Patients may need two or more strabismus surgeries during their lifetime.1–5 Knowing the location of the muscle insertion prior to surgery is important when planning a recurrent surgery. Primary surgery information may not be available for adult patients, and there might be a slippage of the reattached muscle and stretched scars that change the muscle position.6–8
Previous studies have shown that ultrasound biomicroscopy can accurately determine the distances of extraocular muscle insertions from the limbus in unoperated horizontal and vertical muscles7 and horizontal muscles that underwent surgery.8,9 In addition to measuring distance, ultrasound biomicroscopy can differentiate between the presence of a pseudotendon and a true insertion.7,10
Ultrasound biomicroscopy examinations in early reports were performed under general anesthesia and required an eyelid speculum and forceps to manipulate the eye.9 Previous ultrasound biomicroscopy instruments were limited by lower resolution, screen range of view (which was confined to 5 to 6 mm), and the need for using a caliper as a reference point to measure further back from the limbus.9–11 More recent reports extended the range of view to 14 mm from the limbus using wide-field ultrasound biomicroscopy.12,13 However, the examination was performed under general anesthesia with an eyelid speculum and traction, using the immersion technique that has been widely abandoned.14 Mirmohammadsadeghi et al.15 reported good agreement between intraoperative and wide-field preoperative ultrasound biomicroscopy measurements for locating extraocular muscles. Although a wide-field ultrasound biomicroscopy instrument was used under topical anesthesia, an eyelid speculum and tear gel were still needed as a coupling agent.
The introduction of the modern ultrasound biomicroscopy bag/balloon technique (also called the clear shell technique) increased diagnostic accuracy through higher image resolution. This allows easy visualization of the extraocular muscle insertions up to the equator as the most posterior extent. Moreover, this technique is sterile because the ultrasound biomicroscopy probe is covered by a sterile, single-use, water-filled balloon that is easy to use and allows a comfortable examination in an upright position without an eyelid speculum.14,16
Therefore, the purpose of our study was to investigate the accuracy to allocate muscle insertion positions before and after strabismus surgeries using wide-field ultrasound biomicroscopy with the bag/balloon technique.
Patients and Methods
Study Design and Participants
The study protocol was reviewed and approved by the ethics committees of the participating medical center in compliance with the tenets of the Declaration of Helsinki. Written informed consent was obtained from all 25 patients prior to enrollment in the study. All adult patients older than 18 years who required strabismus surgery involving horizontal recti muscles only were included in the study. Patients with additional vertical strabismus were excluded. The magnitude of strabismus was not an exclusion criterion. Three patients were excluded due to incomplete documentation of ultrasound biomicroscopy data.
After a signed consent was obtained, patients underwent a routine strabismus examination performed by the surgeon, which included visual acuity, alignment at distance and near, and eye movement documentation. Both the preoperative ultrasound biomicroscopy examination and the clinical evaluation were performed 1 day before surgery. Postoperative ultrasound biomicroscopy was performed at least 30 days after surgery, when patients arrived for their planned follow-up visit at the strabismus clinic.
The ultrasound biomicroscopy examination with the 50-MHz Aviso S instrument (Quantel Medical, Clermont-Ferrand, France) was performed by one of two experienced ultrasound specialists (DZ, MN). Following topical ocular anesthesia with oxybuprocaine hydrochloride 0.4% (Localin; Fischer Pharmaceutical Labs, Tel Aviv, Israel), the patient's head was tilted back approximately 45°. A sterile balloon (ClearScan; ESI, Plymouth, MA) was filled to the top of the flexible sealing collar with room temperature distilled water. Hydroxyethylcellulose 1.4% eye drops (Celluspan; Fischer Pharmaceutical Labs) were used for the interface between the balloon's surface and the conjunctiva. The ultrasound biomicroscopy balloon was held perpendicular to the insertion on the respective muscle, and the patient was asked to maintain gaze away from the examined muscle.
The limbus was identified as the “corneoscleral junction” and seen as an abrupt transition between the hyporeflective corneal stroma and the hyperreflective sclera.13 The muscle belly was well delineated as a hyporeflective area. The distance between the limbus and the muscle insertion was measured with a built-in caliper in the ultrasound biomicroscopy machine (Figure 1). An average of three readings were taken for each muscle. During recording, the observer was masked to all clinical data.
(A) Representative ultrasound biomicroscopy measurement of the rectus muscle insertion distance from the corneoscleral limbus in a naive eye. A line was drawn perpendicular to the corneal endothelium from the scleral spur (open arrow), indicating the corneoscleral limbus. The distance between the corneoscleral limbus and the insertion of the medial rectus muscle was measured. Asterisk (*) = anterior chamber; star = cornea; bold arrow = medial rectus muscle belly. (B) Ultrasound biomicroscopy image of a lateral rectus muscle prior to reoperation for residual exotropia showing the insertion of the lateral rectus muscle at a distance of 9.5 mm from the limbus (arrow). The limbus is represented by the white line. The actual distance measured intraoperatively was 10 mm. The white banner indicates the location of the original insertion of the lateral rectus muscle. An irregularity of the anterior sclera at the area of the previous surgery can be seen.
Surgical Procedure and Intraoperative Measurements
Extraocular muscle surgery was performed by experienced strabismologists. The surgeon was masked to the ultrasound biomicroscopy records. Caliper measurement was performed prior to disconnecting the muscle from the globe after the muscle insertion was exposed (ie, preoperative caliper measurement), and again at the end of the procedure after the muscle was reconnected to the globe and before the conjunctiva was closed (ie, end of surgery caliper measurement). Prior to each measurement, the accuracy of the caliper was confirmed with a ruler. Then the distance from the limbus to the muscle insertion was measured with the caliper.
In cases of reoperations where the muscle was found far from the original insertion, a tightening procedure was planned to advance the muscle to its original insertion and resect as needed. Patients were followed up at the strabismus clinic according to our routine recommendations.
Differences between caliper and ultrasound biomicroscopy measurements were calculated as absolute values to avoid mutual deduction of opposite differences and to expose differences otherwise annulled by large variations on opposite sides of the mean.
Data were recorded in Microsoft Excel (Microsoft Corporation, Redmond, WA) and analyzed using SPSS software (version 25; SPSS, Inc., Chicago, IL).
Correlations in measurements of continuous variables (eg, distance from the limbus) were calculated with the Pearson correlation coefficient (r). Differences between caliper and ultrasound biomicroscopy measurements were represented in Bland–Altman plots. The differences between the measurements were plotted against the mean of the two measurements, with 1.96 standard deviations (SD) of the difference representing the 95% limits of agreement.
Continuous variables were compared within subjects using the paired t test and between subjects using the independent sample t test.
For small group comparisons and ordinal variables, the Wilcoxon non-parametric test was performed for paired comparisons and the Mann–Whitney non-parametric test was performed for independent samples.
All tests were two-tailed and the threshold for statistical significance was defined as a P value of less than .05.
Thirty-nine muscles of 22 adult patients were included in the analysis. The mean age of the patients was 34.7 ± 15.5 years (range: 18 to 78 years). Relevant demographic and clinical data are detailed in Table 1. Nineteen were medial rectus muscles and 20 were lateral rectus muscles.
Clinical Data and Measurements of All Muscles Included in the Analysis
Table 2 summarizes the mean and range of ultrasound biomicroscopy and caliper measurements prior to surgery according to the operated muscle. No significant differences were found between ultrasound biomicroscopy and caliper measurements of the medial rectus and lateral rectus muscles (P = .116 and .377, respectively). When comparing the absolute mean differences between medial rectus and lateral rectus muscle measurements, the difference was 0.07 mm (P = .708). For 29 of 39 muscles (76.9%), the differences between ultrasound biomicroscopy and caliper measurements were within 1 mm.
Comparison Between Preoperative UBM and Caliper Measurements According to Type of Muscle
We also compared the accuracy of ultrasound biomicroscopy measurements prior to surgery in naive muscles to muscles that underwent previous surgery (re-operation subgroup). Of the 39 muscles included, 27 were naive and 12 previously underwent surgery. In the naive muscle group, the mean difference between ultrasound biomicroscopy and intraoperative caliper measurement was 0.56 ± 0.43 mm. In the reoperation group, the mean difference was 0.91 ± 0.81 mm. The difference in accuracy between naive and reoperated muscles was 0.351 mm (P = .327). Bland–Altman plots comparing preoperative ultrasound biomicroscopy and caliper measurements are demonstrated in Figure 2. The muscles are subdivided according to medial rectus and lateral rectus muscles (Figure 2A) and there is a subclassification between the naive muscle and reoperation groups (Figure 2B).
Bland–Altman plots showing the difference between preoperative ultrasound biomicroscopy (UBM) and preoperative caliper measurements plotted against the average of the two measurements. The dashed lines represent the 95% confidence interval (CI). (A) Color sub-classification according to medial rectus and lateral rectus muscles and (B) color subclassification according to the naive muscle group versus the reoperation group. The three dots outside the CI represent previously operated muscles. MR = medial rectus; LR = lateral rectus
For 25 muscles of 14 patients (56%), follow-up ultrasound biomicroscopy was available 191.9 ± 82.55 days (range: 67 to 365 days) after the surgery. Twelve of those muscles were recessed (moved away from the limbus) and 13 were tightened by resection and/or advancement (muscle remains at the original insertion). We compared the location of muscles at the end of surgery, as measured by the surgeon using a caliper (ie, end of surgery caliper measurement), to the location at the follow-up ultrasound biomicroscopy evaluation (ie, postoperative ultrasound biomicroscopy measurement). For 16 of 25 muscles (64%), the difference between the two methods was less than 1 mm, with a 95% confidence interval between 42.5% and 82.0%.
Table 3 summarizes the mean and range of muscle positions at the end of surgery by caliper and ultrasound biomicroscopy according to type of procedure. The recessed muscles group had one outlier (muscle 12). This was a 7-mm lateral rectus recession performed with the “hang-back suture” technique. Although the end of surgery position was 14 mm, the postoperative ultrasound biomicroscopy measurement was only 6.7 mm. Because surgery was successful and the strabismus was resolved, we believe that a pseudotendon formed between the original and new insertion and it was mistakenly identified as the insertion of the recessed muscle on ultrasound biomicroscopy. Figure 3 shows preoperative and postoperative ultrasound biomicroscopy findings.
Comparison Between End of Surgery Caliper and Postoperative UBM Measurements According to Type of Surgery Performed
Ultrasound biomicroscopy images of muscle 12. (A) Preoperative ultrasound biomicroscopy showing the insertion of the lateral rectus muscle at a distance of 6.3 mm (arrow). The actual distance measured intraoperatively was 6.5 mm. (B) The lateral rectus muscle after a hang-back recession procedure. A hypoechogenic area at a distance of 6.7 mm from the limbus was identified and interpreted as the new insertion of the muscle. The band is irregular and relatively thin. Therefore, the hypoechogeneicity most probably represents a pseudotendon and not the actual muscle tendon. The tendon should be at a distance of 14 mm from limbus, which cannot be observed in the picture.
We compared the absolute mean differences between recession and resection measurements. For the recessed muscles group including the outlier, the difference was 1.09 (P = .014). When excluding the outlier, the difference was 0.562 (P = .026). A Bland–Altman plot of postoperative ultrasound biomicroscopy compared to end of surgery caliper measurements according to type of procedure is demonstrated in Figure 4.
A Bland–Altman plot showing the difference between postoperative ultrasound biomicroscopy (UBM) and end of surgery caliper measurements versus the average of the two measurements. The dashed lines represent the 95% confidence interval (CI). Color subclassification according to type of procedure. One recessed muscle (muscle 12) is outside the CI. Rc = recession; Rs = resection
Additionally, we compared the differences between end of surgery caliper and postoperative ultrasound biomicroscopy measurements between naive and reoperated muscles. Of the 25 muscles included, 16 had surgery for the first time and 9 were reoperations. The difference was 1.21 ± 1.77 mm in the naive muscle group and 0.95 ± 0.84 mm in the reoperation group. The difference between the two groups was 0.26 mm (P = .934). Measuring the same excluding the difference was 0.79 ± 0.61 mm in the naive muscle group (15 muscles) and 0.95 ± 0.84 mm in the reoperation group (9 patients). The difference between the two groups was 0.16 mm (P = .864).
The Pearson correlation was 0.929 (P < .001) between preoperative ultrasound biomicroscopy and intraoperative measurements for all 39 muscles. Between end of surgery caliper and postoperative ultrasound biomicroscopy measurements, the Pearson correlation was 0.792 (P < .001) for 25 muscles, including the outliers. The Pearson correlation between time to final ultrasound biomicroscopy and the absolute difference between postoperative ultrasound biomicroscopy and end of surgery caliper measurement was −0.139 (P = .516). To better understand the difference between measurements across a range of measurements, for 25 muscles that were also evaluated by ultrasound biomicroscopy postoperatively, there was a significant positive correlation between the position at ultrasound biomicroscopy and the end of surgery position (rp = 0.436; P = .030). After excluding the outlier previously described, the correlation lost its significance (rp = 0.389; P = .060).
To our knowledge, this is the first study to prospectively evaluate the accuracy of wide-field ultrasound biomicroscopy's modern bag/balloon technique in locating extraocular muscle insertions in patients with strabismus. High-resolution ultrasound biomicroscopy enabled precise identification of the muscle insertions and stood in high correlation with caliper measurements.
The bag/balloon technique used in the current study has several advantages over previous methods. A sterile, disposable water-filled balloon covers the ultrasound biomicroscopy probe, which eliminates the risk of transmitting infections between patients.14 Probes used in previous studies11,13,15 required an immersion cup or needed to come into direct contact with the eye.15 The inability to easily sterilize the probes resulted in an unacceptable high prevalence of micro-organisms (up to 53%), despite proper cleaning.17 Because the bag/balloon technique avoids direct contact with the eye, there is no risk of corneal erosion. Moreover, the technique is painless and well tolerated by patients and easily performed in an upright position, without needing an eyelid speculum. In contrast, the open shell technique may cause the patient great discomfort due to an eyelid speculum, a rigid scleral shell, and the fluid dripping on the patient's face.
Furthermore, combining the bag/balloon technique with the wide field-of-view ultrasound biomicroscopy instrument made it easy to visualize the entire area of interest (from the limbus and all areas of visible sclera up to the equator). Although the bag/balloon technique can be used to visualize beyond 14 mm from the limbus, we aimed to visualize both the limbus and the muscle insertion in one capture without manipulating the eye with forceps or using the caliper as a reference point to obtain accurate measurements. With our system, the maximum distance that can be visualized in sufficient quality is 14 to 15 mm from the limbus. The probe needs to be positioned in the plane of the respective muscle. Furthermore, the ability to display structures distant to the limbus depends on the aperture of the eyelid and mobility of the eye.
One strength of the current study was the relatively large number of patients included because it allowed comparisons between different subgroups. Preoperatively, ultrasound biomicroscopy measurements showed excellent accuracy for muscle insertion assessment with no significant difference between medial rectus and lateral rectus muscles measurements. Our results oppose those of Thakur et al.,13 who found that ultrasound biomicroscopy was not accurate in assessing the lateral rectus muscle insertion position. However, we believe that this may be explained by the bag/balloon technique used in our study as opposed to the scleral shell technique used by Thakur et al. Another possible reason is that the average lateral rectus recession was 8.45 mm (range: 7 to 9 mm) in the study published by Thakur et al., whereas it was 5.3 mm (range: 3 to 7 mm) in the current study. A subanalysis on muscles that had previous surgery also revealed high accuracy. The mean difference between ultrasound biomicroscopy and caliper measurements was within 1 mm, which was not statistically significant. However, the standard deviation was larger for previously operated muscles compared to naive muscles, representing a wider distribution of differences (Figure 2B).
Postoperative evaluations were performed by ultrasound biomicroscopy at least 4 weeks after surgery to avoid the effects of healing and edema on the accuracy of the measurements. An ultrasound biomicroscopy examination during follow-up showed accuracy in locating resected muscles. For recessed muscles, the mean difference between ultrasound biomicroscopy and end of surgery position after excluding an outlier was 1.15 mm, which was almost within the accepted mistake range of 1 mm. Postoperative ultrasound biomicroscopy analysis of recessed muscles revealed underestimation, compared to the muscle location measured by the caliper at the end of surgery (Figure 4). Therefore, although ultrasound biomicroscopy gives accurate information regarding the type of previous surgery (ie, recession or resection), users should bear in mind while planning a second surgery that ultrasound biomicroscopy may underestimate the magnitude of recession.
After comparing the differences between caliper and ultrasound biomicroscopy measurements, we found a moderate positive correlation between the distance of the muscle from the limbus and the inaccuracy of the measurement. A further in-depth analysis showed that this correlation was driven by one outlier with distinct discrepancies in measurement. Contrary to previous reports,10,13 we identified all of the muscles during follow-up. The farthest muscle (number 10) was positioned 15 mm from the limbus and located by ultrasound biomicroscopy at 11.8 mm. The two most inaccurate ultrasound biomicroscopy measurements were muscle 12 postoperatively and muscle 19 preoperatively. Muscle 12 was positioned 14 mm from the limbus and located at 6.7 mm. Muscle 19, which had been previously operated on, was located by ultrasound biomicroscopy at 5.6 mm, whereas the true location at the beginning of surgery was 10 mm from limbus.
In both cases, we believe that a pseudotendon was created between the original insertion and the new insertion and it was mistaken for a muscle insertion (Figure 3). In retrospect, the hyporeflective band in Figure 3B is very thin and irregular. The surgeon performing the ultrasound biomicroscopy was blinded to the type of surgery and may have assumed it was a resection and did not look for another band. Therefore, a thin and irregular muscle must be suspected as being a pseudotendon, and the surgeon should look for another hyporeflective band behind it. Dai et al.,10 Khan et al.,12 and Mirmohammadsadeghi et al.15 clearly identified pseudotendons in their studies, whereas Thakur et al.13 estimated that their inaccuracy was related to pseudotendons that were not properly identified. After excluding the outlier (muscle 12) from our results, the significance of the correlation between distance from the limbus and measurement error was lost, indicating that accuracy is well maintained (Figure 4).
Patients who previously underwent extraocular muscle surgery present a diagnostic challenge to ultrasound biomicroscopy and a decision challenge to the surgeon. Although ultrasound biomicroscopy was easily performed and interpreted in the current study, the ultrasound is an operator-dependent instrument and appropriate experience is needed for proper acquisition and evaluation of the images.
Ultrasound biomicroscopy is a helpful tool for strabismus surgeons who are planning a recurrent surgery. The bag/balloon technique is a reliable method to evaluate muscle position, regardless of the length of time since surgery. An experienced physician is required to identify previously operated muscle insertions. Larger prospective studies are needed to define the current role of ultrasound biomicroscopy imaging in planning strabismus surgery and postoperative follow-up.