Treatment for central-peripheral rivalry-type diplopia (“dragged-fovea diplopia syndrome”)

Abstract

Purpose:

To report the effectiveness of treatments for central-peripheral rivalry (CPR)-type diplopia due to retinal misregistration.

Design:

Retrospective, interventional case series

Setting:

Adult strabismus clinic at a tertiary medical center

Patient population:

50 adults with retinal misregistration and CPR-type diplopia (minimum frequency of “sometimes” at distance and/or for reading), caused by epiretinal membrane (n=44) or other retinal disorders (n=6).

Interventions:

Included treatments were: Bangerter filter, tape, Fresnel prism, clear prism (incorporated into glasses or loose prism in office), iseikonic manipulation (using iseikonic lenses or contact lenses), MIN lens, or epiretinal membrane peeling (alone or in any combination). Not all treatments were trialed on all patients.

Main outcome measures:

Diplopia frequency evaluated using the Diplopia Questionnaire. Success was defined as “never” or “rarely” diplopic for distance and reading using the Diplopia Questionnaire, at an outcome examination as close as possible to 6 months.

Results:

Overall, 17 (34%, 95% CI 21% to 49%) of 50 patients were classified a success: Fresnel prism 4 (57%, 18% to 90%) of 7 patients; Bangerter filter 4 (14%, 4% to 33%) of 28, epiretinal membrane peeling 8 (44%, 22% to 69%) of 18 patients, and iseikonic manipulation (using a contact lens) 1 (4%, 0% to 22%) of 23 patients.

Conclusions and relevance:

CPR-type diplopia may be relieved in some patients using non-surgical treatment options of Fresnel prism or Bangerter filter. ERM peeling was surprisingly successful and should be considered.

INTRODUCTION

Epiretinal membranes (ERM) and other maculopathies can cause retinal misregistration resulting in central-peripheral rivalry (CPR)-type diplopia^1–3 (also known as foveal-peripheral rivalry,⁴ dragged-fovea diplopia,⁵ or macular diplopia⁶). CPR-type diplopia occurs when peripheral retinal images are fused but central retinal images are out of register one with another. This CPR type of diplopia caused by retinal misregistration is difficult to treat.^6–8 Previous reports have concluded that prism correction is almost always unsuccessful,^4–10 Bangerter filters or Scotch tape may be helpful^{5, 7, 11, 12} and that ERM peeling may be beneficial.³ Nevertheless, most previous studies have not rigorously evaluated diplopia frequency, or used standardized outcome criteria, making it difficult to determine the success of different treatment options. The aim of the present study was therefore to evaluate the effectiveness of non-surgical treatments for CPR-type diplopia, and to re-evaluate the effectiveness of ERM peeling in a larger cohort, utilizing the patient-reported Diplopia Questionnaire¹³ to standardize outcome assessment.

Materials and Methods

Approval for this retrospective study was obtained in advance from the Institutional Review Board at Mayo Clinic, Rochester, MN for retrospective data collection and analysis. All procedures and data collection were conducted in a manner compliant with the Health Insurance Portability and Accountability Act. All research procedures adhered to the tenets of the Declaration of Helsinki.

Patients

We retrospectively identified 50 consecutive patients seen in an adult strabismus practice (JMH) at a tertiary medical center, with a diagnosis of CPR-type diplopia in the context of retinal disease. As in previous studies,^{1, 2} we defined CPR-type diplopia as binocular diplopia in the presence of retinal misregistration where other barriers to single vision had been addressed. Care was taken to exclude monocular diplopia as a cause of the patient’s symptoms. Retinal misregistration can cause CPR-type diplopia^{1, 2, 4–6} but may also cause aniseikonia^{2, 5, 10} and/or metamorphopsia.^{2, 4, 5} In the present study, diplopia was considered to be of CPR type if there was evidence of retinal misregistration based on a positive central-peripheral rivalry response on the optotypeframe test,¹ but also if there was ≥5% aniseikonia by New Aniseikonia Test¹⁴ and/or any metamorphopsia by M-Charts.¹⁵ The optotype-frame test¹ was performed in all 50 patients and found to be positive for retinal misregistration in 46 (92%) of 50. Of the 4 patients demonstrating no retinal misregistration by optotype-frame test, 3 had evidence of retinal misregistration by New Aniseikonia Test¹⁴ (≥5%), and one by M-Charts¹⁵ (not all patients had aniseikonia and metamorphopsia testing). Some included patients had coexisting true strabismus, but treatment for the strabismic component had been optimized and residual diplopia was determined to be CPR type.

The first included examination was the visit at which CPR-type diplopia was first diagnosed in the adult strabismus clinic and a new trial of treatment started. All patients underwent full orthoptic evaluation and completed the Diplopia Questionnaire¹³ at each strabismus clinic examination (full questionnaire and scoring available at www.pedig.org). The Diplopia Questionnaire provides a patient-reported evaluation of diplopia frequency (“never,” “rarely,” “sometimes,” “often,” or “always”) in each of seven gaze positions.¹³ For the present study, we included patients rating diplopia as “sometimes” or more for straight ahead distance or for reading.

Previous or current treatment for CPR-type diplopia was allowed provided the patient met inclusion criteria for diplopia frequency (in any current treatment) and was undergoing a new trial of treatment. We included patients regardless of coexisting eye disease or level of visual acuity since our aim was to evaluate the effectiveness of treatment across the spectrum of patients with CPR-type diplopia.

Treatments

Treatments evaluated (in isolation or combination) were: prism (Fresnel, incorporated, or loose), Bangerter filter (covering the entire lens), translucent tape (typically Scotch Satin Tape™ 3M, allowing the patient to experiment with segmental occlusion at home), iseikonic manipulation (iseikonic lenses [range 1% to 5%] in office or ground into spectacles, or using contact lenses), MIN lens. ERM peeling was included as a treatment since we have previously found that some patients show improvement in CPR-type diplopia following ERM peeling surgery.³ ERM peeling was evaluated as a potential treatment for diplopia, even if performed prior to the first strabismus clinic examination, provided there was clear evidence of diplopia and retinal misregistration prior to peeling surgery. Trials of treatment were not fully standardized therefore not every patient had a trial of every treatment.

Individual treatments for CPR-type diplopia were either trialed “at-home” or “in-office.” Patients could be declared failure for a specific treatment following an in-office trial, but could not be declared a success without an at-home trial. For at-home treatments, diplopia was reevaluated using the Diplopia Questionnaire¹³ at a follow-up examination as close as possible to 6-months (minimum 3 weeks) after treatment was prescribed.

Analysis

Success of a treatment was defined as “never” or “rarely” diplopic for both distance and reading, using the Diplopia Questionnaire at a follow-up examination (therefore assigned for at-home treatments only), with any treatment in place. For every episode of success all available subsequent examinations were evaluated to determine whether there was subsequent failure. We also noted whether success was associated with a new reduction in visual acuity in either eye by >0.3 logMAR. Failure was defined as “sometimes” or more diplopic for straight ahead distance or for reading using the Diplopia Questionnaire, or a documented lack of improvement in-office. Treatments trialed in-office only that appeared successful were classified “indeterminate” since they had not been subjected to an at-home trial. For each patient, we recorded each at-home and in-office treatment episode and classified outcome as either success (at-home treatments only), failure, or indeterminate.

Outcomes were analyzed for broad treatment types and for specific treatment sub-types (see Table 1). To summarize specific treatment outcomes for each patient, if there was more than 1 episode of a given treatment (e.g., different Bangerter filter strengths trialed in different treatment episodes), the outcome was counted only once. To represent broad treatment outcomes only once for each patient, specific treatment sub-types were combined. For each of these analyses, a success outcome superseded failure and failure superseded indeterminate.

Table 1.

Treatments undertaken for central-peripheral rivalry-type diplopia, showing success rates in 50 patients with 138 specific treatment episodes.

Treatment type and specific sub-type		Success a N (%, 95% CI)	Failure a			Indeterminate a N (%)
			Overall N (%, 95% CI)	At- home	In- office
Type: Bangerter filter (n=28)		4 (14, 4 to 33)	17 (61, 41 to 78)	12	5	7 (25)
Sub-type b	Bangerter alone (n=21)	1 (5, 0 to 24)	14 (67, 43 to 85)	9	5	6 (29)
	Bangerter with Fresnel prism (n=4)	2 (50, 7 to 93)	2 (50, 7 to 93)	2	0	0 (0)
	Bangerter + incorporated prism (n=7)	1 (14, 0 to 58)	4 (57, 18 to 90)	4	0	2 (29)
	Tape (n=4)	0 (0, 0 to 60)	2 (50, 7 to 93)	1	1	2 (50)
Type: Fresnel (n=7)		4 (57, 18 to 90)	3 (43, 10 to 82)	3	0	0 (0)
Sub-type b	Fresnel prism alone (n=6)	3 (50, 12 to 88)	3 (50, 12 to 88)	3	0	0 (0)
	Fresnel + incorporated prism (n=1)	1 (100)	0(0)	0	0	0 (0)
Type: Clear prism (n=25)		0 (0, 0 to 14)	23 (92, 74 to 99)	5	18	2 (8)
Sub-type b	Incorporated prism (n=6)	0 (0, 0 to 46)	6 (100, 54 to 100)	5	1	0 (0)
	Loose prism in office (n=20)	N/A	18 (90, 68 to 99)	0	18	2 (10)
Type: Iseikonic treatment (n=23)		1 (4, 0 to 22)	19 (83, 61 to 95)	1	18	3 (13)
Sub-type b	Iseikonic lenses alone (n=22)	0 (0, 0 to 15)	19 (86, 65 to 97)	1	18	3 (14)
	Iseikonic + clear prism (n=4)	0 (0, 0 to 60)	3 (75 19 to 99)	0	3	1 (25)
	Contact lens adjustment (n=1)	1 (100)	0 (0)	0	0	0 (0)
Type: ERM peel with/without adjunctive therapy (n=18)		8 (44, 22 to 69)	10 (56, 31 to 78)	NA	NA	0 (0)
Sub-type b	ERM peel alone (n=10)	3 (30, 7 to 65)	7 (70, 35 to 93)	NA	NA	0 (0)
	ERM peel + Fresnel prism with/without incorporated (n=5)	2 c (40, 5 to 85)	3 (60, 15 to 95)	3	0	0 (0)
	ERM peel + clear prism (n=12)	4 c (33, 10 to 65)	6 (50, 21 to 79)	4	2	2 (17)
	ERM peel + Bangerter filter/tape (n=6)	0 (0, 0 to 46)	5 (83, 36 to 100)	3	2	1 (17)
	ERM peel + Bangerter/tape + clear prism (n=2)	0 (0, 0 to 84)	1 (50, 1 to 99)	1	0	1 (50)
	ERM peel + iseikonic lenses (n=2)	0 (0, 0 to 84)	1 (50, 1 to 99)	0	1	1 (50)
	ERM peel + iseikonic lenses + clear prism (n=3)	0 (0, 0 to 71)	3 (100, 29 to 100)	2	1	0 (0)
Type: MIN lens (n=2)		0 (0, 0 to 84)	1 (50, 1 to 99)	0	1	1 (50)

Abbreviations: ERM = epiretinal membrane; NA = not applicable

^aFor each specific treatment, each patient was counted only once, with any episode of success superseding any episode of failure, and any episode of failure superseding any episode classified indeterminate.

^bSome patients had more than one treatment episode within a broad treatment category.

^cOne patient undergoing epiretinal membrane peeling surgery was a success initially in Fresnel prism and then subsequently in incorporated prism.

The proportion of “success” and “failure” outcomes was calculated overall, and for specific and broad treatment types, with 95% confidence intervals.

RESULTS

Of the 50 patients 24 (48%) were male and for 46 (92%) race was reported as white. Median age was 70.5 years old (range to 51 to 84 years). For 44 (88%) of 50, retinal misregistration was caused by ERM; other causes were macular degeneration (n=5) and central serous chorioretinopathy with secondary choroidal neovascularization (n=1). The 50 patients underwent a total of 138 treatment episodes (Table 1). The outcome examination was a median of 5.8 months (range 3 weeks to 37.6 months) after the pre-treatment examination. Data for 12 (67%) of the 18 patients have been previously reported.³

Overall success

At the first included examination, the highest frequency of diplopia for distance or reading was “sometimes” for 11 (22%), “often” for 15 (30%) and “always” for 24 (48%). Overall, 17 (34%, 95% CI 21% to 49%) of 50 patients were classified as having successful relief of diplopia (“rarely” or “never” diplopic for both distance and reading); only 1 of whom had a coexisting reduction in visual acuity. Four (24%) of the 17 successes showed continued success over longer follow-up, 6 (35%) showed evidence of subsequent failure, and 7 (41%) have had no additional follow-up to date.

Bangerter filter or tape

Overall, 28 patients underwent treatment with Bangerter filter or tape, with a total of 36 treatment episodes (Table 1). The median strength of Bangerter filter used was 0.4 (range 0.8 to 0.1). Four (14%, 95% CI 4% to 33%) of 28 patients were classified a success at follow-up (Table 1): 3 subsequently failed treatment (2 increased to “sometimes” diplopic at distance; 1 had a subsequent retinal detachment with increase in diplopia to “always”). For 17 (61%, 95% CI 41% to 78%) of 28 patients, Bangerter filter or tape treatment outcome was classified as failure (Table 1).

Fresnel prism

Seven patients underwent an at-home trial of Fresnel prism (1 in combination with incorporated prism) (Table 1). All Fresnel prisms were applied vertically and prism strength ranged from 1 to 6 (median 3) prism diopters. Four (57%, 95% CI 18% to 90%) of 7 were classified a success (Table 1). Two of the 4 successes had no additional follow-up; for 1 there was evidence of subsequent failure (“sometimes” diplopic for reading). Three (43%, 95% CI 10% to 82%) of 7 were classified treatment failures (Table 1).

Iseikonic manipulation

Twenty-three patients underwent a trial of iseikonic treatment with a total of 27 treatment episodes. Median magnitude of aniseikonia using the New Aniseikonia Test was 8% (range 1% to 15%). The majority of treatment trials were in-office only, using iseikonic lenses, with no successes (Table 1). Nevertheless, 1 patient trialed a contact lens adjustment designed to address image-size differences and was classified as success (4%, 95% CI 0% to 22%; Table 1; no additional follow-up available).

ERM peeling

Eighteen patients underwent an ERM peeling procedure either historically (with evidence of diplopia and retinal misregistration preoperatively; n=3) or during the course of study follow-up (n=15) (Table 1). There were a total of 40 specific treatment episodes when considering ERM peeling alone or in combination with adjunctive therapies (Table 1). Overall 8 (44%, 95% CI 22% to 69%) of 18 patients were classified a success (Table 1). Three of the 8 successes were with ERM peeling alone, and 5 with prism added postoperatively (Table 1; 1 patient being a success with both Fresnel prism and subsequently with incorporated prism). One success had a reduction in VA of >0.3 logMAR (in the non-operated eye). Three of the 8 successes had no additional follow-up, 3 remained a success at subsequent follow-up examinations, and 2 became a failure subsequently (1 developed an ERM in the other eye and 1 increased to “sometimes” double for distance). All patients that were successful with prism following ERM peeling surgery had failed prism preoperatively.

Clear prism treatment

Twenty-five patients underwent a total of 26 treatment episodes using either loose prisms in the office (n=20) or incorporated prism (n=6) (Table 1). There were no successes with clear prism treatment alone (0%, 95% CI 0% to 14%) (Table 1).

MIN lens treatment

We had only 2 patients who underwent treatment with a MIN lens, one of whom (50%, 95% CI 1 % to 99%) was a failure at outcome and the other was classified as indeterminate regarding treatment outcome (apparent success in office but no office follow-up yet) (Table 1).

DISCUSSION

In this retrospective review of treatments for CPR-type diplopia associated with retinal misregistration, we found Fresnel prism was more successful than anticipated and Bangerter filter less so. These findings are important since treatment with Fresnel prism alone may not be considered for CPR-type diplopia, whereas Bangerter filters or tape are often a first line of treatment.⁵ In addition, we continue to find³ a surprising proportion of patients had successful resolution of diplopia following ERM peeling surgery.

There are a few, isolated reports of prism success in previous studies^{4, 16} but, in general, prism correction is not considered a viable treatment for CPR-type diplopia.^{5–7, 9, 10, 17} Nevertheless, this present study demonstrates that some patients may in fact respond favorably, specifically to Fresnel prism. We recognize that prism correction is non-intuitive as a treatment since it does not differentially affect central versus peripheral fields of vision. However, as described by Iacobucci et al,¹¹ vertical Fresnel prism may have the benefit of preventing a corrective peripheral motor fusion response when prescribed at a strength greater than the patient’s typically small vertical motor fusion capacity (true for 3 of our 4 Fresnel prism successes). In the study by Iacobucci et al¹¹ Fresnel prism was applied in combination with a Bangerter filter (to create a central scotoma). We hypothesize that Fresnel prism alone may be successful because it not only prevents peripheral motor fusion but simultaneously reduces visual acuity^18–20 and contrast sensitivity,^{18, 21} which may be sufficient to eliminate the diplopic image. We recommend that patients with CPR-type diplopia undergo at least an in-office trial of Fresnel prism, since for some this form of treatment may result in relief of diplopia. It would be of interest to evaluate contrast sensitivity, visual acuity and suppression in CPR-type diplopia patients improving and those not improving with Fresnel prism.

Previous studies have reported relatively good success of Bangerter filters (either alone,⁷ or in combination with prism¹¹) and also of Scotch Satin Tape™.⁵ However, few investigators report follow-up data or utilize a standardized outcome assessment. We found that, when using a patient-reported diplopia questionnaire for outcome assessment, most at-home Bangerter filter / tape treatment episodes were classified as failure. This apparently higher failure rate of Bangerter filters may simply reflect a more rigorous approach to outcome assessment and more stringent success criteria than used in previous studies. Alternatively, Bangerter filters may actually be less successful than previously thought: although they cause a mild to moderate reduction in visual acuity (depending on the strength used), they may have less impact on contrast sensitivity.²² Further study of associations between visual acuity, contrast sensitivity and relief of CPR-type diplopia will better direct treatment strategies.

In a recent prospective cohort study we reported 4 (33%) of 12 patients with CPR-type diplopia had resolution following ERM peeling.³ The present study included all 12 plus an additional 6 patients. Eight (44%) of the total of 18 patients met criteria for success with all but 2 maintaining success status at most recent follow-up. It is interesting that while 3 were successful with peeling alone, 5 were successful with incorporated prism added postoperatively, despite failure of prism preoperatively (duration of preoperative prism wear 6 to 18 months; unknown in 1 patient). These patients may have developed true small-angle strabismus due to prolonged CPR-type misalignment, or coincidentally developed non-paretic, small-angle hypertropia.²³ Alternatively, it is possible that maladaptation to preoperative prism occurred, resulting in a need for prism postoperatively.⁵ Against this, 2 of 8 successes were in prism preoperatively (6 and 96 months) but did not require prism for success postoperatively. Whatever the reason, our findings illustrate the importance of reattempting prism correction when diplopia persists after an ERM peel.

Manipulation of image size using iseikonic lenses or contact lenses is a less-frequently reported treatment for CPR-type diplopia. Similar to the experience of others^{10, 24} we found poor success of iseikonic lenses possibly because, unlike optical aniseikonia, retinally-induced aniseikonia is field dependent. It is also possible that the relatively small amounts of iseikonic correction that can be prescribed are inadequate when large amounts of aniseikonia are present or that, as previously suggested by Benegas et al,¹⁰ aniseikonia may be a contributing factor, but not solely responsible for diplopia. Our 1 patient with iseikonic treatment success involved using a contact lens to correct myopia in 1 eye, thus reducing image size differences. Further study directed at understanding aniseikonia-related symptoms and specific factors associated with iseikonic treatment success would be informative.

There are limitations to the present study. Treatments were not trialed systematically nor follow-up standardized across patients. Treatment using segmental occlusion was not fully standardized and therefore success of this treatment approach might be expected to be different from that reported by De Pool et al.⁵ In addition, our use of dichotomous success criteria may provide a somewhat narrow view of treatment effectiveness, in that it does not capture improvement (e.g., from “always” to “sometimes”). Some treatments (e.g. segmental occlusion leaving central vision unobstructed⁵) may be doomed to failure when using the Diplopia Questionnaire as an outcome measure and assessing diplopia straight ahead in the distance, but nevertheless such treatments may be useful to the patient. In future studies it may be instructive to evaluate “improvement” as well as “success.” It is important to note that treatment success is sometimes not long-lasting; in some cases this may be due to evolving retinal pathology (e.g., retinal detachment, ERM in the other eye) and studies with longer follow-up are needed to better understand success rates over the long-term. Nevertheless, even if short-lived, a period of diplopia relief may be worth pursuing. There are additional limitations when using the Diplopia Questionnaire in this patient population since coexisting symptoms such as blur, metamorphopsia and aniseikonia may make it difficult to distinguish diplopia. There is a need for additional patient-derived, patient-reported outcome measures that enable assessment of the various visual symptoms experienced by these patients. Future studies should incorporate retinal imaging, careful clinical evaluation, and condition-specific patient-reported outcome measures to aid understanding of associations between clinical and anatomical factors and treatment response. Further study evaluating the effect of treatments on contrast sensitivity would be helpful in understanding mechanisms of treatment effectiveness.

In patients with CPR-type diplopia due to retinal misregistration it is worth exploring non-surgical treatment options such as Fresnel prism, Bangerter filter, or tape. In cases where CPR-type diplopia is caused by an ERM, we suggest consideration be given to ERM peel surgery, mindful of the patient’s age, systemic and ocular comorbidities, and the everyday-life impact of their symptoms.