Eye Res. 1983; 37: 349–365.
Peters JC, Bhattacharya S, Clark AF, Zode GS. Increased endoplasmic reticulum stress in human glaucomatous trabecular meshwork cells and tissues. Invest Ophthalmol Vis Sci. 2015; 56: 3860–3868.
Ebeigbe JA, Ebeigbe PN. Sex hormone levels and intraocular pressure in postmenopausal Nigerian women. Afr J Med Med Sci. 2013; 42: 317–323.
Tint NL, Alexander P, Tint KM, Vasileiadis GT, Yeung AM, Azuara-Blanco A. Hormone therapy and intraocular pressure in nonglaucomatous eyes. Menopause. 2010; 17: 157–160.
Pasquale LR, Rosner BA, Hankinson SE, Kang JH. Attributes of female reproductive aging and their relation to primary open-angle glaucoma: a prospective study. J Glaucoma. 2007; 16: 598–605.
Decaroli MC, Rochira V. Aging and sex hormones in males. Virulence. 2017; 8: 545–570.
Matlach J, Grehn F, Klink T. Klinefelter syndrome associated with goniodysgenesis. J Glaucoma. 2013; 22: e7–e8.
Pamuk BO, Torun AN, Kulaksizoglu M, et al. 49,XXXXY syndrome with autoimmune diabetes and ocular manifestations. Med Princ Pract. 2009; 18: 482–485.
Ohnesorg T, Keller B, Hrabe de Angelis M, Adamski J. Transcriptional regulation of human and murine 17beta-hydroxysteroid dehydrogenase type-7 confers its participation in cholesterol biosynthesis. J Mol Endocrinol. 2006; 37: 185–197.
Wang YE, Kakigi C, Barbosa D, et al. Oral contraceptive use and prevalence of self-reported glaucoma or ocular hypertension in the United States. Ophthalmology. 2016; 123: 729–736.
Pasquale LR, Kang JH. Female reproductive factors and primary open-angle glaucoma in the Nurses' Health Study. Eye (Lond). 2011; 25: 633–641.
APPENDIX
Members of the ANZRAG Consortium
Department of Ophthalmology, Flinders University, Adelaide, South Australia, Australia: Shiwani Sharma, Sarah Martin, Tiger Zhou, Emmanuelle Souzeau, John Landers, Jude T. Fitzgerald, Richard A. Mills, Rhys Fogarty. 
Department of Anatomical Pathology, Flinders University, Flinders Medical Centre, Adelaide, South Australia, Australia: Sonja Klebe. 
Ophthalmology and Vision Science, Macquarie University, Sydney, New South Wales, Australia: Stuart L. Graham. 
South Australian Institute of Ophthalmology, University of Adelaide, Adelaide, South Australia, Australia: Robert J. Casson, Mark Chehade. 
Centre for Eye Research Australia (CERA), University of Melbourne, Royal Victorian Eye and Ear Hospital, Melbourne, Victoria, Australia: Jonathan B. Ruddle. 
Department of Ophthalmology, University of Sydney, Sydney Eye Hospital, Sydney, New South Wales, Australia: Ivan Goldberg. 
Centre for Vision Research, Westmead Millennium Institute, University of Sydney, Westmead, New South Wales, Australia: Andrew J. White, Paul R. Healey. 
QIMR Berghofer Medical Research Institute, Brisbane, Queensland, Australia: Grant W. Montgomery, Nicholas G. Martin, Graham Radford-Smith, David C. Whiteman, Matthew H. Law. 
University of Queensland Diamantina Institute, Brisbane, Queensland, Australia: Matthew A Brown, Katie Cremin. 
Figure
 
The testosterone pathway is depicted. Gene names are as follows: AKR1C3-ALDO-KETO reductase family 1, member C3, also known as HSD17B5; HSD17B 1, 2, 3, 7, 8, 12, 14–17 beta hydroxysteroid dehydrogenase isoforms; CYP19A1: cytochrome P450, family 19, subfamily A, polypeptide 1; SULT1E1: sulfotransferase family 1E, estrogen-preferring, member 1; SULT1A1: sulfotransferase family 1A, cytosolic, phenol-preferring, member 1; SRD5A1, 2, 3: steroid 5-alpha-reductase 1, 2, 3; HSD3B 1, 2: 3-beta-hydroxysteroid dehydrogenase 1, 2. AR, androgen receptor; ER, estrogen receptor; DHT, dehydroxytestosterone.
Figure
 
The testosterone pathway is depicted. Gene names are as follows: AKR1C3-ALDO-KETO reductase family 1, member C3, also known as HSD17B5; HSD17B 1, 2, 3, 7, 8, 12, 14–17 beta hydroxysteroid dehydrogenase isoforms; CYP19A1: cytochrome P450, family 19, subfamily A, polypeptide 1; SULT1E1: sulfotransferase family 1E, estrogen-preferring, member 1; SULT1A1: sulfotransferase family 1A, cytosolic, phenol-preferring, member 1; SRD5A1, 2, 3: steroid 5-alpha-reductase 1, 2, 3; HSD3B 1, 2: 3-beta-hydroxysteroid dehydrogenase 1, 2. AR, androgen receptor; ER, estrogen receptor; DHT, dehydroxytestosterone.
Table 1
 
The Mean Age Distribution of POAG Cases and Controls in the NEIHBORHOOD, Stratified by Sex and by IOP (HTG or NTG)
Table 1
 
The Mean Age Distribution of POAG Cases and Controls in the NEIHBORHOOD, Stratified by Sex and by IOP (HTG or NTG)
Table 2
 
The Mean Age and Distribution of POAG Cases and Controls in the ANZRAG, Stratified by Sex and HTG or NTG
Table 2
 
The Mean Age and Distribution of POAG Cases and Controls in the ANZRAG, Stratified by Sex and HTG or NTG
Table 3
 
Relation Between the Testosterone Pathway Genetic Variants and POAG HTG and NTG With Sex-Stratified Results
Table 3
 
Relation Between the Testosterone Pathway Genetic Variants and POAG HTG and NTG With Sex-Stratified Results
Table 4
 
Gene Significance Within the Testosterone Pathway for POAG Patients Stratified by Sex in NEIGHBORHOOD (United States) and ANZRAG (Australia)
Table 4
 
Gene Significance Within the Testosterone Pathway for POAG Patients Stratified by Sex in NEIGHBORHOOD (United States) and ANZRAG (Australia)
Table 5
 
Gene Significance Within the Testosterone Pathway for POAG Patients Stratified by IOP in NEIGHBORHOOD (United States) and ANZRAG (Australia)
Table 5
 
Gene Significance Within the Testosterone Pathway for POAG Patients Stratified by IOP in NEIGHBORHOOD (United States) and ANZRAG (Australia)
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×