- Email: [email protected]
Anti-tuberculosis treatment for Devic’s neuromyelitis optica
Journal of Clinical Neuroscience 17 (2010) 1372–1377
Contents lists available at ScienceDirect
Journal of Clinical Neuroscience journal homepage: www.elsevier.com/locate/jocn
Clinical Study
Anti-tuberculosis treatment for Devic’s neuromyelitis optica Yan-qing Feng *, Ning Guo, Fan Huang, Xi Chen, Qiao-song Sun, Jun-xiu Liu Department of Neurology, The First Affiliated Hospital of Sun Yat-Sen University, 58 Zhong Shan Road 2, Guangzhou 510080, China
a r t i c l e
i n f o
Article history: Received 10 June 2009 Accepted 7 February 2010
Keywords: Anti-tuberculosis treatment Neuromyelitis optica Tuberculous infection
a b s t r a c t There are no specific treatments for patients with acute, severe neurological deficits caused by neuromyelitis optica (NMO) who fail to recover after treatment with high-dose corticosteroids. We evaluated the clinical response of anti-tuberculosis treatment (ATT) in patients suffering from steroid-refractory NMO, and investigated the correlation between NMO and tuberculous infection of the central nervous system (CNS). We conducted this prospective, controlled study in southern China. Twelve patients with steroidrefractory NMO were monitored during ATT and compared with a control group of 13 patients with the same type of NMO who received current standard therapies. A molecular diagnostic test was carried out and Extended Disability Status Scale (EDSS) score analysis, visual acuity, the number of relapses and MRI changes were evaluated at study entry and after 1 and 2 years of therapy. ATT may lead to the recovery of important neurological functions and all our patients responded positively to therapy. EDSS score and visual acuity improved and abnormalities in the spinal cord, observed by MRI, markedly decreased over time. ATT also significantly reduced the rate of relapse. By comparison, in the control group, a significant clinical deterioration was observed, and patients did not show favourable EDSS scores and MRI changes. This study suggests that CNS infection with Mycobacterium tuberculosis is an important cause of NMO. Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction
2. Methods
Neuromyelitis optica (NMO) is an inflammatory demyelinating disease that selectively targets optic nerves and the spinal cord either simultaneously or in quick succession. Patients with NMO generally have a poor prognosis. The etiology and nosology of NMO have been controversial. No effective treatment for NMO has been proven; however, it is commonly treated with corticosteroids and patients who experience recurrent attacks are often managed with chronic immunosuppressing treatments. Patients who experience severe attacks of NMO, and fail to respond to corticosteroid therapy, can develop severe, permanent neurological disabilities or die. An association between NMO and evolving pulmonary tuberculosis has been reported in the past 40 years.1–4 This prospective study tests the hypothesis that anti-tuberculosis chemotherapy is an effective treatment for steroid-refractory NMO patients, and defines the correlation between steroid-refractory NMO and tuberculosis infection. We investigated, at the clinical and molecular level, NMO patients who did not respond to treatment with corticosteroids.
2.1. Study design This prospective, controlled and nonrandomized study involved 25 patients and was conducted between January 2003 and December 2007. All patients displayed clinically definite NMO according to Wingerchurk’s 1999 criteria.5 Steroid-refractory NMO was defined as treatment with intravenous (iv) pulsed doses of methylprednisolone for a minimum of 5 days with, at best, trivial improvement and, at worst, continued deterioration. All patients received at least one course of iv immunoglobulin (Ig) 0.4 g/kg/ day for 5 days, and this treatment did not show any effect. Appropriate investigations ruled out alternative diagnoses. Twelve patients (treatment group) underwent anti-tuberculosis treatment (ATT) without steroid therapy. Patients were selected based on the following considerations: (i) consent to receive ATT treatment; (ii) able to undergo lumbar puncture before ATT; and (iii) consent to stop steroid therapy. Thirteen patients (control group) underwent standard therapies. 2.2. Treatment group patients
* Corresponding author. Tel.: +86 20 83759031; fax: +86 20 87332686. E-mail address: [email protected] (Y.-q. Feng). 0967-5868/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.jocn.2010.02.023
All patients in the treatment group (10 female, 2 male; age range = 36.83 ± 15.89) were informed of the potential short- and
1373
Y.-q. Feng et al. / Journal of Clinical Neuroscience 17 (2010) 1372–1377
long-term drug complications. Written informed consent was obtained from all patients and the conduct of the study was approved by the university ethics committee. A detailed neurological examination was carried out, including the motor, sensory, visual and sphincter systems. A Mantoux test and chest X-ray were performed to exclude infection by other agents. Cerebrospinal fluid (CSF) samples (10 mL or more) were obtained from all 12 patients. The routine CSF examination included cell counts, protein and glucose measurements, direct smear and culture for acid-fast bacilli, a Gram stain, an Indian ink preparation and a Mycobacterium tuberculosis antibody test (purified protein derivative–IgG). Before ATT was commenced, all patients had a baseline evaluation. 2.3. Control patients The control group comprised 13 people (11 female, 2 male; age range = 38.31 ± 13.93 years). Individuals suffering from respiratory distress, fever, or other critical illnesses before enrolment were excluded. The inclusion time point was taken from the time when these patients were steroid-refractory, and unresponsive to iv Ig treatment. At the beginning of the trial all individuals underwent a chest X-ray and Mantoux test to exclude active tuberculosis, demographic data were obtained, a full history was taken and their Extended Disability Status Scale (EDSS) score was evaluated. 2.4. Nested polymerase chain reaction Nested polymerase chain reaction (N-PCR) was used to determine whether patients in the treatment group were suffering from direct tuberculosis infection of the central nervous system (CNS). We aspirated 5 mL of CSF and centrifuged samples at 12 000 g for 15 min. DNA extraction was performed using the phenol–chloroform method and purified by ethanol precipitation. Two pairs of oligonucleotide primers specific for the MPB64 gene of M. tuberculosis were prepared for N-PCR.6 The second-round amplification products were analyzed by agarose gel electrophoresis and visualized using an ultraviolet transilluminator. A 200 base-pair (bp) band indicated the presence of the M. tuberculosis gene. 2.5. Treatment and follow-up Prior to ATT initiation, all treatments with corticosteroids and other systemic immunosuppression therapies were discontinued for patients in the treatment group. Our treatment protocols comprised a four anti-tuberculosis–drug regimen (per day: isoniazid 8 mg/kg, rifampicin 10 mg/kg, pyrazinamide 25 mg/kg, streptomycin 20 mg/kg; for 2 months), a three-drug regimen (isoniazid, rifampicin and pyrazinamide; for 4 months with the same dosages), followed by treatment with a combination of isoniazid and rifampicin with the same dosages that ceased after 24 months. Treatment was carried out under extensive observation. Demographic data and the number of new attacks were recorded for each patient. Efficacy was assessed using the EDSS scores at study entry (Y0) and after 1 (Y1) and 2 (Y2) years of therapy. All patients also underwent MRI at the same times. MRI were recorded in T1weighted, T2-weighted and fluid-attenuated inversion recovery (FLAIR) modes, with and without gadolinium enhancement. The clinical and radiological evaluations were unblinded. Safety was assessed quarterly by the treating neurologist. In the control group, treatment with steroids, plasma exchange, iv Ig and immunosuppressive drugs was allowed during the 2-year observational period. Azathioprine was used for five patients, mitoxantrone hydrochloride for three patients, cyclophosphamide for two patients and one patient underwent plasma exchange during the study period. The other two patients received steroid treatment without immunosuppressive drugs.
A relapse was defined as the appearance of a new symptom(s) or worsening of previous symptoms that, in the absence of fever, lasted for more than 24 hours, resulting in an increase of at least 0.5 points in the patient’s EDSS score. Patients were informed to contact the outpatients’ clinic if a relapse occurred. Neurologically confirmed relapses were treated with vitamin B12 in the treatment group, while the control patients were treated with a cycle of highdose steroid and/or iv Ig. 2.6. Statistical analysis Treatment and control group variables were compared using the Student’s t-test and the chi-squared (v2) test. Statistical Package for the Social Sciences (SPSS) version 13.0 analysis system software (SPSS; Chicago, IL, USA) was used to perform these tests. 3. Results 3.1. Baseline results Baseline clinical characteristics data for all patients are shown in Table 1. These data show that patients in the treatment and control groups were well matched on all clinical parameters prior to treatment. The symptoms were associated with myelopathy and optic neuritis (ON) in all patients with NMO who were steroidrefractory. All patients received iv-pulsed doses of methylprednisolone or combined immunosuppressing treatments, and failed to recover. In the treatment group, symptoms indicating motor involvement included paraparesis in nine patients and tetraparesis in the other three patients. Sphincter symptoms were also associated in 10 patients. ON was bilateral in eight patients and unilateral in four patients. Corticosteroid medication was ineffective after some relapses in the 10 relapsing–remitting patients, and had no effect at the start of treatment in two patients. None of the patients had definite evidence of tuberculosis from chest radiography. However, pleural thickening was observed in two patients, pneumonia occurred in three patients and unilateral pleural effusion occurred in one patient. CSF abnormalities were detected in 10 of 12 patients, the most frequent being high protein levels and pleocytosis. CSF protein levels were increased (>0.6 g/L) in seven patients and six patients had elevated CSF leukocyte counts. The CSF glucose concentration was normal for all patients. Seven patients had a Mantoux test result P10 mm. N-PCR showed that only two of 12 CSF samples were positive for M. tuberculosis. Spinal cord MRIs were abnormal for all patients; the most common abnormality was an increased signal intensity spanning several sections of the spinal cord on T2-weighted images, mainly involving the central part of the cord. Cord swelling was observed for nine patients. The control patients were similar in terms of age, sex and disease duration, and were evenly matched for EDSS at the time of entry into the trial. Variables were compared between the treatment
Table 1 Baseline clinical features in patients treated for Devic’s neuromyelitis optica with anti-tuberculosis chemotherapy and control groups
Number Female/male Age of study entry (years) Disease duration (months) Mean baseline EDSS Relapses before ATT
Treatment group
Control group
12 10/2 36.83 ± 15.89 47.33 ± 68.06 6.38 ± 1.25 52
13 11/2 38.31 ± 13.93 52.65 ± 45.27 5.85 ± 1.03 57
ATT = anti-tuberculosis treatment, EDSS = Extended Disability Status Scale score.
1374
Y.-q. Feng et al. / Journal of Clinical Neuroscience 17 (2010) 1372–1377
Table 2 Analysis of the Extended Disability Status Scale scores and relapse changes in the patients treated for Devic’s neuromyelitis optica with anti-tuberculosis chemotherapy and control groups Variable
Treatment group
Control group
p-value
Mean baseline EDSS 1-year EDSS 2-year EDSS Relapses during ATT
6.38 ± 1.25 2.50 ± 1.52 2.21 ± 1.32 8à
5.85 ± 1.03 7.35 ± 1.66 8.15 ± 1.68 15
>0.05 <0.01 <0.01 <0.01
ATT = anti-tuberculosis treatment, EDSS = Extended Disability Status Scale score. Statistically significant (independent Student’s t-test). à Statistically significant (chi-squared test).
and control groups using the Student’s t-test and chi-squared tests. No significant differences between the treatment and control groups were apparent. 3.2. Control group During the 2-year study period, five patients in the control group died. Three of these patients died from respiratory failure,
one patient died suddenly from a presumed pulmonary embolus and one patient died of pneumonia. The EDSS score progressively and significantly (p < 0.01) increased from 5.85 ± 1.03 at Y0, to 7.35 ± 1.66 at Y1 and 8.15 ± 1.68 at Y2 (Table 2). These patients had 15 relapses in total. MRI did not show any decrease on the T2-weighted and FLAIR lesion load during the 2-year follow-up.
3.3. Treatment group In the treatment group, we withdrew fully corticosteroid and immunosuppressive treatments before anti-tuberculosis chemotherapy. The response occurred early in the course of treatment for all patients. Clinical worsening was halted during the 2-week course of anti-tuberculosis chemotherapy. For three patients, a dramatic improvement in limb weakness was observed within 1– 3 days of commencing treatment. Clinical outcomes were evaluated after 1 and 2 years and significantly decreased EDSS scores were observed at Y1 (2.50 ± 1.52; Y2 vs. Y0 and Y2 vs. control, p < 0.01) and at Y2 (2.21 ± 1.32; Y2 vs. Y0 and Y2 vs. control, p < 0.01). These improved clinical outcomes were sustained and continued to improve. All patients felt their motor and sphincter
Fig. 1. MRI changes in the spinal cord of patient 1 diagnosed with neuromyelitis optica (NMO) and treated with anti-tuberculosis treatment (ATT). (A) Sagittal T1-weighted MRI showing thickening of the cord from the inferior medulla to the T2 level, with subtle intraparenchymal hyperintensity. (B) Sagittal T2-weighted MRI showing a contiguous area of increased signal intensity. (C) Axial T2-weighted MRI showing the lesion to involve the whole cord. (D) Sagittal T1-weighted MRI after 1 year of treatment, showing the swelling of the cervical spinal cord and patchy abnormal signal have resolved. (E) Sagittal T2-weighted MRI showing a small area of abnormal signal remains in the central cervical cord. (F) Axial T2-weighted MRI showing a small lesion is visible at the C6 level. (G) Sagittal T2-weighted MRI after 2 years of treatment showing a small lesion remains. (H) Axial T2-weighted cord MRI after 2 years of treatment.
Y.-q. Feng et al. / Journal of Clinical Neuroscience 17 (2010) 1372–1377
dysfunction symptoms markedly improve and also noticed improvement in their sensory symptoms. Visual acuity gradually ameliorated in all patients after 2 years of treatment. During the 24-month trial, six of the 12 patients in the treatment group experienced eight relapses in total. There was a lower rate of relapse incidence than for the control group (chi-squared test, p < 0.05). Six relapses occurred in the first year, particularly during the first 3 months of treatment (four relapses). Two patients experienced relapses of visual decline as their limb weakness improved. Without changing our regimen, these relapses gradually recovered as treatment continued. Overall, throughout the 2 years of treatment, exacerbation rates gradually declined. ATT significantly limited the occurrence of relapses in patients suffering from steroid-refractory NMO. 3.4. Efficacy as measured by MRI 3.4.1. Illustrative patients treated with ATT During the study period, MRI demonstrated that abnormal signal and expansion of the spinal cord progressively decreased in all patients after the initiation of ATT. Signal abnormalities on T2-weighted MRI were markedly decreased in size and prominence and had virtually resolved by the end of treatment. Additionally, regions showing well-defined enhancing signals, and poorly defined marginal enhancement in pre-treatment studies, resolved with treatment. Figs. 1–3 show the MRI changes in the spinal cord of patients 1, 4 and 5. Patient 1 was a 24-year-old female who had experienced an acute attack of bilateral lower extremity weakness and bladder incontinence, and fully recovered after 15 days’ treatment with dexamethasone. Two episodes of relapse occurred during the following 2 months. On admission, visual acuity was reduced in both eyes to perception of light; quadriparesis, voiding difficulties and sensory loss had also occurred. She was treated with pulsed doses of methylprednisolone and intravenous immunoglobulin but did not improve. After ATT her symptoms gradually resolved (Fig. 1). Patient 4 was a 34-year-old female who developed bilateral ON and fully recovered after 1 month of steroid treatment. She later experienced two relapses of back pain, bilateral lower extremity parasthesia and weakness. Partial recoveries were achieved by pulsed methylprednisolone and iv Ig treatments. Three years after
1375
onset, she presented with quadriplegia and voiding difficulty, but the same medication did not take effect. She underwent ATT, and her condition gradually recovered, although two relapses still occurred during the first year of ATT (Fig. 2). Patient 5 was a 33year-old female, who had a history of ON 8 months before, and who developed paraplegia that quickly progressed. Pulse methylprednisolone, intravenous immunoglobulin and azathioprine were initiated with no resolution and she continued to deteriorate. Her cerebrospinal fluid was normal. Following complete paralysis and sensory loss below the T4 level, our treatment regimen commenced. After 1 month of chemotherapy, the patient recovered sphincter function with grade 2 muscle strength in the lower limbs and sensory deficit greatly improved. Six months later, she was able to walk without a cane. Full sensation returned and abnormal MRI signal had completely disappeared after 1 year of ATT treatment. The Extended Disability Status Scale (EDSS) score was 2.0. In the second year, the EDSS score was stable (Fig. 3). 3.4.2. Patients in the control group In the control group, NMO cord lesions were usually persistent or became more severe and extensive over time; three patients developed focal spinal cord atrophy, as detected in the follow-up study. During the observational period, a significant decrease in the intramedullary lesion burden was not detected in the surviving control group patients.
4. Discussion This study shows that anti-tuberculosis chemotherapy can benefit patients with steroid-refractory NMO, suggesting that tuberculosis infection correlates closely with NMO. This study was undertaken in southern China, where tuberculosis infection is epidemic with a prevalence of 44.5%.7 The study was based on trial antituberculosis chemotherapy; steroids and other immunosuppressing drugs were not used. The findings show that ATT is effective and adequate to treat NMO. The effects were associated with significantly improved neurological status and changes in the spinal cord were observed by MRI. The positive results obtained for mycobacterial DNA in CSF with N-PCR revealed that pathogenesis involves direct CNS infection.
Fig. 2. MRI changes in the spinal cord of patient 4 diagnosed with neuromyelitis optica (NMO) and treated with anti-tuberculosis treatment (ATT). (A) Sagittal T2-weighted MRI showing a long segment, high intensity signal abnormality from C1 to T5 in the spinal cord before our treatment regimen commenced. (B) After 5 months’ treatment, the lesion was greatly decreased. (C) By the end of the first year, only traces of the lesion remained. (D) After 2 years’ treatment, all lesions had disappeared.
1376
Y.-q. Feng et al. / Journal of Clinical Neuroscience 17 (2010) 1372–1377
Fig. 3. MRI changes in the spinal cord of patient 5 diagnosed with neuromyelitis optica (NMO) and treated with anti-tuberculosis treatment (ATT). (A) Sagittal T2weighted MRI showing a contiguous area of increased signal intensity spanning from C4 to T4. (B,C) Axial T2-weighted MRI showing the high signal involved the whole cord with cord. (D) Sagittal T2-weighted MRI after 1 year of treatment, showing a normal spinal cord. (E,F) Axial T2-weighted MRI, showing no lesions.
M. tuberculosis infection can involve the spinal cord and optic nerve. In three reported patients with acute myelopathy associated with pulmonary tuberculosis who underwent post-mortem examination,8 myelopathy had been progressive or relapsing, and demyelination of the white and grey matter was found in the spinal cord. A similar report suggested that clinical improvement can be achieved after steroid therapy.9 Gupta et al. studied MRIs of 20 patients with intraspinal tuberculosis, and revealed that cord edema, cavitation and linear enhancement of the surface of the spinal cord were common; these features are also found in Devic’s disease.10 Some reports suggest a correlation between pulmonary tuberculosis and NMO. El Otmani et al. reported a patient with NMO with evolving pulmonary tuberculosis. Following ATT, the patient recovered normal mobility and sphincter control.1 Silber et al. reviewed
eight patients with NMO with pulmonary tuberculosis.3 Two of these patients died; the cause was unknown in one patient, the other died from extensive pulmonary tuberculosis. One patient’s condition did not improve during the 5-month observation period but some degree of improvement occurred in the remaining five patients, ranging from improved arm strength to full recovery of leg power. The authors proposed this syndrome is most likely due to an immune reaction to tuberculosis. It is possible that ATT, initiated to treat pulmonary tuberculosis, simultaneously cured NMO because of their similar etiology. Our patients recovered more completely than those in the study by Siber et al.; possibly this was because patients included in our study had not progressed to an advanced stage of infection. Our study challenges the classic view of NMO as a subtype of multiple sclerosis, and suggests that tuberculosis infection is one of the most important causes of NMO and might be involved in the pathogenesis. Specific associations with viral diseases such as varicella,11 human immunodeficiency virus–1,12 and hepatitis virus A13 have been reported. Many studies also show that NMO has striking neuropathological features that are not attributable solely to demyelination, indicating a pathological mechanism distinct from that observed in patients with multiple sclerosis. For example, Lucchinetti et al. performed a detailed analysis of the autopsy of nine patients with Devic’s disease.14 The lesions were severely destructive with cavitation, necrosis and acute axonal pathology, in both gray and white matter. Inflammatory infiltrates were characterized by extensive macrophage infiltration associated with large numbers of perivascular granulocytes and eosinophils. Similarly, Beck16 (1927), Hassin17 (1937), and Lowenberg18 (1947) all described involvement of both gray and white matter of the spinal cord, marked inflammatory infiltrates and destruction of the spinal cord extending continuously through multiple segments. These changes were thought to be distinct from those associated with multiple sclerosis. The findings from our study suggest that direct infection could explain these common findings in NMO. Our study also analyzed the occurrence of relapses in patients with NMO. A decrease in the relapse rate correlated with a decrease in EDSS after 2 years, suggesting that a relapse represents real lasting efficacy, rather than a reflection of disease progression. There was a tendency towards reduction in the relapse rate as treatment continued. Four relapses occurred during the first 3 months of treatment, and relapses tended to occur in patients with a long history of disease and severe demyelination. The mechanism of relapse requires further investigation but it may be similar to the paradoxical deterioration during tuberculosis infection. Relapses are not restricted to patients with demyelinating disease; they can occur several times in patients with tuberculous myelopathy during treatment.15 Methodological limitations could impact the findings of this study. Overall, the patient groups showed comparable clinical and MRI parameters at the start of the study. However, the number of patients in this trial was small and the inevitable risk of selection bias exists. Also, relapses are less well defined in patients with steroid-refractory NMO. Further studies of the efficacy of antituberculosis chemotherapy would determine if this treatment regimen is effective for relapsing–remitting NMO. An additional study of a longer duration and larger sample size is required for steroidrefractory NMO.
5. Conclusion This small, controlled study shows the long-term clinical efficacy of ATT in patients with steroid-refractory NMO. This treatment can reduce disease activity and halt progression, and can also result in a significant recovery of neurological deficits. Considering NMO
Y.-q. Feng et al. / Journal of Clinical Neuroscience 17 (2010) 1372–1377
is a treatable condition, and that a delay in treatment leads to significant morbidity and mortality, we recommend that anti-tuberculosis chemotherapy should be considered for patients with steroidrefractory NMO. References 1. El Otmani H, Rafai MA, Moutaouakil F, et al. Devic’s optic neuromyelitis and pulmonary tuberculosis. Rev Mal Respir 2005;22:143–6. 2. Dixon GJ, Roberts GB, Tyrrell WF. The relationship of neuropathy to the treatment of tuberculosis with isoniazid. Scott Med J 1956;1:350–4. 3. Silber MH, Willcox PA, Bowen RM, et al. Neuromyelitis optica (Devic’s syndrome) and pulmonary tuberculosis. Neurology 1990;40:934–8. 4. Papais-Alvarenga RM, Miranda-Santos CM, Puccioni-Sohler M, et al. Optic neuromyelitis syndrome in Brazilian patients. J Neurol Neurosurg Psychiatry 2002;73:429–35. 5. Wingerchuk DM, Hogancamp WF, O’Brien PC, et al. The clinical course of neuromyelitis optica (Devic’s syndrome). Neurology 1999;53:1107–14. 6. Bhigjee AI, Padayachee R, Paruk H, et al. Diagnosis of tuberculous meningitis: clinical and laboratory parameters. Int J Infect Dis 2007;11:348–54. 7. National Technic Steering Group of the Epidemiological Sampling Survey for Tuberculosis, Duanmu H. Report on fourth national epidemiological sampling survey of tuberculosis. Zhonghua Jie He He Hu Xi Za Zhi 2002;25:3–7. 8. Hughes RA, Mair WG. Acute necrotic myelopathy with pulmonary tuberculosis. Brain 1977;100:223–38.
1377
9. Reid AC, Bone I. Myelopathy during treatment for pulmonary tuberculosis. Postgrad Med J 1980;56:511–2. 10. Gupta RK, Gupta S, Kumar S, et al. MRI in intraspinal tuberculosis. Neuroradiology 1994;36:39–43. 11. Merle H, Smadja D, Cordoba A. Optic neuromyelitis and bilateral acute retinal necrosis due to varicella zoster in a patient with AIDS. J Fr Ophtalmol 1998;21:381–6. 12. Mirsattari SM, Johnston JB, McKenna R, et al. Aboriginals with multiple sclerosis: HLA types and predominance of neuromyelitis optica. Neurology 2001;56:317–23. 13. Jouhadi Z, Ouazzani I, Abid A, et al. Devic’s optic neuromyelitis and viral hepatitis type A. A paediatric case report. Rev Neurol (Paris) 2004;160: 1198–202. 14. Lucchinetti CF, Mandler RN, McGavern D, et al. A role for humoral mechanisms in the pathogenesis of Devic’s neuromyelitis optica. Brain 2002;125:1450–61. 15. White VL, Al-Shahi R, Gamble E, et al. Transverse myelopathy and radiculomyelopathy associated with pulmonary atypical Mycobacterium infections. Thorax 2001;56:158–60. 16. Beck GM. A case of diffuse myelitis associated with optic neuritis. Brain 1927;50:687–703. 17. Hassin GB. Neuroptica myelitis versus multiple sclerosis: a pathologic study [abstract]. Arch Neurol Psychiat (Chicago) 1937;37:1083–99. 18. Lowenberg K, DeJong RN, Foster DB. Neuromyelitis optica: its relations to progressive necrosis of the spinal cord and acute multiple sclerosis [abstract]. Trans Am Neurol Assoc 1947;67:50–66.
Contents lists available at ScienceDirect
Journal of Clinical Neuroscience journal homepage: www.elsevier.com/locate/jocn
Clinical Study
Anti-tuberculosis treatment for Devic’s neuromyelitis optica Yan-qing Feng *, Ning Guo, Fan Huang, Xi Chen, Qiao-song Sun, Jun-xiu Liu Department of Neurology, The First Affiliated Hospital of Sun Yat-Sen University, 58 Zhong Shan Road 2, Guangzhou 510080, China
a r t i c l e
i n f o
Article history: Received 10 June 2009 Accepted 7 February 2010
Keywords: Anti-tuberculosis treatment Neuromyelitis optica Tuberculous infection
a b s t r a c t There are no specific treatments for patients with acute, severe neurological deficits caused by neuromyelitis optica (NMO) who fail to recover after treatment with high-dose corticosteroids. We evaluated the clinical response of anti-tuberculosis treatment (ATT) in patients suffering from steroid-refractory NMO, and investigated the correlation between NMO and tuberculous infection of the central nervous system (CNS). We conducted this prospective, controlled study in southern China. Twelve patients with steroidrefractory NMO were monitored during ATT and compared with a control group of 13 patients with the same type of NMO who received current standard therapies. A molecular diagnostic test was carried out and Extended Disability Status Scale (EDSS) score analysis, visual acuity, the number of relapses and MRI changes were evaluated at study entry and after 1 and 2 years of therapy. ATT may lead to the recovery of important neurological functions and all our patients responded positively to therapy. EDSS score and visual acuity improved and abnormalities in the spinal cord, observed by MRI, markedly decreased over time. ATT also significantly reduced the rate of relapse. By comparison, in the control group, a significant clinical deterioration was observed, and patients did not show favourable EDSS scores and MRI changes. This study suggests that CNS infection with Mycobacterium tuberculosis is an important cause of NMO. Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction
2. Methods
Neuromyelitis optica (NMO) is an inflammatory demyelinating disease that selectively targets optic nerves and the spinal cord either simultaneously or in quick succession. Patients with NMO generally have a poor prognosis. The etiology and nosology of NMO have been controversial. No effective treatment for NMO has been proven; however, it is commonly treated with corticosteroids and patients who experience recurrent attacks are often managed with chronic immunosuppressing treatments. Patients who experience severe attacks of NMO, and fail to respond to corticosteroid therapy, can develop severe, permanent neurological disabilities or die. An association between NMO and evolving pulmonary tuberculosis has been reported in the past 40 years.1–4 This prospective study tests the hypothesis that anti-tuberculosis chemotherapy is an effective treatment for steroid-refractory NMO patients, and defines the correlation between steroid-refractory NMO and tuberculosis infection. We investigated, at the clinical and molecular level, NMO patients who did not respond to treatment with corticosteroids.
2.1. Study design This prospective, controlled and nonrandomized study involved 25 patients and was conducted between January 2003 and December 2007. All patients displayed clinically definite NMO according to Wingerchurk’s 1999 criteria.5 Steroid-refractory NMO was defined as treatment with intravenous (iv) pulsed doses of methylprednisolone for a minimum of 5 days with, at best, trivial improvement and, at worst, continued deterioration. All patients received at least one course of iv immunoglobulin (Ig) 0.4 g/kg/ day for 5 days, and this treatment did not show any effect. Appropriate investigations ruled out alternative diagnoses. Twelve patients (treatment group) underwent anti-tuberculosis treatment (ATT) without steroid therapy. Patients were selected based on the following considerations: (i) consent to receive ATT treatment; (ii) able to undergo lumbar puncture before ATT; and (iii) consent to stop steroid therapy. Thirteen patients (control group) underwent standard therapies. 2.2. Treatment group patients
* Corresponding author. Tel.: +86 20 83759031; fax: +86 20 87332686. E-mail address: [email protected] (Y.-q. Feng). 0967-5868/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.jocn.2010.02.023
All patients in the treatment group (10 female, 2 male; age range = 36.83 ± 15.89) were informed of the potential short- and
1373
Y.-q. Feng et al. / Journal of Clinical Neuroscience 17 (2010) 1372–1377
long-term drug complications. Written informed consent was obtained from all patients and the conduct of the study was approved by the university ethics committee. A detailed neurological examination was carried out, including the motor, sensory, visual and sphincter systems. A Mantoux test and chest X-ray were performed to exclude infection by other agents. Cerebrospinal fluid (CSF) samples (10 mL or more) were obtained from all 12 patients. The routine CSF examination included cell counts, protein and glucose measurements, direct smear and culture for acid-fast bacilli, a Gram stain, an Indian ink preparation and a Mycobacterium tuberculosis antibody test (purified protein derivative–IgG). Before ATT was commenced, all patients had a baseline evaluation. 2.3. Control patients The control group comprised 13 people (11 female, 2 male; age range = 38.31 ± 13.93 years). Individuals suffering from respiratory distress, fever, or other critical illnesses before enrolment were excluded. The inclusion time point was taken from the time when these patients were steroid-refractory, and unresponsive to iv Ig treatment. At the beginning of the trial all individuals underwent a chest X-ray and Mantoux test to exclude active tuberculosis, demographic data were obtained, a full history was taken and their Extended Disability Status Scale (EDSS) score was evaluated. 2.4. Nested polymerase chain reaction Nested polymerase chain reaction (N-PCR) was used to determine whether patients in the treatment group were suffering from direct tuberculosis infection of the central nervous system (CNS). We aspirated 5 mL of CSF and centrifuged samples at 12 000 g for 15 min. DNA extraction was performed using the phenol–chloroform method and purified by ethanol precipitation. Two pairs of oligonucleotide primers specific for the MPB64 gene of M. tuberculosis were prepared for N-PCR.6 The second-round amplification products were analyzed by agarose gel electrophoresis and visualized using an ultraviolet transilluminator. A 200 base-pair (bp) band indicated the presence of the M. tuberculosis gene. 2.5. Treatment and follow-up Prior to ATT initiation, all treatments with corticosteroids and other systemic immunosuppression therapies were discontinued for patients in the treatment group. Our treatment protocols comprised a four anti-tuberculosis–drug regimen (per day: isoniazid 8 mg/kg, rifampicin 10 mg/kg, pyrazinamide 25 mg/kg, streptomycin 20 mg/kg; for 2 months), a three-drug regimen (isoniazid, rifampicin and pyrazinamide; for 4 months with the same dosages), followed by treatment with a combination of isoniazid and rifampicin with the same dosages that ceased after 24 months. Treatment was carried out under extensive observation. Demographic data and the number of new attacks were recorded for each patient. Efficacy was assessed using the EDSS scores at study entry (Y0) and after 1 (Y1) and 2 (Y2) years of therapy. All patients also underwent MRI at the same times. MRI were recorded in T1weighted, T2-weighted and fluid-attenuated inversion recovery (FLAIR) modes, with and without gadolinium enhancement. The clinical and radiological evaluations were unblinded. Safety was assessed quarterly by the treating neurologist. In the control group, treatment with steroids, plasma exchange, iv Ig and immunosuppressive drugs was allowed during the 2-year observational period. Azathioprine was used for five patients, mitoxantrone hydrochloride for three patients, cyclophosphamide for two patients and one patient underwent plasma exchange during the study period. The other two patients received steroid treatment without immunosuppressive drugs.
A relapse was defined as the appearance of a new symptom(s) or worsening of previous symptoms that, in the absence of fever, lasted for more than 24 hours, resulting in an increase of at least 0.5 points in the patient’s EDSS score. Patients were informed to contact the outpatients’ clinic if a relapse occurred. Neurologically confirmed relapses were treated with vitamin B12 in the treatment group, while the control patients were treated with a cycle of highdose steroid and/or iv Ig. 2.6. Statistical analysis Treatment and control group variables were compared using the Student’s t-test and the chi-squared (v2) test. Statistical Package for the Social Sciences (SPSS) version 13.0 analysis system software (SPSS; Chicago, IL, USA) was used to perform these tests. 3. Results 3.1. Baseline results Baseline clinical characteristics data for all patients are shown in Table 1. These data show that patients in the treatment and control groups were well matched on all clinical parameters prior to treatment. The symptoms were associated with myelopathy and optic neuritis (ON) in all patients with NMO who were steroidrefractory. All patients received iv-pulsed doses of methylprednisolone or combined immunosuppressing treatments, and failed to recover. In the treatment group, symptoms indicating motor involvement included paraparesis in nine patients and tetraparesis in the other three patients. Sphincter symptoms were also associated in 10 patients. ON was bilateral in eight patients and unilateral in four patients. Corticosteroid medication was ineffective after some relapses in the 10 relapsing–remitting patients, and had no effect at the start of treatment in two patients. None of the patients had definite evidence of tuberculosis from chest radiography. However, pleural thickening was observed in two patients, pneumonia occurred in three patients and unilateral pleural effusion occurred in one patient. CSF abnormalities were detected in 10 of 12 patients, the most frequent being high protein levels and pleocytosis. CSF protein levels were increased (>0.6 g/L) in seven patients and six patients had elevated CSF leukocyte counts. The CSF glucose concentration was normal for all patients. Seven patients had a Mantoux test result P10 mm. N-PCR showed that only two of 12 CSF samples were positive for M. tuberculosis. Spinal cord MRIs were abnormal for all patients; the most common abnormality was an increased signal intensity spanning several sections of the spinal cord on T2-weighted images, mainly involving the central part of the cord. Cord swelling was observed for nine patients. The control patients were similar in terms of age, sex and disease duration, and were evenly matched for EDSS at the time of entry into the trial. Variables were compared between the treatment
Table 1 Baseline clinical features in patients treated for Devic’s neuromyelitis optica with anti-tuberculosis chemotherapy and control groups
Number Female/male Age of study entry (years) Disease duration (months) Mean baseline EDSS Relapses before ATT
Treatment group
Control group
12 10/2 36.83 ± 15.89 47.33 ± 68.06 6.38 ± 1.25 52
13 11/2 38.31 ± 13.93 52.65 ± 45.27 5.85 ± 1.03 57
ATT = anti-tuberculosis treatment, EDSS = Extended Disability Status Scale score.
1374
Y.-q. Feng et al. / Journal of Clinical Neuroscience 17 (2010) 1372–1377
Table 2 Analysis of the Extended Disability Status Scale scores and relapse changes in the patients treated for Devic’s neuromyelitis optica with anti-tuberculosis chemotherapy and control groups Variable
Treatment group
Control group
p-value
Mean baseline EDSS 1-year EDSS 2-year EDSS Relapses during ATT
6.38 ± 1.25 2.50 ± 1.52 2.21 ± 1.32 8à
5.85 ± 1.03 7.35 ± 1.66 8.15 ± 1.68 15
>0.05 <0.01 <0.01 <0.01
ATT = anti-tuberculosis treatment, EDSS = Extended Disability Status Scale score. Statistically significant (independent Student’s t-test). à Statistically significant (chi-squared test).
and control groups using the Student’s t-test and chi-squared tests. No significant differences between the treatment and control groups were apparent. 3.2. Control group During the 2-year study period, five patients in the control group died. Three of these patients died from respiratory failure,
one patient died suddenly from a presumed pulmonary embolus and one patient died of pneumonia. The EDSS score progressively and significantly (p < 0.01) increased from 5.85 ± 1.03 at Y0, to 7.35 ± 1.66 at Y1 and 8.15 ± 1.68 at Y2 (Table 2). These patients had 15 relapses in total. MRI did not show any decrease on the T2-weighted and FLAIR lesion load during the 2-year follow-up.
3.3. Treatment group In the treatment group, we withdrew fully corticosteroid and immunosuppressive treatments before anti-tuberculosis chemotherapy. The response occurred early in the course of treatment for all patients. Clinical worsening was halted during the 2-week course of anti-tuberculosis chemotherapy. For three patients, a dramatic improvement in limb weakness was observed within 1– 3 days of commencing treatment. Clinical outcomes were evaluated after 1 and 2 years and significantly decreased EDSS scores were observed at Y1 (2.50 ± 1.52; Y2 vs. Y0 and Y2 vs. control, p < 0.01) and at Y2 (2.21 ± 1.32; Y2 vs. Y0 and Y2 vs. control, p < 0.01). These improved clinical outcomes were sustained and continued to improve. All patients felt their motor and sphincter
Fig. 1. MRI changes in the spinal cord of patient 1 diagnosed with neuromyelitis optica (NMO) and treated with anti-tuberculosis treatment (ATT). (A) Sagittal T1-weighted MRI showing thickening of the cord from the inferior medulla to the T2 level, with subtle intraparenchymal hyperintensity. (B) Sagittal T2-weighted MRI showing a contiguous area of increased signal intensity. (C) Axial T2-weighted MRI showing the lesion to involve the whole cord. (D) Sagittal T1-weighted MRI after 1 year of treatment, showing the swelling of the cervical spinal cord and patchy abnormal signal have resolved. (E) Sagittal T2-weighted MRI showing a small area of abnormal signal remains in the central cervical cord. (F) Axial T2-weighted MRI showing a small lesion is visible at the C6 level. (G) Sagittal T2-weighted MRI after 2 years of treatment showing a small lesion remains. (H) Axial T2-weighted cord MRI after 2 years of treatment.
Y.-q. Feng et al. / Journal of Clinical Neuroscience 17 (2010) 1372–1377
dysfunction symptoms markedly improve and also noticed improvement in their sensory symptoms. Visual acuity gradually ameliorated in all patients after 2 years of treatment. During the 24-month trial, six of the 12 patients in the treatment group experienced eight relapses in total. There was a lower rate of relapse incidence than for the control group (chi-squared test, p < 0.05). Six relapses occurred in the first year, particularly during the first 3 months of treatment (four relapses). Two patients experienced relapses of visual decline as their limb weakness improved. Without changing our regimen, these relapses gradually recovered as treatment continued. Overall, throughout the 2 years of treatment, exacerbation rates gradually declined. ATT significantly limited the occurrence of relapses in patients suffering from steroid-refractory NMO. 3.4. Efficacy as measured by MRI 3.4.1. Illustrative patients treated with ATT During the study period, MRI demonstrated that abnormal signal and expansion of the spinal cord progressively decreased in all patients after the initiation of ATT. Signal abnormalities on T2-weighted MRI were markedly decreased in size and prominence and had virtually resolved by the end of treatment. Additionally, regions showing well-defined enhancing signals, and poorly defined marginal enhancement in pre-treatment studies, resolved with treatment. Figs. 1–3 show the MRI changes in the spinal cord of patients 1, 4 and 5. Patient 1 was a 24-year-old female who had experienced an acute attack of bilateral lower extremity weakness and bladder incontinence, and fully recovered after 15 days’ treatment with dexamethasone. Two episodes of relapse occurred during the following 2 months. On admission, visual acuity was reduced in both eyes to perception of light; quadriparesis, voiding difficulties and sensory loss had also occurred. She was treated with pulsed doses of methylprednisolone and intravenous immunoglobulin but did not improve. After ATT her symptoms gradually resolved (Fig. 1). Patient 4 was a 34-year-old female who developed bilateral ON and fully recovered after 1 month of steroid treatment. She later experienced two relapses of back pain, bilateral lower extremity parasthesia and weakness. Partial recoveries were achieved by pulsed methylprednisolone and iv Ig treatments. Three years after
1375
onset, she presented with quadriplegia and voiding difficulty, but the same medication did not take effect. She underwent ATT, and her condition gradually recovered, although two relapses still occurred during the first year of ATT (Fig. 2). Patient 5 was a 33year-old female, who had a history of ON 8 months before, and who developed paraplegia that quickly progressed. Pulse methylprednisolone, intravenous immunoglobulin and azathioprine were initiated with no resolution and she continued to deteriorate. Her cerebrospinal fluid was normal. Following complete paralysis and sensory loss below the T4 level, our treatment regimen commenced. After 1 month of chemotherapy, the patient recovered sphincter function with grade 2 muscle strength in the lower limbs and sensory deficit greatly improved. Six months later, she was able to walk without a cane. Full sensation returned and abnormal MRI signal had completely disappeared after 1 year of ATT treatment. The Extended Disability Status Scale (EDSS) score was 2.0. In the second year, the EDSS score was stable (Fig. 3). 3.4.2. Patients in the control group In the control group, NMO cord lesions were usually persistent or became more severe and extensive over time; three patients developed focal spinal cord atrophy, as detected in the follow-up study. During the observational period, a significant decrease in the intramedullary lesion burden was not detected in the surviving control group patients.
4. Discussion This study shows that anti-tuberculosis chemotherapy can benefit patients with steroid-refractory NMO, suggesting that tuberculosis infection correlates closely with NMO. This study was undertaken in southern China, where tuberculosis infection is epidemic with a prevalence of 44.5%.7 The study was based on trial antituberculosis chemotherapy; steroids and other immunosuppressing drugs were not used. The findings show that ATT is effective and adequate to treat NMO. The effects were associated with significantly improved neurological status and changes in the spinal cord were observed by MRI. The positive results obtained for mycobacterial DNA in CSF with N-PCR revealed that pathogenesis involves direct CNS infection.
Fig. 2. MRI changes in the spinal cord of patient 4 diagnosed with neuromyelitis optica (NMO) and treated with anti-tuberculosis treatment (ATT). (A) Sagittal T2-weighted MRI showing a long segment, high intensity signal abnormality from C1 to T5 in the spinal cord before our treatment regimen commenced. (B) After 5 months’ treatment, the lesion was greatly decreased. (C) By the end of the first year, only traces of the lesion remained. (D) After 2 years’ treatment, all lesions had disappeared.
1376
Y.-q. Feng et al. / Journal of Clinical Neuroscience 17 (2010) 1372–1377
Fig. 3. MRI changes in the spinal cord of patient 5 diagnosed with neuromyelitis optica (NMO) and treated with anti-tuberculosis treatment (ATT). (A) Sagittal T2weighted MRI showing a contiguous area of increased signal intensity spanning from C4 to T4. (B,C) Axial T2-weighted MRI showing the high signal involved the whole cord with cord. (D) Sagittal T2-weighted MRI after 1 year of treatment, showing a normal spinal cord. (E,F) Axial T2-weighted MRI, showing no lesions.
M. tuberculosis infection can involve the spinal cord and optic nerve. In three reported patients with acute myelopathy associated with pulmonary tuberculosis who underwent post-mortem examination,8 myelopathy had been progressive or relapsing, and demyelination of the white and grey matter was found in the spinal cord. A similar report suggested that clinical improvement can be achieved after steroid therapy.9 Gupta et al. studied MRIs of 20 patients with intraspinal tuberculosis, and revealed that cord edema, cavitation and linear enhancement of the surface of the spinal cord were common; these features are also found in Devic’s disease.10 Some reports suggest a correlation between pulmonary tuberculosis and NMO. El Otmani et al. reported a patient with NMO with evolving pulmonary tuberculosis. Following ATT, the patient recovered normal mobility and sphincter control.1 Silber et al. reviewed
eight patients with NMO with pulmonary tuberculosis.3 Two of these patients died; the cause was unknown in one patient, the other died from extensive pulmonary tuberculosis. One patient’s condition did not improve during the 5-month observation period but some degree of improvement occurred in the remaining five patients, ranging from improved arm strength to full recovery of leg power. The authors proposed this syndrome is most likely due to an immune reaction to tuberculosis. It is possible that ATT, initiated to treat pulmonary tuberculosis, simultaneously cured NMO because of their similar etiology. Our patients recovered more completely than those in the study by Siber et al.; possibly this was because patients included in our study had not progressed to an advanced stage of infection. Our study challenges the classic view of NMO as a subtype of multiple sclerosis, and suggests that tuberculosis infection is one of the most important causes of NMO and might be involved in the pathogenesis. Specific associations with viral diseases such as varicella,11 human immunodeficiency virus–1,12 and hepatitis virus A13 have been reported. Many studies also show that NMO has striking neuropathological features that are not attributable solely to demyelination, indicating a pathological mechanism distinct from that observed in patients with multiple sclerosis. For example, Lucchinetti et al. performed a detailed analysis of the autopsy of nine patients with Devic’s disease.14 The lesions were severely destructive with cavitation, necrosis and acute axonal pathology, in both gray and white matter. Inflammatory infiltrates were characterized by extensive macrophage infiltration associated with large numbers of perivascular granulocytes and eosinophils. Similarly, Beck16 (1927), Hassin17 (1937), and Lowenberg18 (1947) all described involvement of both gray and white matter of the spinal cord, marked inflammatory infiltrates and destruction of the spinal cord extending continuously through multiple segments. These changes were thought to be distinct from those associated with multiple sclerosis. The findings from our study suggest that direct infection could explain these common findings in NMO. Our study also analyzed the occurrence of relapses in patients with NMO. A decrease in the relapse rate correlated with a decrease in EDSS after 2 years, suggesting that a relapse represents real lasting efficacy, rather than a reflection of disease progression. There was a tendency towards reduction in the relapse rate as treatment continued. Four relapses occurred during the first 3 months of treatment, and relapses tended to occur in patients with a long history of disease and severe demyelination. The mechanism of relapse requires further investigation but it may be similar to the paradoxical deterioration during tuberculosis infection. Relapses are not restricted to patients with demyelinating disease; they can occur several times in patients with tuberculous myelopathy during treatment.15 Methodological limitations could impact the findings of this study. Overall, the patient groups showed comparable clinical and MRI parameters at the start of the study. However, the number of patients in this trial was small and the inevitable risk of selection bias exists. Also, relapses are less well defined in patients with steroid-refractory NMO. Further studies of the efficacy of antituberculosis chemotherapy would determine if this treatment regimen is effective for relapsing–remitting NMO. An additional study of a longer duration and larger sample size is required for steroidrefractory NMO.
5. Conclusion This small, controlled study shows the long-term clinical efficacy of ATT in patients with steroid-refractory NMO. This treatment can reduce disease activity and halt progression, and can also result in a significant recovery of neurological deficits. Considering NMO
Y.-q. Feng et al. / Journal of Clinical Neuroscience 17 (2010) 1372–1377
is a treatable condition, and that a delay in treatment leads to significant morbidity and mortality, we recommend that anti-tuberculosis chemotherapy should be considered for patients with steroidrefractory NMO. References 1. El Otmani H, Rafai MA, Moutaouakil F, et al. Devic’s optic neuromyelitis and pulmonary tuberculosis. Rev Mal Respir 2005;22:143–6. 2. Dixon GJ, Roberts GB, Tyrrell WF. The relationship of neuropathy to the treatment of tuberculosis with isoniazid. Scott Med J 1956;1:350–4. 3. Silber MH, Willcox PA, Bowen RM, et al. Neuromyelitis optica (Devic’s syndrome) and pulmonary tuberculosis. Neurology 1990;40:934–8. 4. Papais-Alvarenga RM, Miranda-Santos CM, Puccioni-Sohler M, et al. Optic neuromyelitis syndrome in Brazilian patients. J Neurol Neurosurg Psychiatry 2002;73:429–35. 5. Wingerchuk DM, Hogancamp WF, O’Brien PC, et al. The clinical course of neuromyelitis optica (Devic’s syndrome). Neurology 1999;53:1107–14. 6. Bhigjee AI, Padayachee R, Paruk H, et al. Diagnosis of tuberculous meningitis: clinical and laboratory parameters. Int J Infect Dis 2007;11:348–54. 7. National Technic Steering Group of the Epidemiological Sampling Survey for Tuberculosis, Duanmu H. Report on fourth national epidemiological sampling survey of tuberculosis. Zhonghua Jie He He Hu Xi Za Zhi 2002;25:3–7. 8. Hughes RA, Mair WG. Acute necrotic myelopathy with pulmonary tuberculosis. Brain 1977;100:223–38.
1377
9. Reid AC, Bone I. Myelopathy during treatment for pulmonary tuberculosis. Postgrad Med J 1980;56:511–2. 10. Gupta RK, Gupta S, Kumar S, et al. MRI in intraspinal tuberculosis. Neuroradiology 1994;36:39–43. 11. Merle H, Smadja D, Cordoba A. Optic neuromyelitis and bilateral acute retinal necrosis due to varicella zoster in a patient with AIDS. J Fr Ophtalmol 1998;21:381–6. 12. Mirsattari SM, Johnston JB, McKenna R, et al. Aboriginals with multiple sclerosis: HLA types and predominance of neuromyelitis optica. Neurology 2001;56:317–23. 13. Jouhadi Z, Ouazzani I, Abid A, et al. Devic’s optic neuromyelitis and viral hepatitis type A. A paediatric case report. Rev Neurol (Paris) 2004;160: 1198–202. 14. Lucchinetti CF, Mandler RN, McGavern D, et al. A role for humoral mechanisms in the pathogenesis of Devic’s neuromyelitis optica. Brain 2002;125:1450–61. 15. White VL, Al-Shahi R, Gamble E, et al. Transverse myelopathy and radiculomyelopathy associated with pulmonary atypical Mycobacterium infections. Thorax 2001;56:158–60. 16. Beck GM. A case of diffuse myelitis associated with optic neuritis. Brain 1927;50:687–703. 17. Hassin GB. Neuroptica myelitis versus multiple sclerosis: a pathologic study [abstract]. Arch Neurol Psychiat (Chicago) 1937;37:1083–99. 18. Lowenberg K, DeJong RN, Foster DB. Neuromyelitis optica: its relations to progressive necrosis of the spinal cord and acute multiple sclerosis [abstract]. Trans Am Neurol Assoc 1947;67:50–66.