Customize your JAMA Network experience by selecting one or more topics from the list below.
Veverka KK, Hatt SR, Leske DA, et al. Prevalence and Associations of Central-Peripheral Rivalry–Type Diplopia in Patients With Epiretinal Membrane. JAMA Ophthalmol. 2017;135(12):1303–1309. doi:10.1001/jamaophthalmol.2017.4350
What is the prevalence of central-peripheral rivalry–type diplopia in patients with epiretinal membrane and what are the associated clinical characteristics?
In an unselected cohort of 31 patients with epiretinal membrane, the prevalence of central-peripheral rivalry–type diplopia was 16%. When evaluating clinical associations of 56 patients with epiretinal membrane, those with central-peripheral rivalry–type diplopia had better worse-eye visual acuity and more severe metamorphopsia than patients without central-peripheral rivalry–type diplopia (both differences were statistically significant), but aniseikonia and presence of retinal misregistration were similar.
These data suggest that central-peripheral rivalry–type diplopia is not uncommon in patients with epiretinal membrane and is associated with better visual acuity and more severe metamorphopsia; individual variability is marked.
The prevalence and clinical associations of patients with epiretinal membrane (ERM) who develop central-peripheral rivalry (CPR)–type diplopia are unknown.
To determine the prevalence of CPR-type diplopia in retinal disease clinic patients with ERM and to determine clinical findings associated with CPR-type diplopia.
Design, Setting and Participants
A prospective cross-sectional study of 31 patients with ERM from retinal disease clinics to determine the prevalence of CPR-type diplopia. A retrospective case cohort of 25 additional patients with ERM, selected from adult strabismus clinics, was added (total = 56) to determine clinical associations with CPR-type diplopia. All data were collected between June 2014 and November 2016; prospective cohort data were collected from June 2016 to November 2016.
Main Outcomes and Measures
The presence of diplopia was determined by patient history and diplopia questionnaire responses. Visual acuity and ocular alignment were recorded. Metamorphopsia was documented qualitatively by evaluation of the door frame and Amsler grid and measured quantitatively using M-charts and D-charts. Aniseikonia was determined by subjective description and results of the Awaya new aniseikonia test. Retinal misregistration testing consisted of optotype-frame test and synoptophore; CPR-type diplopia was defined as diplopia associated with evidence of retinal misregistration when other causes did not fully explain diplopia. Outcomes were as follows: prevalence of CPR-type diplopia in patients with ERM seen in retinal disease clinics, and whether or not clinical findings differed between patients with ERM and CPR-type diplopia vs patients with ERM without CPR-type diplopia.
Of the 31 patients with ERM seen in retinal disease clinics, 16 were women and 15 were men; the mean (SD) age was 69 (10) years. The prevalence of any diplopia was 23% (7 of 31; 95% CI, 10% to 41%), with CPR-type diplopia present in 16% (5 of 31; 95% CI, 5% to 34%). For analysis of associations, 12 of 56 patients (21%) had CPR-type diplopia and 37 (66%) had no diplopia. Seven of the 56 patients were excluded for other types of diplopia. Patients with CPR-type diplopia had better worse-eye visual acuity (mean difference, −0.23; 95% CI, −0.37 to −0.09 logMAR, P = .003), and more severe quantitative metamorphopsia (mean M-score difference 0.6; 95% CI, 0.05 to 1.1, P = .01) than patients without diplopia, but similar aniseikonia (Awaya new aniseikonia test; mean difference 0.6%; 95% CI, −2.9% to 4.0%, P = .33) and similar evidence of retinal misregistration (100% vs 73%; P = .09) by any test.
Conclusions and Relevance
Our findings suggest that CPR-type diplopia is not uncommon in patients with ERM. On average, patients with CPR-type diplopia have better visual acuity and more metamorphopsia than those without CPR-type diplopia, but there is considerable individual variability. Aniseikonia and retinal misregistration are similar between patients with ERM associated with CPR-type diplopia and those without CPR-type diplopia. Retinal misregistration with coexistent metamorphopsia appears necessary but is not sufficient for CPR-type diplopia.
Epiretinal membranes (ERM) may cause disruption of the retinal mosaics that can lead to clinical symptoms of aniseikonia,1-6 metamorphopsia,2,6-11 and decreased visual acuity.1,12 Binocular misregistration of the retinal mosaics may lead to central-peripheral rivalry (CPR)–type diplopia,13 also known as macular diplopia14 or dragged-fovea diplopia,2 and may manifest as foveal-peripheral rivalry on synoptophore testing.7 A previous study13 found that CPR-type diplopia was not uncommon in a select group of patients with ERM evaluated by a physician specializing in the treatment of strabismus. To our knowledge, the prevalence of CPR-type diplopia in patients with ERM presenting to physicians specializing in retinal disease has not been studied. Neither have the possible differences between clinical findings and optical coherence tomography (OCT) findings in patients with ERM, with and without CPR-type diplopia.
In the present study, we sought to quantify the prevalence of CPR-type diplopia in a sample of patients with ERM seen in clinics specializing in retinal disease and to compare the clinical findings and OCT findings of patients with ERM and CPR-type diplopia vs patients with ERM and no CPR-type diplopia.
All participants had a clinical diagnosis of ERM deemed to potentially affect vision based on review of OCT images by a retinal specialist (one of us, R.I.) in one or both eyes. Patients with an ERM judged to be minimal or only peripheral were excluded. OCT imaging was required to be performed within 6 months of the study examination. Patients were also excluded from the study if they had undergone prior surgical procedure for ERM (unless there had been a recurrence), if visual acuity was worse than 20/400 in either eye, or if a macular hole, wet age-related macular degeneration, or retinal vein occlusion with macular edema was present in either eye. Approval for analysis and data collection was obtained from the institutional review board at the Mayo Clinic, Rochester, Minnesota, and all patients provided written informed consent. All procedures were performed and data were collected in compliance with the Health Insurance Portability and Accountability Act. All research procedures adhered to the tenets of the Declaration of Helsinki.15
We prospectively recruited 31 patients with ERM from clinics specializing in retinal disease in order to determine prevalence of CPR-type diplopia. In the second part of this study, we incorporated an additional 25 patients from a clinic specializing in strabismus disorders. In our practice, patients without diplopia are often referred to the adult strabismus clinic for objective testing using standardized measures of diplopia, aniseikonia, metamorphopsia, and retinal misregistration. Seven of the 25 patients were recruited prospectively, and 18 patients were identified retrospectively, having undergone testing at previous clinical examinations. Data from the 56 patients (2 of whom had recurrent ERM following a previous ERM peel procedure) were used to compare clinical characteristics of patients with ERM and CPR-type diplopia with patients with ERM who did not have CPR-type diplopia.
Patient information from adjacent examinations within the past 6 months was also included to complete any missing data, provided there were no intervening surgical procedures. All 18 patients identified retrospectively for this study were included in a previous study describing types of diplopia diagnosed in 50 patients with ERM.13
All patients completed a standardized questionnaire16 quantifying diplopia with questions delivered either verbally or in written form. As reported previously,13 we defined diplopia as present if rated “rarely” or more frequently for straight ahead distance or reading, or if the patient was being treated with prism.
Ocular alignment was evaluated by a certified orthoptist using the simultaneous prism cover test and prism alternate cover test at a near distance of one-third meter and a far distance of 3 m.
Retinal misregistration was assessed using the optotype-frame test and superimposition slides designed for the synoptophore, as described previously.13 The optotype-frame test was designed to be performed at 3 m where the distance between the top of the letter and the inside edge of the screen subtended 2.9°.13 In the present study, the optotype-frame test was performed at 6 m (subtending 1.3°) for 18 of the 31 patients from the cohort prospectively recruited to assess prevalence. If the patient was unable to detect a 20/40-sized optotype with each eye, a 20/100-sized optotype was used (able to be detected with 20/400 visual acuity). As described previously,13 a double letter with a single frame or single letter with a double frame was recorded as retinal misregistration.13
To further test retinal misregistration, we used customized superimposition slides for the synoptophore13 based on those reported by Burgess et al.7 A positive finding of retinal misregistration was recorded if one or more peripheral X’s were misaligned using either 5° or 10° targets.
Four types of diplopia have been previously described in patients with ERM13: CPR-type diplopia alone, strabismic diplopia alone, indeterminate, and mixed CPR-type diplopia and strabismus. Central-peripheral rivalry–type diplopia alone was defined as binocular diplopia with retinal misregistration present by optotype-frame test or synoptophore and no other identified barriers to single vision. Strabismic diplopia alone was defined as horizontal strabismus greater than 0 prism diopters (PD) or vertical strabismus greater than 5 PD on simultaneous prism cover test, decompensating horizontal heterophoria, or resolution of diplopia with a surgical procedure or ground-in prism (excluding a Fresnel prism). Indeterminate diplopia was defined as diplopia without sufficient information regarding treatment response to determine diplopia type. Mixed CPR-type diplopia and strabismus was defined as the presence of retinal misregistration with improvement (but not complete resolution) of diplopia using a ground-in or Fresnel prism or a strabismus surgical procedure.
Aniseikonia was quantified using subjective description and the Awaya New Aniseikonia Test (Handaya Co. Ltd).17 For subjective description, the patient was asked to estimate the percent difference (horizontally and vertically separately) in the size of a 20/400 “E” at 3 m when viewed with each eye separately. For the Awaya New Aniseikonia Test,17 the patient reported which pair of semicircles appeared equal in size in each of the vertical and horizontal orientations, and the percentage of aniseikonia was recorded. For both tests, a larger image in the eye with the worse ERM (or affected eye in unilateral cases) was assigned a positive value, and a smaller image a negative value. The most extreme of horizontal and vertical values was used for analysis.
Metamorphopsia was assessed qualitatively by asking the patient to report whether the lines on an Amsler grid or the door frame of the examination room appeared wavy or distorted. Metamorphopsia was considered present if distortion was present by either method.
A subset of patients completed quantitative metamorphopsia testing. Using M-charts,18 we measured metamorphopsia in 0.1° increments up to a maximum of 2.0°. If distortion was still present when the patient viewed the 2.0° line, metamorphopsia was recorded as 2.1°. Testing was performed with the book at 0° and then rotated at 90° to provide vertical and horizontal values. Some patients were also tested with the computerized D-chart method11 and the degree of metamorphopsia was quantified as a raw D-score (overall severity of metamorphopsia) and as a weighted D-score (more weight to distortion of central vision).
Ocular dominance was determined by using the distance hole-in-the-card test19 or by instructing patients to form a hole between their hands at arm’s length. The test was repeated 3 times to determine the dominant eye.
Spectral-domain optical coherence tomography (SD-OCT) (Spectralis SD-OCT; Heidelberg Engineering Inc) measured macular volume and average central macular thickness of each eye using the automated system software. A retina specialist (one of us, R.I.), who was masked to diplopia status and clinical measurements, interpreted the SD-OCT images for presence or absence of tractional retinal detachment, photoreceptor abnormality, tractional retinoschisis, macular pucker, outer nuclear layer pucker, and cystoid macular edema in each eye.
The frequency of CPR-type diplopia alone or mixed CPR-type diplopia and strabismus was calculated to determine the prevalence of CPR-type diplopia. To determine associations with CPR-type diplopia, the clinical characteristics of patients with ERM and CPR-type diplopia and those of mixed CPR-type with strabismus were combined and compared with those of patients with ERM without diplopia. For patients with bilateral ERM, data from the eye with the worse ERM as indicated clinically, or by SD-OCT findings, was used for analysis.
Most clinical factors were nonnormally distributed by Shapiro-Wilk tests and, therefore, nonparametric Wilcoxon rank sum tests were used. Mean values with a 95% CI around the mean difference were also calculated. Because SD-OCT values were normally distributed, unpaired 2-tailed t tests were used for those analyses.
Thirty-one patients with ERM (16 women and 15 men; mean [SD] age, 69  years; range, 30-94 years) were prospectively recruited from clinics specializing in retinal diseases. Seven of 31 patients reported diplopia (23%; 95% CI, 10%-41%) and 24 of 31 (77%) reported no diplopia. The cause of diplopia for 2 of the 7 patients was indeterminate. Five of 31 patients had CPR-type diplopia (overall prevalence 16%; 95% CI, 5%-34%). Within this group, 3 patients had CPR-type diplopia alone (overall prevalence 10%; 95% CI, 2%-26%) and 2 had mixed CPR-type diplopia and strabismus (overall prevalence 6%; 95% CI, 1%-21%). No patient had strabismus alone as the cause of diplopia, or monocular diplopia.
Of the 56 study participants with ERM (with or without CPR-type diplopia) evaluated for associations with CPR-type diplopia, 31 patients were from the prevalence cohort and 25 patients were from the cohort seen in the strabismus clinic. In this combined cohort, 12 patients (mean [SD] age, 68  years; 5 women) had confirmed CPR-type diplopia. Five of the 12 patients were from the prevalence cohort and 7 were from the cohort seen in the strabismus clinic.
Ten of these 12 patients (83%) had diplopia distance straight ahead with the following frequencies: “always” (5 of 12 [42%]), “often” (1 of 12 [8%]), “sometimes” (1 of 12 [8%]), and “rarely” (3 of 12 [25%]). The remaining patients (2 of 12 [17%]) rated diplopia as “never” for distance straight ahead but both reported diplopia for reading, with 1 patient wearing a prism. The comparison group consisted of 37 patients (mean [SD] age, 68  years; 15 women), including 24 patients from the prevalence cohort and 13 patients from the strabismus cohort. Seven of 56 participants were excluded from this analysis because of indeterminate classification or other types of diplopia.
The clinical characteristics of patients with CPR-type diplopia vs without CPR-type diplopia are shown in Table 1. Patients with CPR-type diplopia had better visual acuity in the eye with the worse ERM than did patients without diplopia (mean difference, −0.23; 95% CI, −0.37 to −0.09 logMAR; P = .003). Qualitative metamorphopsia was almost universal (100% vs 70%, P = .09). By quantitative testing using M-charts, patients with CPR-type diplopia had more severe metamorphopsia (mean difference, 0.6; 95% CI, 0.05-1.1; P = .01). All patients with CPR-type diplopia measured using M-charts (n = 7) had marked metamorphopsia (lowest value, 0.6). The severity of metamorphopsia was similar between groups (P > .40) when the computerized D-chart method was used, but testing was more difficult to perform.
Aniseikonia was common in both patient groups, with no significant differences by aniseikonia subjective description (mean difference, 4.0%; 95% CI, −5.3% to 13.2%; P = .06) (Table 1) or by the Awaya New Aniseikonia Test (mean difference, 0.6%; 95% CI, −2.9% to 4.0%; P = .33). Spherical equivalent refractive error ranges and calculated optical aniseikonia were also similar between groups.
There was a high overall prevalence of retinal misregistration and no difference in overall prevalence between patient groups. However, differences in results were noted by testing method. A higher proportion of patients with CPR-type diplopia had retinal misregistration when measured using the optotype-frame test (10 of 12 [83%] vs 4 of 37 [11%]; difference, 73%; 95% CI, 49% to 96%, P < .001) and using the 10° targets on the synoptophore (12 of 12 [100%] vs 24 of 37 [65%]; difference, 35%; 95% CI, 20%-51%; P = .02). Of the 4 patients (11%) without CPR-type diplopia but with retinal misregistration on the optotype-frame test, 3 patients (75%) were also positive on the 10° synoptophore targets.
Ocular dominance testing revealed that 5 of 10 patients (50%) with CPR-type diplopia had eye dominance of the eye with worse ERM compared with 11 of 32 patients (34%) with ERM and without CPR-type diplopia (P = .46, Fisher exact test).
The SD-OCT findings of patients with ERM and CPR-type diplopia vs patients with ERM but without CPR-type diplopia are provided in Table 2. The mean central macular subfield thickness was lower (mean difference, −57 µm; 95% CI, −129 to 14 µm; P = .11) and the cystoid macular edema was less frequent (4 of 12 [33%] vs 23 of 37 [62%]; difference, −29%; 95% CI, −60% to 2%; P = .10) for patients with ERM and CPR-type diplopia, but differences were not statistically significant.
In the present study, CPR-type diplopia was present in 16% of an unselected cohort of patients with ERM deemed to potentially affect vision. Patients with CPR-type diplopia had significantly better visual acuity and greater metamorphopsia by M-chart findings. Aniseikonia and retinal misregistration results were similar and SD-OCT findings showed no significant differences. For both clinical and OCT measures, there was marked individual variability.
We are not aware of previous studies that evaluate the prevalence of CPR-type diplopia in patients with ERM. A previous study13 reported diplopia in patients with ERM seen in an adult strabismus clinic, while other studies only included patients presenting with diplopia1,2,7,9,20-24 or aniseikonia.5,13 In a previous study describing different causes of diplopia in patients with ERM,13 true strabismus was reported to be the cause of diplopia in 28%. Clinicians should be aware of the causes of diplopia in patients with ERM to facilitate appropriate referral and treatment.
We found that patients with ERM and CPR-type diplopia had significantly better visual acuity than similar patients without CPR-type diplopia. This finding is consistent with data reported by De Pool et al2: in that study, 88% of affected eyes had a visual acuity of 20/40 or better. In the present study, worse-eye visual acuity was on average approximately 20/30 among patients with diplopia (no patient worse than 20/100) compared with an average visual acuity of approximately 20/50 among patients without diplopia (Table 1). These data suggest that poorer worse-eye visual acuity may be protective against the perception of diplopia; however, there is considerable individual variability, including poor visual acuity in some patients with diplopia and good visual acuity in some patients without diplopia. Anecdotally, some patients have onset of CPR-type diplopia following ERM peeling procedure associated with improvement of visual acuity. This topic requires further study.
Metamorphopsia is a common overall finding in patients with ERM,25 including those with diplopia.2,7 We are unaware of previous studies comparing metamorphopsia in patients with ERM, with or without diplopia. The high frequency of metamorphopsia in patients with ERM without diplopia suggests that the presence of metamorphopsia alone is not sufficient to induce CPR-type diplopia. It is interesting that CPR-type diplopia was associated with more severe metamorphopsia, which may have therapeutic implications for the ERM peeling procedure.
In previous studies, retinal aniseikonia has been identified as a possible factor in the development of symptomatic diplopia.1,2,4,5 Benegas et al,1 suggested that aniseikonia greater than 5% contributed to the experience of diplopia. Our current study found that similar proportions of patients with diplopia or without diplopia had 5% or greater aniseikonia (P = .17; Table 1). Rutstein5 found that 10 of 12 patients with retinal aniseikonia (range, 1.7%-11.3%) had symptoms of diplopia. In the present study, we found that, on average, aniseikonia was similar between the patient groups. However, some patients with CPR-type diplopia had no aniseikonia and some patients without CPR-type diplopia had high degrees of aniseikonia, suggesting that CPR-type diplopia does not always develop when aniseikonia exceeds a certain threshold.
We defined retinal misregistration as present if it was detected by either the optotype-frame test or synoptophore. A previous study13 used the optotype-frame test alone in the definition of CPR-type diplopia. Nevertheless, most patients in that study were reported to have positive misregistration findings on the synoptophore test.13 In the present study, many patients tested positive for retinal misregistration by the synoptophore but did not test positive when the optotype-frame test was used (Table 1). The synoptophore targets subtend 5° and 10°; by extending further into the periphery, this test may provide a greater opportunity to detect any retinal misregistration. Because the synoptophore test is not readily available in every clinical practice, the optotype-frame test13 may still provide a practical means of assessing retinal registration with the understanding that a false-negative result may occur in some cases.
Data from the present study suggest that a combination of clinical factors (retinal misregistration, good visual acuity, and metamorphopsia) increases the risk of CPR-type diplopia in patients with ERM. However, the specific contribution of these and other factors in the pathogenesis of CPR-type diplopia remains unclear. The presence of retinal misregistration with coexistent metamorphopsia appears necessary for the development of CPR-type diplopia, but presence of these factors alone is not sufficient; many patients without CPR-type diplopia also demonstrate these findings. It is possible that critical thresholds or combinations of factors precipitate the development of CPR-type diplopia. It is also possible that other factors (eg, underlying ocular misalignment, motor fusion amplitudes, ocular dominance, or optical aniseikonia) may influence the individual associations of clinical measures with CPR-type diplopia.
We found that patients with ERM and CPR-type diplopia had numerically lower mean central macular subfield thickness and less frequent cystoid macular edema on SD-OCT imaging than patients with ERM and without diplopia, although differences were not statistically significant. We are unaware of previous studies evaluating SD-OCT measures in patients with ERM regarding associations with CPR-type diplopia. The specific role of these factors in the development of diplopia remains unclear. Some previous studies have evaluated SD-OCT findings in relation to visual acuity in patients with ERM and found that thinner central macular subfield thickness was associated with better best-corrected visual acuity.26 We currently have only a preliminary understanding of how retinal changes related to ERM cause specific symptoms, such as CPR-type diplopia.
There are some limitations to our study. Patients reporting varying degrees of diplopia, including the occurrence of diplopia as “rarely,” may have reduced our ability to detect differences between patients with diplopia and without diplopia. We excluded patients with certain coexisting retinal conditions because the ERM was not the primary cause of visual symptoms. The prevalence and associations with clinical measures of CPR-type diplopia may differ when other retinal conditions dominate. When evaluating associations with CPR-type diplopia, some patients were identified retrospectively. Data from retrospectively identified patients were complete; however, they were collected before full implementation of a standardized testing protocol and there may have been small differences in how testing was performed. We were unable to create multiple logistic regression models because of sample size.
Central-peripheral rivalry–type diplopia is the most common cause of diplopia in unselected retina clinic patients with ERM deemed to potentially affect vision. Patients with ERM and CPR-type diplopia have better visual acuity and more severe metamorphopsia than patients with ERM and no CPR-type diplopia, but there is considerable individual variability. It appears that retinal misregistration with coexistent metamorphopsia is necessary for the development of CPR-type diplopia, but that presence alone is not sufficient to develop this condition.
Corresponding Author: Jonathan M. Holmes, BM, BCh, Department of Ophthalmology, Mayo Clinic, 200 First St SW, E4, Rochester, MN 55905 (firstname.lastname@example.org).
Accepted for Publication: September 2, 2017.
Published Online: November 16, 2017. doi:10.1001/jamaophthalmol.2017.4350
Author Contributions: Drs Iezzi and Holmes are considered co–senior authors. Dr Holmes had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Veverka, Hatt, Leske, Iezzi, Holmes.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Veverka, Hatt, Holmes.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Veverka, Leske, Holmes.
Obtained funding: Holmes.
Administrative, technical, or material support: Leske, Barkmeier, Iezzi, Holmes.
Study supervision: Iezzi, Holmes.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Messrs Veverka and Leske, Ms Hatt, and Dr Holmes reported receiving grants from the National Eye Institute, Research to Prevent Blindness, and the Mayo Foundation. Dr Iezzi reported serving as a paid member of an advisory board for Alcon Surgical. No other disclosures were reported.
Funding/Support: Financial assistance for this study came from the Mayo Foundation, grant EY024333 from the National Institutes of Health, and an unrestricted grant to the Department of Ophthalmology, Mayo Clinic from Research to Prevent Blindness.
Role of the Funder/Sponsor: The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Create a personal account or sign in to: