Article Text
PDF
Abstract
Purpose Glaucoma, a major cause of irreversible blindness, is a highly heritable human disease. Currently, the majority of the risk genes for glaucoma are unknown. We established the Genetics of Glaucoma Study (GOGS) to identify disease genes and improve genetic prediction of glaucoma risk and response to treatment.
Participants More than 5700 participants with glaucoma or a family history of glaucoma were recruited through a media campaign and the Australian Government healthcare service provider, Services Australia, making GOGS one of the largest genetic studies of glaucoma globally. The mean age of the participants was 65.30±9.36 years, and 62% were female. Participants completed a questionnaire obtaining information about their glaucoma-related medical history such as family history, glaucoma status and subtypes, surgical procedures, and prescriptions. The questionnaire also obtained information about other eye and systemic diseases. Approximately 80% of the participants provided a DNA sample and ~70% consented to data linkage to their Australian Government Medicare and Pharmaceutical Benefits Scheme schedules.
Findings to date 4336 GOGS participants reported that an optometrist or ophthalmologist has diagnosed them with glaucoma and 3639 participants reported having a family history of glaucoma. The vast majority of the participants (N=4393) had used at least one glaucoma-related medication; latanoprost was the most commonly prescribed drug (54% of the participants who had a glaucoma prescription). A subset of the participants reported a surgical treatment for glaucoma including a laser surgery in 2008 participants and a non-laser operation in 803 participants. Several comorbid eye and systemic diseases were also observed; the most common reports were ocular hypertension (53% of the participants), cataract (48%), hypertension (40%), nearsightedness (31%), astigmatism (22%), farsightedness (16%), diabetes (12%), sleep apnoea (11%) and migraines (10%).
Future plans GOGS will contribute to the global gene-mapping efforts as one of the largest genetic studies for glaucoma. We will also use GOGS to develop or validate genetic risk prediction models to stratify glaucoma risk, particularly in individuals with a family history of glaucoma, and to predict clinical outcomes (eg, which medication works better for an individual and whether glaucoma surgery is required). GOGS will also help us answer various research questions about genetic overlap and causal relationships between glaucoma and its comorbidities.
- OPHTHALMOLOGY
- Glaucoma
- GENETICS
Data availability statement
Data are available on reasonable request. One of the primary motivators for establishing GOGS was to contribute to international efforts to better understand genetic predisposition to glaucoma. Although the individual-level data will not be available due to ethical considerations, genome wide association study summary statistics will be made available upon publication following meta-analysis with studies from the International Glaucoma Genetics Consortium. Summary statistics will be deposited to a public repository. The full GOGS questionnaire is included as online supplemental material and we encourage collaborations with researchers from other studies to help advance our collective understanding of glaucoma genetics (contact Stuart MacGregor, stuart.macgregor@qimrberghofer.edu.au).
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
STRENGTHS AND LIMITATIONS OF THIS STUDY
Genetics of Glaucoma Study (GOGS) is one of the largest genetic studies of glaucoma in the world, providing a valuable source for glaucoma gene discovery.
GOGS includes a rich set of phenotypes such as family history of glaucoma as well as a range of clinical traits.
Self-reported data complemented with more detailed information through linkage to the Australian Government Medicare and Pharmaceutical Benefits Scheme schedules.
Higher female participation; a well-known limitation for many research studies developed based on volunteer participation.
The vast majority of the participants are of European descent, limiting statistical power for other ancestry groups.
Introduction
Glaucoma, a major cause of irreversible blindness worldwide, is characterised by a progressive degeneration of the optic nerve.1 2 Primary open-angle glaucoma (POAG) is one of the major types of glaucoma, often accompanied by an elevated intraocular pressure (IOP).1 2 POAG is a highly heritable human disease with 70% of the variation in risk attributed to genetics.3 However, the currently known risk genes for POAG collectively explain only ~10% of the familial relative risk.4 Hence, new studies and larger international consortia are required to accelerate discovery of new glaucoma risk genes. Identifying risk genes and genetic variants for POAG helps us understand the underlying biology of the disease. Gene discovery also has clinical implications; identifying disease genes enables genetic risk prediction and assists with target identification in efforts to develop new treatments.
Genetic risk prediction provides an opportunity to identify people at high risk before the clinical onset of the disease. Currently, nearly half of the POAG cases remain undiagnosed until the optic nerve is damaged,2 5 with associated vision loss. These people may benefit from early screening and intervention with ocular hypertension treatments that have been shown to prevent or slow down disease progression.6 Early screening and intervention may also reduce healthcare costs by reducing the need for costly surgical procedures through early disease management, and through reducing screening of those at low risk. Currently, ~10% of the general population have a family history of glaucoma, which raises their risk of glaucoma ninefold, so are recommended to be screened 1–2 times yearly.7 However, the majority of those with family history do not go on to develop glaucoma and a risk stratification approach that also includes genetic risk predictors could prioritise people who require more intensive screening and intervention while people at low risk could undergo less frequent follow-up.
We previously showed that genetic risk prediction built on common glaucoma-associated single-nucleotide polymorphisms (SNPs with minor allele frequency >0.01) is predictive of glaucoma risk, progression and age at diagnosis.8 However, given the polygenic architecture of POAG with a large number of SNPs contributing to the disease risk, each with a small effect, as well as the necessity for multiple testing correction to control the type I error rate in Genome-Wide Association Studies (GWAS), large sample sizes are needed to achieve the statistical power required for gene discovery to narrow down the ‘missing heritability’ gap. We previously estimated that increasing the sample size of glaucoma GWAS studies to 400K cases and a matched number of controls will help identify additional glaucoma risk variants that would explain up to 80% of the genetic variance in glaucoma and improve the predictive ability of the current genetic risk prediction models up to an area under curve of ~0.8.9 Given that developing a single study to target this large sample size for glaucoma is logistically and financially challenging, international collaborations are required. Hence, to improve glaucoma gene discovery and risk prediction, we aimed to develop a large glaucoma genetics study in Australia, which will contribute to international collaborative efforts.
In addition to the potential for early intervention, gene mapping also opens up additional avenues for personalised medicine. Previous studies have shown association of genetic variants with differential response to glaucoma treatment between individuals.10 We have also shown that SNP-based genetic risk prediction is capable of predicting clinical outcomes and the probability of requiring surgical intervention.8 Hence, from the personalised medicine perspective, prediction of response to treatment and surgical interventions based on an individual’s genetic architecture paves the way for establishing personalised treatments where each person could have a unique ‘treatment cocktail’ based on their distinct genetic mechanisms contributing to their glaucoma risk.
The other major clinical significance of gene mapping includes developing new treatments for POAG. Identifying novel disease genes can allow repurposing of existing, approved, drugs or provide potential new targets for developing novel drugs.11 All current treatments for POAG aim at lowering IOP, which is effective in slowing down the disease progression; however, these treatments cannot restore optic nerve damage that has already occurred. Hence, development of neuroprotective treatments by targeting the risk genes that damage the optic nerve independent of IOP would be a major advance in glaucoma prevention.
To achieve the above purposes, we established the Genetics of Glaucoma Study (GOGS), a large study comprising >5700 (as of April 2022) Australian participants with glaucoma or a family history of glaucoma. We will primarily use this data to expand our understanding of genetic risk factors for glaucoma and improve genetic risk prediction. A unique facet of this study is the involvement of a large number of people with a family history of glaucoma; studying this group may help us improve risk prediction accuracy among this high-risk group and inform mitigation strategies to help them manage their risk.
Objectives
Combining data collected in this study with independent glaucoma datasets worldwide, GOGS aim to achieve the following collective goals:
Identify novel genetic risk loci for glaucoma.
Improve genetic risk prediction for glaucoma, with an emphasis on risk prediction among people with a family history of glaucoma, and develop genetic risk assessment tools to identify people at high risk.
Investigate genetic influences on response to glaucoma therapies (such as IOP-lowering medications).
Investigate genetic links between glaucoma and a wide range of eye diseases, as well as systemic diseases such as Alzheimer’s disease, cardiovascular diseases and diabetes.
Investigate causality of modifiable putative risk factors for glaucoma such as body mass index, education levels and tobacco smoking.
Cohort description
Study participants
Participants were recruited via two arms: (1) a media campaign that invited individuals with glaucoma or a family history to take part and (2) direct mailouts to people who have been prescribed a glaucoma medication within the last year (see details below).
Patient and public involvement
Several participants provided feedback on the study questionnaire and two participants, one with glaucoma and one with a family history of glaucoma, were involved in the study press conference to encourage participation in this study. Moreover, we consulted three ophthalmologists to codesign the study questionnaire in order to ensure that relevant clinical and surgical information is obtained from GOGS participants. We provided participants with a newsletter to inform them about the overall study progress, with further newsletters to be provided to update them on the study results and acknowledge their contributions to related publications.
Recruitment strategy
Media campaign: For this pathway, QIMR Berghofer’s External Relations Team advertised the study through the QIMR Berghofer website and Facebook. The adverts invited participants over 18 years old, who self-reported a diagnosis of glaucoma or who reported having a family history of glaucoma to volunteer.
Recruitment through medical and pharmaceutical prescription history: This recruitment pathway was conducted with assistance from Services Australia, formerly known as the Department of Human Services. Services Australia is part of the Australia Government that provides social and healthcare services to Australians. Medical and pharmaceutical services are subsidised by the Australian Government through the Medicare Benefits Schedule (MBS) and Pharmaceutical Benefits Scheme (PBS), respectively. Following regulatory and ethics approval, Services Australia is able to send invitation letters to people with certain healthcare (MBS) and pharmaceutical (PBS) records to invite them to participate in medical research studies.
On obtaining approvals from the QIMR Berghofer and Services Australia human research ethics committees, 32 000 invitation letters were sent by Services Australia on behalf of our study. People were invited to take part based on the following glaucoma-related procedure MBS item codes—10914, 11200, 42752, 42705, 42744, 42746, 42749, 42752, 42755, 42758, 42770, 42 782 and 42794—who were also prescribed any glaucoma medications (please see details in the Glaucoma medication section). This was performed in two waves. In the pilot wave to estimate the participation rate and overall success of this approach, 2000 invitation letters were sent to people over 18 years old who had at least 6 months of continuous glaucoma medication supply on their PBS records. Following the successful recruitment of more than 200 people (a response rate of 10%), we sent 30 000 additional invitation letters in the second wave.
The following changes were made to the selection criteria relative to the pilot wave: (1) we enrolled participants between 18 and 70 years old (rather than over 18 years from the pilot study). This was because people over 70 years old may be less able/willing to participate due to other health issues, and a subset of them may develop glaucoma due to advanced age rather than a high genetic load (rendering them less suitable for the gene discovery and risk prediction aims of this study). (2) Participants had to have at least 2 months of glaucoma medication supply (unlike in our pilot study, supply did not have to be continuous) within the past year at the time of invitation (ie, between 25 February 2020 and 24 February 2021). We opted to remove the continuity of supply requirement as it greatly limited the number of potential recruits. The vast majority of these people (91%) confirmed they have glaucoma (self-report) although a small fraction had ocular hypertension, the major risk factor for glaucoma.
Managing privacy concerns: The invitation letters highlighted that Services Australia had not provided any information to the researchers about people who have received the letters. Letters contained a brief description of the study aims, participation and withdrawal policies (highlighting that the study is completely voluntary and people can withdraw from the study at any time), as well as the study webpage and contact details (study-specific email and phone).
Enrolment procedure
People invited by both routes described above were asked to enrol in the study by filling out a survey through visiting our study webpage (https://www.qimrberghofer.edu.au/genetics-of-glaucoma/) or requesting a hard copy version of the survey. Enrolment involved reading through a participant information sheet that contained information including the study aims, participation eligibility, and the study ethics and policies on participation, withdrawal, confidentiality, storage and sharing of biological samples and questionnaire data, and any use of the data in future studies. Participants then consented that they had read and understood the information sheet, were willing to provide a sample of their saliva for DNA extraction and genotyping, and that their data may be used for other medical research studying health and behaviour in the future, subject to review by the appropriate research ethics committees. Participants were then directed to fill out the study questionnaire obtaining glaucoma and other health-related information. For subsequent research work and data analysis, all participant data were deidentified.
The questionnaire information
The study questionnaire recorded participants’ demographics and medical history related to glaucoma including glaucoma status and subtypes, their age at diagnosis, surgical procedures, prescriptions, and family history. The questionnaire also captured information about other eye conditions (eg, myopia and cataract), and other systemic diseases (eg, diabetes and dementia). A summary of the questionnaire topics is shown in online supplemental table 1 and the full questionnaire is presented as online supplemental material. Questions that were compulsory for the participant to answer have been marked with a red asterisk (*).
Supplemental material
Supplemental material
Consent to supply DNA
As part of the study consent, we obtained participants’ consent to provide a sample of their saliva for DNA extraction and genotyping. For a subset of the participants (N=544) with existing genotypes, we further obtained their consent for us to access their genotypes obtained from a previous genetic test performed through a direct-to-consumer genotyping company (ie, AncestryDNA, 23andMe, MyHeritage, FamilyTreeDNA and LivingDNA) or from QIMRB QSkin study12 or QIMRB Australian Genetics of Depression Study (AGDS).13 If people were unwilling to share their existing genotypes, we sent them a saliva collection kit, similar to the participants without existing genotypes (see details below).
Participants who consented to provide their direct-to-consumer genotyping data were advised to securely upload their existing genotypes to QIMRB server. Following consent, we also obtained internal access to existing genotypes stored at QIMRB for the consented QIMRB QSkin and AGDS participants.
Saliva collection and DNA extraction
An Isohelix GeneFix Saliva Collection Kit along with an MBS/PBS consent form (please see below) was dispatched. Saliva samples were returned to QIMR Berghofer by prepaid post. A reminder email was sent after 2 months to people who had not returned their saliva kits. DNA was extracted in the Sample Processing Lab at QIMR Berghofer.
To reduce the possibility of sample mix-up due to handling and human error, the barcodes of the saliva collection kits were linked to the participant’s study ID associated with their consent forms and questionnaires. On receipt, the kits were unpacked and the barcodes were scanned to automatically link the saliva samples to the participants’ information obtained online. We have additional QC protocols in place to ensure that the participants’ genotypes match with their questionnaire information. For example, we will compare the genotype-inferred and self-reported sex; any mismatches will be excluded from further analysis.
DNA genotyping
At the time of writing, we have plated 3978 DNA samples for genotyping using Illumina Infinium Global Screening Array (San Diego, California, USA). The previously existing genotypes (N=544) were obtained from the following arrays: Illumina OmniExpress (23andMe v3/v4, AncestryDNA, FamilyTreeDNA and MyHeritage) and Illumina Global Screening Array (23andMe v5, LivingDNA, FamilyTreeDNA since 2019, QSkin and AGDS). The minimum number of discrete groups with the maximal SNP overlap will be identified for separate imputation; a batch variable will be used in subsequent analysis. In total, we obtained genotype data from 4522 participants.
MBS and PBS records linkage
Self-reported glaucoma clinical information such as age at diagnosis, treatment procedures or surgeries, and medications prescribed will be supplemented by linkage to the Australian Government MBS/PBS records, access to which provide us more accurate clinical and prescription data. For example, some participants may not recall their exact type of glaucoma surgeries or medications prescribed; by linking data to their PBS and MBS records, we can retrieve such clinical information.
On approval from Services Australia External Request Evaluation Committee, we obtained participants’ consent to access to their glaucoma-related MBS/PBS records for the last 4.5 years. As opposed to the study consent, giving consent to the MBS/PBS component was not required for participation in this study; people could still participate without agreeing to provide their MBS/PBS data. The requested MBS/PBS items are presented in online supplemental table 2). As of April 2022, approximately 70% of the participants consented to their MBS/PBS records access.
Findings to date
Sample characteristics
In total, 5737 people consented to participate and filled out the study questionnaire. The vast majority of participants self-reported their ancestry as European (5104 of 5520; 92.5%). Other self-reported ancestries included Asian (N=205; 3.7%), Indigenous Australian or Torres Strait Islander (N=26; 0.47%), Middle Eastern (N=25; 0.45%), Latino (N=24; 0.43%), African (N=11; 0.2%), Pacific Islander (N=5; 0.09%) and mixed ancestry (N=120; 2.17%). Participants were aged between 19 and 100 at the time of recruitment, with a mean age of 65.30 (SD=9.36), and the majority were female (N=3537; 61.7%). The demographic and study participation characteristics are summarised in table 1.
Demographic and participation characteristics in GOGS
Of 5737 participants, 3453 (60%) were recruited via the media campaign and 2284 (40%) via invitation letters sent through Services Australia. The participation rate from the Services Australia recruitment arm was 7% (2284 of 32 000 invited). The participation characteristics were similar between the media and Services Australia arm (table 1), apart from sex where the female-to-male ratio was higher in the media arm (2.28 and 1.06 for the media and Services Australia arms, respectively).
Of 5737 participants, 4522 (~80%) provided a saliva sample (or access to genotype data from a previous study) and ~70% consented to the study to access their Medicare and PBS data access.
Glaucoma status and subtypes
Of 5737 participants, 4336 self-reported glaucoma and 3639 reported a family history of glaucoma. Figure 1 shows the number of people according to glaucoma and family history status. The proportion of glaucoma cases was higher in the Services Australia arm (91% for Services Australia vs 65% for the media arm). This was expected given that Services Australia participants were invited from people who had at least 2 months of glaucoma medication supply within the last year of recruitment. By contrast, the proportion of participants with a family history of glaucoma was higher in the media arm compared with Services Australia (51% and 71% in Services Australia and media arm, respectively). Among people with glaucoma, 52% and 59% reported a family history in the Services Australia and media arm, respectively.
The distribution of glaucoma and family history in GOGS. The X-axis shows whether the participants have been diagnosed with glaucoma by an ophthalmologist or optometrist (self-reported answers YES, NO and Don’t know). The legend shows self-reported family history of glaucoma (YES, NO and Don’t know). The numbers represent counts in each category. Given that participation in this study required individuals to have glaucoma or a family history, this study did not include individuals who do not have or are unaware of both conditions (eg, no glaucoma and no family history, no glaucoma and unaware of family history), and hence, these categories are missing in this figure. GOGS, Genetics of Glaucoma Study.
The majority of the people who reported glaucoma were not aware which subtype of glaucoma they have. Of the reported subtypes, POAG was the most common (791 of 1134; 69.75%), followed by ACG (185 of 1134; 16.31%), with smaller proportions reporting other glaucoma types (figure 2). Some people reported various combinations of these glaucoma subtypes. Of 4336 people with glaucoma, only 254 were aware of their glaucoma pressure subtypes: 126 reported high-tension glaucoma and 128 normal-tension glaucoma.
Self-reported glaucoma subtypes in GOGS. This figure shows the frequency of glaucoma subtypes among the participants who provided this information (N=1134). Some participants reported multiple subtypes, resulting in more than 35 combinations. The reported statistics in this figure includes the overlapping individuals across the glaucoma subtype categories. ACG, angle closure glaucoma; GOGS, Genetics of Glaucoma Study; POAG, primary open-angle glaucoma.
Glaucoma medication
In addition to obtaining medication data from PBS, we asked the participants whether they had used any of the 19 commonly prescribed glaucoma medications for treatment of glaucoma in Australia (table 2). Of 4393 people who answered to this question, 1878 had used only 1 of those 19 medications, 2436 had used various combinations and 79 reported other medications (table 2). Latanoprost was the most commonly prescribed drug; 54% of the participants (2369 of 4393) had used latanoprost at least once.
Frequency of self-reported glaucoma medications prescribed among the GOG participants
Surgical procedures for glaucoma
We also obtained data on laser or non-laser surgical treatments for glaucoma. Among the participants who provided this information, 2008 reported a laser treatment, of whom 1048 knew of their specific surgical procedure. Of these, the majority (672 of 1048) had a selective laser trabeculoplasty while 312 had a laser peripheral iridotomy and nine underwent a cyclodiode procedure, with some people reporting a combination of these procedures (figure 3). Similarly, 803 reported a non-laser glaucoma operation, 684 of whom knew of their specific surgical procedure. Of these, 238 had stent insertion (MIGS/iStent), 237 trabecectomy, 50 glaucoma drainage device, 44 tube insertion and 1 sclerectomy, with some people reporting a various combination of these surgical procedures.
Frequency of the laser treatment surgical procedures among the GOGS participants who provided this information (N=1048). GOG, Genetics of Glaucoma Study; PI, peripheral iridotomy; SLT, selective laser trabeculoplasty.
Other diseases in GOGS
We also obtained information on whether the participants had any other eye or systemic diseases. For eye diseases, of 5459 who provided information, 4979 had at least one of the eye conditions listed in our survey (figure 4A), 350 did not report any, and 130 reported other conditions. The most commonly reported eye diseases were ocular hypertension (2913 of 5459; 53.36%), cataract (2599 of 5459; 47.61%), nearsightedness (1704 of 5459; 31.21%), astigmatism (1223 of 5459; 22.40%) and farsightedness (899 of 5459; 16.47%). Similarly, of 5674 people who provided information, 3410 had at least one of the systemic diseases listed in the survey (figure 4B), 2049 did not report any and 215 had other diseases. The most commonly reported systemic diseases were hypertension (2264 of 5674; 39.9%), diabetes (706 of 5674; 12.4%) and sleep apnoea (629 of 5674; 11.1%).
Frequency of other diseases among the GOGS participants. (A) Self-reported non-glaucoma eye diseases in 5459 participants. (B) Self-reported systemic diseases in 5674 participants. GOGS, Genetics of Glaucoma Study. AMD, age-related macular degeneration.
Discussion
We developed the GOGS, a large genetic study enriched with glaucoma-related clinical information. The study comprises more than 5700 participants with glaucoma or a family history of glaucoma across Australia. Of these, ~80% provided a DNA sample and ~70% consented to data linkage to their Medicare and PBS data. GOGS is one of the largest studies of its kind internationally and when combined with previous and upcoming glaucoma genetic datasets, we anticipate GOGS improving power to detect novel risk loci for glaucoma. GOGS also allows for developing or validating genetic prediction models for glaucoma risk and response to treatment, as well as clinical outcomes (eg, whether the POAG polygenic risk score is predictive of the necessity, type and success of surgery in glaucoma patients). Another strength of GOGS is the availability of data on family history of glaucoma, enabling us to develop risk prediction models specifically for people in this high-risk group.
The demographic and characteristics of the participants (eg, age, education, smoking status) were similar between the media campaign and Services Australia recruitment arms of this study, apart from a higher rate of female participation through the media arm compared with the Services Australia approach. Although the proportion of sexes was relatively equal for the Services Australia participants, we could not systematically investigate a sex-driven selection bias in the Services Australia arm as we do not know the demographic and sex of invited people due to ethical limitations (but we should note that sex was not a selection criterion for invitation). Although participation rate for the Services Australia arm was relatively low (7%), it proved to be an effective way of recruiting people who are very likely to have glaucoma, with a per-person cost comparable to that of the media approach.
A high proportion of the GOGS participants reported a family history of glaucoma; 51% in the Services Australia arm and 71% in the media arm. While the high rate in the media arm could be due to ascertainment bias (people were invited if they had either glaucoma or a family history of glaucoma), among patients with glaucoma, the rate of reported family history was 52% and 59% in the Services Australia and media arms, respectively. These rates are consistent with the previously reported rates of 40%–60% in several studies.14–16 In particular, direct examination of relatives of patients with POAG in a previous study showed that ~60% have affected relatives,16 suggesting that ascertainment bias may not be the main driver of a high rate of family history among patients with glaucoma in GOGS. Although we cannot rule out various biases such as recall and survival bias in reporting family history, such biases are expected to result in underestimation of family history,17 18 which does not seem to be the case among the glaucoma patients in GOGS.
While all the GOGS participants reported whether they have been diagnosed with glaucoma, the majority were not aware of their glaucoma subtype. We believe that the vast majority of the glaucoma cases with unknown subtype in GOGS are POAG. This is because the vast majority (92%) of GOGS participants are of European ancestry; in a previous study of European ancestry individuals with glaucoma, 87% had POAG19 and we anticipate a similarly high percentage in GOGS. Among the GOGS participants who were aware of their glaucoma subtype (1134 of 4336 glaucoma patients), 791 (~70%) reported POAG. Hence, our genetic studies on self-reported glaucoma in GOGS would predominantly identify POAG risk loci. However, once we have confirmed and updated subtype status using Medicare data, we will also perform subtype-specific genetic studies to identify or validate unique loci to each subtype. GOGS also provides valuable data to identify shared genetic risk loci across various subtypes and evaluate the ability of the POAG polygenic risk scores to predict glaucoma risk for the other subtypes.
GOGS has several strengths. First, it is one of the very largest single site studies of its kind internationally, which has collected rich glaucoma-related phenotypes such as family history (including which relatives have glaucoma) and a range of clinical information on glaucoma subtypes, medication and surgical procedures. In addition, many study participants consented to recontact and follow-up. In the future, we hope to follow-up a subset of people who have both a family history of glaucoma and an unusually high polygenic risk score; given their very high risk, these people will be targeted for follow-up in a future longitudinal study. By collecting clinical and imaging data on those at very high risk, we will be able to assess the practical utility of genetic data for stratifying people into risk groups where they can receive appropriately tailored screening and care.
Second, GOGS is a valuable source for glaucoma gene discovery. We aim to perform GWAS for glaucoma or other related phenotypes available in GOGS, followed by a meta-analysis of the results with those from other glaucoma genetics data currently in hand (eg, UK Biobank,4 8 20 the Canadian Longitudinal Study on Aging 21 and the International Glaucoma Genetics Consortium4) to identify additional risk variants and genes for glaucoma. Integrating the GWAS data with publicly available omics data, we further aim to identify druggable causal genes, which may provide potential drug target candidates for development of novel treatments for glaucoma. We aim to focus on drug target genes that could lead to development of novel neuroprotective treatments, as the current treatment strategies for glaucoma focus solely on lowering IOP.
Third, GOGS is not limited to only clinical information on glaucoma but also contains other glaucoma-related data such as smoking status, educational attainment, IOP levels, presence of other related eye diseases (eg, cataract and myopia) and systemic diseases (eg, diabetes, sleep apnoea and heart disease). This information is helpful to evaluate links between glaucoma and other diseases, and to investigate whether glaucoma genetics contributes to those comorbidities, and whether the comorbidities affect response to treatment for glaucoma. However, we should note that due to the sampling design, this study is not ideal for answering research questions that require a cohort study. We should also note that the statistical power would be limited for some of the phenotypes in GOGS and hence, a joint analysis with other national and international studies would be beneficial. Participants in GOGS have provided consent for their data to be used in future related studies (subject to ethics approval), providing a great opportunity for future collaborations for glaucoma gene discovery.
We did not collect controls (ie, people without a family history of glaucoma and who have not been diagnosed with glaucoma) for this study. This is because we have access to a variety of genetic data obtained from people without a glaucoma diagnosis across Australia, genotyped on the same array used in this study. For example, we have genetic data for >16K people from the QSkin Sun and Health Study (QSkin), a cohort study to investigate the development of skin cancer and melanoma,12 and ~15K people from the AGDS, a study to investigate genetic and environmental risk factors for depression.13 Both of these studies are also housed and led by researchers within the same institute (QIMR Berghofer), making these datasets accessible for use as study controls on GOGS. We will match and select controls for GOGS for any subsequent association analyses using the standard matching and QC protocols for genetics studies.
GOGS has several limitations. First, it has a selection bias towards female participants, as well as a possible bias towards recruitment of people with less severe forms of glaucoma. This is a well-known limitation for many research studies developed based on volunteer participation. Although a higher rate of female participation imposes a lower statistical power for male-specific studies, given the high genetic correlation (rg=0.99, SE=0.06) between males and females for glaucoma,4 sex-stratified analysis is not a primary aim of this study. To investigate whether GOGS has a selection bias based on disease severity, we will generate genetic risk scores for glaucoma patients and compare with the scores for patients with severe forms of glaucoma in our previous study of the Australian and New Zealand Registry of Advanced Glaucoma.22 Second, most of the information collected in GOGS is self-reported data. However, as mentioned earlier, we will complement the self-reported information with more accurate and detailed MBS/PBS data. Finally, although GOGS is a multiethnic study, the vast majority of the participants are of European descent, and hence the statistical power is fairly limited for studies involving other ancestry groups.
Data availability statement
Data are available on reasonable request. One of the primary motivators for establishing GOGS was to contribute to international efforts to better understand genetic predisposition to glaucoma. Although the individual-level data will not be available due to ethical considerations, genome wide association study summary statistics will be made available upon publication following meta-analysis with studies from the International Glaucoma Genetics Consortium. Summary statistics will be deposited to a public repository. The full GOGS questionnaire is included as online supplemental material and we encourage collaborations with researchers from other studies to help advance our collective understanding of glaucoma genetics (contact Stuart MacGregor, stuart.macgregor@qimrberghofer.edu.au).
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and this study was approved by the ethics committees of QIMR Berghofer Medical research Institute and Services Australia. Participants gave informed consent to participate in the study before taking part.
Acknowledgments
We thank all of the GOGS participants for their time and willingness to participate in this study. We also thank QIMR Berghofer External Relations Team for helping with the media campaign, as well as Services Australia for helping with sending invitation letters to people who were prescribed glaucoma medications. We thank Datatime Services for assisting with mailing study documents and kits. We would like to acknowledge QIMR Berghofer staff Richard Parker, Mary Ferguson, Zoe Maloney, Rebekah Cicero, Lucy Winkler, Susan List-Armitage, Lisa Bowdler, and Tabatha Goncalves for sample collection, lab processing, and data entry. We also acknowledge advice and assistance from David Whiteman and Catherine Olsen from QSkin and Nick Martin from the Australian Genetics of Depression study.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Twitter @yeda_wu
Contributors PG and SM designed the GOGS study and are responsible for the overall content as guarantors. J-SO, ML, JEC, DAM and AWH revised the online questionnaire and provided intellectual input into the content. PG analysed the data. PG and SM drafted the manuscript. WH, SDT, YW, NI, RY, MS, J-SO, ML, JEC, DAM and AWH revised the article for intellectual content. All authors have read and approved the final version of the manuscript.
Funding This work was supported by an NHMRC Program Grant (APP1150144) and by NHMRC Fellowships/Investigator grants (PG #1173390, JEC #1154824, DAM #1154543, AWH #2009079 and 1103329, and SM #1154543).
Competing interests SM, JEC and AWH are cofounders of and hold stock in Seonix Bio.
Patient and public involvement Patients and/or the public were involved in the design, or conduct, or reporting, or dissemination plans of this research. Refer to the Methods section for further details.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.