Drug-Induced Ocular Adverse Reactions: Review of the Safety Alerts Issued During the Last Decade

Introduction

Drugs used either in the treatment of ocular or systemic pathologic conditions have the potential to cause adverse events in the eye and to affect ocular functions. The blood–ocular barriers, namely the blood–aqueous barrier and the blood–retinal barrier, constitute a feature of the eye that plays an important role in limiting the ability of drugs to adversely affect this organ.^1,2 However, such structures may be subject to changes in their permeability that adversely impact the regulation of flow of molecules, such as water, ions, nutrients, or drugs, between the retina, retinal pigment epithelium, and blood vessels. ³ Thus, ocular disorders may be induced not only by drugs administered directly into the eye but also by systemically administered drugs, which reach the eye through the vascular supply.

There are several examples of ocular adverse drug events. Intraoperative floppy iris syndrome has been reported in patients treated with tamsulosin and who were undergoing cataract surgery. ⁴ Antineoplastic agents, namely docetaxel and paclitaxel, have been reported to be associated with the development of open-angle glaucoma. ⁵ Close-angle glaucoma has been reported in patients receiving adrenergic agonists, namely adrenaline and salbutamol, anticholinergics, including antidepressants, such as fluoxetine, paroxetine, fluvoxamine, or venlafaxine, and sulfa-based drugs, namely topiramate.^6–12

It is also interesting to note that there are data suggesting that ocularly administered drugs may be associated with the development of systemic adverse events. For example, the vascular endothelial growth factor (VEGF) inhibitors, which are administered through intravitreal injection, have been associated with an increased incidence of thromboembolic events and mortality. ¹³ In this context, the investigation of the regulatory actions resulting from the identification of serious ocular disorders associated with the use of drugs, irrespective of being locally or systemically administered, is of utmost importance given its implication for the clinical practice.

Frequent and well-recognized adverse events are potentially observed in randomized controlled trials. However, rare and long-term latency events are detected with difficulty in such studies due to their structural limitations. ¹⁴ Therefore, most evidence on harm is obtained from postmarketing surveillance, which may comprise spontaneous reporting systems and/or observational studies. ¹⁵ The additional knowledge on the safety profile of drugs arising from clinical practice is continuously evaluated to assure public health protection. To keep patients and healthcare professionals updated, authorities frequently issue drug safety alerts regarding benefit/risk ratio reevaluations being conducted and subsequent regulatory decisions.^16,17

The aim of this study is to evaluate ocular-related adverse events associated with the use of drugs by analyzing the content of safety alerts issued by major regulatory authorities during the last decade (2005–2014).

Methods

The websites of 4 health regulatory authorities were reviewed to identify safety alerts issued on ocular adverse events. Data were extracted from the following publicly accessible addresses: the United States Food and Drug Administration (US FDA) “Drug Safety Communications,” “Potential Signals of Serious Risks/New Safety Information Identified from the FDA Adverse Event Reporting System (FAERS),” “Postmarket Drug and Biologic Safety Evaluations” completed, “MedWatch Safety Alerts for Human Medical Products,” and “Biologics Safety and Availability Communications”; “Advisories, warnings and recalls” and “Canadian Adverse Reaction Newsletter” of Health Canada, European Medicines Agency (EMA); “News and press release archive,” “Meeting Highlights,” and “Agendas” from the “Committee for Medicinal Products for Human Use (CHMP)” and “Pharmacovigilance Risk Assessment Committee (PRAC)” monthly plenary meetings; and the Australian Therapeutics Goods Administration (TGA) “All alerts,” “Medicines Safety Update,” and “All monitoring communications.”^16–29

Safety alerts were included if they have been issued between January 2005 and December 2014. Natural and healthcare products, medical devices, contrast agents, drug–drug interactions, drug–food interactions, medication errors, and evaluations of lack of efficacy and adverse events occurring during off-label use were not considered for inclusion.

Only safety alerts on drugs with market authorization were considered for inclusion. The bibliographic lists of all relevant safety alerts were hand-searched to identify additional eligible safety alerts. The website search and the safety alerts selection were conducted by one researcher (A.P.). A second researcher scanned the included safety alerts (C.A.). Discrepancies were resolved by majority (2 out of 3) decision involving a third investigator (D.M.).

The following information from each safety alert was extracted: (1) date of first publication; (2) regulatory authority issuing the safety alert; (3) adverse event(s) of interest; (4) suspected drug(s); (5) type of data source supporting the evaluation; and (6) drug label section(s) updated. Drug remains under revision was used when the drug(s) were still under evaluation. Benefit/risk ratio unchanged was considered when the regulatory authority(ies) considered no iatrogenic increased risk.

The suspected drugs involved in each safety alert were coded according to the second-level therapeutic subgroup of the Anatomical Therapeutic Chemical (ATC) classification system. ³⁰ The adverse events were coded in the primary system organ class (SOC) and in the high-level term (HLT), according to the Medical Dictionary for Regulatory Activities (MedDRA), version 17.0. ³¹

Updates of the same safety alert were revised to retrieve further information on the regulatory authority decision and/or other data sources evaluated. A safety alert may contain one or more drugs and one or more adverse events. As an example, a safety alert issued by EMA in 2013 assessed ivabradine for phosphenes/vision blurred/diplopia/visual impairment and pegylated interferon alpha-2b for serous retinal detachment. Two safety alerts were considered to be evaluating the same clinical question when they assessed the same adverse event for the same suspected drug(s). Each safety alert could have been supported by more than one type of data source. More than one section of the drug label could have been updated. Regulatory authorities could have decided to keep the suspected drug(s) under revision, despite labeling changes having been carried out.

Results

A total of 38 safety alerts evaluating ocular adverse events were identified, of those, 4 (11%) were issued for ophthalmic medicines. Eighteen safety alerts were published by EMA, 14 by US FDA, 5 by Health Canada, and 1 by Australian TGA. Table 1 describes the characteristics of the safety alerts.

Drugs issued in safety alerts

The therapeutic subgroups of the suspected drugs reported in the safety alerts are presented in Figure 1. Urologicals (n=11; 29%), followed by drugs used in diabetes, antibacterials for systemic use, antineoplastic agents, and ophthalmologicals were the most frequently suspected drugs evaluated in the safety alerts (n=4; 10%, each).

OPEN IN VIEWER

FIG. 1.

Suspected drugs Anatomical Therapeutic Chemical (ATC) classification. A, alimentary tract and metabolism; A10, drugs used in diabetes; C, cardiovascular system; C01, cardiac therapy; C09, agents on the renin–angiotensin system; G, genitourinary system and sex hormones; G03, sex hormones and modulators of the genital system; G04, urologicals; J, anti-infectives for systemic use; J01, antibacterials for systemic use; J05, antivirals for systemic use; L, antineoplastic and immunomodulating agents; L01, antineoplastic agents; L03, immunostimulants; L04, immunosuppressants; M, musculoskeletal system; M05, drugs for treatment of bone diseases; N, nervous system; N03, antiepileptics; N05, psycholeptics; N06, psychoanaleptics; N07, other nervous system drugs; P, antiparasitic products, insecticides, and repellents; P01, antiprotozoals; R, respiratory system; R01, nasal preparations; R07, other respiratory system products; S, sensory organs; S01, ophthalmologicals.

Discussion

The development of ocular adverse events could be, in a certain way, more expected for ophthalmic pharmacological therapies locally administered than for systemic drugs. ³² However, drugs systemically administered to treat or prevent nonophthalmic conditions can also cause ocular adverse events. ³³

The suspected drugs more frequently associated with ocular adverse events subject to regulatory evaluation were the urologicals, namely sildenafil, vardenafil, and tadalafil. These 5-phosphodiesterase inhibitors have been associated with nonarteritic anterior ischemic optic neuropathy, a rare adverse event, which can lead to loss of visual acuity. ³⁴ The most common adverse effects of this drug class are changes in color perception, blurry vision, and increased light sensitivity. ³²

Regulatory authorities also reviewed the safety profile of drugs included in other classes, such as antidiabetic, antibacterial, or antineoplastic drugs. This diversity of suspected drugs has also been identified in other studies dedicated to review ocular toxicity. ³⁵

We identified a relatively low proportion of safety alerts (9%) issued for drugs indicated to treat ophthalmologic conditions. Moreover, these safety alerts were issued in the past 2 years for ophthalmic drugs introduced recently in the market when compared with medicines marketed more than a decade ago, such as pioglitazone, tamsulosine, or fluoroquinolones. Thus, this study suggests that, in the cases of the safety alerts evaluated herein, most of the newly identified ocular adverse events during the postmarketing surveillance are associated with nonophthalmic drugs.

Although ocular barriers may impair drugs causing eye toxicity, such structures may suffer permeability changes that adversely affect visual capacity. ² Moreover, the rich blood supply and relatively small mass of the eye make it particularly susceptible to drug-induced adverse events. ³³ Drug molecules present in the system may become deposited in specific ocular tissues, causing varied symptoms of drug toxicity. ³³

The most frequently evaluated adverse events by regulatory authorities were the visual disorders NEC, a hierarchical classification, including events such as visual impairment, diplopia, and blurred vision. Other severe conditions, such as retinal detachment, ischemic neuropathy, or macular edema, have also been associated with pharmacological treatment side effects during the last years. This indicated that the drugs affected several ocular structures, causing multiple disorders, which is consistent with previous findings. ³⁵ Most of the ocular drug adverse events observed in the clinical practice appear to be nonserious, usually harmful, and resolving with drug suspension. ³⁵

The eyelids are the eye structures more frequently affected by drug toxicity, which commonly manifests as an inflammation resulting from hypersensitivity reactions. ³² However, the ocular adverse events evaluated in this study are considered serious since they may progress and cause significant visual impairment, therefore leading regulatory authorities to reevaluate the benefit/risk ratio of the suspected drugs. As an example, the use of pegylated interferon alpha-2b in combination with ribavirin has been associated with the Vogt–Koyanagi–Harada disease, which is a multisystem syndrome characterized by ocular (uveitis and retinal detachment), neurological, and integumentary involvement. ³⁶ As another example, thiazolidinedione use was associated with an increased risk of macular edema among patients with type 2 diabetes. ³⁷

Spontaneous reports of adverse events supported the majority of the safety alerts addressed in this study. Rare and unpredicted events, as most of the ocular adverse reactions described herein, are prone to be spontaneously reported by healthcare professionals and patients to regulatory authorities when they occur within a relatively short period of time after the initiation of a drug therapy. ³⁸ Moreover, drug-induced ocular adverse events are the second most frequent reasons for claims against ophthalmologists. ³³ This could have influenced ophthalmologists to report adverse events to authorities. However, we do not know the extension to which such factors prompt regulatory decisions. Previous studies investigating data sources used to support regulatory decisions highlighted the value of spontaneous reporting systems in providing evidence to risk assessment. ¹⁵ As an example, the development of acute myopia or acute angle-closure glaucoma in association with the use of hydrochlorothiazide was detected by means of spontaneous reporting in existing few case reports described in literature.^39–41

A boxed warning and a change in the Therapeutic indications label section were the major regulatory actions taken. The use of retigabine (ezogabine in the United States) for the treatment of partial seizures in adults with epilepsy was associated with retinal pigmentation in long-term studies.^42,43 Of the 36 patients (of the ongoing studies) who had eye examinations, 11 developed retinal pigmentations associated with visual acuity reduced. ⁴² These data were sufficient to address a boxed warning alert for the importance of eye monitoring during the treatment and a change in therapeutic indications to patients for whom other antiepileptic medicines have proved inadequate or have not been tolerated.^42,43

The Warnings and Precautions section was the most frequently updated label section. Although no drugs have been withdrawn, there was a need to apply risk minimization measures during the last 10 years to reduce the incidence of ocular adverse effects. Most of the drugs evaluated in this study have been introduced in the market for several years. Unknown and serious adverse events are more likely to be identified for recently approved drugs. A previous study found that half of drug withdrawals occur within 2 years after a market authorization has been granted and that half of the major label changes (defined as drug withdrawal or black box warning inclusion) occur within 7 years after drug approval. ⁴⁴

Despite some ocular adverse events having been well described in premarketing clinical trials, their inclusion in the Adverse Reactions label section was dependent on the clarification of the relationship between these events and the suspected drug. This is the case of luminous phenomena (phosphenes) with ivabradine treatment, which were clearly identified in clinical trials; however, only after the dose relationship and pharmacological effect had been clarified, its inclusion in the drug label was performed.^45–47

None of the 38 safety alerts have been simultaneously raised and evaluated by the 4 regulatory authorities. It should be taken into consideration that the majority of the drugs evaluated in this study are approved in the countries regulated by the selected authorities. Thus, due to this difference on safety regulation, similar populations are exposed to different levels of risk. Moreover, during the last 10 years, it was also found that there is a significant difference in the number of drug-induced ocular disorder evaluations conducted in the selected countries since only 1 safety alert was raised by Australian TGA compared with 15 issued by US FDA during the same period. In a previous study conducted using a similar methodology, it was identified that different regulatory authorities took similar decisions within different time frames. ¹⁵ Although discrepancies in label changes were rarely identified in this study, Hirst et al. described some examples of different regulatory actions conducted for the same medicines in different countries. ⁴⁸ Guidelines and cooperative agreements have been developed to harmonize pharmacovigilance activities worldwide, but drug safety regulation might be affected by healthcare systems, regulatory procedures, and even cultural differences.^49–52

This study has some limitations. Only information posted by 4 regulatory authorities was considered for inclusion. Publications from other countries or regions and in languages other than English were not analyzed. This could lead to the exclusion of drug safety alerts, which might have been released elsewhere during the same period of time. The information evaluated in the study was collected from press releases and communications posted in the regulatory authorities websites. Therefore, some data could not have been entirely published as well as we did not attempt to confirm the findings with the authorities.

Over the last decade, innovative pharmacological treatments have been approved to treat serious ophthalmic conditions, such as biologic molecules for intravitreal injection.^53–55 These therapeutic options, although able to delay the progression of vision loss and improve patients' quality of life, have been associated with relatively frequent and severe iatrogenic profiles.^56–59 However, ocular adverse events newly identified during this decade come mostly from systemic drugs, some of them marketed for several years. Physicians should be aware of drug-induced adverse events in the eye to avoid, as soon as possible, their progression, which can lead to visual impairment.