Comparison of head tilt test between sagging eye syndrome and acquired unilateral trochlear nerve palsy

Acquired trochlear nerve palsy, which is often observed after microcirculatory disturbance or trauma, has been considered to account for most cases of vertical strabismus and cyclotorsion [1]. However, in 2013, Chaudhuri and Demer reported “sagging eye syndrome” (SES), caused by age-related changes in the orbital pulley [2], and it was pointed out that SES is often the cause of vertical strabismus and cyclotorsion that had been considered to be caused by trochlear nerve palsy [2, 3]. SES is caused by rupture of the lateral rectus (LR) – superior rectus (SR) band or bowed LR with inferior displacement of the LR due to aging and is characterized by age-related distance esotropia (ARDE) in bilateral cases and small angle cyclo-vertical strabismus (CVS) in unilateral cases. The latter presents clinical findings similar to those of acquired unilateral trochlear nerve palsy and is often difficult to distinguish between CVS-type SES (SES group) and acquired unilateral trochlear nerve palsy (Trochlear group).

Wei et al. reported the three-step test is often positive in SES clinical alignment patterns may confound SES with acquired superior oblique palsy [4].

In this study, we investigated the clinical differences between SES group and Trochlear group using a head-tilt test, in addition to evaluation of horizontal deviations, vertical deviations and cyclotorsion, to determine the relevant clinical differences.

Of the patients with vertical strabismus and/or cyclotorsion who visited our department between November 2016 and October 2022 for the elimination of diplopia, 15 patients in the SES group (mean age 74.6 ± 5.2 years, 7 males and 8 females) and 14 patients in the Trochlear group (mean age 55.2 ± 15.9 years, 8 males and 6 females) met the following diagnostic criteria.

The diagnosis of SES was made by magnetic resonance imaging (MRI) in all patients (Fig. 1a), which showed inferior displacement of the LR, rupture of the LR-SR band or bowed LR. In addition, axial MRI demonstrated that the LR was markedly elongated in SES, and SO on cross section of MRI were normal in CVS [5]. ARED is present when the inferior displacement of LR pulley was bilaterally symmetrical. The characteristic facial features of SES were defined as a superior sulcus deformity, aponeurotic blepharoptosis, blepharoptosis, and baggy lower eyelid [6].

The diagnosis of unilateral trochlear nerve palsy was made after trauma or acute onset, with subjective symptoms of vertical or cyclo-diplopia in primary position (PP), which were aggravated by downward, unilateral inferior oblique muscle overaction and superior oblique muscle under-action, and no abnormal findings in extraocular muscles on MRI (Fig. 1b). To exclude thyroid ophthalmopathy and myasthenia gravis, thyroid-related autoantibodies and anti-acetylcholine receptor antibodies were measured, and those positive for either were excluded. Patients with diurnal variation were also excluded.

The angle of deviation was measured with the alternate prism cover test (APCT) in the PP, and the data of APCT for distance was used. The data of strabismus angle were those of dominant eye fixation in the SES group and those of the unaffected eye fixation in the Trochlear group. The vertical deviations were measured by head tilt to the hypertropia side (Hyper) and to the hypotropia side (Hypo), according to Bielschowsky head tilt test (BHTT), and the differences in vertical deviation between PP and Hyper, PP and Hypo, and Hyper and Hypo were calculated and compared between the groups. Subjective cyclotorsion was measured using the cross slide of a synoptophore, and the sum of the two eyes was used for the objective cyclotorsion which was measured, one eye at a time, with the Glaucoma Module Premium Edition (GMPE) of the SPECTRALIS OCT (Heidelberg) [4]. We used the Mann–Whitney test to test between the two groups, and The Friedman test between the three groups, and p < 0.05 was considered significant. If the Friedman test showed no significant difference in the SES group, but only in the Trochlear group, the Bonferroni method was used to examine significant differences between the groups for the Trochlear group only. And receiver operating characteristic (ROC) curves were drawn, and the sensitivity and specificity of diagnostic indicators were calculated and added to the method.

In 1995, Demer et al. reported that connective tissue pulleys histologically and functionally control the movement of the rectus muscles [7, 8]. Furthermore, they reported that there is a correlation between the restriction of vertical eye movement and the sagging of the horizontal rectus muscle pulley, that the pulley affects the horizontal, vertical and cyclo-movements by all extraocular muscles, and that connective tissue degeneration occurs with aging [9‐11]. In 2013, this strabismus caused by age-related connective tissue degeneration was named “sagging eye syndrome” [2]. SES has two types (ARDE and CVS), with the identification of SES, many cases of vertical strabismus and/or cyclotorsion of unknown etiology, were found to be caused by SES [3]. On the other hand, acquired unilateral trochlear nerve palsy is often difficult to differentiate from SES when it is not caused by trauma or intracranial lesions. In acquired trochlear nerve palsy caused by microcirculatory disturbance, MRI does not show atrophy of the superior oblique muscle, but only almost age-appropriate changes. Furthermore, since it has been reported that the LR-SR band is elongation in approximately 50% of elderly patients without strabismus [2], it is possible that patients with acquired trochlear nerve palsy also show rupture or elongation of LR-SR band. Regarding differential point, it has been reported that SES is characterized by exocyclotorsion in the eye with hypotropia because the eye with greater sagging of LR becomes hypotropia when the configuration of LR-SR pulley is asymmetric between the two eyes [2, 3]. However, in this study, there was no significant difference in exocyclotorsion in either the SES or Trochlear group.

As in the previously reported [12‐14], the horizontal deviation showed a tendency to be esodeviation in SES, while the vertical deviation showed a tendency to be small in SES. In primary position, the vertical deviation in SES seems to be smaller than that in the acquired trochlear nerve palsy, but since no cutoff value is given, it is difficult to distinguish between SES from acquired trochlear nerve palsy with small vertical deviation.

For a long time, Parks' 3-step method, which includes a head-tilt test, has been used to diagnose trochlear nerve palsy (superior oblique palsy: SOP) [1, 15]. This test is known to have a low positivity rate, especially in patients with atrophy of the SO muscle, and Demer et al. reported a positivity rate of 50% [16]. The 14 cases of acquired trochlear nerve palsy in this study did not show atrophy of the SO muscle on MRI, but Uchiyama, et al. [17] and Lee, et al. [18] reported the positive rates in cases without atrophy of the SO muscle to be 75% and 78%, respectively. Furthermore, it has been reported that about half of the SES cases fulfilled Parks' 3-step method [19]. Some reports define BHTT positively as a change of 1 PD or more [18, 20], while others define BHTT positively as a difference of 5 PD or more in vertical deviation by head tilt [21]. As previously reported, 6 of 15 patients (40%) in the SES group were positive when BHTT was defined as 1 PD or more in this study. Therefore, it is difficult to differentiate SES from acquired trochlear nerve palsy by BHTT alone.

Hence, we decided to compare the differences in vertical deviation of BHTT, between primary position (PP) and hypertropia side (Hyper), between PP and hypotropia side (hypo), and between Hyper and Hypo, thinking that these differences might be useful for differentiating between SES and acquired unilateral trochlear nerve palsy. In the SES group, head tilt did not significantly change the vertical deviation, but in the Trochlear group, head tilt to the Hypo resulted in a significantly smaller vertical deviation. The deviation of PP − Hypo, PP − Hyper, and Hyper − Hypo were all significantly smaller in the SES group than in the Trochlear group, but the most obvious difference between the two groups was found in Hyper − Hypo.

In the ROC analysis logistic regression of Hyper − Hypo, which showed the largest difference between the SES and Trochlear groups in this study, the sensitivity was 100%, specificity was 100%, and AUC was 1.0 at the cutoff value of 6 PD, indicating that SES with a Hyper − Hypo value of 6 PD or more is extremely rare.

The limitation of this study was the small number of samples. The sensitivity and specificity of 100% were probably due to the small number of samples (Fig. 4).

In this study, even without MRI, the difference between Hyper and Hypo may be the most useful index for differentiating SES and acquired unilateral trochlear nerve palsy.