Dorsal Midbrain Syndrome: Clinical and Imaging Features in 75 Cases

Dorsal midbrain syndrome (DMS) consists of a constellation of clinical features, including reduced upgaze, pupillary light-near dissociation, lid retraction, convergence retraction, and eye misalignment. This syndrome results mostly from intrinsic or extrinsic mesodiencephalic tumors or strokes, obstructive hydrocephalus, failure of cerebrospinal fluid shunting to correct obstructive hydrocephalus, and head trauma. Published reports that include imaging corroboration are based on relatively small cohorts and have not included comprehensive patient self-reports on the impact of these abnormalities on quality of life.

Original Contribution Section Editors: Clare Fraser, MD Susan Mollan, MD Dorsal Midbrain Syndrome: Clinical and Imaging Features in 75 Cases Jonah E. Yousif, BS, Eric Liao, MD, Jonathan D. Trobe, MD Background: Dorsal midbrain syndrome (DMS) consists of a constellation of clinical features, including reduced upgaze, pupillary light-near dissociation, lid retraction, convergence retraction, and eye misalignment. This syndrome results mostly from intrinsic or extrinsic mesodiencephalic tumors or strokes, obstructive hydrocephalus, failure of cerebrospinal fluid shunting to correct obstructive hydrocephalus, and head trauma. Published reports that include imaging corroboration are based on relatively small cohorts and have not included comprehensive patient self-reports on the impact of these abnormalities on quality of life. Methods: We conducted a retrospective review of cases of DMS identified between 1998 and 2019 at the University of Michigan using the Electronic Medical Record Search Engine. Patients were included only if they had been evaluated by a neuro-ophthalmologist and had a corroborative imaging abnormality. We collected data on symptoms and on neuro-ophthalmic and neurologic signs. We reviewed brain imaging reports on all 75 patients, and the study neuroradiologist analyzed the imaging in 57 patients. Using a uniform list of questions, we conducted telephone interviews of 26 patients to assess lingering symptoms and their impact on quality of life. Results: There were 75 patients, only 5 of whom were younger than 10 years. Neoplasms accounted for 47%, strokes (mostly thalamic) for 25%, nonneoplastic masses for 12%, nonneoplastic hydrocephalus for 7%, traumatic brain injury for 5%, and demyelination for 4%. Reduced upgaze occurred in 93% of patients, being completely absent or reduced to less than 50% amplitude in 67%. Convergence retraction on attempted upgaze occurred in 52%, horizontal misalignment in 49%, vertical misalignment in 47%, and pupillary light-near dissociation in 37%. Optic Kellogg Eye Center, Department of Ophthalmology and Visual Sciences (JEY, JDT), Department of Radiology (Neuroradiology) (EL), and Department of Neurology (JDT), University of Michigan, Ann Arbor, Michigan. The authors report no conflicts of interest. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Web site (www. jneuro-ophthalmology.com). Address correspondence to Jonathan D. Trobe, MD, Kellogg Eye Center, Department of Ophthalmology and Visual Sciences, University of Michigan, 1000 Wall Street, Ann Arbor, MI 48105; E-mail: [e-mail] e644 neuropathy attributed to chronic papilledema occurred in only 3%. Three or more neuro-ophthalmic signs were present in 84%, and only 4% had a single sign—reduced upgaze. Imaging features did not correlate with the frequency or severity of clinical signs. There was some improvement in the clinical signs among the patients with stroke but no change among the patients with neoplasms. In the 26 telephone interviews, patients with neoplasms reported that imbalance had a greater impact on quality of life than did diplopia. Patients with strokes reported that imbalance had the greatest impact initially but that its effect dissipated. Neither group reported lingering effects of impaired upgaze. Conclusions: This large series expands on the clinical profile of DMS. Neoplasms and strokes were the most common causes. Obstructive hydrocephalus alone, identified as a major cause in the largest previously published series, was uncommon. At least 3 neuro-ophthalmic signs were present in nearly all patients, with upgaze deficit as predominant. Unlike an earlier report, this study found no correlation between brain imaging and clinical signs. Neuroophthalmic signs persisted even after neoplasms were successfully treated and improved only slightly after stroke. Telephone interviews with patients revealed that diplopia and upgaze deficit had less lasting impact on quality of life than did ataxia and concurrent nonneurologic problems. Journal of Neuro-Ophthalmology 2021;41:e644–e654 doi: 10.1097/WNO.0000000000001052 © 2020 by North American Neuro-Ophthalmology Society D orsal midbrain syndrome (DMS) consists of a constellation of clinical features, including reduced upgaze, pupillary light-near dissociation, lid retraction, convergence retraction, and ocular misalignment (1). These abnormalities, which arise from damage to rostral midbrain structures that mediate vertical gaze, vertical alignment, and the pupillary light reflex, are largely caused by intrinsic or extrinsic mesodiencephalic tumors and strokes, demyelination, head trauma, and failure of cerebrospinal fluid shunting in aqueductal stenosis (1,2). The relative prevalence of causes of DMS has varied in the published literature. In a 1990 study of 206 patients with sparse imaging corroboration, Keane (3) found that Yousif et al: J Neuro-Ophthalmol 2021; 41: e644-e654 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution hydrocephalus accounted for nearly 40% of cases, often precipitated by cysticercosis. Strokes accounted for 26% and neoplasms for 22%. Data on resolution of clinical manifestations were not reported. A study of 40 patients by Shields et al (4) found that midbrain stroke accounted for 50% of cases, pineal tumors for 30%, and midbrain tumors for 15%; nonneoplastic hydrocephalus was not listed as a cause. In that study, the authors did not state how many cases were supported by imaging reports or by their imaging review. A study of 71 patients with imaging-supported DMS (5), largely aimed at clinical-imaging correlation, found that 63% had strokes or intrinsic midbrain tumors, 37% had pineal tumors, and only 8% had hydrocephalus alone. Among patients with pineal tumors, 50% showed improvement in symptoms, whereas among patients with intrinsic midbrain lesions, only 24% showed improvement. MRI signal abnormalities did not correlate with resolution of DMS symptoms. A study of MRI abnormalities in 77 patients with pineal tumors (6) found that the presence of intrinsic midbrain signal abnormalities, but not the size of the tumor or its imaging relationship to the dorsal midbrain, was associated with clinical DMS abnormalities. An imaging-supported study of 26 patients with DMS (7) found that 38% were caused by midbrain tumors, 31% by pineal tumors, 19% by strokes and vascular malformations, and only 8% by hydrocephalus alone. Only 33% of patients showed some improvement in clinical signs, with only 8% showing “complete clinical recovery.” The degree of hydrocephalus did not predict the number of clinical DMS signs or their resolution. Reported series of DMS limited to children and young adults with pineal tumors have found that clinical DMS abnormalities endure even when the tumor has been successfully treated (8–12). One study (11) explored the impact on quality of life in 16 patients with pineal tumors using a standard cancer quality of life questionnaire. The authors found that lingering diplopia was the only symptom that had an adverse effect, but the questionnaire did not include any questions related to upgaze deficiency or imbalance. We undertook a retrospective review of cases of DMS examined by neuro-ophthalmologists at a single tertiary care academic medical center with the objective of assessing the causes of DMS, its clinical manifestations and their correlation with imaging features, the evolution of clinical manifestations over the course of time, and their impact on quality of life as reported by the patients. Based on clinical experience, we hypothesized that 1) some clinical manifestations would persist indefinitely; 2) imaging features would predict the severity of persisting clinical signs; 3) optic neuropathy from chronic papilledema would be prominent and cause lasting visual impairment; and 4) lingering ataxia would have a greater impact on quality of life than diplopia. Yousif et al: J Neuro-Ophthalmol 2021; 41: e644-e654 METHODS We obtained Institutional Review Board approval for a retrospective analysis of the electronic medical records at the University of Michigan of patients with DMS and telephone interviews with the patients based on a standardized questionnaire (see Supplemental Digital Content 1, Appendix, http://links.lww.com/WNO/A434). Patient Selection We used the Electronic Medical Record Search Engine of the University of Michigan to search the records of the Neuro- Ophthalmology Clinics from 1998 to 2019 for the terms “dorsal midbrain syndrome,” “parinaud,” “pretectal,” “pineal,” and “midbrain stroke.” Patients were included only if they had been evaluated by a neuroophthalmologist and had had an imaging study concurrent with that examination. Included patients met the following criteria: 1) at least 1 neuro-ophthalmic sign of dorsal midbrain dysfunction, including upgaze reduction, convergence retraction, eye misalignment, and light-near dissociation and 2) imaging evidence of a lesion near or within the dorsal midbrain or hydrocephalus. We excluded 15 patients who had at least 3 neuroophthalmic signs of DMS but who did not have a correlative imaging abnormality. Among them were 6 patients who had shunted hydrocephalus but no imaging signs of enlarging ventriculomegaly at the onset of the DMS, 3 who had a history of traumatic brain injury but no correlative imaging abnormality, and 6 who had neither a definite cause for DMS nor a correlative imaging abnormality. In 10 of these 15 patients, computed tomography (CT) was the only available imaging study. In identifying the cause of DMS, we applied the following definitions: 1) Neoplasms: imaging evidence of a benign or malignant tumor adjacent to or within the dorsal midbrain; hydrocephalus might also be present. 2) Nonneoplastic masses: arteriovenous malformation (AVM), cavernous malformation, brain aneurysm, and pineal cyst adjacent to or within the dorsal midbrain; hydrocephalus might also be present. 3) Strokes: clinical and imaging evidence of an acute ischemic or hemorrhagic lesion adjacent to or within the dorsal midbrain without hydrocephalus. 4) Nonneoplastic hydrocephalus: imaging evidence of obstructive hydrocephalus at the aqueduct or fourth ventricle without an intracranial mass lesion or inflammation. 5) Traumatic brain injury: history of concussive brain trauma and imaging evidence of compression or hemorrhage adjacent to or within the dorsal midbrain without hydrocephalus. e645 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution 6) Demyelinative lesion: clinical and imaging evidence of an inflammatory lesion adjacent to or within the dorsal midbrain without hydrocephalus. Data Collection of Clinical Features From the electronic medical records, we recorded patient gender and age at diagnosis, date and setting (inpatient or outpatient) at which the DMS was first identified, its cause and treatment, and duration of follow-up, defined in months from identification of the DMS to the last followup visit. From the initial neuro-ophthalmologic and last follow-up visit, we recorded symptoms, neuro-ophthalmic signs (best-corrected visual acuity, visual field, pupil size and constriction, ocular ductions and versions, eye alignment, convergence retraction, lid retraction, nystagmus, papilledema, and presence of optic disc pallor), and accompanying neurologic signs (ataxia, extremity weakness, sensory loss, and abnormalities of deep tendon reflexes). Definitions of Clinical Features Among eye movement abnormalities, upgaze and downgaze deficits were graded as follows: 1%–50% reduction (Grade 1), 51%–70% reduction (Grade 2), 71%–100% (Grade 3), and tonic upgaze or downgaze deviation (Grade 4). Convergence retraction was defined as visible backward ± convergent movement of both eyes on attempted upgaze; it was not graded. Horizontal and vertical eye misalignment was graded in prism diopters (PD) with the cover and/or single Maddox rod tests. Torsional misalignment was graded in degrees based on the double Maddox rod test. Horizontal misalignment was recorded as esotropia or exotropia and whether it was comitant or incomitant, with incomitance defined as a difference of at least 5 PD between primary and lateral gaze positions. Vertical misalignment was documented as having been attributed by the examiner to skew deviation or fourth cranial nerve palsy. Pupil reflexes were assessed as to whether there was adequate constriction to a light directed onto the eye with the patient fixating on a distant target. If constriction was judged to be impaired with this maneuver, pupil constriction to a target placed at reading distance was assessed. If the pupil constricted inadequately to a light and to a near target, we diagnosed “pupil hyporeflexia.” If the pupil constricted inadequately to a light but adequately to a near target, we diagnosed “light-near dissociation.” Pathologic anisocoria was defined as a difference in pupil diameter greater than 0.5 mm in dim illumination if there was also impaired pupil constriction to light. Lid retraction was defined was pathologic elevation of the upper lid in both eyes with scleral show; it was not graded. Nystagmus was defined as involuntary rhythmic oscillations of both eyes in any position of gaze. The gaze position and plane of the nystagmus, as well as the direction of the fast phase of jerk nystagmus, were recorded. Optic neuropathy attributable to previous or current papile646 ledema was based on persistently decreased visual acuity or visual field in either eye of any amount or by the presence of optic disc pallor, provided other causes for these phenomena had been excluded. Papilledema was defined as acquired elevation of the optic discs in both eyes that was visible on ophthalmoscopy and attributable to increased intracranial pressure; it was not graded. Data Collection of Imaging Features Imaging reports were available on all patients, but only 57 studies (51 MRI and 6 CT) were available for analysis by the study neuroradiologist (E.L.). As initial studies, we chose those performed on dates closest to the initial neuroophthalmologic visit. As follow-up studies for the neoplasm patients who had documented follow-up visits, we chose the studies closest to those visits. We analyzed the following features: 1) Neoplasms: from the initial studies, we assessed the degree of ventriculomegaly (on a gradient of mild/ moderate/severe) and the site of origin of the neoplasm and its relationship to the dorsal midbrain (on a gradient of lying adjacent but not abutting, compressing, or intrinsic signal abnormality); from the follow-up studies after tumor treatment, we assessed the residual signal abnormality on a gradient of no imaging abnormality, excessive T2 brightness, and dorsal midbrain atrophy. 2) Strokes: from the initial studies, we assessed the location of the stroke within the thalamus or midbrain and whether it was adjacent to or within the dorsal midbrain; there were too few patients with follow-up studies to allow meaningful review of those studies. 3) Nonneoplastic hydrocephalus: in the initial and followup studies, we assessed the degree of ventriculomegaly (on a gradient of mild/moderate/severe), and the presence or absence of an altered configuration of the dorsal midbrain. 4) Arteriovenous and cavernous malformations, aneurysm, and demyelination: in the initial studies, we assessed the site of origin of the lesion and whether it was adjacent to or within the dorsal midbrain. Assessment of Long-Term Self-Reported Morbidity We attempted to contact all 75 patients but were able to conduct telephone interviews on only 26 patients, using a standard questionnaire (see Supplemental Digital Content 1, Appendix, http://links.lww.com/WNO/A434) designed to assess lingering symptoms from DMS. The questioner (J.E.Y.) was not involved in the care of the patients. The distribution of causes of DMS among the interviewed patients was 11 neoplasms, 8 strokes, 4 nonneoplastic masses, and 3 nonneoplastic hydrocephalus. The average time from initial presentation to interview date was 6 years. Yousif et al: J Neuro-Ophthalmol 2021; 41: e644-e654 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution Six of the interviews were conducted with the patients’ caregivers because the patients were deceased, younger than 18 years, or lacked adequate cognition. Seventeen patients had had a follow-up visit, so the comparison of reported symptoms to documented deficits was based on the last follow-up visit. In the remaining 9 patients who did not have a follow-up visit, the comparison was based on the documented deficits at the initial visit. Statistical Analysis We performed bivariate analyses using IBM SPSS version 26 (Armonk, NY). Chi-square tests were used to compare proportions, and Fisher exact tests were used when there were small sample sizes among comparison groups. Linear regressions were used to define correlations between severity of imaging findings and clinical signs. For all statistical analyses, we defined a statistical significance level of 0.05. The study was approved by the Institutional Review Board (#HUM00158189 and #HUM00168753) of the University of Michigan Medical School. RESULTS Demographic Features There were 75 patients, including 47 (63.1%) men. The median age was 35 years (range 4 months–86 years). Only 5 (7%) patients were younger than 10 years. Causes Neoplasms accounted for 35 (47%) cases, including 15 germ cell tumors, 12 gliomas, 2 pineocytomas, 2 embryonal tumors with multilayered rosettes, 1 meningioma, 1 metastasis, and 2 cases of unknown histologic type. Among the neoplasms, 21 originated in the pineal gland, 12 in the midbrain, and 2 in the thalamus (Fig. 1). Among the 28 imaging studies available for our analysis (Fig. 2), 2 had lesions adjacent to (but not abutting) the dorsal midbrain (Fig. 2A), 2 had abutting lesions (Fig. 2B), 4 had compressing lesions (Fig. 2C), and 20 had infiltrating lesions (Fig. 2D). Patient age ranged from 4 months to 86 years (median 19 years). Strokes accounted for 19 (25%) cases, 13 (68%) originating in the thalamus (7 hemorrhagic and 6 ischemic) and 6 (32%) originating in the midbrain itself (4 ischemic and 2 hemorrhagic). Of the 11 (9 MRI and 2 CT) studies available for our imaging review, 10 strokes originated in the thalamus and 1 in the dorsal midbrain (Fig. 3). All 10 thalamic strokes predominantly involved the medial region; 4 extended into the lateral thalamus, 3 into the tegmental midbrain without exerting mass effect, 4 were limited to the rostral region (Fig. 3A), 2 to the caudal region (Fig. 3B), and 4 involved both the rostral and caudal regions (Fig. 3C). Eight thalamic strokes were unilateral (Fig. 3B) and 2 were bilateral (Fig. 3A). Patient age ranged from 18 to 83 years (median 63 years). Yousif et al: J Neuro-Ophthalmol 2021; 41: e644-e654 FIG. 1. Causes of dorsal midbrain syndrome in 75 patients. Nonneoplastic masses accounted for 9 (12%) cases, all of which were available for our imaging review. They included 4 AVMs (Fig. 4A), 3 cavernomas (Fig. 4B), 1 pineal cyst, and 1 brain aneurysm. Three AVMs originated in the midbrain (Fig. 4A) and 1 originated in the thalamus. The cerebral aneurysm arose from the basilar artery. Imaging artifact created by coiling interfered with assessment of the relationship of the aneurysm to the dorsal midbrain. Nonneoplastic hydrocephalus accounted for only 5 (7%) cases, 3 of which were available for our imaging review. Two patients had not been shunted, 1 with a fourth ventricular hemorrhage and 1 with a traumatic brain injury causing obstructive hydrocephalus. The remaining 3 patients with nonneoplastic hydrocephalus were identified at the time of shunt malfunction, including 1 patient with a remote history of neoplasm, 1 with a remote history of traumatic brain injury, and 1 with hydrocephalus of unknown cause (Fig. 4C). Recent traumatic brain injury accounted for DMS in 4 (5%) cases, only 1 of which was available for our imaging review—CT that provided insufficient detail. Imaging reports on the 3 patients whose MRI studies were not available for our review described third or fourth ventricular hemorrhage but did not mention compression of the dorsal midbrain. Brainstem and thalamic demyelination accounted for 3 (4%) cases, 2 of which were available for our review, both showing a high T2/fluid-attenuated inversion recovery (FLAIR) signal within or adjacent to the dorsal midbrain, and diagnosed as acute disseminated encephalomyelitis (ADEM) (Fig. 4D). The third patient was reported to have had multifocal demyelination, including a lesion “adjacent” to the dorsal midbrain. Among the neoplastic and nonneoplastic tumors and strokes, which constituted most causes of DMS, there was no significant correlation between the imaging measures of degree of ventriculomegaly or lesion impact on the dorsal midbrain and the severity or number of DMS signs. Clinical Manifestations Symptoms The most common presenting symptom was diplopia (63%), followed by blurred vision (23%), headache e647 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 2. Postcontrast T1 sagittal MRI images of neoplasms that caused dorsal midbrain syndrome. A. Astrocytoma (arrow) of the tegmental midbrain that is adjacent to the dorsal midbrain. B. Glioblastoma multiforme (arrow) of the caudal thalamus that is abutting but not compressing the dorsal midbrain. C. Pineocytoma (arrow) compressing the dorsal midbrain. D. Embryonal tumor with multilayered rosettes (arrow) infiltrating the dorsal midbrain. (20%), and imbalance (15%). Only 9 patients reported difficulty looking up. Signs The presenting signs are displayed in Table 1. Eye Movement Abnormalities There were 70 (93%) patients with reduced upgaze. Among them, 23 (33%) had Grade 1 reduction, 4 (6%) Grade 2 reduction, and 38 (54%) Grade 3 reduction. Only 5 (7%) patients had tonic downgaze deviation (Grade 4 reduction), including 2 with nonneoplastic masses, 1 with traumatic brain injury, 1 with a thalamic glioblastoma, and 1 with a thalamic hemorrhage. Reduced upgaze was the only neuro-ophthalmic sign in 3 patients. All 5 patients who did not have reduced upgaze had some combination of eye misalignment and light-near dissociation. Thirty-nine patients (52%) had convergence retraction. All of them also had an upgaze deficit, but the severity of the upgaze deficit did not predict whether convergence retraction would be present. There were 18 (24%) patients with reduced downgaze. Among them, 7 (39%) had Grade 1 reduction, 5 (28%) had Grade 2 reduction, and 6 (33%) had Grade 3 reduction. No patient had tonic upward deviation. e648 Downgaze reduction was never the only neuroophthalmic sign. All 18 patients who had this sign also had reduced upgaze and 7 (37%) had light-near dissociation. Eye Misalignment There were 56 (75%) patients with eye misalignment. Among them, 37 (49%) patients had horizontal misalignment, including 28 (76%) with exotropia, 7 (19%) with esotropia, and 2 (5%) with an unspecified pattern. Among the exotropia cases, 61% were comitant, 11% were incomitant, and 28% lacked documentation of this feature. Among the esotropia cases, 42% were comitant, 29% were incomitant, and 29% lacked documentation of this feature. The cause of horizontal incomitance was sparsely documented. Vertical misalignment was present in 35 (47%) patients; it was attributed to skew deviation in 13 (37%), fourth nerve palsy in 13 (37%), and unattributed in 9 (26%). Torsional misalignment was present in 6 (8%) patients, 4 of whom clearly met other criteria for fourth nerve palsy; the remaining 2 patients lacked adequate documentation. Pupil Abnormalities Pupil abnormalities were present in 35 (47%) patients, always accompanied by at least one other DMS sign. Yousif et al: J Neuro-Ophthalmol 2021; 41: e644-e654 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 3. Axial diffusion-weighted images of strokes that caused dorsal midbrain syndrome. A. Bilateral paramedian rostral thalamic infarcts (arrow). B. Right paramedian caudal thalamic infarct demonstrating encephalomalacia. C. Left thalamic hemorrhage involving its rostral and caudal portions (arrow). There were 7 patients with impaired constriction to a light and a near target (“pupil hyporeflexia”), and 28 patients with impaired constriction to light but preserved constriction to a near target (“light-near dissoci- ation”). Among the 35 patients with any type of impaired pupil constriction, only 11 patients had anisocoria, which was 1.5 mm or less in dim illumination in 6 patients. FIG. 4. MRI studies of miscellaneous causes of dorsal midbrain syndrome. A. T2 axial MRI image shows an arteriovenous malformation (arrow) infiltrating the dorsal midbrain. B. FLAIR image shows a pontine cavernoma (arrow) within the midbrain with surrounding edema infiltrating the dorsal midbrain. C. Precontrast T1 sagittal MRI image shows ventriculomegaly and distortion of the dorsal midbrain (arrow) in nonneoplastic hydrocephalus. D. Axial FLAIR image shows a hyperintense signal within the thalamus (arrow) in a patient with acute disseminated encephalomyelitis. FLAIR, fluid-attenuated inversion recovery. Yousif et al: J Neuro-Ophthalmol 2021; 41: e644-e654 e649 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution Lid Retraction Lid retraction was present in 20 (27%) patients, never as an isolated sign. Upgaze was reduced in 95% of those patients, eye misalignment was present in 65%, and pupil abnormalities in 55%. Nystagmus Nystagmus was documented in 15 (20%) patients, in primary gaze position in 7 patients and in eccentric gaze in 8 patients. Its apparent waveform was always jerk and its trajectory was upbeat in 8, downbeat in 1, horizontal in 4, and mixed in 2. Nystagmus was never the only sign. Optic Neuropathy Optic neuropathy was present in only 2 (3%) patients. In both patients, optic disc pallor was present in both eyes. Visual acuity was 20/400 bilaterally in one patient, and counting fingers in the right eye and 20/70 in the left eye in the other patient. There were no patients with normal visual acuity but abnormal visual fields, and no patients with afferent pupil defects. Papilledema Papilledema was present in only 4 (5%) patients, all of whom had had obstructive hydrocephalus from neoplasms in the region of the aqueduct. Visual function was intact in all of them. Constellations of Neuro-Ophthalmic Signs At least 3 neuro-ophthalmic signs of DMS were noted at presentation in 64 (84%) patients, with 23 (31%) displaying 5 or more signs (Fig. 5). The most common combination of 3 signs was reduced upgaze, convergence retraction, and vertical misalignment. Only 3 patients had just one sign of DMS—upgaze reduction. Relationship of Vertical Gaze Deficits to Pupil Abnormalities The presence of intact downgaze did not increase the likelihood of preserved pupil constriction to a near target. Presenting Neuro-Ophthalmic Signs in Relation to Cause of Dorsal Midbrain Syndrome The presenting neuro-ophthalmic signs did not differ according to whether the DMS was caused by neoplasm or stroke, except that lid retraction was more prevalent in patients with stroke than in patients with neoplasm (P = 0.02) (Fig. 6). For other causes of DMS, the number of patients was too small to conduct a meaningful comparison (Table 1). Persistence of Neuro-Ophthalmic Signs Of the 35 patients with neoplasms, 19 completed at least 1 follow-up visit (median interval 10 months). The number and severity of neuro-ophthalmic signs did not change, e650 FIG. 5. Number of neuro-ophthalmic signs at presentation. except that there was a nonsignificant trend toward improvement in convergence retraction (Fig. 7). There were 12 patients with stroke who completed at least 1 follow-up visit (median interval 3 months). The prevalence of lid retraction significantly declined (P = 0.03). There were nonsignificant trends in reduced severity of upgaze reduction and in the number of neuro-ophthalmic signs per person (P = 0.07 and 0.143, respectively). The slight decline over the course of time in the prevalence of several neuro-ophthalmic signs was not significant (Fig. 8). Accompanying Neurologic Signs Ataxia was the most common documented accompanying neurologic sign, present in 21 (28%) patients on the initial visit. Data on the severity and whether the ataxia impaired speech, limb function, or gait were sparse. Of the 17 patients who had ataxia at presentation and had at least one follow-up visit, 11 (65%) still had ataxia at the last followup visit, but its severity was not documented. The causes of DMS in these 11 patients were stroke (5), nonneoplastic masses (3), neoplasm (2), and nonneoplastic hydrocephalus (1). Data on other accompanying neurologic signs had been minimally documented. Patient Self-Reported Symptoms Among the 11 interviewed patients with neoplasms, 10 reported lingering balance or coordination problems that interfered with ambulation, showering, and putting on clothing, but 8 of them stated that these problems had improved over the course of time. Six patients reported that they still had diplopia but that it interfered less with quality of life than did balance or coordination. Regarding the diplopia, they reported that they could distinguish the real from the false image and could often dismiss the false image. Among them were 3 patients with comitant exotropia varying between 2 and 25 PD (median 8 PD), and 3 patients with vertical misalignments of 1, 2, and 4 PD. Ten patients who had documented reduction of upgaze at the initial visit still endorsed difficulty looking up, but all stressed that it was not an impediment to quality of life because they were able to assume a slight chin-up position to allow acceptable viewing. Among them were 6 patients who had documented Grade 3 upgaze reduction and 4 patients who had Grade 1 or Grade 2 upgaze reduction. Yousif et al: J Neuro-Ophthalmol 2021; 41: e644-e654 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 6. Presenting neuro-ophthalmic signs in patients with neoplasms (35) and strokes (19). The “*” indicates statistical significance (P , 0.05). Among the 8 interviewed patients with stroke, 6 stated that balance or coordination difficulties had initially impaired writing, eating, and driving but that this deficit had lessened over the course of time. Five of them reported that heart or kidney disease, short-term memory, or “decreased peripheral vision” were more important factors in reducing quality of life than the neurologic symptoms. Three patients stated that diplopia contributed to a reduced quality of life. Notably, 5 patients who did not report lingering diplopia had large (up to 60 PD) horizontal misalignments and small vertical misalignments (1–2 PD). Five patients reported lingering difficulty looking up but that it did not impair quality of life despite a Grade 3 upgaze deficit documented at the latest visit. DISCUSSION Regarding the relative prevalence of causes of DMS, this study of 75 patients differs in some respects from the benchmark study of 206 patients by Keane (3) and conforms more to the study of 40 patients by Shields et al (4). Keane (3) reported hydrocephalic cysticercosis as a major cause of DMS, an artifact created by a large population of Mexican immigrants. By contrast, our study found that nonneoplastic hydrocephalus accounted for only 5 (7%) cases, and Shields et al (4) did not list hydrocephalus as a cause. If cysticercosis is not in the mix, nonneoplastic hydrocephalus is apparently an uncommon cause of DMS, although the “setting sun” sign of tonic downgaze deviation is widely attributed to obstructive hydrocephalus and shunt malfunction (13,14). Notably, among the 5 patients in our series who had tonic downgaze deviation, the cause was never nonneoplastic hydrocephalus. We cannot exclude the possibility that hydrocephalus was a contributing factor to DMS in our patients with midbrain region tumors, but in the 14 cases in which the hydrocephalus disappeared after tumor treatment, the DMS features TABLE 1. Presenting neuro-ophthalmic signs Neuro-Ophthalmic Sign Eye movement abnormalities Reduced upgaze Convergence retraction Reduced downgaze Eye misalignment Horizontal Vertical Torsional Pupil abnormalities Impaired pupil constriction to light and near target (“pupil hyporeflexia”) Impaired pupil constriction to light but preserved constriction to a near target (“light-near dissociation”) Pathologic anisocoria Lid retraction Nystagmus Optic neuropathy Papilledema Yousif et al: J Neuro-Ophthalmol 2021; 41: e644-e654 No. of Patients Who Manifested That Sign (Percent of Total Cohort), n (%) 72 70 39 18 56 37 35 6 35 7 (96) (93) (52) (24) (75) (49) (47) (8) (47) (9) 28 (37) 11 20 15 2 4 (15) (27) (20) (3) (5) e651 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 7. Persistence of neuro-ophthalmic signs in 20 patients with neoplasms who had at least one follow-up visit. Median follow-up = 10 months. persisted. Moreover, Pollak et al (7) did not find a correlation between the degree of hydrocephalus and the number of clinical signs of DMS. Perhaps mild versions of DMS that do not include tonic downgaze deviation are being overlooked in nonneoplastic hydrocephalus, given that many patients are children, who might not report a deficit in upgaze or diplopia from ocular misalignment. Among the 35 tumors that made up 47% of cases in our study, many were germ cell pineal neoplasms that vanished with treatment, although the DMS signs remained largely unchanged, as noted in earlier reports of pineal tumors (8– 12). One-third of the neoplasms in our study met imaging criteria for low-grade dorsal midbrain (“tectal”) gliomas. They remained clinically and radiologically stable over years of follow-up. With a median age of 19 years, these patients were considerably younger than the patients with DMS caused by stroke. Stroke made up 25% of our cases, occurring in patients with a median age of 63 years and evenly divided between infarction and hemorrhage. As suggested by Keane (3), but not verified by imaging in that report, most strokes in our study were not located in the midbrain itself but in the thalamus. They were always in the medial thalamus and mostly unilateral. Imaging extension into the midbrain was evident in only a minority of cases. Even highdefinition MRI is apparently not sensitive enough to detect clinical damage to the caudally adjacent pretectum and superior colliculus, where the pathways for upgaze and pupil constriction reside (15). Traumatic brain injury was an infrequent cause of DMS in this study, in part because we decided to exclude cases that lacked a clear imaging correlate. Based on the few cases we included, we were unable to define a mechanism for dorsal midbrain damage in concussive brain injury. This report expands on the profile of neuro-ophthalmic manifestations described by Keane (3) and Shields et al (3,4). In our study, 3 or more neuro-ophthalmic signs were present in more than ¾ of patients, and 5 or more signs were present in 1/3 of patients. Keane (3) did not provide comparable information, and Shields et al (4) reported that only 65% had “the classical triad of vertical gaze palsy, convergence retraction, and light-near dissociation.” Reduced upgaze was the most common sign in our study. It was extremely common in the Keane (3) and Shields et al (4) studies, but those studies did not clearly separate upgaze from downgaze deficiency. Despite the fact that the upgaze deficit was at least 50% in 2/3 of our patients, medical records indicated that a minority of patients reported FIG. 8. Persistence of neuro-ophthalmic signs in 12 patients with stroke who had at least one follow-up visit. The “*” indicates statistical significance (P , 0.05). +Median follow-up time = 3 months. e652 Yousif et al: J Neuro-Ophthalmol 2021; 41: e644-e654 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution difficulty looking up. Thus, this manifestation could go undetected unless searched for by examiners. A downgaze deficit affected 1/4 of our cohort, sometimes severely, but never without an upgaze deficit. Convergence retraction, described in 1/3 of patients by Keane (3) and in 4/5 of patients by Shields et al (4), was present in 1/2 of our cohort, but discomfort on looking up was never documented. Unexpectedly, convergence retraction was not more likely if the upgaze deficit was severe. Eye misalignment, reported in nearly 1/2 of patients in previous studies, was recorded in three-fourth of our patients, evenly split between horizontal and vertical misalignment, as noted by Keane (3). It was often comitant in our study, reflecting a supranuclear disturbance of midbrain convergence (1). Vertical misalignment was attributed equally to skew deviation and fourth nerve palsy, a finding that differs from that of Keane (3), who attributed vertical misalignment exclusively to skew deviation. Shields et al (4) attributed vertical misalignment to skew, fourth nerve palsy, or third nerve palsy. Pupil abnormalities occurred in nearly 1/2 of patients in our study, in more than 1/2 of patients in the Shields et al (4) study, and in nearly all patients in the Keane (3) study. Keane (3) listed “fixed/flicker to light” in 1/2 of his cohort, light-near dissociation in 1/3, and anisocoria of greater than 0.5 mm in 1/4. Shields et al (4) listed light-near dissociation as the only type of pupil abnormality. In our study, light-near dissociation was much more common than impaired pupil constriction to light and a near target. Among the 35 patients in our study with any kind of impaired pupil constriction, only 11 had anisocoria, often less than 1.5 mm. This finding explains why pupil abnormalities in DMS might be overlooked if pupil constriction is not tested. Although we expected to find that patients with light-near dissociation would have dorsally situated tumors that might spare the putative ventral pathway to the Edinger–Westphal nuclei that serves pupil constriction to a near target (16), imaging of our patients did not disclose such a feature. Moreover, preservation of downgaze was not associated with light-near dissociation. Lid retraction, commonly known as Collier sign, was documented in only 1/4 of our cohort, showing that the absence of this most obvious of signs does not exclude DMS. Nystagmus was noted in 1/5 of our patients, with the same prevalence as in the studies of Keane (3) and Shields et al (3,4). The waveform was always jerk and the trajectory usually upbeat in our patients, whereas it was uniformly downbeat in the patients of Shields et al (4). Keane (3) reported upbeat and downbeat nystagmus in equal numbers. Papilledema, reported in 1/4 of the patients in the Keane (3) study, was documented in only 4 (5%) of our patients and not listed in the report of Shields et al (3,4). In our study, papilledema was always associated with severe neoplastic obstructive hydrocephalus at the Sylvian aqueduct. We surmise that the high frequency of papilledema in the Keane (3) Yousif et al: J Neuro-Ophthalmol 2021; 41: e644-e654 study is attributable to the large number of patients with cysticercosis-induced obstructive hydrocephalus. Impaired visual function was a minor feature in our series. Only 2 patients had optic disc pallor, both of whom had impaired optic nerve function. Visual acuity was 20/ 400 bilaterally in one patient, and counting fingers in the right eye and 20/70 in the left eye in the other patient. Given the low frequency of these abnormalities, we presume that patients with DMS are coming to medical attention before those abnormalities develop as the result of chronic papilledema. The neuro-ophthalmic abnormalities in our study did not differ substantially according to the cause of the DMS and did not improve markedly. Among the neoplasms, abnormalities persisted despite imaging regression after treatment, an observation that conforms to previous studies of pineal tumors (8–12). In the patients with stroke, there was only mild lessening of clinical abnormalities over the course of time. These observations are important because caregivers must not only caution their patients that DMS deficits may persist indefinitely but also be aware that some deficits are subtle enough to have been left over from an earlier visit at which they might have been overlooked. From the telephone interviews with 1/3 of our cohort, we learned that neither the diplopia nor the upgaze deficit was a major impediment to quality of life. Instead, patients with tumors were more bothered by lingering ataxia, which interfered with writing, eating, walking, and driving. Patients with stroke were initially bothered most by ataxia but over the course of time it became less of an impediment than nonneurologic medical problems. Our detailed analysis of the imaging features disappointingly failed to disclose a meaningful relationship between the size, location, and signal abnormalities of the lesions and the number and severity of DMS signs. In that regard, our study differs from that of Vuppala et al (6), who found that MRI signal abnormalities within the dorsal midbrain predicted whether DMS signs would be present in patients with pineal tumors. This study, the largest of its kind with imaging confirmation, provides a detailed profile of DMS. Its strengths include accrual by means of a search of electronic medical record text, which is likely to be more complete than one drawn from discharge diagnoses or the files of individual examiners. It includes a rigorous definition of clinical features and a detailed analysis of most of the imaging studies by a neuroradiologist together with follow-up information on a substantial portion of the patients with neoplasms and strokes. Its weaknesses are those of a retrospective study, which is dependent on documentation that was not part of standard rules of data collection. REFERENCES 1. Leigh JR, Zee DS. The Neurology of Eye Movements. Oxford, UK: Oxford University Press, 2015:881–886. 2. Miller NR, Newman NJ, Walsh FB, Hoyt WF. Walsh and Hoyt’s Clinical Neuro-Ophthalmology. 5th edition. Philadelphia, PA: Lippincott Williams & Wilkins, 2005. e653 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution 3. Keane JR. The pretectal syndrome: 206 patients. Neurology. 1990;40:684–690. 4. Shields M, Sinkar S, Chan W, Crompton J. Parinaud syndrome: a 25-year (1991–2016) review of 40 consecutive adult cases. Acta Ophthalmol. 2017;95:792–793. 5. Hura N, Vuppala AAD, Sahraian S, Beheshtian E, Miller NR, Yousem DM. Magnetic resonance imaging findings in Parinaud’s syndrome: comparing pineal mass findings to other etiologies. Clin Imaging. 2019;58:170–176. 6. Vuppala AAD, Hura N, Sahraian S, Beheshtian E, Miller NR, Yousem DM. MRI findings in Parinaud’s syndrome: a closer look at pineal masses. Neuroradiology. 2019;61:507–514. 7. Pollak L, Zehavi-Dorin T, Ana E, Milo R, Huna-Baron R. Parinaud syndrome: any clinicoradiological correlation? Acta Neurol Scand. 2017;136:721–726. 8. Goldenberg-Cohen N, Haber J, Ron Y, Kornreich L, Toledano H, Snir M, Cohen IJ, Michowiz S. Long-term ophthalmological follow-up of children with Parinaud syndrome. Ophthalmic Surg Lasers Imaging. 2010;41:467–471. 9. Dinc C, Iplikcioglu AC, Ozek E. Pineal epidermoid tumors: report of five cases. Turk Neurosurg. 2013;23:446–450. e654 10. Hoehn ME, Calderwood J, O’Donnell T, Armstrong GT, Gajjar A. Children with dorsal midbrain syndrome as a result of pineal tumors. J AAPOS. 2017;21:34–38. 11. Hart MG, Sarkies NJ, Santarius T, Kirollos RW. Ophthalmological outcome after resection of tumors based on the pineal gland. J Neurosurg. 2013;119:420–426. 12. Hankinson EV, Lyons CJ, Hukin J, Cochrane DD. Ophthalmological outcomes of patients treated for pineal region tumors. J Neurosurg Pediatr. 2016;17:558–563. 13. Chattha AS, Delong GR. Sylvian aqueduct syndrome as a sign of acute obstructive hydrocephalus in children. J Neurol Neurosurg Psychiatry. 1975;38:288–296. 14. Tzekov C, Cherninkova S, Gudeva T. Neuroophthalmological symptoms in children treated for internal hydrocephalus. J Neurosurg Pediatr. 1991;17:317–320. 15. Horn AK, Leigh RJ. The anatomy and physiology of the ocular motor system. Handb Clin Neurol. 2011;102:21–69. 16. Thompson HS, Miller NR. Disorders of pupillary function, accommodation and lacrimation. In: Miller NR, Newman NJ, eds. Walsh and Hoyt’s Clinical Neuro-Ophthalmology. Baltimore, MD: Lippincott Williams & Wilkins, 1998:961–1040. Yousif et al: J Neuro-Ophthalmol 2021; 41: e644-e654 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited.