Abstract
Purpose:
To determine the risk of initiating ocular hypertension and glaucoma treatment with repeated injections of anti-vascular endothelial growth factors (anti-VEGF).
Methods:
A unique, retrospective cohort study was performed using a large national US medical claims database. The study population included patients who had 1 or more injections of an anti-VEGF agent. Exclusion occurred for any previous glaucoma, glaucoma suspect, glaucoma related procedure, an ocular steroid injection, or not seeing an eye care provider at least once in each year of follow up. Cohorts were divided into quartiles based on the number of injections performed over the follow up period. Patients were observed for 2 and 3 years. The main outcome measure was defined as any new prescription for an ocular antihypertensive medication with a concurrent diagnosis of glaucoma, glaucoma suspect or ocular hypertension. Multivariate logistic regression determined the odds of initiating glaucoma treatment in each injection quartile while controlling for numerous covariates. Sensitivity analysis assessed outcomes that included new medication only as well as a new medication plus diagnosis of glaucoma.
Results:
In total, 17113 and 9992 patients met 2- and 3-year observation end points, respectively. The multivariate odds ratio for initiating glaucoma treatment at 2 years was higher in the highest quartile (OR:1.96,95% CI:1.39-2.76,p<0.001) compared to the lowest. The 3-year comparison had similar results with increased odds in the highest quartile (OR:1.51,95%CI:1.07-2.13,p=0.006) compared to the lowest. Sensitivity analyses also showed similar results with more injections being associated with initiating treatment (p<0.053 for all comparisons).
Conclusions:
Repeated anti-VEGF injections are associated with an increased odds of initiating treatment for ocular hypertension and glaucoma.
Keywords: glaucoma, anti-VEGF, intravitreal injections, epidemiology
Introduction:
Over the past decade, intravitreal injections of anti-vascular endothelial growth factor (anti-VEGF) have revolutionized the treatment of numerous retinal disorders [1-3]. Despite the noted visual benefits of anti-VEGF injections, the long-term implications of this treatment including its effects on intraocular pressure (IOP) regulation and risk of glaucomatous optic neuropathy are still being evaluated. Intravitreal anti-VEGF agents are known to induce transient IOP elevations lasting 30 to 60 minutes, which is thought to be directly related to an increase in intraocular fluid volume and are prolonged in the setting of preexisting glaucoma [4-7]. Multiple studies implicate anti-VEGF injection as a cause of persistent IOP elevation in a subset of susceptible individuals [8, 9]. In particular, post-hoc analyses of the ANCHOR, MARINA, and Diabetic Retinopathy Clinical Research (DRCR) data identified higher rates of IOP elevation in ranibizumab-treated versus sham or verteporfin photodynamic therapy (PDT) treated eyes [10-18]. Higher rates of sustained IOP elevation have also been demonstrated following injections of bevacizumab, pegaptanib, and aflibercept [17, 19-25].
Conflicting findings exist in the form of multiple publications which failed to demonstrate anti-VEGF as a risk factor for sustained IOP elevation in subjects with age-related macular generation (AMD), retinal vein occlusion (RVO), and diabetic macular edema [26-29]. More recently, Atchison et al. (2018) queried the IRIS registry for all comers who received unilateral anti-VEGF injections and concluded that anti-VEGF treatment is associated with a statistically significant decrease in IOP compared to the untreated fellow eye [30]. However, in a contradictory secondary finding, the same study also concluded that a higher proportion of treated eyes experienced clinically significant IOP elevation compared to untreated fellow eyes. Similarly, Eadie et al. identified 74 patients who underwent glaucoma surgery and 740 controls within a large health database [31]. The rates of glaucoma surgeries were compared and found to be highest for those who received ≥ 7 bevacizumab injections per year. The study concluded that sustained IOP elevation secondary to anti-VEGF injections lead to a greater need for glaucoma surgical interventions. A study limitation lies in the fact that the cohort contained subjects both with and without a preexisting diagnosis of glaucoma, and therefore did not address whether repeated anti-VEGF injections increase the risk for development of glaucoma. This study will utilize a large claims-based database which will allow for longer follow up times than used previously, and will also use initiation of glaucoma medications as the outcome measure to circumvent the limitation of using inconsistently collected IOP data.
As the scope for intravitreal anti-VEGF therapy continues to expand, any existing associations between intravitreal injections and glaucoma development require clarification. In this study, we used a large insurance billing claim database to examine the associations between intravitreal anti-VEGF injections and risk for development of glaucoma, glaucoma suspect, or ocular hypertension.
Methods:
This is a retrospective cohort study performed using administrative medical claims and prescription claims de-identified data from the Clinformatics™ Data Mart Database (OptumInsight, Eden Prairie, MN) which collates data from a single large, national U.S. insurer. The data used for this project encompassed the years 2006-2016. Due to the de-identified nature of the data, The University of Pennsylvania’s Institutional Review Board deemed this study exempt from review.
All patients who were given anti-VEGF intravitreal injections after January 1st, 2006 were included. The date of incident injection was considered the index date. Patients were excluded for having any prior ocular hypertensive or glaucoma diagnoses or for receiving glaucoma related procedures or medication prescriptions before the index date. (Please see Supplemental Table 1 for full list of ICD9/ICD10 and CPT codes used during this study). Patients were also required to have at least two uninterrupted years in the dataset prior to the index date and to be treated by an eye care provider. To further limit the inclusion of secondary glaucoma, all patients before or after the index date who received ocular steroids or were diagnosed with neovascular glaucoma were excluded. Two analyses were run, a primary analysis which required all patients to be followed for two years after the index date and a sensitivity analysis which observed patients for three years. After the index date, patients were also required to have been evaluated by an eye care provider at least once during each of the observation years.
Multivariate logistic regression was used to determine the odds of initiating treatment for glaucoma via the use of an ocular antihypertensive medication. For the primary analysis this was defined as a patient having both filled a new prescription for an IOP lowering medication and having a new diagnosis of open angle glaucoma (OAG), glaucoma suspect, or ocular hypertension. To analyze the cohorts, patients were divided into quartiles based on the number of injections each patient received during the observation period. In an effort to limit the possibility of confounding by indication, the patients given the lowest quartile of injections were the reference group. This issue was further addressed by controlling for diseases associated with the injections. Other covariates of interest evaluated in the model included age, gender, race, education level, financial net worth, history of diabetes, history of hypertension, pseudophakia, region of the country and year of cohort entry.
In addition to the primary analysis, two sensitivity analyses were run that varied the outcome definition. In the first, to further concentrate on only those truly diagnosed with glaucoma, the same initiation of a glaucoma medication was mandated but was only counted if a diagnosis of open angle glaucoma was also given. The second analysis used a more broad definition, where an outcome was considered if any IOP lowering medication was started without requiring a concurrent diagnosis. Results of the analyses were considered statistically significant for p<0.05 (two-tailed). Statistical analysis was performed using SAS (version 9.4; SAS Institute Inc., Cary, NC).
Results:
Analyses of 2-year Cohort
A total of 17113 patients met the inclusion criteria for the primary analysis (Figure 1). Two years after the index date, 278 (1.6%) initiated an IOP lowering medication and 16835 remained untreated. Treated patients were more likely to be female and were on average 74 years old, predominately white, and treated for age-related macular degeneration. No statistically significant differences were seen between those who were treated for high IOP and those who were not with respect to age, gender, race/ethnicity, yearly income, education level, history of diabetes, or history of hypertension (p>0.05 for all comparisons; see Table 2 for all baseline characteristics.) Initiators of ocular antihypertensive medications were found to have lower rates of pseudophakia (20% vs. 28%; p=0.01), and were more likely to receive injections for retinal vein occlusions (RVO; 12% vs. 10%; p=0.03) or a combination of diseases (12% vs. 8%; p=0.03). Geographic regional divisions were also significantly different between groups, most noticeably with a higher percentage of treated patients from the Mountain region (19% vs. 13% of the unaffected patients) compared to a higher percentage of unaffected patients from the Upper Midwest region (31% vs. 22% of the treated patients; p=0.002). The 2-year study population was divided into quartiles based on the number of injections and contained 4913(1-3 injections), 3747(4-7), 4262(8-13), and 3913(14+) patients in each quartile with 65, 58, 63, and 92 patients who initiated ocular antihypertensive medications, respectively.
Figure 1:
Flowchart showing inclusion and exclusion criteria for study patients.
Table 2:
Comparisons between baseline characteristics of ocular hypertensive treated and untreated groups at 2 and 3 years.
| Patients, No. (%) |
||||||
|---|---|---|---|---|---|---|
| 2 Years (n = 17113) |
3 Years (n = 9992) |
|||||
| Characteristic | Untreated (n = 16835) |
Treated OHT/POAG Tx (n = 278) |
p- value |
Untreated (n = 9705) |
Treated OHT/POAG Tx (n= 287) |
p- value |
| Age, mean (SD) | 73.7 (11.4) | 74.1 (10.0) | 0.52 | 74.1 (10.9) | 74.0 (9.8) | 0.85 |
| Female | 9850 (59%) | 154 (55%) | 0.56 | 5837 (60%) | 156 (54%) | 0.14 |
| Race/ethnicity | 0.51 | 0.22 | ||||
| White | 13354 (79%) | 211 (78%) | 7880 (81%) | 219 (76%) | ||
| Black | 1118 (7%) | 19 (7%) | 591 (6%) | 21 (7%) | ||
| Asian | 336 (2%) | 8 (3%) | 179 (2%) | 9 (3%) | ||
| Hispanic | 901 (5%) | 20 (7%) | 431 (4%) | 17 (6%) | ||
| Unknown | 1126 (7%) | 20 (7%) | 624 (6%) | 21 (7%) | ||
| Education level | 0.91 | 0.61 | ||||
| < 12th Grade | 63 (< 1%) | 1 (< 1%) | 34 (< 1%) | 0 (0%) | ||
| High School Diploma | 4443 (26%) | 67 (24%) | 2471 (25%) | 79 (28%) | ||
| < Bachelor Degree | 9398 (56%) | 157 (56%) | 5466 (56%) | 153 (53%) | ||
| Bachelor Degree + | 2342 (14%) | 42 (15%) | 1449 (15%) | 44 (15%) | ||
| Unknown | 589 (3%) | 11 (4%) | 285 (3%) | 11 (4%) | ||
| Yearly Income | 0.35 | 0.67 | ||||
| < $40K | 6165 (37%) | 97 (35%) | 3637 (37%) | 96 (33%) | ||
| $40K-$74K | 4361 (26%) | 61 (22%) | 945 (10%) | 24 (8%) | ||
| $75K-$99K | 1744 (10%) | 36 (13%) | 723 (7%) | 23 (8%) | ||
| $100K + | 2574 (15%) | 50 (18%) | 862 (9%) | 27 (9%) | ||
| Unknown | 1991 (12%) | 34 (12%) | 1073 (11%) | 32 (11%) | ||
| Diabetes | 6732 (40%) | 104 (37%) | 0.38 | 3598 (37%) | 109 (38%) | 0.75 |
| Hypertension | 13669 (81%) | 218 (78%) | 0.24 | 7793 (80%) | 231 (80%) | 0.94 |
| Pseudophakia | 4660 (28%) | 56 (20%) | 0.01 | 2714 (28%) | 60 (21%) | 0.008 |
| Geographic Region | 0.002 | 0.23 | ||||
| Upper Midwest | 5144 (31%) | 60 (22%) | 3152 (32%) | 72 (25%) | ||
| Southern Midwest | 2300 (14%) | 31 (11%) | 1308 (13%) | 41 (14%) | ||
| Northeast | 1491 (9%) | 29 (10%) | 748 (8%) | 24 (8%) | ||
| Mountain | 2226 (13%) | 54 (19%) | 1284 (13%) | 44 (15%) | ||
| Pacific | 1947 (12%) | 42 (15%) | 1113 (11%) | 41 (14%) | ||
| South Atlantic | 3646 (22%) | 60 (22%) | 2055 (21%) | 64 (22%) | ||
| Unknown | 81 (<1%) | 2 (1%) | 45 (<1%) | 1 (<1%) | ||
| Cohort Year | 0.03 | 0.05 | ||||
| 2006 | 667 (4%) | 14 (5%) | 500 (5%) | 19 (7%) | ||
| 2007 | 1067 (6%) | 14 (5%) | 750 (8%) | 17 (6%) | ||
| 2008 | 1147 (7%) | 11 (4%) | 820 (8%) | 11 (4%) | ||
| 2009 | 1426 (8%) | 30 (11%) | 1057 (11%) | 40 (14%) | ||
| 2010 | 1732 (10%) | 28 (10%) | 1260 (13%) | 31 (11%) | ||
| 2011 | 2633 (16%) | 52 (19%) | 1630 (17%) | 56 (20%) | ||
| 2012 | 2503 (15%) | 34 (12%) | 1800 (19%) | 55 (19%) | ||
| 2013 | 2674 (16%) | 58 (21%) | 1888 (19%) | 58 (20%) | ||
| 2014 | 2986 (18%) | 37 (13%) | ||||
| Disease Category | 0.03 | 0.09 | ||||
| AMD/CNVM | 10703 (64%) | 162 (58%) | 6555 (68%) | 175 (61%) | ||
| PDR/DME | 2049 (12%) | 29 (10%) | 968 (10%) | 35 (12%) | ||
| RVO/CME | 1672 (10%) | 34 (12%) | 864 (9%) | 28 (10%) | ||
| Combined | 1285 (8%) | 34 (12%) | 758 (8%) | 33 (11%) | ||
| Other | 1126 (7%) | 19 (7%) | 560 (6%) | 16 (6%) | ||
| # of Injections (2yr ∣ 3yr) | 0.001 | 0.09 | ||||
| 1-3 ∣ 1-4 | 4913 (29%) | 65 (23%) | 2811 (29%) | 73 (25%) | ||
| 4-7 ∣ 5-10 | 3747 (22%) | 58 (21%) | 2225 (23%) | 55 (19%) | ||
| 8-13 ∣ 11-19 | 4262 (25%) | 63 (23%) | 2445 (25%) | 79 (28%) | ||
| 14+ ∣ 20+ | 3913 (23%) | 92 (33%) | 2224 (23%) | 80 (28%) | ||
Abbreviations: OHT= Ocular hypertension; POAG= Primary open angle glaucoma; AMD= Age-related macular degeneration; CNVM=choroidal neovascular membrane; PDR=proliferative diabetic retinopathy; DME= diabetic macular edema; RVO= retinal vein occlusion; CME= cystoid macular edema
Analyses of 3-year Cohort
For the 3-year cohort, an additional 7121 patients were excluded due to either lack of time in the dataset or not visiting an eye care provider in the third year, leaving a total of 9992 patients (Table 2). In this cohort, 287 (2.9%) initiated an ocular antihypertensive medication while 9705 remained untreated. Similar demographic distributions were present at baseline in the 3-year cohort as was in the 2-year. However, given the smaller population size, fewer variables met statistical significance between the unaffected and incident glaucoma and ocular hypertension groups. The only significant difference in baseline characteristics was pseudophakia with the incident glaucoma and ocular hypertension patients less likely to be pseudophakic (21% vs. 28%, p=0.008). The 3-year study population was also divided into quartiles based on the number of injections and contained 2811(1-4 injections), 2225(5-10), 2445(11-19), and 2224(20+) patients in each quartile with 73, 55, 79, and 80 new initiations of ocular antihypertensive medications, respectively.
Multivariate logistic regression demonstrated that the highest quartile of anti-VEGF injections significantly corresponded to higher odds of starting an IOP lowering medication at 2 and 3 years (14+ injections at 2 years OR:1.96, 95% CI:1.39-2.76, p<0.001; 20+ injections at 3 years OR:1.51, 95% CI: 1.07-2.13, p=0.006). Results additionally suggested that pseudophakia was protective for initiating an IOP lowering medication, with ORs of 0.68 (95% CI: 0.50-0.92) at 2 years, and 0.70 (95% CI: 0.52-0.94) at 3 years. Compared to a diagnosis of AMD/choroidal neovascular membrane (CNVM), diagnoses of RVO/cystoid macular edema (CME) and combined pathologies conferred higher odds of ocular hypertensive treatment at 2 years, with ORs of 1.47 (1.00-2.18) and 1.95 (1.33-2.85), respectively. The combined group alone remained significant at 3 years (OR:1.75, 95% CI: 1.19-2.58, p=0.047). Geographic regional divisions also impacted the odds of treatment with the Pacific (OR:1.75, 95%CI:1.16-2.64), Northeast (OR:1.57, 95%CI:1.00-2.47), and Mountain (OR:1.99, 95%CI:1.36-2.92) regions all having higher odds of medication initiation compared to the Upper Midwest (p=0.02), but these differences did not remain significant at 3 years (p=0.34). Other factors such as age, gender, education, income, race, history of diabetes, and history of hypertension did not significantly correspond to risk of initiating a ocular antihypertensive medication at 2 or 3 years.
Secondary Outcomes/Sensitivity Analyses
Two sensitivity analyses were also run that tested different variations of the primary outcome. The first used a more stringent definition and only looked at those patients treated and diagnosed specifically with open angle glaucoma. In this restricted group there were fewer outcomes at both 2 and 3 years (180 vs. 278 at 2 years; 204 vs. 287 at 3 years). Despite the reduced outcomes, after controlling for all covariates, the odds were increased at both the 2 year (Or:1.90, 95%CI: 1.21-2.97, p=0.006) and 3 year (OR:1.45, 95%CI:0.96-2.20, p=0.053) time points in the highest quartile compared to the lowest, although the 3 year OR just missed significance. The second sensitivity outcome was much broader and included anyone that started an IOP lowering medication regardless of concurrent diagnoses. In this analysis there were 455 and 415 outcomes at 2 and 3 years, respectively, in the same cohort size as the primary analysis. Here again the multivariate results showed that the highest quartile had a significantly higher odds of initiation IOP lowering treatment compared to the lowest quartile at both time points (2 years OR:1.89, 95%CI: 1.45-2.48, p<0.001; 3 years OR:1.33, 95%CI:1.00-1.77, p=0.02).
Discussion:
Utilizing a national insurance database, we examined the association between repeated anti-VEGF injections and the initiation of an ocular antihypertensive medication at 2 and 3 years. In a cohort of over 17000 patients receiving injections at 2 years and nearly 10000 at 3 years, our results offered additional evidence that the number of anti-VEGF injections is a risk factor for initiating glaucoma treatment. Previous studies have found that risk factors for sustained IOP elevation following intravitreal anti-VEGF injections included total number of injections, shorter intervals between injections, and a previous history of glaucoma [15, 21, 22, 32-35]. Proposed pathophysiologic mechanisms underlying the development of elevated IOP have included mechanical obstruction by microparticles and/or high molecular weight protein aggregates that impede outflow or induce inflammation leading to scarring of the trabecular meshwork (TM) [36-38]. Others postulated theories include IOP changes as a result of direct pharmacologic effects of the drugs on the TM, cumulative damage to the TM with recurrent transient IOP spikes, and/or mechanical trauma from repeated injections [16, 36, 39].
The largest and most significant study to date on the subject utilized the IRIS Registry and found a significant average decrease in IOP with injections [30]. In a seemingly contradictory secondary finding, injected eyes were also found to have nearly twice the rate of clinically significant IOP raise. On first pass, these results may seem difficult to reconcile, but considering the percentage of patients with IOP spikes is low, it is not surprising that when averaged over 100 patients, the 3 that had a spike of at least 6 mmHg would not influence average IOP over such a broad population measure. A similar finding was also reported in the DRCR network analysis of their protocol which similarly showed an average decrease in IOP, and concurrently, a 2.9 fold increased risk for sustained IOP elevation [18]. Several other notable differences between the IRIS Registry study and ours are worth mentioning. First, the average follow-up duration in the IRIS study was less than two years, where as ours required a minimum of two years with a good proportion of our patients also contributing to the analysis at 3 years. Additionally, the IRIS study did not account for the use of IOP lowering medications, which may have masked IOP spikes and even impacted the overall IOP assessment in that study.
While the low rates of incident glaucoma in our study provide reassurance that not every injection is inching all patients towards eventual glaucoma, there does seem to be a susceptible population for whom this procedure carries a risk for IOP elevation and increases the risk for glaucoma. While previous studies have found that those diagnosed with glaucoma were at risk for IOP elevation or glaucoma progression following injections, it is difficult to generalize these results to non-glaucomatous populations [18, 21, 30, 31, 36]. To address this issue, our study removed all individuals who have ever received diagnoses of glaucoma, glaucoma suspect, or ocular hypertension, as well as those who have taken glaucoma medication, received a steroid injection, or underwent a glaucoma related procedure. Despite the removal of these high-risk patients, we still found an elevated risk in patients receiving anti-VEGF injections.
One critique of our study might be that we were unable to evaluate IOP directly, and our overall rate of progression to glaucoma (1.6% at 2 years and 2.9% at 3 years) was lower than the 3-11% previously reported rates of sustained IOP elevation [5, 11-21]. This lower rate is not surprising when the difference in outcome definitions are taken into account. In theory, the outcome defined by our study required not only a sustained elevation in IOP, but that changes to aspects of a patient’s ocular condition are sufficient enough to necessitate treatment initiation. We found two other studies that qualified a new IOP lowering medication as an outcome. While one did not separate patients who had high IOP from those taking a new medication, the other started 2 out of 140 eyes (1.4%) on IOP medications but was not clear about the timeline in which the medications were started making comparisons with our study difficult [17, 18].
Vastly different numbers of injections have been postulated as the “tipping point” for seeing an IOP related ocular change [17, 22, 40]. Our study only found significantly elevated risks in the 14+ (2-year) and 20+ (3-year) injections subgroups. Other injection subgroups also trended toward elevated risks, but given the infrequency of outcomes, most likely did not have enough statistical power to reach significance. Interestingly, Martinez-de-la-Casa et al. (2012) found a significant decrease in RNFL thickness after an average of 4.8 ranibizumab injections over the course of 12 months, suggesting physiologic changes in those susceptible may occur early during anti-VEGF use, even if other measures of glaucoma have not yet detected a difference [41].
Other limitations to our study are worth noting when evaluating our results. First, this study is based on administrative claims billing data, and we therefore did not have direct access to typical chart level data like IOP measurements or visual field testing which are used to determine onset of glaucoma. Instead we defined outcomes within the framework of the available data. Next, the decision to add IOP lowering medications was made by individual physicians based on their own judgment. It is therefore possible that as publications continue to surface linking injections with IOP elevation, physicians became more apt to diagnose and treat IOP elevations. To limit this, we controlled for year of treatment initiation. Given no increase in incident glaucoma was seen over time in our study, this is unlikely to have happened in our dataset. Next, due to the nature of administrative data, we chose to associate patients with injections and not a specific eye. Because of this, a possibility exists that while one eye was injected, the fellow eye was the one started on ocular hypertensive treatment. However, for this to have affected the results of the study, this unusual situation would have needed to occur more frequently in the highest tier of injections compared to the lowest tier, which also seems unlikely. Since a significant portion of our patients likely had bilateral injections over the course of two years, by not linking injections to a specific eye offers the possibility that the number of injections per eye associated with progression to glaucoma may in fact be lower than our results suggest. Additionally, it is possible that in certain instances glaucoma medications were only used for short-term treatment of IOP spikes and were not used in the long term. Nevertheless, our analyses were significant at 2 and 3 years which are beyond the time period of a transient IOP spike. Finally, our primary analysis included
Numerous studies have looked at small groups of disease and anti-VEGF medication specific patients. We used a large medical claims database to find that high numbers of injections are associated with an increased risk of initiating treatment for a new diagnosis of ocular hypertension, glaucoma suspect, or glaucoma, despite excluding those predisposed to developing these conditions. This association held despite using a more stringent definition of a new diagnosis of open angle glaucoma as the outcome in a secondary analysis. This study adds to the growing body of literature supporting an association between anti-VEGF injections and glaucoma. Despite the risk, anti-VEGF injections are known to provide substantial visual benefits to multiple groups of patients. Future work is needed to not only identify those patients with the highest risk and understand the underlying mechanism, but to also target possible early intervention before glaucomatous damage can affect the vision of these patients.
Supplementary Material
Table 3:
Adjusted odds ratio of progressing to requiring ocular hypertensive medications at 2 and 3 years.
| Patients, No. (%) |
||||
|---|---|---|---|---|
| 2 Years |
3 Years |
|||
| Characteristic | OR (95% CI) of Tx for OHT/POAG (n = 17113) |
p-value | OR (95% CI) of Tx for OHT/POAG (n = 9992) |
p-value |
| Age | 1.01 (1.00-1.03) | 0.08 | 1.01 (1.00-1.03) | 0.14 |
| Female | 0.89 (0.69-1.13) | 0.34 | 0.82 (0.64-1.05) | 0.28 |
| Race/ethnicity (white ref) | 0.55 | 0.46 | ||
| Black | 1.25 (0.77-2.05) | 1.30 (0.81 - 2.08) | ||
| Asian | 1.37 (0.66-2.83) | 1.56 (0.78 - 3.14) | ||
| Hispanic | 1.41 (0.87-2.29) | 1.36 (0.81 - 2.28) | ||
| Unknown | 1.06 (0.57-1.99) | 1.09 (0.59 - 2.00) | ||
| Education (HS diploma ref) | 0.99 | 0.63 | ||
| < 12th Grade | 1.01 (0.13-7.59) | ND | ||
| < Bachelor Degree | 0.95 (0.70-1.30) | 0.80 (0.60 - 1.08) | ||
| Bachelor Degree + | 0.89 (0.57-1.39) | 0.77 (0.50 - 1.19) | ||
| Unknown | 1.06 (0.42-2.70) | 1.10 (0.44 - 2.79) | ||
| Yearly Income (< $40K ref) | 0.34 | 0.67 | ||
| $40K-$49K | 0.69 (0.41-1.17) | 0.98 (0.62 - 1.54) | ||
| $50K-$59K | 1.05 (0.64-1.69) | 1.20 (0.75 - 1.92) | ||
| $60K-$74K | 1.01 (0.63-1.63) | 1.20 (0.77 - 1.89) | ||
| $75K-$99K | 1.37 (0.91-2.07) | 1.41 (0.93 - 2.15) | ||
| $100K + | 1.35 (0.89-2.04) | 1.36 (0.89 - 2.08) | ||
| Unknown | 1.04 (0.66-1.66) | 1.05 (0.65 - 1.69) | ||
| Geographic Region (Upper Midwest ref) | 0.01 | 0.34 | ||
| Southern Midwest | 1.13 (0.73-1.76) | 1.31 (0.88 - 1.95) | ||
| Northeast | 1.57 (1.00-2.47) | 1.31 (0.82 - 2.11) | ||
| Mountain | 1.99 (1.36-2.92) | 1.45 (0.98 - 2.14) | ||
| Pacific | 1.75 (1.16-2.64) | 1.61 (1.08 - 2.41) | ||
| South Atlantic | 1.27 (0.88-1.85) | 1.24 (0.87 - 1.77) | ||
| Unknown | 1.95 (0.46-8.18) | 0.85 (0.11 - 6.31) | ||
| Cohort Year (2006 ref) | 0.02 | 0.06 | ||
| 2007 | 0.58 (0.27-1.23) | 0.56 (0.29 - 1.09) | ||
| 2008 | 0.41 (0.19-0.92) | 0.34 (0.16 - 0.71) | ||
| 2009 | 0.89 (0.47-1.70) | 0.92 (0.53 - 1.62) | ||
| 2010 | 0.66 (0.34-1.27) | 0.57 (0.32 - 1.03) | ||
| 2011 | 0.71 (0.39-1.32) | 0.75 (0.44 - 1.30) | ||
| 2012 | 0.51 (0.27-0.96) | 0.67 (0.39 - 1.16) | ||
| 2013 | 0.80 (0.43-1.46) | 0.67 (0.39 - 1.16) | ||
| 2014 | 0.44 (0.23-0.84) | |||
| Diabetes | 0.94 (0.71-1.26) | 0.69 | 0.93 (0.69 - 1.23) | 0.59 |
| Hypertension | 0.89 (0.65-1.22) | 0.48 | 1.04 (0.76 - 1.43) | 0.80 |
| Pseudophakia | 0.68 (0.50-0.92) | 0.01 | 0.70 (0.52 - 0.94) | 0.02 |
| Disease Cat. (AMD ref) | 0.008 | 0.047 | ||
| PDR/DME | 1.17 (0.73-1.89) | 1.52 (0.96 - 2.40) | ||
| RVO/CME | 1.47 (1.00-2.18) | 1.29 (0.85 - 1.96) | ||
| Combined | 1.95 (1.33-2.85) | 1.75 (1.19 - 2.58) | ||
| Other | 1.40 (0.84-2.31) | 1.19 (0.69 - 2.05) | ||
| No. of Injections (1-3 2yr ref ∣ 1-4 3yr ref) | <0.001 | 0.006 | ||
| 4-7 ∣ 5-10 | 1.23 (0.86-1.77) | 0.98 (0.68 - 1.41) | ||
| 8-13∣ 11-19 | 1.20 (0.84-1.73) | 1.35 (0.96 - 1.89) | ||
| 14+ ∣ 20+ | 1.96 (1.39-2.76) | 1.51 (1.07 - 2.13) | ||
Abbreviations: OR= odds ratio; OHT= Ocular hypertension; POAG= Primary open angle glaucoma; AMD= Age-related macular degeneration; CNVM=choroidal neovascular membrane; PDR=proliferative diabetic retinopathy; DME= diabetic macular edema; RVO= retinal vein occlusion; CME= cystoid macular edema
Acknowledgments
Funding/Support: National Institutes of Health K23 Award (1K23EY025729 - 01), National Institute of Healthy-National Eye Institute K12 Award (2K12EY015398-11A1; PI: Maureen Maguire), and University of Pennsylvania Core Grant for Vision Research (2P30EYEY001583). 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 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. None of the organizations had any role in the design or conduction of the study.
Footnotes
Publisher's Disclaimer: This Author Accepted Manuscript is a PDF file of a an unedited peer-reviewed manuscript that has been accepted for publication but has not been copyedited or corrected. The official version of record that is published in the journal is kept up to date and so may therefore differ from this version.
Conflicts of interest: Qi Cui declares that she has no conflicts of interest to report; Iga Gray declares that she has no conflicts of interest to report; Yinxi Yu declares that she has no conflicts of interest to report; Brian VanderBeek declares that he has no conflicts of interest to report
Ethical approval: All analyses performed in this study involving data from humans were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. This study was deemed exempt from IRB review by the University of Pennsylvania Institutional Review Board due to the de-identified nature of the data used within the study. The need for informed consent was also waived by the University of Pennsylvania IRB again due to the de-identified nature of the data used.
References:
- 1.Rosenfeld PJ, Brown DM, Heier JS, et al. (2006) Ranibizumab for neovascular age-related macular degeneration. N Engl J Med 355(14):1419–31. [DOI] [PubMed] [Google Scholar]
- 2.Brown DM, Kaiser PK, Michels M, et al. (2006) Ranibizumab versus verteporfin for neovascular age-related macular degeneration. N Engl J Med 355(14):1432–44. [DOI] [PubMed] [Google Scholar]
- 3.Group CR, Martin DF, Maguire MG, et al. (2011) Ranibizumab and bevacizumab for neovascular age-related macular degeneration. N Engl J Med 364(20):1897–908. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Kim JE, Mantravadi AV, Hur EY, Covert DJ. (2008) Short-term intraocular pressure changes immediately after intravitreal injections of anti-vascular endothelial growth factor agents. Am J Ophthalmol 146(6):930–4 e1. [DOI] [PubMed] [Google Scholar]
- 5.Bakri SJ, Pulido JS, McCannel CA, et al. (2009) Immediate intraocular pressure changes following intravitreal injections of triamcinolone, pegaptanib, and bevacizumab. Eye (Lond) 23(1):181–5. [DOI] [PubMed] [Google Scholar]
- 6.El Chehab H, Agard E, Russo A, et al. (2016) Intraocular Pressure Spikes after Aflibercept Intravitreal Injections. Ophthalmologica ;236(1):43–7. [DOI] [PubMed] [Google Scholar]
- 7.El Chehab H, Le Corre A, Agard E, et al. (2013)Effect of topical pressure-lowering medication on prevention of intraocular pressure spikes after intravitreal injection. Eur J Ophthalmol ;23(3):277–83. [DOI] [PubMed] [Google Scholar]
- 8.Agard E, Elchehab H, Ract-Madoux G, et al. (2015) Repeated intravitreal anti-vascular endothelial growth factor injections can induce iatrogenic ocular hypertension, especially in patients with open-angle glaucoma. Can J Ophthalmol ;50(2):127–31. [DOI] [PubMed] [Google Scholar]
- 9.Leleu I, Penaud B, Blumen-Ohana E, et al. (2018) [Late and sustained intraocular pressure elevation related to intravitreal anti-VEGF injections: Cases requiring filtering surgery (French translation of the article)]. J Fr Ophtalmol ;41(9):789–801. [DOI] [PubMed] [Google Scholar]
- 10.Jalil A, Fenerty C, Charles S (2007) Intravitreal bevacizumab (Avastin) causing acute glaucoma: an unreported complication. Eye (Lond) 21(12):1541. [DOI] [PubMed] [Google Scholar]
- 11.Bakri SJ, McCannel CA, Edwards AO, Moshfeghi DM (2008) Persisent ocular hypertension following intravitreal ranibizumab. Graefes Arch Clin Exp Ophthalmol 246(7):955–8. [DOI] [PubMed] [Google Scholar]
- 12.Zhou Y, Zhou M, Xia S, et al. (2016) Sustained Elevation of Intraocular Pressure Associated with Intravitreal Administration of Anti-vascular Endothelial Growth Factor: A Systematic Review and Meta-Analysis. Sci Rep 6:39301. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Kahook MY, Kimura AE, Wong LJ, et al. (2009) Sustained elevation in intraocular pressure associated with intravitreal bevacizumab injections. Ophthalmic Surg Lasers Imaging 40(3):293–5. [DOI] [PubMed] [Google Scholar]
- 14.Adelman RA, Zheng Q, Mayer HR (2010) Persistent ocular hypertension following intravitreal bevacizumab and ranibizumab injections. J Ocul Pharmacol Ther 26(1): 105–10. [DOI] [PubMed] [Google Scholar]
- 15.Hoang QV, Tsuang AJ, Gelman R, et al. (2013) Clinical predictors of sustained intraocular pressure elevation due to intravitreal anti-vascular endothelial growth factor therapy. Retina 33(1):179–87. [DOI] [PubMed] [Google Scholar]
- 16.Bakri SJ, Moshfeghi DM, Francom S, et al. (2014) Intraocular pressure in eyes receiving monthly ranibizumab in 2 pivotal age-related macular degeneration clinical trials. Ophthalmology 121(5):1102–8. [DOI] [PubMed] [Google Scholar]
- 17.Vo Kim S, Fajnkuchen F, Sarda V, et al. (2017) Sustained intraocular pressure elevation in eyes treated with intravitreal injections of anti-vascular endothelial growth factor for diabetic macular edema in a real-life setting. Graefes Arch Clin Exp Ophthalmol 255(11):2165–71. [DOI] [PubMed] [Google Scholar]
- 18.Bressler SB, Almukhtar T, Bhorade A, et al. (2015) Repeated intravitreous ranibizumab injections for diabetic macular edema and the risk of sustained elevation of intraocular pressure or the need for ocular hypotensive treatment. JAMA Ophthalmol 133(5):589–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Choi DY, Ortube MC, McCannel CA, et al. (2011) Sustained elevated intraocular pressures after intravitreal injection of bevacizumab, ranibizumab, and pegaptanib. Retina 31(6):1028–35. [DOI] [PubMed] [Google Scholar]
- 20.Baek SU, Park IW, Suh W. (2016) Long-term intraocular pressure changes after intravitreal injection of bevacizumab. Cutan Ocul Toxicol 35(4):310–4. [DOI] [PubMed] [Google Scholar]
- 21.Good TJ, Kimura AE, Mandava N, Kahook MY. (2011) Sustained elevation of intraocular pressure after intravitreal injections of anti-VEGF agents. Br J Ophthalmol 95(8):1111–4. [DOI] [PubMed] [Google Scholar]
- 22.Hoang QV, Mendonca LS, Della Torre KE, et al. (2012) Effect on intraocular pressure in patients receiving unilateral intravitreal anti-vascular endothelial growth factor injections. Ophthalmology 119(2):321–6. [DOI] [PubMed] [Google Scholar]
- 23.Mathalone N, Arodi-Golan A, Sar S, et al. (2012) Sustained elevation of intraocular pressure after intravitreal injections of bevacizumab in eyes with neovascular age-related macular degeneration. Graefes Arch Clin Exp Ophthalmol 250(10):1435–40. [DOI] [PubMed] [Google Scholar]
- 24.Al-Abdullah AA, Nowilaty SR, Asghar N, et al. (2015) Intraocular pressure trends after intravitreal injections of anti-vascular endothelial growth factor agents for diabetic macular edema. Retina 35(3):440–8. [DOI] [PubMed] [Google Scholar]
- 25.Freund KB, Hoang QV, Saroj N, Thompson D. (2015) Intraocular Pressure in Patients with Neovascular Age-Related Macular Degeneration Receiving Intravitreal Aflibercept or Ranibizumab. Ophthalmology 122(9):1802–10. [DOI] [PubMed] [Google Scholar]
- 26.Kim YJ, Sung KR, Lee KS, et al. (2014) Long-term effects of multiple intravitreal antivascular endothelial growth factor injections on intraocular pressure. Am J Ophthalmol 157(6):1266–71 e1. [DOI] [PubMed] [Google Scholar]
- 27.Nariani A, Williams B, Hariprasad SM. (2016) Long-term effect of anti-vascular endothelial growth factor injections on intraocular pressure. Indian J Ophthalmol 64(9):643–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Wehrli SJ, Tawse K, Levin MH, et al. (2012) A lack of delayed intraocular pressure elevation in patients treated with intravitreal injection of bevacizumab and ranibizumab. Retina 32(7):1295–301. [DOI] [PubMed] [Google Scholar]
- 29.Boyer DS, Goldbaum M, Leys AM, et al. (2014) Effect of pegaptanib sodium 0.3 mg intravitreal injections (Macugen) in intraocular pressure: posthoc analysis from V.I.S.I.O.N. study. Br J Ophthalmol 98(11):1543–6. [DOI] [PubMed] [Google Scholar]
- 30.Atchison EA, Wood KM, Mattox CG, et al. (2018) The Real-World Effect of Intravitreous Anti-Vascular Endothelial Growth Factor Drugs on Intraocular Pressure: An Analysis Using the IRIS Registry. [DOI] [PubMed] [Google Scholar]
- 31.Eadie BD, Etminan M, Carleton BC, et al. (2017) Association of Repeated Intravitreous Bevacizumab Injections With Risk for Glaucoma Surgery. JAMA Ophthalmol 135(4):363–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Yannuzzi NA, Patel SN, Bhavsar KV, et al. (2014) Predictors of sustained intraocular pressure elevation in eyes receiving intravitreal anti-vascular endothelial growth factor therapy. Am J Ophthalmol 158(2):319–27 e2. [DOI] [PubMed] [Google Scholar]
- 33.Beato J, Pedrosa AC, Pinheiro-Costa J, et al. (2016) Long-Term Effect of Anti-VEGF Agents on Intraocular Pressure in Age-Related Macular Degeneration. Ophthalmic Res 56(1):30–4. [DOI] [PubMed] [Google Scholar]
- 34.Pershing S, Bakri SJ, Moshfeghi DM. (2013) Ocular hypertension and intraocular pressure asymmetry after intravitreal injection of anti-vascular endothelial growth factor agents. Ophthalmic Surg Lasers Imaging Retina 44(5):460–4. [DOI] [PubMed] [Google Scholar]
- 35.Segal O, Ferencz JR, Cohen P, et al. (2013) Persistent elevation of intraocular pressure following intravitreal injection of bevacizumab. Isr Med Assoc J 15(7):352–5. [PubMed] [Google Scholar]
- 36.Dedania VS, Bakri SJ. (2015) SUSTAINED ELEVATION OF INTRAOCULAR PRESSURE AFTER INTRAVITREAL ANTI-VEGF AGENTS: What Is the Evidence? Retina 35(5):841–58. [DOI] [PubMed] [Google Scholar]
- 37.Kahook MY, Ammar DA. (2010) In vitro effects of antivascular endothelial growth factors on cultured human trabecular meshwork cells. J Glaucoma 19(7):437–41. [DOI] [PubMed] [Google Scholar]
- 38.Kahook MY, Liu L, Ruzycki P, et al. (2010) High-molecular-weight aggregates in repackaged bevacizumab. Retina 30(6):887–92. [DOI] [PubMed] [Google Scholar]
- 39.Apte RS, Gordon M, Kass MA (2017) Anti-VEGF Injections and Glaucoma Surgery. JAMA Ophthalmol 135(4):368–9. [DOI] [PubMed] [Google Scholar]
- 40.Tseng JJ, Vance SK, Della Torre KE, et al. (2012) Sustained increased intraocular pressure related to intravitreal antivascular endothelial growth factor therapy for neovascular age-related macular degeneration. J Glaucoma 21(4):241–7. [DOI] [PubMed] [Google Scholar]
- 41.Martinez-de-la-Casa JM, Ruiz-Calvo A, Saenz-Frances F, et al. (2012) Retinal nerve fiber layer thickness changes in patients with age-related macular degeneration treated with intravitreal ranibizumab. Invest Ophthalmol Vis Sci 53(10):6214–8. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.