Idiopathic severe recurrent transverse myelitis: a restricted variant of neuromyelitis optica

Clinical Neurology and Neurosurgery 107 (2005) 132–135

Case report

Idiopathic severe recurrent transverse myelitis: a restricted variant of neuromyelitis optica Koon Ho Chan∗ , Kin Lun Tsang, Gardnia Chung Yan Fong, Raymond Tak Fai Cheung, Shu Leong Ho Division of Neurology, University Department of Medicine, 4th Floor, Professorial Block, Pokfulam Road, Queen Mary Hospital, The University of Hong Kong, Hong Kong, SAR, China Received 12 January 2004; received in revised form 19 May 2004; accepted 28 May 2004

Abstract Recurrent idiopathic transverse myelitis occur in multiple sclerosis (MS) and neuromyelitis optica (NMO). In NMO, acute optic neuritis and myelitis occur, either monophasic or relapsing, without clinical manifestations of involvement of other parts of the central nervous system (CNS). Recent evidence suggests that NMO is different from multiple sclerosis. The authors reported two patients having severe recurrent transverse myelitis sparing the optic nerves and cerebral hemispheres. Both patients had longitudinally extensive myelitis in some attacks with poor neurological outcome despite aggressive immunomodulatory therapy. One patient had prominent clinical features of brainstem injury with radiological and histological confirmation of brainstem involvement, and the other patient had trigeminal neuralgia suggestive of possible brainstem dysfunction. Histologically, prominent necrosis and neutrophilic infiltration of spinal cord tissue without eosinophils or hyalinized vessels were observed, and oligoclonal bands were absent in their cerebrospinal fluid. It is likely to be a distinct idiopathic inflammatory demyelinating disorder restricted to the spinal cord and brainstem different from MS, but within the spectrum of NMO with probably an autoimmune basis. © 2004 Elsevier B.V. All rights reserved. Keywords: Neuromyelitis optica; Idiopathic transverse myelitis; Multiple sclerosis; Spinal cord; Brainstem

1. Introduction Acute transverse myelitis (ATM) is commonly due to postinfectious immune-mediated inflammation of the spinal cord [1]. Relapsing transverse myelitis occur in multiple sclerosis (MS), neuromyelitis optica (NMO) and other conditions including systemic lupus erythematosis and herpes simplex infection. Anti-phospholipid antibody syndrome and spinal arteriovenous malformation also cause recurrent myelopathy. Clinically, NMO patients developed ATM and optic neuritis sparing other parts of the central nervous system (CNS), which can be monophasic or relapsing [2,3]. In this communication, we described two patients having relapsing extensive myelitis sparing the optic nerves and cere∗

Corresponding author. Tel.: +852 2855 5318; fax: +852 2855 5322. E-mail address: [email protected] (K.H. Chan).

0303-8467/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.clineuro.2004.05.001

bral hemispheres with clinical manifestations of brainstem dysfunction.

2. Results 2.1. Patient 1 A 37-year-old woman presented with four limbs pinsand-needles sensation and weakness for 5 days. She took amitriptylline for left trigeminal neuralgia diagnosed 3 years ago with normal contrast magnetic resonance imaging (MRI) brain. She had no major medical illnesses. There were no preceding fever or flu-like illness. Examination revealed coherent speech, intact cranial nerves, tetraparesis of grade 3/5 power and impaired sensation with a sensory level at C4.

K.H. Chan et al. / Clinical Neurology and Neurosurgery 107 (2005) 132–135

133

Fig. 1. MRI of cervical spine of patient 1 showing T2 hyperintense lesion affecting C1–3 levels (A and B) which is gadolinium enhancing (C). Repeated MRI cervical spine months later revealed marked atrophy of the affected cervical cord region (D). MRI thoracic spine of patient 1 showed T2 hyperintense lesion extending from T7 to T9 (E) which was enhanced with gadolinium (F). MRI of cervical spine of patient 2 showed T2 hyperintense signal abnormality that extended from lower medulla oblongata to high cervical cord (G), which was minimally enhanced with gadolinium (H) probably due to late imaging.

MRI scan showed an T2 hyperintense lesion of the cervical cord at C1–3 with swelling and contrast enhancement (Fig. 1A–C). Cerebrospinal fluid (CSF) analysis showed pleocytosis of 28 uL−1 (97% lymphocytes), protein 0.64 g/L and glucose 3.2 mmol/L (blood 6.9 mmol/L); culture for bacteria, mycobacterium tuberculosis, fungi, polymerase chain reaction for herpes simplex virus, Ebstein Barr virus and MTB, serologies for varicella zoster virus, cytomegalovirus, adenovirus, human immunodeficiency virus, mycoplasma pneu-

monia, coxiella burnetti, borrelia burdorferi, venereal disease research laboratory and cytology were normal or negative. CSF was negative for oligoclonal bands with increased IgG to 8.0 mg/dL. Serum anti-nuclear antibody was positive at 1/40; anti-double stranded DNA antibody, anti-neutrophil cytoplasmic antibody and anti-extractable nuclear antigen antibodies were negative. Anti-cardiolipin antibodies were not elevated. ESR was 75 mm/h and C-reactive protein was elevated to 10 mg/dL. Cervical cord biopsy revealed conges-

134

K.H. Chan et al. / Clinical Neurology and Neurosurgery 107 (2005) 132–135

tion, hemosiderin deposits and polymorph infiltration without eosinophils, tumour cells or hyalinized vessels. MRI brain and brainstem, and visual evoked potentials (VEPs) were normal. She was treated with intravenous methylprednisone (IVMP) 1 gm daily for 5 days. Sensation and power improved to mild residual left hemiparesis with numbness weeks later. A month later, both legs deteriorated acutely to paraplegia with dense sensory loss below T6 level. MRI showed a T2 hyperintense and enhancing intramedullary lesion from T7 to T9 (Fig. 1D and E). IVMP 1 gm daily for 5 days was not followed by clinical improvement. Being a Jehovah’s witness, she refused plasmapheresis and intravenous immunoglobulin (IVIg) 0.4 gm/kg/day for 5 days was given. Sensation improved but paraplegia persisted. Two more attacks occurred in the next two months affecting T1 and C5–6 segments, respectively, and were treated similarly with IVMP followed by IVIg. Azathioprine was started and stepped up to 75 mg daily. Upper limbs power and sensation improved. CT thorax, abdomen and pelvis showed no tumours. MRI months later showed cord atrophy at lesion sites (Fig. 1F) while repeated MRI brain and VEPs were normal. She remained paraplegic with functional upper limbs and free of relapse for 2 years, azathioprine was reduced to 50 mg daily. Six months later, she had a recurrent attack affecting thoracic cord at T7–9 with partial sensory recovery after pulse IVMP for 5 days. Contrast MRI brain and VEPs were again normal. 2.2. Patient 2 A previously healthy 19-year-old girl presented with right temporal headache, vomiting, diplopia on left lateral gaze and photophobia, followed by dizziness and vertigo a month later. Right partial ptosis with normal pupils was noted without other neurological signs. Contrast CT brain, routine blood counts and biochemistry, thyroid function test, serum IgG, A and M levels, autoimmune markers and microbiological studies as for patient 1 were unremarkable. Serological, electrophysiological and edrophonium tests for myasthenia gravis were negative. Two months later she noticed neck pain, limbs stiffness, urinary incontinence and paroxysmal upper limbs spasms. Generalized hyperreflexia, upgoing plantars and sensory loss of limbs, neck and trunk were present. Contrast MRI brain and EEG were normal. MRI spine 6 weeks later revealed a T2 hyperintense lesion from C1–3 extending to lower medulla oblongata with only mild contrast enhancement (Fig. 1G and H) probably due to delayed neuroimaging. Lumbar puncture showed normal opening pressure. CSF was acellular, protein 0.3 g/L, glucose 4.6 mmol/L and IgG 3.2 mg/dL. CSF was negative for oligoclonal bands, cytology and microbiological studies as for patient 1. VEPs were normal. She was treated with IVMP 1 gm daily for 5 days and carbamazepine. Sensation and spasms improved but 3 weeks later, she developed worsened spasms with nystagmus on horizontal gaze, and sluggish gag reflex followed by aspiration pneumonia. Limb power remained full. Repeated MRI showed no cerebral hemispheric lesions but increased

swelling of the cervical cord lesion. IVIg 0.4 gm/kg/day for 5 days was given. Her pneumonia resolved with antibiotics, nystagmus subsided and spasms lessened. MRI 8 weeks later showed cord atrophy of C1–3 region. Repeated VEPs were normal. She succumbed a month later upon sudden cardiac arrest. Post-mortem examination showed atrophy of the cervical cord and multiple irregular patches of necrosis from below the inferior olivary nuclei to upper cervical segments. Microscopy showed extensive demyelination involving whole cross-section of brainstem at medulla oblongata and upper cervical segments. There was perivascular lymphocytic cuffing with infiltration of lymphocytes and macrophages into the glial matrix. Abundant foamy macrophages together with fluid accumulation caused early microcystic space in some areas. Gliosis and giant reactive astrocytes were noted. The cerebral hemispheres, optic nerves, thoracic and lumbar cord were normal. No cardiopulmonary pathologies were found.

3. Discussion Our patients probably had a distinct autoimmune disorder different from MS with recurrent attacks of inflammatory demyelination and necrosis restricted to the cord and brainstem sparing the cerebral hemispheres and optic nerves as evidenced by (1) severe relapsing transverse myelitis with longitudinal extent between one to three vertebral segments, (2) normal MRI brain and VEP, (3) absence of CSF oligoclonal bands, (4) severe residual neurological deficits and (5) distinct histological findings. The longitudinally extensive myelitis in some attacks of our patients, poor neurological recovery despite aggressive immunomodulatory therapy, sparing of optic nerves and cerebral hemispheres, and marked neutrophilic infiltration in patient 1 were inconsistent with prototypical MS. In MS, degree of necrosis is variable and neutrophilic infiltration rare [4]. Diagnosis of NMO requires all absolute criteria including (1) optic neuritis, (2) acute myelitis and (3) no evidence of clinical disease outside the optic nerve and spinal cord, plus one major or two minor supportive criteria [3]. The major criteria include (1) negative brain MRI at onset, (2) spinal cord MRI with signal abnormality extending over three or more vertebral segments and (3) CSF pleocytosis of >50 WBC/mm3 or >5 neutrophils/mm3 . Weinshenker described the most specific finding of NMO being a longitudinally extensive (three or more vertebral segments), central, gadolinium-enhancing cord lesion on magnetic resonance imaging (MRI) in the setting of myelitis [4], which is very rare in MS [4,5]. The normal MRI brain, absence of oligoclonal band, extensive myelitis with MRI abnormalities extending over three vertebral segments in some attacks and poor prognosis of our patients were compatible of NMO. In addition, MRI lesions extending to the brainstem [3] and pathological changes in brainstem associated with fatal dysautonomia have been described in NMO [13]. However,

K.H. Chan et al. / Clinical Neurology and Neurosurgery 107 (2005) 132–135

clinical manifestation of brainstem involvement excluded NMO. Importantly, NMO is pathologically characterized by prominent cord necrosis affecting grey and white matters, marked eosinophilic and neutrophilic infiltration, extensive complement activation in a unique perivascular pattern, vascular proliferation and hyalinization [6]. The marked neutrophilic infiltration of cord tissue in patient 1 and prominent necrosis in patient 2 were compatible with NMO, however there was no eosinophilic infiltration, and no hyalinized vessels which were present in all 9 patients reported by Lucchinnetti et al. [6]. The distinctive features of our patients were clinical manifestations of brainstem dysfunction. Dizziness, vertigo, nystagmus, dysphagia and impaired gag reflex in patient 2 were explained by involvement of the vestibular nucleus, inferior cerebellar peduncle, dorsal motor nucleus of vagus, nucleus ambiguus and nucleus solitaritus in the medulla oblongata. Her partial ptosis was probably due to involvement of sympathetic pathway in the medulla oblongata. She might die of dysautonomia. Trigeminal neuralgia in patient 1 suggested underling brainstem dysfunction though it remained possible that it was unrelated to her recurrent ATM which developed 3 years after onset of trigeminal neuralgia and MRI brain was normal. Tippet et al described isolated relapsing myelitis with longer durations between relapses, better prognosis and intact brainstem function [7]. Progressive necrotic myelopathy should be considered but it presents after age of 40 without brainstem involvement [8]. Our patients most likely suffered from an unique inflammatory demyelinating disorder within the spectrum of NMO but restricted to the spinal cord and brainstem. The boundaries of the spectrum of NMO are not well defined, and it is likely that some patients with isolated recurrent extensive transverse myelitis or isolated recurrent optic neuritis are in the spectrum of NMO. Brainstem involvement in NMO are not well documented and a thorough study on brainstem pathology in typical NMO patients would be valuable. We do not know the reason for the restricted clinical and pathological involvement to the spinal cord and brainstem in our patients. As these regions are not known to possess different glial cells or myelin proteins compared to other parts of the CNS, regional differences in the permeability of the blood–brain barrier [15,16] or ability in antigen processing and presentation to activate T cells may explain this unique topography of involvement. Patient 1 had no relapse for 2 years while taking azathioprine which is effective in NMO [11] and a recurrent attack 6 months after reduction of azathioprine dosage, implying that the pathogenesis of her myelitis may bear similarities to that of NMO. Recent evidence suggests an important role of humoral factors targeting the CNS perivascular regions in the pathogenesis of NMO [6]. Lennon et al. [9] discovered a marker autoantibody for NMO that binds selectively to an element associated with CNS capillaries, pia and subpia, which helps to distinguish clinically definite NMO from

135

typical MS [10]. As plasmapheresis is effective for steroidrefractory NMO [12,14], early aggressive plasmapheresis in severe myelitis as our cases may be effective.

Acknowledgements The authors thank Audrey Wong and Jess Chan for clinical and secretarial assistance.

References [1] Berman M, Feldman S, Alter M, Zilber N, Kahana E. Acute transverse myelitis: incidence and etiologic considerations. Neurology 1981;31:966–71. [2] Devic C. Myelite subaigue compliquee de neurite optique. Bull Med 1894;35:18–30. [3] Wingerchuk DM, Hogancamp WF, O’Brien PC, Weinshenker BG. The clinical course of neuromyelitis optica (Devic’s syndrome). Neurology 1999;53:1107–14. [4] Weinshenker BG. Neuromyelitis optica: what is it and what it might be. Lancet 2003;361:889–90. [5] Kidd D, Thorpe JW, Thompson AJ, et al. Spinal cord MRI using multi-array coils and fast spin echo II. Findings in multiple sclerosis. Neurology 1993;43:2632–7. [6] Lucchinetti CF, Mandler RN, McGavern D, Bruck W, Gleich G, Ransohoff RM, et al. A role of humoral mechanisms in the pathogenesis of Devic’s neuromyelitis optica. Brain 2002;125:1450–61. [7] Tippet DS, Fishman PS, Panitch HS. Relapsing transversed myelitis. Neurology 1991;41:703–6. [8] Katz JD, Ropper AH. Progressive necrotic myelopathy. Arch Neurol 2000;57:355–61. [9] Lennon VA, Luccinetti CF, Weinshenker BG. Identification of a marker autoantibody of neuromyelitis optica. Neurology 2003;60(Suppl 1):A519 [abstract]. [10] Weinshenker BG, Wingerchuk DM, Lucchinetti CF, Lennon VA. A marker autoantibody discriminates neuromyelitis optica and multiple sclerosis. Neurology 2003;60(Suppl 1):A520 [abstract]. [11] Mandler RN, Ahmed W, Dencoff JE. Devic’s neuromyelitis optica: a prospective study of seven patients treated with prednisone and azathioprine. Neurology 1988;51(4):1219–20. [12] Weinshenker BG, O’Brien PC, Patterson TM, Noseworthy JH, Lucchinetti CF, Dodick DW, et al. A randomized trial of plasma exchange in acute central nervous system inflammatory disease. Ann Neurol 1999;46(6):878–86. [13] Baudoin D, Gambarelli D, Gayraud D, Bensa P, Nicoli F, Sudan N, et al. Devic’s neuromyelitis optica: a clinicpathological review of the literature in connection with a case showing fatal dysautonomia. Clin Neuropathol 1998;17:175–83. [14] Keegan M, Pineda AA, McClelland RL, Darby CH, Rodriguez M, Weinshenker BG. Plasma exchange for severe attacks of CNS demyelination: predictors of response. Neurology 2002;58:143–6. [15] Lassmann H. Comparative neuropathology of chronic experimental allergic encephalomyelitis and multiple sclerosis. Schriftenr Neurol 1983;25:1–135. [16] Butler C, Baker D, O’Neill JK, Turk JL. Mononuclear cell trafficking and plasma protein extravasation into the CNS during chronic relapsing experimental allergic encephalomyelitis in Biozzi AB/H mice. J Neurol Sci 1991;104:9–12.