(1)
Charlotte Eye Ear Nose & Throat Associates, Charlotte, NC, USA
Abstract
In the United States, screening for 4-aminoquinoline retinopathy (4AQR) is the standard of care. In this environment, a few cases occur despite attention to dosing by ideal body weight, other predisposing risk factors, and proper use and interpretation of ancillary tests. However, most clinically serious cases are iatrogenic, and arise from overdosing, missed screening for retinopathy, or tardy recognition of evidence of toxicity on ancillary testing. Symptoms and visual acuity are unreliable indicators of retinopathy, which develops as a continuum with three broadly defined stages: premaculopathy, early retinopathy, and advanced retinopathy. The boundaries between stages have been inconsistently specified. The probability of reversing retinopathy with cessation of 4-aminoquinolines decreases and the probability of progression despite cessation increases as the stage of retinopathy increases. No treatment exists for 4AQR.
Abbreviations
4AQR
4-aminoquinoline retinopathy
4AQs
4-aminoquinolines (chloroquine and hydroxychloroquine)
BAL
British anti-Lewisite
C
Chloroquine
EOG
Electro-oculogram
ERG
Electroretinogram
FAF
Fundus autofluorescence
HC
Hydroxychloroquine
mfERG
Multifocal electroretinogram
RA
Rheumatoid arthritis
RPE
Retinal pigment epithelium
SAP
Standard automated perimetry
SD-OCT
Spectral domain optical coherence tomography
SLE
Systemic lupus erythematosus
VF-25
Visual function 25 questionnaire
This chapter covers the natural history of chloroquine and hydroxychloroquine retinopathy. There is no evidence to suggest that the natural histories of these maculopathies differ from each other. Therefore, all the statements in this chapter apply to both drugs. Hydroxychloroquine and chloroquine will be referred to as 4-aminoquinolines (4AQs) and the retinopathy that they can cause will be termed 4-aminoquinoline retinopathy (4AQR). Commonly used abbreviations in this chapter are collected in “Abbreviations” for reference. Each term will be first used in its full form, along with its abbreviation.
6.1 Clinical Setting and Picture for Development of Chloroquine and Hydroxychloroquine Retinopathy
Although scenarios vary in which chloroquine and hydroxychloroquine retinopathy occur, certain characteristics are common. Most patients with retinopathy have been overdosed, and height and weight have not been checked to determine the ideal body weight (IBW) (see Chap. 7) [1, 2]. Occasionally renal or hepatic insufficiency has not been noted and an adjustment to dosing made to account for these characteristics. These most common mistakes led Morsman to write, “Examining the patient’s medical records is more useful than examining their eyes” [3]. In other cases, appropriate screening tests have been performed but misinterpreted as normal (see Chap. 8) [1]. Many patients have been screened at appropriate intervals, but the diagnosis of retinopathy has been missed (see Chap. 9) [1, 4, 5]. Although it is true that cases of 4AQR can occur in which everything has been done correctly, in most cases a failure analysis will indicate that the condition is iatrogenic.
6.2 Symptoms
Symptoms have been reported in 3–71 % of patients taking 4AQs [6–8]. Comparisons across case series are fraught with pitfalls because different examiners expend different amounts of effort eliciting symptoms. Symptoms can be expressed spontaneously by the patient or elicited by questioning by the clinician. When subgrouped in this way, half are in each category [6]. The common symptoms associated with chloroquine and hydroxychloroquine use and their relative frequencies are noted in Table 6.1. Trouble with reading is the most common symptom [9].
Table 6.1
Symptoms in patients taking 4-aminoquinolines and patients with 4-aminoquinoline retinopathy
Class of patient | Drug | N | Symptom | % | Study |
|---|---|---|---|---|---|
Taking 4AQs | C | 198 | Trouble reading | 16.2 | Percival [17] |
Blurred images | 8.6 | ||||
Haloes | 1.0 | ||||
Grittiness | 6.1 | ||||
28 | Blurred vision, colored rings, lack of accommodation | 6.8 | Reed [18] | ||
165 | Haloes | 7.3 | Hobbs [19] | ||
165 | Blurring | 5.5 | |||
165 | Halos and visual disturbances | 1.3 | |||
C or HC | 56 | Haloes | 12.5 | Henkind [6] | |
Difficulty adjusting to sunlight | 7.1 | ||||
Blurred vision | 7.1 | ||||
Vague awareness of eyes | 7.1 | ||||
Eyes tired | 5.4 | ||||
Diplopia | 3.6 | ||||
Dim vision | 3.6 | ||||
437 | Decrease vision or blurred vision | 2.9 | Bray [7] | ||
HC | 758 | Blurred vision, photophobia, lachrymation, red circles, lack of accommodation, diplopia | 1.1 | Grierson [20] | |
25 | Nyctalopia, color vision problems, glare | 2.0 | Chen [21] | ||
4AQR | C | 8 | Trouble reading | 6.3 | Okun [22] |
Photophobia | 6.3 | ||||
Blurred distance vision | 7.5 | ||||
Peripheral field defects | 3.8 | ||||
Floaters/flashes of light | 2.5 | ||||
16 | Blurred vision | 3.1 | Araiza-Casillas [23] | ||
Lack of accommodation | 1.3 | ||||
Decreased peripheral vision | 1.0 | ||||
C or HC | 3 | Trouble reading | 100 | Hickley [24] | |
Lack of accommodation | 100 | ||||
Decreased color vision | 6.7 | ||||
16 | Trouble reading | 6.3 | Michaelides [9] | ||
Decreased color vision | 1.3 | ||||
Nyctalopia | 1.3 | ||||
Blurred distance vision | 1.0 | ||||
HC | 7 | Decreased vision | 4.3 | Payne [25] | |
Flashing lights | 1.4 | ||||
Decreased color contrast | 2.9 | ||||
Disappearing letters | 1.4 | ||||
13 | Change in color perception | NG | Bienfang [26] | ||
Trouble finding a seat in a dim room | |||||
Scotomata on a TV screen | |||||
6a | Decreased or distorted vision | 100 | Bienfang [5] | ||
7 | Blurred vision | 42.9 | Mititelu [8] | ||
Shimmering lights | 28.6 | ||||
Trouble with night vision | 28.6 | ||||
Trouble reading | 14.3 |
The frequency of symptoms increases in patients who develop 4AQR, but many patients with retinopathy are asymptomatic [6, 9–14]. Easterbrook reported that 14 % of patients had symptoms at the stage where only relative paracentral scotomata were found, but 60 % of patients with absolute paracentral scotomata had symptoms [10, 15]. Thus, 40–86 % of patients with retinopathy are asymptomatic, depending on the stage of retinopathy [9, 10]. Patients often have retained excellent single-letter visual acuity scores, but have reading difficulties because of loss of paracentral photoreceptors [16]. In a study in which the visual function-25 (VF-25) questionnaire was given to patients with chloroquine retinopathy, the median composite score of patients was 33.9, the range being 10–81.9 [4].
A failure analysis of cases of 4AQR showed a median delay of 12 months between the onset of symptoms referable to 4AQR and cessation of 4AQs [4]. Although some observers are enthusiastic about the usefulness of symptoms as a screening item [5], most have found symptoms to be unhelpful in detecting retinopathy. On the other hand, noting symptoms is inexpensive. Therefore, the clinician would do well to search for new symptoms at each visit by a patient taking 4AQs. New symptoms warrant increased suspicion of retinopathy and the need for review of risk factors such as daily dosing based on IBW [15].
6.3 Signs of 4-Aminoquinoline Retinopathy
6.3.1 Visual Acuity
Visual acuity does not change in premaculopathy and may be excellent in advanced retinopathy with a bull’s-eye lesion and dense paracentral scotomata [4–6, 9, 22, 27–29]. Decrease in visual acuity secondary to chloroquine or hydroxychloroquine indicates advanced maculopathy and progression from a paracentral to a central scotoma on standard automated perimetry (SAP) [30]. Approximately 8.6 % (17 of 198) of patients with 4AQR have less than 20/20 best corrected visual acuity in one or both eyes [17].
Easterbrook suggested that visual acuity is a useful factor for predicting progression of retinopathy. His idea is that if patients have visual acuity of 20/20, then there is a good chance that the relative scotoma will not progress [31]. On the other hand, he estimates the probability at greater than 50 % that further deterioration in the retinopathy will occur if the visual acuity is less than 20/20 at the time that retinopathy is diagnosed [14].
6.3.2 Stages of 4-Aminoquinoline Retinopathy
Much of the early literature on chloroquine and hydroxychloroquine retinopathy contained observations and interpretations that are now recognized as erroneous. For example, macular edema and subretinal fluid were thought to be early signs of chloroquine retinopathy, but more sophisticated examination techniques, including spectral domain optical coherence tomography (SD-OCT), have shown that edema and subretinal fluid do not occur in 4AQR [6, 15, 18, 19, 32–36]. Nonspecific pigmentary changes of the macula were often attributed to 4AQR, but now it is thought that many of these observations reflected the common presence of pigmentary mottling in aging patients [6, 32, 37–39]. Peripheral retinopathy was claimed by some observers to predate central and paracentral changes, but it is now known that the order of changes is the reverse [40, 41]. A case of peripheral onset chloroquine retinopathy without maculopathy was reported, but not shown, and there remains doubt that this represented true chloroquine retinopathy [42]. Bull’s-eye macular lesions were reported to go away with cessation of treatment, but subsequent experience suggests that these were mistaken observations, and that bull’s-eye lesions do not go away [43, 44]. Although early observers claim to have had patients with bull’s-eye lesions without scotoma or visual loss [32, 33], these reports probably represent mistakes as well. The presence of a bull’s-eye lesion is always accompanied by an annular scotoma and usually by symptoms [6, 45, 46].
Patients with 4AQR have been categorized in several ways. Some authors categorize by groupings of factors. In this classification, patients are classified first based on funduscopic changes, and then are further subdivided into subgroups based on which ancillary tests are abnormal. A second system classification is probabilistic. Retinopathy is graded as possible (also called questionable), probable, or definite [46]. The former method has the problem of uncertain reproducibility, as most studies do not report test-retest statistics using chosen definitions. The latter methodology depends on the panel of graders, which raises the issue of generalizability. Both problems limit the applicability of the information generated to a clinical setting and suffer from the lack of standardization across observers. For example, case 2 illustrated in a paper by Okun and colleagues is termed early chloroquine retinopathy [22], yet the photograph (Fig. 2 from that reference) shows a definite bull’s-eye maculopathy, which would imply that it is an advanced case in most categorizations [47]. Table 6.2 lists commonly understood characteristics of the three major stages of retinopathy: premaculopathy, early maculopathy, and advanced maculopathy (see also Chap. 4).
Table 6.2
Stages of 4-aminoquinoline retinopathy
Characteristic | Premaculopathy | Early maculopathy | Advanced maculopathy |
|---|---|---|---|
Visual field | Scotoma to red target only | Scotoma to red or white target | Absolute scotoma to white object |
Symptoms | Rare | Sometimes | Usually present |
Visual acuity | Not affected | Rarely affected | More often affected, but can be normal |
Macula | Pigment stippling or loss of foveal reflex at worsta; often normal | Pigment stippling or loss of foveal reflexa | Bull’s-eye lesion |
Reversibility | Usually | Sometimes | No |
Onset or progression after cessation of drug | Rare | Uncommon | Possible |
There are varying definitions of premaculopathy (see Chap. 4) [7, 25, 38, 46, 48–54]. Some definitions specify that patients are asymptomatic but have macular pigmentary changes that are difficult to separate from normal aging changes [7, 25, 52, 54–57]. One common definition is the presence of central or pericentral scotomata to a red test object without an accompanying scotoma to a white test object [32, 34, 41, 58]. Others define this stage in a similar way as having a relative but not an absolute paracentral scotoma [10]. Some add to this definition those patients who lose the foveal reflex during therapy, a questionable additional criterion given the lack of any evidence that foveal reflex assessment is reproducible [17, 38]. Another definition is a state of interference at the metabolic level that has not yet produced morphologic changes and that is reversible [17]. An example of premaculopathy is shown in Fig. 6.1.
Fig. 6.1
This 38-year-old woman had been taking hydroxychloroquine at a dose of 400 mg/day for 21 years for systemic lupus erythematosus. She had no visual complaints and was seen for routine retinopathy screening evaluations. Her height was 66 inches and her weight 210 pounds. She had no liver or renal disease. (a) Her maculas had some drusen temporally, but no stigmata of hydroxychloroquine retinopathy were present. (b) Fundus autofluorescence imaging was normal bilaterally. (c) Relative paracentral scotomata were present on three consecutive 10-2 visual fields using a white target. (d) Multifocal electroretinography was normal bilaterally based on normal amplitudes in all rings and a normal R1/R2 ratio. (e) Spectral domain optical coherence tomography showed slight decreased reflectivity of the pericentral retinal pigment epithelial layer (yellow arrows) bilaterally. Although her daily dosage of hydroxychloroquine was 5.94 mg/kg based on an ideal body weight of 148 pounds, a typically nontoxic dosage, it was recommended that she decreased her hydroxychloroquine daily dosage to 300 mg. She had no relapse in lupus activity. She was asked to return for a recheck in 6 months rather than 1 year
From zero to 14 % of patients with premaculopathy were found to have ophthalmic symptoms in various studies [7, 10]. Reversibility (see next section) has been claimed for some patients at this stage of retinopathy, but documentation has been scant in support of this assertion [32, 41, 54, 58–60]. For example, Easterbrook reported on 44 eyes with relative paracentral scotomata after taking chloroquine or hydroxychloroquine [10]. Over a period of at least 4 years of follow-up, 5 (11 %) showed a decrease in size of the relative scotomas, but no eye returned to a normal visual field [10]. In a primate model of chloroquine retinopathy, histopathologic changes are evident before funduscopic changes occur [61].
The rate at which premaculopathy develops in hydroxychloroquine and chloroquine users has not been clearly defined in light of the variability of the definitions. Estimates of 3.3–17 % have been published for hydroxychloroquine users and 5.6 % for chloroquine users [7, 57].
There are few definitive case reports in which premaculopathy has progressed to advanced retinopathy (Fig. 6.2) [7, 57]. More frequently reported is progression to a more severe stage of retinopathy short of a bull’s-eye maculopathy [10]. For example, of 44 eyes having relative paracentral scotomas that were followed at least 4 years, 4 (9 %) showed an increase in size of the relative scotomas, and 2 (4 %) showed progression from relative to absolute scotomas [10]. Fine macular stippling has no prognostic value with respect to subsequent unequivocal retinopathy. Many older normal patients have this finding [17, 57]. Cessation of 4AQs at the stage of premaculopathy generally has a good prognosis, although exceptions occur (Fig. 6.2) [7, 10].
Fig. 6.2
Images of a 69-year-old woman with systemic lupus erythematosus who had taken hydroxychloroquine at a dosage of 400 mg/day from 1995 until 2006 when the drug was stopped secondary to retinopathy. She had no renal or liver disease. Her height was 5 feet 11 inches and her actual body weight 187 pounds; her ideal body weight 173 pounds. (a) She had a bull’s-eye macular lesion of the right eye (white arrow) consistent with advanced retinopathy and mild, nonspecific retinal pigment epithelial (RPE) mottling of the left macula consistent with early maculopathy. (b) Spectral domain optical coherence tomography (SD-OCT) of the right eye shows loss of the outer segments and inner segment/outer segment junction in the perifovea (white arrows). In the left eye the inner segment/outer segment junction is attenuated but not lost (yellow arrows). (c) Fundus autofluorescence images show a hypoautofluorescent ring around the fovea of the right eye corresponding to RPE atrophy. In the left eye there is a subtle locus of hyperautofluorescence (orange arrow). (d) Multifocal electroretinogram (mfERG) in the retina view (since the patient faces the observer, the patient’s right eye is the left panel). The averaged N1P1 amplitudes for rings R1 and R2 are subnormal, but the R1/R2 ratio is normal for the right eye. All of the N1P1 amplitudes are low-normal for the left eye and the R1/R2 ratio is normal. (e) 10-2 visual fields over time for the right eye show development of a ring scotoma that improves once the drug is stopped (2006). (f) 10-2 visual fields over time for the left eye showing parallel changes to the right eye. The scotoma depth is not as great in this eye. Note that the mfERG, SD-OCT, and 10-2 VFs agree in their indication of severity of retinopathy
Patients with premaculopathy can pose a dilemma for clinicians who do not wish to discontinue a beneficial drug but also wish to avoid the harm of retinopathy (Fig. 6.1). In such cases close follow-up and assuring that toxic dosing, the only modifiable risk factor, does not occur may be the best that can be done.
Early retinopathy is more advanced than premaculopathy but is also inconsistently defined, leading to numerous cases that might be placed in either category [29, 63]. Patients often do not have symptoms at this stage [64]. Some authors term this moderate rather than early retinopathy [65]. The presence of a relative scotoma to a white test object is part of the usual definition [11]. Some have included a minimum duration criterion, e.g., treatment for at least 9 months [66]. Fundus changes can be present, but no bull’s-eye lesion [22, 67]. In a primate retina model of chloroquine retinopathy, definite histopathological changes predated fundus changes [9, 61, 68]. Some authors describe patients in this group as having color vision problems and yet no visual field defects [29]. Various components of this stage include blunting or loss of the foveal reflex, parafoveal retinal pigment epithelial (RPE) irregularity, and parafoveal depigmentation (Fig. 6.2) [65, 69–71]. Loss of a foveal reflex is not a reliable sign since patients can have retinopathy and retain a foveal reflex [31]. Moreover, these changes can be seen in normal patients as they age [63]. Although some have said that early fundus changes are reversible in up to 50 % of such cases [54], skepticism is appropriate as there has been no photographic documentation to analyze [53]. Patients with early retinopathy may have multifocal electroretinographic (mfERG) abnormalities with or without visual field abnormalities (Fig. 6.2) [29].
With cessation of 4AQs, regression of scotomas is said to be less frequent than if scotomas are only present to a red test object [32]. Fundus changes are often absent at this stage. The mean onset of this stage in a series of four patients from among 262 screened was 3.5 years [72]. Over 4 years of follow-up of 44 such patients, there were no patients who progressed to loss of visual acuity [10].
The presence of an absolute paracentral scotoma is more advanced than the presence of a relative white scotoma. Sixty per cent of such patients will have symptomatic color vision problems, dropped letters, or symptomatic field loss [10]. Over a period of at least 4 years of follow-up, 63 % of such eyes will lose central visual acuity, and 13 % will experience a visual acuity decline from 20/40 or better to 20/200 [10].
The classically recognized bull’s-eye lesion of 4AQR signifies a very advanced stage of retinopathy that never reverts to normal [31, 46, 52, 62, 71, 73–75]. Patients at this stage will have more severe mfERG abnormalities [29]. The earliest that a bull’s-eye lesion has developed after beginning 4AQ therapy has not been established. There is a report of a lesion developing after 10 months of taking HC at 400 mg/day after a normal baseline examination, but photodocumentation was not provided [76]. Early fundus changes can progress to a bull’s-eye lesion within 3 years [77].
The bull’s-eye lesion has been described as having a hyperpigmented center with a surrounding hypopigmented ring [38, 54, 78, 79]. Occasionally, the lesion forms a circle (e.g., Fig. 6.2 and case 1 of Weiner and colleagues [80]). However, the lesion is usually horizontally oval, with more RPE atrophy in the inferior half of the oval (Fig. 6.2) [5, 50, 51, 81]. The earliest and broadest zone of RPE atrophy is often found in the inferotemporal quadrant [50, 80, 82]. The central zone in which the RPE remains intact is not oval, however, but rather more circular (Fig. 6.5) [30, 83]. Correspondingly, the density of the associated scotoma is often greater for the superior paracentral visual field (Figs. 6.2 and 6.4). Most cases of 4AQR are symmetrical, but asymmetrical cases have been reported (Fig. 6.2) (Fig. 4, [9, 11, 20, 22, 25, 33, 77, 81, 84, 85]). Asymmetry may be detected in ancillary testing as well. For example, Maturi noted mfERG abnormalities in one eye but not the fellow eye in four patients taking hydroxychloroquine and suspected of having early retinopathy [86]. Similarly, Amsler grid testing can show asymmetric scotomata [87].
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
