Enhancing spinal cord lesions detected with T1-weighted magnetic resonance imaging in idiopathic recurrent transverse myelitis at first or second attack. C indicates cervical spinal cord; CST, corticospinal tract; SCT, spinocerebellar tract; STT, spinothalamic tract; and T, thoracic spinal cord.
Enhancing spinal cord lesions detected with T1-weighted magnetic resonance imaging in multiple sclerosis–associated recurrent transverse myelitis at the first or second attack. C indicates cervical spinal cord; T, thoracic spinal cord.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Kim K. Idiopathic Recurrent Transverse Myelitis. Arch Neurol. 2003;60(9):1290–1294. doi:10.1001/archneur.60.9.1290
To determine whether idiopathic recurrent transverse myelitis (RTM) can be distinguished from multiple sclerosis–associated RTM (MSRTM) on the basis of clinical manifestations of myelopathy, or findings from magentic resonance imaging or cerebrospinal fluid examination.
A retrospective analysis of 37 cases was conducted. Patients were classified as having idiopathic RTM on the basis of recurrent myelitis confirmed by clinical manifestations of myelopathy and magnetic resonance imaging findings. On review patients with idiopathic RTM had normal cranial magnetic resonance imagings and did not demonstrate paraclinical evidence of spatial dissemination beyond the spinal cord of the disease process. Patients were classified as having MSRTM on the basis of criteria of Poser et al for clinically definite multiple sclerosis involving the central nervous system. Fifteen patients met study criteria for idiopathic RTM. Twenty-two patients had MSRTM.
Asan Medical Center, Seoul, South Korea, from January 1, 1992, through December 31, 2001.
Main Outcome Measures
Presenting symptoms and clinical manifestations, relapsing times, magnetic resonance imaging features (involved spinal cord segments in T2-weighted images and gadolinium 64–enhanced lesions on T1-weighted images), IgG index, and oligoclonal bands in cerebrospinal fluid were compared.
Idiopathic RTM occurred preponderantly in male patients and presented more often with acute transverse myelitis than did MSRTM. More than 2 relapses occurred in 6 cases (40%) of idiopathic RTM. The involved segments of spinal cord on T2-weighted images were not significantly different in idiopathic RTM and MSRTM, with enhancing lesions mostly in the posterior columns, and the spinothalamic and spinocerebellar tracts of white matter. Additionally, almost all patients with idiopathic RTM had normal cerebrospinal fluid indexes.
Idiopathic RTM might be a disease entity distinct from MSRTM, differing in its male preponderance, absence of oligoclonal bands, frequent multiple relapses, and frequent presentation as acute transverse myelitis.
RECURRENT TRANSVERSE myelitis (RTM) usually heralds or is accompanied by lesions producing dysfunction elsewhere in the central nervous system and almost always signifies the presence of multiple sclerosis (MS).1-4 A series of 3 cases of RTM in patients who never developed other evidence of MS or collagen vascular disease and had no abnormalities on magnetic resonance imagings (MRIs) of the brain suggested that RTM may be a disorder distinct from MS.5 Recurrent transverse myelitis as an independent entity, with neither abnormal cranial MRI nor oligoclonal bands (OCBs) in cerebrospinal fluid (CSF), is an interesting topic in terms of the causative factors and therapeutic considerations. Further studies in which the criteria for an idiopathic RTM of unknown origin are clearly defined and that include long-term follow-up are needed to distinguish idiopathic RTM from MS-associated RTM (MSRTM). In this study, the medical records of 15 patients with idiopathic RTM who were followed up for longer than 5 years after their first attack of myelitis were reviewed for clinical manifestations, specific spinal MRI findings, and CSF study results to verify a novel disease entity distinguished from clinically definite myelopathic MS.
The medical records of 15 consecutive patients (12 men and 3 women) with idiopathic RTM were reviewed. Diagnostic criteria for idiopathic RTM consisted of 2 or more attacks at different neuroanatomically distinct spinal cord levels without association with a febrile illness, each lasting over 24 hours and confirmed by both high signal intensity of the T2-weighted image (T2WI) at the responsible spinal cord level and a normal MRI of the brain within 7 days of an episode.5-7 Patients were excluded if there was a history of recent carcinoma, abdominal surgery radiation therapy, epidural spinal block, arteriovenous malformation of the spinal cord, paraneoplastic spinal cord syndrome with or without paraneoplastic antibodies (anti-Hu, anti–collapsin response-mediated protein 58), or clinically definite myelopathic MS. Patients with systemic lupus erythematosus or Sjögren syndrome were also excluded. If the results of an antinuclear antibody test were positive, anti–double-strain DNA, antisoluable nuclear antigen, anti-Ro antibodies, anti-SS/A antibodies, and rheumatoid factor tests were performed for other collagenases. Serum antibody tests for human immunodeficiency virus and hepatitis B or A, VDRL, polymerase chain reaction of tuberculosis, and spinal tap in the first 24 hours were performed in all patients. All 15 patients had MRIs of the spinal cord and brain during each attack. The mean age for patients with idiopathic RTM was 43 years.
Patients in the MSRTM group consisted of 22 consecutive patients (6 men and 16 women) with onset of MS symptoms referable to the spinal cord, optic nerve, or brainstem, all of whom eventually had RTM and fulfilled criteria by Poser et al9 for clinically definite MS. The mean age for patients with MSRTM was 41 years. All patients underwent MRI of the brain and spinal cord. Patients with idiopathic RTM or MSRTM underwent CSF studies for OCBs and IgG indexes within 10 days after the first symptoms of myelopathy. All patients received high doses of intravenous corticosteroids after the CSF tests.
Magnetic resonance imaging of the spine was performed with a minimum of relative T1-weighted imaging (T1WI), enhanced T1WI, and T2WI sagittal 3-mm cuts and 1-mm gaps using either a 1.5-magneton unit (Siemens Corp, Berlin, Germany) or a 1.5-T unit (GE Global Research, Milwaukee, Wis). Magnetic resonance images of the brain were obtained with T1WIs and T2WIs axially with 5-mm cuts and 5-mm gaps. Records were also reviewed for the results of CSF IgG indexes and OCBs. Visual evoked potentials (VEPs), brainstem auditory evoked potentials, and somatosensory evoked potentials were performed in 13 patients from the idiopathic RTM group and in 15 patients from the MSRTM group. The mean follow-up time for the idiopathic RTM group was 5.9 years and for the MSRTM group was 7.0 years.
Acute transverse myelitis (ATM) was defined as a spinal cord syndrome characterized by paraplegia with or without sensory symptoms and bladder dysfunction that typically manifests itself over a period of hours to 1 week.5,6,10,11 Magnetic resonance imaging and clinical information on all patients are given in Table 1. Fisher exact test of statistical procedures for assessment of data was selected.
Fifteen cases of idiopathic RTM and 22 cases of MSRTM were reviewed. Twelve men and 3 women composed the idiopathic RTM group that had a mean age of 43 years. None of these patients had upper respiratory tract infection or gastroenteritis precede their myelitis. Eight (53%) of the 15 patients with idiopathic RTM had paresthesias and numbness, 5 (33%) had weakness, and 2 (14%) had pain. Six patients (40%) had symmetric sensory and motor dysfunction of their upper or lower extremities and bladder dysfunction, thereby meeting the criteria for ATM during their first or second attack.
Six men and 16 women composed the MSRTM group that had a mean age of 41 years. Patients in the MSRTM group had similar initial symptoms, with 14 (64%) of the 22 patients having paresthesia and numbness; 8 (36%) of the 22, weakness. Only 1 patient with MSRTM had ATM during her second attack. The mean interval between recurrent spinal cord syndromes was 13 months in the idiopathic RTM group and 15 months in the MSRTM group. Relapsing myelitis was observed from 2 to 4 times in the idiopathic RTM group and from 2 to 3 times in the MSRTM group.
In the idiopathic RTM group 5 patients had upper cervical cord lesions visible on T2WI, 1 of whom had recurrent upper cervical myelitis. Ten patients had recurrent thoracic myelitis. The mean number of segments involved on T2WI was 3.02. Two patients had focal involvement of the spinal cord (no more than 1 segment was involved) during the first or second attack (Table 1). In 7 (47%) of the 15 patients with idiopathic RTM, gadolinium 64–enhanced T1-weighted lesions were noted. The lesions were located in the posterior columns, spinothalamic tracts, and spinocerebellar tracts in order of decreasing frequency (Figure 1).
In the MSRTM group 13 (59%) of the 22 patients had high signal T2-weighted lesions in the upper cervical cord. Eight patients (36%) had recurrent thoracic lesions only. Six patients (27%) had focal involvement of the spinal cord during the first or second attack. The average number of segments involved on T2WI was 3.36. In 10 patients (45%) gadolinium 64–enhanced T1-weighted lesions were noted. These lesions were located in the posterior columns, spinothalamic tracts, and spinocerebellar tracts in order of decreasing frequency. In gadolinium 64–enhanced T1WIs 10 patients had 1 or more lesions in myelinated tracts in the cervical or thoracic spinal cord (Figure 2).
No more than 10 white blood cells/µL of CSF were found in all patients in this study. Protein and other CSF contents of all patients were within the normal range. High IgG indexes (>0.7) were found in 3 patients with idiopathic RTM and in 9 patients with MSRTM. Oligoclonal bands were absent in the 15 patients with idiopathic RTM but were present in 7 patients (32%) with MSRTM. The VEPs and brainstem auditory evoked potentials were normal in 13 (87%) of the 22 patients examined in the idiopathic RTM group. The VEPs were abnormal in 9 (mostly opticospinal form) (60%) of the 15 patients examined in the MSRTM group.
In clinical manifestations, idiopathic RTM with no paraclinical evidence of disseminated lesions and a normal brain MRI had presented as ATM during the first or second attack in 6 (40%) of 15 patients. Miller et al12 described 4 patients younger than 50 years said to have chronic relapsing isolated noncompressive partial myelopathy. They had neither OCBs nor lesions on brain MRI like patients with idiopathic RTM. Tippet et al5 described 3 patients with recurrent ATM who never developed other evidence of MS or partial myelopathic symptoms. These cases suggest that recurrent myelitis may occur separately from ATM or partial transverse myelitis.
Patients with idiopathic RTM were preponderantly male, in contrast to MSRTM (P<.01, Fisher exact test). To my knowledge, the male preponderance has not previously been reported. Are there androgen effects on this disease or is there an X-linked genetic control defect on immunologic disease?
There is no difference in the location of spinal cord lesions on T2WI and the frequency of sites with lesions between the idiopathic RTM and MSRTM groups. The lesions in the middle or upper cervical cord regions were more frequently involved in MS studies.6,13,14 However, to my knowledge, there have been no reports that a larger proportion of patients with MS have lesions of transverse myelitis in the thoracic cord. Duanne et al10 and al Deeb et al15 reported that the lesion of ATM more frequently occurs at the thoracic spinal cord than cervical cords. In this study, the thoracic lesion is more frequently involved in idiopathic RTM and MSRTM than the upper or middle cervical cord, suggesting that there is not a prevalent lesion of cord for idiopathic RTM or MSRTM. It is also suggested that the larger area of the thoracic spinal cord (T1 through T12 segments) has more chance to contact precipitating inflammatory cells than that of the cervical cord (C1 through C7). Among the few cases of relapsing acute myelopathy without clinical lesions above the spinal cord, Nilsson et al16 described a patient with 6 attacks of spinal cord symptoms and a cranial MRI that showed no abnormality; however, the VEPs were positive and there were OCBs in the CSF. Despite the increased IgG index in 3 patients with idiopathic RTM, normal VEPs and brainstem auditory evoked potentials suggest no dissemination lesion beyond the spinal cord.
In 1 of the 6 cases of idiopathic myelitis reported by Jeffery et al,7 there was recurrence. As in other reports5,7 the number of relapses is variable in this case. This variable relapsing rate may be caused by misinterpretation of pseudorelapse, by differences in follow-up duration, or by different precipitating factors.
There is no difference in the enhancement patterns on T1WI between idiopathic RTM and MSRTM. Eccentric or peripheral enhancement was detected mostly in myelinated spinal cord white matter. As in the patients of other studies6 characteristic patterns of myelopathic MS lesions involving discrete multiple segments or oval enhancement were not detected in idiopathic RTM. Few focal spinal lesions were detected in idiopathic RTM, as in other cases of MSRTM.6
Although decreased positivity of OCBs (33%) was reported in the Asian form of MS (opticospinal form),17,18 the absence of OCBs in the CSF of all 15 patients with idiopathic RTM clearly is different from MSRTM (P<.001, Fisher exact test) where 7 (32%) of 22 patients had OCBs. This suggests that there may not be an abnormality in IgG production in the central nervous system of patients with idiopathic RTM.19
This study suggests that idiopathic RTM might be a distinct disease entity from MSRTM, with male preponderance, absent OCBs, frequent multiple relapses, and frequently presenting with ATM. Is this an autoimmune disease against the spinal cord white matter? Might autoantigen-mimicking virus or superantigen or immune modulating cytokines provoke sensitized T-helper cells or T-memory cells to spinal cord? Is this a different HLA-Ag of idiopathic RTM from that of the MSRTM? To firmly establish idiopathic RTM as a novel disease entity, more cases must be gathered using strict criteria for idiopathic RTM and longer follow-up periods (>5 years after the second episode of myelitis), and study for HLA-Ag is also needed for differentiating the opticospinal form of MS from idiopathic RTM.
Corresponding author and reprints: Kwang-kuk Kim, MD, PhD, Department of Neurology, Univeristy of Ulsan College of Medicine, Asan Medical Center, 388-1 Pungnap-dong, Songpa-gu, Seoul 138-736, South Korea (e-mail: firstname.lastname@example.org).
Accepted for publication March 6, 2003.