Target in Sight: A Comprehensive Review of Hydroxychloroquine-Induced Bull’s Eye Maculopathy

Abstract

Purpose of Review:

We review the latest screening and diagnostic techniques, and the most recent recommendations on the management of hydroxychloroquine retinopathy.

Recent Findings:

Hydroxychloroquine (HCQ) has been shown to cause retinal toxicity in a dose-dependent fashion. Early diagnosis is critical as the resultant retinopathy is not reversible. New imaging modalities, such as adaptive optics (AO), microperimetry, and retro-mode imaging, may show promise in the timely diagnosis of HCQ retinopathy.

Summary:

Automated visual fields and spectral-domain optical coherence tomography (SD-OCT) are the primary tests used in routine screening for HCQ retinopathy, but fundus autofluorescence (FAF) and multifocal electroretinogram (mfERG) have also been shown to be useful. A baseline ophthalmologic examination is recommended in all patients beginning long-term hydroxychloroquine therapy within the first year of starting therapy. Automated visual fields and SD-OCT should be included during this baseline exam in patients with pre-existing macular conditions. Afterwards, annual screening can be deferred for the first 5 years of HCQ treatment unless the patient has a major risk factor.

Hydroxychloroquine (HCQ) was first developed in the 1960s as an antimalarial drug, but it is now seldom used for that purpose due to widespread resistance.[1] It use has expanded to include autoimmune conditions due to its immunomodulatory properties.

Indications :

Hydroxychloroquine most common indication is the treatment of rheumatologic conditions such as systemic lupus erythematosus (SLE), rheumatoid arthritis (RA), and mixed connective tissue disorders.[2–4] In fact, it is estimated that over 50% of patients with SLE are treated with HCQ due to its demonstrated survival benefit, reduced incidence of diabetes, dyslipidemia, thrombosis and other cardiovascular complications, and its antithrombotic effect in patients with antiphospholipid antibody syndrome [2–7]. HCQ is also under investigation for treatment of several autoimmune dermatologic conditions and a range of infectious diseases, and an adjunct to a variety of chemotherapy agents in oncology. [8–10]

Recently, during the coronavirus disease 2019 (COVID-19) pandemic, HCQ was recommended by the Centers for Disease Control and Prevention (CDC) in hospitalized patients with mild-moderate COVID-19, based on early preclinical studies. In vitro studies demonstrated that HCQ had activity against the virus and many retrospective studies showed improved morbidity and/or mortality among patients hospitalized with COVID-19. [11–14] Subsequent large-scale randomized control trials (RCTs) failed to demonstrate a benefit in patients treated with HCQ and the recommendation was eventually withdrawn [15–17].

Contraindications:

Despite a relatively benign safety profile with documented safe use in pregnancy and in children, there are contraindications for the use of HCQ: hypersensitivity to HCQ or related compounds and pre-existing maculopathy. [18] Notably, HCQ is often continued during pregnancy in patients with SLE and its use has been associated with improved maternal and neonatal outcomes. [19]

Dosing:

Historically, HCQ was dosed based on ideal body weight (IBW), however, recent studies suggest that actual body weight (ABW) based dosing may better prevent systemic and retinal toxicity which are especially likely to occur in patients with compromised hepatic or renal function. [20– 22].

When used in autoimmune disorders, the duration of HCQ therapy is not clearly defined, and it is not uncommon for patients to take HCQ for decades or even for the remainder of their lives. Most patients require chronic suppressive therapy with a combination of disease modifying drugs with the goal of long-term disease control. [23]

Systemic Toxicity:

Mild adverse reactions to HCQ are relatively common, especially on drug initiation. [6] Gastrointestinal symptoms (which are often self-limited and usually resolve upon drug cessation) including nausea, vomiting, or diarrhea can affect up to one-third of patients. Cutaneous hyperpigmentation occurs in approximately one-fourth, and myopathy is a common, though rarely clinically significant, side-effect.[24] HCQ has been associated with hypoglycemia, which can be severe in patients taking antidiabetic medications. [25]

More severe HCQ-related toxicity is very rare. Cases of hepatic and renal failure as well as bone marrow suppression have been reported [6, 26]. Cardiac side effects, including cardiomyopathy and conduction anomalies such as prolonged Q-T interval, were a concern during the COVID-19 pandemic, but several prospective studies have demonstrated that these effects rarely result in adverse cardiac events. [27, 28]

Hydroxychloroquine-Induced Retinal Toxicity:

Mechanism

Numerous studies and animal models have attempted to characterize the mechanisms by which HCQ exerts its effects on the body, but a complete understanding of the pathophysiology of HCQ retinopathy remains elusive. HCQ is highly lipophilic, with a large volume of distribution, and long half-life, allowing the drug to easily penetrate cell membranes and lysosomes, disrupt toll-like receptor pathways and interfere with cyclic GMP-AMP mediators of gene expression [6, 29–35]. Other drug effects include inhibition of enzymes, cytokine release, cell signaling, apoptosis, antigen presentation, and the ability to directly inhibit natural killer cells [36–49].

The mechanisms for HCQ’s toxic effects on the retina are also incompletely elucidated, but may include inhibition of all-trans retinol recycling in the retinal pigment epithelium (RPE) via organic anion transporting polypeptide 1A2 (OATP1A2)[50, 51]; impaired phagocytosis of photoreceptor outer segments, disrupted lysosome function and autophagy in RPE cells[52]; autophagy-lysosomal defects leading to alterations in sphingolipid metabolism and early damage in the inner retinal neurons and retinal ganglion cells [53]; excess Ca²⁺ influx by stimulation of the transient receptor potential melastatin 2 (TRPM2) cation channel and led subsequent retinal oxidative toxicity[54].

Inter-individual variability and comorbid disease states may also play a role in the retinal toxicity of HCQ by affecting pharmacokinetics. The long half-life of 40–50 days can be as short as 5 days in some contexts and can differ significantly in individuals with severe COVID-19 infection [55–57]. SLE could confer additional risk for HCQ retinopathy over other indications for HCQ use [58].

Recent case reports showing retinopathy within 2 years of initiation– even as early as two months– suggest additional mechanisms and risk factors for HCQ toxicity, and even suggest the need for supplementary screening tests beyond those currently recommended for some patients [59, 60].

Risk Factors

The most recent American Academy of Ophthalmology (AAO) guidelines regarding HCQ retinopathy were published in 2016. According to those guidelines, major risk factors for HCQ retinopathy include HCQ dose >5.0 mg/kg ABW or chloroquine dose >2.3 mg/kg ABW, duration of use >5 years, renal impairment calculated by reduced glomerular filtration rate (GFR), tamoxifen use, and macular disease [21, 61]. Isolated drusen should not be seen as a contraindication to initiation of HCQ therapy.

The duration of HCQ use is perhaps the most widely appreciated risk factor for retinopathy among physicians. Although a few cases of early HCQ retinal toxicity have been reported since the updated 2016 guidelines, the risk is still considered insignificant before 5 years of continuous use [21, 58, 60, 61].

A. Dosing Risk Factors:

Another major consideration is the daily dosing of HCQ and cumulative lifetime dose of the drug, and the 2016 AAO recommendations were updated to incorporate updated practices based on new evidence. Traditionally, HCQ was dosed according to the ideal body weight (IBW) of the patient; however, multiple retrospective reviews demonstrated that patients received daily doses in excess to those recommended by AAO [62–64]. The 2016 guidelines recommended a change to using actual body weight (ABW) for safer and more accurate dosing [21, 61].

Notably, a recent small case-control study showed that elevated blood levels of HCQ were not associated with an increased risk of retinal toxicity, and thus should not be used as an indication to reduce dosages [65].

B. Ocular Risk Factors:

Early HCQ retinopathy screening should be initiated in patients with pre-existing maculopathy based on a presumed increase in risk for HCQ retinopathy [22, 61]. No study has demonstrated increased incidence of HCQ retinopathy in patients with pre-existing maculopathy, but the 2016 AAO clinical guidelines recommend a baseline examination with SD-OCT and 10–2 visual field testing for non-Asian patients (24–2 in Asian populations) within the first year of initial treatment [61].

C. Systemic Risk Factors:

Currently, only renal failure, as defined by GFR, tamoxifen use, and pharmacogenomic factors are considered major systemic risk factors for HCQ retinopathy. Age was previously considered a major risk factor in the 2011 AAO guidelines, but this was removed in the 2016 revision. This exclusion has been supported by a 2021 study showing no association between HCQ retinopathy and age or sex [66].

Renal dysfunction, specifically, a reduction in GFR of 50%, has been associated with a doubling of the rate of HCQ retinopathy [21]. Additionally, tamoxifen use has been shown to directly lead to retinal toxicity, as well as predispose or potentiate the mechanisms of HCQ toxicity [67–69]. The association between renal dysfunction, tamoxifen use and HCQ retinopathy has been redemonstrated in more recent studies as well [65].

In addition to these systemic factors, pharmacogenomic variables likely explain part of the apparent idiosyncratic nature of HCQ retinopathy. Previous studies have demonstrated that pathologic variants of the gene ABCA4 may predispose to retinopathy, while other variants of the same gene are thought to be protective [70, 71]. Further evidence of pharmacogenomic contributions to HCQ retinopathy comes from a recent case-control study which showed that the geographical origin of patients (patients from the West Indies and sub-Saharan Africa) was associated with higher risk of retinopathy [65].

Epidemiology

Though the study HCQ retinopathy epidemiology is difficult due in large part to the lack of large multi-center, generalizable data, the results of the Melles and Marmor study, the largest epidemiological study to date, remain the primary source of accurate information regarding the incidence of this disease [21]. According to this study, the overall prevalence of HCQ retinopathy is 7.5%. When stratifying by duration, the incidence is less than 1% at 5 years and less than 2% at 10 years. However, the incidence increases to around 20% after 20 years of use for those taking 4.0–5.0 mg/kg ABW/day and can exceed 50% at 20 years for those taking greater than 5.0 mg/kg ABW/day.

Interestingly, the clinical presentations of HCQ retinopathy vary by race and geographic origin. In a 2015 study, patients of European descent were more likely to exhibit a classic parafoveal pattern of deficits, while Asian patients were more likely to manifest an extramacular or pericentral pattern of disease with abnormalities near the vascular arcades [72]. 55% of Asian patients vs. 2% of Caucasian patients will develop a peripheral pericentral toxicity pattern instead of the well-known parafoveal pattern [73]. Black and Hispanic patients had predominantly parafoveal deficits, but pericentral/mixed deficits occurred more commonly in these races than in Caucasians [72]. A 2020 case control study showed that patients from the West Indies and sub-Saharan Africa had an increased risk of HCQ retinopathy (OR 8.7, P = 0.007) [65]. Future studies should use large, diverse datasets from multiple centers to further characterize differences in rates of HCQ retinopathy among different populations.

Clinical Features

Though patients taking HCQ with detectable retinopathy are frequently asymptomatic, symptomatic patients typically present with color deficits, missing central vision, difficulty reading, reduced or blurred vision, glare, flashing lights, and metamorphopsia [61, 74, 75]. Symptomatic patients have detectable abnormalities in the RPE that can evolve into a complete bull’s eye maculopathy on clinical exam. Early signs of bull’s eye maculopathy can be seen on fundus exam as macular edema and/or bilateral granular depigmentation of the RPE in the macula. Over time, this can progress to concentric rings of hypo- and hyperpigmentation surrounding the fovea, creating the classic “bull’s eye”. In late stages, attenuation of retinal arterioles and optic disc pallor can also be evident [74–76]. Figure 1. These changes are generally irreversible and are usually accompanied by severe vision loss and occasionally complicated by cystoid macular edema (CME), epiretinal membrane (ERM), and other sequelae [76].

Figure 1.

A. Color fundus photo of the left eye demonstrating “Bull’s Eye” maculopathy. Paracentral depigmentation of the retinal pigment epithelium (RPE) that spares the central fovea can be visualized. B. Fundus autofluorescence (FAF) photo of the left eye showing parafoveal hypo autofluorescence due to the low level RPE metabolic activity with underlying local atrophy with secondary photoreceptor loss. C. Optical coherence tomography (OCT) of the macula showing significant loss of the parafoveal photoreceptor inner segment/outer segment (IS/OS) junction and thinning of the outer retina with RPE irregularity. D. Humphrey visual field central 10–2 white-on-white pattern demonstrates dense paracentral scotoma.

Differential Diagnosis

The appearance of hydroxychloroquine maculopathy on exam, while characteristic, is not specific. The bull’s eye maculopathy seen in Figure 1 corresponds to foveal and parafoveal damage to the RPE [77]. The differential diagnosis for a finding of bull’s-eye maculopathy includes Stargardt’s disease, cone/rod and cone dystrophies, age-related macular degeneration, benign concentric annular dystrophy, neuronal ceroid lipofuscinosis, fenestrated sheen macular dystrophy, central areolar choroidal dystrophy, chronic macular hole, and macular telangiectasia [78]. Summary of the differential diagnosis for bull’s-eye maculopathy can be found in Table 1.

Table 1.

Differential Diagnosis of Bull’s Eye Maculopathy

Diagnosis	Clinical Characteristics
HCQ Retinopathy	• Typically presents after 5 years of HCQ use
Stargardt Disease	• VA loss usually in childhood/early adulthood • DFE reveals characteristic round, yellow white pisciform flecks in the posterior pole
Rod/Cone Dystrophy	• Presentation in first two decades of life • Presents with central scotomas and photophobia
Age-related Macular Degeneration	• Incidence higher in populations >70 • DFE reveals presence of Drusen
Benign Concentric Annular Macular Dystrophy	• Presents in childhood • Typically will have family history
Batten Disease	• Typically presents with rapid vision loss early in life • Associated with intellectual and motor deterioration and seizures
Fenestrated Sheen Macular Dystrophy	• Characterized by a yellow sheen with red fenestrations that is limited to the macula
Central Areolar Choroidal Dystrophy	• Macular changes in third and fifth decades • Progress gradually leading to severe vision loss between the fourth and seventh decades
Chronic Macular Hole	• Sharp border between the depigmented macular area and the surrounding normally pigmented RPE • HCQ retinopathy has relatively indistinct borders
Macular Telangiectasia	• Presence of ectatic vessels • Evidence of venules diving at a right angle into deeper retinal layer

Stargardt disease is the most common hereditary macular dystrophy. It is associated with loss of visual acuity typically starting in childhood or early adulthood. On fundoscopy, it can be differentiated from HCQ retinopathy by the characteristic round, yellow white pisciform flecks (due to lipofuscin A2E deposits) seen in the posterior pole.

Cone dystrophy and rod/cone dystrophy refer to a group of disorders that affect the cones and rods and cones, respectively. These dystrophies manifest with decreased visual acuity with central scotomas, reduced color vision, and photophobia, which help differentiate them from HCQ toxicity. These patients tend to present in the first two decades of life [79].

‌ Age-related macular degeneration (AMD) is the leading cause of blindness in the industrialized world. It is characterized by the accumulation of lipid and protein deposits (drusen), along with progressive degeneration of photoreceptors and adjacent tissues [80]. Advanced disease may involve parafoveal areas of retinal pigment epithelium (RPE) loss which can occasionally appear as a bull’s-eye maculopathy. AMD can be differentiated from HCQ toxicity by the presence of drusen as well as patient demographics. The pathophysiology of AMD is multifactorial in nature, but age is a primary risk factor. The prevalence of age-related macular degeneration sharply increases in those 75 years or older.

Benign concentric annular macular dystrophy (BCAMD) is an extremely rare, autosomal dominant macular dystrophy that presents in childhood with peripheral vision loss and decreased night vision [81]. These patients often have relatively preserved visual acuity compared to other focal RPE dystrophies. Family history of BCAMD and presentation in childhood aid in diagnosis.

Neuronal ceroid lipofuscinosis (NCL), also known as Batten disease is a member of a group of neurodegenerative disorders characterized by lysosomal accumulation of auto fluorescent lipopigments [82]. It can be easily differentiated from HCQ toxicity clinically as the most common form, juvenile neuronal ceroid lipofuscinosis (JNCL), presents with rapid vision loss early in life due to retinal degeneration. This is followed by progressive intellectual and motor deterioration and seizures and is invariably fatal by the third decade of life.

Fenestrated sheen macular dystrophy (FSMD) is a rare, slowly progressive, autosomal dominant macular dystrophy characterized by a yellow sheen with red fenestrations that is limited to the macula. Patients develop progressive annular hypopigmentation around the area of the sheen at the level of the RPE [83]. Eventually, a hyperpigmented area appears centrally giving rise to a “bull’s eye”-appearing lesion. Family history and the characteristic sheen are key in the diagnosis.

Central areolar choroidal dystrophy (CACD) is an incredibly rare hereditary dystrophy of the macula. It presents as a well-circumscribed lesion of the macula with loss of retinal and choroidal tissue. The RPE, choriocapillaris, and neurosensory retina are absent in dystrophy with some cases of CACD displaying a bull’s eye appearance of pigment epithelial atrophy. Family history and clinical picture play a role in diagnosis as macular changes usually begin between the third and fifth decades and progress gradually leading to severe vision loss between the fourth and seventh decades [84].

Chronic macular hole should be considered in the differential diagnosis of bull’s eye retinopathy as selective depigmentation of the RPE may occur under the cuff of subretinal fluid that surrounds a chronic macular hole. The pigmentation of the RPE underlying the macular hole is usually preserved and corresponds to the center of the bull’s eye pattern [85]. Unlike the bull’s eye maculopathy associated with HCQ retinopathy which has relatively indistinct borders, chronic macular hole generally has a sharp border between the depigmented macular area and the surrounding normally pigmented RPE.

Macular telangiectasia (MacTel) is a multifactorial macular disease characterized by abnormalities of capillaries of the fovea/perifoveal region and associated with atrophy of foveal photoreceptors which can progress to cystic slit-like changes and cavitations in all retinal layers. It is divided into 3 main groups with type 2 being the most common. The term “MacTel” usually refers specifically to type 2. Eyes with MacTel have a bull’s-eye retinopathy which can be differentiated from others by the presence of ectatic vessels with evidence of venules diving at a right angle into deeper retinal layer [78].

Screening and Monitoring

Current practice in the United States is guided by the 2016 AAO guidelines, which recommend the use of both automated visual fields and SD-OCT for routine screening. Fundus autofluorescence (FAF) and multifocal electroretinogram (mfERG) were also noted as useful screening tests [86]. While visual fields are potentially more sensitive, they are also subjective and thus variably reliable. SD-OCT is preferred as it is widely available, objective, highly specific, and considered sensitive to visually significant levels of toxicity. Given the variation in disease presentation in Asian vs non-Asian eyes, wider-angle scans or scans directed across the vascular arcade region are particularly important for these populations. The Royal College of Ophthalmologists (RCOphth) 2020 clinical guidelines support the inclusion of (FAF) along with SD-OCT in primary screening, reserving the use of automated visual fields for patients with a structural abnormality on imaging [61].

The AAO and RCOphth agree that at least one objective test (SD-OCT/FAF) should confirm findings on subjective testing before a diagnosis of HCQ retinopathy is made. The groups recommend that all patients beginning long-term HCQ therapy undergo baseline ophthalmologic examination within the first year of starting therapy in order to assess for any pre-existing ocular conditions, specifically of the macula, and to establish a record of the fundus appearance and functional status. Visual fields and SD-OCT are not necessary at this baseline exam unless other abnormalities are present that might affect screening tests. [86]

With proper dosing, the risk of HCQ retinopathy is low within the first decade of treatment, and annual screening can be deferred for the first 5 years of HCQ treatment unless the patient has a major risk factor: HCQ dose >5.0 mg/kg ABW, chloroquine dose >2.3 mg/kg ABW, macular disease, GFR of less than 60ml/min/1.73m², or tamoxifen use. Patients with any of these risk factors should be screened annually after beginning HCQ treatment [61].

Since initial damage can be recognized as a distinct, focal interruption of the photoreceptor outer segment, screening should aim to uncover previously unrecognized areas of retinopathy, rather than looking for gradual or chronic changes.

1. Spectral-domain optical coherence tomography (SD-OCT) is one of the most widely performed imaging modalities for the diagnosis and monitoring of macular disease. In hydroxychloroquine toxicity, there are parafoveal and, more commonly in Asian populations, peripheral thickness and volume changes that can be visualized on SD-OCT. These changes often present in the inferior/inferotemporal retina with the parafoveal retina more noticeably affected. Though, as previously alluded to, wide-angle SD-OCT imaging or SD-OCT scans through arcades are required to detect pericentral disease often seen in patients of Asian descent. In addition to outer nuclear layer thinning, morphological changes that can be seen on SD-OCT in hydroxychloroquine retinopathy include disruption of the photoreceptor layer, disruption of the IS/OS junction, loss of space between ellipsoid zone and interdigitation zone, loss of interdigitation zone, retinal pigment epithelium (RPE) loss and accumulation of debris and increased choroidal reflectance secondary to RPE loss [86].

2. Central visual field testing is frequently used in the screening of hydroxychloroquine retinopathy. It has the benefit of being a functional test and is very sensitive in reliable patients. However, it is a subjective test and its efficacy as a screening tool relies on the patient. 10–2 visual field testing has high resolution within the macula and is excellent for non-Asian patients, while wider test patterns (24–2 or 30–2) are needed for Asian patients in whom toxicity often manifests beyond the macula. Even a single central spot of reduced sensitivity in a 24–2 or 30–2 should be taken seriously, given the decreased sensitivity compared to 10–2 fields [61]. Uncertain visual field changes should trigger re-testing or evaluation with objective testing, such as SD-OCT.

3. Fundus autofluorescence (FAF) is an imaging technique that uses a monochromatic light source (typically 488nm) to elicit the auto-reflectance properties of lipofuscin within the RPE. In hydroxychloroquine retinopathy, there is often early hyperfluorescence indicating RPE stress, and later hypofluorescence indicating RPE loss [61]. The distribution of disease may be parafoveal or pericentral, as described in the SD-OCT section. Fundus autofluorescence can reveal early photoreceptor damage as findings of FAF may precede retinal thinning on SD-OCT. It is also valuable in that it gives a topographic view of damage across the posterior fundus.

Multifocal electroretinography (mfERG) is an objective electrodiagnostic test that involves the projection of an array of hexagonal light sources onto the retina. It has been shown to be sensitive and specific in the screening of hydroxychloroquine retinopathy however it is a costly test, and its availability is limited. Findings may include amplitude reduction, prolonged implicit time, ring response reduction or ring ratios greater than normal limits, as well as color difference plots indicating decreased response time [61, 86].

Tests not recommended for screening

Fundus examination and photography are useful for baseline examination when a patient initiates HCQ treatment as tests to screen for pre-existing macular conditions. However, for yearly screening for HCQ retinopathy, macula OCT and visual fields are more sensitive modalities and are recommended over fundus examination or photography. Other tests not recommended for yearly screening include time-domain optical coherence tomography, fluorescein angiography, full-field electroretinogram, Amsler grid, color vision testing, and electrooculogram [61]. While color vision changes can be found in HCQ retinopathy, regular color vision screening is not recommended in patients as changes in color vision have not been shown to be sensitive or specific. Further discussion can be found in Table 2.

Table 2.

Summary of Screening Recommendations.

Tests Recommended for Screening	Reasoning	Tests Not Recommended for Screening	Reasoning
Spectral Domain Optical Coherence Tomography (SD-OCT)	Retinal thickness and volume changes secondary to retinopathy can be visualized on SD-OCT	Fundus Examination and Photography	Photoreceptor damage is detectable with other techniques such as SD-OCT and FAF well before visible changes in the fundus
Fundus Autofluorescence (FAF)	FAF can reveal early photoreceptor damage	Time-Domain Optical Coherence Tomography	The resolution is insufficient to detect early toxicity
10–2 Automated Visual Field Testing (non-Asian populations)	Functional test that is very sensitive in reliable patients	Fluorescein Angiography	RPE defects may be visualized but later in the course of toxicity compared to other techniques
24–2 or 30–2 Automated Visual Field Testing (Asian populations)	Even a single central spot of reduced sensitivity in a 24–2 or 30–2 should be taken seriously, given the decreased sensitivity compared to 10–2 fields	Full-Field Electroretinogram	Abnormalities may be seen in only very late toxicity. It may be useful to judge the extent of damage beyond the macula
Multifocal Electroretinogram (mERG)	It has been shown to be sensitive and specific in the screening of hydroxychloroquine retinopathy	Amsler Grid	Other techniques such as HVF testing are more reliable in the screening of scotomas
		Color Vision Testing	Color vision errors may occur but are not sensitive or specific
		Electrooculogram	The electrooculogram has not been shown to be a reliable screening test.

Testing under investigation:

• Adaptive optics (AO), microperimetry, and retro-mode imaging all show promise in the diagnosis of HCQ retinopathy [87]. Adaptive optics allows direct visualization of individual photoreceptor cells and studies involving eyes with HCQ retinopathy have shown that structural disruption of photoreceptors seen on AO corresponds to ellipsoid zone (EZ) defects on SD-OCT. Another study showed that these changes on AO can be seen even before EZ changes are detectable on SD-OCT. However, AO equipment is not readily accessible, and the imaging is difficult to perform, analyze, and interpret.

• Microperimetry assesses the pointwise retinal sensitivity in the macula by integrating computerized threshold perimetry with real-time fundus imaging. Studies have evaluated microperimetry in the detection of HCQ retinal toxicity and showed inferior sensitivity but superior specificity when compared with mfERG. With this high specificity, microperimetry has been suggested as an ancillary test to exclude the diagnosis of hydroxychloroquine retinopathy if other screening tests provided conflicting results. However, microperimetry is not widely available limiting its practicality [87].

• Retro-mode imaging enables visualization of the outer retina using laser ophthalmoscopy. It has been shown that changes in parafoveal reflectance can be detected in patients with hydroxychloroquine retinopathy with high specificity. However, retro-mode imaging is limited in availability and thus its use remains currently impractical [87].

Management of HCQ Retinopathy

Hydroxychloroquine retinopathy is not reversible, and cellular damage appears to continue for a period even following drug discontinuation, though this is typically mild if the toxicity is recognized before there is RPE damage [88]. Yet, there is no therapy shown to be effective in treating or reducing the risk of developing HCQ retinopathy.

Patients taking hydroxychloroquine with age-related maculopathy or macular dystrophies are sometimes advised to avoid sun exposure and supplement lutein and zeaxanthin (which are foveal protectants). The value of these recommendations has yet to be demonstrated.

Primary prevention is currently the best approach in the management of HCQ toxicity. Once signs of retinopathy are recognized, the decision to stop medication should be made in conjunction with the patient and the prescribing physician so that systemic control of disease is optimized.

Conclusion

Hydroxychloroquine is known to cause irreversible retinal toxicity in a dose-dependent fashion and given the lack of treatments, primary prevention is the best strategy to avoid visual complications. Patients and prescribing physicians need to be mindful of this risk and the importance of regular screening. It is our hope that this review will further inform the screening and examination of patients taking HCQ prior to the presentation of visual symptoms.

Availability of data and materials

Not applicable