Neurol India Home 

Year : 1999  |  Volume : 47  |  Issue : 4  |  Page : 253--4

Role of MRI in acute transverse myelitis.

UK Misra, J Kalita 

Correspondence Address:
U K Misra

How to cite this article:
Misra U K, Kalita J. Role of MRI in acute transverse myelitis. Neurol India 1999;47:253-4

How to cite this URL:
Misra U K, Kalita J. Role of MRI in acute transverse myelitis. Neurol India [serial online] 1999 [cited 2021 Feb 26 ];47:253-4
Available from:

Full Text

Myelopathy is clinically classified as compressive and noncompressive. This distinction is based on demonstration or ruling out a compression. The method has paralleled the technological advances; myelography, CT scanning and now MR imaging. In this issue of Neurology India there are two papers addressing noncompressive myelopathy[1] and radiological changes in acute transverse myelitis(ATM).[2] The commonest cause of noncompressive myelopathy is ATM. Both these papers highlight importance of MRI in the diagnosis of ATM.

The diagnostic criteria of ATM proposed by Dawson and Potts in 1991 include `an acute or subacute spinal cord dysfunction characterised by paraplegia, horizontal level of sensory impairment and sphincter dysfunction in which the secondary causes such as compressive lesions, tuberculosis, syphilis, AV malformation, trauma and malignant infiltration have been excluded'.[3] It is important to exclude the secondary causes as these can mimic the syndrome of ATM. Both these studies have reported high frequency of MRI abnormality; 27 out of 31 in the study by Prabhakar et al[1] and 12 out of 13 by Murthy et al.[2] These two studies have confirmed the findings of Choi et al[4] which included (1) hyperintense spinal cord signal changes in T2 weighted images, extending over 3-4 vertebral segments and (2) hyperintense lesion located in central 2/3rd of spinal cord with a central dot.

The MR changes in ATM depend on the underlying pathology which includes oedema, demyelination, necrosis haemorrhage and myelomalacia. The changes not only vary with the severity and rapidity of disease progression but are also dependent on timing of MRI study, strength of magnetic coil and extent of MRI examination i.e. if only a clinically relevant area or whole of spinal cord is imaged. In the paper by Choi et al the patient population was not homogenous and many patients did not fulfill the clinical criteria of ATM;[5] and they restricted the MRI study to clinically relevant level which may be responsible for relatively shorter extent of signal changes in 3-4 vertebral segments, whereas in our study there were more extensive signal changes. The mean level of signal changes was 10 spinal segment above the sensory level.[6] ATM is commonly regarded to involve the thoracic spinal cord and affects the cervical region in 10% patients only. We reported cervical spinal cord involvement in 7 out of 10 patients on clinical, neurophysiological and MRI study.[6] Murthy et al also found high frequency of cervical spinal cord involvement (9 out of 13 patients), based only on MRI findings. Autopsy studies in ATM have also confirmed that the histologic changes of oedema and demelination extend 12-16 segments above the sensory level although the necrotic changes are restricted to thoracic or lumbar region.[7] We feel that the descrepancy between different MRI studies regarding the extent of signal changes may be due to timing of MRI study, strength of magnet, heterogeneity of study population and restricted area of MRI scanning.

In the literature, there is emphasis on differentiating ATM from the first attack of multiple sclerosis (MS) by MRI features. In MS the spinal cord involvement is less than 2/3rd in axial section, vertical extent is also less than 2 segment and there is central homogeneous enhancement.[2],[4] Jafferay et al highlighted the importance of cord swelling in parainfectious myelitis and its absence in MS.[8] Even the group of parainfectious ATM is not homogenous because it can follow an array of infections and vaccinations. The clinical and MRI findings in different subsets of parainfectious ATM have not been well documented although isolated report of MRI findings in mumps have revealed extensive signal changes with good outcome and disseminated HSV1 myelopathy in AIDS patients with fatal outcome and restricted spinal cord involvement by Herpes zoster.[1],[9],[10] Extent of MRI changes has been reported as a poor prognostic predictor of ATM.[6],[11] However in a multivariate analysis extent of MRI signal changes did not [figure] as an important predictor of outcome. On the contrary severity of weakness and evidences of denervation on EMG are most useful predictors of outcome. The EMG changes are apparant only after 2-3 weeks, therefore, in the early stage unrecordable MEP, especially on spinal stimulation may help in predicting the outcome.[10]

In a patient with non compressive myelopathy in tropical countries other conditions such as nutritional, HIV, HTLV1, lathyrism and Konzo etc should also be considered in defferential diagnosis because many of these conditions may result in somewhat similar MRI findings as in ATM. These conditions however have distinct epidemiological and clinical clues especially lack of horizontal sensory loss. Tropical neurophysicians should make every effort to differentiate ATM not only from multiple sclerosis but also from other infections such as tuberculosis, cysticercosis, schistosomiasis etc, as these require specific therapy.


1Prabhakar S, Syal P, Paramjeet S et al : Noncompressive myelopathy clinical and radiological study. Neurol India 1999; 4 : 294-299.
2Murthy JMK, Reddy JJ, Meena AK et al : Acute transverse myelitis : MR characteristics. Neurol India 1999; 4 : 290-293.
3Dawson DM, Potts FM : Acute nontraumatic myelopathies. In : Neurology Clinic : Disorders of spinal cord, Wilsey RM, Young RR Eds. WB Saunders Philadelphia. 1991; 551-603.
4Choi KH, Lee KS, Chung SO et al : Idiopathic transverse myelitis : MR characteristics. Arn J Neuroradiol 1996; 17 : 1151-1160.
5Misra UK, Kalita J : MR changes in transverse myelitis. Am J Neuroradiol 1997; 18 : 1189-1190.
6Misra UK, Kalita J, Kumar S : A clinical, MRI and neurophysiological study of acute transverse myelitis. J Neurol Sci 1996, 138 : 150-156.
7Hoffman HL, Acute necrotic myelopathy. Brain 1955; 78 : 377-393.
8Jefferay DR, Mandler RN, Davis L : Transverse myelitis: retrospective analysis of 33 cases with differentiation of cases associated with multiple sclerosis and parainfectious events. Arch Neurol 1993; 50 : 532-535.
9Bansal R, Kalita J, Misra UK et al : A rare presentation of mumps. Pediatr Neurosurg 1998; 28 : 204-206.
10Kalita J, Misra UK, Mandal SK : Prognostic predictors of acute transverse myelitis. Acta Neurol Scand 1998; 98 : 60-62.
11Saleh M, Deeb AI, Basim A et al : Acute transverse myelitis: A localised fonn.of post infectious encephalawelitis. Brain 1997; 7 : 1115-1122.