Morphometry of the Trochlear Nerve and Superior Oblique Muscle Volume in Congenital Superior Oblique Palsy

Abstract

Purpose.: To infer the pathogenic mechanism of congenital superior oblique palsy (SOP) by evaluating trochlear nerve diameter (CN4D) and superior oblique muscle (SO) volume in patients with congenital SOP.

Methods.: The medical records of 125 patients diagnosed with unilateral congenital SOP and 34 age-matched healthy controls were reviewed retrospectively. Using thin-section high-resolution magnetic resonance imaging, we evaluated the presence of the trochlear nerve, CN4D, SO volume, and their relationship.

Results.: Of the 125 patients with congenital SOP, 87 showed absence of the trochlear nerve (n = 87, 70%, absent group) and 38 showed bilateral presence of the trochlear nerve (n = 38, 30%, present group). The nonparetic side CN4D was smaller than controls in the absent group (P = 0.001), and larger than controls in the present group (P = 0.001). Trochlear nerve diameter positively correlated with SO volume in controls (P = 0.014, R² = 0.174) and in the nonparetic sides of congenital SOP (present group P = 0.023, R² = 0.135; absent group, P = 0.008, R² = 0.079). The paretic side SO volume did not show a linear correlation between CN4D and SO volume in the present group (P = 0.243).

Conclusions.: In congenital SOP patients without a trochlear nerve, the nonparetic side CN4D was also reduced in contrast to those with a trochlear nerve. The relatively weaker relationship of CN4D and nonparetic side SO volume in the absent group than that of the present group suggests different pathogenic mechanisms underlying these two entities of congenital SOP.

Introduction

Superior oblique palsy (SOP), recognized as one of the most common causes of vertical diplopia,¹ is characterized by hypertropia of the paretic eye and binocular vertical diplopia worsening in lateral gaze away from the affected eye and during ipsilateral head tilt.^2,3 With congenital origin, associated findings are longstanding head tilt in old photographs, facial asymmetry, intermittent exotropia, amblyopia in the affected eye and atypical very-late onset diplopia.^4,5 Morphometric differences between the paretic and nonparetic superior oblique muscles (SO) in congenital SOP have been investigated using magnetic resonance imaging (MRI).^6–11 Recently, we were able to dichotomize SOP according to the presence or absence of the trochlear nerve and found clinical differences between these groups.^12,13 However, the underlying pathogenic mechanism of these two groups of congenital SOP is not clear. These studies raised questions about the heterogeneity of SOP. There has been no report describing the relationship between the trochlear nerve diameter (CN4D) and SO volume in congenital SOP. The current study attempts to infer the pathogenic mechanism of congenital SOP by evaluating CN4D and SO volume in comparison with control subjects. 

Methods

Participants

The medical records were reviewed retrospectively for 125 consecutive patients diagnosed with unilateral congenital SOP (97 patients were involved in our previous study¹²) and 34 age-matched controls at Seoul National University Bundang Hospital (Seongnam, Korea) between August 2009 and June 2012. These study subjects were included from our previous study of superior oblique muscle volume analysis and we additionally measured the trochlear nerve diameters in those subjects.¹⁴ We collected age, sex, and laterality of the paretic side. Diagnosis of SOP was established by clinical examination including: positive Bielschowsky head tilt test, underdepression and overelevation in adduction on the affected side, large fusional amplitudes of vertical deviation, a history or an evidence of long standing strabismus or anomalous head position from infancy. We excluded patients who had primary overaction of the inferior oblique muscle on the affected side, no evidence of anomalous head posture or facial asymmetry, and any evidence of acquired disease including a history of head or ocular trauma, skew deviation, ocular tilt reaction, and plagiocephaly. Orthotropic subjects and patients with esotropia and exotropia without oblique overaction or underaction were enrolled as the control group. Oblique overaction was evaluated by the presence of overelevation or overdepression in adduction, and oblique underaction was evaluated by the presence of underelevation or underdepresion in adduction compared with the fellow eye. All aspects of the research protocol were in compliance with the Declarations of Helsinki and were approved by the institutional review board of Seoul National University Bundang Hospital. 

MR Imaging

Magnetic resonance images were obtained from a 3T system (Intera Achieva; Philips Healthcare, Best, The Netherlands) with an 8-channel head coil, initially with T2 weighted imaging of the entire brain and the orbit and subsequently with high-resolution cranial nerve imaging of the brainstem. Then, we performed the initial orbital imaging to evaluate the extraocular muscles, with fixation of head position and closed eyes, without the use of visual targets for fixation, with the turbo spin echo technique. Additional high-resolution cranial nerve imaging to visualize the cisternal segment of the trochlear nerve was performed with a 3-dimensional balanced turbo field echo sequence at the pontomesencephalic junction, including the inferior margin of the inferior colliculus, which is known to be the level of the root exit of the trochlear nerve. The scanning plane was set to an oblique axial direction perpendicular to the long axis of the aqueduct, which was approximately parallel to the course of the trochlear nerve. The sequence parameters were as follows: repetition time/echo time, 9.9/5.0 ms; flip angle, 60°; field of view, 150 × 150 mm; matrix, 500 × 500; section thickness, 0.25 mm; number of slices, 60; sensitivity encoding (SENSE) factor, 2; number of signal averaging, 2; and acquisition time, 7 minutes and 14 seconds. The voxel size was 0.3 × 0.3 × 0.25 mm.^12–14 We obtained the area of the SO manually using DTU-710 (WACOM, Saitama, Japan) and Picture Archiving and Communication System (PACS) software (INFINITT, Seoul, Korea), which provides automatic measurement for area and length. The optic nerve-globe junction was defined as the standard point (“point 0”), and SO areas were measured in the coronal sections of five points, including the standard point and points that were 2- and 4-mm anterior or posterior to the standard point (Fig. 1).¹⁴ The volume of the SO was defined as the sum of SO areas at the five points, as above, multiplied by 2 (mm), which was the interval between points. The CN4D was defined as the average value of the diameters measured at three points along the cisternal segment of the nerve. Three individual examiners (DSL, HKY, and JWL) measured the data blindly and measurements were repeated three times for each examiner. The average of the nine values was used for the main analysis. The measurements of CN4D, ratio of trochlear nerve diameter (left to right for controls, paretic to nonparetic for SOP) and SO volume were investigated. In the absent group, CN4D was measured only for the nonparetic side. 

Figure 1

 

Magnetic resonance imaging in a patient with right superior oblique palsy and an ipsilateral absent trochlear nerve. (A) T2 weighted coronal image of the superior oblique muscle (SO). The right SO is hypoplastic compared with the left. The area surrounded by the curvilinear line was measured in five different points. (B) High-resolution imaging showing only the left trochlear nerve and absence of the right trochlear nerve. The trochlear nerve diameter was obtained at three points (double dots) and averaged.

Statistical Analyses

Statistical analyses were performed using SPSS software for Windows version 17.0 (SPSS, Inc., Chicago, IL, USA). We compared groups using the ANOVA with Bonferroni post hoc test and Pearson's χ² test. Linear regression was performed to elucidate the association between parameters. 

The interobserver intraclass correlation coefficients (ICC) and their 95% confidence intervals were calculated among the measurements. 

Results

Patients' Characteristics

The participants had a mean age of 14.5 ± 17.51 years, and 92 of subjects were men and 67 were women. Of the 125 patients with unilateral congenital SOP, 87 patients (70%) were found with an ipsilateral absence of the trochlear nerve and variable degree of SO hypoplasia (absent group).¹⁴ The other 38 patients (30%) had normal anatomy of the SO and trochlear nerve on both sides (present group).¹⁴ Control subjects without SOP had normal anatomical features of the SO and trochlear nerves (control group).¹⁴ The mean age and sexual distribution were not significantly different among the groups (P = 0.896, 0.283, respectively).¹⁴ There was no significant difference in the laterality of SOP between the absent and present group (P = 0.085) (Table).¹⁴ 

Table.

 

Trochlear Nerve Diameter and Superior Oblique Volume Measured in Congenital Superior Oblique Palsy Patients Without (Absent Group) and With (Present Group) an Ipsilateral Trochlear Nerve and in Controls (Mean ± SD)

	Control, n= 34	Absent, n= 87	Present, n= 38	PValue	Post Hoc Test
Age14 (range)	16.1 ± 20.7 (1–69)	15.2 ± 17.9 (1–64)	11.5 ± 13.0 (1–58)	0.896*
Sex,14 M:F	16:18	55:32	21:17	0.283†
Paretic side,14 L:R		39:48	21:17	0.085†
Paretic CN4D, mm	0.39 ± 0.049 (left)		0.40 ± 0.050	0.461‡
Nonparetic CN4D, mm	0.39 ± 0.045 (right)	0.37 ± 0.055	0.41 ± 0.050	0.001*	B < A, C
CN4D ratio, L/R§, p/np‖	0.99 ± 0.09		0.97 ± 0.12	0.518‡
Paretic SO volume,14 mm3	41.6 ± 8.15	21.5 ± 8.69	47.0 ± 9.50	< 0.001*	B < A < C
Nonparetic SO volume,14 mm3	40.9 ± 9.87	49.3 ± 11.15	46.8 ± 9.60	0.001*	A < B
SO volume ratio,14 L/R§, p/np‖	1.03 ± 0.148	0.44 ± 0.146	1.03 ± 0.099	< 0.001*	B < A, C

 p, paretic side; np, nonparetic side; L, left; R, right.

*  P value by ANOVA.

†  P value by Pearson's χ².

‡  P value by independent t-test.

§  Left to right ratio for control group.

‖  Paretic side to nonparetic side ratio for absent and present group.

Trochlear Nerve Diameter

Trochlear nerve diameter was not significantly different between both eyes in the present group and controls (P = 0.398, P = 0.122 by paired t-test, respectively). Trochlear nerve diameter of the nonparetic side was significantly different among the three groups (P = 0.001 by ANOVA): Nonparetic side CN4D was smaller than controls in the absent group and larger than controls in the present group. Paretic side CN4D of the present group was not significantly different from controls (P = 0.461 by independent t-test) (Table; Fig. 2). The paretic to nonparetic side CN4D ratio of the present group showed no significant difference compared with the left to right ratio of the control group (P = 0.518). The calculated interobserver ICCs were 0.667 (0.590–0.731, P < 0.001) in CN4D. 

Figure 2

 

Trochlear nerve diameter (CN4D) of the paretic side and nonparetic side of congenital SOP patients and controls (mean ± SE). (A) Trochlear nerve diameter of the nonparetic side was significantly different among the three groups. Trochlear nerve diameter was thinner than controls in the absent group and thicker than controls in the present group (P = 0.001 by ANOVA). (B) There was no significant difference of CN4D between controls and the paretic side of the present group (P = 0.461 by independent t-test).

Superior Oblique Volume

The SO volume of the paretic side was smallest in the absent group, larger in the control group, and largest in the present group (P < 0.001).¹⁴ The SO volume of the nonparetic side was smallest in the control group, larger in the present group, and largest in the absent group (P = 0.001).¹⁴ The paretic to nonparetic side SO volume ratio was significantly lower in the absent group compared with other groups (P < 0.001) (Table).¹⁴ The calculated interobserver ICC was 0.770 (0.747–0.790, P < 0.001) of the measured SO area.¹⁴ 

Quantitative Relationships of SO Volume and Trochlear Nerve Diameter

The quantitative relationships of SO volume and CN4D are displayed in Figure 3. There was a positive linear relationship between SO volume and CN4D in the control group (P = 0.014, R² = 0.174). In the nonparetic eye of congenital SOP patients, the relationship between SO volume and CN4D showed a positive linear correlation in the absent group (P = 0.008, R² = 0.079) and the present group (P = 0.023, R² = 0.135, respectively). However, in the paretic eye of the present group, SO volume and CN4D showed no significant relationship (P = 0.243). 

Figure 3

 

The relationship between CN4D and SO volume was positively correlated in (A) controls, (B) the nonparetic side of the absent group, and (C) the nonparetic side of the present group in congenital SOP. (D) The paretic side of the present group showed no significant relationship between CN4D and SO volume.

Discussion

The small diameter of the trochlear nerve has been the biggest barrier against its in vivo evaluation with MRI techniques. The development of a high-resolution and thin-section MRI sequence raised the detection rate of the trochlear nerve to 100% in healthy controls,¹⁵ which was the basis of the next step to analyze the trochlear nerve morphometrically. To our knowledge, this report is the first description of the relationship between CN4D and SO volume in congenital SOP. 

The followings are the new findings of our study. First, CN4D of the nonparetic side was smaller than that of controls in the absent group, but not in the present group. Second, CN4D was positively correlated with SO volume in the nonparetic side of congenital SOP patients and controls. Third, there was no significant correlation between CN4D and SO volume in the paretic side of the present group. These results can be partial evidence of the normal relation between CN4D and SO volume in controls, lack of normal relationships on the paretic side of congenital SOP, and the discrepancy between present and absent group to infer different pathogenic mechanisms. 

The mean CN4D of the control group measured in our study was similar to the values reported in previous studies: 0.54 mm (0.35–0.96 mm)¹⁵ and 0.4 mm (0.2–0.6 mm).¹⁶ In the absent group, the diameter of the trochlear nerve was abnormal not only in the paretic side, but also in the nonparetic side. In contrast, the CN4D in the present group showed no significant difference with those in the controls. This discrepancy suggests that the mechanism of SO palsy might be different in the absent and present groups. In the absent group, a deficit during embryonic development may be the underlying mechanism of SOP. The SO volume of the paretic side in the absent group was abnormally small. We assumed that the SO of the paretic side could not receive proper neurotropic stimulation during development because of the absence of the trochlear nerve. Even in the nonparetic side, the size of the trochlear nerve was smaller and the linear correlation was much weaker than that of controls. These findings may suggest the possibility of a developmental abnormality in both sides of the trochlear nerve. Conversely, in the present group, the pathogenic mechanism of SOP may be explained by heterogeneous anatomical problems rather than a developmental deficit. The diameters of both trochlear nerves appeared normal, and there was no significant atrophy of the SO in the paretic side. These results imply that neurotropic stimulation for muscle growth is sufficient in the present group and other anatomical causes may be responsible for SO dysfunction. The lack of a linear relationship between the CN4D and SO volume on the paretic side in the present group can be explained by heterogenous anatomical problems such as heterotropic muscle pulleys,¹⁷ tendon anomaly,⁶ or successful morphologic catch-up growth after hidden birth injury. However, the R² values and correlations between the CN4D and SO volume in controls and both groups of SOP patients were weak, which limits the interpretation of our results. The variable state of contraction of the SO due to different gaze positions may have yielded variable estimates of the SO volume, leading to lower correlations with the CN4D data. A prospective study with controlled eye positions may lead to improved correlations. 

Some limitations need to be considered in this analysis. First, this is a retrospective study, which cannot elucidate the cause and effect relationship and errors resulting from unintended selection bias. Second, we obtained MRI images without using visual targets for fixation. In our study, some participants who were too young to remain stable during MRI were sedated. Although closed eyes tend to rest in an upgaze position because of Bell's phenomenon, variable positions of the eyeball could have affected the muscle size. If the eyes were in an upgaze position, expected in most patients, the cross-sectional area of the SO would be smaller than that in the primary position, which could underestimate the SO volume.¹⁰ The state of contraction of the SO would have been quite variable among the subjects due to different gaze positions, yielding to quite variable estimates of the SO volume. In turn, this could lead to lower correlations with the CN4D data. However, Demer and Miller¹⁰ reported that contracture of a normal SO changes its volume significantly but the paretic SO does not. Third, high-resolution cranial nerve imaging of the brain stem with 0.25-mm thickness requires a long scanning time, which can induce motion artifacts interfering with the identification of the trochlear nerve. Furthermore, subjects with a hypoplastic trochlear nerve, with a diameter less than 0.125 mm, maybe classified as the absent group, which could confound the data. However, such misclassification might be rare, because the trochlear nerve runs obliquely in the cisternal area and any of the portion of the nerve might be detected by MRI. Magnetic resonance imagining sequence with thinner sections and a faster scanning time could overcome these problems, and prospective studies are needed to reveal the more accurate cause and effect relationship. Finally, another weak point of our study is that most of the control group had horizontal strabismus and pure orthotropic controls were only five patients. However, if we divided the control group into orthotropia and strabismus without oblique involvement, we could not achieve sufficient statistical power for age-matched analysis. 

In conclusion, the CN4D of the nonparetic side in the absent group was smaller compared with controls which may be partial evidence of an underlying developmental deficit. The weak correlation of the paretic side CN4D and SO volume in the present group may suggest heterogeneous anatomical deficiencies as the underlying mechanism rather than a developmental problem in these patients. 

Acknowledgments

Supported by grants from the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT, and Future Planning (2013R1A1A2010606; Seoul, South Korea). 

Disclosure: D.S. Lee, None; H.K. Yang, None; J.H. Kim, None; J.-M. Hwang, None 

Footnotes

 DSL and HKY contributed equally to the work presented here and should therefore be regarded as equivalent authors.