Comparing Hypertropia in Upgaze and Downgaze Distinguishes Congenital From Acquired Fourth Nerve Palsies

Isolated fourth nerve palsies are commonly caused by decompensation of a congenitally dysfunctional superior oblique muscle ('decompensated congenital palsies'). Distinguishing such palsies at initial presentation from palsies caused by presumed microvascular ischemia ('ischemic palsies') has value for patient reassurance and in forestalling ancillary testing. Abnormally large vertical fusional amplitudes traditionally have been used to identify decompensated congenital palsies, but that may not be a reliable distinguishing feature. This study was undertaken to determine if the amount of hypertropia in upgaze and downgaze might be a more efficient separator. We also studied traumatic and tumorous fourth nerve palsies to see if they could be distinguished from decompensated congenital palsies by using this hypertropia comparison. Retrospective review of case records of patients diagnosed with isolated fourth nerve palsies at the University of Michigan Neuro-Ophthalmology Clinics over the past 15 years. We recorded the age, gender, vascular risk factors, duration of follow-up, cause, side of palsy, and alignment measurements in all patients. Inclusion criteria were met by 118 patients. Hypertropia was equal or greater in upgaze than downgaze in 50 of the 58 decompensated congenital palsies (86%) in whom those data were recorded. Hypertropia was never greatest in upgaze in the 15 patients with traumatic palsies. Vertical fusional amplitudes were increased in only 15 of 27 patients (56%) with decompensated palsies in whom those data were recorded. Torsional misalignment on double Maddox rod testing was present in 16 (94%), 13 (87%), and 3 (100%) patients with ischemic, traumatic, and tumorous palsies, but also in 19 patients (54%) with decompensated congenital palsies in whom those data were recorded. Hypertropia greater in upgaze than downgaze or equal in upgaze and downgaze was an efficient separator of congenital from ischemic and tumorous fourth nerve palsies, being characteristic of patients with decompensated congenital palsies and never present in patients with ischemic, traumatic, or tumorous palsies. Vertical fusional amplitudes and torsional misalignment did not effectively differentiate between the patient groups. Comparing the hypertropia in upgaze and downgaze improved differential diagnosis and reduces the potential for unnecessary ancillary tests.

Original Contribution Comparing Hypertropia in Upgaze and Downgaze Distinguishes Congenital From Acquired Fourth Nerve Palsies Yair Ivanir, MD, Jonathan D. Trobe, MD Background: Isolated fourth nerve palsies are commonly caused by decompensation of a congenitally dysfunctional superior oblique muscle ("decompensated congenital palsies"). Distinguishing such palsies at initial presentation from palsies caused by presumed microvascular ischemia ("ischemic palsies") has value for patient reassurance and in forestalling ancillary testing. Abnormally large vertical fusional amplitudes traditionally have been used to identify decompensated congenital palsies, but that may not be a reliable distinguishing feature. This study was undertaken to determine if the amount of hypertropia in upgaze and downgaze might be a more efficient separator. We also studied traumatic and tumorous fourth nerve palsies to see if they could be distinguished from decompensated congenital palsies by using this hypertropia comparison. Methods: Retrospective review of case records of patients diagnosed with isolated fourth nerve palsies at the University of Michigan Neuro-Ophthalmology Clinics over the past 15 years. We recorded the age, gender, vascular risk factors, duration of follow-up, cause, side of palsy, and alignment measurements in all patients. Results: Inclusion criteria were met by 118 patients. Hypertropia was equal or greater in upgaze than downgaze in 50 of the 58 decompensated congenital palsies (86%) in whom those data were recorded. Hypertropia was never greatest in upgaze in the 15 patients with traumatic palsies. Vertical fusional amplitudes were increased in only 15 of 27 patients (56%) with decompensated palsies in whom those data were recorded. Torsional misalignment on double Maddox rod testing was present in 16 (94%), 13 (87%), and 3 (100%) patients with ischemic, traumatic, and tumorous palsies, but also in 19 patients (54%) with decompensated congenital palsies in whom those data were recorded. Conclusions: Hypertropia greater in upgaze than downgaze or equal in upgaze and downgaze was an efficient separator of congenital from ischemic and tumorous fourth nerve Departments of Ophthalmology and Visual Sciences (YI), Kellogg Eye Center, University of Michigan; and Department of Neurology (YI, JDT), University of Michigan, Ann Arbor, Michigan. The authors report no conflicts of interest. Address correspondence to Jonathan D. Trobe, MD, Kellogg Eye Center, 1000 Wall Street, Ann Arbor, MI 48105; E-mail: [e-mail] Ivanir and Trobe: J Neuro-Ophthalmol 2017; 37: 365-368 palsies, being characteristic of patients with decompensated congenital palsies and never present in patients with ischemic, traumatic, or tumorous palsies. Vertical fusional amplitudes and torsional misalignment did not effectively differentiate between the patient groups. Comparing the hypertropia in upgaze and downgaze improved differential diagnosis and reduces the potential for unnecessary ancillary tests. Journal of Neuro-Ophthalmology 2017;37:365-368 doi: 10.1097/WNO.0000000000000460 © 2017 by North American Neuro-Ophthalmology Society F ourth nerve palsy is attributed to decompensation of a congenital weakness, extraaxial ischemia, head or neurosurgical trauma, intracranial tumor, or meningeal inflammation (1-5). Diagnosis of the palsy is based on finding hypertropia that is greater in contralateral than ipsilateral gaze and greater in ipsilateral than contralateral head tilt (Parks-Bielschowsky 3-step test) (1,6). However, this test does not distinguish among causes of fourth nerve palsy. In particular, it does not separate a decompensated congenital palsy from other causes (6). That distinction is traditionally based on the presence of abnormally large vertical fusional amplitudes, ipsilateral facial atrophy, lack of subjective excyclotorsion, and a longstanding head tilt (7). However, these features may not always be reliable in separating decompensated congenital weakness from other causes of fourth nerve palsy (8-10). A more robust differentiating sign would be useful. Based on long-term observation, we hypothesized that a decompensated congenital palsy might be unique in having a hypertropia greater in upgaze than in downgaze. Accordingly, we reviewed the records of patients with a diagnosis of fourth nerve palsy who were examined at a single tertiary referral center with attention to ocular alignment measurements, as well as the presence of other factors traditionally used to separate decompensated congenital palsies from other causes. 365 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution METHODS We obtained Institutional Review Board approval for a retrospective evaluation of the neuro-ophthalmologic electronic medical records at the University of Michigan from 2000 to 2015 for isolated fourth nerve palsy (additional search terms: "trochlear" or "superior oblique" palsy). Adult (age .18 years) patients were included only if no other nerve palsy was diagnosed and if no other neurological findings were present. Patients were excluded if: 1) they had findings typical of bilateral superior oblique palsies except for 3 patients with traumatic bilateral superior oblique palsies, who were included; or 2) a finding or diagnosis of any confounding factor such as multiple sclerosis, myasthenia gravis, or thyroid eye disease. All patients had a positive 3-step test. A decompensated congenital palsy was diagnosed if the patient had a history of intermittent vertical double vision in the past and a persistent palsy for at least 6 months. An ischemic palsy was diagnosed if the palsy had resolved within 6 months, and if the patient had vascular risk factors. A traumatic palsy was diagnosed if the patient had sustained sufficient head trauma or had developed the palsy immediately after a neurosurgical procedure in the region of the fourth nerve. A tumorous palsy was diagnosed based on imaging evidence of an intracranial mass compressing the fourth nerve. We recorded the age, gender, vascular risk factors, duration of follow-up, cause, side of palsy, and alignment measurements in all patients. Ocular misalignment was measured at a testing distance of 20 feet (6 m) using the prism-and-cover test or the Maddox rod with prism bar. Patients were included only if they had alignment measurements performed in at least primary, left, and right gazes. Upgaze and downgaze alignment measurements were taken from the midline or from left gaze in the case of a right hypertropia and from right gaze in the case of a left hypertropia. Torsional alignment was measured with the double Maddox rod test and recorded in degrees. Excyclodeviation was considered increased if it measured $2°. Vertical fusional amplitudes were measured either by break in fusion with increasing vertical prism strength or by calculating the difference between the phoric and tropic measurements in primary gaze position. Vertical fusional amplitudes were considered increased if they were $6 prism diopters (PD) and above. RESULTS Inclusion criteria were met by 118 patients, of whom 65 were men and 53 were women, with a mean age of 57 years (range, 13-87 years) (Table 1). There were 62 patients (53%) with a decompensated congenital palsy. Their mean age was 55.7 years. Patients had undergone the following negative work up: 23 MRIs, 9 computed tomographies (CTs), and 16 serum acetylcholine receptor antibody titers (AchR-Ab). There were 31 patients (26%) with an ischemic palsy. Their mean age was 64 years. All patients achieved complete resolution of the palsy within a median of 3 months. Patients had undergone the following negative work up: 14 MRIs, 3 CTs, and 4 AchR-Ab titers. There were 20 patients (17%) with a traumatic palsy (17 following head trauma and 3 following neurosurgical procedures). Their mean age was 50 years. All patients with traumatic palsies had undergone brain imaging, usually a CT in the emergency department. Over a minimum follow-up period of 6 months, 3 achieved complete resolution, 3 partial resolution, and 14 no resolution. There were 5 patients (4%) with a tumorous palsy, including 3 tentorial meningiomas, 1 pineal tumor, and 1 temporal lobe oligodendroglioma. Their average age was TABLE 1. Demographic features of patients with fourth nerve palsies Type of Palsy Decompensated congenital Side of Palsy Right/Left/ Bilateral Brain Imaging CT/MRI No. Patients Age, yr Male/ Female 62 Average = 55.7 34/28 32/30/0 9/23 19/12 19/12/0 3/14 9/11 7/10/3 15/5 3/2 4/1/0 2/5 Ischemic 31 Traumatic 20 Tumorous 5 SD = 16.27 Range = 13-83 Average = 64.2 SD = 12.9 Range = 24-84 Average = 50.47 SD = 19.38 Range = 18-78 Average = 57 SD = 23.76 Range = 28-87 CT, computed tomography; SD, standard deviation. 366 Ivanir and Trobe: J Neuro-Ophthalmol 2017; 37: 365-368 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution TABLE 2. Alignment and vertical fusional amplitude measurements in patients with fourth nerve palsies Type of Palsy Hypertropia in PD in Primary Gaze, Mean ± SD (Range) Hypertropia in PD in Downgaze, Mean ± SD (Range) Hypertropia in PD in Upgaze, Mean ± SD (Range) Decompensated 7.08 ± 6.55 (0-25) 6.63 ± 6.35 (0-30) 6.13 ± 7.31 (0-30) congenital Ischemic 3.3 ± 1.93 (0-8) 1.29 ± 1.91 (0-7) 6.5 ± 4.58 (2-20) Traumatic 5.2 ± 6.8 (1-25) 2.83 ± 4.39 (0-16) 7.5 ± 4.54 (3-16) Tumorous 5.0 ± 4.9 (0-12) 1.75 ± 1.7 (0-4) 9.25 ± 6.7 (3-16) $2° Torsional Misalignment by Double Maddox Rod (Yes/No/Not Recorded) $6PD Vertical Fusional Amplitude (Yes/No/Not Recorded) 19/16/27 15/12/35 16/1/14 13/2/5 3/0/2 0/13/18 NA NA NA, not available; PD, prism diopters; SD, standard deviation. 57 years. Over a minimum follow-up period of 6 months (mean, 24 months), none of their palsies resolved. The alignment measurements in primary position, upgaze and downgaze are listed in Table 2. Hypertropia in primary gaze position was generally greater in patients with decompensated congenital palsies. Data comparing the amount of hypertropia (in PD) in upgaze and downgaze were available in 58 of 62 patients (94%) in the decompensated group. Hypertropia was greatest in upgaze in 34 patients (59%) with decompensated congenital palsies (Table 3). Hypertropia was equal in upgaze and downgaze in 8 (14%). Hypertropia was greatest in downgaze in 16 patients (27%). Data comparing hypertropia in upgaze and downgaze were available in all 32 patients in the ischemic group and in all 5 patients in the tumorous group. Hypertropia was greatest in downgaze in all patients in both groups. Data comparing hypertropia in upgaze and downgaze were available in 15 of 20 patients (75%) in the traumatic group. Hypertropia was never greatest in upgaze in traumatic palsies (Table 3). Hypertropia was greatest in downgaze in 11 patients (73%). Hypertropia was equal in upgaze and downgaze in 4 (27%). The hypertropia difference between upgaze and downgaze was statistically significant when comparing the decompensated palsies to the ischemic, traumatic, and tumorous palsies (Table 4). Torsional misalignment was assessed in 35 (56%), 17 (55%), 15 (75%), and 3 (60%) of patients in the decompensated, ischemic, traumatic, and tumorous groups, respectively (Table 2). Torsional misalignment of at least 2° was present in 94, 87, and 100% of the ischemic, traumatic, and tumorous palsies, respectively, in whom it was assessed. However, it was also present in 54% of the patients with a decompensated congenital palsy in whom it was assessed. Thus, although the frequency of torsional misalignment was less in the congenital than in the acquired palsies, the difference was not significant. Moreover, the amount of torsional misalignment did not differ between the congenital and acquired palsies or among the acquired palsies. Vertical fusional amplitude was assessed in only 27 (44%) and 13 (42%) patients in the decompensated and ischemic palsies, respectively (Table 2), and not assessed in any patients with traumatic or tumorous palsies. It was abnormally large in 56% of the patients with decompensated palsies and never in patients with ischemic palsies. Thus, although this feature occurred in more than half of the decompensated palsies and in none of the acquired palsies, it occurred too infrequently in the congenital palsies to be clinically useful in separating out this entity. DISCUSSION In our retrospective cohort, hypertropia was either greater in upgaze than downgaze or equal in upgaze and downgaze in 73% of patients with decompensated palsies. By contrast, hypertropia was always greater in downgaze than in upgaze in the ischemic and tumorous palsies, even after a mean follow-up period of 24 months for the tumorous palsies. Thus, this measurement proved to be an efficient separator TABLE 3. Comparison of upgaze and downgaze hypertropia among patients with fourth nerve palsies in whom alignments in these gaze positions were recorded Hypertropia Upgaze . downgaze Downgaze . upgaze Downgaze = upgaze Decompensated Congenital Palsy Ischemic Palsy Traumatic Palsy Tumorous Palsy 34 16 8 0 31 0 0 11 4 0 5 0 Ivanir and Trobe: J Neuro-Ophthalmol 2017; 37: 365-368 367 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution TABLE 4. Difference in hypertropia in upgaze and downgaze among patients with fourth nerve palsies in whom measurements in these gaze positions were recorded Hypertropia Difference (in PD) Up vs Downgaze Decompensated Congenital Palsy Ischemic Palsy Traumatic Palsy Tumorous Palsy 0.45 6.76 - 24.91 4.07 0.0002 24.67 4.54 0.014 27.5 5.51 0.0246 Mean SD P value vs decompensated congenital palsy PD, prism diopters; SD, standard deviation. of congenital palsies from ischemic and tumorous palsies. Hypertropia was equal in upgaze and downgaze in a minority of patients with traumatic palsies. This feature proved less helpful in separating congenital from traumatic palsies, but the presence of recent severe head trauma would have allowed that separation. Patients who had hypertropia greatest in upgaze always had a congenital palsy. We did find that 28% of patients with congenital palsies had hypertropia greatest in downgaze, so that a minority of patients with decompensated palsies could not be separated from the acquired palsies using this feature. Why should patients with congenital palsies mostly have a hypertropia greater in upgaze or a hypertropia at least as great in upgaze as in downgaze? The answer is not known, but 2 possibilities can be considered: 1) congenital superior oblique palsy evokes an adaptional response in the form of overaction of the inferior oblique muscle; or 2) congenital superior oblique palsy is not an innervational disorder like the acquired conditions, but rather a separate anatomic disorder that usually does not show reduced infraduction in adduction. Other clinical features traditionally used to separate congenital from acquired fourth nerve palsies were not reliable in our patients. For example, vertical fusional amplitudes were recorded in less than half of the cohort of patients with decompensated and ischemic palsies, and not recorded in any of those patients with traumatic and tumorous palsies. Moreover, fusional amplitudes were abnormally large in only about half of the patients with decompensated palsies in whom it was recorded. Nor was the presence of torsional misalignment on double Maddox rod testing a reliable differentiating measure. It was assessed in only about half of patients in the decompensated and ischemic groups. Although it was present in a majority of the acquired palsies, it was also present in more than half of the patients with decompensated congenital palsies. The fact that fusional amplitude and torsional alignment measurements were not often recorded may indicate that these tests are too subjective, time-consuming, or difficult to perform. The retrospective nature of the study poses some limitations. The measurements were gathered from several examiners whose assessment techniques were not 368 standardized, and some measurements traditionally considered valuable in differentiating congenital from acquired palsies were incompletely gathered. A prospective study addressing these limitations would be valuable in strengthening the findings of this study. Despite these limitations, our study offers credible evidence that comparing the degree of hypertropia in upgaze and downgaze is a valuable separator of congenital from acquired fourth nerve palsies. Finding that the hypertropia is greatest in upgaze allows the physician to reassure the patient of a benign condition and to avoid stressful, expensive, and time-consuming studies. STATEMENT OF AUTHORSHIP Category 1: a. Conception and design: Y. Ivanir and J. D. Trobe; b. Acquisition of data: Y. Ivanir and J. D. Trobe; c. Analysis and interpretation of data: Y. Ivanir and J. D. Trobe. Category 2: a. Drafting the manuscript: Y. Ivanir and J. D. Trobe; b. Revising it for intellectual content: Y. Ivanir and J. D. Trobe. Category 3: a. Final approval of the completed manuscript: Y. Ivanir and J. D. Trobe. REFERENCES 1. Ellis FD, Helveston EM. Superior oblique palsy: diagnosis and classification. Int Ophthalmol Clin. 1976;16:127-135. 2. Tiffin PA, MacEwen CJ, Craig EA, Clayton G. Acquired palsy of the oculomotor, trochlear and abducens nerves. Eye (Lond). 1996;10:377-384. 3. Brazis PW. Palsies of the trochlear nerve: diagnosis and location, recent concepts. Mayo Clin Proc. 1993;68:501-509. 4. Akagi T, Miyamoto K, Kashii S, Yoshimura N. Cause and prognosis of neurologically isolated third, fourth, or sixth cranial nerve dysfunction in cases of oculomotor palsy. Jpn J Ophthalmol. 2008;52:32-35. 5. Knapp P, Moore S. Diagnosis and surgical options in superior oblique surgery. Int Ophthalmol Clin. 1976;16:137-149. 6. Engel JM. Treatment and diagnosis of congenital fourth nerve palsies: an update. Curr Opin Ophthalmol. 2015;26:353-356. 7. Parks MM. Ocular Motility and Strabismus. Hagerstown, MD: Harper & Row, 1975. 8. Muthusamy B, Irsch K, Peggy Chang HY, Guyton DL. The sensitivity of the Bielschowsky head-tilt test in diagnosing acquired bilateral superior oblique paresis. Am J Ophthalmol. 2014;157:901-907. 9. Trobe JD. Cyclodeviation in acquired vertical strabismus. Arch Ophthalmol. 1984;102:717-720. 10. Von Noorden GK, Murray E, Wong SY. Superior oblique paralysis. A review of 270 cases. Arch Ophthalmol. 1986;104:1771-1776. Ivanir and Trobe: J Neuro-Ophthalmol 2017; 37: 365-368 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited.