Assessment the real-world safety of intravitreal dexamethasone implant (Ozurdex): novel insights from a comprehensive pharmacovigilance analysis utilizing the FAERS database

Abstract

Objective

The intravitreal dexamethasone implant (Dex) is widely used for various ocular conditions, including diabetic macular edema (DME), retinal vein occlusion (RVO), and non-infectious uveitis. Despite its efficacy, concerns remain regarding its safety profile. This study aims to analyze the adverse events (AEs) associated with Dex reported in the FDA Adverse Event Reporting System (FAERS) database from 2010 to 2024.

Methods

Data were extracted from FAERS, focusing on cases where Dex was the primary suspect drug. The dataset was processed to eliminate duplicates and incomplete entries. Disproportionality analysis, including Reporting Odds Ratio (ROR) and Proportional Reporting Ratio (PRR), was used to detect safety signals. AEs were categorized by system organ class (SOC) and preferred term (PT).

Results

A total of 1,588 adverse event reports (AERs) were analyzed, revealing a significant upward trend. The Eye disorders was the most commonly reported SOC, with strong disproportionality signals (ROR: 45.11; PRR: 23.71). Key AEs identified at the PT level included Corneal decompensation, Choroidal hematoma, and Posterior capsule rupture, which were not listed on the drug label. Considering the reported numbers, the Endophthalmitis was the most common AE. Additionally, a significant proportion of AEs were observed within the first seven days post-administration, emphasizing the need for monitoring.

Conclusion

While Dex remains an effective treatment option for ocular conditions, its use is associated with significant risks, particularly regarding unexpected and severe complications such as corneal decompensation. Continuous pharmacovigilance and detailed patient monitoring are essential to mitigate these risks. Future studies should focus on prospective designs and comprehensive clinical data to better understand the safety profile of Dex.

Introduction

The intravitreal dexamethasone implant (Dex, Ozurdex 0.7 mg, Allergan Inc, Irvine, CA, USA) has emerged as an alternative treatment option for various ocular conditions [1], including diabetic macular edema (DME), retinal vein occlusion-related (RVO) and non-infectious uveitis [2,3,4]. This biodegradable implant delivers sustained-release dexamethasone directly to vitreous, providing targeted and prolonged anti-inflammatory effects.

DME, a major cause of vision impairment in diabetic patients, is characterized by the fluid accumulation in the macula due to leakage from damaged retinal blood vessels [5]. The occurrence of RVO, which can lead to retinal edema and hemorrhage, is one of the most common retinal vascular disorders worldwide [6]. Inflammatory factors have been identified to play a critical role in the pathogenesis in DME and RVO, including the up-regulation of extensive pro-inflammatory cytokines, chemokines and adhesion molecules [7,8,9,10,11]. Substantial clinical investigations have demonstrated that Dex could inhibit inflammatory responses and reduce vascular permeability, leading to decreased macular thickness and improving visual outcomes [12,13,14,15]. For patients with non-infectious uveitis, Dex provides persistent control of ocular inflammation, preserving the retinal functions and reducing the recurrence frequency [16].

Despite its therapeutic benefits, the use of Dex may result in several potential clinical adverse effects, such as the intraocular pressure (IOP) elevation and formation of cataract. Persistent corticosteroid-induced ocular hypertension could progress to glaucoma, requiring the use of IOP-lowering medications and surgical intervention. Another common side effect is the increased risk of cataract formation. Corticosteroid is widely acknowledged as an essential cause of posterior subcapsular cataract, which should be resolved by surgery. Additionally, the intravitreal injection leads to the increased risk of endophthalmitis, a severe and vision-threatening disorder.

To comprehensively assess the clinical adverse events (AEs) associated with Dex, this study utilizes data from the Food and Drug Administration Adverse Event Reporting System (FAERS). The FAERS database is a valuable resource for post-market surveillance, containing millions of reports on AEs and medication errors submitted by healthcare professionals, consumers, and manufacturers [17, 18]. By analyzing FAERS data, researchers can identify potential safety concerns, monitor the incidence of AEs, and detect trends or patterns in drug safety. This approach provides insights beyond clinical trial data, encompassing a broader patient population and longer follow-up periods.

Understanding the real-world evidence of AEs through FAERS data is critical for optimizing the use of Dex in clinical practice. Comprehensive patient evaluation and post-implantation monitoring are essential strategies to mitigate the incidence of AEs. Overall, this investigation aims to explore and analyze the AEs associated with the Dex, providing convincing evidence for safety profile and clinical application.

Methods

Data source

The primary data source for this study was the FAERS database. FAERS is a critical pharmacovigilance tool utilized by the FDA to monitor AEs and medication errors reported by healthcare professionals, consumers, and manufacturers. FAERS plays a vital role in assessing the post-marketing safety of pharmaceutical products. This study specifically focused on AEs associated with the intravitreal dexamethasone implant, as recorded in FAERS. FAERS comprises multiple data files, including demographic information (DEMO), drug information (DRUG), adverse event reports (REAC), patient outcomes (OUTC), report sources (RPSR), drug therapy (THER), and indications for drug use (INDI).

Data extraction involved downloading the raw data files directly from the FDA’s official website. Search terms “Dex/Ozurdex” and its generic name “dexamethasone intravitreal implant” were used to capture of all relevant reports comprehensively. Considering the widespread use of Dex implant worldwide, all indications for the Dex were included in this study to ensure the integrity and reliability of the results. The time frame for data collection extended from the drug’s approval date to the most recent update, ensuring the inclusion of all potential AEs. The extracted data included reports from a global pool, capturing a wide range of patient demographics and healthcare settings, thereby providing a comprehensive perspective on the safety profile of Dex.

Data processing

Data processing consisted of several key steps to refine and prepare the dataset for analysis. Initially, the raw data files were imported into RStudio, where initial data underwent a cleaning process according to the FDA recommendations. Duplicate entries, often arising from follow-up reports, were identified using unique case identifiers and a manual review, ensuring that no redundant data were included in the final dataset. This step was crucial to maintain the accuracy and reliability of the analysis. Reports with incomplete or ambiguous information, such as missing patient demographics or unspecified AEs, were excluded from further analysis to ensure the dataset’s integrity and robustness.

The Medical Dictionary for Regulatory Activities (MedDRA) was employed to standardize the terminology of AEs, ensuring consistency and accuracy in their classification, which is essential for reliable analysis. Reports were filtered to include only those where Dex was listed as the primary suspect drug (PS), ensuring that the focus remained on events directly associated with Dex use. Each AE was classified by its system organ class (SOC) and preferred term (PT), facilitating a structured and systematic analysis of the data. Additionally, detailed information, including patient age, gender, reporter sources, serious outcomes et al., was extracted and summarized to provide context to the AE data. The final dataset was formatted appropriately for statistical analysis.

Statistical analysis

Due to the absence of denominator data, it is impossible to determine the incidence of AEs directly. For the statistical analysis, disproportionality methods were employed to detect signals of AEs associated with Dex. Disproportionality analysis is a widely used method in pharmacovigilance for identifying potential safety signals. The primary methods used in this study were the Reporting Odds Ratio (ROR) and Proportional Reporting Ratio (PRR) [19, 20].

The criteria for signal detection were a lower limit of the 95% confidence interval (CI) of ROR greater than 1, a PRR of at least 2, and a chi-squared value (χ²) of at least 4. These thresholds indicate a statistically significant association between Dex and the reported AEs. Additionally, the Bayesian Confidence Propagation Neural Network (BCPNN) and Empirical Bayes Geometric Mean (EBGM) were applied to corroborate the findings and ensure robustness [21, 22]. These methods were all based on the 2 × 2 contingency table, using specific formulas and thresholds (Supplementary Table S1-S2). R software (version 4.4.0) was utilized for comprehensive database analysis.

Results

Descriptive analysis

This study contains 56 quarters of AEs from FAERS database, spanning from the second quarter of 2010 to the first quarter of 2024. As shown in Fig. 1, a total of 45,122,449 drug-AE reports (AERs) were collected after eliminating the duplication records, among which 1,588 AERs and 3,630 AEs at the PT level were finally determined with Dex as the PS. Figure 2 indicated an overall upward trend in the number of AERs, with notable peaks in 2022 (13.92%) and 2023 (13.10%).

Fig. 1

The flow diagram of selecting Dex-related AEs from FAES database

Fig. 2

The number and percentage of adverse event reports in each year from the FAERS database

Detailed demographics of AERs were summarized in Table 1, including the clinical characteristics, indications, incidence of serious outcomes and onset time of events. Regarding gender distribution, male and female patients accounted for 41.37% and 36.27% of the known data, respectively. Among the records that documented the age of patients, the age group over 60 years accounted for approximately 26.83% of cases, with a median age of 67 years. The majority of reports were submitted by physicians (51.51%), followed by consumers (29.35%), pharmacists (10.14%) and other health-professionals (8.31%). In terms of reporting regions, the United States had the highest proportion of AERs (39.23%), followed by Europe (27.96%), Australia (3.40%), and Turkey (2.33%). Based on data with clearly recorded therapeutic indications, the most common indications were DME (15.68%), RVO (15.37%), and uveitis (7.30%). To ensure data integrity, we provided comprehensive documentation of the reporting regions and associated therapeutic indications in Supplementary Table 3. In addition, serious outcomes associated with reports should be emphasized, such as hospitalization (12.76%), disability (5.23%) and death (4.75%). Notably, our study revealed that a significant proportion (30.73%) of AEs occurred within the first seven days, underscoring the critical need for continuous monitoring to ensure patient safety and effective risk management.

Table 1 Clinical characteristics of reports with Dex from the FAERS database (2010 Q2-2024 Q1)

Signal detection

The signal strengths of AEs related to Dex at the system organ class (SOC) level are described in Table 2. The Eye disorders emerged as the most prominently reported SOC, with a total of 1,761 case reports. Disproportionality analysis showed significant signal strengths for this category, with an ROR (95% CI) of 45.11 (42.26, 48.14) and a PRR (95% CI) of 23.71 (22.8, 24.66). Additionally, the IC (IC025) and EBGM (EBGM05) values were 4.56 (4.48) and 23.67 (22.41) respectively, suggesting a strong association between Dex and reported eye disorders​. Among the other SOCs, the Injury, poisoning and procedural complications (561 cases), General disorders and administration site conditions (458 cases) and Infections and infestations (351 cases) were common systems, without significant signal strengths.

Table 2 Signal strength of reports at the system organ class (SOC) level

The PT level analysis provides a granular view of the specific AEs, offering insights beyond the broader of SOC classifications. In this analysis, the top 30 PTs ranked by ROR were listed in Table 3. The results revealed that Corneal decompensation (cases = 35, ROR = 6285.32, PRR = 6211.93, IC = 12.08, EBGM = 4321.65), Choroidal haematoma (cases = 3, ROR = 6091.26, PRR = 6085.16, IC = 12.06, EBGM = 4259.91) and Posterior capsule rupture (cases = 19, ROR = 5776.52, PRR = 5739.9, IC = 12, EBGM = 4087.79) were the top three signal AEs. Based on the reported case numbers, the most common AEs were Endophthalmitis (cases = 226, ROR = 811.95, PRR = 750.8, IC = 9.48, EBGM = 713.14), IOP increased (cases = 199, ROR = 271.02, PRR = 253.09, IC = 7.96, EBGM = 248.68) and Visual acuity reduced (cases = 134, ROR = 79.9, PRR = 76.37, IC = 6.25, EBGM = 75.97), as presented in Supplementary Table 4.

Table 3 The top 30 signal strengths of reports ranked by ROR at the PTs level and sorted by SOC

Interestingly, this analysis determined a number of AEs not mentioned in the drug label. Considering the signal intensity, the top five unexpected AEs were Corneal decompensation (ROR = 6285.32), Choroidal haematoma (ROR = 6091.26), Posterior capsule rupture (ROR = 5776.52), Iris neovascularization (ROR = 1016.23) and Necrotizing retinitis (ROR = 446.87). Additionally, the top five PTs based on the case numbers were Corneal decompensation (cases = 35), Hypopyon (cases = 20), Posterior capsule rupture (cases = 19), Necrotizing retinitis (cases = 15) and Macular hole (cases = 14).

Discussion

In this study, we analyzed AEs related to Dex reported in the FAERS database from 2010 to 2024. A substantial upward trend in AE reports was observed, indicating increasing pharmacovigilance data. The Eye disorders was identified as the most commonly reported category, with significant disproportionality signals confirming a strong association between Dex and these ocular disorders. At the PT level, rare but noteworthy events such as corneal decompensation, choroidal hematoma, and posterior capsule rupture, which are not listed on the drug label, showed high signal intensities.

Two recent large-scale retrospective studies reported AE incidence rates of 1% and 1.69% following Dex implant administration, highlighting the importance of continuous monitoring in clinical practice [23, 24]. Epidemiological data indicate that the rapidly increasing prevalence and incidence of DME and RVO in the United States and Europe warrant significant attention, particularly in elderly populations [25,26,27]. A university-based study has confirmed that among patients experiencing AEs following Dex implant, a higher proportion are male and elderly, consistent with our findings [23]. Among the reported regions, the United States accounted for the highest number of cases, which could be attributed to several potential factors. Firstly, the FAERS database is established by the FDA in the United States, leading to a stronger tendency to report and record AEs [18]. Secondly, the Dex implant was used in the United States relatively early, leading to a higher number of surgeries and reported adverse effects [3, 28]. Additionally, Kiristioglu and colleagues found that DME is the most common indication for the Dex implant and that most AEs are identified within one week, findings that align with the results of our study [23].

The most noteworthy finding emerging from this analysis is the strong signal strength associated with corneal decompensation, which requires greater attention. In recent years, numerous investigations have demonstrated the significant association between corneal decompensation and Dex implant anterior chamber migration [29, 30]. Scholars have reported that the incidence of anterior chamber migration ranges from 0.26 to 0.63% [23, 31]. Despite the rare incidence of implant migration, it potentially leads to vision-threatening outcomes. Khurana and colleagues demonstrated that corneal edema typically occurs in the early stage of migration, which may be attributed to the high dose of dexamethasone [32]. The toxicity of dexamethasone and the direct mechanical trauma further contribute to endothelial injury and corneal decompensation [33]. In an observational study of 15 patients who experienced anterior chamber migration and underwent immediate surgical removal, 14 patients developed corneal edema, and 6 patients required corneal transplantation [32]. Previous studies have indicated that the key risk factors for migration into the anterior segment include a history of vitrectomy, absence of the lens capsule, insufficient zonular support, and peripheral iridotomy [24, 31, 32]. After a vitrectomy, the vitreous cavity is filled with aqueous fluid, increasing the mobility of the implant and raising the risk of anterior chamber migration [34]. Additionally, patients with a history of intraocular surgery have a higher risk of irreversible corneal edema due to a reduced endothelial cell count. In rare cases, non-standard surgical techniques during implantation and subsequent migration of the Dex implant may also result in posterior capsule rupture, further increasing the risk of migration into the anterior segment. To minimize the risk of corneal decompensation, the Dex implant should be used with caution in patients with these risk factors. Regarding the choroidal haematoma, no previous studies have reported this AE. We hypothesize that transient hypotony following Dex implant could lead to fluctuations in choroidal microvasculature perfusion, subsequently causing a choroidal haematoma. A prospective, multi-center investigation involving DME patients also observed rare instances of vitreous hemorrhage and macular hole, findings consistent with our results [35]. Ophthalmologists should be vigilant about these relatively rare complications to mitigate potential risks effectively.

Extensive studies have previously reported that IOP elevation is the most common AE following Dex implant administration [36,37,38,39]. Notably, most cases of ocular hypertension can be effectively managed, with only a small number of patients requiring glaucoma surgery. Additionally, elevated IOP was determined as the most frequent AE of Dex treatment in non-infectious uveitic edema, which could generally be controlled with topical IOP-lowering medications [40]. Endophthalmitis, a severe and rapidly progressive intraocular inflammation, has been rarely reported in previous investigations. The incidence of endophthalmitis was approximately 0.1% after Dex implant, reported by multi-center studies [41,42,43]. In the present analysis, endophthalmitis emerged as the most commonly reported AE. A population-based cohort study from France indicated that the Dex implant is associated with a higher risk of endophthalmitis, compared with anti-vascular endothelial growth factor agents [44]. One potential explanation for the high number of reported cases could be the use of a 22-gauge needle for Dex implant insertion. The larger gauge may cause wound gaping, thereby increasing the risk of infection [45]. Additionally, corticosteroids may induce localized immunosuppression, potentially facilitating bacterial proliferation [46].

This study has several limitations that need to be considered and acknowledged. Firstly, the FAERS database operates as a spontaneous reporting system where AERs can be submitted not only by healthcare professionals but also by consumers. However, consumers’ limited medical expertise may lead to underreporting, incomplete data, and reporting biases. Additionally, while disproportionality analysis methods such as ROR and PRR are useful in detecting safety signals, they cannot establish a direct causal relationship. Furthermore, due to the limitations of the database, we are unable to determine the presence of concomitant medications or underlying diseases, nor can we perform stratified analyses to assess the effects of potential confounding factors. Lastly, although the study highlights significant signals related to ocular complications, the retrospective design limits the ability to determine the temporal relationship between Dex administration and AE onset. Future research should incorporate prospective designs and detailed clinical data to validate these findings and better understand the safety profile of Dex in diverse patient populations.

Based on the analysis of AEs associated with Dex in the FAERS database, this study highlights significant safety concerns that warrant careful consideration in clinical practice, particularly regarding high-risk complications such as corneal decompensation. Dex remains an effective treatment for various ocular conditions, the potential risks, including those not listed in the drug label, underscore the need for continuous pharmacovigilance. In conclusion, clinicians should be aware of both common and rare AEs when prescribing Dex and ensure appropriate follow-up and management to mitigate risks.