The Clinical and Imaging Profile of Skew Deviation: A Study of 157 Cases

Background: Skew deviation, a vertical misalignment of the eyes caused by a lesion in the vestibulo-ocular pathway, is a common manifestation of brainstem dysfunction, yet comprehensive information about its clinical profile is lacking. The aim of this study was to document presenting symptoms, causes, ocular alignment features, accompanying neurologic signs, pertinent brain imaging abnormalities, and measures used to relieve diplopia.

Original Contribution Section Editors: Clare Fraser, MD Susan Mollan, MD The Clinical and Imaging Profile of Skew Deviation: A Study of 157 Cases Eyal Walter, MD, Jonathan D. Trobe, MD Downloaded from http://journals.lww.com/jneuro-ophthalmology by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8K2+Ya6H515kE= on 05/04/2022 Background: Skew deviation, a vertical misalignment of the eyes caused by a lesion in the vestibulo-ocular pathway, is a common manifestation of brainstem dysfunction, yet comprehensive information about its clinical profile is lacking. The aim of this study was to document presenting symptoms, causes, ocular alignment features, accompanying neurologic signs, pertinent brain imaging abnormalities, and measures used to relieve diplopia. Methods: We searched the electronic medical records text from 2000 to 2018 for “skew” or “skew deviation” at a tertiary care academic center, including only patients diagnosed under supervision of faculty neuro-ophthalmologists. After excluding patients with features suggesting an imitator of skew deviation, we collected data useful in answering the selected clinical issues. Results: In a cohort of 157 patients, vertical misalignment ranged between 1 and 30 prism diopters (PD) (median 5 PD) and was comitant in 100 (64%) patients. Diplopia was reported by 87% and blurred vision by 11% of patients. Blurred vision was more common when vertical misalignment was less than 3 PD. At least one accompanying neurologic sign was present in 133 (85%) patients, most often nystagmus, followed by gaze paresis, ataxia, saccadic pursuit, and internuclear ophthalmoplegia. Stroke affecting the thalamus, brainstem, or cerebellum accounted for 82 (52%) of cases. Stroke was usually ischemic, mostly unprovoked, but also occurring after intracranial or extracranial surgical procedures. Brainstem tumor and operative injury caused most of the remaining cases. A subgroup of 17 (11%) patients had skew deviation as the only new clinical sign and had no pertinent brain imaging abnormalities. Resolution of skew deviation, documented in 58 (42%) of 137 patients who had at least one follow-up visit, usually occurred within 3 months, but sometimes not until after 12 months. Of 110 patients who still had diplopia on follow-up examinations, the diplopia was successfully relieved with prism spectacles in 68 (62%). The cause of the skew Department of Ophthalmology and Visual Sciences (EW, JDT), Kellogg Eye Center, University of Michigan, Ann Arbor, Michigan; and Departments of Neurology (JDT) and Neurosurgery (JDT), University of Michigan, Ann Arbor, Michigan. The authors report no conflicts of interest. Address correspondence to Jonathan D. Trobe, MD, University of Michigan, Kellogg Eye Center, Department of Ophthalmology and Visual Sciences, 1000 Wall Street, Ann Arbor, MI 48105; E-mail: [e-mail] Walter and Trobe: J Neuro-Ophthalmol 2021; 41: 69-76 deviation, amount of vertical misalignment, and degree of incomitance did not predict the success in relieving diplopia. Accompanying neurologic signs, including ataxia, endured in 44% of patients and were often more debilitating than the diplopia of skew deviation. Conclusions: The amplitude of misalignment in skew deviation varies widely but is generally 5 PD or less. When misalignment is 3 PD or less, patients report blurred vision rather than diplopia. Skew deviation is usually accompanied by other neurologic signs reflecting brainstem dysfunction. Yet there is a small subgroup in which vertical misalignment is an isolated sign, and there are no supporting brain imaging abnormalities. The main cause of skew deviation is ischemic stroke, which affects not only the brainstem but also the thalamus. Diplopia from skew deviation frequently persists, in which case prism spectacles may be successful in palliating it. Accompanying neurologic signs, especially ataxia, may outlast skew deviation and be more debilitating. Journal of Neuro-Ophthalmology 2021;41:69–76 doi: 10.1097/WNO.0000000000000915 © 2020 by North American Neuro-Ophthalmology Society S kew deviation is a vertical misalignment of the eyes, or hypertropia, that is part of an abnormal ocular righting reflex called the “ocular tilt reaction,” which includes ocular torsion and head tilt (1,2). It occurs when a lesion upsets the balanced input to the interstitial nuclei of Cajal in the mesodiencephalic junction. Experimental and clinical lesions of the labyrinth, vestibular nerve, brainstem, cerebellum, and thalamus can produce this phenomenon (3). Common causes are stroke, tumors, and inflammatory demyelination. Patients generally do not notice the head tilt or a deviation in the perceived position of the horizon (“subjective visual vertical”), reporting only diplopia or blurred vision resulting from the hypertropia (3). Although many single case reports and a thorough review (3) have documented skew deviation from a variety of causes, there are only 3 reported clinical series. Two of these reports appeared before the era of high-definition brain imaging (4,5), and the third was dedicated to the study of ocular torsion (6). 69 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution Smith et al (4) described 12 cases of skew deviation, all but one of which had accompanying neurologic manifestations. Skew deviation could be comitant or incomitant, and the authors surmised that pontine lesions probably accounted for most cases. They reported 2 additional cases with isolated hypertropia and features of a decompensated superior oblique palsy that they considered imitators of skew deviation. When there are no accompanying neurologic abnormalities, they cautioned, the distinction from other causes of hypertropia is difficult. Keane (5) described the clinical features of 100 patients with skew deviation who were examined over a 4-year period as inpatients at Los Angeles County Hospital. Diagnosis was based on clinical features and, in a minority, on autopsy. No brain imaging was reported. In that cohort, 60 were judged to have suffered stroke, 14 had posterior fossa tumors, 11 had clinically presumed multiple sclerosis, and the remainder had other diagnoses. Accompanying neurologic manifestations were common, especially internuclear ophthalmoplegia, horizontal gaze paresis, sixth nerve and seventh nerve palsies, and homonymous hemianopia. The author did not specify how many patients had skew deviation as an isolated neurologic manifestation. Based on the pattern of accompanying neurologic abnormalities, skew deviation was localized to all levels of the brainstem and perhaps the cerebellum. Most patients exhibited some improvement in ocular alignment over time, but no further details about recovery were provided. In 1993, Brandt and Dieterich (6) reported that 56 (36%) of their 155-patient cohort with imagingconfirmed unilateral brainstem infarction had skew deviation. In that study, the authors’ aim was to document the prevalence of ocular torsion and the location of the lesion in relation to the direction of the torsion and the side of the higher eye. All patients had ocular torsion in at least one eye, as determined by foveal displacement on fundus photographs, and all demonstrated a tilt of the subjective visual vertical, as measured in an earlier study by having the patients position a line according to their perception of verticality in relation to the horizon (7). The authors did not state whether the patients had a head tilt or whether they had reported a tilt of the subjective visual vertical outside the experimental setting. Infarcts were distributed throughout the brainstem and thalamus, with skew occurring in thalamic infarction when the lesion extended caudally into the rostral midbrain. The hypertropic eye was always ipsilateral to lesions rostral to the midpons and contralateral to lesions caudal to the midpons. Hypertropia ranged from 2 prism diopters (PD) to 40 PD, averaging 8 PD. Given these reports, there remain several issues worthy of more detailed study: 1) in what setting is the diagnosis of skew deviation most often made? 2) what symptoms do patients typically report? 3) what are the features of the hypertropia? 4) how often is skew deviation accompanied by other neurologic manifestations and which ones are most 70 common? 5) what are the causes of skew deviation and how often can these causes be verified on brain imaging? 6) how long does skew deviation persist? 7) how long do accompanying neurologic manifestations persist? and 8) how often, and under what circumstances, is a spectacle prism successful in relieving the diplopia of skew deviation? In an attempt to answer these questions, we undertook a retrospective review of patients diagnosed with skew deviation by neuro-ophthalmologists at a single tertiary-care academic institution in which adequate documentation was available. METHODS We obtained Institutional Review Board approval for a retrospective analysis of the electronic medical records at the University of Michigan for patients with skew deviation. We defined skew deviation as an acute onset of vertical misalignment without other plausible cause, using the following exclusion criteria: 1) inadequate ocular alignment data, 2) the Parks 3-step test suggesting fourth nerve palsy, 3) lack of brain imaging, 4) imaging or clinical evidence of an orbital abnormality, 5) ptosis present at any encounter, 6) positive acetylcholine receptor antibody test or electromyography suggestive of myasthenia gravis, 7) nonabrupt onset of diplopia, 8) examinations performed without supervision of a University of Michigan faculty neuroophthalmologist, and 9) no follow-up visits to document an alternative diagnosis when original brain imaging was negative. We used the Electronic Medical Record Search Engine (EMERSE) of the University of Michigan (8) to search the records of the Neuro-Ophthalmology Clinics from 2000 to 2018 for the terms “skew” and “skew deviation.” We obtained 468 hits and excluded 329 patients, leaving a cohort of 157 patients, on whom we recorded the following data: 1) age at diagnosis and sex; 2) setting in which the diagnosis was made, including emergency department (ED), inpatient ward, follow-up outpatient visit in the NeuroOphthalmology Clinics after hospital discharge, or direct outpatient referral to the Neuro-Ophthalmology Clinics; 3) presenting chief complaint, including “diplopia,” “blurred vision,” or “other;” 4) amplitude of the vertical misalignment in PD recorded in primary gaze position, right gaze, left gaze, up gaze, down gaze, right head tilt, and left head tilt, considering the tropia to be comitant if the same eye was hypertropic in all positions of gaze, and the amplitude did not vary by more than 30% in the different positions of gaze (alignment measurements in the lying position had not been recorded on any patient); 5) degrees of torsional misalignment measured with the double Maddox rod test (deviation of the subjective visual vertical had not been recorded on any patient); 6) accompanying neurologic signs, including saccadic pursuit, nystagmus, ataxia, extremity weakness, sensory loss, and abnormalities Walter and Trobe: J Neuro-Ophthalmol 2021; 41: 69-76 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution of deep tendon reflexes, where noted; 7) cause, including imaging-confirmed stroke, imaging-unconfirmed but clinically presumed stroke, posterior fossa tumor, skew immediately after an intracranial or extracranial procedure, imaging-based demyelinating brain lesions, nontumorous cerebellar disorder, traumatic brain injury, other miscellaneous, and undetermined; 8) other pertinent imaging abnormalities; 9) time to resolution of skew deviation and accompanying neurologic signs; and 10) measures used to relieve diplopia, including occluders, prisms, and eye muscle surgery. RESULTS Demographics The cohort consisted of 157 patients, among them 81 (52%) men and 76 women with an average age of 58 years. Diagnostic Setting Initial diagnosis was made on 102 (65%) outpatients and 55 (35%) inpatients or ED patients. Among the 102 outpatients, 88 had been referred to the clinics by an ophthalmologist (37), unspecified provider type (21), neurologist (18), physiatrist (5), primary care physician (5), or neurosurgeon (2). An additional 14 (9%) outpatients were diagnosed after an earlier hospital admission or ED visit at our institution. Symptoms There were 136 (87%) patients who reported diplopia, 18 (11%) who reported blurred vision, and only 3 (2%) who reported no visual symptoms. Among the 136 patients who reported diplopia, the vertical misalignment averaged 8 PD (median 5 PD), whereas among the 18 patients who reported blurred vision, the vertical misalignment averaged only 3 PD (median 2 PD), being larger than 2 PD in the primary position in only 3 patients. The 3 patients who reported no visual symptoms had vertical misalignment averaging 5 PD, but one patient had pseudophakic monovision and the other 2 patients had reduced visual acuity in one eye. No patient reported a tilt of the visual environment, but documentation of whether that symptom was explored was not available. In 91 (58%) patients, diplopia or blurred vision were accompanied by nonvisual symptoms, including dizziness (22), imbalance (22), headache (21), weakness (11), slurred speech (9), nausea (7), numbness (5), extremity weakness (4), oscillopsia (3), drowsiness (2), hearing loss (2), tinnitus (2), facial weakness (2), facial numbness (2), facial pain (1), and visual field loss (1). In the 63 (40%) patients who had diplopia or blurred vision as the only recorded new symptoms, 41 (65%) had an abnormal neurologic examination, including nystagmus (27), vertical or horizontal gaze deficits (8), internuclear Walter and Trobe: J Neuro-Ophthalmol 2021; 41: 69-76 ophthalmoplegia (7), ataxia (3), facial weakness (3), slow saccades (2), visual field defect (2), dorsal midbrain syndrome (2), Horner syndrome (1), or increased deep tendon reflexes (1). Features of Vertical Ocular Misalignment The vertical ocular misalignment was comitant in 100 (64%) patients and incomitant in 57 (36%) patients. Among the 57 patients with incomitant misalignment, 16 patients had hypertropia that reversed from a right hypertropia in right gaze to a left hypertropia in left gaze (“alternating skew deviation”). There were no patients with intermittent (“paroxysmal”) skew deviation. Head tilt was not recorded in any patient. The amplitude of vertical misalignment in primary gaze position ranged from 1 to 30 PD with an average of 7 PD and a median of 5 PD. The amount of vertical misalignment did not differ substantially according to the cause of skew deviation. Torsional alignment was measured in only 33 patients. Among them, there were 6 patients with excyclodeviation (average 4°) and 4 patients with incyclodeviation (average 3°). None of these 10 patients reported perceiving torsional misalignment or a shift in the subjective visual vertical, but there was inadequate documentation that these symptoms were queried. Binocular torsion was not documented. There was no correlation between the amount of hypertropia and the frequency of a pertinent imaging abnormality or accompanying neurologic signs. Thus, among the 51 patients with a hypertropia of more than 7 PD, 42 (82%) had a pertinent imaging abnormality and 44 (86%) had accompanying neurologic signs. Among the 41 patients with hypertropia of 3 PD or less, a pertinent imaging abnormality was seen in 31 (76%) and 38 (93%) had accompanying neurologic signs. Causes Stroke Stroke was imaging-confirmed in 67 patients and clinically presumed in 15 patients, for a total of 82 (52%) patients. In 13 of these patients, the stroke followed within 24 hours of a procedure. In 10 of them, the stroke occurred after an extracranial procedure (3 cardiac catheterizations, 3 transaortic valve replacements, 1 each of coronary artery bypass graft, mitral valve surgery, popliteal artery bypass surgery, and cesarean section). In 3 patients, the stroke occurred after an intracranial procedure (1 trans-sphenoidal surgery, 1 meningioma resection, and 1 basilar artery stenting). In the 67 patients with imaging-confirmed stroke, there was a pertinent lesion in the thalamus (17), pons (16), cerebellum (14), midbrain (12), medulla (9), unspecified location within the brainstem (7), or “cerebellopontine angle” (2). Ten of these patients had lesions in 2 locations that could have accounted for skew deviation. The imaged 71 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution strokes were ischemic in 43 patients and hemorrhagic in 14 patients. Among these patients, 62 (93%) had at least one other neurologic sign, and 47 (70%) had 2 or more other neurologic signs. The distribution of strokes did not differ in the patients with comitant, incomitant, and alternating skew deviation. In the 15 patients without pertinent imaging abnormalities, stroke was clinically presumed in 13 patients on the basis of age, risk factors, and accompanying neurologic manifestations. In the remaining 2 patients, stroke was clinically presumed because it immediately followed surgery and could not be explained by direct surgical injury. Tumor There were 15 patients (10%) who developed skew deviation in the setting of a posterior fossa tumor (3 metastatic breast cancer, 2 glioblastoma multiforme involving the brainstem, 2 primary central nervous system (CNS) lymphoma, 1 pilocytic astrocytoma, 1 medulloblastoma, 1 cavernoma, 1 fourth ventricular cyst, 1 meningioma, 1 metastatic non-small-cell lung cancer, 1 metastatic lymphoma, and 1 metastatic urothelial carcinoma). All but 1 patient (primary CNS lymphoma) had neurologic signs in addition to skew deviation. Operative Injury There were 15 patients (10%) who developed skew deviation immediately after intracranial surgery for brainstem or thalamic tumors in which imaging-confirmed surgical trauma rather than stroke was judged to be the cause of the skew deviation (3 cavernomas, 2 subependymomas, 2 meningiomas, and 1 ependymoma, 1 schwannoma, 1 glioblastoma, 1 anaplastic astrocytoma, 1 pilocytic astrocytoma, and 1 medulloblastoma). Skew deviation immediately followed 2 intracranial procedures for nontumorous conditions (removal of an intraventricular shunt and posterior fossa decompression for trigeminal neuralgia). All but 1 patient (pilocytic astrocytoma) had accompanying neurologic signs. Nontumorous Cerebellar Disorders Skew deviation occurred in 10 (6%) patients with nontumorous cerebellar disorders, including 5 patients with spinocerebellar atrophy, 3 with a paraneoplastic disorder (1 anti-GAD 65, 1 anti-Yo, and 1 with metastatic uterine cancer but negative paraneoplastic panel and no lesions on MRI), and 2 with inherited cerebellar degeneration. All 10 patients had other neurologic signs, including nystagmus, saccadic pursuit, or ataxia. Eight patients had MRI abnormalities consistent with marked cerebellar volume loss, and 2 had normal MRI scans. Demyelination Skew deviation developed in 7 (4%) patients with MRI abnormalities compatible with brainstem demyelination (5 pons, 2 cerebellum, 1 midbrain, and 1 thalamus). Among 72 them, 4 had had no previous diagnosis of multiple sclerosis at the time of diagnosis of skew deviation. All 7 patients had accompanying neurologic signs, including internuclear ophthalmoplegia, nystagmus, ataxia, saccadic pursuit, gaze paresis, or increased deep tendon reflexes. Traumatic Brain Injury Skew deviation occurred in 4 (3%) patients after head trauma. All had other neurologic signs, including nystagmus, quadriplegia, or square wave jerks. Among these 4 patients, 3 had diffuse axonal injury signs on computed tomography (CT) or MRI, and 1 had a normal CT as the only brain imaging study. Posterior Circulation Aneurysm Skew deviation occurred in 2 (1%) patients with angiographically proven brainstem saccular aneurysms with mass effect (1 midbasilar and 1 vertebrobasilar junction). Both had other neurologic signs, including sixth nerve palsy, facial hypesthesia, ataxia, or nystagmus. Miscellaneous Causes Miscellaneous causes accounted for 5 cases of skew deviation (1 Chiari malformation, 1 brainstem encephalitis, 1 neurosarcoidosis, 1 Parkinson disease, and 1 autoimmune encephalomyelitis). All 5 patients had other neurologic signs, and 2 (Chiari malformation and neurosarcoidosis) had a pertinent imaging abnormality. Undetermined Cause Skew deviation without accompanying neurologic signs or supportive imaging abnormalities was diagnosed in 17 (11%) patients on 2 separate clinic visits at least 3 months apart. In this group, hypertropia averaged 6 PD (range 2–16 PD, median 5 PD). Accompanying Neurologic Signs Among the 157 patients, 133 (85%) had at least one other neurologic sign besides skew deviation, with 81 (52%) having 2 or more neurologic signs (Table 1). These signs included nystagmus in 63 patients (gaze-evoked horizontal jerk in 16, torsional in 10, upbeat in 6, downbeat in 5, convergence retraction in 2, Bruns type in 1, mixed in 9, and unspecified in 14), gaze or ductional deficits in 34 (vertical gaze paresis or palsy in 18, horizontal gaze paresis or palsy in 6, complete gaze palsy in 2, and abduction deficit in 8), ataxia in 24 (affecting gait in 4, speech in 2, appendicular movement in 3, trunk in 1, mixed in 1, and unspecified in 14), saccadic pursuit in 23, and internuclear ophthalmoplegia in 18 (bilateral in 3). Skew was part of a dorsal midbrain syndrome (documented as a constellation of light-near dissociation, vertical gaze palsy, and convergence retraction nystagmus) in 6 and was accompanied by saccadic deficits in 5 (slowed saccades in 3 and saccadic dysmetria in 2), Horner syndrome in 4, visual field deficits Walter and Trobe: J Neuro-Ophthalmol 2021; 41: 69-76 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution TABLE 1. Types of accompanying neurologic signs in 133 patients with skew deviation Accompanying Neurologic Signs No. of Patients With This Sign Nystagmus Gaze paresis Ataxia Saccadic pursuit Internuclear ophthalmoplegia Abduction deficit Dorsal midbrain syndrome Hemiparesis Horner syndrome Visual field defect Saccadic slowing Dysarthria Impaired cognition Wallenberg syndrome Saccadic dysmetria Afferent pupil defect Quadriplegia 63 26 24 23 18 8 6 4 4 4 3 3 3 3 2 1 1 in 4, lateral medullary (Wallenberg) syndrome in 3, dysarthria in 3, hemiparesis or hemiplegia in 4, impaired cognition in 3, and by quadriplegia and afferent pupillary defect in 1 case each. Among the 24 (15%) patients with skew deviation as an isolated neurologic sign, a pertinent imaging abnormality was present in 7 (38%) patients. That left 17 patients, or 11% of the entire cohort, with skew deviation as an isolated neurologic sign and no pertinent imaging abnormality. By the last follow-up visit, no new diagnostic information had appeared in those 17 patients and hypertropia persisted in 7 (41%) of them. Brain Imaging Brain imaging included MRI alone in 118 patients, CT alone in 22, and MRI and CT in 17 (Table 2). Across the entire cohort, correlative lesions were present in 119 patients (76%). There was no association between correlative imaging lesions and amount of hypertropia. In other words, imaging-positive patients had an average hypertropia of 8 PD (median 5 PD), whereas imaging-negative patients had an average hypertropia of 6 PD (median 4 PD). In the 15 patients with clinically presumed stroke and no correlative lesions on imaging, 9 had undergone MRI and 6 had undergone only CT. Among the 15 patients with operative injury as the cause of skew deviation, all had nonstroke lesions on MRI. In the 7 patients with demyelination as the cause of skew deviation, all had pertinent brainstem or diencephalic demyelinating lesions on MRI. Of the 10 patients with a cerebellar disorder as the cause of skew deviation, all but 2 had an abnormal MRI. Of the 4 patients with traumatic brain injury as the cause of Walter and Trobe: J Neuro-Ophthalmol 2021; 41: 69-76 skew deviation, 1 had diffuse axonal injury evident on MRI, 2 had a diffuse axonal injury evident on CT, and 1 had a normal CT. Resolution of Skew Deviation At least one follow-up visit was recorded in 137 patients, with an interval to last follow-up visit ranging between 0.25 and 132 months (average 14 months, median 4.5 months) (Tables 2 and 3). Resolution of skew was documented in 58 (42%) patients, one of whom needed eye muscle surgery to produce it. The likelihood of spontaneous resolution of skew deviation varied greatly according to its cause (Table 2). It was most likely to resolve in demyelination (86%), less often in undetermined causes (59%), traumatic brain injury (50%), post-operative injury (40%), and stroke (39%), rarely in brainstem tumor (13%), and not at all in posterior circulation aneurysms and nontumorous cerebellar disorders. Resolution of skew deviation occurred within 3 months in 33 patients, between 3 and 6 months in 7, between 6 and 12 months in 6, and after 12 months in 12 (Table 3). Resolution of Accompanying Neurologic Signs Among 137 patients with at least one follow-up visit, 120 patients had accompanying neurologic signs at the initial visit (Tables 2 and 3). Accompanying signs had partially resolved in 37 (31%) patients and completely resolved in 31 (26%) of these patients by the last follow-up examination, leaving 52 (43%) of the cohort with persistent accompanying neurologic signs. When resolution of accompanying neurologic occurred, it did so within 3 months of diagnosis in 27 patients, between 3 and 6 months in 10 patients, between 6 and 12 months in 8 patients, and after more than 12 months in 23 patients. Thus, accompanying neurologic signs generally lingered longer than did skew deviation. Among these signs was ataxia, which was often more debilitating than the diplopia (Table 3). As with skew deviation, the likelihood of recovery of accompanying neurologic signs varied according to their cause. Partial or complete resolution occurred most often in demyelination (86%), less often in tumor (53%), stroke (52%), aneurysm (50%), and operative injury (33%), rarely in nontumorous cerebellar disorders (10%), and not at all in traumatic brain injury. Among the 17 patients with skew deviation of undetermined origin, 7 (41%) had failed to resolve at the last follow-up visit (Table 2). Measures to Relieve Diplopia Of the 110 patients who still had diplopia on follow-up examinations, prism spectacles successfully relieved diplopia in 68 (62%). The cause of the skew deviation, the size of the primary position vertical misalignment, and the presence or absence of incomitance did not predict success in 73 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution TABLE 2. Clinical and imaging features in our cohort of 157 patients with skew deviation Cause of Skew Deviation (No. of Cases) Stroke (82) Hemorrhagic (14) Ischemic (68) Procedure-related (13) Tumor (15) Operative injury (15) Nontumorous cerebellar disorders (10) Demyelination (7) Traumatic brain injury (4) Posterior circulation aneurysm (2) Miscellaneous (5) Undetermined (17) *One † One ‡ One § One Presence of Pertinent Imaging Abnormality Presence of Accompanying Neurologic Signs Resolution of Skew Deviation by Last Follow-up Visit‡ Improvement or Resolution of Accompanying Neurologic Signs by Last Follow-up Visit§ 82% 100% 78% 85% 100% 100% 80% 94% 100% 93% 77% 93% 93% 100% 39% 29% 41% 46% 13% 40% 0% 52% 43% 54% 62% 53% 33% 10% 100% 75%* 100% 100% 100% 100% 86%† 50% 0% 86%† 0% 50% 40% 0% 100% 0% 0% 59% 80% NA patient with negative imaging had CT only. patient had last follow-up visit at 0.5 months. hundred thirty-seven patients had at least follow-up visit. hundred twenty patients with accompanying neurologic signs at initial visit had at least one follow-up visit. relieving diplopia. Thus, among the 68 patients in whom prism spectacles relieved diplopia, the primary gaze position hypertropia averaged 5 PD (range 1 PD–30 PD, median 3 PD). In that group, 47 (69%) had a comitant hypertropia and 21 (31%) had an incomitant hypertropia, including 5 patients who had a right hypertropia on right gaze and a left hypertropia on left gaze. Among the 42 patients in whom prism spectacles did not relieve diplopia, the average primary gaze position hypertropia averaged 6 PD (range 1–30 PD, median 4 PD). In that group, 24 (57%) had a comitant deviation and 18 (43%) had an incomitant deviation, including 8 patients who had a right hypertropia on right gaze and a left hypertropia on left gaze. In the 42 patients in whom prism glasses did not relieve diplopia, 2 had diplopia relieved with an opaque contact lens in 1 eye, and 6 had diplopia relieved by eye muscle surgery (1 surgery for 5 patients and 3 surgeries for 1 patient). DISCUSSION This reported series on patients with skew deviation, the largest to date, confirms and further quantitates clinical impressions derived from smaller reported series (4,5) and adds some valuable new information. Based on this and earlier studies, skew deviation is usually embedded in a complex of neurologic manifestations reflecting dysfunction at any level of the brainstem, as well as the thalamus. At least one other neurologic sign accompanied skew deviation in 85% of our cohort. Nystagmus was most common, appearing in nearly half of the cohort. Gaze paresis, saccadic pursuit, and internuclear ophthalmoplegia were other common neuro-ophthalmic abnormalities, as noted in an earlier series (5). Ataxia was the most common nonophthalmic manifestation. Importantly, in the 63 patients reporting diplopia or blurred TABLE 3. Time to resolution of skew deviation and accompanying neurologic signs Time Interval From Initial Diagnostic Visit #3 months .3 #6 months .6 #12 months .12 months Total Resolution of Accompanying Neurologic Signs† Complete Resolution of Skew Deviation* Partial Resolution Complete Resolution 33 7 6 12 58 (42%) 10 7 5 15 37 (30%) 17 3 3 8 31 (26%) *In 137 patients with at least one follow-up visit. † In 120 patients with at least one follow-up visit who displayed accompanying neurologic signs at the initial visit. 74 Walter and Trobe: J Neuro-Ophthalmol 2021; 41: 69-76 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution vision as the only symptom, neurologic signs were found in 65%, making the search for these accompanying features particularly important from a diagnostic point of view. Because these accompanying neuro-ophthalmic signs were often subtle, and the neurologic examination sometimes incomplete, we suspect that they may have been even more prevalent and often overlooked. Stroke accounted for half of causes, with lesions distributed throughout the brainstem, cerebellum, and thalamus. The thalamus emerged as the most common site of stroke, a finding that has not been emphasized. Although brain imaging could not verify the full extent of the thalamic lesion, it probably reached into the rostral termination of the pathway mediating vertical ocular alignment. As suggested by Keane (5) in the preimaging era, strokes proved to be more often ischemic than hemorrhagic. Although most were unprovoked, an important number immediately followed intracranial and extracranial procedures, including heart catheterization, heart valve surgery, and bypass grafting, as well as popliteal artery bypass surgery, and even cesarean section. Brainstem stroke in these settings is a known phenomenon, perhaps because of autonomic dysregulation or emboli (9). Notably, there were 9 patients in our study with a strong clinical presumption of a provoked brainstem stroke in whom MRI was negative. Even high-definition brain MRI does not always detect stroke in the brainstem (10). Aside from stroke, brainstem tumor and operative injury were common causes. Demyelination, traumatic brain injury, posterior circulation aneurysms, and miscellaneous causes accounted a much smaller proportion of cases. An important subgroup of our cohort consisted of 24 patients with skew deviation who had no accompanying neurologic manifestations. In that group, sometimes called “ambulatory skew deviation,” only 7 patients had pertinent brain imaging abnormalities, leaving 17 with neither imaging nor clinically supportive information toward a diagnosis. CT was the only study in 6 of them; MRI might have detected the lesion. In the 11 patient studies with MRI, we surmise that skew deviation might have been caused by a brainstem stroke too small to be detected even on MRI (10). In this study, the pattern of ocular misalignment was incomitant in 1/3 of cases. Within that group, 1/3 showed reversal of the hypertropia with lateral gaze (“alternating skew deviation”). Although these phenomena have been well described, this report provides the first quantitation of their relative prevalence in a large group of patients. Previous reports have mentioned that the hypertropia of skew deviation may show improvement over time, but no specific details have been provided, probably because of lack of follow-up. In this study, with a median follow-up interval of 4.5 months, spontaneous resolution of hypertropia and diplopia was observed in fewer than 50% of patients, varying widely according to cause (Table 2). If skew deviWalter and Trobe: J Neuro-Ophthalmol 2021; 41: 69-76 ation was of undetermined origin, it had a slightly better than 50% chance of disappearing. Among all causes of skew deviation, recovery usually occurred within the first 3 months but was often delayed for longer than 12 months (Table 3). The frequently accompanying neurologic signs often proved to be more debilitating and enduring than the skew deviation. Improvement was noted in only slightly more than half of the cohort. When improvement did occur, it was often delayed 12 months or more. We acknowledge that follow-up was incomplete and that patients may not have returned because they had recovered or sought care elsewhere. A critical finding in this study was that 62% of the patients with enduring diplopia could have it relieved with spectacle prisms. Surprisingly, neither the cause of the skew, the size of the hypertropia, nor the degree of incomitance predicted success in relieving diplopia. Surprisingly, diplopia associated with large hypertropia could sometimes be palliated with spectacle prisms, suggesting that some patients may be reporting relief of diplopia even when ocular misalignment persists. This study further delineates the clinical and imaging profile of skew deviation. Its strengths are a large patient cohort with adequate clinical documentation confirmed on examination by neuro-ophthalmologists, strict exclusion criteria to avoid inclusion of clinical imitators of skew deviation, high prevalence of high-definition neuroimaging, and a search engine capable of detecting any mention of skew deviation in electronic medical record text. But there are also weaknesses inherent in such a retrospective study. Examination techniques were performed by 3 different neuro-ophthalmologists in a nonstandard fashion, often generating incomplete data. Symptoms of alteration in the subjective visual vertical were not elicited, and measurements of postural differences in hypertropia, ocular torsion, and a disturbed visual vertical were not often performed. Acknowledging these study strengths and weaknesses, the relevant conclusions from this study are as follows: 1) Most patients with skew deviation will report a visual disturbance—usually diplopia, but if the hypertropia is less than 3 PD, they may describe blurred vision. 2) The hypertropia of skew deviation will be incomitant in an important minority of cases; within that subgroup, the hypertropia often reverses in gaze from side to side (“alternating skew deviation”). 3) Skew deviation is usually accompanied by other neurologic manifestations, especially nystagmus, saccadic pursuit, horizontal gaze paresis, internuclear ophthalmoplegia, and ataxia; such accompanying signs, often subtle, are important in distinguishing skew deviation from other causes of acute hypertropia, especially if visual symptoms are isolated or most prominent. 75 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution 4) An important subgroup, amounting to 11% in this study, will consist of patients with skew deviation without other neurologic abnormalities or correlative imaging abnormalities; perhaps they have had a stroke too small to show up on current brain imaging. 5) Stroke accounts for most cases of skew deviation, occurring at any level of the brainstem, including the cerebellum, and especially in the thalamus; other important causes are posterior fossa tumors, operative injury, brainstem demyelination, nontumorous cerebellar disorders, and traumatic brain injury. 6) Skew deviation may be persistent even among patients in whom it is the only clinical manifestation. 7) Diplopia associated with persistent skew deviation may be relieved with spectacle prisms in slightly more than half of cases, even among patients with incomitance and large vertical misalignment. 8) The neurologic manifestations accompanying skew deviation, especially ataxia, may be more debilitating and enduring than the diplopia. The important management implications of this profile are the following: 1) Attention must be paid to detection of accompanying neurologic features, which are common and critical for diagnosis. 2) In the small proportion of cases where acute hypertropia is the only clinical abnormality and high-definition brain imaging is negative, the diagnosis of skew deviation is viable but must be questioned. 3) Patients should be advised that skew deviation may last for months or even indefinitely, and that accompanying neurologic manifestations, especially ataxia, may be more debilitating. 76 4) Prism spectacles may alleviate diplopia even when the misalignment is greater than 3 PD and incomitant; when that intervention fails, eye muscle surgery may be effective in relieving diplopia. STATEMENT OF AUTHORSHIP Category 1: a. conception and design: E. Walter and J. D. Trobe; b. acquisition of data: E. Walter and J. D. Trobe; c. analysis and interpretation of data: E. Walter and J. D. Trobe. Category 2: a. drafting the manuscript: E. Walter and J. D. Trobe; b. revising it for intellectual content: E. Walter and J. D. Trobe. Category 3: a. final approval of the completed manuscript: E. Walter and J. D. Trobe. REFERENCES 1. Brandt T, Dieterich M. Different types of skew deviation. J Neurol Neurosurg Psychiatry. 1991;54:549–550. 2. Wong AMF. Understanding skew deviation and a new clinical test to differentiate it from trochlear nerve palsy. J AAPOS. 2010;14:61–67. 3. Brodsky MC, Donahue SP, Vaphiades M, Brandt T. Skew deviation revisited. Surv Ophthalmol. 2006;51:105–128. 4. Smith JL, David NJ, Klintworth G. Skew deviation. Neurology. 1964;14:96–105. 5. Keane JR. Ocular skew deviation. Arch Neurol. 1975;32:185–190. 6. Brandt T, Dieterich M. Skew deviation with ocular torsion : a vestibular brainstem sign of topographic diagnostic value. Ann Neurol. 1993;33:528–534. 7. Dieterich M, Brandt T. Ocular torsion and perceived vertical in oculomotor, trochlear and abducens nerve palsies. Brain. 1993;116:1095–1104. 8. Hanauer DA, Mei Q, Law J, Khanna R, Zheng K. Supporting information retrieval from electronic health records : a report of University of Michigan’s nine-year experience in developing and using the Electronic Medical Record Search Engine (EMERSE). J Biomed Inform. 2015;55:290–300. 9. Suzuki T, Nishio M, Chikuda M, Takayanagi K. Skew deviation as a complication of cardiac catheterization. Am J Ophthalmol. 2001;132:282–283. 10. Edlow BL, Hurwitz S, Edlow JA. Diagnosis of DWI-negative acute ischemic stroke. Neurology. 2017;89:256–262. Walter and Trobe: J Neuro-Ophthalmol 2021; 41: 69-76 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited.