- Email: [email protected]
Efficacy of inferior oblique anterior transposition placement grading for dissociated vertical deviation
Efficacy of Inferior Oblique Anterior Transposition Placement Grading for Dissociated Vertical Deviation Jonathan H. Engman, BA,1 James E. Egbert, MD,1 C. Gail Summers, MD,1,2 Terri L. Young, MD1,2,3 Objective: To determine if graded anterior placement of a transposed inferior oblique muscle is beneficial for treating variable amounts of dissociated vertical deviation (DVD). Design: Retrospective, consecutive, comparative case series. Participants: Patients who underwent inferior oblique muscle anterior transposition (IOAT) for DVD at one institution between 1991 and 1999. Methods: Chart review. All patients had IOAT procedures of graded placement at 1, 2, or 3 mm anterior to the inferior rectus muscle insertion or standard placement at the level of the inferior rectus muscle insertion. Main Outcome Measures: The effect of graded and standard placement was assessed by measuring the difference between preoperative and postoperative DVD and was defined as DVD correction. The success of surgery was judged by the residual DVD at long-term follow-up of 6 months or more. Excellent, fair, and poor outcomes were defined as residual DVD of 0 to 5 prism diopters (PD), 6 to 12 PD, and 13 or more PD, respectively. Results: Fifty-five patients (106 eyes) underwent IOAT for DVD. The comparison of DVD correction for the standard versus graded group yielded significance at long-term follow-up (P ⫽ 0.001). This result became nonsignificant after adjusting for preoperative DVD (P ⫽ 0.178). The power to detect a 5-PD difference between graded and standard placement was 90%. The surgical success was similar for patients receiving graded and standard IOAT. Patients with 0 to 15 PD of preoperative DVD fared better than those with more than 15 PD of preoperative DVD. Conclusions: This study does not demonstrate increased correction of DVD with graded IOAT versus standard IOAT. We do not recommend placement of the inferior oblique muscle anterior to the inferior rectus muscle insertion. Inferior oblique muscle anterior transposition for DVD was clinically more effective for smaller amounts of DVD. Ophthalmology 2001;108:2045–2050 © 2001 by the American Academy of Ophthalmology. Dissociated vertical deviation (DVD) is a condition marked by elevation, abduction, and extorsion of the nonfixating eye. It is usually a bilateral phenomenon, often asymmetric, and commonly associated with congenital esotropia and other entities that disrupt binocular vision in early life.1 Recently, DVD has been documented in subjects with normal binocular vision.2 The mechanism of DVD is not well understood. Originally received: October 22, 2000. Accepted: June 13, 2001.
Manuscript no. 200542.
1
Department of Ophthalmology, University of Minnesota Academic Health Center, Minneapolis, Minnesota.
2
Department of Pediatrics, University of Minnesota Academic Health Center, Minneapolis, Minnesota.
3
Division of Ophthalmology, Children’s Hospital of Philadelphia and the University of Pennsylvania, Philadelphia, Pennsylvania. Presented in part at the American Academy of Ophthalmology annual meeting, Dallas, Texas, October 2000. Supported in part by an unrestricted grant from Research to Prevent Blindness, Inc., New York, New York. Correspondence to Terri L. Young, MD, Division of Ophthalmology, The Children’s Hospital of Philadelphia, 34th Street and Civic Center Boulevard, Philadelphia, PA 19104-4399. E-mail: [email protected].
© 2001 by the American Academy of Ophthalmology Published by Elsevier Science Inc.
Various surgical methods have been used to correct DVD,3– 8 but none are completely effective. Inferior oblique anterior transposition (IOAT) has been proposed as a treatment for DVD.3,8 –10 Recently, there has been debate regarding the optimal placement of the muscle.3,10 –12 Reports documenting the success of IOAT in the treatment of DVD have differed in the site of placement of the inferior oblique (IO) muscle. It is not known whether placement of the muscle more anterior to the temporal border of the inferior rectus (IR) muscle insertion results in more effective correction of DVD. The purpose of this study was to determine the efficacy of graded IOAT placement for the treatment of DVD. First, the efficacy was assessed by analyzing the quantitative effect of the surgery as measured by change in amount of DVD in individual eyes. Second, efficacy was determined by analyzing the success of surgery for each patient. Placement grading of IOAT to positions anterior to the temporal border of the IR muscle insertion was compared with standard placement at the level of the IR muscle insertion to determine the relative quantitative and qualitative effects. The effect of coexisting clinical characteristics on the amount of correction of DVD was also studied. ISSN 0161-6420/01/$–see front matter PII S0161-6420(01)00801-6
2045
Ophthalmology Volume 108, Number 11, November 2001
Materials and Methods The records of patients who had undergone IOAT for DVD by one of the authors (TLY, JEE, or CGS) between 1991 and 1999 were reviewed. Patients with neurologic, genetic, or craniofacial abnormalities were excluded. Patients with preexisting superior oblique muscle palsy or who underwent concurrent or previous cyclovertical muscle procedures were also excluded. The ability to cooperate with alternate prism cover testing was required for inclusion. Previous and concurrent horizontal muscle surgery, presence of amblyopia, presence of a V pattern, preoperative DVD, age at surgery, degree of inferior oblique overaction (IOOA), and preoperative horizontal tropias were recorded. The site of placement of the IO muscle and concurrent surgery were recorded. After surgery, the amount of DVD, coexisting tropias, and complications were recorded. Ocular deviations and DVD were measured by alternate prism cover test in primary position while fixating at 6 m with the refractive error corrected. Visual acuity was measured with correction using Snellen charts, HOTV matching, or linear Allen figures. Amblyopia was defined as a difference in visual acuity between the eyes of two or more lines. The fixation pattern of preverbal children was measured by induced tropia testing with a 12-prism diopter (PD) vertical prism.13,14 In these children, amblyopia was defined by the presence of an asymmetric response to induced prismatic dissociation. The presence of a V pattern was defined as a difference in horizontal deviation of 15 PD or more as measured by versions in up and down gaze. Inferior oblique overaction was measured in the field of action of the IO muscle and was graded on a 9-point scale (⫺4 to ⫹4) as described by Doughty et al.15 Nondissociated vertical deviations were distinguished from DVD and neutralized with vertical prisms before measuring the DVD. Operations were performed by one of three surgeons (TLY, JEE, or CGS). The indication for surgery was presence of manifest DVD in one or both eyes. The technique used was similar to that described by Elliot and Nankin,16 except that the IO muscle was placed 0, 1, 2, or 3 mm anterior to the insertion of the IR muscle. One eye underwent IOAT to a position 1 mm posterior to the insertion of the IR muscle and, for analytic purposes, was grouped with the eyes that underwent IOAT to the temporal border of the IR muscle insertion. Placement of the IO muscle was left to each surgeon’s judgment. Placement at the temporal border of the IR muscle insertion was defined as standard placement, and placement 1, 2, or 3 mm anterior to the temporal border of the IR muscle insertion was defined as graded placement. Postoperative measurements were recorded at 1 to 6 weeks (early), 6 weeks to 6 months (intermediate), and latest follow-up at 6 months or longer (long term). Three separate time points were chosen in an attempt to assess short- and long-term stability. Postoperative visits at less than 1 week were not included because of the possibility of patient discomfort affecting measurements. For the quantitative analysis of IOAT, the eye was the unit of observation. The quantitative effect was defined as the difference between the preoperative DVD and the DVD at each postoperative observation period. A positive number represents a decrease in the amount of DVD, and a negative number represents an increase in the amount of DVD. Only eyes that underwent IOAT and also had measurable preoperative DVD were included. Some patients received bilateral IOAT even though DVD was manifest in only one eye before surgery. In such cases, the eye that did not have manifest DVD before surgery was not included in the quantitative analysis. A minimum of one postoperative visit was required for inclusion. Descriptive statistics, including the mean and standard devia-
2046
tion, were used to summarize the quantitative change in DVD. The two-sample and the analysis of variance tests were used to compare the mean correction in DVD among the studied variables. One of these variables was the position of IO muscle placement, which was classified into standard or graded. Regression analysis was used to adjust this comparison for the initial preoperative DVD value. Power analysis was performed to measure the ability to detect a difference of 5 PD between the graded and standard IOAT. To determine the success of the surgery qualitatively, a method similar to that described by Burke et al9 was used. This outcome analysis was assessed per patient, and the eye with the greater deviation was used to determine success. A postoperative DVD of 0 to 5 PD was deemed an excellent result, 6 to 12 PD a fair result, and 13 PD or more a poor result. Only those patients who returned for a long-term (ⱖ6 months) visit were included in the qualitative analysis. Results were reported as absolute number of patients and percent of patients in each category.
Results Fifty-five patients underwent IOAT for DVD. Fifty-one patients underwent bilateral IOAT and four patients underwent unilateral IOAT. Demographic and preoperative characteristics are shown in Table 1. Twenty-four eyes were included in the graded group, whereas 70 eyes were in the standard group. The average preoperative DVD was 11.04 ⫾ 6.59 PD for the graded group and 7.47 ⫾ 5.33 PD for the standard group. The average preoperative DVD for all eyes was 8.38 ⫾ 5.86 PD (range, 2–30 PD). The age at surgery was comparable for both groups. In the graded IOAT group, 8, 11, and 5 eyes underwent IOAT to 1, 2, and 3 mm anterior to the IR muscle insertion, respectively. Table 2 shows the results of the quantitative analysis of mechanical effect of IOAT for DVD. Overall, there was DVD correction of 6.9 ⫾ 4.8, 4.8 ⫾ 5.3, and 4.8 ⫾ 6.1 PD at early, intermediate, and long-term follow-up, respectively. For the standard group, there was DVD correction of 6.5 ⫾ 5.1, 3.9 ⫾ 4.1, and 3.3 ⫾ 5.1 PD at early, intermediate, and long-term follow-up, respectively. For the graded group, there was DVD correction of 7.6 ⫾ 4.2, 7.4 ⫾ 7.2, and 8.9 ⫾ 7.1 PD at early, intermediate, and long-term follow-up, respectively. The comparison of amount of DVD correction for the standard versus the graded group yielded P values of 0.450, 0.054, and 0.001 at early, intermediate and long-term follow-up, respectively. All results became nonsignificant after adjusting for preoperative DVD, with P values of 0.780, 0.887, and 0.178 at early, intermediate, and long-term follow-up, respectively. The power to detect a 5-PD difference between the graded and standard groups was 90%. For preoperative DVD of 0 to 15 PD, there was DVD correction of 5.6 ⫾ 3.5, 3.6 ⫾ 3.6, and 3.1 ⫾ 4.1 PD at early, intermediate, and long-term follow-up, respectively. For preoperative DVD of more than 15 PD, there was DVD correction of 14.2 ⫾ 4.7, 12.9 ⫾ 7.3, and 14.6 ⫾ 6.9 PD at early, intermediate, and long-term follow-up, respectively. There was a statistically significant difference between the 0 to 15 PD and the more than 15 PD groups at each follow-up period (P ⬍ 0.001, P ⫽ 0.003, and P ⬍ 0.001). At early follow-up, the amount of preoperative IOOA had a statistically significant effect on DVD correction. There were three patients with DVD only, without IOOA. For eyes with ⫹1 or ⫹2 IOOA, there was 5.6 ⫾ 3.6 PD of correction, and for eyes with ⫹3 or ⫹4 IOOA, there was 10.2 ⫾ 5.8 PD of correction (P ⬍ 0.001). This statistically significant effect disappeared at intermediate and long-term follow-up (P ⫽ 0.368 and P ⫽ 0.082). No statistically significant effect was observed when comparing unilateral with bilateral cases, the presence or absence of amblyopia, V pattern,
Engman et al 䡠 Inferior Oblique Anterior Transposition Placement Grading Table 1. Preoperative Demographics and Characteristics of Patients and Eyes Included in the Qualitative and Quantitative Analyses Qualitative Analysis Characteristic No. of patients No. of eyes Underwent previous horizontal operation Underwent concurrent horizontal operation Preoperative amblyopia Preoperative V pattern Preoperative DVD (PD) Age at surgery (yrs) Average placement of IO insertion (mm anterior to IR insertion)
Cumulative Patients
Patients Who Returned for Long-term Follow-up
55 110 32 patients 47 patients 7 patients 13 patients 7.16 ⫾ 6.17 6.8 ⫾ 5.1
37 74 24 patients 29 patients 7 patients 9 patients 7.05 ⫾ 6.37 7.2 ⫾ 5.34
0.42
0.44
Quantitative Analysis Graded Group
Standard Group
24 14 eyes 16 eyes 1 eye 6 eyes 11.04 ⫾ 6.59 6.9 ⫾ 4.8 1.87 8 at 1 mm 11 at 2 mm 5 at 3 mm
70 36 eyes 53 eyes 5 eyes 18 eyes 7.47 ⫾ 5.33 7.0 ⫾ 5.3 0
DVD ⫽ dissociated vertical deviation; IO ⫽ inferior oblique muscle; IR ⫽ inferior rectus muscle; PD ⫽ prism diopters.
previous or concurrent surgery, refractive error, or preoperative esotropia and exotropia (Table 2). Residual DVD amounts at each postoperative period for standard and graded groups are shown in Table 3. For the standard group, the residual DVD was 2.2 ⫾ 3.7, 2.6 ⫾ 3.6, and 3.9 ⫾ 4.2 PD at early, intermediate, and long-term follow-up, respectively. For the graded group, the residual DVD was 2.4 ⫾ 2.9, 4.1 ⫾ 5.0, and 3.6 ⫾ 3.6 PD at early, intermediate, and long-term follow-up, respectively. The long-term success of IOAT for DVD is shown in Table 4. Thirty-six patients returned for long-term follow-up (mean, 20.7 months; range, 6 –71 months) and were included in the qualitative analysis. Calculation of statistical significance was not performed because of the insufficient number of patients. Before surgery, 11 patients (30%) had 0 to 5 PD, 15 patients (42%) had 6 to 12 PD, and 10 patients (28%) had 13 PD or more of DVD. Overall outcomes determined for the 36 patients who received IOAT for DVD in one or both eyes were 56% with an excellent outcome (0 –5 PD), 36% with a fair outcome (6 –12 PD), and 8% with a poor outcome (ⱖ13 PD). For patients who had 0 to 15 PD of preoperative DVD, 64% had an excellent outcome, 29% had a fair outcome, and 7% had a poor outcome. For patients with preoperative DVD of more than 15 PD, 25% had an excellent outcome, 63% had a fair outcome, and 12% had a poor outcome. For patients with preoperative DVD of 6 to 15 PD, 53%, 35%, and 12% had excellent, fair, and poor outcomes, respectively. For patients who underwent a graded procedure in one or both eyes, 53%, 40%, and 7% had excellent, fair, and poor outcomes, respectively. For patients who underwent standard IOAT, 57%, 33%, and 10% had excellent, fair, and poor results, respectively. At long-term follow-up, three patients who underwent bilateral operations experienced a consecutive hypotropia in primary position. Two of these patients underwent bilateral IOAT to the level of the IR muscle insertion and had 4 PD of hypotropia. One underwent a bilateral, asymmetric, graded IOAT to a position 2 mm anterior to the IR muscle insertion and had 6 PD of hypotropia. None of the four patients receiving unilateral IOAT experienced a new-onset vertical deviation in primary gaze.
Discussion Inferior oblique muscle anterior transposition has been shown to be effective in the treatment of DVD. The site for
optimal placement of the inferior oblique muscle is still under consideration. Mims and Wood3 performed IOAT to a position 2 to 4 mm anterior to the lateral end of the IR muscle insertion for IOOA and subsequently discovered a beneficial reduction or prevention of DVD. Kratz et al11 hypothesized that anterior grading of the posterior fibers of the IO muscle based on severity of DVD would improve the effect of the anteriorization. They found that the graded group showed less residual DVD than the standard group, with 1.15 PD and 2.44 PD (P ⬍ 0.02), respectively. However, their positions were more posterior (1 mm posterior to 1 mm anterior to IR muscle insertion) than those of Mims and Wood and included only the posterior fibers of the IO muscle. For DVD and IOOA, Seawright and Gole10 performed graded IOAT to positions located between 2 mm posterior to and 2 mm anterior to the temporal border of the IR muscle insertion. Grading was based on the presence or amount of preoperative IOOA, V pattern, hypertropia, and DVD. The aggregate results of 21 patients were reported; however, the efficacy of grading was not evaluated. Guemes and Wright12 reported the results of graded IOAT 0 to 4 mm posterior to the IR muscle insertion for four patients with DVD and IOOA. All four patients had between 1 and 10 PD of improvement in DVD after surgery. This study suggests that the amount of DVD correction from IOAT is dependent on the amount of preoperative DVD, and more anterior placement of the IO muscle does not increase its effectiveness. The position of IO muscle placement (standard vs. graded) appeared to have a significant effect on DVD correction at long-term follow-up, but after adjusting for preoperative DVD, the statistical significance disappeared. The sample size was sufficient to detect a difference of 5 PD with a power of 90%. Eyes with more than 15 PD of preoperative DVD had a statistically significant greater correction in DVD when compared with eyes with 0 to 15 PD of preoperative DVD. The eyes that underwent graded procedures had an average of 3.57 PD more preoperative DVD than eyes that underwent standard procedures. Postoperative DVD measurements were similar between the two groups at all observation points.
2047
Ophthalmology Volume 108, Number 11, November 2001 Table 2. Analysis of Quantitative Effect of Inferior Oblique Muscle Anterior Transposition for Dissociated Vertical Deviation Results Reported as Amount of Dissociated Vertical Deviation Change from Preoperative to Postoperative Follow-up Times
Variable Analyzed Position of IO Placement Standard ⫹ graded Standard Graded After adjusting for preoperative DVD Preoperative DVD 0–15 PD ⬎15 PD Presence of V pattern Present Not present IOAT Unilateral Bilateral Amblyopia Present Not present Previous horizontal surgery Yes No Concurrent horizontal surgery Yes No Refractive error Myopic Hyperopic Plano Preoperative deviations Esotropic Exotropic Preoperative IOOA 0 ⫹1 or ⫹2 ⫹3 or ⫹4
Dissociated Vertical Deviation Correction at Early Follow-up (PD)
Dissociated Vertical Deviation Correction at Intermediate Follow-up (PD)
Dissociated Vertical Deviation Correction at Long-term Follow-up (PD)
6.9 ⫾ 4.8 (n ⫽ 59) 6.5 ⫾ 5.1 (n ⫽ 39) 7.6 ⫾ 4.2 (n ⫽ 20) P ⫽ 0.450 P ⫽ 0.780
4.8 ⫾ 5.3 (n ⫽ 77) 3.9 ⫾ 4.1 (n ⫽ 57) 7.4 ⫾ 7.2 (n ⫽ 20) P ⫽ 0.054 P ⫽ 0.887
4.8 ⫾ 6.1 (n ⫽ 61) 3.3 ⫾ 5.1 (n ⫽ 45) 8.9 ⫾ 7.1 (n ⫽ 16) P ⫽ 0.001 P ⫽ 0.178
5.6 ⫾ 3.5 (n ⫽ 50) 14.2 ⫾ 4.7 (n ⫽ 9) P ⬍ 0.001
3.6 ⫾ 3.6 (n ⫽ 67) 12.9 ⫾ 7.3 (n ⫽ 10) P ⫽ 0.003
3.1 ⫾ 4.1 (n ⫽ 52) 14.6 ⫾ 6.9 (n ⫽ 9) P ⬍ 0.001
7.1 ⫾ 6.1 (n ⫽ 14) 6.8 ⫾ 4.4 (n ⫽ 45) P ⫽ 0.867
3.8 ⫾ 4.0 (n ⫽ 20) 5.1 ⫾ 5.6 (n ⫽ 57) P ⫽ 0.312
2.5 ⫾ 5.4 (n ⫽ 16) 5.6 ⫾ 6.2 (n ⫽ 45) P ⫽ 0.084
6.2 ⫾ 1.7 (n ⫽ 4) 6.9 ⫾ 5.0 (n ⫽ 55) P ⫽ 0.788
7.5 ⫾ 0.7 (n ⫽ 2) 4.7 ⫾ 5.3 (n ⫽ 75) P ⫽ 0.462
5.0 ⫾ 1.7 (n ⫽ 3) 4.8 ⫾ 6.3 (n ⫽ 58) P ⫽ 0.948
5.8 ⫾ 2.2 (n ⫽ 6) 7.0 ⫾ 5.0 (n ⫽ 53) P ⫽ 0.578
5.4 ⫾ 3.8 (n ⫽ 5) 4.7 ⫾ 5.4 (n ⫽ 72) P ⫽ 0.787
5.6 ⫾ 2.6 (n ⫽ 5) 4.7 ⫾ 6.4 (n ⫽ 56) P ⫽ 0.755
6.5 ⫾ 4.3 (n ⫽ 30) 7.3 ⫾ 5.3 (n ⫽ 29) P ⫽ 0.541
3.9 ⫾ 5.2 (n ⫽ 42) 5.8 ⫾ 5.2 (n ⫽ 35) P ⫽ 0.120
3.9 ⫾ 5.6 (n ⫽ 35) 5.9 ⫾ 6.7 (n ⫽ 26) P ⫽ 0.208
7.1 ⫾ 5.2 (n ⫽ 44) 6.3 ⫾ 3.6 (n ⫽ 15) P ⫽ 0.614
4.3 ⫾ 4.8 (n ⫽ 54) 5.9 ⫾ 6.1 (n ⫽ 23) P ⫽ 0.219
4.8 ⫾ 5.3 (n ⫽ 40) 4.8 ⫾ 7.6 (n ⫽ 21) P ⫽ 0.994
5.6 ⫾ 2.9 (n ⫽ 7) 7.0 ⫾ 5.0 (n ⫽ 49) 8.3 ⫾ 5.8 (n ⫽ 3) P ⫽ 0.674
4.5 ⫾ 3.2 (n ⫽ 8) 4.7 ⫾ 5.4 (n ⫽ 65) 6.5 ⫾ 6.8 (n ⫽ 4) P ⫽ 0.797
2.0 ⫾ 7.4 (n ⫽ 5) 4.8 ⫾ 5.9 (n ⫽ 54) 12.0 ⫾ 8.5 (n ⫽ 2) P ⫽ 0.149
6.7 ⫾ 4.9 (n ⫽ 49) 7.7 ⫾ 4.9 (n ⫽ 9) P ⫽ 0.593
5.2 ⫾ 5.0 (n ⫽ 56) 3.6 ⫾ 6.0 (n ⫽ 20) P ⫽ 0.244
5.3 ⫾ 5.8 (n ⫽ 43) 3.4 ⫾ 7.1 (n ⫽ 17) P ⫽ 0.283
3.3 ⫾ 2.3 (n ⫽ 3) 5.6 ⫾ 3.6 (n ⫽ 38) 10.2 ⫾ 5.8 (n ⫽ 18) P ⬍ 0.001*
3.5 ⫾ 2.1 (n ⫽ 2) 4.2 ⫾ 4.4 (n ⫽ 49) 6.0 ⫾ 6.6 (n ⫽ 26) P ⫽ 0.368
⫺1.5 ⫾ 4.9 (n ⫽ 2) 3.9 ⫾ 5.0 (n ⫽ 36) 6.7 ⫾ 7.3 (n ⫽ 23) P ⫽ 0.082
DVD ⫽ dissociated vertical deviation; IO ⫽ inferior oblique muscle; IOAT ⫽ inferior oblique anterior transposition; IOOA ⫽ inferior oblique overaction; PD ⫽ prism diopters. *For preoperative inferior oblique overaction, the group of ⫹3 or ⫹4 is significantly different from the group ⫹1 or ⫹2 for DVD correction.
Patients with preoperative DVD of more than 15 PD fared worse when comparing qualitative outcomes. Only 25% of patients with preoperative DVD of more than 15 PD experienced an excellent outcome as compared with 64% of patients with 0 to 15 PD of preoperative DVD. However, 30% of the patients who returned for long-term follow-up had 0 to 5 PD of DVD before surgery, and therefore were likely to have an excellent outcome. When patients with 6 to 15 PD of DVD before surgery were analyzed, 53% had an excellent outcome. The surgical success analysis is limited by the small sample size and by using the basic deviation as
2048
the unit of measurement. The small sample size prevented a meaningful statistical analysis. The basic deviation does not differentiate between the phoric and tropic component, and hence ignores the amount of deviation controlled by the patient. If a patient has improved control of residual DVD after surgery, this desirable result would not be reflected in outcomes defined by the basic deviation. Inferior oblique muscle anterior transposition can cause consecutive hypotropia and elevation deficits.15–19 These untoward effects may be more pronounced with greater amounts of IOAT. Guemes and Wright12 suggested avoid-
Engman et al 䡠 Inferior Oblique Anterior Transposition Placement Grading Table 3. Residual Dissociated Vertical Deviation at Early, Intermediate, and Long-term Follow-up for Eyes that Underwent Inferior Oblique Anterior Transposition for Dissociated Vertical Deviation Surgical Method
Residual Dissociated Vertical Deviation at Early Follow-up (PD)
Residual Dissociated Vertical Deviation at Intermediate Follow-up (PD)
Residual Dissociated Vertical Deviation at Long-term Follow-up (PD)
Standard IOAT Graded IOAT Standard ⫹ graded
2.2 ⫾ 3.7 (n ⫽ 39) 2.4 ⫾ 2.9 (n ⫽ 20) 2.3 ⫾ 3.4 (n ⫽ 59)
2.6 ⫾ 3.6 (n ⫽ 57) 4.1 ⫾ 5.0 (n ⫽ 20) 3.0 ⫾ 4.0 (n ⫽ 77)
3.9 ⫾ 4.2 (n ⫽ 45) 3.6 ⫾ 3.6 (n ⫽ 16) 3.8 ⫾ 4.1 (n ⫽ 61)
DVD ⫽ dissociated vertical deviation; IOAT ⫽ inferior oblique anterior transposition; PD ⫽ prism diopters.
ance of anteriorization of the posterior fibers of the IO muscle to avoid creating a J deformity, thereby minimizing the likelihood of limited elevation. Mims and Wood19 described an antielevation syndrome after IOAT and concurred with Guemes and Wright12 that the position of the posterior IO muscle fibers is important in preventing an elevation deficit. Only 3 of 55 patients experienced a consecutive hypotropic deviation in our series, one after a graded IOAT. It does not appear that the incidence of this occurrence is high after IOAT, independent of grading. Despite this small number of patients with a subsequent hypodeviation in primary position, many other patients may have had limited elevation in upgaze. This was not assessed in this retrospective study. In conclusion, this retrospective study of surgical outTable 4. Success of Inferior Oblique Anterior Transposition for Dissociated Vertical Deviation at Long-term (ⱖ6 Months) Follow-up by Patient According to Preoperative Characteristic or Surgical Method
Preoperative Characteristic or Surgical Method Preoperative distribution of patients by DVD (36 patients) Overall success of IOAT (36 patients) Preoperative DVD of 0–15 PD (28 patients) Preoperative DVD of ⬎15 PD (8 patients) Received graded IOAT (15 patients) Received standard IOAT (21 patients) Received unilateral IOAT (3 patients) Received bilateral IOAT (33 patients) V pattern before surgery (9 patients) No V pattern before surgery (27 patients) Amblyopia before surgery (7 patients) No amblyopia before surgery (29 patients)
Number and Percent of Patients with Excellent, Fair, or Poor Results Excellent (0–5 PD)
Fair (6–12 PD)
Poor (ⱖ13 PD)
11 (30%)
15 (42%)
10 (28%)
20 (56%)
13 (36%)
3 (8%)
18 (64%)
8 (29%)
2 (7%)
2 (25%)
5 (63%)
1 (12%)
8 (53%)
6 (40%)
1 (7%)
12 (57%)
7 (33%)
2 (10%)
1 (33%)
2 (67%)
0 (0%)
19 (58%)
11 (33%)
3 (9%)
5 (56%)
3 (33%)
1 (11%)
15 (56%)
10 (37%)
2 (7%)
4 (57%)
3 (43%)
0 (0%)
16 (55%)
10 (35%)
3 (10%)
DVD ⫽ dissociated vertical deviation; IOAT ⫽ inferior oblique anterior transposition; PD ⫽ prism diopters.
comes for patients with DVD showed that IOAT was more effective for smaller amounts of DVD. Because we could not demonstrate increased correction of DVD with graded IOAT, we do not see an advantage in placing the IO muscle anterior to the temporal border of the IR muscle insertion versus placing the IO muscle at the temporal border of the IR muscle insertion. Acknowledgments: The authors thank Bruce Lindgren and Xu Wuo for their assistance with statistical interpretation of data.
References 1. Velez G. Dissociated vertical deviation. Graefes Arch Clin Exp Ophthalmol 1988;226:117– 8. 2. Van Rijn LJ, ten Tusscher MPM, de Jong I, Hendrikse F. Asymmetrical vertical phorias indicating dissociated vertical deviation in subjects with normal binocular vision. Vision Res 1998;38:2973– 8. 3. Mims JL III, Wood RC. Bilateral anterior transposition of the inferior obliques. Arch Ophthalmol 1989;107:41– 4. 4. Schwartz T, Scott W. Unilateral superior rectus muscle recession for the treatment of dissociated vertical deviation. J Pediatr Ophthalmol Strabismus 1991;28:219 –22. 5. Duncan LB, von Noorden GK. Surgical results in dissociated vertical deviations. J Pediatr Ophthalmol Strabismus 1984;21: 25–7. 6. Caputo A, Raab EL, Ruttum M. Management of dissociated vertical deviation. J Pediatr Ophthalmol Strabismus 1999;36: 208 –12. 7. Sargent RA. Dissociated hypertropia: surgical treatment. Ophthalmology 1979;86:1428 – 40. 8. Jampolsky A. Management of vertical strabismus. Trans New Orleans Acad Ophthalmol 1986;34:141–71. 9. Burke JP, Scott WE, Kutschke PJ. Anterior transposition of the inferior oblique muscle for dissociated vertical deviation. Ophthalmology 1993;100:245–50. 10. Seawright AA, Gole GA. Results of anterior transposition of the inferior oblique. Aust N Z J Ophthalmol 1996;24:339 – 45. 11. Kratz RE, Rogers GL, Bremer DL, Leguire LE. Anterior tendon displacement of the inferior oblique for DVD. J Pediatr Ophthalmol Strabismus 1989;26:212–7. 12. Guemes A, Wright KW. Effect of graded anterior transposition of the inferior oblique muscle on versions and vertical deviation in primary position. J AAPOS 1998;2:201– 6. 13. Wright KW, Walonker F, Edelman P. 10-Diopter fixation test for amblyopia. Arch Ophthalmol 1981;99:1242– 6. 14. Wright KW, Edelman PM, Walonker F, Yiu S. Reliability of fixation preference testing in diagnosing amblyopia. Arch Ophthalmol 1986;104:549 –53. 15. Doughty DD, Lennarson LW, Scott WE. A graphic portrayal of versions. Perspect Ophthalmol 1978;2:55–9.
2049
Ophthalmology Volume 108, Number 11, November 2001 16. Elliot RL, Nankin SJ. Anterior transposition of the inferior oblique. J Pediatr Ophthalmol Strabismus 1981;18:35– 8. 17. Bremer DL, Rogers GL, Quick LD. Primary-position hypotropia after anterior transposition of the inferior oblique. Arch Ophthalmol 1986;104:229 –32.
18. Kushner BJ. Restriction of elevation in abduction after inferior oblique anteriorization. J AAPOS 1997;1:55– 62. 19. Mims JL III, Wood RC. Antielevation syndrome after bilateral anterior transposition of the inferior oblique muscles: incidence and prevention. J AAPOS 1999;3:333– 6.
Discussion by Judith E. Gurland, MD In the 1970s, the surgery that Dr. Philip Knapp would do for what he and Sally Moore called a “double hyper” was to approach the inferior oblique from an inferotemporal incision, cut the muscle free close to its insertion, tamponade the bleeding, and rotate the eye down and out. The deviation subsequently was gone more often than not. What Dr. Knapp called a double hyper, and what we now call dissociated vertical deviation (DVD) has been referred to by different terms. These include dissociated vertical deviation, dissociated vertical divergence, alternating sursumduction, and dissociated double hypertropia. Along with lots of names for the condition, a variety of surgical procedures to cure it have been proposed. These include superior rectus recession with or without posterior fixation suture, inferior rectus resection, inferior oblique disinsertion, inferior oblique recession, and inferior oblique anterior transposition. Engman et al have undertaken an admirable task in attempting to evaluate the efficacy of graded inferior oblique transposition to correct DVD. They tried to correlate the amount of transposition with the amount of DVD corrected and looked at the maintainance of the correction over a period of time. In theory, moving the insertion of the inferior oblique to a position adjacent to the lateral border of the inferior rectus should convert the action of the muscle to that of a depressor. In theory also, the amount of correction of the vertical deviation should change in a consistent way depending on the new position of the inferior oblique muscle. The amount of correction achieved with surgery on horizontal rectus muscles varies directly with the amount of surgery that is performed. Extrapolating data from published nomograms, we can see, for example, that a medial rectus muscle that is recessed 3 mm—as may be the case for a 15 prism diopter (PD) esotropia— From the Bronx Lebanon Hospital Center, Albert Einstein College of Medicine, Bronx, New York. Address correspondence to Judith E. Gurland, MD, Bronx Lebanon Hospital Center, Albert Einstein College of Medicine, 1650 Selwyn Avenue, Bronx, NY 10457.
2050
will correct approximately 2.5 PD of esotropia per millimeter recessed. The medial rectus recessed 6 mm for a 50-PD esotropia can be expected to correct more than 4 D per millimeter recessed. We likewise should be able to predict a surgical response for moving the inferior oblique muscle. The study by Engman et al is a retrospective, nonrandomized, study carried out by chart review. The setting was a universitybased practice. There are many drawbacks inherent in such a study. The preoperative DVD measurements—which are imprecise to begin with—were obtained by different observers, and not necessarily by the person who subsequently performed the operation. There were at least three operating surgeons and an unknown number of participating residents. The final surgical procedure was left to the discretion of the attending surgeon. Ideally, a prospective study on a larger number of patients, divided in a randomized fashion, would yield more scientifically valid results. Perhaps analysis of pooled data from previously published reports would provide useful information. The cause and mechanism of DVD are poorly understood, and there may be coexisting problems— other muscle imbalance such as hypertropia, oblique dysfunction, V patterns, amblyopia, or other neurologic, genetic, or craniofacial abnormalities—that confound a simple analysis of the deviation. Despite these drawbacks, lessons can be learned from such a study. Increased correction of DVD with graded transposition was not demonstrated to be efficacious, but worked about as well as the standard procedure. Dr. Knapp seemed to know which patients would respond to a free tenotomy, and perhaps after we are able to determine the cause of this condition, we will be able to tailor the surgery to the cause. The authors have attempted to bring a scientific approach to an imprecise situation. They have demonstrated that a number of techniques can achieve similar results; what seems to matter is that whatever procedure you chose to perform, do it well. We will keep looking for the ideal, predictable procedure; that is the nature of our surgical personalities. However, I believe that DVD, which may not have a uniform cause in all cases, can be expected to have many treatments.
Manuscript no. 200542.
1
Department of Ophthalmology, University of Minnesota Academic Health Center, Minneapolis, Minnesota.
2
Department of Pediatrics, University of Minnesota Academic Health Center, Minneapolis, Minnesota.
3
Division of Ophthalmology, Children’s Hospital of Philadelphia and the University of Pennsylvania, Philadelphia, Pennsylvania. Presented in part at the American Academy of Ophthalmology annual meeting, Dallas, Texas, October 2000. Supported in part by an unrestricted grant from Research to Prevent Blindness, Inc., New York, New York. Correspondence to Terri L. Young, MD, Division of Ophthalmology, The Children’s Hospital of Philadelphia, 34th Street and Civic Center Boulevard, Philadelphia, PA 19104-4399. E-mail: [email protected].
© 2001 by the American Academy of Ophthalmology Published by Elsevier Science Inc.
Various surgical methods have been used to correct DVD,3– 8 but none are completely effective. Inferior oblique anterior transposition (IOAT) has been proposed as a treatment for DVD.3,8 –10 Recently, there has been debate regarding the optimal placement of the muscle.3,10 –12 Reports documenting the success of IOAT in the treatment of DVD have differed in the site of placement of the inferior oblique (IO) muscle. It is not known whether placement of the muscle more anterior to the temporal border of the inferior rectus (IR) muscle insertion results in more effective correction of DVD. The purpose of this study was to determine the efficacy of graded IOAT placement for the treatment of DVD. First, the efficacy was assessed by analyzing the quantitative effect of the surgery as measured by change in amount of DVD in individual eyes. Second, efficacy was determined by analyzing the success of surgery for each patient. Placement grading of IOAT to positions anterior to the temporal border of the IR muscle insertion was compared with standard placement at the level of the IR muscle insertion to determine the relative quantitative and qualitative effects. The effect of coexisting clinical characteristics on the amount of correction of DVD was also studied. ISSN 0161-6420/01/$–see front matter PII S0161-6420(01)00801-6
2045
Ophthalmology Volume 108, Number 11, November 2001
Materials and Methods The records of patients who had undergone IOAT for DVD by one of the authors (TLY, JEE, or CGS) between 1991 and 1999 were reviewed. Patients with neurologic, genetic, or craniofacial abnormalities were excluded. Patients with preexisting superior oblique muscle palsy or who underwent concurrent or previous cyclovertical muscle procedures were also excluded. The ability to cooperate with alternate prism cover testing was required for inclusion. Previous and concurrent horizontal muscle surgery, presence of amblyopia, presence of a V pattern, preoperative DVD, age at surgery, degree of inferior oblique overaction (IOOA), and preoperative horizontal tropias were recorded. The site of placement of the IO muscle and concurrent surgery were recorded. After surgery, the amount of DVD, coexisting tropias, and complications were recorded. Ocular deviations and DVD were measured by alternate prism cover test in primary position while fixating at 6 m with the refractive error corrected. Visual acuity was measured with correction using Snellen charts, HOTV matching, or linear Allen figures. Amblyopia was defined as a difference in visual acuity between the eyes of two or more lines. The fixation pattern of preverbal children was measured by induced tropia testing with a 12-prism diopter (PD) vertical prism.13,14 In these children, amblyopia was defined by the presence of an asymmetric response to induced prismatic dissociation. The presence of a V pattern was defined as a difference in horizontal deviation of 15 PD or more as measured by versions in up and down gaze. Inferior oblique overaction was measured in the field of action of the IO muscle and was graded on a 9-point scale (⫺4 to ⫹4) as described by Doughty et al.15 Nondissociated vertical deviations were distinguished from DVD and neutralized with vertical prisms before measuring the DVD. Operations were performed by one of three surgeons (TLY, JEE, or CGS). The indication for surgery was presence of manifest DVD in one or both eyes. The technique used was similar to that described by Elliot and Nankin,16 except that the IO muscle was placed 0, 1, 2, or 3 mm anterior to the insertion of the IR muscle. One eye underwent IOAT to a position 1 mm posterior to the insertion of the IR muscle and, for analytic purposes, was grouped with the eyes that underwent IOAT to the temporal border of the IR muscle insertion. Placement of the IO muscle was left to each surgeon’s judgment. Placement at the temporal border of the IR muscle insertion was defined as standard placement, and placement 1, 2, or 3 mm anterior to the temporal border of the IR muscle insertion was defined as graded placement. Postoperative measurements were recorded at 1 to 6 weeks (early), 6 weeks to 6 months (intermediate), and latest follow-up at 6 months or longer (long term). Three separate time points were chosen in an attempt to assess short- and long-term stability. Postoperative visits at less than 1 week were not included because of the possibility of patient discomfort affecting measurements. For the quantitative analysis of IOAT, the eye was the unit of observation. The quantitative effect was defined as the difference between the preoperative DVD and the DVD at each postoperative observation period. A positive number represents a decrease in the amount of DVD, and a negative number represents an increase in the amount of DVD. Only eyes that underwent IOAT and also had measurable preoperative DVD were included. Some patients received bilateral IOAT even though DVD was manifest in only one eye before surgery. In such cases, the eye that did not have manifest DVD before surgery was not included in the quantitative analysis. A minimum of one postoperative visit was required for inclusion. Descriptive statistics, including the mean and standard devia-
2046
tion, were used to summarize the quantitative change in DVD. The two-sample and the analysis of variance tests were used to compare the mean correction in DVD among the studied variables. One of these variables was the position of IO muscle placement, which was classified into standard or graded. Regression analysis was used to adjust this comparison for the initial preoperative DVD value. Power analysis was performed to measure the ability to detect a difference of 5 PD between the graded and standard IOAT. To determine the success of the surgery qualitatively, a method similar to that described by Burke et al9 was used. This outcome analysis was assessed per patient, and the eye with the greater deviation was used to determine success. A postoperative DVD of 0 to 5 PD was deemed an excellent result, 6 to 12 PD a fair result, and 13 PD or more a poor result. Only those patients who returned for a long-term (ⱖ6 months) visit were included in the qualitative analysis. Results were reported as absolute number of patients and percent of patients in each category.
Results Fifty-five patients underwent IOAT for DVD. Fifty-one patients underwent bilateral IOAT and four patients underwent unilateral IOAT. Demographic and preoperative characteristics are shown in Table 1. Twenty-four eyes were included in the graded group, whereas 70 eyes were in the standard group. The average preoperative DVD was 11.04 ⫾ 6.59 PD for the graded group and 7.47 ⫾ 5.33 PD for the standard group. The average preoperative DVD for all eyes was 8.38 ⫾ 5.86 PD (range, 2–30 PD). The age at surgery was comparable for both groups. In the graded IOAT group, 8, 11, and 5 eyes underwent IOAT to 1, 2, and 3 mm anterior to the IR muscle insertion, respectively. Table 2 shows the results of the quantitative analysis of mechanical effect of IOAT for DVD. Overall, there was DVD correction of 6.9 ⫾ 4.8, 4.8 ⫾ 5.3, and 4.8 ⫾ 6.1 PD at early, intermediate, and long-term follow-up, respectively. For the standard group, there was DVD correction of 6.5 ⫾ 5.1, 3.9 ⫾ 4.1, and 3.3 ⫾ 5.1 PD at early, intermediate, and long-term follow-up, respectively. For the graded group, there was DVD correction of 7.6 ⫾ 4.2, 7.4 ⫾ 7.2, and 8.9 ⫾ 7.1 PD at early, intermediate, and long-term follow-up, respectively. The comparison of amount of DVD correction for the standard versus the graded group yielded P values of 0.450, 0.054, and 0.001 at early, intermediate and long-term follow-up, respectively. All results became nonsignificant after adjusting for preoperative DVD, with P values of 0.780, 0.887, and 0.178 at early, intermediate, and long-term follow-up, respectively. The power to detect a 5-PD difference between the graded and standard groups was 90%. For preoperative DVD of 0 to 15 PD, there was DVD correction of 5.6 ⫾ 3.5, 3.6 ⫾ 3.6, and 3.1 ⫾ 4.1 PD at early, intermediate, and long-term follow-up, respectively. For preoperative DVD of more than 15 PD, there was DVD correction of 14.2 ⫾ 4.7, 12.9 ⫾ 7.3, and 14.6 ⫾ 6.9 PD at early, intermediate, and long-term follow-up, respectively. There was a statistically significant difference between the 0 to 15 PD and the more than 15 PD groups at each follow-up period (P ⬍ 0.001, P ⫽ 0.003, and P ⬍ 0.001). At early follow-up, the amount of preoperative IOOA had a statistically significant effect on DVD correction. There were three patients with DVD only, without IOOA. For eyes with ⫹1 or ⫹2 IOOA, there was 5.6 ⫾ 3.6 PD of correction, and for eyes with ⫹3 or ⫹4 IOOA, there was 10.2 ⫾ 5.8 PD of correction (P ⬍ 0.001). This statistically significant effect disappeared at intermediate and long-term follow-up (P ⫽ 0.368 and P ⫽ 0.082). No statistically significant effect was observed when comparing unilateral with bilateral cases, the presence or absence of amblyopia, V pattern,
Engman et al 䡠 Inferior Oblique Anterior Transposition Placement Grading Table 1. Preoperative Demographics and Characteristics of Patients and Eyes Included in the Qualitative and Quantitative Analyses Qualitative Analysis Characteristic No. of patients No. of eyes Underwent previous horizontal operation Underwent concurrent horizontal operation Preoperative amblyopia Preoperative V pattern Preoperative DVD (PD) Age at surgery (yrs) Average placement of IO insertion (mm anterior to IR insertion)
Cumulative Patients
Patients Who Returned for Long-term Follow-up
55 110 32 patients 47 patients 7 patients 13 patients 7.16 ⫾ 6.17 6.8 ⫾ 5.1
37 74 24 patients 29 patients 7 patients 9 patients 7.05 ⫾ 6.37 7.2 ⫾ 5.34
0.42
0.44
Quantitative Analysis Graded Group
Standard Group
24 14 eyes 16 eyes 1 eye 6 eyes 11.04 ⫾ 6.59 6.9 ⫾ 4.8 1.87 8 at 1 mm 11 at 2 mm 5 at 3 mm
70 36 eyes 53 eyes 5 eyes 18 eyes 7.47 ⫾ 5.33 7.0 ⫾ 5.3 0
DVD ⫽ dissociated vertical deviation; IO ⫽ inferior oblique muscle; IR ⫽ inferior rectus muscle; PD ⫽ prism diopters.
previous or concurrent surgery, refractive error, or preoperative esotropia and exotropia (Table 2). Residual DVD amounts at each postoperative period for standard and graded groups are shown in Table 3. For the standard group, the residual DVD was 2.2 ⫾ 3.7, 2.6 ⫾ 3.6, and 3.9 ⫾ 4.2 PD at early, intermediate, and long-term follow-up, respectively. For the graded group, the residual DVD was 2.4 ⫾ 2.9, 4.1 ⫾ 5.0, and 3.6 ⫾ 3.6 PD at early, intermediate, and long-term follow-up, respectively. The long-term success of IOAT for DVD is shown in Table 4. Thirty-six patients returned for long-term follow-up (mean, 20.7 months; range, 6 –71 months) and were included in the qualitative analysis. Calculation of statistical significance was not performed because of the insufficient number of patients. Before surgery, 11 patients (30%) had 0 to 5 PD, 15 patients (42%) had 6 to 12 PD, and 10 patients (28%) had 13 PD or more of DVD. Overall outcomes determined for the 36 patients who received IOAT for DVD in one or both eyes were 56% with an excellent outcome (0 –5 PD), 36% with a fair outcome (6 –12 PD), and 8% with a poor outcome (ⱖ13 PD). For patients who had 0 to 15 PD of preoperative DVD, 64% had an excellent outcome, 29% had a fair outcome, and 7% had a poor outcome. For patients with preoperative DVD of more than 15 PD, 25% had an excellent outcome, 63% had a fair outcome, and 12% had a poor outcome. For patients with preoperative DVD of 6 to 15 PD, 53%, 35%, and 12% had excellent, fair, and poor outcomes, respectively. For patients who underwent a graded procedure in one or both eyes, 53%, 40%, and 7% had excellent, fair, and poor outcomes, respectively. For patients who underwent standard IOAT, 57%, 33%, and 10% had excellent, fair, and poor results, respectively. At long-term follow-up, three patients who underwent bilateral operations experienced a consecutive hypotropia in primary position. Two of these patients underwent bilateral IOAT to the level of the IR muscle insertion and had 4 PD of hypotropia. One underwent a bilateral, asymmetric, graded IOAT to a position 2 mm anterior to the IR muscle insertion and had 6 PD of hypotropia. None of the four patients receiving unilateral IOAT experienced a new-onset vertical deviation in primary gaze.
Discussion Inferior oblique muscle anterior transposition has been shown to be effective in the treatment of DVD. The site for
optimal placement of the inferior oblique muscle is still under consideration. Mims and Wood3 performed IOAT to a position 2 to 4 mm anterior to the lateral end of the IR muscle insertion for IOOA and subsequently discovered a beneficial reduction or prevention of DVD. Kratz et al11 hypothesized that anterior grading of the posterior fibers of the IO muscle based on severity of DVD would improve the effect of the anteriorization. They found that the graded group showed less residual DVD than the standard group, with 1.15 PD and 2.44 PD (P ⬍ 0.02), respectively. However, their positions were more posterior (1 mm posterior to 1 mm anterior to IR muscle insertion) than those of Mims and Wood and included only the posterior fibers of the IO muscle. For DVD and IOOA, Seawright and Gole10 performed graded IOAT to positions located between 2 mm posterior to and 2 mm anterior to the temporal border of the IR muscle insertion. Grading was based on the presence or amount of preoperative IOOA, V pattern, hypertropia, and DVD. The aggregate results of 21 patients were reported; however, the efficacy of grading was not evaluated. Guemes and Wright12 reported the results of graded IOAT 0 to 4 mm posterior to the IR muscle insertion for four patients with DVD and IOOA. All four patients had between 1 and 10 PD of improvement in DVD after surgery. This study suggests that the amount of DVD correction from IOAT is dependent on the amount of preoperative DVD, and more anterior placement of the IO muscle does not increase its effectiveness. The position of IO muscle placement (standard vs. graded) appeared to have a significant effect on DVD correction at long-term follow-up, but after adjusting for preoperative DVD, the statistical significance disappeared. The sample size was sufficient to detect a difference of 5 PD with a power of 90%. Eyes with more than 15 PD of preoperative DVD had a statistically significant greater correction in DVD when compared with eyes with 0 to 15 PD of preoperative DVD. The eyes that underwent graded procedures had an average of 3.57 PD more preoperative DVD than eyes that underwent standard procedures. Postoperative DVD measurements were similar between the two groups at all observation points.
2047
Ophthalmology Volume 108, Number 11, November 2001 Table 2. Analysis of Quantitative Effect of Inferior Oblique Muscle Anterior Transposition for Dissociated Vertical Deviation Results Reported as Amount of Dissociated Vertical Deviation Change from Preoperative to Postoperative Follow-up Times
Variable Analyzed Position of IO Placement Standard ⫹ graded Standard Graded After adjusting for preoperative DVD Preoperative DVD 0–15 PD ⬎15 PD Presence of V pattern Present Not present IOAT Unilateral Bilateral Amblyopia Present Not present Previous horizontal surgery Yes No Concurrent horizontal surgery Yes No Refractive error Myopic Hyperopic Plano Preoperative deviations Esotropic Exotropic Preoperative IOOA 0 ⫹1 or ⫹2 ⫹3 or ⫹4
Dissociated Vertical Deviation Correction at Early Follow-up (PD)
Dissociated Vertical Deviation Correction at Intermediate Follow-up (PD)
Dissociated Vertical Deviation Correction at Long-term Follow-up (PD)
6.9 ⫾ 4.8 (n ⫽ 59) 6.5 ⫾ 5.1 (n ⫽ 39) 7.6 ⫾ 4.2 (n ⫽ 20) P ⫽ 0.450 P ⫽ 0.780
4.8 ⫾ 5.3 (n ⫽ 77) 3.9 ⫾ 4.1 (n ⫽ 57) 7.4 ⫾ 7.2 (n ⫽ 20) P ⫽ 0.054 P ⫽ 0.887
4.8 ⫾ 6.1 (n ⫽ 61) 3.3 ⫾ 5.1 (n ⫽ 45) 8.9 ⫾ 7.1 (n ⫽ 16) P ⫽ 0.001 P ⫽ 0.178
5.6 ⫾ 3.5 (n ⫽ 50) 14.2 ⫾ 4.7 (n ⫽ 9) P ⬍ 0.001
3.6 ⫾ 3.6 (n ⫽ 67) 12.9 ⫾ 7.3 (n ⫽ 10) P ⫽ 0.003
3.1 ⫾ 4.1 (n ⫽ 52) 14.6 ⫾ 6.9 (n ⫽ 9) P ⬍ 0.001
7.1 ⫾ 6.1 (n ⫽ 14) 6.8 ⫾ 4.4 (n ⫽ 45) P ⫽ 0.867
3.8 ⫾ 4.0 (n ⫽ 20) 5.1 ⫾ 5.6 (n ⫽ 57) P ⫽ 0.312
2.5 ⫾ 5.4 (n ⫽ 16) 5.6 ⫾ 6.2 (n ⫽ 45) P ⫽ 0.084
6.2 ⫾ 1.7 (n ⫽ 4) 6.9 ⫾ 5.0 (n ⫽ 55) P ⫽ 0.788
7.5 ⫾ 0.7 (n ⫽ 2) 4.7 ⫾ 5.3 (n ⫽ 75) P ⫽ 0.462
5.0 ⫾ 1.7 (n ⫽ 3) 4.8 ⫾ 6.3 (n ⫽ 58) P ⫽ 0.948
5.8 ⫾ 2.2 (n ⫽ 6) 7.0 ⫾ 5.0 (n ⫽ 53) P ⫽ 0.578
5.4 ⫾ 3.8 (n ⫽ 5) 4.7 ⫾ 5.4 (n ⫽ 72) P ⫽ 0.787
5.6 ⫾ 2.6 (n ⫽ 5) 4.7 ⫾ 6.4 (n ⫽ 56) P ⫽ 0.755
6.5 ⫾ 4.3 (n ⫽ 30) 7.3 ⫾ 5.3 (n ⫽ 29) P ⫽ 0.541
3.9 ⫾ 5.2 (n ⫽ 42) 5.8 ⫾ 5.2 (n ⫽ 35) P ⫽ 0.120
3.9 ⫾ 5.6 (n ⫽ 35) 5.9 ⫾ 6.7 (n ⫽ 26) P ⫽ 0.208
7.1 ⫾ 5.2 (n ⫽ 44) 6.3 ⫾ 3.6 (n ⫽ 15) P ⫽ 0.614
4.3 ⫾ 4.8 (n ⫽ 54) 5.9 ⫾ 6.1 (n ⫽ 23) P ⫽ 0.219
4.8 ⫾ 5.3 (n ⫽ 40) 4.8 ⫾ 7.6 (n ⫽ 21) P ⫽ 0.994
5.6 ⫾ 2.9 (n ⫽ 7) 7.0 ⫾ 5.0 (n ⫽ 49) 8.3 ⫾ 5.8 (n ⫽ 3) P ⫽ 0.674
4.5 ⫾ 3.2 (n ⫽ 8) 4.7 ⫾ 5.4 (n ⫽ 65) 6.5 ⫾ 6.8 (n ⫽ 4) P ⫽ 0.797
2.0 ⫾ 7.4 (n ⫽ 5) 4.8 ⫾ 5.9 (n ⫽ 54) 12.0 ⫾ 8.5 (n ⫽ 2) P ⫽ 0.149
6.7 ⫾ 4.9 (n ⫽ 49) 7.7 ⫾ 4.9 (n ⫽ 9) P ⫽ 0.593
5.2 ⫾ 5.0 (n ⫽ 56) 3.6 ⫾ 6.0 (n ⫽ 20) P ⫽ 0.244
5.3 ⫾ 5.8 (n ⫽ 43) 3.4 ⫾ 7.1 (n ⫽ 17) P ⫽ 0.283
3.3 ⫾ 2.3 (n ⫽ 3) 5.6 ⫾ 3.6 (n ⫽ 38) 10.2 ⫾ 5.8 (n ⫽ 18) P ⬍ 0.001*
3.5 ⫾ 2.1 (n ⫽ 2) 4.2 ⫾ 4.4 (n ⫽ 49) 6.0 ⫾ 6.6 (n ⫽ 26) P ⫽ 0.368
⫺1.5 ⫾ 4.9 (n ⫽ 2) 3.9 ⫾ 5.0 (n ⫽ 36) 6.7 ⫾ 7.3 (n ⫽ 23) P ⫽ 0.082
DVD ⫽ dissociated vertical deviation; IO ⫽ inferior oblique muscle; IOAT ⫽ inferior oblique anterior transposition; IOOA ⫽ inferior oblique overaction; PD ⫽ prism diopters. *For preoperative inferior oblique overaction, the group of ⫹3 or ⫹4 is significantly different from the group ⫹1 or ⫹2 for DVD correction.
Patients with preoperative DVD of more than 15 PD fared worse when comparing qualitative outcomes. Only 25% of patients with preoperative DVD of more than 15 PD experienced an excellent outcome as compared with 64% of patients with 0 to 15 PD of preoperative DVD. However, 30% of the patients who returned for long-term follow-up had 0 to 5 PD of DVD before surgery, and therefore were likely to have an excellent outcome. When patients with 6 to 15 PD of DVD before surgery were analyzed, 53% had an excellent outcome. The surgical success analysis is limited by the small sample size and by using the basic deviation as
2048
the unit of measurement. The small sample size prevented a meaningful statistical analysis. The basic deviation does not differentiate between the phoric and tropic component, and hence ignores the amount of deviation controlled by the patient. If a patient has improved control of residual DVD after surgery, this desirable result would not be reflected in outcomes defined by the basic deviation. Inferior oblique muscle anterior transposition can cause consecutive hypotropia and elevation deficits.15–19 These untoward effects may be more pronounced with greater amounts of IOAT. Guemes and Wright12 suggested avoid-
Engman et al 䡠 Inferior Oblique Anterior Transposition Placement Grading Table 3. Residual Dissociated Vertical Deviation at Early, Intermediate, and Long-term Follow-up for Eyes that Underwent Inferior Oblique Anterior Transposition for Dissociated Vertical Deviation Surgical Method
Residual Dissociated Vertical Deviation at Early Follow-up (PD)
Residual Dissociated Vertical Deviation at Intermediate Follow-up (PD)
Residual Dissociated Vertical Deviation at Long-term Follow-up (PD)
Standard IOAT Graded IOAT Standard ⫹ graded
2.2 ⫾ 3.7 (n ⫽ 39) 2.4 ⫾ 2.9 (n ⫽ 20) 2.3 ⫾ 3.4 (n ⫽ 59)
2.6 ⫾ 3.6 (n ⫽ 57) 4.1 ⫾ 5.0 (n ⫽ 20) 3.0 ⫾ 4.0 (n ⫽ 77)
3.9 ⫾ 4.2 (n ⫽ 45) 3.6 ⫾ 3.6 (n ⫽ 16) 3.8 ⫾ 4.1 (n ⫽ 61)
DVD ⫽ dissociated vertical deviation; IOAT ⫽ inferior oblique anterior transposition; PD ⫽ prism diopters.
ance of anteriorization of the posterior fibers of the IO muscle to avoid creating a J deformity, thereby minimizing the likelihood of limited elevation. Mims and Wood19 described an antielevation syndrome after IOAT and concurred with Guemes and Wright12 that the position of the posterior IO muscle fibers is important in preventing an elevation deficit. Only 3 of 55 patients experienced a consecutive hypotropic deviation in our series, one after a graded IOAT. It does not appear that the incidence of this occurrence is high after IOAT, independent of grading. Despite this small number of patients with a subsequent hypodeviation in primary position, many other patients may have had limited elevation in upgaze. This was not assessed in this retrospective study. In conclusion, this retrospective study of surgical outTable 4. Success of Inferior Oblique Anterior Transposition for Dissociated Vertical Deviation at Long-term (ⱖ6 Months) Follow-up by Patient According to Preoperative Characteristic or Surgical Method
Preoperative Characteristic or Surgical Method Preoperative distribution of patients by DVD (36 patients) Overall success of IOAT (36 patients) Preoperative DVD of 0–15 PD (28 patients) Preoperative DVD of ⬎15 PD (8 patients) Received graded IOAT (15 patients) Received standard IOAT (21 patients) Received unilateral IOAT (3 patients) Received bilateral IOAT (33 patients) V pattern before surgery (9 patients) No V pattern before surgery (27 patients) Amblyopia before surgery (7 patients) No amblyopia before surgery (29 patients)
Number and Percent of Patients with Excellent, Fair, or Poor Results Excellent (0–5 PD)
Fair (6–12 PD)
Poor (ⱖ13 PD)
11 (30%)
15 (42%)
10 (28%)
20 (56%)
13 (36%)
3 (8%)
18 (64%)
8 (29%)
2 (7%)
2 (25%)
5 (63%)
1 (12%)
8 (53%)
6 (40%)
1 (7%)
12 (57%)
7 (33%)
2 (10%)
1 (33%)
2 (67%)
0 (0%)
19 (58%)
11 (33%)
3 (9%)
5 (56%)
3 (33%)
1 (11%)
15 (56%)
10 (37%)
2 (7%)
4 (57%)
3 (43%)
0 (0%)
16 (55%)
10 (35%)
3 (10%)
DVD ⫽ dissociated vertical deviation; IOAT ⫽ inferior oblique anterior transposition; PD ⫽ prism diopters.
comes for patients with DVD showed that IOAT was more effective for smaller amounts of DVD. Because we could not demonstrate increased correction of DVD with graded IOAT, we do not see an advantage in placing the IO muscle anterior to the temporal border of the IR muscle insertion versus placing the IO muscle at the temporal border of the IR muscle insertion. Acknowledgments: The authors thank Bruce Lindgren and Xu Wuo for their assistance with statistical interpretation of data.
References 1. Velez G. Dissociated vertical deviation. Graefes Arch Clin Exp Ophthalmol 1988;226:117– 8. 2. Van Rijn LJ, ten Tusscher MPM, de Jong I, Hendrikse F. Asymmetrical vertical phorias indicating dissociated vertical deviation in subjects with normal binocular vision. Vision Res 1998;38:2973– 8. 3. Mims JL III, Wood RC. Bilateral anterior transposition of the inferior obliques. Arch Ophthalmol 1989;107:41– 4. 4. Schwartz T, Scott W. Unilateral superior rectus muscle recession for the treatment of dissociated vertical deviation. J Pediatr Ophthalmol Strabismus 1991;28:219 –22. 5. Duncan LB, von Noorden GK. Surgical results in dissociated vertical deviations. J Pediatr Ophthalmol Strabismus 1984;21: 25–7. 6. Caputo A, Raab EL, Ruttum M. Management of dissociated vertical deviation. J Pediatr Ophthalmol Strabismus 1999;36: 208 –12. 7. Sargent RA. Dissociated hypertropia: surgical treatment. Ophthalmology 1979;86:1428 – 40. 8. Jampolsky A. Management of vertical strabismus. Trans New Orleans Acad Ophthalmol 1986;34:141–71. 9. Burke JP, Scott WE, Kutschke PJ. Anterior transposition of the inferior oblique muscle for dissociated vertical deviation. Ophthalmology 1993;100:245–50. 10. Seawright AA, Gole GA. Results of anterior transposition of the inferior oblique. Aust N Z J Ophthalmol 1996;24:339 – 45. 11. Kratz RE, Rogers GL, Bremer DL, Leguire LE. Anterior tendon displacement of the inferior oblique for DVD. J Pediatr Ophthalmol Strabismus 1989;26:212–7. 12. Guemes A, Wright KW. Effect of graded anterior transposition of the inferior oblique muscle on versions and vertical deviation in primary position. J AAPOS 1998;2:201– 6. 13. Wright KW, Walonker F, Edelman P. 10-Diopter fixation test for amblyopia. Arch Ophthalmol 1981;99:1242– 6. 14. Wright KW, Edelman PM, Walonker F, Yiu S. Reliability of fixation preference testing in diagnosing amblyopia. Arch Ophthalmol 1986;104:549 –53. 15. Doughty DD, Lennarson LW, Scott WE. A graphic portrayal of versions. Perspect Ophthalmol 1978;2:55–9.
2049
Ophthalmology Volume 108, Number 11, November 2001 16. Elliot RL, Nankin SJ. Anterior transposition of the inferior oblique. J Pediatr Ophthalmol Strabismus 1981;18:35– 8. 17. Bremer DL, Rogers GL, Quick LD. Primary-position hypotropia after anterior transposition of the inferior oblique. Arch Ophthalmol 1986;104:229 –32.
18. Kushner BJ. Restriction of elevation in abduction after inferior oblique anteriorization. J AAPOS 1997;1:55– 62. 19. Mims JL III, Wood RC. Antielevation syndrome after bilateral anterior transposition of the inferior oblique muscles: incidence and prevention. J AAPOS 1999;3:333– 6.
Discussion by Judith E. Gurland, MD In the 1970s, the surgery that Dr. Philip Knapp would do for what he and Sally Moore called a “double hyper” was to approach the inferior oblique from an inferotemporal incision, cut the muscle free close to its insertion, tamponade the bleeding, and rotate the eye down and out. The deviation subsequently was gone more often than not. What Dr. Knapp called a double hyper, and what we now call dissociated vertical deviation (DVD) has been referred to by different terms. These include dissociated vertical deviation, dissociated vertical divergence, alternating sursumduction, and dissociated double hypertropia. Along with lots of names for the condition, a variety of surgical procedures to cure it have been proposed. These include superior rectus recession with or without posterior fixation suture, inferior rectus resection, inferior oblique disinsertion, inferior oblique recession, and inferior oblique anterior transposition. Engman et al have undertaken an admirable task in attempting to evaluate the efficacy of graded inferior oblique transposition to correct DVD. They tried to correlate the amount of transposition with the amount of DVD corrected and looked at the maintainance of the correction over a period of time. In theory, moving the insertion of the inferior oblique to a position adjacent to the lateral border of the inferior rectus should convert the action of the muscle to that of a depressor. In theory also, the amount of correction of the vertical deviation should change in a consistent way depending on the new position of the inferior oblique muscle. The amount of correction achieved with surgery on horizontal rectus muscles varies directly with the amount of surgery that is performed. Extrapolating data from published nomograms, we can see, for example, that a medial rectus muscle that is recessed 3 mm—as may be the case for a 15 prism diopter (PD) esotropia— From the Bronx Lebanon Hospital Center, Albert Einstein College of Medicine, Bronx, New York. Address correspondence to Judith E. Gurland, MD, Bronx Lebanon Hospital Center, Albert Einstein College of Medicine, 1650 Selwyn Avenue, Bronx, NY 10457.
2050
will correct approximately 2.5 PD of esotropia per millimeter recessed. The medial rectus recessed 6 mm for a 50-PD esotropia can be expected to correct more than 4 D per millimeter recessed. We likewise should be able to predict a surgical response for moving the inferior oblique muscle. The study by Engman et al is a retrospective, nonrandomized, study carried out by chart review. The setting was a universitybased practice. There are many drawbacks inherent in such a study. The preoperative DVD measurements—which are imprecise to begin with—were obtained by different observers, and not necessarily by the person who subsequently performed the operation. There were at least three operating surgeons and an unknown number of participating residents. The final surgical procedure was left to the discretion of the attending surgeon. Ideally, a prospective study on a larger number of patients, divided in a randomized fashion, would yield more scientifically valid results. Perhaps analysis of pooled data from previously published reports would provide useful information. The cause and mechanism of DVD are poorly understood, and there may be coexisting problems— other muscle imbalance such as hypertropia, oblique dysfunction, V patterns, amblyopia, or other neurologic, genetic, or craniofacial abnormalities—that confound a simple analysis of the deviation. Despite these drawbacks, lessons can be learned from such a study. Increased correction of DVD with graded transposition was not demonstrated to be efficacious, but worked about as well as the standard procedure. Dr. Knapp seemed to know which patients would respond to a free tenotomy, and perhaps after we are able to determine the cause of this condition, we will be able to tailor the surgery to the cause. The authors have attempted to bring a scientific approach to an imprecise situation. They have demonstrated that a number of techniques can achieve similar results; what seems to matter is that whatever procedure you chose to perform, do it well. We will keep looking for the ideal, predictable procedure; that is the nature of our surgical personalities. However, I believe that DVD, which may not have a uniform cause in all cases, can be expected to have many treatments.