Abstract
Background/aims
Glucagon-like peptide-1 receptor (GLP-1R) agonists regulate blood glucose and are commonly used to treat type 2 diabetes mellitus. Recent work showed that treatment with the GLP-1R agonist NLY01 decreased retinal neuroinflammation and glial activation to rescue retinal ganglion cells in a mouse model of glaucoma. In this study, we used an insurance claims database (Clinformatics Data Mart) to examine whether GLP-1R agonist exposure impacts glaucoma risk.
Methods
A retrospective cohort of patients who initiated a new GLP-1R agonist was 1:3 age, gender, race, classes of active diabetes medications and year of index date matched to patients who initiated a different class of oral diabetic medication. Inverse probability of treatment weighting (IPTW) was used within a multivariable Cox proportional hazard regression model to test the association between GLP-1R agonist exposure and a new diagnosis of primary open-angle glaucoma, glaucoma suspect or low-tension glaucoma.
Results
Cohorts were comprised of 1961 new users of GLP-1R agonists matched to 4371 unexposed controls. After IPTW, all variables were balanced (standard mean deviation <|0.1|) between cohorts. Ten (0.51%) new diagnoses of glaucoma were present in the GLP-1R agonist cohort compared with 58 (1.33%) in the unexposed controls. After adjustment, GLP-1R exposure conferred a reduced hazard of 0.56 (95% CI: 0.36 to 0.89, p=0.01), suggesting that GLP-1R agonists decrease the risk for glaucoma.
Conclusions
GLP-1R agonist use was associated with a statistically significant hazard reduction for a new diagnosis of glaucoma. Our findings support further investigations into the use of GLP-1R agonists in glaucoma prevention.
INTRODUCTION
Glaucoma is the leading cause of irreversible blindness globally. Primary open-angle glaucoma (POAG) is the most prevalent form and is projected to affect approximately 80 million people worldwide by 2040.1 In about 60% of cases, POAG is associated with ocular hypertension.2 However, the fraction of POAG associated with elevated intraocular pressure (IOP) varies significantly across different ethnic and racial groups.3 All available therapies for glaucoma target IOP reduction through either medical or surgical means. However, disease progression can continue despite IOP normalisation, and aggressive IOP lowering is associated with vision-threatening complications.4 5 New therapies, targeting mechanisms of glaucoma beyond elevated IOP, are urgently needed to prevent permanent vision loss in patients who have exhausted existing treatment options.2
Glucagon-like peptide 1 (GLP-1) is an incretin hormone that regulates blood glucose, weight and satiety. Agents that increase GLP-1 receptor (GLP-1R) signalling, including GLP-1 analogues and dipeptidyl peptidase 4 inhibitors, have been developed for the treatment of type 2 diabetes mellitus (DM). In the United States, GLP-1R agonists first gained Food and Drug Administration-approval in 2005 and include exenatide, liraglutide, albiglutide, dulaglutide, semaglutide and lixisenatide. In addition to their effects in the periphery, GLP-1R agonists cross the blood–brain barrier and exert effects in the central nervous system, influencing feeding behaviour, cellular proliferation, mitochondrial function and neuroinflammation.6 GLP-1R agonists are neuroprotective in mouse models of neurodegenerative diseases including Parkinson’s7 and Alzheimer’s8 9 disease. A randomised control trial of 45 patients with Parkinson’s disease found a statistically significant improvement in both motor and non-motor deficits among subjects treated with exenatide.10 In a randomised control trial targeting Alzheimer’s disease, liraglutide improved cognitive function.11 Additional trials testing the novel GLP-1R agonist NLY01 in Parkinson’s (Clinical-Trials.gov, identifier NCT04154072) and Alzheimer’s disease are ongoing.
Recent work has identified shared mechanisms of neurodegeneration in animal models of Parkinson’s disease and glaucoma. In both diseases, neuroinflammation driven by glial cell activation contributes to neuron death.7 12-14 Treatment with NLY01 ameliorated neuroinflammation in animal models of both diseases to reduce dopaminergic neuron7 or retinal ganglion cell (RGC)12 death. In this study, we used an insurance claims database to examine whether GLP-1R agonist use impacted the risk for a new glaucoma diagnosis.
MATERIALS AND METHODS
Data source
Optum’s deidentified Clinformatics Data Mart Database was used for this study. This database contains the medical claims from commercial and Medicare Advantage insurance plans obtained from a large US insurer. All outpatient medical claims (inclusive of office visits, associated diagnoses and laboratory testing) and demographic data for each beneficiary during their enrolment was accessible. The subset of data available for this study included all patients in the database from 1 January 2008 to 30 June 2019.
Cohorts
All patients over 18 years of age who initiated a new GLP-1R agonist (ie, exenatide, liraglutide, albiglutide, dulaglutide, semaglutide and lixisenatide) were eligible for inclusion in the GLP-1R cohort. The date the initial prescription was filled was assigned as the index date of each patient. Exclusion occurred for those with less than 2 years in the plan prior to the index date, and for those who did not see an eye care provider at least once prior to the index date. Patients were also excluded for any history of a glaucoma, glaucoma suspect or ocular hypertension diagnosis, a prior glaucoma procedure/surgery, or an active glaucoma medication prescription in the 6 months prior to the index date. See online supplemental table 1 for all glaucoma-related codes used in this study.
An unexposed comparison cohort was created from patients who initiated a new oral diabetic medication during their time in the insurance plan. Patients who initiated a new GLP-1R agonist were matched 1:3 on age, gender, race and year of index date to the unexposed cohort. For additional equalisation, cohorts were also matched on the number of active diabetic medications at the time of the index date. Active medication was defined as the number of active prescriptions within 45 days of the index date (ie, if a patient initiating a GLP-1R agonist was already using two other diabetic medications, then the matching unexposed patient was also required to be initiating a third, non-GLP-1R agonist, class of diabetic medication). Matched unexposed patients were also required to meet all inclusion and exclusion criteria outlined above.
Outcomes and covariates
The primary outcome of interest was an International Classification of Disease (ICD)-9 or ICD-10 incident diagnosis code for POAG, glaucoma suspect or low-tension glaucoma at any time after the index date. Covariates assessed at the time of the index date were age, gender, race, geographic location, yearly income and education level. Ocular health factors included were the level of diabetic retinopathy (ie, none, non-proliferative or proliferative disease) and any history of anti-vascular endothelial growth factor (VEGF) intravitreal injections prior to the index date, as we have previously shown that anti-VEGF treatments are associated with an increased risk for glaucoma.15 Systemic health states included were hypertension, hypercholesterolaemia, kidney disease (ie, none, chronic or end-stage renal disease) and the Diabetes Complications Severity Index (DCSI). The DCSI is a validated metric that is calculated using diagnosis codes from six categories of diabetic complications.16 We also assessed for statin, metformin and beta-blocker usage at the time of index, as these medications have previously been associated with modulation of glaucoma risk.17 Lab values were also assessed for the level of haemoglobin A1c (HbA1c). Healthcare utilisation and, more specifically, eye care utilisation were accounted for by counting every distinct day with at least one claim for an office visit in the year prior to the index date, and served as proxy for both overall health status and the likelihood of seeking healthcare after the index date. Although smoking is not often directly coded for in administrative datasets, we used a combination of smoking diagnosis codes, use of antismoking drugs and current procedural terminology codes for smoking cessation counselling as proxies for smoking. This method has previously found smoking rates in administrative databases to be 10%–11%, similar to that reported in the general population.18 To account for differences in these baseline covariates between the users of GLP-1R agonists and unexposed cohorts, inverse probability of treatment weighting (IPTW) using the estimated propensity scores from all covariates was performed in all subsequent analyses. IPTW is a statistical method used to create cohorts that are otherwise similar when examining the effect of a treatment or exposure, in this case GLP-1R agonist use, as estimated by propensity score. Each study subject is then given a certain weight when conducting statistical tests or regression modelling which reduces or removes the impact of confounding variables.
Statistical analysis
Descriptive statistics were used to analyse the data, with continuous variables reported using mean and SD and categorical variables using frequency and percentage. A multivariable Cox proportional hazard regression was performed to determine the association between hazard of developing glaucoma and GLP-1R agonist exposure. Patients were censored if any of the following exclusion criteria occurred after the index date: initiation of the comparison cohort of medication, a gap of >60 days or more in active prescription for the cohort medication, a new diagnosis for a different form of glaucoma not included as the outcome, the patient exited from the insurance plan or the end of observation was reached. Two sensitivity analyses were also conducted. The first analysis excluded the ‘glaucoma suspect’ diagnosis codes from the outcome definition. The second analysis censored all unexposed patients at the same time their matched exposed patients were censored to equalise the time in analysis between cohorts. All data analyses were performed using SAS software (V.9.4; SAS Institute).
RESULTS
After inclusion and exclusion criteria were met, 1961 new users of GLP-1R agonists (figure 1) were matched to 4371 unexposed controls. After matching and IPTW, all covariates including race, gender, age, education, income, geographic location, diagnosis of hypertension, diagnosis of hypercholesterolaemia, presence and severity of kidney disease, history of smoking, beta blocker, statin and metformin usage, mean values of HbA1c, presence and severity of diabetic retinopathy, history of anti-VEGF injections, DCSI and mean days of healthcare and eye care utilisation were balanced and comparable between cohorts (standard mean deviation <|0.1| for all comparisons; see table 1 for baseline and weighted characteristics of the study population). After IPTW, the unexposed cohort was 51.9% female, 78.3% white, 12.7% black and 7.5% Hispanic, averaged 56.2 years of age and had a mean HbA1c of 7.98±2.45. This compared with the exposed group after IPTW at 52.8% female, 78.9% white, 11.4% black and 8.0% Hispanic, averaged 55.0 years of age and had a mean HbA1c of 8.00±3.14.
Figure 1.
Flowchart showing inclusion/exclusion criteria for study patients. GLP-1R, glucagon-like peptide 1 receptor.
Table 1.
Baseline characteristics
| Variable | Non-user (n=4371) |
GLP-1R agonist user (n=1961) |
Standardised mean difference |
||
|---|---|---|---|---|---|
| Unweighted | Weighted | Unweighted | Weighted | Weighted | |
| Race | 0.0454 | ||||
| Asian | 69 (1.58%) | 1.53% | 33 (1.68%) | 1.75% | |
| Black | 542 (12.40%) | 12.65% | 234 (11.93%) | 11.35% | |
| Hispanic | 330 (7.55%) | 7.49% | 157 (8.01%) | 7.99% | |
| White | 3430 (78.47%) | 78.32% | 1537 (78.38%) | 78.91% | |
| Gender | 0.0171 | ||||
| Female | 2271 (51.96%) | 51.90% | 1028 (52.42%) | 52.76% | |
| Male | 2100 (48.04%) | 48.10% | 933 (47.58%) | 47.24% | |
| Age | −0.0768 | ||||
| Mean (SD) | 55.63 (10.59) | 56.17 (12.77) | 55.43 (10.43) | 54.96 (18.32) | |
| Median (Q1–Q3) | 56.00 (49.00–63.00) | 57.00 (50.00–63.00) | 56.00 (49.00–62.00) | 56.00 (49.00–62.00) | |
| Education | 0.0265 | ||||
| Less than 12th grade | 14 (0.32%) | 0.29% | 5 (0.25%) | 0.24% | |
| High-school diploma | 1321 (30.22%) | 30.35% | 589 (30.04%) | 31.06% | |
| Less than bachelor degree | 2380 (54.45%) | 54.56% | 1079 (55.02%) | 54.24% | |
| Bachelor degree plus | 638 (14.60%) | 14.48% | 286 (14.58%) | 14.03% | |
| Unknown | 18 (0.41%) | 0.32% | 2 (0.10%) | 0.42% | |
| Income | 0.0267 | ||||
| Unknown | 533 (12.19%) | 12.22% | 246 (12.54%) | 11.69% | |
| <US$40K | 661 (15.12%) | 15.22% | 296 (15.09%) | 15.20% | |
| US$40K-US$49K | 260 (5.95%) | 5.94% | 108 (5.51%) | 6.18% | |
| US$50K-US$59K | 336 (7.69%) | 7.43% | 132 (6.73%) | 7.41% | |
| US$60K-US$74K | 481 (1 1.00%) | 10.56% | 189 (9.64%) | 11.12% | |
| US$75K-US$99K | 708 (16.20%) | 16.16% | 322 (16.42%) | 16.36% | |
| US$100K+ | 1392 (31.85%) | 32.47% | 668 (34.06%) | 32.03% | |
| Geographic location | 0.0481 | ||||
| Mountain | 326 (7.46%) | 7.46% | 148 (7.55%) | 7.10% | |
| Northeast | 314 (7.18%) | 6.60% | 113 (5.76%) | 6.20% | |
| Pacific | 254 (5.81%) | 5.41% | 88 (4.49%) | 5.66% | |
| South Atlantic | 1491 (34.11%) | 34.16% | 677 (34.52%) | 33.66% | |
| Southern Midwest | 695 (15.90%) | 17.72% | 405 (20.65%) | 18.61% | |
| Unknown | 4 (0.09%) | 0.06% | 0 (0.00%) | 0.00% | |
| Upper Midwest | 1287 (29.44%) | 28.58% | 530 (27.03%) | 28.77% | |
| Hypertension | 0.0326 | ||||
| No | 864 (19.77%) | 17.12% | 239 (12.19%) | 15.91% | |
| Yes | 3507 (80.23%) | 82.88% | 1722 (87.81%) | 84.09% | |
| Hypercholesterolaemia | 0.0605 | ||||
| No | 636 (14.55%) | 12.44% | 169 (8.62%) | 10.51% | |
| Yes | 3735 (85.45%) | 87.56% | 1792 (91.38%) | 89.49% | |
| Kidney disease | 0.0391 | ||||
| No | 3105 (71.04%) | 65.37% | 1104 (56.30%) | 63.52% | |
| CKD | 1188 (27.18%) | 32.43% | 798 (40.69%) | 34.07% | |
| ESRD | 78 (1.78%) | 2.20% | 59 (3.01%) | 2.41% | |
| Smoking | 0.0110 | ||||
| No | 3310 (75.73%) | 74.95% | 1451 (73.99%) | 74.48% | |
| Yes | 1061 (24.27%) | 25.05% | 510 (26.01%) | 25.52% | |
| Beta blocker use | 0.0194 | ||||
| No | 3640 (83.28%) | 82.50% | 1605 (81.85%) | 81.75% | |
| Yes | 731 (16.72%) | 17.50% | 356 (18.15%) | 18.25% | |
| Statin use | 0.0223 | ||||
| No | 3302 (75.54%) | 74.08% | 1405 (71.65%) | 73.09% | |
| Yes | 1069 (24.46%) | 25.92% | 556 (28.35%) | 26.91% | |
| Metformin use | −0.0235 | ||||
| No | 1073 (24.55%) | 36.26% | 1162 (59.26%) | 37.39% | |
| Yes | 3298 (75.45%) | 63.74% | 799 (40.74%) | 62.61% | |
| Diabetic retinopathy | 0.0272 | ||||
| None | 3847 (88.01%) | 85.27% | 1589 (81.03%) | 84.51% | |
| NPDR | 411 (9.40%) | 11.84% | 308 (15.71%) | 12.18% | |
| PDR | 113 (2.59%) | 2.89% | 64 (3.26%) | 3.31% | |
| Anti-VEGF treatment | 0.0350 | ||||
| No | 4297 (98.31%) | 98.24% | 1928 (98.32%) | 97.75% | |
| Yes | 74 (1.69%) | 1.76% | 33 (1.68%) | 2.25% | |
| HbA1c level | 0.0061 | ||||
| Mean (SD) | 7.82 (1.90) | 7.98 (2.45) | 8.12 (1.80) | 8.00 (3.14) | |
| Median (Q1–Q3) | 7.30 (6.40–8.70) | 7.60 (6.50–8.90) | 7.80 (6.90–9.00) | 7.70 (6.70–9.00) | |
| DCSI | 0.0253 | ||||
| Mean (SD) | 1.66 (1.98) | 1.84 (2.56) | 2.12 (2.16) | 1.92 (3.62) | |
| Median (Q1–Q3) | 1.00 (0.00–3.00) | 1.00 (0.00–3.00) | 2.00 (0.00–3.00) | 1.00 (0.00–3.00) | |
| Number of active DM med classes (mean (SD)) | 1.66 (0.74) | 1.89 (0.82) | |||
| Days prior to censoring/events | |||||
| Mean (SD) | 266.5 (299.8) | 143.9 (195.1) | |||
| Median (Q1–Q3) | 151 (91–326) | 84 (28–168) | |||
| Days of healthcare usage | 0.0138 | ||||
| Mean (SD) | 7.39 (6.26) | 8.18 (8.85) | 9.28 (6.70) | 8.32 (10.55) | |
| Median (Q1–Q3) | 6.00 (3.00–10.00) | 6.00 (3.00–10.00) | 8.00 (5.00–12.00) | 7.00 (4.00–11.00) | |
| Days of eye care usage | 0.0391 | ||||
| Mean (SD) | 0.72 (0.96) | 0.70 (1.13) | 0.65 (1.10) | 0.79 (2.96) | |
| Median (Q1–Q3) | 1.00 (0.00–1.00) | 1.00 (0.00– 1.00) | 0.00 (0.00–1.00) | 0.00 (0.00–1.00) | |
| Number of events in regular analysis (new diagnosis of glaucoma or glaucoma suspect) | −0.1013 | ||||
| No event | 4313 (98.67%) | 98.66% | 1951 (99.49%) | 99.60% | |
| Had an event | 58 (1.33%) | 1.34% | 10 (0.51%) | 0.40% | |
| Number of events in sensitivity analysis 1 | −0.1001 | ||||
| No event | 4347 (99.45%) | 99.43% | 1960 (99.95%) | 99.98% | |
| Had an event | 24 (0.55%) | 0.57% | 1 (0.05%) | 0.02% | |
| Number of events in sensitivity analysis 2 | −0.1334 | ||||
| No event | 3340 (98.21%) | 98.22% | 1951 (99.49%) | 99.60% | |
| Had an event | 61 (1.79%) | 1.78% | 10 (0.51%) | 0.40% | |
CKD, chronic kidney disease; DCSI, Diabetes Complications Severity Index; DM, diabetes mellitus; ESRD, end-stage renal disease; HbA1c, haemoglobin A1c; NPDR, non-proliferative diabetic retinopathy; PDR, proliferative diabetic retinopathy; VEGF, vascular endothelial growth factor.
During the follow-up period, 58 (1.33%) new diagnoses of glaucoma or glaucoma suspect were present in unexposed controls compared with 10 (0.51%) new diagnoses of glaucoma or glaucoma suspect in the GLP-1R agonist cohort. After IPTW of all covariates, Cox regression analysis revealed a 0.56 HR (95% CI: 0.36 to 0.89, p=0.01; table 2) for incident glaucoma among patients who initiated GLP-1R agonist versus unexposed controls.
Table 2.
Multivariable Cox regression analysis with inverse probability of treatment weighting
| Variable | Category | HR (95% CI) | P value |
|---|---|---|---|
| GLP-1R agonist | 0.01 | ||
| User | 0.56 (0.36 to 0.89) | ||
| Non-user | Ref |
GLP-1R, glucagon-like peptide 1 receptor.
In the first sensitivity analysis, which removed ‘glaucoma suspect’ diagnoses codes from the outcome definition, a reduced number of outcomes occurred in both cohorts, consisting of 1 (0.05%) new diagnosis of glaucoma in the GLP-1R agonist cohort and 24 (0.57%) new diagnoses of glaucoma in the control cohort. Despite this reduction, however, an even stronger protective association for incident glaucoma was found following GLP-1R agonist exposure (HR=0.08, 95%CI: 0.02 to 0.42, p=0.003). The second sensitivity analysis, which matched the time in analysis between cohorts, resulted in three additional outcomes in the unexposed cohort, and a strongly protective HR of 0.21 for incident glaucoma and glaucoma suspect in the GLP-1R agonist cohort (95% CI: 0.13 to 0.33, p<0.001).
DISCUSSION
GLP-1R agonists are a common second-line therapy used in the treatment of type 2 DM. In a mouse model of ocular hypertension, a novel GLP-1R agonist, NLY01, reduced RGC loss.12 Using a national database, we examined the association between exposure to GLP-1R agonists and a new diagnosis of glaucoma or glaucoma suspect. Our analysis identified approximately 6400 patients who fulfilled our inclusion criteria. Cox regression analysis showed a significant reduction in hazard for a new diagnosis of glaucoma in patients exposed to a GLP-1R agonist relative to unexposed patients. These results suggest that, among patients with DM, GLP-1R agonists may reduce the risk for glaucoma.
Preclinical data from a mouse model of ocular hypertension showed that NLY01, a novel GLP-1R agonist designed to maximise central nervous system penetration, can reduce retinal inflammation and RGC death. Although the mechanism of protection has not been definitively determined, evidence suggests that NLY01 acts on microglia, macrophages, and perhaps astrocytes to reduce local retinal inflammation and prevent activation of the complement cascade after IOP elevation. Specifically, GLP-1R agonism reduces microglia/macrophage production of proinflammatory signalling molecules, retinal levels of complement component 3 (C3) and reactive astrocyte formation.12 Together, these immunomodulatory effects may reduce RGC death in ocular hypertension.
Preclinical and randomised clinical trial data suggest that GLP-1R agonists may also be protective in neurodegenerative diseases of the brain, including Alzheimer’s8 9 and Parkinson’s7 disease. Although data on the neuroprotective role of GLP-1R agonists are relatively new, this class of medication is bolstered by more than 15 years of safety data obtained from their widespread use in the treatment of DM, making them an attractive therapeutic option for patients suffering from slowly progressive neurodegenerative diseases who may require decades of therapy to preserve neurological function of the retina or brain.
Our study focused on patients with DM because GLP-1R agonists are FDA-approved to treat DM. However, it should be noted that multiple studies have shown an association between DM and an increased risk of POAG.19-21 Despite this association, the mechanistic link between DM and glaucoma is not well understood, and it is not known whether GLP-1R agonists directly alter glaucoma risk independent of their impact on DM. Of note, our cohorts were comparable with respect to diabetes severity as evidenced by the DCSI score and HbA1c levels, reducing the potential confounding effect of DM in this study.
Although our results showed a statistically significant reduction in hazard for a new glaucoma or glaucoma suspect diagnosis among patients treated with GLP-1R agonists, our analysis identified only 68 patients across both the unexposed and exposed cohorts who met outcome criteria. This was largely due to censoring for patients with a gap in active prescription longer than 60 days. Although this requirement limited the number of patients who met outcome criteria, altering this requirement would bias the study towards the null by including more patients not actively taking a GLP-1R agonist in the exposed cohort. Despite limited numbers, we still found a reduction in disease hazard in those treated with GLP-1R agonists, attesting to the robustness of this association.
Population studies suggest that greater than 50% of glaucoma remains undiagnosed.22-25 In our study, the percentage of patients in the control group that progressed to glaucoma or glaucoma suspect was only 1.33%. This is lower than the 2% prevalence of glaucoma in individuals ≥40 years of age in the USA, and suggests the possibility of underdiagnosis in our cohorts as well.26 However, it is not possible to know exactly what the expected rate of glaucoma diagnosis is in our cohort of patients with diabetes with their specific mix of age frequencies and individual days of follow-up without parsing-out each patient assessed. This is precisely why we used a matched control cohort from the same database. Further, even though our cohorts were completely balanced in baseline covariates, the control group was still found to have 10 times the number of new glaucoma diagnosis compared with the GLP-1R agonist group (0.57% vs 0.05%). While some between-group differences in underdiagnosis could have occurred, it is unlikely that a resulting difference of this magnitude would have occurred only due to a difference in underdiagnosis of glaucoma. Because a difference in care seeking between cohorts would also impact our results, we included two variables to assess both healthcare and eye care utilisation. Both showed good balance after IPTW, suggesting that the likelihood of obtaining eye care was similar between cohorts and was therefore unlikely to impact study outcomes. In addition, because existing evidence from randomised clinical trials suggest that GLP-1R agonists may increase the rate of diabetic retinopathy,27 clinicians familiar with these findings would, if anything, increase eye examination frequency in patients treated with GLP-1R agonists as opposed to other anti-glycaemic agents.
Further, lack of specific clinical data including visual acuity prevented us from examining whether GLP-1R agonists reduced the risk of glaucoma progression or improved visual outcomes. Future work using a larger database containing ocular data, including indicators of glaucoma severity, would address these limitations to provide a more complete assessment of the potential for GLP-1R agonists to provide neuroprotection in glaucoma.
Preclinical data showing GLP-1R agonist-mediated RGC rescue, in combination with our findings, support further investigation into the use of GLP-1R agonists for glaucoma prevention and treatment. While our results are exciting, we understand they comprise a single study in addressing the impact of GLP-1R agonists in glaucoma. However, given the favourable side effect profile of GLP-1R agonists, including low incidence of hypoglycaemia,28 and the availability of extensive safety data for this class of medications, our results provide at least a preliminary impetus for clinicians to preferentially consider GLP-1R agonists in treating patients with DM at high risk for glaucoma.
Supplementary Material
Funding
National Institutes of Health-National Eye Institute K23 Award (K23EY025729), National Institutes of Health-National Eye Institute K12 Award (K12EY015398; PI: Joshua L Dunaief), National Institutes of Health-National Eye Institute K08 Award (K08EY029765), National Institutes of Health-National Eye Institute Vision Training Grant (T32EY007035; PI: Diego Contreras), National Institutes of Health-National Eye Institute F30 Award (F30EY032339) and the University of Pennsylvania Core Grant for Vision Research (P30EYEY001583). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. Additional funding was provided by Research to Prevent Blindness (no grant/award number) and the Paul and Evanina Mackall Foundation. Funding from each of the above sources was received in the form of block research grants to the Scheie Eye Institute or University of Pennsylvania Perelman School of Medicine. None of the funding organisations had any role in the design or conduct of the study.
Footnotes
Competing interests Patent pending (Penn Center for Innovation; JS, JLD and QNC).
Ethics approval The University of Pennsylvania Institutional Review Board declared this study exempt because it involves anonymised data with removal of protected health information.
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.
Data availability statement
Data are available upon reasonable request.
REFERENCES
- 1.Tham Y-C, Li X, Wong TY, et al. Global prevalence of glaucoma and projections of glaucoma burden through 2040. Ophthalmology 2014;121:2081–90. [DOI] [PubMed] [Google Scholar]
- 2.Dielemans I, Vingerling JR, Wolfs RC, et al. The prevalence of primary open-angle glaucoma in a population-based study in the Netherlands. The Rotterdam study. Ophthalmology 1994;101:1851–5. [DOI] [PubMed] [Google Scholar]
- 3.Cedrone C, Mancino R, Cerulli A, et al. Epidemiology of primary glaucoma: prevalence, incidence, and blinding effects. Prog Brain Res 2008;173:3–14. [DOI] [PubMed] [Google Scholar]
- 4.Quigley HA. 21st century glaucoma care. Eye 2019;33:254–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Gedde SJ, Feuer WJ, Shi W, et al. Treatment outcomes in the primary tube versus trabeculectomy study after 1 year of follow-up. Ophthalmology 2018;125:650–63. [DOI] [PubMed] [Google Scholar]
- 6.Drucker DJ. Mechanisms of action and therapeutic application of glucagon-like peptide-1. Cell Metab 2018;27:740–56. [DOI] [PubMed] [Google Scholar]
- 7.Yun SP, Kam T-I, Panicker N. Block of A1 astrocyte conversion by microglia is neuroprotective in models of Parkinson’s disease. Nat Med 2018:1–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Bomba M, Ciavardelli D, Silvestri E, et al. Exenatide promotes cognitive enhancement and positive brain metabolic changes in PS1-KI mice but has no effects in 3xTg-AD animals. Cell Death Dis 2013;4:e612. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Park J-S, Kam T-I, Lee S, et al. Blocking microglial activation of reactive astrocytes is neuroprotective in models of Alzheimer’s disease. Acta Neuropathol Commun 2021;9:78. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Athauda D, Maclagan K, Skene SS, et al. Exenatide once weekly versus placebo in Parkinson’s disease: a randomised, double-blind, placebo-controlled trial. Lancet 2017;390:1664–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Femminella GD, Frangou E, Love SB, et al. Evaluating the effects of the novel GLP-1 analogue liraglutide in Alzheimer’s disease: study protocol for a randomised controlled trial (ELAD study). Trials 2019;20:191. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Sterling JK, Adetunji MO, Guttha S, et al. GLP-1 Receptor agonist NLY01 reduces retinal inflammation and neuron death secondary to ocular hypertension. Cell Rep 2020;33:108271. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Williams PA, Marsh-Armstrong N, Howell GR, et al. Neuroinflammation in glaucoma: a new opportunity. Exp Eye Res 2017;157:20–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Guttenplan KA, Stafford BK, El-Danaf RN, et al. Neurotoxic reactive astrocytes drive neuronal death after retinal injury. Cell Rep 2020;31:107776. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Cui QN, Gray IN, Yu Y, et al. Repeated intravitreal injections of antivascular endothelial growth factors and risk of intraocular pressure medication use. Graefes Arch Clin Exp Ophthalmol 2019;257:1931–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Young BA, Lin E, Von Korff M, et al. Diabetes complications severity index and risk of mortality, hospitalization, and healthcare utilization. Am J Manag Care 2008;14:15–23. [PMC free article] [PubMed] [Google Scholar]
- 17.Wu A, Khawaja AP, Pasquale LR, et al. A review of systemic medications that may modulate the risk of glaucoma. Eye 2020;34:12–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Desai RJ, Eddings W, Liao KP, et al. Disease-modifying antirheumatic drug use and the risk of incident hyperlipidemia in patients with early rheumatoid arthritis: a retrospective cohort study. Arthritis Care Res 2015;67:457–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Pasquale LR, Kang JH, Manson JE, et al. Prospective study of type 2 diabetes mellitus and risk of primary open-angle glaucoma in women. Ophthalmology 2006;113:1081–6. [DOI] [PubMed] [Google Scholar]
- 20.Chopra V, Varma R, Francis BA, et al. Type 2 diabetes mellitus and the risk of open-angle glaucoma the Los Angeles Latino eye study. Ophthalmology 2008;115:227–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Jung Y, Han K, Park H-YL, et al. Type 2 diabetes mellitus and risk of open-angle glaucoma development in Koreans: an 11-year nationwide propensity-score-matched study. Diabetes Metab 2018;44:328–32. [DOI] [PubMed] [Google Scholar]
- 22.Shaikh Y, Yu F, Coleman AL. Burden of undetected and untreated glaucoma in the United States. Am J Ophthalmol 2014;158:1121–9. [DOI] [PubMed] [Google Scholar]
- 23.Tielsch JM, Sommer A, Katz J, et al. Racial variations in the prevalence of primary open-angle glaucoma. the Baltimore eye survey. JAMA 1991;266:369–74. [PubMed] [Google Scholar]
- 24.Leske MC, Connell AM, Schachat AP, et al. The Barbados eye study. prevalence of open angle glaucoma. Arch Ophthalmol 1994;112:821–9. [DOI] [PubMed] [Google Scholar]
- 25.Heijl A, Bengtsson B, Oskarsdottir SE. Prevalence and severity of undetected manifest glaucoma: results from the early manifest glaucoma trial screening. Ophthalmology 2013;120:1541–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Malihi M, Moura Filho ER, Hodge DO, et al. Long-Term trends in glaucoma-related blindness in Olmsted County, Minnesota. Ophthalmology 2014;121:134–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Marso SP, Daniels GH, Brown-Frandsen K, et al. Liraglutide and cardiovascular outcomes in type 2 diabetes. N Engl J Med 2016;375:311–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Filippatos TD, Panagiotopoulou TV, Elisaf MS. Adverse effects of GLP-1 receptor agonists. Rev Diabet Stud 2014;11:202–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data are available upon reasonable request.