Doc Ophthalmol DOI 10.1007/s10633-015-9486-x
ORIGINAL RESEARCH ARTICLE
Influence of chloroquine intake on the multifocal electroretinogram in patients with and without maculopathy Richard Bergholz • Klaus Ru¨ther • Jan Schroeter Christoph von Sonnleithner • Daniel J. Salchow
•
Received: 11 July 2014 / Accepted: 22 January 2015 Ó Springer-Verlag Berlin Heidelberg 2015
Abstract Purpose To evaluate the effect of long-term chloroquine intake on the multifocal electroretinogram (mfERG) in female patients with and without maculopathy. Methods Retrospective analysis of the mfERGs recorded in three different groups: (1) patients with bilateral maculopathy having taken chloroquine, (2) patients without maculopathy having taken chloroquine, and (3) healthy control subjects (age-matched to group 2) who never took chloroquine. MfERGs of each group were averaged, and the data of each patient group were compared to the control group. The main
Part of the data of this manuscript was presented as a poster at the annual meeting of the Deutsche Ophthalmologische Gesellschaft (DOG) 2014. R. Bergholz (&) C. von Sonnleithner D. J. Salchow Department of Ophthalmology, Charite´ – Universita¨tsmedizin Berlin, Campus Virchow-Klinikum, Augustenburger Platz 1, 13353 Berlin, Germany e-mail: [email protected] K. Ru¨ther Department of Ophthalmology, Sankt GertraudenKrankenhaus, Berlin, Germany J. Schroeter Institute of Transfusion Medicine, University Tissue Bank, Cornea Bank Berlin, Charite´ – Universita¨tsmedizin Berlin, Berlin, Germany
outcome measures were N1 and P1 characteristics and the ring ratio analysis. Results In group 1, 11 female subjects (22 eyes) were included, group 2 consisted of nine patients (18 eyes) and group 3 of seven healthy female subjects (14 eyes). Compared with healthy controls, patients in group 1 showed significantly reduced response densities of both N1 and P1 across all ring eccentricities except ring 5. Implicit times were significantly delayed only concerning N1 (ring 4 and the sum response of the left eye of group 1). P1 implicit times showed no significant alterations in either group. Ring ratios of the response densities were significantly higher mainly concerning group 1 (N1: ring 5/ring 2 and ring 5/ring 4 of the right eye; P1: all ring ratios of the right eye and all ratios except ring 5/ring 1 and ring 5/ring 4 of the left eye). The only ring ratio being significantly higher in group 2 was P1 ring 5/ring 1 ratio of the right eye. Conclusions In the absence of clinically apparent maculopathy, chloroquine intake was not associated with major alterations of the mfERG. Keywords Chloroquine Hydroxychloroquine Maculopathy Toxic retinal damage Multifocal electroretinography Introduction Although chloroquine and hydroxychloroquine (CQ and HCQ) are considered relatively safe medications
123
Doc Ophthalmol
for the treatment of rheumatic diseases and for the prophylaxis of malaria, their long-term use may lead to vision-threatening toxic retinal damage. Therefore, ophthalmologic screening of patients taking these drugs is recommended from the sixth year after initiation of therapy [1, 2]; in addition to a thorough clinical examination, retinal imaging and multifocal electroretinography (mfERG) are recommended. Macular function as measured by mfERG was reported to be impaired in asymptomatic patients taking HCQ without clinical evidence of maculopathy [3, 4]. This would demonstrate the sensitivity of the mfERG to detect subtle impairment of the retinal function but also complicate its clinical interpretation [5]. The question rises whether there is a clear threshold for maculopathy or whether toxic retinal damage from these drugs is a gradual process with insidious onset. The aim of this study was to evaluate the effect of long-term CQ intake on the mfERG in asymptomatic patients. This was done by comparing the mfERGs of these patients with the data of patients who suffer from CQ maculopathy as well as with healthy controls without maculopathy who have never taken either drug.
Methods The study complied with the rules of the Declaration of Helsinki. As this is a retrospective study, no approval of the ethics committee was requested and no informed consent was given by the patients. The internal eye hospital database was searched for patients taking or having taken CQ. Search period was from January 1, 2010, to January 1, 2013 (during this period, spectral-domain optical coherence tomography (SD-OCT) and fundus autofluorescence imaging were consistently performed on patients taking CQ). Patients were assigned to the following groups: 1. 2. 3.
Patients with bilateral maculopathy and a history of CQ intake. Patients without maculopathy taking CQ. Healthy control subjects without maculopathy who never took CQ. These subjects were chosen from the reference database established in our laboratory for the mfERG device used in this study. The control subjects were chosen from this
123
pool by age-matching them to the patients in group 2 as good as possible. To maximize the comparability of CQ patients without maculopathy and normal controls, age-matching between group 1 and group 3 was disregarded. To be classified as having no maculopathy (group 2), patients had to fulfill the following criteria: normal visual acuity (best-corrected decimal visual acuity of 1.0 or better), full visual fields as measured by static perimetry (macular perimetry with the Octopus 101 or Octopus 900 visual field analyzer, Haag-Streit, Switzerland), normal ocular fundus examination, fundus autofluorescence (Heidelberg Retina Angiograph, Heidelberg Instruments, Germany) and optical coherence tomography (Spectralis OCT, Heidelberg Instruments, Germany) with normal macular thickness and normal macular structure including continuous external limiting membrane and ellipsoid zone (IS/OS). Patients in group 1 had to show typical signs of toxic maculopathy in both eyes: (peri-)central scotoma on visual fields, bull’s eye pigmentary alterations in the macula and unequivocal imaging findings (pericentral outer retinal thinning on macular OCT, bull’s eye autofluorescence pattern). Control subjects in group 3 had no history of (H)CQ intake, normal visual acuity of at least 1.0 on the decimal scale and no signs of ocular disease. Subjects and patients with myopia, hyperopia or spherical equivalent higher than ±3.0 diopters and with astigmatism higher than 1.5 diopters were excluded from the study. In the database, there were just 10 male patients with a history of or ongoing CQ therapy and just one male patient with a manifest CQ maculopathy. Therefore, we limited this study on female patients. Multifocal electroretinogram MfERGs were recorded with a RETI-port system (Roland Consult, Brandenburg, Germany) and contact lens electrodes according to the standard of the International Society for Clinical Electrophysiology of Vision (ISCEV) [6]. A 61-hexagon stimulus was presented on a 51-cm cathode ray tube monitor with a frame frequency of 60 Hz. It encompassed the central visual field of 44° horizontally and vertically. A small cross in the center of the stimulus area served as fixation target. The
Doc Ophthalmol
scaled-size hexagons (distortion factor 4) were lightmodulated according to a binary pseudorandom m-sequence. The light state had a luminance of 120 cd/m-2 and the dark state of 1 cd/m-2. Refractive errors were not corrected. Pupil dilation was achieved with one drop of tropicamide–phenylephrine mixture into the conjunctival sac of each eye about 15 min before the exam. After topical anesthesia of the corneas with oxybuprocain hydrochlorid drops, the contact lens electrodes (ERG-jetÒ, CareFusion, San Diego, USA) were applied. Amplifier gain was set to 50,000 and the band pass from 5 to 100 Hz. There were eight recording cycles of 47 s each amounting to a total recording time of 6 min and 16 s. The 61 traces were grouped centrifugally into five rings, and the averaged first-order kernel was analyzed. The N1 amplitude was measured from the electroneutral starting point to the first negative deflection. The P1 amplitude was measured from the N1 trough to the following maximum positive peak. Implicit times were quantified from the moment of stimulus presentation to the peak of each deflection. Response densities were calculated as the quotient of amplitude/area [nanovolts per degree squared]. Ring ratios were calculated for the N1 and P1 response densities as described by Adam et al. [7]. Ring 5 served as a reference ring and as a dividend of rings 1–4. These ratios were also compared between each patient group and the control group using the t-test for independent groups. The R1/Rx ratios as described by Lyons et al. [8] were not calculated: In the first place, the central response is most prone to random distortion as it consists of the smallest amount of response averages when compared with the other ring eccentricities. Furthermore, in toxic maculopathy very often not only the pericentral responses but also the central response is affected. This stands in contrast to the relatively stable ring 5 responses even in toxic maculopathy, which is another reason to choose this eccentricity as a comparison. For each study group, the patient responses were averaged for right and left eye separately. N1 and P1 response densities, implicit times and the response density ring ratios were compared between the control group (group 3) and each study group (group 1 and 2) for the sum responses and each ring eccentricity (ring 1 [central response] and ring 2–5). Statistical analysis was done with the unpaired t-test for independent
groups; p values below 0.05 were regarded as statistically significant.
Results Eleven female patients were assigned to group 1, nine female patients to group 2 and seven healthy female subjects to group 3. Demographic and clinical data of these subjects are summarized in Table 1. Mean age for group 1 was 59.82 years, for group 2 51.0 years and for group 3 50.15 years. Patients without maculopathy had a longer mean duration of CQ intake (131.33 months) but a lower mean daily dose per ideal weight (4.21 mg/kg) than patients with maculopathy (115.82 months and 5.03 mg/kg, respectively). Mean daily dose of group 1 was 275 mg, of group 2 250 mg. These clinical data were compared between groups using the Mann–Whitney U test rendering the following results:
Table 1 Demographic and clinical data of the control subjects and patients Subject group Number of subjects/sex
1 11 females
2 9 females
3 7 females
Age (years) Mean
59.82
51.00
50.15
6.29
7.81
6.04
Duration of intake (months) Mean 115.82
131.33
n.a.
84.08
n.a.
SD
SD
53.04
Daily dose per ideal weight (mg/kg) Mean
5.03
4.21
n.a.
SD
1.45
0.26
n.a.
Eye
OD
OS
OD
OS
OD
OS
Refractive error (spherical equivalent, diopters) Mean
0.61
0.68
-0.01
0.01
-0.36
-0.27
SD
0.95
1.03
1.04
0.75
1.25
0.37
Best-corrected visual acuity (decimal) Mean
0.52
0.45
1.02
1.01
1.02
1.05
SD
0.25
0.30
0.07
0.08
0.15
0.10
Group 1—patients with maculopathy and a history of chloroquine intake, group 2—patients taking chloroquine without maculopathy, and group 3—healthy control subjects without maculopathy and without a history of chloroquine intake
123
Doc Ophthalmol
1.
2. 3. 4.
Age: Significant difference between group 1 and 3 (p \ 0.01*) but no significant difference between group 2 and 3 (p = 0.21). Duration of intake: No significant difference between group 1 and 2 (p = 0.824). Daily dose: No significant difference between group 1 and 2 (p = 0.72). Daily dose per ideal weight: No significant difference between group 1 and 2 (p = 0.059).
Standard mfERG analyses Figure 1 shows sample trace arrays of one subject of each group. Figures 2, 3, 4 and 5 show N1 and P1 response density and implicit time for each group, each eye and for each ring eccentricity. Statistically significant differences were found only between group 1 and controls (group 3). No significant differences were observed between group 2 and controls. As to group 1 (patients with maculopathy), both deflections (N1 and P1) showed significantly smaller response densities across all ring eccentricities except ring 5, as well as smaller sum responses than in group 3. As to implicit times, the N1 trough was significantly delayed only for ring 4 and the sum response of the left eye of group 1. No significant differences could be observed for P1 or between group 2 and group 3 in general. Group 2 showed a tendency toward smaller response densities and longer implicit times, but neither parameter reached a statistically significant level. Ring ratio analysis of the response densities For the N1 trough, ring 5/ring 2 ratio and ring 5/ring 4 ratio of the right eye were significantly higher in group 1 compared to group 3 (controls). For the P1 peak, all ring ratios of the right eye of group 1 were significantly higher than in group 3 as well as ring 5/ring 2 and ring 5/ring 3 ratio of the left eye. Concerning group 2, only the P1 ring 5/ring 1 ratio of the right eye was significantly higher than in group 3 (see Figs. 6, 7).
Discussion Toxic maculopathy is a recognized complication of CQ and HCQ intake. To investigate whether its
123
clinical manifestation is preceded by electrophysiologically measurable changes in retinal function, we analyzed mfERG responses of patients taking CQ but without toxic maculopathy and compared them to normal controls as well as to patients with established retinal damage due to CQ intake. Other authors reported a measurable impairment of the mfERG in patients taking HCQ without maculopathy: [3, 4] Lai et al. reported on reduced N1 and P1 amplitudes as well as increased P1 implicit times. Maturi et al. also measured reduced response densities in 11 of 19 patients and abnormal implicit times in 3 of 19 patients. While Lai et al. followed a longitudinal approach, we performed a cross-sectional study like Maturi et al. averaging the mfERGs of several subjects to minimize artificial distortion and the innate variability of the recordings. To our knowledge, our study is the first to include asymptomatic patients taking CQ. Nowadays, there are fewer patients taking CQ as the risk of acquiring maculopathy is higher than for HCQ. It would be expected that compared to HCQ the influence of CQ on the mfERG should be bigger even in patients without clinical maculopathy. As expected, the mfERGs of patients with CQ maculopathy showed significantly reduced response densities of both N1 and P1. This finding applied to all eccentricities but ring 5. The implicit times showed only few significant alterations even in group 1 and these only concerning N1. Regarding the standard analysis, asymptomatic patients taking CQ without maculopathy (group 2) did not show any significant alterations of their mfERG. Calculating the ring ratios of the response densities as described by Adam et al. [7] can help to minimize the influence of artificial distortion and noise on the intergroup comparison. Since CQ toxicity affects mainly the pericentral macula, the ratios of the peripheral reference ring divided by the more central rings response density are expected to rise. Ring ratio analysis showed significant increases in patients with maculopathy, particularly for the P1 response density. In contrast, there was just one significant difference of ring ratios between patients without maculopathy and controls (higher P1 ring 5/ring 1 ratio of the right eye). Therefore, this ratio probably is not more sensitive in detecting a deterioration of the macular function than the standard mfERG analysis.
Doc Ophthalmol Fig. 1 MfERG sample trace arrays of one subject of each group
There is a trend toward increased implicit times and decreased amplitudes in patients without maculopathy compared to control subjects. This trend did not reach significant levels in the statistical analysis. Nevertheless, these findings may be a sign of impaired retinal function even before the establishment of toxic maculopathy.
The mean duration of CQ intake was longer in patients without maculopathy than in patients with maculopathy. This is somewhat surprising, since it is known that longer CQ intake increases the risk of CQ maculopathy [9]. However, the CQ mean daily dose per ideal weight in patients without maculopathy was lower than in patients with maculopathy (4.21 vs.
123
Doc Ophthalmol
Fig. 2 Multifocal ERG N1 response densities of the different ring eccentricities and the sum response. Error bars show the 95 % confidence interval. p values are given above the error
bars. Significant differences compared to the control group are highlighted with asterisks
Fig. 3 Multifocal ERG N1 implicit times of the different ring eccentricities and the sum response
5.03 mg/kg, respectively). This corroborates that not only the duration of intake, but also the daily dose per ideal weight is a relevant risk factor [10]. There are certain drawbacks to this study: As this is a retrospective study, our data might have been distorted through selection bias. Furthermore, recall bias might have influenced the information on
123
daily and total drug dosage as well as the duration of intake. We had to confine our study to female patients because there was only one male patient with manifest CQ maculopathy in our database. The reason for this is unclear, but rheumatic diseases are significantly more prevalent in women than in men, and more women than
Doc Ophthalmol
Fig. 4 Multifocal ERG P1 response densities of the different ring eccentricities and the sum response
Fig. 5 Multifocal ERG P1 implicit times of the different ring eccentricities and the sum response
men receive antirheumatic medications such as CQ. Since our study included only female subjects, the findings may not necessarily apply to men. However, there is evidence that women probably do not have a higher vulnerability to toxic retinal damage than men [2]. The fact that age-matching was done only between group 2 and 3 and not between group 1 and 3: The age difference between the two latter groups might have influenced the comparison of patients with maculopathy
and the normal control subjects. But as the main objective of this study was to evaluate the mfERG impairment of CQ patients without maculopathy, we hold that this drawback should be admissible. While other authors measured significant amplitude reductions and implicit time prolongations even in asymptomatic patients taking the less dangerous HCQ, we were unable to detect any significant differences concerning implicit times or response densities between
123
Doc Ophthalmol
Fig. 6 Ring ratios of the N1 response density as described by Adam et al. [7]. Ring 5 served as a reference
Fig. 7 Ring ratios of the P1 response density as described by Adam et al. [7]. Ring 5 served as a reference
asymptomatic chloroquine patients and controls [3, 4]. One reason for this discrepancy might be that these other studies on the topic were performed before the advent of high-resolution retinal imaging. In every patient included in our study, OCT and fundus autofluorescence imaging were necessarily normal. The reported findings on HCQ toxicity are more likely
123
attributable to small sample sizes or incorrect recall of patient dosages although it cannot be ruled out completely that HCQ in fact leads to stronger alterations of the mfERG in asymptomatic patients than CQ. In summary, long-term intake of CQ not associated with clinically apparent maculopathy does not necessarily lead to major alterations of the mfERG.
Doc Ophthalmol Acknowledgments The authors would like to thank Anja Brune and Carola Lehrhaft for the technical assistance with the electrophysiologic recordings. Conflict of interest
None.
References 1. Marmor MF, Kellner U, Lai TY, Lyons JS, Mieler WF (2011) Revised recommendations on screening for chloroquine and hydroxychloroquine retinopathy. Ophthalmology 118:415–422 2. Pluta JP, Ruther K (2009) Retinal damage by (hydroxy)chloroquine intake: published evidence for an efficient ophthalmological follow-up. Klin Monbl Augenheilkd 226:891–896 3. Lai TY, Chan WM, Li H, Lai RY, Lam DS (2005) Multifocal electroretinographic changes in patients receiving hydroxychloroquine therapy. Am J Ophthalmol 140: 794–807 4. Maturi RK, Yu M, Weleber RG (2004) Multifocal electroretinographic evaluation of long-term hydroxychloroquine users. Arch Ophthalmol 122:973–981
5. Marmor MF (2005) The dilemma of hydroxychloroquine screening: new information from the multifocal ERG. Am J Ophthalmol 140:894–895 6. Hood DC, Bach M, Brigell M, Keating D, Kondo M, Lyons JS et al (2012) ISCEV standard for clinical multifocal electroretinography (mfERG) (2011 edition). Doc Ophthalmol 124:1–13 7. Adam MK, Covert DJ, Stepien KE, Han DP (2012) Quantitative assessment of the 103-hexagon multifocal electroretinogram in detection of hydroxychloroquine retinal toxicity. Br J Ophthalmol 96:723–729 8. Lyons JS, Severns ML (2009) Using multifocal ERG ring ratios to detect and follow Plaquenil retinal toxicity: a review: review of mfERG ring ratios in Plaquenil toxicity. Doc Ophthalmol 118:29–36 9. Bergholz R, Schroeter J, Ruther K (2010) Evaluation of risk factors for retinal damage due to chloroquine and hydroxychloroquine. Br J Ophthalmol 94:1637–1642 10. Ruther K, Foerster J, Berndt S, Schroeter J (2007) Chloroquine/hydroxychloroquine: variability of retinotoxic cumulative doses. Ophthalmologe 104:875–879
123
ORIGINAL RESEARCH ARTICLE
Influence of chloroquine intake on the multifocal electroretinogram in patients with and without maculopathy Richard Bergholz • Klaus Ru¨ther • Jan Schroeter Christoph von Sonnleithner • Daniel J. Salchow
•
Received: 11 July 2014 / Accepted: 22 January 2015 Ó Springer-Verlag Berlin Heidelberg 2015
Abstract Purpose To evaluate the effect of long-term chloroquine intake on the multifocal electroretinogram (mfERG) in female patients with and without maculopathy. Methods Retrospective analysis of the mfERGs recorded in three different groups: (1) patients with bilateral maculopathy having taken chloroquine, (2) patients without maculopathy having taken chloroquine, and (3) healthy control subjects (age-matched to group 2) who never took chloroquine. MfERGs of each group were averaged, and the data of each patient group were compared to the control group. The main
Part of the data of this manuscript was presented as a poster at the annual meeting of the Deutsche Ophthalmologische Gesellschaft (DOG) 2014. R. Bergholz (&) C. von Sonnleithner D. J. Salchow Department of Ophthalmology, Charite´ – Universita¨tsmedizin Berlin, Campus Virchow-Klinikum, Augustenburger Platz 1, 13353 Berlin, Germany e-mail: [email protected] K. Ru¨ther Department of Ophthalmology, Sankt GertraudenKrankenhaus, Berlin, Germany J. Schroeter Institute of Transfusion Medicine, University Tissue Bank, Cornea Bank Berlin, Charite´ – Universita¨tsmedizin Berlin, Berlin, Germany
outcome measures were N1 and P1 characteristics and the ring ratio analysis. Results In group 1, 11 female subjects (22 eyes) were included, group 2 consisted of nine patients (18 eyes) and group 3 of seven healthy female subjects (14 eyes). Compared with healthy controls, patients in group 1 showed significantly reduced response densities of both N1 and P1 across all ring eccentricities except ring 5. Implicit times were significantly delayed only concerning N1 (ring 4 and the sum response of the left eye of group 1). P1 implicit times showed no significant alterations in either group. Ring ratios of the response densities were significantly higher mainly concerning group 1 (N1: ring 5/ring 2 and ring 5/ring 4 of the right eye; P1: all ring ratios of the right eye and all ratios except ring 5/ring 1 and ring 5/ring 4 of the left eye). The only ring ratio being significantly higher in group 2 was P1 ring 5/ring 1 ratio of the right eye. Conclusions In the absence of clinically apparent maculopathy, chloroquine intake was not associated with major alterations of the mfERG. Keywords Chloroquine Hydroxychloroquine Maculopathy Toxic retinal damage Multifocal electroretinography Introduction Although chloroquine and hydroxychloroquine (CQ and HCQ) are considered relatively safe medications
123
Doc Ophthalmol
for the treatment of rheumatic diseases and for the prophylaxis of malaria, their long-term use may lead to vision-threatening toxic retinal damage. Therefore, ophthalmologic screening of patients taking these drugs is recommended from the sixth year after initiation of therapy [1, 2]; in addition to a thorough clinical examination, retinal imaging and multifocal electroretinography (mfERG) are recommended. Macular function as measured by mfERG was reported to be impaired in asymptomatic patients taking HCQ without clinical evidence of maculopathy [3, 4]. This would demonstrate the sensitivity of the mfERG to detect subtle impairment of the retinal function but also complicate its clinical interpretation [5]. The question rises whether there is a clear threshold for maculopathy or whether toxic retinal damage from these drugs is a gradual process with insidious onset. The aim of this study was to evaluate the effect of long-term CQ intake on the mfERG in asymptomatic patients. This was done by comparing the mfERGs of these patients with the data of patients who suffer from CQ maculopathy as well as with healthy controls without maculopathy who have never taken either drug.
Methods The study complied with the rules of the Declaration of Helsinki. As this is a retrospective study, no approval of the ethics committee was requested and no informed consent was given by the patients. The internal eye hospital database was searched for patients taking or having taken CQ. Search period was from January 1, 2010, to January 1, 2013 (during this period, spectral-domain optical coherence tomography (SD-OCT) and fundus autofluorescence imaging were consistently performed on patients taking CQ). Patients were assigned to the following groups: 1. 2. 3.
Patients with bilateral maculopathy and a history of CQ intake. Patients without maculopathy taking CQ. Healthy control subjects without maculopathy who never took CQ. These subjects were chosen from the reference database established in our laboratory for the mfERG device used in this study. The control subjects were chosen from this
123
pool by age-matching them to the patients in group 2 as good as possible. To maximize the comparability of CQ patients without maculopathy and normal controls, age-matching between group 1 and group 3 was disregarded. To be classified as having no maculopathy (group 2), patients had to fulfill the following criteria: normal visual acuity (best-corrected decimal visual acuity of 1.0 or better), full visual fields as measured by static perimetry (macular perimetry with the Octopus 101 or Octopus 900 visual field analyzer, Haag-Streit, Switzerland), normal ocular fundus examination, fundus autofluorescence (Heidelberg Retina Angiograph, Heidelberg Instruments, Germany) and optical coherence tomography (Spectralis OCT, Heidelberg Instruments, Germany) with normal macular thickness and normal macular structure including continuous external limiting membrane and ellipsoid zone (IS/OS). Patients in group 1 had to show typical signs of toxic maculopathy in both eyes: (peri-)central scotoma on visual fields, bull’s eye pigmentary alterations in the macula and unequivocal imaging findings (pericentral outer retinal thinning on macular OCT, bull’s eye autofluorescence pattern). Control subjects in group 3 had no history of (H)CQ intake, normal visual acuity of at least 1.0 on the decimal scale and no signs of ocular disease. Subjects and patients with myopia, hyperopia or spherical equivalent higher than ±3.0 diopters and with astigmatism higher than 1.5 diopters were excluded from the study. In the database, there were just 10 male patients with a history of or ongoing CQ therapy and just one male patient with a manifest CQ maculopathy. Therefore, we limited this study on female patients. Multifocal electroretinogram MfERGs were recorded with a RETI-port system (Roland Consult, Brandenburg, Germany) and contact lens electrodes according to the standard of the International Society for Clinical Electrophysiology of Vision (ISCEV) [6]. A 61-hexagon stimulus was presented on a 51-cm cathode ray tube monitor with a frame frequency of 60 Hz. It encompassed the central visual field of 44° horizontally and vertically. A small cross in the center of the stimulus area served as fixation target. The
Doc Ophthalmol
scaled-size hexagons (distortion factor 4) were lightmodulated according to a binary pseudorandom m-sequence. The light state had a luminance of 120 cd/m-2 and the dark state of 1 cd/m-2. Refractive errors were not corrected. Pupil dilation was achieved with one drop of tropicamide–phenylephrine mixture into the conjunctival sac of each eye about 15 min before the exam. After topical anesthesia of the corneas with oxybuprocain hydrochlorid drops, the contact lens electrodes (ERG-jetÒ, CareFusion, San Diego, USA) were applied. Amplifier gain was set to 50,000 and the band pass from 5 to 100 Hz. There were eight recording cycles of 47 s each amounting to a total recording time of 6 min and 16 s. The 61 traces were grouped centrifugally into five rings, and the averaged first-order kernel was analyzed. The N1 amplitude was measured from the electroneutral starting point to the first negative deflection. The P1 amplitude was measured from the N1 trough to the following maximum positive peak. Implicit times were quantified from the moment of stimulus presentation to the peak of each deflection. Response densities were calculated as the quotient of amplitude/area [nanovolts per degree squared]. Ring ratios were calculated for the N1 and P1 response densities as described by Adam et al. [7]. Ring 5 served as a reference ring and as a dividend of rings 1–4. These ratios were also compared between each patient group and the control group using the t-test for independent groups. The R1/Rx ratios as described by Lyons et al. [8] were not calculated: In the first place, the central response is most prone to random distortion as it consists of the smallest amount of response averages when compared with the other ring eccentricities. Furthermore, in toxic maculopathy very often not only the pericentral responses but also the central response is affected. This stands in contrast to the relatively stable ring 5 responses even in toxic maculopathy, which is another reason to choose this eccentricity as a comparison. For each study group, the patient responses were averaged for right and left eye separately. N1 and P1 response densities, implicit times and the response density ring ratios were compared between the control group (group 3) and each study group (group 1 and 2) for the sum responses and each ring eccentricity (ring 1 [central response] and ring 2–5). Statistical analysis was done with the unpaired t-test for independent
groups; p values below 0.05 were regarded as statistically significant.
Results Eleven female patients were assigned to group 1, nine female patients to group 2 and seven healthy female subjects to group 3. Demographic and clinical data of these subjects are summarized in Table 1. Mean age for group 1 was 59.82 years, for group 2 51.0 years and for group 3 50.15 years. Patients without maculopathy had a longer mean duration of CQ intake (131.33 months) but a lower mean daily dose per ideal weight (4.21 mg/kg) than patients with maculopathy (115.82 months and 5.03 mg/kg, respectively). Mean daily dose of group 1 was 275 mg, of group 2 250 mg. These clinical data were compared between groups using the Mann–Whitney U test rendering the following results:
Table 1 Demographic and clinical data of the control subjects and patients Subject group Number of subjects/sex
1 11 females
2 9 females
3 7 females
Age (years) Mean
59.82
51.00
50.15
6.29
7.81
6.04
Duration of intake (months) Mean 115.82
131.33
n.a.
84.08
n.a.
SD
SD
53.04
Daily dose per ideal weight (mg/kg) Mean
5.03
4.21
n.a.
SD
1.45
0.26
n.a.
Eye
OD
OS
OD
OS
OD
OS
Refractive error (spherical equivalent, diopters) Mean
0.61
0.68
-0.01
0.01
-0.36
-0.27
SD
0.95
1.03
1.04
0.75
1.25
0.37
Best-corrected visual acuity (decimal) Mean
0.52
0.45
1.02
1.01
1.02
1.05
SD
0.25
0.30
0.07
0.08
0.15
0.10
Group 1—patients with maculopathy and a history of chloroquine intake, group 2—patients taking chloroquine without maculopathy, and group 3—healthy control subjects without maculopathy and without a history of chloroquine intake
123
Doc Ophthalmol
1.
2. 3. 4.
Age: Significant difference between group 1 and 3 (p \ 0.01*) but no significant difference between group 2 and 3 (p = 0.21). Duration of intake: No significant difference between group 1 and 2 (p = 0.824). Daily dose: No significant difference between group 1 and 2 (p = 0.72). Daily dose per ideal weight: No significant difference between group 1 and 2 (p = 0.059).
Standard mfERG analyses Figure 1 shows sample trace arrays of one subject of each group. Figures 2, 3, 4 and 5 show N1 and P1 response density and implicit time for each group, each eye and for each ring eccentricity. Statistically significant differences were found only between group 1 and controls (group 3). No significant differences were observed between group 2 and controls. As to group 1 (patients with maculopathy), both deflections (N1 and P1) showed significantly smaller response densities across all ring eccentricities except ring 5, as well as smaller sum responses than in group 3. As to implicit times, the N1 trough was significantly delayed only for ring 4 and the sum response of the left eye of group 1. No significant differences could be observed for P1 or between group 2 and group 3 in general. Group 2 showed a tendency toward smaller response densities and longer implicit times, but neither parameter reached a statistically significant level. Ring ratio analysis of the response densities For the N1 trough, ring 5/ring 2 ratio and ring 5/ring 4 ratio of the right eye were significantly higher in group 1 compared to group 3 (controls). For the P1 peak, all ring ratios of the right eye of group 1 were significantly higher than in group 3 as well as ring 5/ring 2 and ring 5/ring 3 ratio of the left eye. Concerning group 2, only the P1 ring 5/ring 1 ratio of the right eye was significantly higher than in group 3 (see Figs. 6, 7).
Discussion Toxic maculopathy is a recognized complication of CQ and HCQ intake. To investigate whether its
123
clinical manifestation is preceded by electrophysiologically measurable changes in retinal function, we analyzed mfERG responses of patients taking CQ but without toxic maculopathy and compared them to normal controls as well as to patients with established retinal damage due to CQ intake. Other authors reported a measurable impairment of the mfERG in patients taking HCQ without maculopathy: [3, 4] Lai et al. reported on reduced N1 and P1 amplitudes as well as increased P1 implicit times. Maturi et al. also measured reduced response densities in 11 of 19 patients and abnormal implicit times in 3 of 19 patients. While Lai et al. followed a longitudinal approach, we performed a cross-sectional study like Maturi et al. averaging the mfERGs of several subjects to minimize artificial distortion and the innate variability of the recordings. To our knowledge, our study is the first to include asymptomatic patients taking CQ. Nowadays, there are fewer patients taking CQ as the risk of acquiring maculopathy is higher than for HCQ. It would be expected that compared to HCQ the influence of CQ on the mfERG should be bigger even in patients without clinical maculopathy. As expected, the mfERGs of patients with CQ maculopathy showed significantly reduced response densities of both N1 and P1. This finding applied to all eccentricities but ring 5. The implicit times showed only few significant alterations even in group 1 and these only concerning N1. Regarding the standard analysis, asymptomatic patients taking CQ without maculopathy (group 2) did not show any significant alterations of their mfERG. Calculating the ring ratios of the response densities as described by Adam et al. [7] can help to minimize the influence of artificial distortion and noise on the intergroup comparison. Since CQ toxicity affects mainly the pericentral macula, the ratios of the peripheral reference ring divided by the more central rings response density are expected to rise. Ring ratio analysis showed significant increases in patients with maculopathy, particularly for the P1 response density. In contrast, there was just one significant difference of ring ratios between patients without maculopathy and controls (higher P1 ring 5/ring 1 ratio of the right eye). Therefore, this ratio probably is not more sensitive in detecting a deterioration of the macular function than the standard mfERG analysis.
Doc Ophthalmol Fig. 1 MfERG sample trace arrays of one subject of each group
There is a trend toward increased implicit times and decreased amplitudes in patients without maculopathy compared to control subjects. This trend did not reach significant levels in the statistical analysis. Nevertheless, these findings may be a sign of impaired retinal function even before the establishment of toxic maculopathy.
The mean duration of CQ intake was longer in patients without maculopathy than in patients with maculopathy. This is somewhat surprising, since it is known that longer CQ intake increases the risk of CQ maculopathy [9]. However, the CQ mean daily dose per ideal weight in patients without maculopathy was lower than in patients with maculopathy (4.21 vs.
123
Doc Ophthalmol
Fig. 2 Multifocal ERG N1 response densities of the different ring eccentricities and the sum response. Error bars show the 95 % confidence interval. p values are given above the error
bars. Significant differences compared to the control group are highlighted with asterisks
Fig. 3 Multifocal ERG N1 implicit times of the different ring eccentricities and the sum response
5.03 mg/kg, respectively). This corroborates that not only the duration of intake, but also the daily dose per ideal weight is a relevant risk factor [10]. There are certain drawbacks to this study: As this is a retrospective study, our data might have been distorted through selection bias. Furthermore, recall bias might have influenced the information on
123
daily and total drug dosage as well as the duration of intake. We had to confine our study to female patients because there was only one male patient with manifest CQ maculopathy in our database. The reason for this is unclear, but rheumatic diseases are significantly more prevalent in women than in men, and more women than
Doc Ophthalmol
Fig. 4 Multifocal ERG P1 response densities of the different ring eccentricities and the sum response
Fig. 5 Multifocal ERG P1 implicit times of the different ring eccentricities and the sum response
men receive antirheumatic medications such as CQ. Since our study included only female subjects, the findings may not necessarily apply to men. However, there is evidence that women probably do not have a higher vulnerability to toxic retinal damage than men [2]. The fact that age-matching was done only between group 2 and 3 and not between group 1 and 3: The age difference between the two latter groups might have influenced the comparison of patients with maculopathy
and the normal control subjects. But as the main objective of this study was to evaluate the mfERG impairment of CQ patients without maculopathy, we hold that this drawback should be admissible. While other authors measured significant amplitude reductions and implicit time prolongations even in asymptomatic patients taking the less dangerous HCQ, we were unable to detect any significant differences concerning implicit times or response densities between
123
Doc Ophthalmol
Fig. 6 Ring ratios of the N1 response density as described by Adam et al. [7]. Ring 5 served as a reference
Fig. 7 Ring ratios of the P1 response density as described by Adam et al. [7]. Ring 5 served as a reference
asymptomatic chloroquine patients and controls [3, 4]. One reason for this discrepancy might be that these other studies on the topic were performed before the advent of high-resolution retinal imaging. In every patient included in our study, OCT and fundus autofluorescence imaging were necessarily normal. The reported findings on HCQ toxicity are more likely
123
attributable to small sample sizes or incorrect recall of patient dosages although it cannot be ruled out completely that HCQ in fact leads to stronger alterations of the mfERG in asymptomatic patients than CQ. In summary, long-term intake of CQ not associated with clinically apparent maculopathy does not necessarily lead to major alterations of the mfERG.
Doc Ophthalmol Acknowledgments The authors would like to thank Anja Brune and Carola Lehrhaft for the technical assistance with the electrophysiologic recordings. Conflict of interest
None.
References 1. Marmor MF, Kellner U, Lai TY, Lyons JS, Mieler WF (2011) Revised recommendations on screening for chloroquine and hydroxychloroquine retinopathy. Ophthalmology 118:415–422 2. Pluta JP, Ruther K (2009) Retinal damage by (hydroxy)chloroquine intake: published evidence for an efficient ophthalmological follow-up. Klin Monbl Augenheilkd 226:891–896 3. Lai TY, Chan WM, Li H, Lai RY, Lam DS (2005) Multifocal electroretinographic changes in patients receiving hydroxychloroquine therapy. Am J Ophthalmol 140: 794–807 4. Maturi RK, Yu M, Weleber RG (2004) Multifocal electroretinographic evaluation of long-term hydroxychloroquine users. Arch Ophthalmol 122:973–981
5. Marmor MF (2005) The dilemma of hydroxychloroquine screening: new information from the multifocal ERG. Am J Ophthalmol 140:894–895 6. Hood DC, Bach M, Brigell M, Keating D, Kondo M, Lyons JS et al (2012) ISCEV standard for clinical multifocal electroretinography (mfERG) (2011 edition). Doc Ophthalmol 124:1–13 7. Adam MK, Covert DJ, Stepien KE, Han DP (2012) Quantitative assessment of the 103-hexagon multifocal electroretinogram in detection of hydroxychloroquine retinal toxicity. Br J Ophthalmol 96:723–729 8. Lyons JS, Severns ML (2009) Using multifocal ERG ring ratios to detect and follow Plaquenil retinal toxicity: a review: review of mfERG ring ratios in Plaquenil toxicity. Doc Ophthalmol 118:29–36 9. Bergholz R, Schroeter J, Ruther K (2010) Evaluation of risk factors for retinal damage due to chloroquine and hydroxychloroquine. Br J Ophthalmol 94:1637–1642 10. Ruther K, Foerster J, Berndt S, Schroeter J (2007) Chloroquine/hydroxychloroquine: variability of retinotoxic cumulative doses. Ophthalmologe 104:875–879
123
