The Need for a West Nile Virus MRI Registry

Hugh J. Robertson; James J. Sejvar

West Nile virus (WNV) infection was first recorded in North America in 1999 in the vicinity of New York City. By December 2002, the infection had been reported in most U.S. states and several Canadian provinces. Each year, from 1999 through 2001, there were fewer than 70 hospitalized cases in the United States, with the mortality rate varying from approximately 9% to 16%. In 2002, however, a large outbreak of WNV infection occurred in the United States, with more than 4,000 serologically confirmed cases and 277 deaths. Substantial numbers of patients with severe neurologic disease were reported in outbreaks of WNV infection in previously known endemic areas (Romania, 1996; Russia, 1999; Israel, 2000).

Sporadic reports of small numbers of WNV cases have noted a variety of MR imaging findings. One early report noted periventricular white matter T2 focal areas of hyperintensity and meningeal contrast enhancement (1). In a recent single-case report, Rosas and Wippold noted bilateral T2 hyperintensities within the basal ganglia and thalami. No hemorrhage was noted within these lesions, and no meningeal abnormality was reported (2).

The WNV is a flavivirus closely related to the viruses causing central nervous system infection in St. Louis, Japanese, Kunjin, and Murray Valley encephalitis. Among these encephalitides, only Japanese encephalitis has been reviewed for imaging findings in case reports with moderately large numbers of patients (3, 4). Hemorrhagic lesions of the cerebral basal ganglia and thalami were frequent MR imaging findings in Japanese encephalitis. St. Louis encephalitis is a recurrent seasonal infection in the southern United States, but MR imaging findings of substantia nigra signal intensity abnormality are noted in only one case report (5). Descriptions of MR imaging findings in some other varieties of encephalitis with seasonal occurrence in the United States are scarce to nonexistent. Einsiedel et al recently reported the imaging findings in a single case of Murray Valley encephalitis with severe neurologic disease involving the brain and spinal cord. High signal intensity lesions on T2 sequences were noted in the thalami, substantia nigra, red nuclei, reticular formation, and the cervical spinal cord. No hemorrhage was reported, but the authors considered the distribution of the lesions to be similar to that of Japanese encephalitis (6).

Many of the patients with WNV encephalitis in the 2002 epidemic in Louisiana were treated at rural or small urban hospitals. CT scanning was sometimes the only available diagnostic imaging study. Magnets of a variety of field strengths were used, when locally available, frequently for only one brain scan early in the course of the disease. Almost all of the CT scans and many of the MR imaging brain scans were considered normal. Nine patients with MR imaging scan abnormalities each had distinctly different findings. A fatal case presented with diffuse high T2 signal intensity in the vermis and cerebellar white matter, putamina, and adjacent white matter but no evident involvement of the thalami. Another patient presented with focal cerebral white matter lesions, some with restricted diffusion, closely resembling multiple sclerosis. Clinical detection of Parkinsonian tremor and sometimes flaccid paresis occurred in some patients with coincident apparently normal CT scans and MR images of the brain and cervical spine. Subsequent follow-up MR imaging in some but not all patients with Parkinsonian features showed T2 high signal intensity abnormalities in the cerebral basal ganglia and thalami. Imaging experience in WNV infection has confirmed what was already known in the study of patients with a variety of encephalitides. CT lacks the sensitivity for detection of some pathologic findings. In a case report at RSNA 2002, Butman noted progressive MR imaging findings in the basal ganglia, thalami, pons, and dentate nuclei in serial MR imaging scans over a period of 5 weeks. The first MR imaging scan obtained soon after the onset of the illness was considered normal (7).

Some observations on the imaging of WNV meningoencephalomyelitis are possible. Fluid-attenuated inversion recovery MR imaging, diffusion-weighted, and T1 post–intravenous contrast sequences are most useful in the detection of disease. A gradient echo sequence may assist in the detection of hemorrhage in lesions. Serial MR imaging scans over an interval of several days to weeks may be necessary to show any abnormality and can document developing and changing scan findings. Hypertension or diabetes in elderly, debilitated WNV patients may be causes of focal cerebral white matter microvascular ischemic changes (leukoaraiosis), which could be similar to WNV white matter disease. A previous MR imaging brain scan could be essential for comparison in correctly assessing scans obtained in such patients with WNV infection. Positron-emission tomography and molecular imaging are potentially useful future diagnostic tools that may extend further the margins of disease detection and possibly provide earlier diagnostic findings in the study of encephalitis.

A WNV MR imaging registry has been established by the Centers for Disease Control and Prevention, in Atlanta, and the Louisiana State University Health Sciences Center, in New Orleans. It is hoped that the registry data will provide comprehensive information on the imaging characteristics of WNV infection. Parallel studies of other varieties of encephalitis may be possible. All scans sent to the registry are rendered anonymous. The original scan annotated data are deleted, and a scan is identifiable only by a randomized number. The scan images are reviewed independently by three experienced neuroradiologists on a fully secure universal picture archiving communication system system. All scan findings are systematically tabulated for statistical analysis. The submission of scans to the WNV registry does not affect the use of such cases by contributors in scientific case reports and publications. The contributors to the registry will be acknowledged in any future publications resulting from the registry data.

Physicians who are aware of patients with laboratory-confirmed WNV infection who have had MR imaging scans are asked to contact the WNV MR imaging registry at westnileunipacs.com.

References

↵
Weiss D, Carr D, Kellachan J, et al. Clinical findings of West Nile virus infection in hospitalized patients, New York and New Jersey, 2000. Emerg Infect Dis 2001;7:654–658
PubMed
↵
Rosas H, Wippold FJ. West Nile virus: case report with MR imaging findings. AJNR Am J Neuroradiol 2003;24:1376–1378
Abstract/FREE Full Text
↵
Kalita J, Misra U. Brainstem auditory evoked potential in Japanese encephalitis. J Neurol Sci 1999;165:24–27
PubMed
↵
Kalita J, Misra U. Comparison of CT scan and MRI findings in the diagnosis of Japanese encephalitis. J Neurol Sci 2000;174:3–8
CrossRef PubMed
↵
Cerna F, Mehrad B, Luby J, et al. St. Louis encephalitis and the substantia nigra: MR imaging evaluation. AJNR Am J Neuroradiol 2000;20:1281–1283
↵
Einsiedel L, Kat E, Ravindran J, et al. MR findings in Murray Valley encephalitis. AJNR Am J Neuroradiol 2003;24:1379–1382
Abstract/FREE Full Text
↵
Butman J. MR imaging helps confirm West Nile virus encephalitis. RSNA News 2003;13:4–5