Changes in Lateral Comitance After Asymmetric Horizontal Strabismus Surgery

Importance  Asymmetric horizontal strabismus surgery is often performed to correct primary gaze alignment without considering the symptoms that may result from misalignment in the patient’s side gaze. Surgical choices influence alignment in side gaze and may contribute to functional and social deficits.

Objective  To identify the surgical procedures associated with changes of alignment in side gaze to help inform surgical planning for patients with horizontal strabismus.

Design, Setting, and Participants  The medical records of 1081 horizontal strabismus surgical procedures that were performed at Boston Children’s Hospital during a 2-year period were retrospectively reviewed. Only records with strabismus measurements recorded in the right and left gaze before and after surgery were included. Data analysis was conducted from September 1, 2012, through June 7, 2015.

Main Outcomes and Measures  Change in comitance (CIC), determined by measuring the horizontal comitance (the difference between right- and left-gaze strabismus measurements) before and after surgery.

Results  The review identified 569 patients who met the inclusion criteria. Of the 491 patients with comitant preoperative alignment, 59 developed postoperative incomitance, of whom 53 (89.9%) had asymmetric surgery. Of the 78 patients with incomitant preoperative alignment, 36 patients’ (46.2%) deviation had improved to comitance after surgery; 32 (88.9%) of these patients had asymmetric surgery. Asymmetric 2-muscle surgery had a median CIC of 4.0 while symmetric 2-muscle surgery had a median CIC of 1.5 (difference in CIC, 2.5; 95% CI, 2.0-3.0; P < .001). A CIC of 25 prism diopters or more was observed in 6 patients who underwent asymmetric surgery (0 with symmetric surgery). New postoperative incomitance was symptomatic in at least 17 patients (28.8%).

Conclusions and Relevance  Asymmetric strabismus surgery can treat incomitant deviations, but it can also create symptomatic incomitant deviations in patients who were previously comitant. Surgical planning should include consideration of the potential for CIC, including the potential for unsatisfactory appearance in side gaze. Patients with binocular vision will be sensitive to diplopia in any gaze direction; in such cases, the consequences of asymmetric surgery should be considered with particular care.

The goal of horizontal strabismus surgery is often to provide acceptable alignment in primary gaze, but lateral gaze incomitance can produce functional and social deficits that are not accounted for by primary gaze measurements. To our knowledge, the effect of various surgical strategies on the reduction of preoperative lateral incomitance and the production of postoperative lateral incomitance has not been well described in the literature. In this study, we evaluate the effect of horizontal strabismus procedures on lateral comitance (the difference in deviation between right and left gaze) to help inform surgical planning for patients with horizontal strabismus.

Box Section Ref ID

This retrospective data analysis was conducted with the approval of the Boston Children’s Hospital Institutional Review Board. The medical records of all horizontal strabismus procedures performed by 14 strabismus surgeons (D.G.H. and others) at Boston Children’s Hospital from July 1, 2010, through August 31, 2012, were reviewed. Only patients who had strabismus measurements in straight-ahead, right, and left gaze within 6 months before surgery and 1 year after surgery were included. The minimum follow-up period was 6 weeks. Excluded procedures included Kestenbaum operations for torticollis, muscle transpositions, and Faden procedures. Patients with congenital cranial dysinnervation disorders and complete cranial nerve palsies were also excluded; patients with partial cranial nerve palsies were not excluded.

After qualifying medical records were retrieved, each patient was assigned a study ID number. All relevant deidentified data were then extracted from the medical record for analysis, linked only by the unique study ID number. Considering that all data were deidentified and that no direct interaction with patients was necessary to perform the retrospective analysis, patients were not required to provide informed consent.

Data analysis was conducted from September 1, 2012, through June 7, 2015. Study data were collected and managed using REDCap (Research Electronic Data Capture),¹ hosted at Boston Children’s Hospital. Only the first procedure that met inclusion criteria for each patient was included even if multiple procedures occurred during the analyzed period. A procedure was defined as symmetric when the operation was on the same muscle(s) in both eyes and the amount of surgery performed between eyes did not differ by more than 0.5 mm. All other procedures were considered asymmetric, including bilateral recession or resection with more than 0.5-mm difference between eyes, unilateral recess-resect procedures, and 3-muscle horizontal surgery. The procedures were divided into 1-muscle, 2-muscle symmetric, 2-muscle asymmetric, 3-muscle, and 4-muscle surgery for analysis. Only 3 of the procedures (<1%) involved all 4 horizontal muscles; these were therefore excluded from further subanalysis. Adult patients were defined as those who were older than 12 years and pediatric patients were those who were 12 years or younger. Postoperative alignment in side gaze was measured in all patients 1½ to 3 months after surgery and again at the final postoperative visit (when more than 1 visit was available).

The main outcome measure was the change in comitance (CIC), or the difference between right and left gaze alignment, before and after surgery. We defined CIC as large if it was 15 prism diopters (Δ) or greater and very large if it was 25Δ or greater. Data were analyzed using the Mann-Whitney test, Kolmogorov-Smirnov test, and Kruskal-Wallis test for unpaired distributions, Wilcoxon match-pairs signed rank test for paired distributions, and linear regression for continuous variables. The differences in the medians of the distributions were reported with approximate 95% CIs for direct comparisons between 2 medians. Prism, version 6.0f for Mac (GraphPad Software), was used to conduct the calculations.

The medical record review identified 1081 patients, 569 of whom qualified for inclusion (median age, 24 years; range, 1-85 years) (Table 1). Of these, 262 (46.0%) were male and 282 (49.6%) were children. Approximately half the patients had esotropia (282 [49.6%]), and 108 (19.0%) had prior strabismus surgery. Simultaneous vertical muscle surgery was performed in 168 patients (29.5%). The mean time from preoperative measurements to surgery was 1 month, and the mean follow-up period was 9.4 months (range, 1.5-35.8 months).

Overall, asymmetric surgery had a higher median CIC and larger range of CIC than symmetric surgery (difference in CIC, 2.5; 95% CI, 2.0-3.0; P < .001) (Figure 1A). Subanalysis of the procedure type revealed that asymmetric 2-muscle surgery was associated with the largest CIC while symmetric 2-muscle surgery was associated with the smallest (P < .001) (Figure 1B). In this subgroup, the same trend was seen in patients who underwent first-time strabismus surgery (eFigure 1A in the Supplement) whereas 3-muscle surgery resulted in the largest CIC in the reoperation group (eFigure 1B in the Supplement). Analysis of patients who had 6 months or more of follow-up showed that asymmetric surgery still produced the highest CIC (difference in CIC, 3.0; 95% CI, 2.0-4.0; P < .001) (eFigure 2A in the Supplement), with 1-muscle, 2-muscle asymmetric, and 3-muscle surgery having similar CICs at the final follow-up (P = .67) (eFigure 2B in the Supplement). Adults had more CIC overall than children (difference in CIC, 2.0; 95% CI, 1.5-2.5; P < .001) (Figure 1C). Both adults and children had the least CIC after 2-muscle symmetric surgery (eFigure 3 in the Supplement). There was no meaningful correlation between the millimeters of surgery performed and the net CIC (mean [SD] slope, 0.036 [0.072]; 95% CI, −0.104 to 0.177; P = .62) (eFigure 4 in the Supplement).

Asymmetric surgery tended to create incomitance in patients with comitant deviations while correcting incomitance in those with incomitant deviations (Figure 1D). The preoperative deviation was comitant in 491 patients (86.3%) before surgery. Of these, new incomitant deviations developed in 59 patients, most of whom had asymmetric surgery (Figure 2). The remaining 78 patients had incomitant deviations before surgery. Of these, comitance was restored in 36 patients (46.2%), 32 (88.9%) of whom had asymmetric surgery (Table 2).

Four patients with incomitant deviations unexpectedly had comitance restored after symmetric surgery. The first was a 4-year-old girl with a history of exotropia who underwent bilateral lateral rectus muscle recessions of 5.0 mm, a right inferior oblique myectomy, and a left inferior rectus muscle recession. She had 10Δ of preoperative incomitance that then reduced to 2Δ of postoperative comitance. The second patient was a 7-year-old girl with partially accommodative esotropia who had 15Δ of preoperative incomitance (30Δ in right gaze and 45Δ in left gaze). She underwent 5.25-mm bilateral medial rectus muscle recessions with a 0Δ comitant result that persisted through her last appointment 14 months after surgery. The third patient was an 11-year-old girl with a history of alternating exotropia and 10Δ of preoperative incomitance who underwent bilateral lateral rectus muscle recessions of 7.25 mm with bilateral inferior oblique muscle recessions. Her postoperative incomitance was reduced to 6Δ. The final patient was a 38-year-old woman with a history of alternating exotropia who underwent bilateral lateral rectus muscle recessions of 7.0 mm with a left inferior oblique muscle anteriorization. She had 10Δ of preoperative incomitance that reduced to 2Δ after surgery.

For patients with more than 1 postoperative visit documented, there was no overall CIC from the first postoperative visit (6-12 weeks) to the last postoperative visit (difference in CIC, 0; 95% CI, 0-0; P = .75). However, differences were observed over time in subgroups in which the comitance changed after surgery. In comitant patients who developed incomitance after surgery, the amount of incomitance improved over time (difference in CIC, −4.25; 95% CI, 2.0-7.0; P < .001), while in incomitant patients whose comitance improved, the improvement in comitance remained stable over time (difference in CIC, 1.75; 95% CI, −4.0 to 7.0; P = .63) (eFigure 5 in the Supplement).

Many patients who had asymmetric surgery had large (≥15Δ) CICs. In the comitant group, 17 of 250 patients who had asymmetric surgery (6.8%) had a large CIC, including 2 with a very large (≥25Δ) CIC; in contrast, only 1 of 241 patients who had symmetric surgery (0.4%) had a transiently large CIC. This patient was a 7-year-old boy with estropia who underwent bilateral medial rectus muscle recessions with a right superior rectus muscle recession; he had developed 16Δ of incomitance at the first postoperative visit 2.3 months after surgery that completely resolved by 6 months after surgery. In the incomitant group, 12 of 70 patients who had asymmetric surgery (17.1%) had a large CIC, including 4 with a very large CIC. In all these patients, the incomitance was reduced. Of the 7 patients who had symmetric surgery, 1 patient (14.3%), the 7-year-old girl described in the Results section, had a large improvement in comitance. No patients with symmetric surgery had a very large CIC.

New-onset postoperative incomitance was bothersome to many patients. Of the 59 patients with new incomitant deviations, 17 (28.8%) had a persistent concern about the alignment in side gaze, including 15 with diplopia (secondary to esotropia) in 1 field of gaze and 2 with concerns about an unsatisfactory appearance (secondary to exotropia). The mean age was 42 years (range, 7-71 years). Sixteen of these patients (94.1%) had unilateral surgery. Twelve of the 17 patients (70.6%) had recess-resect procedures for exotropia. Most patients (11 of 17 [64.7%]) had stereopsis of at least 200 seconds of arc. In addition, 5 of the remaining 42 patients with new-onset postoperative incomitance described persistent diplopia, but it was not noted in the medical records whether diplopia was present in primary gaze, side gaze, or both. More patients who had surgery on the left eye (11 of 28 [39.3%]) reported having symptoms than those who had surgery in the right eye (5 of 20 [25.0%]) (Table 3).

All patients with double vision had diplopia in the direction of the eye that was operated on (eg, diplopia in left gaze after surgery on the left eye) and had documented duction deficits and esotropia in the diplopic direction of the gaze. With recess-resect procedures and single-muscle recessions, duction deficits were invariably toward the side of the recessed muscle. Adding a resection of the antagonist muscle seemed to exacerbate this effect because the ratio of symptomatic recess-resect procedures to single-muscle recession was 15.0 to 1.0, whereas the ratio in all asymmetric cases was 2.6 to 1.0.

The mean CIC was the same in the groups with and without simultaneous vertical strabismus surgery (difference in CIC, 0.50; 95% CI, 0.0-1.0; P = .21) (eFigure 6A in the Supplement). In the group who underwent simultaneous vertical and horizontal strabismus surgery, the CIC was larger if combined rectus and oblique muscle surgery was performed (P = .006) (eFigure 6B in the Supplement).

Lateral incomitant strabismus can be distressing to patients from a functional and social standpoint, yet we encountered a dearth of published information regarding the effects of asymmetric horizontal strabismus surgery on lateral gaze. In our practice, we encounter patients who developed unintended symptomatic strabismus in lateral gaze after asymmetric surgery, creating concern that many strabismus surgeons perform asymmetric surgery without recognizing the possibility of induced incomitance.

Several studies have examined unilateral muscle recession to correct comitant horizontal strabismus, but few have measured CIC. Single-muscle recessions have been recommended for esotropia^2-5 and exotropia.^6-8 Pollard and Manley² noted that 20% of patients who underwent unilateral medial rectus muscle recession for esotropia of less than 18Δ developed incomitance with overcorrection in the direction of gaze toward the muscle that underwent surgery. While others noted similar incomitance or limitations of ductions toward the muscle that underwent surgery, lateral incomitance was not measured in prism diopters and CIC was not assessed.^5,6,8,9

Yoon and Kim¹⁰ evaluated symmetric lateral comitance (a decreased angle in side gaze) in 155 patients who had surgery for intermittent exotropia. Their article focused on whether lateral incomitance of more than 5Δ—in particular, a reduced angle in side gaze in either direction compared with primary gaze—was a risk factor for poor outcome. Patients with a difference in comitance from right to left gaze before surgery were excluded. They found that the likelihood of induced lateral incomitance doubled in patients who underwent unilateral lateral rectus muscle recession compared with those who underwent bilateral lateral rectus muscle recession (P = .006).

Deacon and colleagues¹¹ prospectively evaluated the effect of unilateral surgery on the postoperative incomitance of 12 patients with exotropia. They measured comitance in terms of percentage as well as in prism diopters. They found that the surgical effect was 120% of that achieved in primary gaze in the direction of gaze toward the eye that underwent surgery but only 75% in the opposite direction. Overall, surgery had the most effect with gaze toward the eye that was operated on in 91% of the patients. Symptomatic diplopia persisted more than 6 months after surgery in 25% of cases. Unlike our study, Deacon and colleagues evaluated only patients who underwent unilateral horizontal rectus muscle procedures (recess-resect or single-muscle recession) for exotropia; they did not compare their results with those of symmetric surgery or evaluate patients with esotropia.

We found that lateral incomitance can be unintentionally produced in previously comitant patients using asymmetric surgery. Similar to the patients described by Deacon and colleagues,¹¹ at least 28.8% of our patients with new incomitant deviations had a persistent concern about a cosmetically or functionally unacceptable result in lateral gaze. Most symptomatic patients were adults, likely because children may be too young to express double vision, less likely to observe double vision in side gaze, and more able to cope with diplopia using suppression. Most of the symptomatic patients had stereopsis; this is expected because the primary symptom was diplopia, which requires some level of binocular function to observe. We confirm Deacon and colleagues’ observation that diplopia was disproportionately in the direction of gaze of the eye that underwent surgery.

Our results support the concept that asymmetric horizontal rectus muscle surgery can be a useful tool when treating lateral incomitance. In our series, 46.2% of incomitant deviations were successfully treated with asymmetric surgery, including single-muscle and recess-resect procedures. Many authors have published suggestions regarding the surgical treatment of incomitance, including Faden operations,¹² combined recession and resection on a single rectus muscle,^13,14 and unilateral rectus muscle recession.¹⁵ There is evidence to suggest that combined recession and resection of a single rectus muscle may also be effective at reducing lateral incomitance in patients who have no diplopia in the primary gaze.¹⁶ Our results suggest that asymmetric strabismus surgery may also improve lateral incomitance in some cases.

We found no predictable correlation between the amount of unilateral surgery and the amount of postoperative incomitance induced; therefore, no recommendation can be made regarding the amount of recession required to reduce incomitance.

While concomitant vertical surgery did not seem to be a risk factor for inducing CIC, there was a larger CIC noted in patients who had surgery on the oblique muscles and the vertical rectus muscles together with horizontal surgery. This change may be related to the tertiary horizontal actions of the vertical rectus and oblique muscles.

There are several limitations of this uncontrolled retrospective study. First, symptoms of incomitance were not assessed or recorded in a structured manner during clinic visits. For this reason, symptoms related to horizontal comitance may have been underreported. Even without prospective assessment of symptoms, many patients reported issues unprompted, revealing that this outcome is an important source of postoperative discomfort and disability. Second, the study evaluated strabismus procedures that were performed by a large number of surgeons using differing techniques. While in a small sample size this difference might cause difficulty extrapolating trends, our pooled data have a large number of measurements that allow for relationships to be identified. Finally, only 569 of 1081 patients identified met the inclusion criteria, primarily owing to a lack of preoperative or postoperative measurements of side gaze. This exclusion may have increased the representation of patients with symptoms in side gaze, artificially increasing the percentage of patients with preoperative and postoperative lateral incomitance. We do not believe, however, that it confounded the overall association of CIC with asymmetric surgery and with induced and reduced incomitance, especially given that a large number of patients in the study had either symmetric surgery or asymptomatic, asymmetric surgery.

This study of more than 500 patients (to our knowledge, the largest to ever evaluate CIC before and after strabismus surgery) demonstrates that asymmetric surgery, including 1- and 2-muscle surgery, can treat—or create—incomitant strabismus. Surgeons who are planning a procedure should assess the preoperative deviation in side gazes and consider the likely results not just in primary gaze but also in side gaze. It is especially important to anticipate CIC when planning surgery for adults with good binocular vision who will be dissatisfied with postoperative diplopia in side gaze despite a good result when looking straight ahead.

Submitted for Publication: March 1, 2015; final revision received and accepted June 23, 2015.

Corresponding Author: David G. Hunter, MD, PhD, Department of Ophthalmology, Boston Children’s Hospital, 300 Longwood Ave, Fegan Bldg, Fourth Floor, Boston, MA 02115 (david.hunter@childrens.harvard.edu).

Published Online: August 20, 2015. doi:10.1001/jamaophthalmol.2015.2721.

Author Contributions: Dr Graeber 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: Both authors.

Acquisition, analysis, or interpretation of data: Both authors.

Drafting of the manuscript: Both authors.

Critical revision of the manuscript for important intellectual content: Both authors.

Statistical analysis: Both authors.

Obtained funding: Both authors.

Administrative, technical, or material support: Both authors.

Study supervision: Hunter.

Conflict of Interest Disclosures: Both authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.

Funding/Support: Funding for this study was provided by the Children’s Hospital Ophthalmology Foundation.

Role of the Funder/Sponsor: Children’s Hospital Ophthalmology Foundation had no direct role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication.

Previous Presentations: This study was presented in part at the American Association for Pediatric Ophthalmology and Strabismus annual meeting; April 3-7, 2013; Boston, Massachusetts; and in full at the American Association for Pediatric Ophthalmology and Strabismus annual meeting March 25-29, 2015; New Orleans, Louisiana.