ORIGINAL CONTRIBUTION

Papilledema in the Assessment of Ventriculomegaly

Iris Ben-Bassat Mizrachi, MD, Jonathan D. Trobe, MD, Stephen S. Gebarski, MD, and Hugh J. L. Garton, MD

Abstract: To determine whether ventriculomegaly ("compensated"), particularly if imaging results are stable is associated with ongoing increased intracranial over time. pressure (ICP), physicians often rely on corroborative We describe 2 patients who presented with headache imaging features such as altered periependymal and alterations in mental status. Family members reported signal, distortion of ventricular shape, subarachnoid that the mental status change was recent. Ventriculomegaly space flattening, and an increase in ventricular size was present on imaging, but elevated ICP was discounted over time. In 2 patients with new headache and for lack of supportive imaging findings. Influenced by altered mental status, symptoms and ventriculome­ the lack of imaging support for increased ICP, caregivers galy were dismissed as long-standing and not reflec­ attributed the mental status changes in the first patient tive of current ICP elevation. In the first patient, ICP to congenital mental impairment and in the second patient was considered normal because there were no cor­ to a known bipolar disorder. When papilledema was later roborative imaging features of elevated ICP. In the discovered, ICP monitoring disclosed markedly elevated second patient, ICP was considered normal because ICP. Ventriculoperitoneal shunting resolved the papilled­ ventricular size was stable over a 1-year period. The ema and markedly improved mentation in both patients. diagnosis of ICP elevation was finally made by ICP These patients are presented to highlight the limitations monitoring after papilledema was recognized. Ven- of brain imaging alone in the diagnosis of elevated ICP and triculoperitoneal shunting rapidly resolved the pap­ to underscore the importance of ophthalmoscopy. illedema and markedly improved mentation. Brain imaging may often be an unreliable guide to the pres­ ence of elevated ICP In such patients, the finding of CASE REPORTS papilledema is a critical determinant of management. Case 1 (/ Neuro-Ophthalmol 2006;26:260-263) A 41-year-old woman complained of persistent headache after a flu-like illness. Family members reported that her thinking had declined recently. She had a shunt n the presence of ventriculomegaly, corroborative imag­ placed at birth for infantile hydrocephalus associated with ing features such as altered periependymal signal, dis­ a Dandy-Walker malformation, but the shunt had been tortion of ventricular shape, subarachnoid space (SAS) removed at age 1 because of malfunction. After shunt re­ flattening, and an increase in ventricular size over time are moval, she had had no neurologic problems except mild often used as determinants of elevated intracranial pressure developmental cognitive impairment. (ICP) (1,2). Although headache, vomiting, nausea, lethargy, Other than altered cognition, the emergency room and double vision are expected if ICP is high, these neurologic examination was considered normal. Brain CT symptoms may be absent, particularly if ICP rises slowly. scanning showed marked dilatation of the lateral and third In such patients, ventriculomegaly is apt to be dismissed ventricles with slight effacement of the brain sulci (Fig. as long-standing and unrelated to elevated ICP 1A). Older neuroimaging results from another institution were not available, so it was not possible to evaluate potential changes in these imaging abnormalities. The ven­ Goldschleger Eye Institute, Chaim Sheba Medical Center, Tel Hashomer, Israel (IBM); and Departments of Ophthalmology (JDT), triculomegaly was judged to be chronic and unreflective of Neurology (JDT), Radiology (SSG), and Neurosurgery (HJLG), University a current elevation in ICP of Michigan, Ann Arbor, Michigan. Six weeks later, ophthalmologic examination re­ Supported in part by a fellowship from the American Physician vealed a visual acuity of 20/20 in the right eye and 20/70 in Fellowship for Medicine in Israel. Address correspondence to Iris Ben-Bassat Mizrachi, MD, Gols- the left eye; the subnormal visual acuity in the left eye was chleger Eye Institute, Sheba Medical Center, Tel Hashomer, Israel 52621; attributed to known amblyopia. She had normal pupillary E-mail: [email protected] size and reactions, full extraocular movements with jerk

260 J Neuro-Ophthalmol, Vol. 26, No. 4, 2006 Papilledema in Ventriculomegaly J Neuro-Ophthalmol, Vol. 26, No. 4, 2006

FIG. 1. Case 1. A. At clinical presentation, a precontrast axial CT scan shows a Dandy-Walker malformation, marked ventriculomegaly, and poor sulcal definition. B. Mild optic disc edema (papilledema) is discovered 6 weeks later. C. Six weeks after ventriculoperitoneal shunting, the CTscan shows better sulcal definition (arrows) but minimal reduction in ventricular size. D. Six weeks after shunting, papilledema has resolved. upbeat nystagmus in upgaze and sidebeat nystagmus with cavernomas attributed to Osler-Weber-Rendu disease in horizontal gaze bilaterally, findings considered typical (Fig. 2A). Comparison with an MRI performed 1 year of congenital nystagmus. On ophthalmoscopy, mild earlier showed no change. Therefore, the ventriculomegaly optic disc edema was found in both eyes (Fig. IB). When was diagnosed as chronic and unreflective of current confronted, she denied having had visual symptoms or ICP elevation. pulsatile tinnitus. Over the ensuing months, the patient's mental state The patient underwent ICP monitoring that revealed declined to the point that she became incoherent. She was pressures ranging from 300 to 700 mm H20. At the time of fired from her job as a hotel clerk. Six months after her ventriculoperitoneal shunt placement, CSF opening pres­ initial evaluation, ophthalmologic examination disclosed sure was estimated to be 900 mm H20. Six weeks after marked optic disc edema (Fig. 2B) leading immediately to shunting, her mental status had returned to baseline, and inpatient ICP monitoring that demonstrated mean pressures CT scanning showed reduced ventricular size and better of 300 mm H20 spiking to 500 mm H20. She denied ever sulcal definition consistent with improved cerebrospinal having visual symptoms or pulsatile tinnitus. fluid (CSF) dynamics (Fig. 1C); the papilledema had re­ She promptly underwent ventriculoperitonal shunting. solved (Fig. ID). One month later, her flight of ideas had resolved, neurologic examination was normal, MRI showed a decrease in ventricular size (Fig. 2C), and papilledema had disappeared Case 2 (Fig 2D). She was rehired by the hotel and has maintained A 54-year-old woman complained of imbalance and her job without difficulty ever since. a reduced ability to concentrate for several weeks. When her neighbors witnessed a fall in the street, she was taken to a hospital emergency room. She had been diagnosed in the past with bipolar disorder, diabetes mellitus, systemic DISCUSSION lupus erythematosus, and Osler-Weber-Rendu disease. In our two patients, the diagnosis of elevated ICP was Neurologic examination showed poor attention, flight of delayed because ventriculomegaly was considered chronic ideas, and a wide-based gait. Brain MRI showed dilatation and unreflective of ongoing ICP elevation. In the first of the lateral and third ventricles and periependymal high patient, there were no other imaging features to corroborate T2 signal, as well as numerous mass lesions consistent elevated ICP. In the second patient, imaging results were

261 J Neuro-Ophthalmol, Vol. 26, No. 4, 2006 Mizrachi et al

FIG. 2. Case 2. A. At clinical presentation, axial FLAIR MRI shows dilation of the lateral and third ventricles with apical high signal {large arrows); multiple cavernomas, part of Osler-Weber-Rendu disease, are incidental findings {small arrows). B. Marked optic disc edema (papilledema) is discovered 6 months later. C. Four weeks after ventriculoperitoneal shunting, MRI shows reductions in ventriculomegaly and periependymal high signal. D. Four weeks after shunting, papilledema has resolved. stable over 1 year. In both patients, the diagnosis of elevated brain will accommodate an increased volume of CSF ICP was entertained only when papilledema was later without producing significantly increased ICP in normal detected. pressure hydrocephalus. Fourth, altered CSF dynamics in Passing off ventriculomegaly as "compensated" is a relatively non-compliant brain may result in ventriculo­ understandable. Ventriculomegaly is, after all, a common megaly, sometimes with no other imaging abnormalities imaging finding in patients who do not have elevated ICP of high ICP yet prominent symptoms. The relationship between ventricular size, intraventricular To decide whether ICP is high, heavy reliance is often pressure, and ICP is poorly understood. When ventriculo­ placed on neuroimaging signs such as periependymal megaly is long-standing, brain viscoelastic properties and low attenuation signal on CT and high signal on T2 MRI. the pressure-volume relationship may be altered such that When CSF is under pressure, the ependymal lining of ventricles remain expanded even when ICP is normal (3,4). the ventricles stretches. As a result of this stretching and The relationship between ventricular size and ICP is typi­ a ventriculofugal pressure gradient, CSF seeps into the cally described in terms of the intracranial compliance adjacent white matter, causing the characteristic periepen­ or elastance (change in pressure for a given change in dymal signal abnormalities (3,4). The ventricular dilatation volume). In response to an increasing volume of CSF, also causes a decrease in size of the subarachnoid space. patients with increased intracranial compliance typically In our Case 1, compromise of the SAS was present but demonstrate normal ICP, whereas those with normal or dismissed. (It became very evident when the pre-shunt brain decreased compliance develop increased ICP. CT scan was compared to the post-shunt brain CT scan.) In Ventriculomegaly may result from a number of con­ our Case 2, periependymal high T2 signal was present but ditions (1,2). First, there may be a loss of brain substance was considered part of a chronic "compensated" state as occurs after trauma. Ventriculomegaly then represents because it had been present on a scan 1 year earlier. "ex vacuo" dilatation, not increased ICP. Second, despite Most studies that relate ventricular size to ICP have the presence of a normally functioning CSF diversion focused on shunt-dependent patients. In a study of 100 device and normal ICP, the ventricles may remain shunt-dependent patients reviewed by radiologists for a asymptomatically enlarged because the interstitial changes question of shunt malfunction (5), 33% were false nega­ induced by prior CSF obstruction that was not resolved tively interpreted as indicating no evidence of shunt failure. ("arrested hydrocephalus"). Third, a highly compliant The basis for this misinterpretation was not identified.

262 © 2006 Lippincott Williams & Wilkins Papilledema in Ventriculomegaly J Neuro-Ophthalmol, Vol. 26, No. 4, 2006

The authors emphasized that radiologists should use 2. Bradley WG Jr. Hydrocephalus. In: Bradley WG Jr, Bydder G, "enlargement of ventricles over time" as a criterion for eds. MSI Atlas of the Brain. New York: Raven Press; 1990: 265-301. elevated ICP. However, the absence of ventricular enlarge­ 3. Taylor Z, Miller K. Reassessment of brain elasticity for analysis ment over time does not rule out raised ICP (6-8), as of biomechanisms of hydrocephalus. J Biomech 2004;37: exemplified by our Case 2. 1263-9. 4. Kiefer M, Eymann R, Von Tiling S, et al. The ependyma in chronic Our cases demonstrate the complexity of the diag­ hydrocephalus. Childs Nerv Syst 1998;14:263-70. nosis of elevated ICP, particularly in patients whose history 5. Iskandar BJ, McLaughlin C, Mapstone TB, et al. Pitfalls in the is not reliable and in whom the ventricles are not expected diagnosis of ventricular shunt dysfunction: radiology reports and ventricular size. Pediatrics 1998;101:1031-6. to "deflate" even when ICP is normalized. Brain 6. Katz DM, Trobe JD, Muraszko KM, et al. Shunt failure without imaging—even serial brain imaging—may not be enough ventriculomegaly proclaimed by ophthalmic findings. J Neurosurg to decide whether the brain is under pressure. In such 1994;81:721-5. 7. Dahlerup B, Gjerris F, Harmsen A, et al. Severe headache as the only patients, the finding of papilledema becomes a critical symptom of long-standing shunt dysfunction in hydrocephalic determinant of management. children with normal or slit ventricles revealed by computer tomography. Childs Nerv Syst 1985;1:49-52. REFERENCES 8. Engel M, Carmel PW, Chutorian AM. Increased intra­ ventricular pressure without ventriculomegaly in children with 1. Bradley WG Jr. Diagnostic tools in hydrocephalus. Neurosurg Clin N shunts: "normal volume" hydrocephalus. Neurosurgery 1979;5: Am 2001;12:661-84. 549-52.

263