Article Text

Download PDFPDF
Recess–resect surgery with myopexy of the lateral rectus muscle to correct esotropia with high myopia
  1. Michael Gräf1,
  2. Birgit Lorenz2
  1. 1Department of Ophthalmology, Universitaetsklinikum Giessen and Marburg UKGM, Giessen, Germany
  2. 2Department of Ophthalmology, Justus-Liebig-University Giessen, Universitaetsklinikum Giessen and Marburg UKGM, Giessen, Germany
  1. Correspondence to Professor Michael Gräf, Department of Ophthalmology, Universitaetsklinikum Giessen and Marburg UKGM, Giessen Campus, Friedrichstr. 18, 35392 Giessen, Germany; michael.h.graef{at}


Background Esotropia due to high myopia can be caused by inferior shift of the lateral rectus muscle (LRM). Innovative surgical methods have been developed to elevate the muscle and, thus, augment its abducting force. However, their efficacy is not yet proven to exceed that of recess–resect surgery (RR).

Methods Data on high myopic esotropia were evaluated who received RR together with elevation of the LRM in the horizontal meridian by equatorial myopexy. Age, gender, axial length, amount of surgery, preoperative and 3 months postoperative strabismus angles and efficacy (mm/°) were analysed. Medians and ranges (minimum–maximum) are given.

Results The age of the 46 patients (37 females, 9 males) was 57 years (36–76). Axial length was 31.5 mm (26.0–35.6), total amount of RR 10.0 mm (5–24). Esotropia 16° (4–60) was reduced to 2° (−22 to 34), hypotropia from 3° (−3 to 30) to 0° (−8 to 18). Efficacy (eyes without prior surgery only) was 1.33°/mm (0.20–3.55).

Conclusions Geometrical considerations rather than empirical data yield some evidence of an augmenting effect of equatorial myopexy on RR. Efficacy of RR with LRM myopexy was higher in unilateral than bilateral high myopia. Strict indication for myopexy appears appropriate.

  • Muscles
  • Optics and Refraction
  • Treatment Surgery

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


In 1864, von Graefe described a specific form of convergent strabismus depending on myopia that typically occurred in young females.1 In the early 20th century, Bielschowsky reported on esotropia of myopes as a rare kind of convergent strabismus. He also mentioned the manifestation in adolescents and young adults.2 There were further descriptions of that and similar types of strabismus, but their specific causes remained obscure. Demer and von Noorden highlighted the possibility that esotropia and restricted motility associated with high myopia may be misdiagnosed as thyroid-associated orbitopathy.3 Since ocular motility is frequently reduced in this condition, neurogenic disorders, especially sixth nerve palsy, must also be differentiated.

Surgery for esotropia associated with high myopia was also a problem. Von Graefe recommended tenotomy of the medial rectus muscle assuming that its contracture was caused by too much near activity without wearing glasses.1 Bielschowsky presumed that the medial rectus muscles were downregulated by compensatory innervation. He preferred bilateral lateral rectus muscle (LRM) resection.2 Both methods did not prevail. Combined recess and resect surgery (RR) sometimes showed a weak effect, and was reported to induce or increase hypotropia. Kolling, by in situ photographs, demonstrated the path of the LRM to be oblique from its insertion dorsally downwards.4 CT and MRI scans showed nasal displacement of the vertical recti muscles and, particularly, inferonasal displacement of the LRM.4 Therefore, high-graded horizontal RR with supraposition of the muscles was recommended for large esotropia with hypotropia.4 At the Bielschowsky meeting, 1995, Herzau reported on a series of such operations.5 His decisive innovation was to normalise the pulling direction of the slipped LRM by adding an equatorial myopexy of this muscle.5 This initiated further studies illustrating the principle of the LRM myopexy for this specific purpose, which completely differs from the aim of retro-equatorial myopexy as described by Cüppers.6 Krzizok et al analysed muscle paths of myopes before and after myopexy by MRI. Imaging contributed to the understanding of the pathophysiology of esotropia and hypotropia associated with high myopia.7 MRI also demonstrated that the path of the LRM could be normalised by equatorial myopexy.7 Case reports on this technique and modifications appeared.8–11 Yokoyama et al elaborated the fundamental idea by joining the lateral and superior rectus muscles by a muscle suture anterior to their pulleys. Avoiding the scleral suture is per se an advantage, in particular, for surgeons who rarely perform myopexy.12–14 Yamaguchi and colleagues presented MRI findings showing the superotemporal shift of the posterior part of the globe out of the normal extraocular muscle cone, and demonstrated that the surgical union of the superior rectus muscles and LRMs restored their normal anatomic relationship.14 Full correction of large esotropia usually required additional medial rectus recession.14 Another technique, transposing the upper and lateral halves of the lateral and superior rectus muscles, respectively, to the upper temporal quadrant, needs scleral sutures, but closer to the corneal limbus.15 Owing to many reports on these modifications,16–18 LRM myopexy lost its popularity. In this study, results of RR with repositioning of the course of the LRM by scleral equatorial myopexy were evaluated on a large sample, analysing the clinical results and efficacy of this method.

Materials and methods

Data collection

The study was approved by the local institutional board of the Justus Liebig University. Patient documents were evaluated retrospectively. Patients were identified in operation schedules of the years 2003–2013, which always contained the axial length of the eyes. Exclusion criteria were sixth nerve palsy, neurogenic disorders and thyroid-associated orbitopathy. Files of patients were evaluated who received RR with equatorial LRM myopexy for esotropia with high myopia. The following data were drawn from the files:

  • age at surgery

  • gender

  • axial length

  • cycloplegic refraction

  • visual acuity

  • type and amount of surgery

  • maximal strabismus angle

  • postoperative binocular function.


The age of the 39 female and 10 male patients was 56 years (36–76). No patient had known connective tissue disorders like Marfan or Stickler syndrome. Eleven patients had undergone prior muscle surgery on the operated eye, and 16 had received cataract surgery. Axial length was 31.5 mm (26.0–35.6), refraction (spherical equivalent) of phakic eyes (16 eyes were pseudophakic) was −18.0 dpt (−28.0 to −6.5), visual acuity 0.32, that is, Snellen fraction of 20/63 (light perception to 1.25 or 20/16, respectively). Three patients had not attended the visit after 3 months. Thus, results of 46 patients were available.


Strabismus angles were measured with the alternating prism and cover test at 5 m and 0.3 m. The prismatic effect of the glasses was taken into account, which is significant in high refractive errors. For example, in front of a −18 dpt glass, a 25° (47 prism dioptres) prism neutralises 16° strabismus angle.19 This effect does not occur in case of contact lens correction or uncorrected myopia. If fixation was lacking due to amblyopia or maculopathy, the corneal light reflex of the deviated eye was positioned symmetrically to the reflex of the fellow eye using base-out and, in case of vertical deviation, additional vertical prisms. When binocular vision allowed subjective measurements, values localised at the Harms tangent scale with the dark-red glass in front of the fixating eye were taken to quantify the vertical deviation. In the horizontal plane, objective strabismus angles were taken, as the dark-red glass can induce convergence. Findings before and 3 months after surgery were evaluated.


Surgery was performed under general anaesthesia using a limbal approach. Polyglactin 6–0 was used for muscle sutures, polyester 4–0 for the myopexy and polyglactin 9–0 for the conjunctiva. For the medial rectus recession, the upper and lower forth of the tendon were sutured, and the distance of the sutures from the limbus was measured before severing the tendon from the sclera. To define the target point at the sclera, the amount of recession was added, and the measurement was performed from the limbal reference point. The LRM was sutured in the desired distance from its insertion (corresponding to the amount of resection, upper and lower third), taken in a muscle clamp, severed from the eye and fixed at the insertion site with the prepared sutures. The anterior tendon was resected. Alternatively, plication of the LRM was performed. The distance of the myopexy from the limbus was measured with a curved ruler or in two equal steps with a calliper, straight backwards along the upper end of the insertion site. The long posterior ciliary artery and the vortex vein served for orientation. The suture was placed parallel to the limbus over a length of 2–3 mm through the sclera and through the upper third of the muscle. The notch was closed on the muscle surface just tight enough to fix the muscle in the desired position.

Outcome measurements

For comparison with literature data, efficacy of the surgery was calculated as the ratio of strabismus angle reduction in degrees and the total amount of surgery (recession + resection) in millimetres.


Individual results, ranges, 0.5 quantiles (medians), 0.1 quantiles and 0.9 quantiles were calculated. The Mann–Whitney U test and, for comparison with other studies, mean values and SDs, were calculated, and the t test was applied using SPSS V.22.


All 49 patients had received unilateral RR with equatorial myopexy of the LRM to the horizontal median. Medians and ranges (in parenthesis, minimum–maximum) are given. The mean total amount of RR was 10.0 mm (5.0–24.0), equally distributed between the medial rectal muscles and LRMs, for example, 5 mm recession and 5 mm resection. Esotropia was reduced from 16° (4–60) to 2° (−22 to 34), hypotropia from 3° (−3 to 30) to 0° (−8 to 18). Esotropia at 0.3 m was reduced from 15.5° (−1 to 60) to 0° (−22 to 36). Additional supraposition of both muscles by 4–6 mm was performed in 11 eyes. In the entire group, esotropia was reduced by 1.44°/mm (−0.40 to 3.55) of total RR, at mean 1.49±0.68°/mm. There was no complication like scleral perforation, intraocular pressure rise or anterior segment ischaemia in any of the patients. For the following calculations, patients were excluded who had received previous muscle surgery. Among the remaining 35 patients, the total amount of RR was 10.0 mm (6.0–24.0) with an equal amount of recession and resection/plication in most cases. The degree-per-millimetre ratio (efficacy) was 1.33 (0.20–3.55). Table 1 contains mean values for comparison with literature data. After 3 months, 22 patients (63%) had a residual horizontal angle of ≤5°, and four patients (11%) had an exodeviation of >5°. Nine patients (26%) had esodeviations of >5°, and in six of them, the preoperative esotropia was 36°–60°.

Table 1

Mean values and SDs (mean±SD) of patients with RR with LRM equatorial myopexy. All cases without prior surgery

Efficacy of surgery was significantly lower in the 23 bilateral than in the 12 unilateral high myopes (axial length of the fellow eye <26 mm), that is, 1.33°/mm (0.50–2.19) versus 1.95 (0.20–3.55) (p=0.001). Axial length 31.7 mm (26.8–33.9), esotropia 14° (5–60) and RR 10.0 mm (6.0–24) in the bilateral group did not differ significantly from their respective values of 29.1 mm (26.0–35.3), 16° (4.0–50) and 10.0 mm (7.0–21.0) in the unilateral group. Neither in the bilateral nor in the unilateral group, there was a statistically significant difference in mean efficacy between patients examined by the light reflex method (30.4% and 66.7%) and patients examined with the cover test (69.6% and 33.3%) with p=0.418 and p=0.711, respectively. Binocular vision was achieved in three patients with unilateral (one random dot stereopsis, two Titmus circles), and 12 patients with bilateral high myopia (five random dot stereopsis, four Titmus circles, one Titmus fly, two Bagolini). There was also no significant difference in efficacy between patients with and without postoperative binocular vision.


The outcomes of this study confirm that RR with equatorial myopexy of the superior part of the LRM is suitable to correct esotropia and hypotropia with high myopia. Of the 46 patients for whom the 3 months’ postoperative data were available, 28 (60.9%) had a residual angle of deviation within ±5° at 5 m. Excluding the seven patients with esotropia >40° who could be expected to need bilateral surgery for full strabismus correction, the success rate increases to 71.8%. In the entire group, including patients with previous surgery, the mean efficacy of 1.55°/mm corresponded to that of RR on normal-size eyes. For esotropia ≥20°, ≥30° and ≥40°, respectively, mean efficacy remained 1.6°/mm, similar to that of patients without previous surgery (table 1).

Krzizok et al7 reported on 12 patients whose LRM was shown to be dislocated both in situ and by MRI. Seven patients received RR with LRM myopexy and five without LRM myopexy because their sclera appeared too thin (table 2). For cases with myopexy, mean efficacy calculated from their data, excluding one patient with extreme esotropia of 140°, exceeded that in our patients, in particular in the subgroup with bilateral high myopia (table 1). Hypotropia also improved. Mean efficacy in their patients without myopexy was not significantly different (p=0.792, table 2), and two also achieved vertical alignment. Just in one case, hypotropia increased, despite supraposition of the horizontal muscles.7 All except that woman had bilateral high myopia. These results appear to relativise the need of myopexy.

Table 2

Data of former study before and after RRwith (n=6) and without (n=5) LRM myopexy (recalculated from data in,7 patient with 140° esotropia excluded)

The difference in efficacy compared with our patients without prior surgery (p=0.0009) needs to be addressed. Suture instability can hardly explain it because, in the former study, RR without myopexy was also highly efficient.7 Mean esotropia was 34° in the former study (without the 140° case), but only 22° in our patients. Accordingly, the total amount of surgery was 18 mm compared with 12 mm in our group. However, even after including only esotropia >20°, resulting in a mean esotropia of 39° and mean RR of 17.5 mm, efficacy in our group did not reach >1.3°/mm, and there was no further increase, when the cut-off was set at 30° or 40° or 30 mm axial length. By exclusion of patients with prior surgery in the former study, efficacy decreased to 1.5°/mm. A great deal of the effect of surgery is probably coming from the myopexy portion of the procedure, and the effectiveness of the myopexy would be expected to depend on the degree of preoperative pathological dislocation of the LRM path.14 Therefore, the value of the degree-per-millimetre ratio as a comparative parameter should not be overstated.

Sturm et al11 reported a mean efficacy of 1.8°/mm in an evaluation of six patients operated with and without LRM myopexy, including one case of extreme esotropia. They favoured a surgical procedure depending on the individual situation, and emphasised that conventional RR yielded good results, even for strabismus fixus.11 Pure medial rectus recession or even tenotomy is not sufficient in case of large esotropia.20 ,21 Instead of a large recession, a tendon elongation (eg, by a bovine pericardium graft) can be performed and combined with surgery on the LRM.4 Recession of nasal conjunctiva, which may be shrunken, may be needed to maintain the full effect of muscle surgery.22

Why did efficacy in unilateral high myopia exceed that in bilateral high myopia? Efficacy was similar in light reflex and cover test subgroups for both unilateral and bilateral high myopia. Sensorimotoric influence is also unlikely; there was no systematic eso-drift compared with the first postoperative day. Contracture of the medial rectus muscle might have been more pronounced in unilateral high myopia than in bilateral high myopia, with a larger rate of patients with alternating fixation. Provided recession of a contracted muscle yields more effect than recession of a non-contracted muscle, this might be a decisive factor explaining the difference in efficacy.

Yokoyama's method yields an abducting and elevating effect by surgical union of the corresponding margins of the downward-displaced LRM and the nasally displaced superior rectus muscles near the equator with a non-absorbable suture that does not involve the sclera.12 ,13 Yamaguchi et al14 reported on 14 patients of whom nine were operated on both eyes. In seven cases, supplemental medial rectus recession was performed simultaneously, and in three cases, after months because of residual esotropia.14 Mean esotropia was 58.8±36.0°, mean postoperative deviation was 0.7±9.0°. There are several reports on that procedure,8 ,11 ,16 but a study on a larger sample is lacking. Yamada's method also requires additional medial rectus muscle recession for large esotropia.15 The retransposition and transposition procedures are, meanwhile, widely used in the treatment of esotropia with high myopia.8 ,17 ,18 Especially in large esotropia, which is often accompanied by hypotropia, their elevating effect can help avoid surgery on vertical muscles. In the former study from our hospital, horizontal RR with LRM myopexy reduced hypotropia by up to 25°.7 We achieved an elevation by 20° and 30° in two cases. These are obviously mechanical effects. In the former study, hypotropia also improved after pure RR.7 However, better long-term stability due to myopexy seems likely. Small vertical deviation can disappear by sensorimotoric regulation after correction for esotropia.

Our patients represent the clinical routine over a longer period of time. MRI findings were not available in all patients, and were not considered in this study. MRI is helpful to differentiate Graves’ orbitopathy and neurological disorder like sixth nerve palsy or myasthenia gravis that can also occur parallel to high myopia. MRI is also useful to quantify the muscle dislocation in case of esotropia associated with high myopia.7 ,14 It supports the indication of muscle repositioning, and visualises the anatomical effect of surgery.7 ,14 Severe myopic staphylomata may develop very bizarre globe shapes so that the eye cannot rotate in the orbit without collision with the orbital walls.3 ,23 In such cases, it would probably be harmful to attempt such a surgery. For these reasons, clinicians may find it prudent to ask for preoperative imaging in highly myopic patients with severe strabismus before selecting a surgical procedure. However, we do not suggest that CT/MRI is indispensable regarding the indication of surgery. In many cases, the decision whether or not myopexy or any transposition makes sense can be made primarily on the basis of history, strabismus angle pattern and intraoperative finding. Significant dislocation of the LRM, recognisable by the downward-sloped muscle path that makes the insertion of the inferior oblique muscle to appear above its superior margin, is an argument to relocate the path of the LRM. If that criterion is not fulfilled, we prefer pure RR.

In conclusion, our results of RR with equatorial myopexy of the superior third of the LRM were satisfactory. Myopexy of the LRM is useful for the particular purpose to normalise the path of the dislocated muscle. Both augmenting and sustained effect of the myopexy appear logical. Thus, in case of significant inferior displacement of the LRM, myopexy is warranted, and represents a valuable alternative to muscle union and transposition procedures.



  • Contributors Collection of data and statistical calculations were mainly performed by MG. Both authors had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis as well as the decision to submit for publication. The discussion and the entire paper are the result of intensive discussions between the authors. Lingual corrections were performed by BL. MG.

  • Competing interests None declared.

  • Ethics approval Ethikkommission, Fachbereich 11, Humanmedizin, Justus-Liebig-Universität Giessen, Giessen, Germany.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Linked Articles

  • At a glance
    Keith Barton James Chodosh Jost Jonas