Strabismus Surgery Outcomes for Internuclear Ophthalmoplegia with Exotropia

Abstract

Purpose:

Exotropia (XT) in internuclear ophthalmoplegia (INO) is a difficult problem to treat. The purpose of this study is to describe surgical approaches in treating INO and glean insight into various pre-operative factors that may affect outcomes for XT in INO.

Methods:

We conducted a retrospective review from 1/1/1995 – 5/1/2021 and identified seven patients with INO who underwent strabismus surgery for XT. Patient age, sex, etiology of INO, pre-operative alignment and sensorimotor exam, presence of diplopia, surgery performed, subsequent surgeries, use of adjustable sutures, post-operative alignment, presence of post-operative diplopia, presence of post-operative diplopia with use of prism correction, and length of follow-up were all collected. Initial surgeries undertaken included unilateral medial rectus (MR) plication and lateral rectus (LR) recession, bilateral medial rectus (MR) plications or resections, or bilateral MR plications combined with either unilateral or bilateral LR recessions.

Results:

Chart review yielded ten charts, however two were excluded due to manifest esotropia (ET), and one was excluded due to incomplete records. Seven total patients were used in final analysis. The cohort age range was from 29–79 years. Pre-operative horizontal distance alignment ranged from 35 to 95 XT with an average exodeviation of 67.8 ± 22.6 prism diopters (PD). Horizontal adduction deficit ranged from −1 to −4 and was present bilaterally in all patients. . A variety of initial surgical approaches were undertaken. After 2 muscle surgery, distance deviation had an average change of 57.3 PD. After 3 muscle surgery, distance deviation had an average change of 75 PD. After 4 muscle surgery, distance deviation had an average change of 60PD. Three patients required additional surgery for XT.

Time to follow-up ranged from 1 to 58 months. Horizontal distance alignment in primary gaze at latest follow-up ranged from 30 ET to 30 XT with an average of 0 (orthotropia) ± 16.0 PD. One patient had a consecutive esotropia of 30 PD, one had a persistent exotropia of 30 PD, and five patients were orthotropic at distance. All patients reported relief of diplopia in primary gaze at near and distance either with or without use of prism. Horizontal ductions improved to some degree in all patients.

Conclusions:

Horizontal rectus surgery can treat many cases of XT in INO. Surgeons should consider INO etiology and concomitant vertical deviations when considering surgery. The degree of pre-operative adduction limitation is another important factor, though did not always dictate final motor and sensory outcomes.

Introduction:

Internuclear ophthalmoplegia (INO) is a neurological condition characterized by a decreased ability to adduct one or both eyes that results from a lesion in the brainstem, specifically, the medial longitudinal fasciculus and surrounding subnuclei.¹ The etiology of INO is multifactorial, most commonly secondary to either demyelinating lesions in multiple sclerosis or stroke; however, other neoplastic, infectious, inflammatory, degenerative, and traumatic causes have been reported.^2,3 When bilateral and associated with exotropia (XT) in primary gaze, the condition is characterized as “wall-eyed bilateral internuclear ophthalmoplegia” (WEBINO). Patients suffering from WEBINO can often have disabling diplopia, adaptive head turns, or decreased quality of life from eye misalignment. Due to the rarity of the entity, there exists a paucity of data regarding surgical outcomes of correcting exotropia related to INO.

Various surgical approaches have been described to correct XT in INO, including bilateral medial rectus resect procedures, unilateral recess-resect procedures, botulinum toxin to lateral recti, and vertical muscle transpositions. There is no study comparing the efficacy of different surgical approaches. Furthermore, there is no specific study examining how various pre-operative factors, such as INO etiology, presence of vertical deviations, or presence of restricted muscles and timing of neurological injury, may affect these outcomes.

Small case series of strabismus outcomes for INO have previously been reported in the literature. (Table 1) Buckley and Elston described three patients with large angle exotropia secondary to INO from strokes, who underwent bilateral medial rectus muscle resections and lateral rectus muscle recessions resulting in restoration of binocular vision in primary gaze and “good cosmesis.”⁴ The study was limited by lack of reporting alignment outcomes and time frame for which measurements were taken. Successful outcomes for INO with large angle XT have been described in patients who had vertical rectus muscle transposition surgery.^5,6 Other studies have described variable motor success with botulinum toxin injections to the extraocular muscles.^2,7 In perhaps the most comprehensive study of sensory and motor outcomes in patients with INO and XT, Roper-Hall et al. describe eight patients who underwent a unilateral recess-resect procedure (seven patients) or bilateral LR recession (one patient) with favorable alignment outcomes.⁸ The authors reported extensively on additional motor and sensory outcomes, presenting clinical improvement in near point of convergence, adduction ability, abducting nystagmus, and ocular dysmetria.

Table 1:

Summary of major case series addressing surgery outcomes for WEBINO

Study	Number of patients	Intervention	Repeat surgical Interventions	Average Pre-Operative Distance Deviation	Average Post Operative Distance Deviation	Other Outcomes Measured	Follow up from most recent intervention
Adams et al 2009	3	Bilateral MR resections with w/ either unilateral o bilateral adjustable LR recessions	1 – 7 years post op for 50 XT	(40–64 XT range)	(3–18 ET range)	Presence of diplopia	3–12 months
Buckley et al 1997	3 (of 11)	Bilateral MR resections w/ either unilateral or bilateral adjustable LR recessions	None	Not reported	Not reported	Presence of diplopia,	6–12 months
Dawson et al 2002	9	7 patierrts- Botulinum Injections to LR (2 - bilateral, 5 unilateral) 2 patients - bilateral LR recessions and unilateral MR resection	Botox group - 3 required maintenance injections Surgery Group - None	Botox Group - 45 XT Surgery Group - 59 XT	Botox Group - 6 XT (8 ET - 20 XT) Surgery Group - 2 ET (4 ET - 0)	Presence of diplopia	Botox Group - variable for maintenance njections Surgery Group - 9 months
Murthy et al 2007	16	Botulinum toxin injections to LR	5 required maintenance injections 5 required occlusion 2 underwent surgery (not specified)	48.4 XT (25 – 90 XT range)	21.4 XT (16 ET - 50 XT range)	Presence of diplpoia, cosmesis, head posture (4), stereopsis (2)	variable for maintenance injections, not time frame not specified
Nathan and Donahue 2012	5	unilateral vertical muscle transposition to MR +/− unilateral or bilateral LR recesssion (4/5) and other eye MR resection (1/5)	1 - underwent LR recession (initially only transposition)	59 XT (30 – 90 XT range)	Orthotropia (0 – 2 XT range)	Presence of diplopia, adduction limitation	10 – 56 weeks
Roper-Hall et al 2008	8	Unilateral MR resection / LR Recssion (7/8), with plan for staged procedure; Bilatral LR Recession (1/8)	3 - underwent opposite eye MR Resection / LR Recession	61.-9 XT (25 – 100 XT range)	After 1 procedure: 20.4 XT (5 ET - 45 XT range) After 2 procedures: 11.4 XT (5 ET - 45 XT range)	Sensory tests: Worth 4 dot, stereoacuity, Presence of diplopia; INO signs: adduction limiation, ocular dysmetria/ adduction lag, abducting nystagmus, convergence, skew deviation, face turn;	Not reported

The objective of this study is to provide insight into surgical outcomes in INO and factors affecting these outcomes.

Methods:

The study was approved by our local Institutional Review Board and was conducted in compliance with the US Health Insurance Portability and Accountability Act of 1996 and in adherence to declaration of Helsinki.

This study is a retrospective chart review. Billing and diagnostic codes from medical records from our clinic were cross-referenced for cases with diagnosed “internuclear ophthalmoplegia” and “strabismus surgery” or “exotropia surgery” with a date range of 01/01/1995 – 05/01/2021. Inclusion criteria included patients above age 18 who were diagnosed with internuclear ophthalmoplegia who underwent strabismus surgery for exotropia. Exclusion criteria were patients below age 18 and patients with pre-existing strabismus unrelated to INO. WEBINO is an even more rare finding in children; therefore, patients younger than 18 were excluded to make the findings most relevant to adult strabismus surgeons.

The following information was extracted on chart review: patient age, sex, etiology of INO (if available), pre-operative alignment measurements, pre- and post-operative sensorimotor examination details, presence of diplopia, surgery performed, subsequent surgeries, use of adjustable sutures, post-operative adjustable suture adjustments, post-operative alignment in primary gaze, presence of post-operative diplopia, presence of post-operative diplopia with use of prism correction, and length of follow up. Adduction deficit was graded based on duction limitation as described by Scott and Kraft⁹ (duction of 0 = normal, 1 = rotate from midline to 75% of full rotation, 2 = rotate from midline to 50% of full rotation, 3 = rotate from midline to 25% of full rotation, 4 = rotate to midline but not into given field).⁹ Clinical evaluation and surgical intervention were performed by two different surgeons. The type and amount of surgery were determined on an individual case basis by the surgeon.

Results:

Ten total patients were identified. Upon review of charts, two patients had a clinical INO but given the extent of their brainstem lesion, had manifested with a concomitant CN VI palsy and esotropia (ET). These two patients were excluded from analysis. Another patient reportedly had a diagnosis of exotropia with INO, but surgical records could not be located, and was therefore excluded. Therefore, seven patients were included in the final data.

Table 2 shows the basic demographic data of our cohort. There were 4 women and 3 men, with ages ranging from 29–79. Three patients had an INO secondary to a stroke, one patient had multiple sclerosis, one patient had a dorsal midbrain hematoma (as a complication from a VP shunt for hydrocephalus), and two patients had clinical INO of unknown etiology (one presumed autoimmune related neuropathy, one presumed small stroke in the setting of normal neuroimaging, but ultimately not diagnosed).

Table 2:

Demographics and etiology of INO

Patient	Age	Sex	Etiology of INO
1	29	female	dorsal midbrain hematoma
2	57	male	unknown (presumed autoimmune)
3	58	male	stroke
4	69	male	stroke
5	64	female	stroke
6	79	female	unknown
7	47	female	multiple sclerosis

Two patients underwent bilateral medial rectus plication (patient 1 and 4), one underwent unilateral lateral rectus recession and medial rectus plication (patient 2), one underwent bilateral medial rectus plication and unilateral lateral rectus recession (patient 3), one underwent bilateral medical rectus resection and unilateral lateral rectus recession (patient 5), and two underwent bilateral medial rectus plication and bilateral lateral rectus recession (patients 6 and 7). Surgeons at our institution traditionally perform plication procedures, however, the patient 5 underwent resection to allow inferior transposition to help treat a vertical deviation. It is also noted that patient 5 underwent surgery in 2007, which pre-dated the description of the plication procedure. The choice between 3 and 4 muscles was mainly dependent upon the degree of pre-operative deviation. In only two patients were forced duction testing formally documented, both of which reported negative forced ductions. The other five did not mention any restricted muscles in the operative reports, and therefore was not a significant factor in surgical planning.

Table 3 shows the pre- and post- operative alignment and motility data for our patients. Pre-operative horizontal distance alignment ranged from 35 to 95 XT with an average exodeviation of 67.8 ± 22.6 PD. Near horizontal deviation ranged from 25 to 100 XT’ with an average exodeviation of 65.7 ± 27.3 PD. Two patients underwent bilateral medial rectus plication, one underwent unilateral lateral rectus recession and medial rectus plication, two underwent bilateral medial rectus plication and unilateral lateral rectus recession, and two underwent bilateral medial rectus plication and bilateral lateral rectus recession.

Table 3:

Patient pre- and post-operative alignment measurements and adduction deficits. Initial surgeries undertaken as well as alignment and motility measurements at post op month 1 and latest follow up

Patient	Pre-op horizontal distance alignment	Pre-op vertical distance alignment	Pre-op near alignment	Pre-op adduction OD	Pre-op adduction OS	initial Surgery	Post op horizontal alignment distance POM 1	Post op vertical alignment distance POM1	Post op alignment near POM1	Additional surgery	Time to additional surgery (months)	Distance horizontal alignment latest follow up	Distance vertical alignment latest follow up	Near alignment at latest follow up	Addition OD at latest follow up	Addition OS at latest follow up	Total length of follow up (months)
1	35 XT	6 LHT	25 XT	−3	−3	R MR plication 6mm L MR plication 6mm (address vertical)	12 ET	ortho	18 ET	No	N/A	30 ET	5 LHT	unclear	−1	−1	48
2	85 XT	ortho	50 XT	−4	−1	R LR Recession 9mm R MR plication 9mm	16 XT	4 RHT	not recorded	Yes, 1	28	ortho, (18 XT in left gaze)	ortho	ortho	−3	full	58
3	95 XT	ortho	100 XT	−2	−4	R MR Plication 7mm L MR Plication 9mm	ortho	ortho	ortho	No	N/A	ortho	ortho	ortho	full	−1	1
4	40 XT	5 RHT	40 XT	−1.5	−1	R MR plication 6mm L MR plication 6mm R IR plication 4.5mm (address vertical)	ortho	ortho	ortho (10 X)	Yes, 1	12	ortho (8 X)	ortho	20 XT	−1	−1	13
5	55 XT	8RHT	60 XT	−1	−1	RMR Resection 5mm L MR Resection 5mm R LR Recession 7mm (adj) Inferoplacement of R MR and R LR (address vertical)	ortho	ortho	ortho (10 X')	No	N/A	ortho	ortho	ortho (10 X')	full	full	1
6	90 XT	ortho	95 XT	−4	−4	R LR Recession 7.5mm R MR Plication 6mm L LR Recession 7.5mm L MR Plication 6 mm	45 XT'	10 LHT	50	Yes, 3	5,5,7	30 XT	5 LhT	40 XT'	−3	−3.5	20
7	75 XT	ortho	90 XT	−4	−3	R LR Recession 6.5mm R MR Plication 6mm L LR Recession 6.5mm L MR Plication 6mm	ortho (4 X)	ortho	ortho (14 X)	No	N/A	25 X(T)(control 2)	ortho	30X'	−2	−2	5

At post op month one, four patients measured orthotropic at distance. Two patients had a residual exotropia, and one patient measured an esotropia of 12 prism diopters.

Three patients required additional surgery, one underwent a Scott procedure for left lateral incomitance 28 months post initial surgery, one required bilateral LR recession one year after initial surgery, and one patient underwent 3 additional surgeries, including maximal MR resections and plications and a unilateral inferior and superior rectus transposition and ipsilateral lateral rectus botulinum toxin injection. The vertical rectus transposition procedure was performed due to a failure of traditional rectus surgery to improve adduction at all, and creation of an adducting force was felt to be required.

Average follow-up time ranged from 1 month to 58 months. Horizontal distance alignment in primary gaze at last follow up ranged from 30 ET to 30 XT with an average of 0 (orthotropia) ± 16.0 PD. Horizontal near alignment at latest follow up ranged from orthotropia to 40 XT’ (n=6, one patient did not have alignment at near evaluated) with an average exodeviation of 8.5 ± 16.0 PD. At latest follow up, one patient had a consecutive esotropia of 30 PD at distance, one had a persistent exotropia of 30 PD at distance (patient who underwent four surgeries), one patient had an intermittent exotropia of 25 PD with good control, and four patients were orthotropic at distance. Horizontal adduction deficit improved to some degree in all patients after surgery. Between both eyes for all patients, the average adduction deficit improved from approximately −2.6 to approximately −1.3 (numbers denote averages of adduction deficit grading scale as referenced in the Methods section).

A subanalysis was performed to quantify the amount of change in distance deviation after 2 muscle versus 3 muscle versus 4 muscle surgery at the one month post operative visit. After 2 muscle surgery (patients 1,2 and 4), distance deviation changed from 53.3 PD XT to 4 PD ET for a change of 57.3PD. Within the two muscle surgery group, unilateral recess/plicate (patient 2) left the patient undercorrected (16 XT at post op month 1) versus the bilateral plications (patients 1 and 3, who were 12 ET and ortho at post op month 1). After 3 muscle surgery (patients 3 and 5), distance deviation changed from 75 PD XT to 0 PD (orthotropia) for a change of 75PD. After 4 muscle surgery (patients 6 and 7), distance deviation changed from 82.5 PD XT to 22.5 PD for a change of 60PD.

Table 4 shows pre- and post-operative complaints of diplopia. All seven patients reported binocular diplopia prior to surgery. At post-operative month one visit after the surgery, six out of seven patients reported relief of diplopia with or without use of prisms (only one required prism correction). At latest follow up, all seven patients reported some degree of binocular single vision with or without use of prisms (three required use of prisms). The one patient that remained exotropic despite four surgeries tolerated 40 PD base-in Fresnel prisms over reading glasses and reported binocular single vision but required occlusion for distance viewing.

Table 4:

Pre- and post-operative complaint of diplopia

Patient	Pre op diplopia	Post op diplopia (w/o prism) POMI	Post op diplopia (w/ prism) POMI	Post op diplopia w/o prism (latest follow up)	Post op diplopia w/ prism (latest follow up)
1	yes	yes	no	yes	no
2	yes	no*	no	no*	no
3	yes	yes^	not specified	yes^	not specified
4	yes	no	no	yes^	no
5	yes	no	no	no	no
6	yes	yes	not tested	yes	no*
7	yes	no	no	no*	no

^{* =}intermittent / some relief with effort

^{^ =}None at distance but present at near

Discussion:

We found that patients with INO who underwent strabismus surgery overall had improved motor and sensory outcomes, albeit by varying degrees among individual patients. The variety of differing initial surgical approaches, coupled with the small amount of patients precluded any formal analysis on the relative success of one surgical approach over another. The large range of pre-operative exodeviation and adduction limitation further make any speculation of optimal surgical approaches difficult. However, the study further substantiates the importance of medial rectus strengthening procedures, which can contribute to adduction improvement in INO. All patients showed at least some improvement in horizontal adduction deficits after strabismus surgery. Previous authors stressed the importance of operating on the medial rectus muscle for improved adduction function.⁸ Similarly in our cohort, patients with large bilateral medial rectus strengthening procedures showed adduction improvement (patients 1,3,4, 5, and 7). Only one patient did not undergo a bilateral MR strengthening procedure (patient #2), and this patient later required a second surgery for a large XT in left lateral gaze, presumably due to reduced adduction contributing to significant lateral incomitance. However, the degree of pre-operative adduction deficit did not necessarily correlate with final motor and alignment outcomes. For example, patients with −4 adduction limitation in at least one eye (patients 2,3, 6 and 7), all demonstrated varying alignment outcomes. Also, the amount of medial rectus strengthening did not correlate with either final degree of adduction function or alignment. Patients 2 and 3 had supramaximal plications of 9mm on the MR corresponding to a −4 adduction limitation. While patient 3 had a significant improvement in horizontal ductions (−4 to −1), patient 2 required further surgery and still demonstrated a −3 adduction limitation (−4 to −3) in that eye. Patients 1 and 7 both had significant adduction limitation (−3, −3 and −4, −3 respectively), but had different alignment outcomes (30 ET versus 25 X(T) respectively). Patient 6 had the highest degree of adduction limitation and had suboptimal alignment outcomes and need for several additional surgeries, further indicating the importance of considering degree of adduction deficit in surgical planning. As an alternative to medial rectus muscle strengthening procedures, some authors have suggested that vertical muscle transposition procedures may be more effective if the INO is accompanied by a large angle XT and significantly reduced adducting saccades.⁵ While we are unable to make this observation in our cohort, this surgical strategy makes intuitive sense for patients with a high degree of adduction deficit.

Our subanalysis on amount of PD change after 2, 3, or 4 horizontal muscle surgery suggests an increased effect with three muscle surgery compared to 2 muscle surgery (average change of 57.3 PD after 2 muscle surgery versus change of 75 PD after 3 muscle surgery). However, 4 muscle surgery produced less of an effect than three muscle surgery. The small numbers of patients and outliers make generalizability and drawing any conclusions from these specific results limited.

Our study contributes to the limited literature on the topic by 1) suggesting that INO etiology may play a role in final alignment outcomes, and 2) demonstrating that surgically addressing vertical deviations in these patients can lead to satisfactory outcomes. Regarding INO etiology, two of the three patients that required additional surgery (patient 2 and patient 6) had an unknown etiology of their INO, suggesting that ocular misalignment in these cases may be less predictable than those with INO and XT caused by stroke. There have not been dedicated studies to determine the etiology of INO as a possible factor contributing to alignment outcomes, and our small number of patients precluded subgroup analysis. An interesting finding in our study was the treatment of concomitant vertical deviations while also correcting the XT. Often in cases of large angle XT, small vertical deviations are expected and theorized to arise from the compartmentalization of horizontal muscle fibers¹⁰. Many surgeons may elect to primarily treat the horizontal deviation in these cases. However, vertical deviations in INO with XT may present with an additional level of complexity as they often result from neurological damage to vertical gaze control centers leading to a skew deviation. Three patients in our cohort had concomitant hyperdeviations (patients 1,4 and 5), and all required vertical prism to fuse in clinic pre-operatively. All three cases were addressed intraoperatively in a different manner. Patient 1 demonstrated a 6 PD right hypertropia treated with a left superior rectus recession of 3mm. Patient 4 had a 5 PD right hypertropia treated with a right inferior rectus plication of 4.5mm. Patient 5 had inferior placement of right medial rectus and lateral rectus. The specific choice of vertical muscle was surgeon dependent. All three patients demonstrated no post operative vertical misalignment. These findings suggest that various procedures can be used to successfully address vertical misalignment, and should be considered in the surgical plan if fusion requires vertical prism correction pre-operatively.

Patients 1 and 6 were two major outliers in our cohort with suboptimal final outcomes. Patient 6 required four surgeries (which included maximal medial rectus plications, lateral rectus recessions, medial rectus resections, one superior rectus and inferior rectus transposition surgery, and botulinum toxin injection to the right lateral rectus muscle) but remained exotropic of 40 PD at her final follow up. It is possible that because there was little to no adduction function noted pre-operatively, this may have contributed to the persistent exotropia despite multiple operations. Interestingly, the cause of this patient’s INO was unknown, but was presumed to be a small stroke not identified despite high quality neuroimaging. Patient 1 was esotropic at post-operative month one (12 ET), and remained so at her most recent follow up 58 months after surgery when she measured 30 ET at distance. She was diagnosed with a consecutive esotropia which she preferred to correct with prisms rather than undergo a second strabismus procedure. There was no evidence of a slipped muscle, which would not be expected in an overcorrection. Interestingly, she had a smaller pre-operative near deviation compared to distance deviation, which would not be expected in cases of bilateral adduction deficit. Being the youngest patient in the cohort, it is possible she demonstrated relative improvement or preservation of medial rectus tone after surgery relative to the older patients in the cohort, The two factors of younger age and near-distance disparity with lesser near angle are two possible factors that may contribute to overcorrection.

Our series has several limitations. Like previous case series, our series was limited by its retrospective nature and small number of patients, which did not allow for control of a variety of potential confounders. One of the major hindrances was the varied follow-up times, with two of our patients not having follow up after their post-operative month one visit. Other factors that may limit motor outcomes, but were not reported in our chart review, include the degree and timing of neurological insult as well as the presence of restricted or paretic muscles.

The combination of all of these factors stress the need for individualized patient planning, and make systematically studying INO with XT outcomes difficult. Furthermore, both the rarity of disease prevalence and its variable clinical course make prospective studies challenging. Future studies should seek to prospectively compare different surgical procedures with motor alignment and sensory outcomes and correlate these outcomes with salient surgical considerations of the disease such as INO etiology, degree of pre-operative adduction, timing of neurological insult relative to surgical correction, presence of vertical deviations, and presence of paretic or restricted muscles.