A randomised comparison of bilateral recession versus unilateral recession–resection as surgery for infantile esotropia
=========================================================================================================================

* J-R Polling
* M J C Eijkemans
* J Esser
* U Gilles
* G H Kolling
* E Schulz
* B Lorenz
* P Roggenkämper
* V Herzau
* A Zubcov
* M P M ten Tusscher
* D Wittebol-Post
* G C Gusek-Schneider
* J R M Cruysberg
* H J Simonsz

## Abstract

**Objective:** Infantile esotropia, a common form of strabismus, is treated either by bilateral recession (BR) or by unilateral recession–resection (RR). Differences in degree of alignment achieved by these two procedures have not previously been examined in a randomised controlled trial.

**Design:** Controlled, randomised multicentre trial.

**Setting:** 12 university clinics.

**Participants and intervention:** 124 patients were randomly assigned to either BR or RR. Standardised protocol prescribed that the total relocation of the muscles, in millimetres, was calculated by dividing the preoperative latent angle of strabismus at distance, in degrees, by 1.6.

**Main outcome measure:** Alignment assessed as the variation of the postoperative angle of strabismus during alternating cover.

**Results:** The mean preoperative latent angle of strabismus at distance fixation was +17.2° (SD 4.4) for BR and +17.5° (4.0) for RR. The mean postoperative angle of strabismus at distance was +2.3° (5.1) for BR and +2.9° (3.5) for RR (p = 0.46 for reduction in the angle and p = 0.22 for the within-group variation). The mean reduction in the angle of strabismus was 1.41° (0.45) per millimetre of muscle relocation for RR and 1.47 (0.50) for BR (p = 0.50 for reduction in the angle). Alignment was associated with postoperative binocular vision (p = 0.001) in both groups.

**Conclusions:** No statistically significant difference was found between BR and RR as surgery for infantile esotropia.

Infantile esotropia (prevalence 0.2–0.35%)1 is usually corrected by posterior relocation of the insertion of the medial rectus of both eyes (bilateral recession, BR) or unilateral recession of the medial rectus muscle combined with a resection of part of the tendon of the lateral rectus muscle (recession–resection, RR).2 3 A good surgical result is a slight esotropia, because overcorrection into exotropia may lead to double vision.3–6 Complete restoration of binocular vision or stereopsis is rare.3–7 To date, there have been no randomised controlled trials to compare BR with RR.8 9 In a recent Cochrane review on interventions for infantile esotropia, it was concluded that the preference of BR or RR is a subject of discussion, and further research was necessary.2

The reduction in the angle of strabismus achieved by surgery depends primarily on the distance of relocation of the medial or lateral rectus muscle.10–18 The degree of alignment is subject to a large variation.19 Factors contributing to this variation are variation of the preoperative angle of strabismus, variations of measurements of the angle of strabismus, variation of the actual relocation of the muscle insertion during surgery, variation in anatomy, mechanical properties of eye muscles and other unknown parameters.3 6 13 16 19

We examined whether the reduction in the angle of strabismus was more predictable with either surgical technique, assuming that alignment can be more easily achieved when a particular surgical technique gives a more predictable result. Primary outcome was alignment assessed as the variation of the angle of strabismus after surgery, a large variation indicating an unreliable surgical technique.

## METHODS

### Design and patients

We performed a controlled, randomised, multicentre trial and recruited patients from 12 university clinics in Germany and The Netherlands. Approval was obtained from the Medical Ethics Committee at the Erasmus MC and confirmed by the local ethics committees of the participating clinics. Written informed consent was obtained from both parents of the children, and one parent in case of single parenthood. Inclusion of study subjects occurred over a period of 48 months. Eligible were all children aged 3 to 8 years with a normal psychophysical development and onset of esotropia before age 1 year, who visited one of the clinics during the study period. The decision not to include children under the age of 3 years was made because the angle of strabismus could not be accurately examined in very young children. Exclusion criteria were: previous strabismus surgery, an angle of strabismus larger than 24° or smaller than 10°, any binocular vision, more than one line logMAR acuity difference between the two eyes, hypermetropia >+6 dioptres or myopia >−3 dioptres, up- or downshoot in (25°) adduction >8°, V-pattern (esotropia measured in 25° up- and downgaze) >8°, A-pattern >5° and manifest vertical strabismus >4°. Cases with significant convergence excess with an angle of strabismus at near fixation more than 1.5 times as large as the angle at distance fixation were excluded.

### Study procedures

Procedures for orthoptic examination and surgery were protocolised. Each participating centre was visited by the research orthoptists who gave instructions on the procedures before and twice during the study. The preoperative orthoptic examination consisted of the following procedures. Refractive error was determined by retinoscopy and fully corrected (under correcting hypermetropia by no more than half a dioptre) before the angle of strabismus was determined. If present, amblyopia was treated before entry in the study. The measurement of the manifest and latent angle of strabismus was performed by prism cover–uncover test and alternating prism cover test at near with a fixation object, and at distance with a fixation light. A prism-adaptation test was allowed for a maximum of 4 h, with a maximum of three prisms in increasing strength. After the final preoperative examination, the child was randomly assigned by permuted block randomisation to either BR or RR.

Postoperatively, two orthoptic examinations were performed: immediately after the bandage was taken off on the day after surgery, the angle of strabismus was estimated by observing the corneal light reflex at 1 m distance. At 3 months, an extensive orthoptic examination was performed, used for final analysis. The angles of strabismus at near and distance fixation were measured. The angle of strabismus at distance fixation was also measured in 25° right and left gaze, to detect incomitance of the angle of strabismus. Binocular vision was examined at near and distance fixation by means of Bagolini striated glasses. Retinoscopy was repeated, and if the spherical equivalent differed by more than 0.75 dioptres, new glasses were prescribed, and orthoptic examination was performed 6 weeks later.

### Surgical techniques

The consensus surgical protocol contained regulations for the surgical plan and for the surgical technique. The total relocation of the two muscles in millimetres for both was calculated as follows: the preoperative latent angle of strabismus at distance (in degrees) divided by 1.6. The preoperative latent angle of strabismus measured the day before surgery at distance fixation was the basis for the distance of relocation of the operated muscles. The study committee, consisting of the participating paediatric ophthalmologists, determined this procedure, including the fixed ratio of 1.6 mm per degree of angle of strabismus and the use of the angle measured the day before surgery. The relocation of the muscles by recession or resection was divided equally over the two muscles in either BR or RR. All ophthalmologists had a minimum of 5 years’ experience with both recessions and resections. To evaluate the interoperator differences and its accuracy, photos were taken in approximately every tenth patient during two stages of the operation in either recession or resection: (1) after fitting the sutures through the muscle and (2) before closing the conjunctiva. A millimetre ruler next to the muscle was photographed with the eye.

### Main outcome measures

The primary parameter to assess whether the reduction in the latent angle of strabismus at distance fixation was more predictable with either technique was the variation of the latent angle at distance fixation, 3 months postoperatively. Secondary outcome was the variation of the mean reduction in the angle of strabismus, divided by the distance of muscle relocation. Another secondary outcome was the presence of binocular vision after surgery. Note that having demonstrable binocular vision prior to surgery was one of the exclusion criteria.

### Statistical analysis

For the main analysis—the assessment of a difference between the variance of the postoperative angles after BR and that after RR—the F test was used. For detection of a difference between the mean reduction in the angles of strabismus of surgery after BR and after RR, the F test was also used. The Student t test was used to assess differences between the RR and BR group with respect to the mean values of quantitative parameters of interest such as angles of strabismus, and the χ2 test was used for differences in qualitative parameters such as the presence or absence of postoperative binocular vision. Using retrospective data, we were able to establish a relationship between the postoperative angle of strabismus and the chance of a reoperation. Assuming normal distributions of postoperative angles, a clinically relevant reduction in reoperations of 30% would correspond with a reduction in standard deviation (SD) of 1.8°. The resulting F statistic would be 1.93 (assuming an SD of 5° in both groups on average). Inclusion of 120 patients would give the study 80% power at alpha = 0.05 to detect such an F ratio.

## RESULTS

During a period of 36 months, 124 consecutive children (average age 5.8) who met the eligibility criteria were recruited by the 12 participating clinics (figs 1, 2) The parents of three children refused to participate. In one child, the randomisation procedure failed. In one patient the angle of strabismus increased preoperatively, and for this reason the surgeon removed the patient from the study. One patient, who was assigned to BR, did not show up at the final orthoptic examination. The remaining 118 patients were treated according to permuted block randomisation assignment. Fifty-eight patients underwent RR, and 60 patients underwent BR. The baseline characteristics of the two groups are summarised in table 1. There were no statistically significant differences in demographic or orthoptic characteristics between the two groups.

![Figure 1](http://bjo.bmj.com/https://bjo.bmj.com/content/bjophthalmol/93/7/954/F1.medium.gif)

[Figure 1](http://bjo.bmj.com/content/93/7/954/F1)

Figure 1 
Eligibility criteria. BR, bilateral recession; RR, recession–resection.

![Figure 2](http://bjo.bmj.com/https://bjo.bmj.com/content/bjophthalmol/93/7/954/F2.medium.gif)

[Figure 2](http://bjo.bmj.com/content/93/7/954/F2)

Figure 2 
Numbers of patients undergoing operations in various clinics.

View this table:
[Table 1](http://bjo.bmj.com/content/93/7/954/T1)

Table 1 Baseline characteristics of randomised children who completed the final orthoptic examination

The mean total recession in the BR group was 10.2 mm (SD 2.0, range 6–15 mm), and the mean total amount of millimetres resected and recessed in the RR group was 10.4 (SD 2.1, range 6–15 mm). In 12 patients, the inferior oblique muscles were recessed bilaterally as well, seven in the BR group and five in the RR group. The mean interval between surgery and the last postoperative visit was 98 days in the BR group (range 46–201) and 103 days (range 43–435) in the RR group (p = 0.58). Dissociated vertical deviation (DVD) was reported in 10 cases. Latent nystagmus (LN) was reported in 26 cases. A combination of DVD and LN was reported in 19 cases.

### Outcome

Analysis of the BR and RR groups with inferior oblique muscles surgery showed no differences in postoperative horizontal angle. Therefore, the patients with additional oblique surgery were analysed together with those who underwent a standard BR or RR.

The mean latent angle of strabismus at near and distance were similar in both groups, 3 months postoperatively (fig 3). The mean latent angle at distance was +2.3° (SD 5.1) for BR, and +2.9° (3.5) for RR (p = 0.46). The mean latent angle at near was +5.0° (4.8) for BR, and +4.8° (4.1) for RR (p = 0.79).

![Figure 3](http://bjo.bmj.com/https://bjo.bmj.com/content/bjophthalmol/93/7/954/F3.medium.gif)

[Figure 3](http://bjo.bmj.com/content/93/7/954/F3)

Figure 3 
Postoperative latent angles for bilateral recession (BR) and recession–resection (RR).

To determine the postoperative variation of the angle of strabismus, the within-group variations of the BR and of the RR were compared. For manifest and latent angles, there were no statistically significant differences for the within-group variation tested with the F test (latent at near p = 0.83, and latent at distance p = 0.22). The mean ratio between the reduction in the latent angle of strabismus at distance fixation divided by the distance of muscle relocation was 1.41° (0.45) per millimetre for RR and 1.47° (0.50) per millimetre for BR. No statistically significant difference in this ratio was found between the BR and RR (p = 0.50, t test).

Eight patients treated with RR and nine patients with BR had an exotropia 3 months postoperatively. The presence of binocular postoperatively vision was found with the Bagolini striated glasses in 41.1% of BR, and in 35.7% of RR (p = 0.35). The variation of the latent postoperative angle at distance (SD) was 2.0° in patients with binocular vision as compared with 4.5° in patients without binocular vision (p = 0.001) (fig 4). We did not find any difference in incomitance between the BR and RR. Traction and scarring of the conjunctiva overlying the resected lateral rectus muscle were not reported as a complaint.

![Figure 4](http://bjo.bmj.com/https://bjo.bmj.com/content/bjophthalmol/93/7/954/F4.medium.gif)

[Figure 4](http://bjo.bmj.com/content/93/7/954/F4)

Figure 4 
Postoperative latent angles for Bagolini-negative and Bagolini-positive glasses.

As a quality measure, we analysed the adherence to the protocol. The guideline of dividing the angle of strabismus measured in degrees the day before surgery by 1.6° to obtain the distance of muscle relocation per millimetre was not strictly implemented in 33% of the cases. Reasons for deviating from the guideline were collected by means of a questionnaire among the participating paediatric ophthalmologists and were categorised as follows: (1) measurements of the angle of strabismus preceding the day before surgery had been considerably larger or smaller (in hindsight it would have been better to take the average of most or all preoperative measured angles of strabismus); (2) the angle at near fixation or the manifest angle at distance fixation had been considerably larger or smaller; or (3) there was significant hypermetropia.

We verified the results of the primary outcome measure by determining, as a secondary outcome measure, the variation of the reduction in the angle per millimetre of muscle relocation. We determined its coefficient of variation (CV). CV is a measure of dispersion of a probability distribution; it is defined as the ratio of the standard deviation to the mean. In those who had adhered fully to the protocol, the CV of the reduction in the angle of strabismus was 32% in RR and 34% in BR for ophthalmologists who adhered to the protocol. It was 36% in RR and 35% in BR for those who did not adhere fully to the protocol. Similarly, the CV of the ratio between the reduction in the angle of strabismus and the relocation of the muscle was 32% in RR and 34% in BR. In those who had not fully adhered to the protocol, the CV of the ratio was 36% in RR and 34% in BR. In summary, if paediatric ophthalmologists had not adhered fully to the protocol, (preoperative angle of strabismus divided by 1.6° per millimetre to obtain the distance of BR or RR) the CV of the postoperative angle and the CV of the mean reduction in the angle of strabismus were not better than in cases where the paediatric ophthalmologists had operated in full accordance with the protocol.

## DISCUSSION

In this controlled prospective randomised trial in children with infantile esotropia, we could not demonstrate any differences between BR and RR. Primary outcome was alignment assessed as the variation of the angle of strabismus after surgery, a large variation indicating an unreliable surgical technique in both cases. We could not demonstrate that either technique gave a more predictable result, that is, no differences were found between BR and RR for variation of the postoperative angle of strabismus. As a secondary measure, the variation of the ratio between reduction in the angle of strabismus and the millimetres of muscle relocation for BR and RR was determined. No differences were found.

We had set in this trial rigorous inclusion criteria to guarantee the enrolment of a homogeneous group of infantile esotropes. This enabled us to study the effect of surgery in a more precise manner but made recruitment difficult. An age of at least 3 years was a criterion in our study to allow for accurate orthoptic measurements. It is questionable that differences between BR and RR could be demonstrated for children under 3 years.

For both BR and RR, 38.4% of subjects had some degree of binocular vision, postoperatively assessed with Bagolini striated glasses test. These children had a significantly better ocular alignment, which may have been either a cause or a consequence of the gain of binocular vision.

## Footnotes

*   **Competing interests:** None.

*   **Funding:** The Netherlands Society for Prevention of Blindness, Haags Oogheelkundig Fonds, Stichting Blindenhulp and the Rotterdamse Vereniging Blindenbelangen supported this study.

*   **Ethics approval:** Ethics approval was provided by the Medical Ethics Committee at the Erasmus MC.

*   **Patient consent:** Obtained from the parents.

## REFERENCES

1.  Simonsz HJ, Kolling GH. Motor development and surgery for infantile esotropia. J AAPOS 2008;12:115–16
    
    [CrossRef](http://bjo.bmj.com/lookup/external-ref?access_num=10.1016/j.jaapos.2008.01.010&link_type=DOI) 
    
    [PubMed](http://bjo.bmj.com/lookup/external-ref?access_num=18423340&link_type=MED&atom=%2Fbjophthalmol%2F93%2F7%2F954.atom) 

2.  Elliott S, Shafiq A. Interventions for infantile esotropia. Cochrane Database Syst Rev 2005;(1):CD004917.
    
    

3.  Ansons AM, Davis H. Diagnosis and management of ocular motility disorders. 3rd edn. Malden: Blackwell Science, 2001.
    
    

4.  Birch EE, Stager DR, Berry P, et al. Prospective assessment of acuity and stereopsis in amblyopic infantile esotropes following early surgery. Invest Ophthalmol 1990;31:758–65.
    
    [Abstract/FREE Full Text](http://bjo.bmj.com/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NDoiaW92cyI7czo1OiJyZXNpZCI7czo4OiIzMS80Lzc1OCI7czo0OiJhdG9tIjtzOjI3OiIvYmpvcGh0aGFsbW9sLzkzLzcvOTU0LmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ==) 

5.  Birch E, Stager D, Everett M. Random dot stereoacuity following surgical correction of infantile esotropia. J Ped Ophthalmol Strabismus 1995;32:231–5.
    
    [PubMed](http://bjo.bmj.com/lookup/external-ref?access_num=7494159&link_type=MED&atom=%2Fbjophthalmol%2F93%2F7%2F954.atom) 
    
    [Web of Science](http://bjo.bmj.com/lookup/external-ref?access_num=A1995RJ93900005&link_type=ISI) 

6.  Von Noorden GK, Campos EC. Principles of surgical treatment. In: Binocular vision and ocular motility. 6th edn. St Louis: Mosby, 2002.
    
    

7.  Swan KC. Problems of exotropia. J Pediatr Ophthalmol 1965;2:25–32.
    
    

8.  Jeoung JW, Lee MJ, Hwang JM. Bilateral lateral rectus recession versus unilateral recess–resect procedure for exotropia with a dominant eye. Am J Ophthalmol 2006;141:683–8.
    
    [CrossRef](http://bjo.bmj.com/lookup/external-ref?access_num=10.1016/j.ajo.2005.11.021&link_type=DOI) 
    
    [PubMed](http://bjo.bmj.com/lookup/external-ref?access_num=16564803&link_type=MED&atom=%2Fbjophthalmol%2F93%2F7%2F954.atom) 
    
    [Web of Science](http://bjo.bmj.com/lookup/external-ref?access_num=000236621800011&link_type=ISI) 

9.  Arnoult JB, Yeshurun O, Mazow ML. Comparative study of the surgical management of congenital esotropia of 50 prism dioptres or less. J Pediatr Ophthalmol 1976;13:129–31.
    
    [PubMed](http://bjo.bmj.com/lookup/external-ref?access_num=1018189&link_type=MED&atom=%2Fbjophthalmol%2F93%2F7%2F954.atom) 

10. Scott WE, Reese PD, Hirsch CR, et al. Surgery for large angle congenital esotropia. Arch Ophthalmol 1986;104:374–7.
    
    [CrossRef](http://bjo.bmj.com/lookup/external-ref?access_num=10.1001/archopht.1986.01050150074030&link_type=DOI) 
    
    [PubMed](http://bjo.bmj.com/lookup/external-ref?access_num=3954636&link_type=MED&atom=%2Fbjophthalmol%2F93%2F7%2F954.atom) 
    
    [Web of Science](http://bjo.bmj.com/lookup/external-ref?access_num=A1986A365500020&link_type=ISI) 

11. Parks MM. Concomitant esodeviations. In: Duane’s clinical ophthalmology. Philadelphia: Lippincott, 1995.
    
    

12. Simonsz HJ, van Dijk B. Analysis of the dosage controversy in recess-resect and Faden surgery with the Robinson computer model of eye movements. Documenta Ophthalmol 1988;67:237–52.
    
    [CrossRef](http://bjo.bmj.com/lookup/external-ref?access_num=10.1007/BF00144278&link_type=DOI) 
    
    [Web of Science](http://bjo.bmj.com/lookup/external-ref?access_num=A1987M485300002&link_type=ISI) 

13. Simonsz HJ, Spekreijse H. Robinson’s computerized strabismus model comes of age. Strabismus 1996;4:25–41.
    
    [CrossRef](http://bjo.bmj.com/lookup/external-ref?access_num=10.3109/09273979609087734&link_type=DOI) 

14. De Decker W. Round-Table-Gespräch “Der operative Eingriff in der Schielbehandlung.” *Ber Dtsch Ophthalmol Ges* 1970:500–19.
    
    

15. Reinecke RD. Muscle surgery. *Duane’s clinical ophthalmology*. Philadelphia: Lippincott, 1995.
    
    

16. von Noorden G. Round-Table-Gespräch “Der operative Eingriff in der Schielbehandlung.” *Ber Dtsch Ophthalmol Ges* 1970:500–19.
    
    

17. Kaufmann H. Dosierung von Schieloperationen. In: Theorie der modernen Schielbehandlung. *Proc XIV Essener Fortbildung* 1984:292–9.
    
    

18. Simonsz HJ, Kolling GH, Unnebrink K. Final report of the early vs late infantile starbisums surgery study (ELISSS), a controlled, prospective, multicenter study. Strabismus 2005;13:169–99.
    
    [CrossRef](http://bjo.bmj.com/lookup/external-ref?access_num=10.1080/09273970500416594&link_type=DOI) 
    
    [PubMed](http://bjo.bmj.com/lookup/external-ref?access_num=16361188&link_type=MED&atom=%2Fbjophthalmol%2F93%2F7%2F954.atom) 

19. Schutte S, Polling JR, van der Helm FC, et al. Human error in strabismus surgery: quantification with a sensitivity analysis. Graefes Arch Clin Exp Ophthalmol 2009;247:399–409.
    
    [CrossRef](http://bjo.bmj.com/lookup/external-ref?access_num=10.1007/s00417-008-0961-x&link_type=DOI) 
    
    [PubMed](http://bjo.bmj.com/lookup/external-ref?access_num=18953557&link_type=MED&atom=%2Fbjophthalmol%2F93%2F7%2F954.atom)