Patients and Methods
A retrospective chart review was performed on all patients diagnosed as having Duane retraction syndrome and exotropia in primary position from January 2012 to December 2017. The study protocol was approved by the ethics committee of the L.V. Prasad Eye Institute and adhered to the tenets of the Declaration of Helsinki. All patients who underwent surgical correction for strabismus during the abovementioned period were included in the study. The main indications for surgical correction in patients with exotropic Duane retraction syndrome were primary position deviation greater than 15 prism diopters (PD), AHP of 15 degrees or greater, and/or significant overshoots in adduction. Exclusion criteria were prior strabismus surgery, history of ocular trauma, and incomplete records of preoperative and postoperative strabismus evaluation.
The recorded baseline data included age, gender, laterality of condition, best corrected visual acuity (BCVA) in each eye, refractive status, presence or absence of amblyopia and history of any previous strabismus surgery or trauma. Orthoptic findings were recorded for both preoperative and final postoperative visits.
In the current study, Duane retraction syndrome was subclassified clinically into three types based on the degree of asymmetry between motility limitation of abduction and adduction, apart from other characteristics of Duane retraction syndrome. No electromyographic study was done. Type 1 Duane retraction syndrome was diagnosed in patients who had marked abduction limitation with no and/or mild adduction limitation. Type 2 Duane retraction syndrome was diagnosed in patients with adduction limitation greater than abduction limitation, whereas type 3 was diagnosed in patients who had a similar limitation of abduction and adduction limitation.11
Complete ocular motility examination was performed, and a deficit of adduction or abduction was graded from −1 to −4.12 A −5 limitation of motility was defined as a limitation of adduction/abduction not reaching midline. The primary position ocular deviation (both horizontal and vertical) was measured using the prism bar cover test (when visual acuity was better than 20/100 in either eye) and the Krimsky test in patients with poor vision or who were uncooperative for cover tests. These measurements were taken with AHP corrected with a prism in front of the normal eye in unilateral cases and in front of the habitually deviated eye in bilateral cases.
The degree of upshoot/downshoot (overshoot) and globe retraction was recorded according to the grading system described by Kekunnaya et al.13 Overshoot (upshoot/downshoot) on attempted adduction was graded from grade 0 to grade 4. Grade 0 indicates no overshoot where a straight line drawn from the fellow eye pupil parallel to the intermedial canthal line bisects the pupil of involved eye. Grade 4 indicates cornea disappearing under the eyelid. Globe retraction was measured with the involved eye in the maximum adducted position. Then the central palpebral fissure height was compared to that of the fellow eye. Grade 0 indicates no narrowing and grade 4 indicates more than 75% narrowing.13
Additional findings noted were AHP (measured by goniometer while the patient was reading the distance vision chart), near stereoacuity (measured by Randot stereoacuity test), and the presence or absence of nystagmus or dissociated vertical deviation.
All surgeries were performed under general anesthesia and the fornix approach was preferred. Passive forced duction testing was performed under general anesthesia before each surgery in either eye and tightness of medial and lateral rectus muscles was graded from 0 to +4, with grade 0 indicating no restriction and grade +4 indicating an inability to move the eye past midline on forced duction testing. Forced duction testing was repeated again following lateral rectus recession.14
Bilateral lateral rectus recession was performed for exotropia of 20 PD or greater, whereas unilateral lateral rectus muscle recession in the affected eye was performed for exodeviations of less than 20 PD. In bilateral asymmetric lateral rectus recessions, a larger recession was performed in the unaffected fellow eye. The amount of lateral rectus recession performed was at the discretion of the surgeon, but it was guided by the amount of deviation in primary position and preoperative tightness of the lateral rectus and medial rectus muscles. Additional medial rectus muscle recession was performed in the affected eye when globe retraction was severe (> grade 3). The presence of overshoots in adduction was managed by Y-splitting the lateral rectus muscle with the two arms at least 10 to 14 mm apart. The Y split of the lateral rectus muscle was performed first and then the two arms of the lateral rectus muscle were fixed and/or sutured to the sclera at the desired measurement point for recession. Surgical success was defined as postoperative horizontal deviation within 8 PD of orthotropia, correction of AHP to less than 5 degrees, two-step decrease in overshoot, and globe retraction at the final postoperative visit.
Statistical analyses were performed using statistical software (version 10.0; STATA, StataCorp) and Microsoft Excel; Microsoft Corporation). The Shapiro-Wilk test was used to assess the normality of the data. The paired t test was used to compare the difference between preoperative and postoperative mean characteristics. A P value less than .05 was considered statistically significant.
During the study period, 73 patients diagnosed as having Duane retraction syndrome and exotropia in the primary position underwent strabismus surgery for various indications. Of these, 67 patients (91%) had unilateral exotropic Duane retraction syndrome. The mean age at time of surgery was 14.11 years (range: 2 to 33 years). Of the 67 patients with unilateral Duane retraction syndrome, the left eye was involved in 42 (69%) and the right eye was involved in 25 (37.3%). There was no gender preponderance (38 male and 35 female).42 None of the patients were amblyopic. Type 3 Duane retraction syndrome was the most common (56 patients [77%]), followed by type 1 (14 patients [19%]) and type 2 (1 patient [1.3%]). Two patients (2.6%) presented with synergistic divergence on attempted adduction. The main indication for the surgery was exotropia in primary position with AHP (90%). The most common compensatory head posture was face turn to the contralateral (normal) side, which was noted in 69 patients (94%). Overshoots in adduction were noted in 52 patients (71%). Of those, 42 patients (81%) had grade 2 or higher overshoot. Severe globe retraction of grade +3 and +4 was noted in 5 patients (7%); however, globe retraction of grade 2 or less was noted in 27 patients (37%) (Table 1).
Comparison of Mean Preoperative Clinical Characteristics Among Different Types of Exotropic Duane Retraction Syndrome
Distribution of Surgeries
Forty-eight patients (65%) underwent unilateral lateral rectus recession in the affected eye only, whereas bilateral lateral rectus recession was performed in 25 patients (34%). Of the 25 patients who underwent bilateral lateral rectus recession, 23 underwent asymmetrical recession and 2 patients underwent symmetrical lateral rectus recession. The differential recession of the medial and lateral rectus muscles of the affected eye was performed in 5 patients to alleviate severe co-contraction and globe retraction of grade 3 or more. Y-splitting of the lateral rectus muscle of the affected eye was performed to correct overshoots in 42 patients (56%) who also underwent simultaneous recession of the lateral rectus muscle for correction of horizontal deviation. No patient underwent contralateral lateral rectus recession for fixation duress, ipsilateral medial rectus resection alone, or ipsilateral lateral rectus periosteal fixation.
Improvement in Primary Position Deviation
The mean exodeviation in primary position reduced from 23 ± 14.04 PD (range: 5 to 90 PD) to 9 ± 8.7 PD (range: 0 to 50 PD) preoperatively, (P < .0001). The mean reduction in primary position deviation was 17 ± 12 PD (range: 13 to 75 PD). The mean preoperative primary position horizontal deviation was 17 PD (range: 5 to 45 PD) and 34 PD (range: 20 to 90 PD) in the unilateral surgery and bilateral surgery groups, respectively. Average postoperative primary position horizontal deviation improved to 4 PD (range: 0 to 28 PD) and 10 PD (range: 0 to 50 PD) of exotropia in the unilateral and bilateral surgery groups, respectively (Table 2). Mean duration of follow-up after surgery was 22.5 weeks (range: 8 to 209 weeks). At the last follow-up visit, 27 patients (40%) were orthotropic and surgical success was achieved in 74% of patients.
Preoperative and Postoperative Horizontal Deviation, Abduction Limitation, and Overshoot in Patients With Exotropic Duane Retraction Syndrome
Improvement in AHP
Sixty-four patients (88%) had significant AHP at the initial presentation. Of these patients, 36 (67%) had AHP of more than 15 degrees and underwent surgery. Postoperative AHP reduced to less than 5 degrees in 54 patients (82%).
Improvement in Overshoots
Among the 42 patients with grade 2 or higher overshoots, 41 underwent Y-splitting of the lateral rectus muscle. The mean postoperative horizontal deviation in this group was 4.42 ± 5.14 PD. Postoperatively overshoots improved two or more grades in 29 patients (71%) and disappeared completely in 18 patients (44%).
Mean preoperative and postoperative abduction limitation in the affected eye in all three types of exotropic Duane retraction syndrome was 2.26 ± 0.90 (range: 0 to 4) and 2.68 ± 0.90 (range: 0 to 4), respectively (P = .02). Adduction limitation was observed preoperatively in patients with type 2 and type 3 exotropic Duane retraction syndrome. Mean adduction limitation was 1.56 ± 1.06 preoperatively and improved to 1.1 ± 1.08 postoperatively. The mean reduction in adduction limitation in patients undergoing unilateral lateral rectus recession was 0.30 ± 0.85 (range: 1 to 2.50).
Five patients underwent simultaneous unilateral recession of the medial and lateral rectus muscles. Three of the patients had type 3 Duane retraction syndrome and 2 patients had type 1. The average globe retraction in this group was grade 2.6 preoperatively, which improved to 1.6 postoperatively. Three of the patients had intraoperatively 3+ tight medial rectus muscle on forced duction testing. There was no worsening of abduction or adduction in these patients postoperatively at their last follow-up visit. Surgical success was achieved in 4 patients.
Reoperation was performed in 14 patients (19%). The most common indication for a reoperation was residual exotropia followed by residual overshoots. The mean preoperative exodeviation in this group was 34 PD (range: 10 to 90 PD), which was larger than the average preoperative deviation of the entire study population. In this group, all patients had type 3 Duane retraction syndrome and 2 patients had synergistic divergence. Two patients in this group underwent simultaneous medial rectus recession for co-contraction along with the lateral rectus recession. The mean exotropia following the second surgery was 14 PD (range: 15 to 25 PD). All patients underwent reoperation within 1 year postoperatively. None of the patients required more than two surgeries to achieve surgical success by their last follow-up visit.
The prevalence of exotropic Duane retraction syndrome varies from 14% to 40% as reported in various populations. A prior study from our population reported that approximately one-third of all patients with Duane retraction syndrome were exotropic irrespective of Huber's classification.15 Although the higher prevalence (almost similar prevalence) is reported by some of the larger studies, the possible reason for exotropia in Duane retraction syndrome remains speculative. Unlike a paretic strabismus, in Duane retraction syndrome the lateral rectus muscle gets its innervation from a branch of the third nerve, and thus the lateral rectus muscle is not paretic and does have an inherent tone.16 Thus the amount of deviation in primary position possibly depends on the tightness of the lateral rectus muscle. Helveston also suggested that a tight lateral rectus muscle might also predispose to the greater incidence of phenomena such as globe retraction and overshoots.16
Surgical management of exotropic Duane retraction syndrome is challenging because all components of the syndrome (head posture, horizontal and vertical deviation, globe retraction, and overshoots) need to be tackled simultaneously. Unlike esotropic Duane retraction syndrome, transposition procedures are not feasible for the patients with exotropic Duane retraction syndrome. Few guidelines exist on the surgical management of exotropic Duane retraction syndrome. Unilateral lateral rectus recession is preferred for exodeviation of less than 20 PD and bilateral lateral rectus recession for deviation of more than 20 PD in primary position.17 For bilateral lateral rectus recession, many authors recommend asymmetrical lateral rectus recession, with the larger recession on the contralateral side because large ipsilateral lateral rectus muscle weakening would further worsen preexisting abduction limitation.18 On the contrary, a large contralateral lateral rectus recession might lead to fixation duress and increased tone to the medial rectus muscle in the affected eye, and it might worsen globe retraction and overshoots. Therefore, it should be avoided in patients with severe globe retraction or overshoots. In patients with severe globe retraction and overshoots, large exotropia and tight lateral rectus supramaximal recession of the ipsilateral lateral rectus muscle or its periosteal fixation can be performed.19 Differential recession of the ipsilateral medial and lateral rectus muscles may also be done for severe globe retraction, wherein a small medial rectus recession with a larger amount of lateral rectus recession is performed to correct exotropia.20 We have used either one or a combination of these techniques in our patients based on the degree of primary position deviation, severity of overshoots, and globe retraction. To the best of our knowledge, this is the largest series of surgical outcomes of exotropic Duane retraction syndrome.
A retrospective case series by Kubota et al21 described the surgical outcomes of 124 patients with Duane retraction syndrome, including 48 patients with exotropic Duane retraction syndrome. They reported a satisfactory ocular alignment in 83% of their patients.21 In our patients, surgical success was not only defined by the correction of horizontal misalignment, but also the correction of overshoots and AHP. In the current study, surgical success was achieved in 54 patients (74%), of whom half were orthotropic in primary position at their last follow-up visit. In their retrospective study of 67 patients with exotropic Duane retraction syndrome, Fouad et al22 found higher surgical success with a smaller angle of deviation. However, unlike in our series, a significant number of patients (n = 13) underwent unilateral muscle surgery with larger angle of deviation (> 25 PD) resulting in undercorrection and an increased reoperation rate.22 In the current study, only 1 patient with more than 25 PD underwent unilateral surgery and achieved a satisfactory outcome.
In the current study, unilateral lateral rectus recession was performed in 65% of patients and 35 patients (73%) achieved surgical success. Although there was worsening of abduction limitation in the affected eye postoperatively, it was not significant (no patients had diplopia in the affected direction). On the contrary, lateral rectus recession seemed to improve adduction limitation marginally in our patients. This suggests that the adduction limitation in patients with exotropic Duane retraction syndrome might be due to a tight lateral rectus muscle.
We performed a bilateral lateral rectus recession in patients with unilateral Duane retraction syndrome and an exodeviation greater than 20 PD. The ipsilateral (affected) eye received lateral rectus recession equal to or at least 1 mm less than that of the unaffected eye to prevent worsening of abduction limitation. However, the maximum amount of asymmetry in recession was 3 mm and the maximum contralateral lateral rectus recession was 7 mm, except in synergistic divergence cases. We did not perform contralateral surgery alone in any of our patients. None of the patients showed worsening of the co-contraction of the affected eye with our approach.
There is sparse literature on the surgical outcomes of patients with bilateral exotropic Duane retraction syndrome. Theodorou and Burke23 showed satisfactory alignment in 11 patients with bilateral exotropic Duane retraction syndrome over a decade of follow-up. However, in most cases, they performed unilateral rectus recession with conjunctival recession. In the current study, we operated on 6 patients with bilateral exotropic Duane retraction syndrome. The mean preoperative exotropia was 39.93 PD (range: 14 to 85 PD). Five patients underwent bilateral lateral rectus recession with a mean follow-up of 29 weeks (range: 6 to 104 weeks). One patient showed synergistic divergence in either eye on attempted abduction. All of our patients showed satisfactory surgical outcome (exotropia less than 8 PD in primary position) at their last follow-up visit.
The unique idea of splitting the lateral rectus muscle for correction of overshoots by reducing the leash effect of the tight lateral rectus muscle was first advocated by Jampolsky.24 Combining lateral rectus recession with Y splitting results in a significant reduction in overshoot and horizontal deviation at the same time.25 Rao et al26 performed Y-splitting of the lateral rectus muscle with recession in 10 patients who had Duane retraction syndrome with overshoots and reported elimination of overshoots postoperatively and preoperatively. Although simple recession of the tight lateral rectus muscle also reduces the overshoot, the result is better with Y-splitting, especially with an upshoot of grade 2 or more because the latter procedure removes the lateral rectus muscle from the crest of the globe. This was shown by Farid27 through the comparison of Y-splitting and isolated lateral recession for upshoot in exotropic Duane retraction syndrome.
In the current study, we performed the Y-splitting procedure in patients with overshoots of greater than grade 2. The mean overshoot in these patients was 2.63 (range: 2 to 4) and reduced significantly in 33 patients (86%). In patients with globe retraction and an overshoot of grade 2 or less, lateral rectus recession alone achieved a satisfactory result. However, a two-step reduction in overshoot and globe retraction was seen in only 22 patients (43%) with Y-splitting and lateral rectus recession. Thus, residual overshoot and globe retraction of grade 2 or more was seen in only in 13 patients (39%).
Periosteal fixation of the lateral rectus muscle has been proposed as an alternative procedure for severe globe retraction and overshoots in patients with exotropic Duane retraction syndrome. Sukhija et al28 reported outcomes of lateral rectus recession and Y-splitting versus periosteal fixation of the lateral rectus muscle in 15 patients with exotropic Duane retraction syndrome. Their study did not find any significant difference in postoperative horizontal deviation and overshoots between the two groups. They preferred lateral rectus periosteal fixation to Y-splitting and recession in view of surgical ease. However, preoperative and postoperative abduction limitation was not compared in their study. In the current study, no patient underwent periosteal fixation irrespective of the amount of preoperative horizontal deviation, other than the 2 patients with synergistic divergence.
Differential recession of the lateral and medical rectus muscles in the affected eye was first described by von Noorden.29 He found asymmetric recession of both horizontal recti equivalent to posterior fixation of the muscle in correcting overshoots and globe retraction. We suggest that unilateral recession of both horizontal muscles should be performed only in cases of severe globe retraction (> grade 3) and tight medial rectus muscle on forced duction test intraoperatively. In this situation, weakening both horizontal recti not only reduces the globe retraction, but also might reduce the bridle effect of the tight muscle and overshoots. It is suggested that the lateral rectus to medial rectus recession dose should be approximately 3:2 in orthotropic patients with Duane retraction syndrome to avoid any consecutive esodeviation or exodeviation; however, to tackle the primary position exotropia, we added to the amount of recession to the lateral rectus muscle. In our study, 5 patients underwent medial rectus recession along with differential (larger) lateral rectus recession in eyes with significant globe retraction. Four patients achieved surgical success. In addition, 3 patients showed two-step reductions in globe retraction at their last visit. However, recession of the lateral and medial rectus muscles compromises ocular rotation in both directions and limits the binocular field.
To preserve ocular rotations, ipsilateral lateral rectus periosteal fixation with vertical rectus transposition to the lateral rectus muscle insertion site has been suggested in such cases.30 Sharma et al31 compared periosteal fixation alone with combined periosteal fixation and partial vertical rectus transposition in 13 patients and reported better success in terms of abduction limitation with the latter procedure, whereas adduction limitation and binocular visual fields improved similarly in both groups. In the current study, we did not find any worsening of ocular motility postoperatively after simultaneous medial and lateral rectus recession, suggesting that this approach might be fairly useful in these patients.
Irrespective of type, exotropic Duane retraction syndrome is often associated with face turn to the normal side. This could be due to a tight lateral rectus muscle causing the affected eye to be in an abducted position to maintain good stereopsis. Snir et al32 showed improvement in motor alignment in 77% of patients following contralateral lateral recession in 11 patients with exotropic Duane retraction syndrome. In our group, 34 patients with unilateral exotropic Duane retraction syndrome with face turn to the normal side of more than 15 degrees underwent ipsilateral lateral rectus recession, of which 19 patients (56%) had 5 degrees or less of face turn postoperatively.
Overcorrections are common in unilateral/bilateral esotropic Duane retraction syndrome, especially in patients who underwent combined vertical rectus transposition with medial rectus recession.33 However, there is little available information about overcorrection or undercorrection rates after a single surgery for exotropic Duane retraction syndrome. Sharma et al31 found overcorrection in 2 of 6 patients who underwent only lateral rectus periosteal fixation. We did not observe any overcorrection in our patients following the first surgery25 In our study, 14 patients (19%) required reoperations for undercorrections, and none of the patients had consecutive esotropia.
Our study suffers from the inherent limitations of a retrospective study and the absence of a control group. We classified Duane retraction syndrome into the various types based on clinical observations alone because we did not perform electromyography due to technical difficulties. We did not consider preoperative stereopsis and fusion for surgical decision, although binocular vision testing was performed in all patients. There was incomplete data for stereoacuity comparisons before and after surgery. The main goal of surgical intervention was to rectify AHP and correct primary position deviation and overshoots, if any.
Despite the limitations and given the rarity of exotropic Duane retraction syndrome, the current study reports surgical outcomes of a large number of patients with exotropic Duane syndrome. Unlike other forms of strabismus, the surgical dose of lateral rectus muscle weakening in exotropic Duane retraction syndrome needs to be determined on a case-by-case basis. Intraoperative forced duction testing, innervation abnormalities of the medial and lateral rectus muscle of the involved eye, concomitant overshoots, globe retraction, and AHP need to be taken into consideration when planning surgery. Our study suggests that lateral rectus muscle weakening with or without Y-splitting gave a predictable result without significant worsening of abducting or adducting forces in our patients. In addition, asymmetrical recession of the lateral rectus and medial rectus muscles in the affected eye might be performed in patients with significant globe retraction/overshoot and tight medial rectus muscle.