Dysregulation in fibroblast growth factor receptor (FGFR) signaling is a well-established oncogenic driver across multiple malignancies. Currently available FGFR inhibitors (FGFRi), including erdafitinib, pemigatinib, and futibatinib, are pan–FGFR–selective agents used in clinical practice for FGFR-altered urothelial carcinoma, cholangiocarcinoma, and other solid tumors. ^,, As their use expands across tumor types and into earlier lines of therapy, there is increasing recognition of class-specific toxicities that require careful monitoring.

Ocular adverse events (oAEs) are among the frequently reported toxicities associated with FGFRi, particularly dry eye and a range of corneal and retinal abnormalities. Clinical trials of erdafitinib, pemigatinib, and futibatinib have documented visual disturbances in up to one quarter of treated patients, including blurred vision, dry eye, retinal detachment, central serous retinopathy (CSR), and keratitis. ^,, Additional reports of bilateral retinopathy, serous retinal detachment, and subretinal fluid accumulation have been described. ^,, However, published data may underestimate the true incidence of FGFRi-associated ocular toxicities due to limitations of clinical trials and case reports, such as strict inclusion criteria and small sample sizes.

Postmarketing pharmacovigilance resources, such as the US Food and Drug Administration Adverse Event Reporting System (FAERS), provide an opportunity to identify rare, unexpected, or disproportionately reported ocular AEs that may not be captured in pre-approval studies. While prior FAERS-based analyses have evaluated the overall safety profile of FGFRi and identified both systemic and select ocular AEs, this study is the first to specifically and comprehensively characterize ocular adverse events associated with FGFRi in a real-world setting. Given the potential for visual impairment associated with FGFRi, ongoing real-world safety surveillance of ocular adverse events is essential.

In this study, we used the FAERS database to characterize the spectrum and disproportionality of reported ocular adverse events associated with the 3 US Food and Drug Administration (FDA)-approved FGFRi: erdafitinib, pemigatinib, and futibatinib. Our objectives were to identify the most frequently reported ocular toxicities, determine which events exhibited strong pharmacovigilance signals, and provide insights to guide clinical monitoring and timely management in patients receiving FGFRi.

A retrospective, population-based pharmacovigilance analysis was conducted using the US FAERS, a publicly available database that compiles spontaneous adverse events submitted by clinicians, patients, and pharmaceutical manufacturers across the US and internationally. FAERS serves as an essential tool for postmarketing safety surveillance and supports the detection of potentially novel or underestimated drug-associated harms. Each FAERS case report may include demographic variables, drug exposure information, indications, outcomes, and reporter characteristics.

All AE reports available in FAERS were accessed through OpenVigil 2.1 (Kiel, Germany), a validated interface designed for standardized querying, deduplication, and signal detection within large pharmacovigilance datasets. Erdafitinib, pemigatinib, and futibatinib were selected for analysis because they represent the FDA-approved FGFRi in clinical practice. Cases were included if erdafitinib, pemigatinib, or futibatinib was listed as the primary suspect medication in order to limit confounding from concomitant therapies. Reports were queried for each medication from initial FDA approval (Erdafitinib: April 19, 2019, pemigatinib: April 17, 2020, and futibatinib: September 30, 2022) through the most recent data update (June 30, 2025). ^,, Reports in which the drug was coded as a secondary suspect or interacting agent were excluded. While this approach may underestimate the total number of potential FGFR inhibitor-associated adverse events, it reduces misattribution in the setting of polypharmacy and improves the specificity of the observed drug–event relationship, consistent with standard practice in FAERS analyses. ^, Duplicate entries were identified and removed according to FDA case identifier guidance to ensure that only unique individual case safety reports were retained. All adverse events were standardized using the Medical Dictionary for Regulatory Activities Preferred Terms. The extracted dataset included all AE terms linked to the 3 FGFRi, along with available demographic fields such as age, sex, reporting country, and report year. To maintain minimum reporting stability, only AE terms with at least 3 individual reports were considered for disproportionality assessment.

Signal detection was performed using reporting odds ratios (RORs) with corresponding 95% confidence intervals (CIs). RORs quantify the likelihood of a given AE being reported for a specific drug relative to its reporting frequency for all other drugs within FAERS during the same period. Bayesian disproportionality metrics were calculated using the information component (IC) and its lower 95% credibility interval (IC ₀₂₅ ), derived from the same 2 × 2 contingency tables used for the reporting odds ratio analysis. The IC is a logarithmic observed-to-expected measure that incorporates Bayesian shrinkage, reducing the influence of low case counts and improving stability in sparse data settings. A signal was considered statistically significant if the lower bound of the IC ₀₂₅ exceeded 0. All statistical analyses were conducted using R version 4.5.2 (R Core Team). All analyses were descriptive and hypothesis-generating, consistent with best practices for pharmacovigilance research.

A disproportionate reporting signal was considered present following the established criteria described by Evans et al. :

• 1.

  The AE had ≥ 3 reports,

• 2.

  The χ² statistic exceeded 4, and

• 3.

  The proportional reporting ratio was > 2,

During the study period, a total of 1582 unique adverse events linked to FGFRi were reported. Of these reports, 562 were male, 414 were female, and 606 lacked sex information. The mean age of those containing information on age was 63.7 ± 16.8 years. The US accounted for the majority of AEs (1170; 73.9%), followed by France (129; 8.2%), Canada (39; 2.5%), and other countries (244; 15.4%) ( Table 1 ). Erdafitinib accounted for 918 reports (58%), pemigatinib for 576 (36.4%), and futibatinib for 88 (5.6%). Ocular AEs accounted for 200 of the total adverse events (12.6%), with 150 ocular AEs reported for erdafitinib (75%), 44 for pemigatinib (22%), and 6 for futibatinib (3%).

ERDAFITINIB

The oAEs with the greatest frequency of reports included dry eye (n = 31), eye disorder (n = 24), retinal detachment (n = 15), visual impairment (n = 14), and increased lacrimation (n = 12). The oAEs reported with the highest disproportionality signal included corneal thinning (ROR 321; 95% CI 102-1009), ocular toxicity (ROR 225; 95% CI 117-439), xerophthalmia (ROR 138; 95% CI 44-432), chorioretinopathy (ROR 121; 95% CI 60-243), detachment of the retinal pigment epithelium (ROR 116; 95% CI 48-279), serous retinal detachment (ROR 113; 95% CI 42-303), retinal edema (ROR 48; 95% CI 18-130), ulcerative keratitis (ROR 46; 95% CI 19-111), and retinal detachment (ROR 42; 95% CI 25-69), all of which remained statistically significant following Bayesian shrinkage (IC025 > 0). Table 2 and Figure 1 present a comprehensive list of oAEs associated with erdafitinib, categorized by report counts and ROR.

TABLE 2

Ocular Adverse Events Associated with Erdafitinib by Number of Absolute Reports

Adverse Event	n = 150	ROR (95% CI)	IC025
Anterior segment
Dry eye	31	17.27 (12.07-24.7)	2.97
Increased lacrimation	12	5.69 (3.22-10.07)	1.25
Ulcerative keratitis	5	45.99 (19.07-110.92)	1.25
Corneal thinning	3	321.12 (102.14-1009.48)	0.52
Xerophthalmia	3	138.46 (44.34-432.33)	0.51
Posterior segment
Retinal detachment	15	41.56 (24.93-69.27)	2.82
Chorioretinopathy	8	121.03 (60.18-243.42)	2.11
Detachment of retinal pigment epithelium	5	115.63 (47.85-279.41)	1.34
Maculopathy	5	24.33 (10.09-58.63)	1.13
Retinal edema	4	48.46 (18.12-129.59)	0.91
Serous retinal detachment	4	112.96 (42.15-302.72)	0.97
Non-specific/visual symptoms
Eye disorder	24	18.19 (12.13-27.29)	2.82
Visual impairment	14	2.71 (1.59-4.59)	0.52
Ocular toxicity	9	225.36 (116.83-438.56)	2.34
Blindness	8	4.58 (2.28-9.19)	0.74

CI = confidence interval; n = number of reports; ROR = reporting odds ratio.

Ocular adverse events associated with erdafitinib by signal strength. CI = confidence interval; n = number of reports; ROR = reporting odds ratio.

PEMIGATINIB

Forty-four oAEs were reported for pemigatinib. The most commonly reported events included dry eye (n = 20), subretinal fluid (n = 5), retinal detachment (n = 5), serous retinal detachment (n = 4), and ocular hyperemia (n = 4). The oAEs with the greatest disproportionality signals were serous retinal detachment (ROR 180; 95% CI 67-484), subretinal fluid (ROR 165; 95% CI 68-398), detachment of the retinal pigment epithelium (ROR 110; 95% CI 35-344), retinal detachment (ROR 22; 95% CI 9-53), and dry eye (ROR 18; 95% CI 11-28). All remained statistically significant following Bayesian shrinkage (IC025 > 0), with the exception of ocular hyperemia, which did not meet the IC025 threshold. The results are presented in Table 3 and Figure 2 .

TABLE 3

Ocular Adverse Events Associated with Pemigatinib by Number of Absolute Reports

Adverse Event	n = 44	ROR (95% CI)	IC025
Anterior segment
Dry eye	20	17.76 (11.37-27.75)	2.64
Ocular hyperemia	4	3.46 (1.29-9.25)	−0.09
Posterior Segment
Subretinal fluid	5	164.52 (68.01-397.95)	1.35
Retinal detachment	5	21.87 (9.06-52.76)	1.10
Serous retinal detachment	4	180.49 (67.27-484.34)	0.99
Detachment of retinal pigment epithelium	3	110.21 (35.35-343.62)	0.51
Non-specific/visual symptoms
Ocular discomfort	3	12.83 (4.12-39.9)	0.247

Only gold members can continue reading. Log In or Register to continue

Share this:

• X
• Facebook

Related

Related posts:

1. Anesthesia-Related Neurologic Risk in Patients of Venezuelan Descent: Implications for Ophthalmology
2. En Face Optical Coherence Tomography Imaging as a Structural Surrogate for Posterior Pole Inflammatory Retinal Vascular Activity
3. Progression in Preperimetric Glaucoma Eyes With and Without Microvasculature Dropout
4. Comment on “Macular OCTA Biomarkers Predict Regional Retinal Nonperfusion Patterns on Ultra-Widefield Angiography in Diabetes”

Stay updated, free articles. Join our Telegram channel

Join