Journal of Neuro-Ophthalmology, December 2014, Volume 34, Issue 4

Measurement of Ocular Cyclotorsion in Superior Oblique Palsy Using a Single Maddox Rod

Measurement of Ocular Cyclotorsion in Superior Oblique Palsy Using a Single Maddox Rod Yehoshua Almog, MD, Arie Y. Nemet, MD, Yokrat Ton, MD Background: To describe a modified technique to measure ocular cyclotorsion using only 1 Maddox rod and compare it with the traditional double Maddox rod test (DMR). Methods: Ocular cyclotorsion was prospectively measured in 48 consecutive patients with superior oblique palsy. Four measurements were taken; 2 using 2 Maddox rods: (A) a red rod in front of the affected eye and a white rod in front of the fellow eye and (B) red rod in front of the sound eye and red in front of the affected eye. Two additional sets of measurements were taken using 1 red Maddox rod in front of 1 eye only, where horizontal lines in the room served as a reference: (C) the red rod in front of the affected eye and (D) the red rod in front of the sound eye. Results: Mean ocular cyclotorsion in 48 patients was 6.7°, 6.8°, 6.6°, and 6.1° in measures A, B, C, and D, respec-tively. The agreement between A and C did not differ statis-tically, whereas the agreement between B and D was statistically different (P , 0.0001). There was high repeat-ability in both comparisons. Conclusions: Ocular cyclotorsion may be measured using a single Maddox rod instead of two. It is simpler to conduct than the DMR. Journal of Neuro-Ophthalmology 2014;34:362-365 doi: 10.1097/WNO.0000000000000148 © 2014 by North American Neuro-Ophthalmology Society Avariety of objective and subjective methods are used to measure cyclodeviation of the eye (1). Maddox rod tests with or without prisms are often used by clinicians to assess the angle of horizontal or vertical deviation at different gaze positions for the purpose of diagnosing and quantifying the severity of a paretic extraocular muscle. The double Maddox rod test (DMR) is frequently used as a subjective method of measuring ocular cyclotorsion of the eye. It complements the Parks-Bielchowsky 3-step test for identifying a paretic vertical muscle and quantifying torsional diplopia (2). One red Maddox rod is placed in front of 1 eye, and a white rod in front of the other eye on the same axis. If cyclodeviation exists, the red and white streaks appear to the patient on different axes. By adjusting the axis settings of one of the 2 Maddox rods until they appear parallel to the patient, the degree of cyclodeviation can be measured. Alternatively, the patient can be asked to adjust the axes of both Maddox rods until they appear horizontal and, therefore, parallel. We report a modified method for measuring cyclodeviation using a single Maddox rod (SMR). We are unaware of published reports of SMR for measuring cyclodeviation. PATIENTS AND METHODS Ocular cyclodeviation was prospectively measured in 48 consecutive cases of superior oblique palsy examined in the Neuro-ophthalmology Outpatient Clinic. Diagnosis of superior oblique palsy was based on measurements of the vertical deviation at different positions of gaze using the Maddox rod and prism technique. Patients were included in this study if they had vertical deviation that increased on gaze contralateral to the hypertropic eye and on head tilt to the same side of the hypertropic eye. Patients with less than 3° of cyclotorsion on the DMR test were excluded. Measurements of cyclodeviation were performed on each patient by 1 unmasked, experienced neuro-ophthalmologist (Y.A.). The DMR test was done first, followed by SMR test. The DMR test was done in a darkened room. A red rod was placed in test spectacles vertically in front of 1 eye and a white rod was placed in front of the fellow eye. The patient was asked to look at the fixation light and to rotate the red rod until the red line appeared parallel to the white reference line. Ocular deviation was determined by reading Department of Ophthalmology, Meir Medical Center and Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel. Presented at the North American Neuro-ophthalmology Society meeting, February 25-March 2, 2006, Tucson, AZ. The authors report no conflicts of interest. Address correspondence to Arie Y. Nemet, MD, Meir Medical Center, 59 Tchernichovsky Street, Kfar Saba 44281, Israel; E-mail: nemet. [e-mail] 362 Almog et al: J Neuro-Ophthalmol 2014; 34: 362-365 Original Contribution Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. the number of degrees on the test spectacles. Two measure-ments were taken with the DMR test. In the first, the red rod was placed in front of the affected eye (measurement A) and then in front of the sound eye (measurement B). For the SMR test, only 1 red Maddox rod was used. Existing lines in the surrounding nondarkened room (e.g., windowsill, doorpost, picture frame) served as a horizontal reference. Cyclodeviation was measured using the red rod in front of the affected eye (measurement C) and in front of the sound eye (measurement D). The study was approved by the local institutional review board. Statistical Analysis Descriptive statistics were presented as mean ± standard deviation (SD). Measures were taken from the same subject using 2 different techniques. The intracorrelation coeffi-cient (ICC), which measures the consistency of repeated measures, was calculated. ICC values closer to 1 indicate better measurement consistency. ICC values .0.75 repre-sent "excellent reliability." The mean difference between the techniques was calculated together with the SD of the difference, standard error of the mean and 95% confidence interval (CI). A 95% CI that does not include zero indicates a significant systematic change in the mean between the 2 techniques. The distributional assumption of the differences was examined using the Kolmogorov-Smirnov test. In addition, the Bland-Altman visual plot and calculation were done to visualize the level of agreement between the measurements. The mean difference was organized as a y-axis reference line with 2 additional lines at 1.96 · SD as the limit of agree-ment. The differences in the measurements were plotted against their corresponding means. A random scatter of points above and below the reference line was considered acceptable as a measure of repeatability. The coefficient of variation (CV) was computed. This makes it possible to determine the typical variation between 2 test occasions. A lower percent CV is associated with higher repeatability. Differences between the 2 tests were analyzed using a Wilcoxon rank test. RESULTS A total of 48 consecutive patients diagnosed with superior oblique palsy were included. The mean age of the patients was 57.5 ± 12.5 years (range, 27-78 years) and 69% were men. Each patient was examined using 4 measurements (A, B, C, and D). The mean difference between the DMR and SMR tests was normally distributed (Kolmogorov-Smirnov test). Clinical and demographic data are summarized in Table 1. The agreement between A and C did not differ statistically, whereas the agreement between B and D was statistically significant. However, the CV% was lower, which indicates high repeatability in both comparisons. Similar results were found in correlations between techni-ques, which were highly correlated in both comparisons (P , 0.0001). Table 2 presents the mean ocular cyclotor-sion in all the combinations and the agreement between techniques A-C and B-D. Bland-Altman plots evaluating agreement between A and C, B and D are shown in Figure 1. The difference between the measurements of techniques A and C was a maximum of 1° in 82.6% of cases and in 72.9% of cases between techniques B and D. Comparison of cyclodeviation meas-urements between the DMR and SMR tests are presented in Figure 2. TABLE 1. Clinical and demographic data of patients with superior oblique palsy (N = 48) Age (yr) Mean ± SD 57.8 ± 12.5 Range 27-78 Gender (N, %) Men 33 (68.8) Women 15 (31.2) Etiology Traumatic 4 Diabetic (ischemic) 8 Congenital 3 Vitamin B deficiency 1 Unknown 32 TABLE 2. Mean ocular cyclotorsion in all the combinations and the agreement between techniques A-C and B-D Technique Mean ocular Cyclotorsion ± SD (degrees) Mean Difference ± SD SEM 95% CI of the Difference ICC Correlation Between Lower Upper Techniques* CV (%) A 6.67 ± 2.8 0.10 ± 1.34 0.19 20.284 0.488 0.93 0.88 14.5 C 6.57 ± 2.6 B 6.82 ± 2.6 0.69 ± 1.46 0.21 0.274 1.113 0.92 0.85 16.1 D 6.12 ± 2.8 *Paired samples statistics. CV, coefficient of variation; ICC, intraclass correlation; SD, standard deviation; SEM, standard error of the mean. Almog et al: J Neuro-Ophthalmol 2014; 34: 362-365 363 Original Contribution Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. DISCUSSION Our study demonstrated that an SMR can provide results equivalent to that of the DMR test. There was no clinical difference in the angles of cyclotorsion recorded when measured with 1 or 2 Maddox rods. It is likely that a number of existing horizontal lines can be found at each ophthalmologist's office to serve as a horizontal reference line and replace one of the Maddox rods. The DMR test has some limitations and disadvantages. It results in a subjective measurement of the perceived tilt of an image (1, 3). In addition, it provides the algebraic sum of the cyclotorsion of the 2 eyes and does not quantify the amount of cyclotorsion of each eye. Furthermore, it is not practical to use this test in eccentric gazes. The method of using an SMR has the same limitations and is not an alter-native to other cyclotorsion tests (4-8) but rather is a simpler technique. Patients with congenital superior oblique palsy may have anomalous torsional retinal correspondence and lack subjective cyclotorsion. As we included only those pa-tients with cyclotorsion 3° or more, we believe that such subjects were excluded. We chose consecutive cases of supe-rior oblique palsy. It is unclear why the etiology was unknown in the majority of patients. We measured the patients with the Maddox rod in front of the affected eye (combinations A and C) and in front of the sound eye (combinations B and D). Kushner and Hariharan (1) reported that there is no immediate motor correction for torsional misalignment when fixation shifts from the normal eye to the contralateral torqued eye; con-sequently, objective torsion can be studied with either eye fixating. There was a significant difference between meas-urements B and D. However, the differences in the angles were very small and without clinical significance. Simons et al (9) found that most patients with superior oblique palsy used visual environmental cues seen through the semitransparent white Maddox as a horizontal refer-ence and directed all the cyclodeviation to the side of the red rod. They suggested to darken the room and to use a red Maddox rod on each side to achieve a higher sensi-tivity in the patient's localized excylotorsion of the paretic eye. In our SMR test, we intentionally allowed the patients to use the horizontal visual cues in the room; conse-quently, all the cyclotorsion was detected in the eye with the red rod. Thus, the SMR test measures the relative cyclotorsion between the 2 eyes and does quantitate the cyclotorsion of only the paretic eye. FIG. 1. Bland-Altman plots show the mean plotted against the differences between 2 methods. The solid line indicates the mean difference. The upper and lower lines represent 95% limit of agreement (LOA). Most measurements are within limits of agreement. The comparison between A to C shows 8.2% outliers (4/48), LOA range of 22.5 to 2.7 and between B to D 6.25% outliers (3/48), LOA range of 22.2 to 3.6. FIG 2. Left, comparison of cyclodeviation measurements with the Maddox rod in front of the affected eye (techniques A and C). Right, comparison with Maddox rod: in front of the sound eye (techniques B and D). 364 Almog et al: J Neuro-Ophthalmol 2014; 34: 362-365 Original Contribution Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Limitations of our study include lack of an etiology of superior oblique palsy in the majority of patients, the potential role of anomalous torsional retinal correspondence, and the fact that the examiner was not masked to the testing technique. REFERENCES 1. Kushner BJ, Hariharan L. Observations about objective and subjective ocular torsion. Ophthalmology. 2009;116:2001- 2010. 2. Wright KW, Spiegel PH, Thompson LS, Peter H, Thompson LS, eds. Handbook of Pediatric Strabismus and Amblyopia. New York, NY: Springer, 2006:165. 3. Guyton DL. Ocular torsion reveals the mechanisms of cyclovertical strabismus: the Weisenfeld lecture. Invest Ophthalmol Vis Sci. 2008;49:847-857. 4. Spierer A. Measurement of cyclotorsion. Am J Ophthalmol. 1996;122:911-912. 5. Ruttum M, von Noorden GK. The Bagolini striated lens test for cyclotropia. Doc Ophthalmol. 1984;58:131-139. 6. Freedman SF, Gearinger MD, Enyedi LB, Holgado S, Toth CA. Measurement of ocular torsion after macular translocation: disc fovea angle and Maddox rod. J AAPOS. 2003;7:103-107. 7. Madigan WP Jr, Katz NN. Ocular torsion-direct measurement with indirect ophthalmoscope and protractor. J Pediatr Ophthalmol Strabismus. 1992;29:171-174. 8. Fujikado T, Ohji M, Kusaka S, et al. Visual function after foveal translocation with 360-degree retinotomy and simultaneous torsional muscle surgery in patients with myopic neovascular maculopathy. Am J Ophthalmol. 2001;131:101-110. 9. Simons K, Arnoldi K, Brown MH. Color dissociation artifacts in double Maddox rod cyclodeviation testing. Ophthalmology. 1994;101:1897-1901. Almog et al: J Neuro-Ophthalmol 2014; 34: 362-365 365 Original Contribution Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited.