The Survival Analysis after Symmetric and Asymmetric Surgery in Basic Intermittent Exotropia with Equal Dominance


 Background: Long-term surgical outcomes were compared between bilateral lateral rectus recession (BLR) and unilateral lateral rectus recession combined with medial rectus resection in the same eye (R&R) for therapy of basic type intermittent exotropia (IXT) with equal dominance.Methods: Eighty-six subjects(3–11 years old) with the basic IXT with equal dominance who underwent surgery were reviewed retrospectively, at least 18-month follow-up. including 48 patients underwent R&R surgery and 38 underwent BLR surgery at a single center. Surgical outcomes between groups were compared.Results: No statistical difference was detected between BLR group and R&R group at all intervals with the exception of the last examination, demonstrating a higher success rate and a lower undercorrection rate in the BLR group than R&R group at the last visit (76.3% vs 60.4%, P =0.04; 15.8% vs 35.4%, P =0.04 ). Postoperative recurrence continued in the R&R group, and it was lower in the BLR group. Stereoacuity and control showed overall improvement following both surgeries for basic intermittent exotropia. but this improvement had no statistical difference between groups(P>0.05). Cumulative probability of survival was lower in the R&R group than in the BLR group (P =0.027, log-rank test).Conclusions: The BLR group had a better outcome than R&R group at final visit, which may be due to the difference in the recurrence rate over time.

3 better for treatment of IXT still exists. 18 Some authors have reported a higher proportion of patients with a ideal outcome with R&R. 111516 while others have reported more success with BLR surgery. 917 Moreover, some experts and savants have revealed a notable exotropic drift over time with the R&R surgery. 89 Jeoung demonstrated that the unilateral R&R surgery resulted in a better outcome than BLR recession procedure in the patients with exotropia with a dominant eye. 7 But, in the patients with intermittent exotropia with equal dominance, convincing surgical procedure has not been established. As far as we know, this topic has not been studied alone. This study aimed to compare the surgical outcome of BLR versus R&R procedure in the patients with basic IXT with equal dominance.

Methods
Patients who underwent BLR or R&R surgery for basic type of IX(T) with equal dominance(the difference between deviation at distance and at near less than 10PD after 45min monocular covering )between May 2017 and March 2018 were reviewed. Medical informed consent document was signed by all subjects or their parents, and This study was approved by the Institutional Review Board of Children's Hospital of Nanjing Medical University.
Eye dominance was determined based on the results of repeated cover-uncover tests alternatively , at least three times per eye. During the tropical period of strabismus, which always showed a fixation shift to the dominant eye after the occluder was removed. otherwise, one eye have equal advantages with the other.
All surgeries were performed using the same operating table by the same surgeon (TABLE 1) under general anesthesia based on the largest distance deviation ever measured. The choice of surgical procedure was made by surgeon who wre not biased towards BLR or R&R in basic exotropia. R&R or BLR were performed after discussion with the patient and his or her parents.Meanwhile,the three decides which eye to perform the R&R operation on.
Patients had a history of systemic anomaly, a history of squint or other ocular surgery, amblyopia, and/or restrictive or paretic strabism were excluded.
Retrospectively reviewed the patient's medical records and extract the age at the time of operation, gender, best corrected visual acuity, refractive errors, deviation angle preoperatively(PD) at near and distance, associated strabismus(oblique function disorder, dissociated vertical deviation [DVD],A-V pattern and vertical deviation), follow-up time, lateral incomitancy, and operation type performed (R&R/BLR). The difference of vision acuity between two eyes equal or greater than 2 lines was named amblyopia, a spherical or cylindrical difference between the two eyes greater than 1.50D was named anisometropia. the angle of exodeviation for primary gaze was more than lateral gaze by 5PD was named lateral incomitancy. the one who had oblique muscle overactions, A-V pattern, or DVD not requiring operative treatment was contained.
The whole involved subjects underwent a comprehensive ophthalmic examination before operation.
the prism -alternate cover testing was performed to measure the squint angle at distance(6 m) and near(1/3 m) in primary and lateral gaze, with accurate spectacle correction if necessary.
Postoperative deviations at postoperative 1 week, 1 month, 6 12 and 18 months to the last examination were collected. Patients with diplopia caused by postoperative overcorrection received full-time alternating patch treatment for four weeks until diplopia disappeared. If the esotropia does not decrease after the patch has been applied for 4 consecutive weeks, cycloplegic refraction was put on the agenda again; and hyperopia that exceeds 0.50D was corrected completely. Patients who exhibited constant esodeviation with hyperopia of less than 0.50D were prescribed base-out Fresnel prisms. Incorporating prism into ordinary glasses was prescribed when the prism had to be worn for several months. The strength of the prism glass must be changed several times to keep fusion. We changed the strength of the prism glasses based on the minimum deviation angle to maintain fusion, and recommended weaning the prism when orthotropia was achieved in patients without the prism.
Reoperation was performed for overcorrected subjects if esotropia of ≥15 PD persisted 6 months after operation.
Surgery results were divided into 3 categories: recurrence/undercorrection (exotropia/phoria 10PD ), success(esotropia/phoria ≤5PD to exotropia/phoria≤10PD ), and overcorrection (esotropia/phoria 5 PD ) according to postoperative deviation angle at distance. Reoperation was needed to treat 5 recurrent exotropia when the maximum angle of exodeviation was at least 15 PD or greater or the patient's fusion control was poor, in which the increase in the apparent phase of exotropia was noticed frequently by clinicians or parents. Sensory status was assessed by a nearby Titmus stereo test. Stereopsis of ≤80 seconds of arc was defined as good.
The control of deviation is assessed by using the Modified Newcastle Control Score (MNCS) 19 . The MNCS combines both subjective and objective methods of control into a simple grading system that could distinguish and quantify the various severity levels in IDEX. the total MNCS ranges from 0 to 9, MNCS 4 was a suitable surgical threshold.
STATISTIAL ANALYSIS SPSS 19.0 was used to conduct statistical analysis. Continuous data and categorical data were expressed as mean(SD) and counts(percentages) respectively. continuous data was compared using Chi square test for continuity correction and Fisher's exact test. Categorical data was compared using a independent t-test. K-M survival analysis and log-rank test were used to compare surgical outcome. p <0.05 was deemed statistically significant.

Preoperative Patient Characteristics:
Among 194 patients who underwent surgery, 108 were excluded, The reasons were as follows: patients with a dominant eye(80); true divergence excess and convergence insufficiency (10); insufficient follow-up period (10); severe amblyopia (5) Mean postoperative deviations at distance did not differ between groups at all intervals with the exception of 18 months and the last examination ( P =0.03 at 18 months; P =0.01 at the last vistit, independent t test). Mean deviations after surgery at near also didn't differ at postoperative 1 week and 1 month but was less exotropic in the BLR group than in the R&R group after that, with a statistical significance at 6 months,18 months and the last examination after surgery (P =0.03, P =0.02, P =0.01, independent t test). but, there was no significant statistical difference at 12 monts(P =0.08, independent t test)(FIGURE 1).
Postoperative outcomes demonstrated that the overcorrection rate decreased and the recurrence rate increased in each group with the time passing after operation. Surgical results at each postoperative follow-up time by 18 months after surgery were not different between the two groups (P 0.05, χ2 test). However, the BLR group was associated with a numerically higher success rate than R&R group  (TABLE 3).
Using Kaplan-Meier survival analysis, the estimate mean surgical failure time in the BLR group was 16.6 ± 0.9 months and in the R&R group was 14.8 ± 1.0 months. The cumulative probability of surgical cure in the R&R group was significantly higher than that in the BLR group. (P= 0.027, log rank test; FIGURE 2).
Postoperative Exodrift: The amount of exodrift between each visit did not differ between BLR group and R&R group (P 0.05, independent t test). The largest exodrift occurred between postoperative 1 month and 6 months at distance and near in both groups , and this was 3.75 △ ±4. The results showed that the MNCS scores variation was of no statistical difference between groups, In the BLR and R&R groups, the proportions of subjects with control of ≤ 3 at baseline were 24% and 25%, and it becames 79% (30 of 38) and 79% (38 of 48) at the 18-month visit. Fifteen patients in total postsurgery control score decreased compared with presurgery, 5 in the BLR group and 10 patients in the R&R group (TABLE 5) . 8 We studied the influence of surgical type on long-term surgical outcomes of patients with basic type intermittent exotropia with equal dominance. BLR group was associated with a numerically higher success rate than R&R group at the last visit (18.13 ± 1.2 months, 18.24 ± 1.1 months, respectively). This may be attributable to a difference in recurrence rate over time: the exodeviation and recurrent rate tended to increase more quickly in the R&R group than in the BLR group as the follow-up period became longer.

Discussion
There was no unitary viewpoint on the effectiveness of two surgical methods in treating basic type IXT. Wang Lihua reported that the success rate of R&R operation was significantly higher than that of BLR operation (R&R 85.1% vs BLR 65.8%, p = 0.04) after a mean follow-up of 15 months 11 . Besides, Chia 16 noted that subjects receiving R&R showed better outcome at 1 year follow-up (R&R 74.2% vs BLR 42.2%). However, Maruo and associates' report suggested that 17 BLR produced a better outcome at 4-year follow-up (success 66.7% with BLR vs 32.8% with R&R). In a recent multicenter, randomized clinical trial, long-term outcomes were compared between 101 patients who underwent BLR and 96 patients who underwent R&R to treat basic type IXT, significant differences were not observed in suboptimal surgical outcome by 3 years between the two groups. 20 But, to the best of our knowledge, this is the first study to explore the influence of surgical type on surgical outcomes of patients with basic type intermittent exotropia with equal dominance, and we found BLR had a better outcome than R&R at more than 18-month follow-up. R&R showed more exotropic deviation from sixth months at distant fixation, resulting in a higher recurrence rate and lower success rate. This is different from Jeoung 's research, he found that better surgical results were obtained with the unilateral R&R procedure. 5 Subsequent exotropic 8,9 21−23 shift with R&R was common, the reasons attributed for higher recurrence in the R&R group may be as follows. Firstly, a medial rectus resection may cause an initial tethering effect, which will result in initial success. But, noncomitance of lateral gaze could cause a gradual loss of fusion and exotropic shift. Secondly, the resected medial rectus tension exists for a long time can result in muscle stretching, which may cause a weakness of the tethering effect. Also, other factors, such as age at surgery, should be considered. Previously, S H Lim and associates 24 found that patients who underwent resection procedures with younger age tended to acquire a higher recurrence rate than patients with older age.
In diplopia patients, occlusion treatment was performed for 4 weeks after operation, and prism glasses were prescribed when patients with symptoms non improvement thereafter. Among these initial postoperative esotropia patients, 3 had prolonged diplopia. In all overcorrection cases, the MNCS score was greater than 4, and stereopsis was greater than 200 seconds of arc before surgery.
These results indicate that the postoperative overcorrection was related to the preoperative control ability and sensory factors.
The difference of initial postoperotive mean esodeviation was not found between the persistence of postoperative esotropia and transient postoperative esotropia. In our study, the persistence of postoperative esotropia was independent of the initial postoperative angle of deviation.We also found that postoperative esodeviation usually returns to orthophoria within 6 months, and its resolution may be related to postoperative wound healing, which usually takes 4 to 6 weeks, and exotropia drift, most occur in the first 6 weeks after exotropia operation 8,9 . The chances of restoring to orthophoria are small for more than 6 months, we consider a second surgical intervention for these patients to avoid the risk of loss of stereopsis and amblyopia in young children.
It is noteworthy that one patient in the R&R group and one patient in the BLR group who changed from orthophoria to overcorrection between postoperative 12 months and 18 months were put on conservative therapy. However, the two participants were both performed lateral rectus muscle advancement on the previous recessed eye as esodeviation becoming larger and larger.
Postoperative results were all satisfactory. Two children were both slight hyperopia without refractive correction before surgery.
Outcome assessments would be different because of a series of outcome measures reported, including sensory function (stereopsis), 7 motor alignment (angle of deviation) 6−17 , control or a combination of the above 20 . Our study mentioning "success" was determined based on motor standards with the deviation of esotropia/phoria ≤ 5PD and exotropia/phoria ≤ 10PD. Therefore, patients with exotropia/phoria > 10 PD can be classified as recurrence even they had well fusion control ability. So, recurrence of exotropia doesn't always mean reoperation and unsatisfactory outcome. The recurrence rate should be distinguished from the rate of reoperation or satisfactory level.
Our study used a conventional dose of surgery. Different surgical dose may produce different results.
We were encouraged to use a larger amount of recession than we used for BLR recessions (augmenting). 25−28 Although the augmented recession may decrease failure rates due to residual exotropia,It may do so at the cost of increased failure rates due to overcorrection. This study has some limitations. First, it was a retrospective study. Therefore, we cannot control all influential factors such as age at surgery, refractive error, amblyopia, fixation preference, and preoperative angle of deviations. Next, because of the chosen patients who were followed up for more than 18 months, ascertainment bias may exist. Patients with unsatisfactory outcomes might be followed up for a longer time, and those with satisfactory results might not return to the eye hospital, this could cause a higher recurrence rate. However, most patients included in our study were followed for at least 18 months.     Kaplan-Meier Survival curves were compared between the bilateral lateral rectus recession group (BLR group) and the unilateral recess-resect procedure group (R&R group) for intermittent exotropia with the equal dominance by using the log-rank test. Results there was a remarkable difference (P<0.05) in the cumulative probability of survival between the groups, showed that the survival probability of BLR group was better than R&R group.