Discussion
In this study, we used a large, nationwide insurance claims database to assess rates of new glaucoma and OHT development following periocular and intravitreal steroid injections over a multiyear period in patients with no prior history of glaucoma or OHT. We showed that patients receiving ocular steroid injections are at considerable risk of developing glaucoma/OHT, particularly in the setting of sustained-release intravitreal steroid implants and/or multiple steroid injections over time. Although the association of steroids and OHT/glaucoma is well known, our utilisation of a large-scale, nationwide database with an extended follow-up period allowed us to investigate clinical questions—such as how the total number of steroid injections over time influences the risk of OHT/glaucoma development, and how different types and routes of steroid injections impact OHT/glaucoma risk within a single population—that smaller clinic-based or hospital-based studies have not addressed. These findings help to provide practical insights into the potential risks, clinical course and follow-up needs of patients receiving periocular and intravitreal steroid injections.
Compared with other types of ocular steroid injections in our study, intravitreal steroid injections and sustained-release intravitreal steroid implants were associated with the highest risk of glaucoma/OHT development. Multiple previous studies have demonstrated an association between the intravitreal injection procedure and spikes in IOP, including for non-steroidal medications such as anti-vascular endothelial growth factor (VEGF) agents.13–15 This IOP elevation is thought to be due, at least in part, to an acute volume effect of injecting fluid into a relatively confined space within the eye.13 Another study of anti-VEGF injections suggested that trauma from the injection procedure led to trabecular meshwork damage, resulting in less efficient aqueous outflow and elevated IOP.16 In the context of anti-VEGF injections, however, the prevalence of IOP elevation with associated glaucomatous damage is relatively low (<5%).17 Our study and others have demonstrated a higher rate of documented glaucoma/OHT after even a single steroid injection, suggesting that additional factors may be involved in the development of IOP elevation following steroid administration.4 18
In addition to mechanical changes associated with the intravitreal injection procedure, the pharmacodynamics and concentration of steroids within the eye may also contribute to the substantial rate of glaucoma/OHT that we observed following intravitreal steroid administration. Weijtens et al demonstrated that the intravitreal route of medication delivery enabled a maximum concentration of intraocular corticosteroids compared with subconjunctival or retrobulbar injections.19 20 Additionally, these high intravitreal steroid concentrations have been shown to persist within the human eye for multiple months to years, even after a single intravitreal injection.21 Sustained-release intravitreal steroid implants have also been associated with OHT and glaucoma.22 The prolonged action of steroids within the vitreous is especially notable for sustained-release implants such as Iluvien and Retisert, which deliver fluocinolone acetonide intraocularly for up to 36 months.23 While the exact mechanism of steroid-induced glaucoma is not yet fully elucidated, corticosteroids are hypothesised to interact with various enzymes and cytoskeletal components within the trabecular meshwork to cause increased aqueous outflow resistance and therefore elevated IOP.2 3 Studies have demonstrated that corticosteroids can lead to microstructure alterations within the trabecular meshwork through the upregulation of myocilin and through binding to glucocorticoid receptor beta.24 25 Others have postulated that physical obstruction of the trabecular meshwork by glucocorticoid crystals contributes to reduced aqueous outflow.26 Our data support that the intravitreal route of steroid delivery facilitates more direct and prolonged interaction between corticosteroids and trabecular meshwork, ultimately leading to higher rates of increased aqueous outflow resistance and sustained IOP elevation compared with subconjunctival or retrobulbar routes.
The risk of postinjection glaucoma/OHT was also significantly higher for patients receiving multiple steroid injections over time. Other studies have similarly shown increased rates of IOP elevation and glaucoma following repeated steroid administration, suggestive of a cumulative or dose-dependent effect of steroids on glaucoma development.4 6 Our findings contrast with a smaller retrospective study by Jonas et al, which demonstrated no increased risk of IOP elevation after repeated triamcinolone acetonide intravitreal injections.27 Of note, we analysed the total number of any type of ocular steroid injections (including both intravitreal and periocular routes). Comparing the number of individual injection types was not possible given that many of the patients receiving multiple steroid injections were given different types of injections over time. Interestingly, we found that the largest increase in glaucoma/OHT risk occurred after a total of three steroid injections. In addition to reflecting a potential cumulative effect of repeated steroid injections on glaucoma/OHT development, this finding may also represent a temporal delay that can occur between steroid administration and subsequent IOP elevation. For instance, multiple studies have shown that glaucoma can take weeks and even months to develop after initial steroid exposure.18 28 These results highlight the importance of maintaining long-term vigilance in IOP monitoring even in the absence of initial IOP elevation.
We also found that the rate of glaucoma/OHT development and, in particular, the need for IOP-lowering drops did not significantly increase with four or more steroid injections compared with three injections. A multicentre retrospective study even found that patients receiving four or more injections had a marginally lower risk of IOP elevation compared with patients receiving fewer injections.29 While these findings may seem contrary to a dose-dependent relationship between steroids and glaucoma, they are perhaps also reflective of provider decision-making and the patients’ known risk factors, such as family history of glaucoma, corneal hysteresis, myopia and/or thinner corneas.10 30 31 Clinicians tend to avoid repeated steroid injections in patients with previous known steroid-induced IOP elevation, or consider IOP lowering intervention prior to repeat injections.29 A further prospective investigation of how serial steroid injections modulate the risk of glaucoma/OHT would be useful in order to facilitate informed decision-making with patients who require repeated intraocular or periocular steroid treatments.
Our study underscores the potential for using large-scale healthcare claims data to investigate patterns in steroid-related glaucoma development. However, certain limitations are inherent to our claims-based study. For example, because insurance claims data do not include IOP measurements, detailed ocular examination findings or glaucoma family history, we relied on diagnostic coding and treatment initiation to determine glaucoma/OHT outcomes and comorbidities within our cohort. While prior studies have demonstrated strong agreement between administrative claims data and corresponding patient records in ophthalmology,32 33 the accuracy of this coding is dependent on providers with a wide range of clinical backgrounds who therefore may differ in their approach to the diagnosis and treatment of ocular diseases.34 Additionally, the ICD-9 coding system lacks laterality information, which limited our analyses to a patient—rather than eye—level. Consequently, the development of postinjection glaucoma/OHT in a patient could potentially be confounded by diagnoses or treatments in their contralateral eye, as well as systemic diagnoses or medications. Despite this limitation, the patient-level rates of new OHT/glaucoma in our study may be useful for both clinicians and patients to better understand the overall risks associated with ocular steroid injections, and they align with general trends of steroid-related OHT/glaucoma development that have been identified in prior hospital-based and clinic-based studies.4 6–8 Due to the claims-based nature of the database and lack of ocular examination findings, we were unable to ascertain the degree of uveitis control or the presence of neovascularisation in each patient; this could potentially confound our findings, as neovascularisation of the angle can cause IOP elevation and active ocular inflammation can either raise or lower IOP (although we excluded patients with documented neovascular or uveitic glaucoma and also controlled for conditions such as diabetic retinopathy and uveitis in our multivariate analyses to reduce the risk of confounding).35 The database could also not differentiate between types and doses of triamcinolone (such as Kenalog and Triescence) or between variations in technique of sub-Tenon injections (anterior vs posterior sub-Tenon approach), which could also influence the extent of postinjection IOP elevation.4 7 Finally, our study was based on retrospective analyses of compiled deidentified data and was, therefore, purely observational in nature. Future prospective studies could incorporate longitudinal postinjection IOP measurements and examination findings (including the duration of steroid-related IOP elevation and treatments) in order to address these limitations. Evaluation of large, multicentre databases that contain both clinical and genetic data would be useful to further understand the risk factors for this condition, as recent evidence suggests a likely genetic influence on steroid-induced OHT and glaucoma development.36
In summary, this study demonstrates that patients receiving periocular and/or intravitreal steroid injections are at considerable risk of developing new glaucoma/OHT, with a substantial proportion of patients requiring topical, laser or even surgical intervention to lower IOP. Whereas many prior studies have used smaller, select populations to assess OHT/glaucoma risk, our study assessed a massive population of patients throughout the USA in order to determine the rate of OHT/glaucoma development after ocular steroid injections.4–7 Based on our findings, we suggest that patients receiving ocular steroid injections be closely monitored for glaucoma and OHT development, particularly in the setting of intravitreal and/or repeated steroid administration. There are currently no well-defined, broadly accepted guidelines regarding the specific frequency and duration of glaucoma screening in patients receiving intravitreal or periocular steroid injections. Future work to better characterise and risk-stratify these patients would be useful both to provide practical, individualised patient counselling and to help guide clinicians as they treat and monitor their patients over time.
