Postoperative shift in ocular alignment following single vertical rectus recession on adjustable suture in adults without thyroid eye disease Emily Bratton, MD,a Mary Ellen Hoehn, MD,b,c,d Wonsuk Yoo, PhD,e,f Kyle Fitzgerald Cox, MD,b and Natalie C. Kerr, MDb,c,d PURPOSE
To determine whether overcorrection shifts occur after vertical rectus recession on adjustable suture in the absence of thyroid eye disease.
METHODS
The medical records of patients without thyroid eye disease who underwent vertical rectus recession surgery from 2001 to 2008 were retrospectively reviewed for shifts in alignment between suture adjustment at postoperative day 1 and 2 months’ follow-up. Superior rectus and inferior rectus recessions were compared. In addition, we compared the use of a nonabsorbable polyester suture to an absorbable polyglactin 910 suture in nonthyroid patients undergoing inferior rectus recessions.
RESULTS
A total of 59 patients were included (superior rectus, 30; inferior rectus, 29). We found a mean undercorrection shift of 1.1 (range, 17.5D undercorrection to 16D overcorrection) and 1.0D (range, 12D undercorrection shift to 6D overcorrection shift) for superior and inferior rectus recessions, respectively, between 1 day and 2 months postoperatively. There was no trend toward overcorrection following unilateral vertical rectus adjustable suture recessions in patients without thyroid eye disease, suggesting that thyroid myopathy may account for overcorrection shifts seen with this surgery. ( J AAPOS 2015;19:247-251)
CONCLUSIONS
L
ate overcorrection following inferior rectus muscle recession is a well-known problem. Risk factors include thyroid eye disease (TED),1-4 the use of adjustable sutures,1,2,4,5 and unilateral inferior rectus muscle recession.1,4,5 The effect of operations on multiple muscles at the time of inferior rectus recession is unknown.1,4 In addition, late overcorrection following superior rectus recession has not been described.1 Overcorrection typically occurs 4-6 weeks after inferior rectus recession.1,3,6 In fact, it has been shown that patients aligned successfully at 4-6 weeks after surgery maintain
Author affiliations: aDepartment of Ophthalmology, Rocky Mountain Lions Eye Institute, University of Colorado, Aurora; bDepartment of Ophthalmology, Hamilton Eye Institute, University of Tennessee College of Medicine, Memphis; cDepartment of Pediatrics, University of Tennessee College of Medicine, Memphis; dDivision of Ophthalmology, St. Jude Children’s Research Hospital, Memphis, Tennessee; eDivision of Biostatistics and Epidemiology, University of Tennessee College of Medicine, Memphis; fDepartment of Community Health and Preventive Medicine, Morehouse School of Medicine, Atlanta, Georgia Supported by an unrestricted grant to the University of Tennessee from Research to Prevent, Blindness Inc, New York, New York. Presented in part at the Annual Meeting of American Academy of Ophthalmology, San Francisco, California, October 24-27, 2009. Submitted April 11, 2013. Revision accepted March 23, 2015. Correspondence: Natalie C. Kerr, MD, FACS, Hiatt Professor of Ophthalmology, Hamilton Eye Institute, University of Tennessee Health Science Center, 930 Madison, Suite 470, Memphis TN 38163 (email: [email protected]). Copyright Ó 2015 by the American Association for Pediatric Ophthalmology and Strabismus. 1091-8531/$36.00 http://dx.doi.org/10.1016/j.jaapos.2015.03.015
Journal of AAPOS
alignment in long-term follow-up.7 This critical period coincides with the point when absorbable sutures lose their strength,8 and it has been proposed that absorbable sutures may predispose patients to late overcorrection.4,9 However, research on fixed, adjustable, and nonabsorbable sutures in a patient population of thyroid and nonthyroid patients has indicated alternative mechanisms for late overcorrection during this critical period.1,3,4 The present study investigated postoperative shifts in patients without TED following suture adjustment during the critical period between postoperative day 1 and 2 months. We compared inferior rectus recessions to superior rectus recessions to determine whether the inferior rectus is uniquely associated with late overcorrection shifts in patients without TED. We also compared the use of a nonabsorbable polyester fiber suture (Mersilene, Ethicon Inc, Somerville, NJ) to an absorbable polyglactin 910 suture (Vicryl, Ethicon Inc) in nonthyroid patients undergoing inferior rectus recession on adjustable suture to examine the possible role of suture absorption in postoperative shifts.
Subjects and Methods This study was approved by the University of Tennessee Investigational Review Board and complied with the US Health Insurance Portability and Accountability Act of 1996. The medical records of consecutive adult patients (.18 years of age) who underwent vertical rectus muscle recessions at the University of Tennessee Helath Science Center by one of two surgeons
247
248
Bratton et al
(MEH and NCK) between 2001 and 2008 were reviewed retrospectively. Patients with bilateral or multiple vertical muscle recessions, transposition, incomplete follow-up, thyroid eye disease (TED), poor visual acuity (\20/40 best-corrected visual acuity in one or both eyes), or in whom fixed sutures were used were excluded. Patients were not sex or age matched within study groups. The following preoperative data were collected: case number, date of birth, date of first visit, preoperative diagnosis or etiology of the motility disorder, and preoperative measurement of ocular deviation. We recorded the measurement of ocular deviations using the alternate prism and cover test in primary position at distance with best spectacle correction. All patients were evaluated at least twice before surgery at 2-month intervals and were stable at .6 months from any inciting event or change in symptoms before surgery. As shown in Table 1, preoperative diagnoses included paralytic conditions (unrelated to a cranial nerve palsy, including myasthenia gravis, trauma, and drainage implant causing muscle paralysis); cranial nerve palsies; neurologic conditions, including bilateral skew deviation and internuclear ophthalmoplegia; restrictive or orbital disease; nonstrabismic eye surgery; and hypertropia of undetermined etiology. In some patients, preoperative hypertropia had more than one cause. The most common diagnosis for all patients was trochlear nerve palsy. In all cases, the inferior or superior rectus muscle was the only vertical muscle operated. All superior rectus recessions were performed with polyglactin 910 sutures. Inferior rectus recessions were performed with either polyglactin 910 or polyester sutures. All inferior rectus recessions from 2005 through 2008 were performed with polyester sutures. The surgeons used a standard follow-up protocol of 2 weeks and then 2 months after muscle recession in order to record postoperative alignment. The muscle recessed, amount of recession, and suture material used were recorded. In all cases, the surgeon attempted to adjust the patient to orthotropia. Patients were adjusted within the first 24 hours postoperatively. The surgeon measured the deviation of ocular alignment at that time. Postoperative measurements were recorded at 1 day and 2 months, assessed using distance measurements only in the primary position of gaze. The surgeon determined the postoperative measurements in all cases. Overcorrection shift and undercorrection shift were reported if the patient had a prism diopter shift of any value toward overcorrection or undercorrection after the time of adjustment and 2 months postoperatively. A separate measure, resultant undercorrection and overcorrection, was recorded if the patient’s final alignment was $2D of over- or undercorrection. The threshold analyzed, $2D, was chosen because of previous studies. The $5D threshold has common usage in the literature dealing with strabismus surgery in inferior rectus and/or thyroid eye disease studies.4,10-13 The $2D threshold was chosen, first, because some studies used restoration of single binocular vision/relief of diplopia as their measure of success,14-17 and, second, because of limited vertical fusion in adult patients, who are susceptible to postoperative diplopia if the postoperative alignment shifts more than 2D. This approach seemed a conservative and reasonable alignment threshold to allow comparison to other
Volume 19 Number 3 / June 2015 Table 1. Preoperative diagnoses for patients with vertical rectus recessiona Diagnosis
Superior rectus muscle (n 5 30)
Inferior rectus muscle (n 5 29)
Paralytic/CNP/neurological All CNP CN III CN IV CN VI Paralyticb Childhood strabismus Restrictive/orbital diseasec Eye surgeryd Unknown
10 6 1 4 1 4 13 3 6 1
12 11 2 14 0 1 8 8 5 0
CNP, cranial nerve palsy. a Some patients had multiple diagnoses. b Not due to CNP. c Not thyroid eye disease. d Not involving muscles or orbit. studies and to provide further insight into clinically important overcorrection shifts.4 Means with standard deviations were determined for continuous variables; frequencies and corresponding proportions were determined for categorical data. The distributions of primary outcomes were examined for skewness/normality using the Shapiro-Wilk or Kolmogorov-Smirnov statistic. The MannWhitney-Wilcoxon test was used to compare the superior rectus and inferior rectus muscle groups as well as the two suture groups. To assess correlations between outcomes of overcorrection shift, no shift, and undercorrection shift or measurements with preoperative deviation, regression analysis was performed for the study groups. Pearson or Spearman correlation analyses and linear regression analyses were used to assess the association between the preoperative vertical deviation and postoperative shift over muscle type and suture types. A P value of \0.05 was considered statistically significant. All analyses were performed using Statistical Analysis Software (SAS Institute, Cary, NC, v. 9.2).
Results A total of 119 cases of vertical rectus muscle recessions performed on adults during the study period were identified. Of these, 60 did not meet inclusion criteria. The mean age of patients was 55 18 years. The average age of those undergoing superior rectus recession (n 5 30) was 51 18.7 years; of those undergoing inferior rectus recession (n 5 29), 59 15 years. The mean age of the polyester group was 49 14.5 years; of the polyglactin 910 group, 65 12.2 years. The average preoperative vertical deviation for superior rectus and inferior rectus recession groups were 12.62D 7.9D and 14.74D 8.8D, respectively, and are shown in Table 2. There was no significant difference in the average preoperative deviation between these two groups (P 5 0.18). The mean preoperative vertical deviation for inferior rectus muscle was 17.8D 12.0D for the polyester group and 13.2D 6.7D for the polyglactin 910 group; the
Journal of AAPOS
Volume 19 Number 3 / June 2015
Bratton et al
249
Table 2. Preoperative vertical deviation Surgery
Suture type
Mean preoperative vertical deviation, PD
All Inferior rectus muscles Inferior rectus muscle Inferior rectus muscle Superior rectus muscle
Nonabsorbable Polyester fiber Polyglactin 910 (absorbable) Polyglactin 910 (absorbable)
14.74 17.75 13.16 12.62
Range, PD 2.0-40.0 5.0-40.0 2.0-22.5 4.0-30.0
Number of patients 29 10 19 30
PD, prism diopter.
difference between the suture groups was not statistically significant (P 5 0.35). The amount of recession varied based on the individual patient and their preoperative measurement. For the superior rectus muscle group, the mean amount of recession was 7.2 2.4 mm. Overall, the mean inferior rectus recession was 6.3 2.1 mm. There was no significant difference in the amount of recession between superior rectus and inferior rectus muscles (P 5 0.17). For the inferior rectus surgeries performed with polyglactin 910 sutures, the mean recession was 6.3 2.2 mm. The mean recession for inferior rectus recessions performed with polyester sutures was 6.6 1.9 mm). There was also no significant difference in the inferior rectus group between recessions performed with polyglactin 910 and polyester sutures (P 5 0.89). The mean deviation immediately following adjustment (postoperative day 1) for superior rectus recessions was 1.2D of undercorrection (range, 15D of undercorrection to 12.5D of overcorrection) and 0.9D of undercorrection (range, 9D of undercorrection to 4D of overcorrection) for inferior rectus recessions. No significant difference was observed in mean deviations in superior rectus and inferior rectus muscle groups at postoperative day 1 (P 5 0.42). We analyzed the postoperative drift from one day (after adjustment) to 2 months postoperative, comparing polyester (mean undercorrection shift of 1.5D, range 12D undercorrection to 4D overcorrection) and polyglactin 910 (mean undercorrection shift of 0.8D, range 8D undercorrection to 6D overcorrection) in patients with inferior rectus recessions and found no significant difference (P 5 0.87). For surgeries executed with polyglactin 910 sutures, 4 patients had an overcorrection, 7 had an undercorrection, and 8 were orthotropic. For polyester sutures, 5 patients had a shift of overcorrection and 5 had a shift of undercorrection. Because we found no significant difference when inferior rectus recessions were performed with polyester and polyglactin 910 sutures, we pooled the data for polyester and polyglactin 910 suture inferior rectus recessions and compared superior rectus recessions with all inferior rectus recessions. The mean shift between postoperative day 1 and 2 months demonstrated an undercorrection shift for both superior rectus and inferior rectus recessions. For superior rectus recessions, a mean undercorrection shift of 1.1D occurred (range, 17.5D-1D undercorrection shift; range, 1D-16D overcorrection shift). In 11 patients there was an overcorrection shift; in 11, an undercorrection shift; and in 8, no shift. For all inferior rectus recessions, a mean
Journal of AAPOS
undercorrection shift of 1.0D occurred (range, 12D-1D of undercorrection shift and range of 0.5D-6D of overcorrection shift). In 9 patients there was an overcorrection shift; in 12, an undercorrection shift; and in 8, no shift. Inferior rectus recessions performed with adjustable sutures thus showed no tendency for overcorrection shifts 2 months postoperatively. We found no significant difference between superior rectus and inferior rectus groups (P 5 0.65). Preoperative deviation played no apparent role in postoperative shift. There was no significant correlation between preoperative vertical deviation and postoperative shift with regression analysis. The P values for the analyses were 0.28 for superior rectus recessions, 0.39 for all inferior rectus recessions, 0.31 for inferior rectus recessions performed with polyglactin 910 sutures, and 0.65 for inferior rectus recessions performed with polyester. Similarly, we compared the preoperative vertical deviation of our study groups with respect to the final outcome of undercorrection, orthotropia, and overcorrection (not shift, but the measurement of strabismus at the 2 month postoperative visit). Regression analysis showed no significant correlation between preoperative deviation and outcome at 2 months, yielding P values of 0.31 for superior rectus recessions, 0.74 for all inferior rectus recessions, 0.78 for inferior rectus recessions performed with polyglactin 910 suture, and 0.93 for inferior rectus recessions performed with polyester. Concurrent horizontal muscle surgery at the time of vertical rectus recession did not influence postoperative shift in superior rectus or inferior rectus groups (superior rectus overcorrection, P 5 0.06; undercorrection, P 5 0.31; inferior rectus overcorrection, P 5 0.10; undercorrection, P 5 0.16).
Discussion Our study found no tendency for overcorrection following unilateral vertical rectus recessions on adjustable suture when TED patients were excluded. Most authors report overcorrection based on final outcome instead of a drift or shift postoperatively when reporting the results of vertical rectus recessions.1,3,5 This study reports the change in measurement, or shift in deviation, between postoperative day 1 after adjustment and 2 months, as well as final outcome measures of over- or undercorrection. In this study, a normal distribution between undercorrection, overcorrection, and orthotropia for both superior rectus
250
Bratton et al
FIG 1. The number of patients with undercorrection, overcorrection and no correction shift at two months after inferior rectus recession with either polyglactin 910 or polyester suture. No significant difference (P 5 0.68) in mean correction shift was found between suture groups.
and inferior rectus muscle recessions is observed (Figure 1). Analyzing the shift in deviation during this critical period also yields an even distribution between overcorrection shifts, undercorrections shifts, and no shift. The present study is limited by its retrospective nature. We did not age-match our patients, and there was a statistically significant age difference between the superior rectus and inferior rectus groups. However, the age range for both groups encompassed young adults to the very elderly, and although the age difference was statistically significant, it was probably not clinically important. The type of suture used for inferior rectus recessions was not randomized, however all patients undergoing inferior rectus recessions after 2005 were managed consistently: polyglactin 910 was used for all patients prior to 2005 and polyester was used in all patients thereafter. If the experience of the surgeons played a role in better outcomes as the study progressed, potentially masking outcome trends, then one would have expected to find an improvement in outcomes for the polyester group compared with the polyglactin 910 group. No such difference was found. Retrospective reviews are often limited by the inability to obtain follow-up data. However, in this study, only 10 of 119 patients were excluded for failure to follow up. Although late overcorrection occurs after 2 months, we did not analyze long-term data because the proposed mechanism at later time points may be different2 than that of shifts in alignment occurring between the time of adjustment and 2 months.1,3,9,18 Factors that limit the interpretation of data in case series of patients undergoing vertical muscle recessions can be found in previous studies, such as combining both fixed and adjustable sutures into the same study population1,3 as well as bilateral and unilateral vertical muscle surgery, and including patients with and without TED.1,6,9,19
Volume 19 Number 3 / June 2015 Inclusion of these variables can lead to contradictory results. For example, Sprunger and Helveston1 studied 67 patients both with and without TED who underwent inferior rectus recession with either adjustable or nonadjustable sutures. Their findings indicated the use of adjustable suture on TED patients would result in an increased incidence of late overcorrection after inferior rectus recession. Similarly, Wright3 observed late overcorrection of the inferior rectus muscle in 7 non-TED patients after use of either fixed or adjustable sutures. In a smaller study (n 5 21), Scotcher and colleagues19 included patients with and without TED who underwent additional vertical muscle surgery at the time of inferior rectus muscle recession and reported undercorrection more frequently than overcorrection. On the other hand, Parsa and colleagues9 suggested that the use of nonabsorbable, adjustable suture decreased the incidence of late overcorrection in a large patient population (n 5 120) that included patients with and without TED. Kerr4 reported an overcorrection shift associated with use of adjustable absorbable suture in patients with TED, adding further evidence that use of nonabsorbable adjustable suture reduced overcorrection. Inferior rectus recession for non-TED patients did not show overcorrection. Even with our strict inclusion criteria we were able to include 59 patients, a large number for a study of adult strabismus surgery.3,5,19 Several theories have been proposed to explain reported late overcorrection after inferior rectus recession. Sprunger and Helveston1 suggested that the unique relationship of the inferior rectus muscle and the inferior oblique and Lockwood’s ligament offers the opportunity for the newly inserted inferior rectus muscle to slip over the globe during horizontal movement of the eyes. These authors also speculated that a restriction of downgaze in thyroid ophthalmopathy in the hypotropic eye may inhibit adhesion of the inferior rectus to the globe and result in late overcorrection of the inferior rectus muscle in TED patients. Chatzistefanou and colleagues6 used multipositional magnetic resonance imaging to confirm that the small arc of contact of the inferior rectus muscle to the globe (with and without TED), insertion of the muscle posterior to the point of tangency (TED only) and the relationship of the muscle to other structures during downgaze all contributed to the increased propensity of the inferior rectus muscle to slip after recession.8 Wright3 used fixed and adjustable sutures in non-TED patients to investigate the mechanism of late overcorrection and proposed that extensive scarring of Lockwood’s ligament resulted in increased slack on the anterior portion of the inferior rectus muscle, causing hyperdeviations postoperatively. As a result of these studies, surgeons have altered their surgical technique in various ways to minimize postoperative overcorrection.7,20 For example, Shikoda and colleagues20 altered the adjustable suture technique by adding a nonabsorbable safety stitch. Cruz and Davitt7 proposed bilateral inferior rectus muscle recession to prevent hypotropia in dysthyroid ophthalmopathy, as
Journal of AAPOS
Volume 19 Number 3 / June 2015 suggested by Sprunger. Several studies using adjustable suture techniques have proposed leaving non-TED patients undercorrected in order to prevent late overcorrection.5,19,20 Parsa and colleagues9 used nonabsorbable sutures and excised Tenon’s tissue between the muscle and the sclera to prevent overcorrection by allowing postoperative inferior rectus adhesion to the globe. Kerr4 supports the use of a nonabsorbable, adjustable suture to reduce overcorrection and the need for undercorrection or bilateral surgery is unnecessary for prevention. In conclusion, our findings indicate that there is no tendency toward overcorrection following unilateral inferior rectus recession on adjustable sutures in patients without TED. Our study therefore does not support altering surgical technique of inferior rectus recession, such as leaving patients undercorrected from the desired postoperative alignment at the time of suture adjustment or using permanent suture material, in non-TED patients. Our conclusions cannot be generalized to patients with TED, fixed sutures, or multiple vertical muscle recessions. References 1. Sprunger DT, Helveston EM. Progressive overcorrection after inferior rectus recession. J Pediatr Ophthalmol Strabismus 1993;30:145-8. 2. Hudson HL, Feldon SE. Late hypercorrection of hypotropia in Graves’ ophthalmopathy: predictive factors. Ophthalmology 1992; 99:356-60. 3. Wright KW. Late overcorrection after inferior rectus recession. Ophthalmology 1996;103:1503-7. 4. Kerr NC. The role of thyroid eye disease and other factors in the overcorrection of hypotropia following unilateral adjustable suture recession of the inferior rectus. Trans Am Ophthalmol Soc 2011;109. 5. Vazquez CW, Mu~ noz M. Overcorrection after adjustable suture suspension-recession of the inferior rectus muscle in non-thyroid eye disease. Binocul Vis Strabismus Q 1999;14:103-6. 6. Chatzistefanou KI, Kushner BJ, Gentry LR. Magnetic resonance imaging of the arc of contact of extraocular muscles: implications regarding the incidence of slipped muscles. J AAPOS 2000;2:84-93.
Journal of AAPOS
Bratton et al
251
7. Cruz OA, Davitt BV. Bilateral inferior rectus muscle recession for correction of hypotropia in dysthyroid ophthalmopathy. J AAPOS 1999;3:157-9. 8. Kushner BJ. Pearls and pointers in adjustable sutures for strabismus surgery. In: Haik BG, ed. Transactions of symposium on oculoplastic surgery, strabismus and pediatric ophthalmology. Thorofare, NJ: SLACK, Inc.; 1990:215-25. 9. Parsa CF, Soltan-Sanjari M, Guyton DL. Non-absorbable suture should be used for adjustable inferior rectus muscle recessions. In: De Faber J-T, ed. 29th European Strabismological Association Meeting: Transactions, Izmir, June 1-4, 2004. New York, NY: Routledge; 2005:81-4. 10. Sharma P, Reinecke RD. Single-stage adjustable strabismus surgery for restrictive strabismus. J AAPOS 2003;7:358-62. 11. Nguyen VT, Park DJ, Levin L, Feldon SE. Correction of restricted extraocular muscle motility in surgical management of strabismus in Graves’ ophthalmopathy. Ophthalmology 2002;109:384-8. 12. Prendiville P, Chopra M, Gauderman WJ, Feldon SE. The role of restricted motility in determining outcomes for vertical strabismus surgery in Graves’ ophthalmology. Ophthalmology 2000;107: 545-9. 13. Thomas SM, Cruz OA. Comparison of two different surgical techniques for the treatment of strabismus in dysthyroid ophthalmopathy. J AAPOS 2007;11:258-61. 14. Fells P, Kousoulides L, Pappa A, Munro P, Lawson J. Extraocular muscle problems in thyroid eye disease. Eye (Lond) 1994;8: 497-505. 15. Lueder GT, Scott WE, Kutschke PJ, Keech RV. Long-term results of adjustable suture surgery for strabismus secondary to thyroid ophthalmopathy. Ophthalmology 1992;99:993-7. 16. Hudson HL, Feldon SE. Late overcorrection of hypotropia in Graves ophthalmopathy. Predictive factors. Ophthalmology 1992; 99:356-60. 17. Oguz V, Yolar M, Pazarli H. Extraocular muscle surgery in dysthyroid orbitomyopathy: influence of previous conditions on surgical results. J Pediatr Ophthalmol Strabismus 2002;39:77-80. 18. Gardner TA, Kennerdell JS. Treatment of dysthyroid myopathy with adjustable suture recession. Ophthalmic Surg 1990;21:519-21. 19. Scotcher SM, O’Flynn EA, Morris RJ. Inferior rectus recession—an effective procedure? Br J Ophthalmol 1997;81:1031-6. 20. Shokida MF, Gabriel J, Sanchez C. Safety stitch: a modification to postoperatively adjustable suture strabismus surgery of the inferior rectus muscle. Binocul Vis Strabismus Q 2007;22:210-15.
To determine whether overcorrection shifts occur after vertical rectus recession on adjustable suture in the absence of thyroid eye disease.
METHODS
The medical records of patients without thyroid eye disease who underwent vertical rectus recession surgery from 2001 to 2008 were retrospectively reviewed for shifts in alignment between suture adjustment at postoperative day 1 and 2 months’ follow-up. Superior rectus and inferior rectus recessions were compared. In addition, we compared the use of a nonabsorbable polyester suture to an absorbable polyglactin 910 suture in nonthyroid patients undergoing inferior rectus recessions.
RESULTS
A total of 59 patients were included (superior rectus, 30; inferior rectus, 29). We found a mean undercorrection shift of 1.1 (range, 17.5D undercorrection to 16D overcorrection) and 1.0D (range, 12D undercorrection shift to 6D overcorrection shift) for superior and inferior rectus recessions, respectively, between 1 day and 2 months postoperatively. There was no trend toward overcorrection following unilateral vertical rectus adjustable suture recessions in patients without thyroid eye disease, suggesting that thyroid myopathy may account for overcorrection shifts seen with this surgery. ( J AAPOS 2015;19:247-251)
CONCLUSIONS
L
ate overcorrection following inferior rectus muscle recession is a well-known problem. Risk factors include thyroid eye disease (TED),1-4 the use of adjustable sutures,1,2,4,5 and unilateral inferior rectus muscle recession.1,4,5 The effect of operations on multiple muscles at the time of inferior rectus recession is unknown.1,4 In addition, late overcorrection following superior rectus recession has not been described.1 Overcorrection typically occurs 4-6 weeks after inferior rectus recession.1,3,6 In fact, it has been shown that patients aligned successfully at 4-6 weeks after surgery maintain
Author affiliations: aDepartment of Ophthalmology, Rocky Mountain Lions Eye Institute, University of Colorado, Aurora; bDepartment of Ophthalmology, Hamilton Eye Institute, University of Tennessee College of Medicine, Memphis; cDepartment of Pediatrics, University of Tennessee College of Medicine, Memphis; dDivision of Ophthalmology, St. Jude Children’s Research Hospital, Memphis, Tennessee; eDivision of Biostatistics and Epidemiology, University of Tennessee College of Medicine, Memphis; fDepartment of Community Health and Preventive Medicine, Morehouse School of Medicine, Atlanta, Georgia Supported by an unrestricted grant to the University of Tennessee from Research to Prevent, Blindness Inc, New York, New York. Presented in part at the Annual Meeting of American Academy of Ophthalmology, San Francisco, California, October 24-27, 2009. Submitted April 11, 2013. Revision accepted March 23, 2015. Correspondence: Natalie C. Kerr, MD, FACS, Hiatt Professor of Ophthalmology, Hamilton Eye Institute, University of Tennessee Health Science Center, 930 Madison, Suite 470, Memphis TN 38163 (email: [email protected]). Copyright Ó 2015 by the American Association for Pediatric Ophthalmology and Strabismus. 1091-8531/$36.00 http://dx.doi.org/10.1016/j.jaapos.2015.03.015
Journal of AAPOS
alignment in long-term follow-up.7 This critical period coincides with the point when absorbable sutures lose their strength,8 and it has been proposed that absorbable sutures may predispose patients to late overcorrection.4,9 However, research on fixed, adjustable, and nonabsorbable sutures in a patient population of thyroid and nonthyroid patients has indicated alternative mechanisms for late overcorrection during this critical period.1,3,4 The present study investigated postoperative shifts in patients without TED following suture adjustment during the critical period between postoperative day 1 and 2 months. We compared inferior rectus recessions to superior rectus recessions to determine whether the inferior rectus is uniquely associated with late overcorrection shifts in patients without TED. We also compared the use of a nonabsorbable polyester fiber suture (Mersilene, Ethicon Inc, Somerville, NJ) to an absorbable polyglactin 910 suture (Vicryl, Ethicon Inc) in nonthyroid patients undergoing inferior rectus recession on adjustable suture to examine the possible role of suture absorption in postoperative shifts.
Subjects and Methods This study was approved by the University of Tennessee Investigational Review Board and complied with the US Health Insurance Portability and Accountability Act of 1996. The medical records of consecutive adult patients (.18 years of age) who underwent vertical rectus muscle recessions at the University of Tennessee Helath Science Center by one of two surgeons
247
248
Bratton et al
(MEH and NCK) between 2001 and 2008 were reviewed retrospectively. Patients with bilateral or multiple vertical muscle recessions, transposition, incomplete follow-up, thyroid eye disease (TED), poor visual acuity (\20/40 best-corrected visual acuity in one or both eyes), or in whom fixed sutures were used were excluded. Patients were not sex or age matched within study groups. The following preoperative data were collected: case number, date of birth, date of first visit, preoperative diagnosis or etiology of the motility disorder, and preoperative measurement of ocular deviation. We recorded the measurement of ocular deviations using the alternate prism and cover test in primary position at distance with best spectacle correction. All patients were evaluated at least twice before surgery at 2-month intervals and were stable at .6 months from any inciting event or change in symptoms before surgery. As shown in Table 1, preoperative diagnoses included paralytic conditions (unrelated to a cranial nerve palsy, including myasthenia gravis, trauma, and drainage implant causing muscle paralysis); cranial nerve palsies; neurologic conditions, including bilateral skew deviation and internuclear ophthalmoplegia; restrictive or orbital disease; nonstrabismic eye surgery; and hypertropia of undetermined etiology. In some patients, preoperative hypertropia had more than one cause. The most common diagnosis for all patients was trochlear nerve palsy. In all cases, the inferior or superior rectus muscle was the only vertical muscle operated. All superior rectus recessions were performed with polyglactin 910 sutures. Inferior rectus recessions were performed with either polyglactin 910 or polyester sutures. All inferior rectus recessions from 2005 through 2008 were performed with polyester sutures. The surgeons used a standard follow-up protocol of 2 weeks and then 2 months after muscle recession in order to record postoperative alignment. The muscle recessed, amount of recession, and suture material used were recorded. In all cases, the surgeon attempted to adjust the patient to orthotropia. Patients were adjusted within the first 24 hours postoperatively. The surgeon measured the deviation of ocular alignment at that time. Postoperative measurements were recorded at 1 day and 2 months, assessed using distance measurements only in the primary position of gaze. The surgeon determined the postoperative measurements in all cases. Overcorrection shift and undercorrection shift were reported if the patient had a prism diopter shift of any value toward overcorrection or undercorrection after the time of adjustment and 2 months postoperatively. A separate measure, resultant undercorrection and overcorrection, was recorded if the patient’s final alignment was $2D of over- or undercorrection. The threshold analyzed, $2D, was chosen because of previous studies. The $5D threshold has common usage in the literature dealing with strabismus surgery in inferior rectus and/or thyroid eye disease studies.4,10-13 The $2D threshold was chosen, first, because some studies used restoration of single binocular vision/relief of diplopia as their measure of success,14-17 and, second, because of limited vertical fusion in adult patients, who are susceptible to postoperative diplopia if the postoperative alignment shifts more than 2D. This approach seemed a conservative and reasonable alignment threshold to allow comparison to other
Volume 19 Number 3 / June 2015 Table 1. Preoperative diagnoses for patients with vertical rectus recessiona Diagnosis
Superior rectus muscle (n 5 30)
Inferior rectus muscle (n 5 29)
Paralytic/CNP/neurological All CNP CN III CN IV CN VI Paralyticb Childhood strabismus Restrictive/orbital diseasec Eye surgeryd Unknown
10 6 1 4 1 4 13 3 6 1
12 11 2 14 0 1 8 8 5 0
CNP, cranial nerve palsy. a Some patients had multiple diagnoses. b Not due to CNP. c Not thyroid eye disease. d Not involving muscles or orbit. studies and to provide further insight into clinically important overcorrection shifts.4 Means with standard deviations were determined for continuous variables; frequencies and corresponding proportions were determined for categorical data. The distributions of primary outcomes were examined for skewness/normality using the Shapiro-Wilk or Kolmogorov-Smirnov statistic. The MannWhitney-Wilcoxon test was used to compare the superior rectus and inferior rectus muscle groups as well as the two suture groups. To assess correlations between outcomes of overcorrection shift, no shift, and undercorrection shift or measurements with preoperative deviation, regression analysis was performed for the study groups. Pearson or Spearman correlation analyses and linear regression analyses were used to assess the association between the preoperative vertical deviation and postoperative shift over muscle type and suture types. A P value of \0.05 was considered statistically significant. All analyses were performed using Statistical Analysis Software (SAS Institute, Cary, NC, v. 9.2).
Results A total of 119 cases of vertical rectus muscle recessions performed on adults during the study period were identified. Of these, 60 did not meet inclusion criteria. The mean age of patients was 55 18 years. The average age of those undergoing superior rectus recession (n 5 30) was 51 18.7 years; of those undergoing inferior rectus recession (n 5 29), 59 15 years. The mean age of the polyester group was 49 14.5 years; of the polyglactin 910 group, 65 12.2 years. The average preoperative vertical deviation for superior rectus and inferior rectus recession groups were 12.62D 7.9D and 14.74D 8.8D, respectively, and are shown in Table 2. There was no significant difference in the average preoperative deviation between these two groups (P 5 0.18). The mean preoperative vertical deviation for inferior rectus muscle was 17.8D 12.0D for the polyester group and 13.2D 6.7D for the polyglactin 910 group; the
Journal of AAPOS
Volume 19 Number 3 / June 2015
Bratton et al
249
Table 2. Preoperative vertical deviation Surgery
Suture type
Mean preoperative vertical deviation, PD
All Inferior rectus muscles Inferior rectus muscle Inferior rectus muscle Superior rectus muscle
Nonabsorbable Polyester fiber Polyglactin 910 (absorbable) Polyglactin 910 (absorbable)
14.74 17.75 13.16 12.62
Range, PD 2.0-40.0 5.0-40.0 2.0-22.5 4.0-30.0
Number of patients 29 10 19 30
PD, prism diopter.
difference between the suture groups was not statistically significant (P 5 0.35). The amount of recession varied based on the individual patient and their preoperative measurement. For the superior rectus muscle group, the mean amount of recession was 7.2 2.4 mm. Overall, the mean inferior rectus recession was 6.3 2.1 mm. There was no significant difference in the amount of recession between superior rectus and inferior rectus muscles (P 5 0.17). For the inferior rectus surgeries performed with polyglactin 910 sutures, the mean recession was 6.3 2.2 mm. The mean recession for inferior rectus recessions performed with polyester sutures was 6.6 1.9 mm). There was also no significant difference in the inferior rectus group between recessions performed with polyglactin 910 and polyester sutures (P 5 0.89). The mean deviation immediately following adjustment (postoperative day 1) for superior rectus recessions was 1.2D of undercorrection (range, 15D of undercorrection to 12.5D of overcorrection) and 0.9D of undercorrection (range, 9D of undercorrection to 4D of overcorrection) for inferior rectus recessions. No significant difference was observed in mean deviations in superior rectus and inferior rectus muscle groups at postoperative day 1 (P 5 0.42). We analyzed the postoperative drift from one day (after adjustment) to 2 months postoperative, comparing polyester (mean undercorrection shift of 1.5D, range 12D undercorrection to 4D overcorrection) and polyglactin 910 (mean undercorrection shift of 0.8D, range 8D undercorrection to 6D overcorrection) in patients with inferior rectus recessions and found no significant difference (P 5 0.87). For surgeries executed with polyglactin 910 sutures, 4 patients had an overcorrection, 7 had an undercorrection, and 8 were orthotropic. For polyester sutures, 5 patients had a shift of overcorrection and 5 had a shift of undercorrection. Because we found no significant difference when inferior rectus recessions were performed with polyester and polyglactin 910 sutures, we pooled the data for polyester and polyglactin 910 suture inferior rectus recessions and compared superior rectus recessions with all inferior rectus recessions. The mean shift between postoperative day 1 and 2 months demonstrated an undercorrection shift for both superior rectus and inferior rectus recessions. For superior rectus recessions, a mean undercorrection shift of 1.1D occurred (range, 17.5D-1D undercorrection shift; range, 1D-16D overcorrection shift). In 11 patients there was an overcorrection shift; in 11, an undercorrection shift; and in 8, no shift. For all inferior rectus recessions, a mean
Journal of AAPOS
undercorrection shift of 1.0D occurred (range, 12D-1D of undercorrection shift and range of 0.5D-6D of overcorrection shift). In 9 patients there was an overcorrection shift; in 12, an undercorrection shift; and in 8, no shift. Inferior rectus recessions performed with adjustable sutures thus showed no tendency for overcorrection shifts 2 months postoperatively. We found no significant difference between superior rectus and inferior rectus groups (P 5 0.65). Preoperative deviation played no apparent role in postoperative shift. There was no significant correlation between preoperative vertical deviation and postoperative shift with regression analysis. The P values for the analyses were 0.28 for superior rectus recessions, 0.39 for all inferior rectus recessions, 0.31 for inferior rectus recessions performed with polyglactin 910 sutures, and 0.65 for inferior rectus recessions performed with polyester. Similarly, we compared the preoperative vertical deviation of our study groups with respect to the final outcome of undercorrection, orthotropia, and overcorrection (not shift, but the measurement of strabismus at the 2 month postoperative visit). Regression analysis showed no significant correlation between preoperative deviation and outcome at 2 months, yielding P values of 0.31 for superior rectus recessions, 0.74 for all inferior rectus recessions, 0.78 for inferior rectus recessions performed with polyglactin 910 suture, and 0.93 for inferior rectus recessions performed with polyester. Concurrent horizontal muscle surgery at the time of vertical rectus recession did not influence postoperative shift in superior rectus or inferior rectus groups (superior rectus overcorrection, P 5 0.06; undercorrection, P 5 0.31; inferior rectus overcorrection, P 5 0.10; undercorrection, P 5 0.16).
Discussion Our study found no tendency for overcorrection following unilateral vertical rectus recessions on adjustable suture when TED patients were excluded. Most authors report overcorrection based on final outcome instead of a drift or shift postoperatively when reporting the results of vertical rectus recessions.1,3,5 This study reports the change in measurement, or shift in deviation, between postoperative day 1 after adjustment and 2 months, as well as final outcome measures of over- or undercorrection. In this study, a normal distribution between undercorrection, overcorrection, and orthotropia for both superior rectus
250
Bratton et al
FIG 1. The number of patients with undercorrection, overcorrection and no correction shift at two months after inferior rectus recession with either polyglactin 910 or polyester suture. No significant difference (P 5 0.68) in mean correction shift was found between suture groups.
and inferior rectus muscle recessions is observed (Figure 1). Analyzing the shift in deviation during this critical period also yields an even distribution between overcorrection shifts, undercorrections shifts, and no shift. The present study is limited by its retrospective nature. We did not age-match our patients, and there was a statistically significant age difference between the superior rectus and inferior rectus groups. However, the age range for both groups encompassed young adults to the very elderly, and although the age difference was statistically significant, it was probably not clinically important. The type of suture used for inferior rectus recessions was not randomized, however all patients undergoing inferior rectus recessions after 2005 were managed consistently: polyglactin 910 was used for all patients prior to 2005 and polyester was used in all patients thereafter. If the experience of the surgeons played a role in better outcomes as the study progressed, potentially masking outcome trends, then one would have expected to find an improvement in outcomes for the polyester group compared with the polyglactin 910 group. No such difference was found. Retrospective reviews are often limited by the inability to obtain follow-up data. However, in this study, only 10 of 119 patients were excluded for failure to follow up. Although late overcorrection occurs after 2 months, we did not analyze long-term data because the proposed mechanism at later time points may be different2 than that of shifts in alignment occurring between the time of adjustment and 2 months.1,3,9,18 Factors that limit the interpretation of data in case series of patients undergoing vertical muscle recessions can be found in previous studies, such as combining both fixed and adjustable sutures into the same study population1,3 as well as bilateral and unilateral vertical muscle surgery, and including patients with and without TED.1,6,9,19
Volume 19 Number 3 / June 2015 Inclusion of these variables can lead to contradictory results. For example, Sprunger and Helveston1 studied 67 patients both with and without TED who underwent inferior rectus recession with either adjustable or nonadjustable sutures. Their findings indicated the use of adjustable suture on TED patients would result in an increased incidence of late overcorrection after inferior rectus recession. Similarly, Wright3 observed late overcorrection of the inferior rectus muscle in 7 non-TED patients after use of either fixed or adjustable sutures. In a smaller study (n 5 21), Scotcher and colleagues19 included patients with and without TED who underwent additional vertical muscle surgery at the time of inferior rectus muscle recession and reported undercorrection more frequently than overcorrection. On the other hand, Parsa and colleagues9 suggested that the use of nonabsorbable, adjustable suture decreased the incidence of late overcorrection in a large patient population (n 5 120) that included patients with and without TED. Kerr4 reported an overcorrection shift associated with use of adjustable absorbable suture in patients with TED, adding further evidence that use of nonabsorbable adjustable suture reduced overcorrection. Inferior rectus recession for non-TED patients did not show overcorrection. Even with our strict inclusion criteria we were able to include 59 patients, a large number for a study of adult strabismus surgery.3,5,19 Several theories have been proposed to explain reported late overcorrection after inferior rectus recession. Sprunger and Helveston1 suggested that the unique relationship of the inferior rectus muscle and the inferior oblique and Lockwood’s ligament offers the opportunity for the newly inserted inferior rectus muscle to slip over the globe during horizontal movement of the eyes. These authors also speculated that a restriction of downgaze in thyroid ophthalmopathy in the hypotropic eye may inhibit adhesion of the inferior rectus to the globe and result in late overcorrection of the inferior rectus muscle in TED patients. Chatzistefanou and colleagues6 used multipositional magnetic resonance imaging to confirm that the small arc of contact of the inferior rectus muscle to the globe (with and without TED), insertion of the muscle posterior to the point of tangency (TED only) and the relationship of the muscle to other structures during downgaze all contributed to the increased propensity of the inferior rectus muscle to slip after recession.8 Wright3 used fixed and adjustable sutures in non-TED patients to investigate the mechanism of late overcorrection and proposed that extensive scarring of Lockwood’s ligament resulted in increased slack on the anterior portion of the inferior rectus muscle, causing hyperdeviations postoperatively. As a result of these studies, surgeons have altered their surgical technique in various ways to minimize postoperative overcorrection.7,20 For example, Shikoda and colleagues20 altered the adjustable suture technique by adding a nonabsorbable safety stitch. Cruz and Davitt7 proposed bilateral inferior rectus muscle recession to prevent hypotropia in dysthyroid ophthalmopathy, as
Journal of AAPOS
Volume 19 Number 3 / June 2015 suggested by Sprunger. Several studies using adjustable suture techniques have proposed leaving non-TED patients undercorrected in order to prevent late overcorrection.5,19,20 Parsa and colleagues9 used nonabsorbable sutures and excised Tenon’s tissue between the muscle and the sclera to prevent overcorrection by allowing postoperative inferior rectus adhesion to the globe. Kerr4 supports the use of a nonabsorbable, adjustable suture to reduce overcorrection and the need for undercorrection or bilateral surgery is unnecessary for prevention. In conclusion, our findings indicate that there is no tendency toward overcorrection following unilateral inferior rectus recession on adjustable sutures in patients without TED. Our study therefore does not support altering surgical technique of inferior rectus recession, such as leaving patients undercorrected from the desired postoperative alignment at the time of suture adjustment or using permanent suture material, in non-TED patients. Our conclusions cannot be generalized to patients with TED, fixed sutures, or multiple vertical muscle recessions. References 1. Sprunger DT, Helveston EM. Progressive overcorrection after inferior rectus recession. J Pediatr Ophthalmol Strabismus 1993;30:145-8. 2. Hudson HL, Feldon SE. Late hypercorrection of hypotropia in Graves’ ophthalmopathy: predictive factors. Ophthalmology 1992; 99:356-60. 3. Wright KW. Late overcorrection after inferior rectus recession. Ophthalmology 1996;103:1503-7. 4. Kerr NC. The role of thyroid eye disease and other factors in the overcorrection of hypotropia following unilateral adjustable suture recession of the inferior rectus. Trans Am Ophthalmol Soc 2011;109. 5. Vazquez CW, Mu~ noz M. Overcorrection after adjustable suture suspension-recession of the inferior rectus muscle in non-thyroid eye disease. Binocul Vis Strabismus Q 1999;14:103-6. 6. Chatzistefanou KI, Kushner BJ, Gentry LR. Magnetic resonance imaging of the arc of contact of extraocular muscles: implications regarding the incidence of slipped muscles. J AAPOS 2000;2:84-93.
Journal of AAPOS
Bratton et al
251
7. Cruz OA, Davitt BV. Bilateral inferior rectus muscle recession for correction of hypotropia in dysthyroid ophthalmopathy. J AAPOS 1999;3:157-9. 8. Kushner BJ. Pearls and pointers in adjustable sutures for strabismus surgery. In: Haik BG, ed. Transactions of symposium on oculoplastic surgery, strabismus and pediatric ophthalmology. Thorofare, NJ: SLACK, Inc.; 1990:215-25. 9. Parsa CF, Soltan-Sanjari M, Guyton DL. Non-absorbable suture should be used for adjustable inferior rectus muscle recessions. In: De Faber J-T, ed. 29th European Strabismological Association Meeting: Transactions, Izmir, June 1-4, 2004. New York, NY: Routledge; 2005:81-4. 10. Sharma P, Reinecke RD. Single-stage adjustable strabismus surgery for restrictive strabismus. J AAPOS 2003;7:358-62. 11. Nguyen VT, Park DJ, Levin L, Feldon SE. Correction of restricted extraocular muscle motility in surgical management of strabismus in Graves’ ophthalmopathy. Ophthalmology 2002;109:384-8. 12. Prendiville P, Chopra M, Gauderman WJ, Feldon SE. The role of restricted motility in determining outcomes for vertical strabismus surgery in Graves’ ophthalmology. Ophthalmology 2000;107: 545-9. 13. Thomas SM, Cruz OA. Comparison of two different surgical techniques for the treatment of strabismus in dysthyroid ophthalmopathy. J AAPOS 2007;11:258-61. 14. Fells P, Kousoulides L, Pappa A, Munro P, Lawson J. Extraocular muscle problems in thyroid eye disease. Eye (Lond) 1994;8: 497-505. 15. Lueder GT, Scott WE, Kutschke PJ, Keech RV. Long-term results of adjustable suture surgery for strabismus secondary to thyroid ophthalmopathy. Ophthalmology 1992;99:993-7. 16. Hudson HL, Feldon SE. Late overcorrection of hypotropia in Graves ophthalmopathy. Predictive factors. Ophthalmology 1992; 99:356-60. 17. Oguz V, Yolar M, Pazarli H. Extraocular muscle surgery in dysthyroid orbitomyopathy: influence of previous conditions on surgical results. J Pediatr Ophthalmol Strabismus 2002;39:77-80. 18. Gardner TA, Kennerdell JS. Treatment of dysthyroid myopathy with adjustable suture recession. Ophthalmic Surg 1990;21:519-21. 19. Scotcher SM, O’Flynn EA, Morris RJ. Inferior rectus recession—an effective procedure? Br J Ophthalmol 1997;81:1031-6. 20. Shokida MF, Gabriel J, Sanchez C. Safety stitch: a modification to postoperatively adjustable suture strabismus surgery of the inferior rectus muscle. Binocul Vis Strabismus Q 2007;22:210-15.
