Disease activity and treatments. The origin on the time axis represents enrollment in the study, and therefore the axis range 0 to −24 months represents the pre–mitoxantrone hydrochloride dosing period and the range 0 to 24 months represents the study period. Lines extending beyond −24 months indicate patients who had other previous relapses. AZ indicates azathioprine; IFN, interferon; Ig, immunoglobulin; IV, intravenous; and PP, plasmapheresis.
Cumulative doses and timing of mitoxantrone hydrochloride infusions in patients 1 and 2 (A) and patients 3 through 5 (B).
Serial brain and spine magnetic resonance images of patient 1 showing the response to mitoxantrone hydrochloride therapy. Images for each time point show the month and year on top. Gad indicates postgadolinium image; T2, T2-weighted image; and FLAIR, fluid-attenuated inversion-recovery image. A, Images at far left (10/01) were obtained immediately before the start of mitoxantrone therapy. Note the multiple enhancing lesions in the cervical spinal cord. Images obtained after initiation of therapy (1/02) show regression of lesions and no enhancement. However, the 2/02 images show worsening of lesions in the cervical cord and a large diffuse lesion in the thoracic cord. B, Serial brain images show a normal image at baseline but the appearance of 2 lesions in the brain 4 months later.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Weinstock-Guttman B, Ramanathan M, Lincoff N, et al. Study of Mitoxantrone for the Treatment of Recurrent Neuromyelitis Optica (Devic Disease). Arch Neurol. 2006;63(7):957–963. doi:10.1001/archneur.63.7.957
Copyright 2006 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2006
Neuromyelitis optica is a severe demyelinating disease that selectively involves the optic nerves and the spinal cord but usually spares the brain. It is considered to have a B-cell–induced pathogenesis. Mitoxantrone hydrochloride, a synthetic anthracenedione approved for worsening relapsing-remitting multiple sclerosis and secondary progressive multiple sclerosis, has been shown to primarily suppress the humoral response.
To evaluate the benefit of mitoxantrone treatment in patients with relapsing neuromyelitis optica.
Prospective 2-year study.
Academic multiple sclerosis center.
Five patients (3 women and 2 men) with an age range of 20 to 51 years and an Expanded Disability Status Scale score of 2.5 to 6.5 (mean ± SD, 4.40 ± 1.88).
Monthly intravenous infusions of mitoxantrone hydrochloride, 12 mg/m2, for 6 months followed by 3 additional treatments every 3 months.
Main Outcome Measures
Expanded Disability Status Scale score measured every 3 months and during relapses; findings on orbital, brain, and spinal cord magnetic resonance images performed at baseline and at 3, 6, 12, 18, and 24 months; and visual evoked potentials and results of ophthalmologic evaluations performed at baseline and annually.
During the 2 years of treatment, 2 patients each had a relapse once within the initial 5 months of treatment (1 severe and 1 moderate). Improvement was seen clinically and on magnetic resonance images in 4 patients. Patients generally tolerated the treatment well, although 1 patient had a reversible decrease in cardiac ejection fraction.
Our results suggest a beneficial effect of mitoxantrone treatment for relapsing neuromyelitis optica.
clinicaltrials.gov Identifier: NCT00304291
Neuromyelitis optica (NMO), also known as Devic disease, is a severe demyelinating condition that selectively affects the optic nerves and the spinal cord with relative sparing of the brain. The disease is characterized by monophasic or recurrent attacks of severe optic neuritis and/or extensive longitudinal myelitis; the cerebrospinal fluid findings typically include polymorphonuclear pleocytosis without intrathecal immunoglobulin secretion or oligoclonal bands.1 By magnetic resonance (MR) imaging, NMO with recurrent attacks (RNMO) can be distinguished from multiple sclerosis (MS) by its extensive spinal cord involvement (≥3 segments of spinal cord gray and white matter) and a paucity of brain findings.1 Pathologic and serologic data support a B-cell–mediated mechanism that underscores its distinction from classic MS.2,3
Recurrent NMO carries a poor prognosis, and, therefore, early diagnosis and effective therapeutic intervention are warranted. Previous studies have reported partially effective therapies.1,4-6 Mitoxantrone hydrochloride (Novantrone; Serono Inc, Rockland, Mass) is an anthracenedione antineoplastic agent approved by the US Food and Drug Administration as a treatment for secondary progressive and worsening relapsing-remitting MS.7,8 Mitoxantrone potentially suppresses T-helper lymphocytes and the humoral immune system via both macrophage and B-cell attenuation.9
Our first case of RNMO treated with mitoxantrone was in a 50-year-old white woman in whom 9 months of treatment with glatiramer acetate followed by a combination of glatiramer with azathioprine had previously failed. After 7 treatments with mitoxantrone hydrochloride, 12 mg/m2, given every 3 months (total, 84 mg/m2), the patient's Expanded Disability Status Scale (EDSS) score improved from 7 to 4 and remained stable during 3 years of follow-up with maintenance azathioprine (200 mg/d) monotherapy. On the basis of this experience and the proven effect of mitoxantrone treatment in MS, we tested the effect of mitoxantrone in patients with RNMO.
In this open-label pilot study of 5 patients with RNMO, the inclusion criteria required recurrent longitudinal myelitis (≥3 segments of spinal cord involvement on MR images) with or without recurrent optic neuritis (unilateral or bilateral) but with normal brain MR images. Patients with recurrent longitudinally extensive myelitis without optic neuritis have an underlying abnormality and serologic findings similar to those of NMO, and it is appropriate to consider this a form of NMO.10 Cerebrospinal fluid required no intrathecal IgG synthesis or oligoclonal bands. Age was required to be 18 to 55 years, with EDSS score of 7 or less.
Exclusion criteria were any of the following: cardiac risk factors (eg, history of congestive heart failure and left ventricular ejection fraction [LVEF] <50%); systemic diseases such as systemic lupus erythematosus, Sjögren syndrome, antiphospholipid antibody syndrome, sarcoidosis, rheumatoid arthritis, or vitamin B12 deficiency; and previous treatment with mitoxantrone or anthracyclines.
Initially, the treatment consisted of a 3-cycle induction period of monthly intravenous (IV) infusions each consisting of mitoxantrone, 12 mg/m2 (not exceeding 20 mg), with IV methylprednisolone sodium succinate, 1000 mg, and prophylactic ondansetron hydrochloride (Zofran; GlaxoSmithKline, Philadelphia, Pa), 24 mg. Three months after completing the last induction dose, the patients were administered mitoxantrone hydrochloride, 12 mg/m2, with the same combination every 3 months for up to 2 years or a maximum dose of 100 mg/m2. We refer to this protocol as “3QM.”
However, the induction protocol was modified to 6 cycles of monthly mitoxantrone hydrochloride (the dose was 12 mg/m2 but did not exceed 20 mg) because the first 2 patients had relapses after 5 months (during the first 2 months of the longer interval after 3-monthly mitoxantrone infusions). As before, 3 months after the last monthly mitoxantrone treatment, the patients were subsequently given the same combination of drugs every 3 months (mitoxantrone hydrochloride dose of 12 mg/m2) for up to 2 years or a maximum dose of 100 mg/m2. This protocol is referred to as “6QM.”
Patients were assessed clinically by EDSS score every 3 months and during relapses. Brain, optic nerve, and spinal cord MR imaging was performed at baseline and every 3 to 6 months until the end of the 2-year study. Ophthalmologic evaluation, echocardiography, chest x-ray, and visual evoked potential (VEP) testing were performed at 12 and 24 months.
The LVEF was evaluated by a multiple gated acquisition cardiac scan at baseline and before the cumulative dose of 100 mg/m2. Complete blood cell count with differential cell count, platelet count, chemistry studies, and urinalysis were performed before each treatment and 10 days after treatment. Study discontinuation was defined a priori as 3 patients reaching one of the following treatment failure criteria: (1) an acute severe attack leading to blindness and/or paraplegia; (2) 2 moderate relapses in a 12-month period (increase in EDSS score of ≥1 point in patients with an EDSS score of ≤5.5 or of ≥0.5 point if the EDSS score was ≥6); (3) 4 minor attacks (change in functional systems without affecting the final EDSS score); or (4) progression of disability, defined as a 1-point increase in EDSS score if the score was 5.5 or less or an increase of 0.5 point if the EDSS score was 6 or greater, sustained for 6 months.
Patients were examined during relapses (within 7 days from onset of symptoms). If the relapse consisted of optic neuritis, an ophthalmologic evaluation was required.
Three months after reaching the maximum mitoxantrone hydrochloride dose of 100 mg/m2, patients were treated with azathioprine and IV methylprednisolone until the end of the 24-month study period. The starting azathioprine dose was 50 mg twice daily for 1 month and was increased to 150 mg once daily after a follow-up visit at the end of 1 month, at which time laboratory tests for liver function and complete blood cell count were assessed. The methylprednisolone dosing regimen was 3 days of 1000 mg IV every 2 months.
The MR imaging of the brain, optic nerves, and spinal cord was performed by means of the same scanning pulse sequences, protocol, and 1.5-T MR imaging platform. The MR imaging of the orbits included 3-mm axial and coronal T2-weighted imaging, with precontrast and postcontrast fat-suppressed T1-weighted images. The MR imaging of the spine included cervical and thoracic sagittal and axial 3-mm fast spin-echo T2-weighted images and precontrast and postcontrast sagittal and axial T1-weighted images using gadopentetate dimeglumine (0.1 mmol/kg IV). The MR imaging of the brain included T1-weighted, T2-weighted, and fluid-attenuated inversion-recovery 5-mm axial images to rule out intracranial lesions.
Review of MR images, performed by an experienced observer (R.B.), included qualitative analysis of T2 hyperintense lesions and the presence of enhancement of the optic nerves and spinal cord.
Five patients were enrolled in this study (Table 1). The number of mitoxantrone infusions per patient ranged between 3 and 10 (mean, 7.6), and follow-up was 2 years. Figure 1 shows the disease activity and timing of treatments for the preceding 2 years and during the trial.
Mitoxantrone hydrochloride doses normalized by body surface area (in milligrams per square meter) and timing of infusions are shown in Figure 2. The cumulative total dose in 3 patients (patients 2, 3, and 5) was 96 mg/m2 to 104 mg/m2; patient 4 had reached a cumulative dose of 86 mg/m2 when treatment was discontinued because of decreased LVEF in the 1-year scheduled multiple gated acquisition cardiac scan (see the “Toxic Reactions and Adverse Effects” section).
Two patients experienced relapses. Patient 1 (Figure 3) completed the 3QM induction protocol. Four months later, after initially experiencing substantial clinical benefit, he had a severe relapse consisting of a cervical myelopathy with quadriplegia (treatment failure). After 5 courses of plasmapheresis, he improved partially and was able to stand and take a few steps. Unfortunately, this patient died 2 months later of pulmonary embolism considered unrelated to mitoxantrone treatment. Patient 2 had a moderate recurrent sensory myelopathy after the initial 3 mitoxantrone induction treatments (at month 5).
After these events, the induction protocol was changed to monthly mitoxantrone infusions for 6 months (the 6QM protocol). All remaining patients, including patient 2, were treated according to 6QM. Their conditions steadily improved and remained stable per clinical and MR imaging measures.
There was a trend toward improvements in EDSS score, which decreased from a mean ± SD of 4.40 ± 1.88 at baseline to 2.25 ± 0.65 at 24 months. We present a detailed descriptive summary of results in lieu of the repeated-measures statistical analysis (Table 2) because of the limitations of missing data in case 1 and the small sample size.
No additional relapses occurred in all 4 patients after switching to the 6QM protocol for the remainder of the 2-year study period. With the inclusion of the 2 treatment relapses after mitoxantrone treatment on the 3QM protocol, the overall treatment regimen led to a decrease in relapse rate: the mean ± SD number of relapses per patient was 2.4 ± 0.89 in the 2 years preceding the mitoxantrone regimen compared with 0.4 ± 0.55 in the 2 years of the study. Two patients who presented with clinical optic neuritis also had abnormal VEPs and optic disc pallor. Patient 1 had severe optic disc pallor and severely abnormal VEP (P100 latency not reproducible) before mitoxantrone treatment; no post–mitoxantrone treatment data were available on this individual. Patient 2 had bilaterally abnormal VEP (P100 latency values: left eye, 147 milliseconds; right eye, 112 milliseconds; normal, ≤110 milliseconds) before mitoxantrone treatment; at 1 year, the VEP showed improvement (P100 latency values: left eye, 110 milliseconds; right eye, 107 milliseconds). The results of neuro-ophthalmologic tests and VEPs for patient 2 remained stable at 2 years. The remaining patients had normal VEPs and results of neuro-ophthalmologic examinations.
The overall treatment regimen consisted of mitoxantrone with IV methylprednisolone for 18 months followed by maintenance therapy with oral azathioprine; it is, therefore, difficult to separate the contribution of mitoxantrone from that of the combination therapy. However, we have highlighted the role of mitoxantrone as an induction therapy because our patients previously received multiple doses of IV methylprednisolone without therapeutic benefit, and it was only after the addition of mitoxantrone to IV methylprednisolone that we observed disease stabilization or improvement.
The MR imaging results are given in Table 3 and shown in Figure 3. Orbital MR images were normal in all patients throughout the study (no lesions). Brain MR images were normal in all patients at baseline. Spinal MR images were characterized by diffuse or multifocal intramedullary lesions involving the cervical or thoracic spinal cord, with or without enhancement and cord swelling (Table 3, Figure 3). The lesions were large, extending over multiple spinal segments and often associated with hypointensity on T1-weighted images (not shown).
After mitoxantrone treatment, spinal lesions on T2-weighted MR images regressed substantially in 3 patients and slightly regressed in 1 patient (Table 3 and Figure 3). Patients 1 and 2 developed new spinal lesions at the time of a clinical relapse (Figure 3). Only patient 1 developed brain lesions during follow-up at the time of a severe relapse (Figure 3). None of the other patients developed new brain, optic nerve, or spinal lesions on MR images.
The therapy was generally well tolerated. Leukopenia at 10 days after therapy was transient, with recovery within 3 weeks (mean ± SD pretreatment leukocyte values, 5560 ± 1820/μL; range, 3700/μL to 11 300/μL; compared with posttreatment mean ± SD of 1640 ± 300/μL; range, 1100-2100/μL). Only 1 patient (case 5) with recurrent urinary tract infections was started on a regimen of preventive antibiotic therapy (4 days) at the time of mitoxantrone infusions. No other infections were noted during the study. One patient developed subclinical cardiac insufficiency (case 4): the LVEF decreased from 60% at baseline to 44% by multiple gated acquisition cardiac scan after 7 cycles of mitoxantrone hydrochloride (total dose, 86 mg/m2). Mitoxantrone treatment was discontinued in this patient and she was followed up every 3 months; her LVEF had recovered to 56% by 9 months after mitoxantrone discontinuation.
In this pilot study, we showed that treatment with mitoxantrone may help stabilize RNMO. Four of 5 patients in the study responded to therapy. One patient we previously treated also responded to mitoxantrone. Although our initial protocol consisted of the 3QM regimen, this was changed to the 6QM regimen after 2 initially treated patients had a relapse (after 4 and 5 months). The mitoxantrone treatment regimen described by Edan et al8 for aggressive relapsing MS is similar and consists of monthly infusions for 6 months.
The diagnosis of NMO carries a poor prognosis. The overall mortality is 35% to 50% within 5 years, with a mortality of 20% in the acute stages.1 Wingerchuk et al1 reported that most patients with RNMO become nonambulatory and blind in at least one eye, and one third in their study died of respiratory failure secondary to cervical spinal cord involvement. Therefore, the mitoxantrone regimen investigated in this study may be useful for this debilitating disease.
Patient 1, who was African American, had an unusual, severe presentation and eventually died of a pulmonary embolism. The severe myelitis and attendant immobility that can occur in NMO require a greater awareness of the increased risk of deep vein thrombosis and pulmonary embolism. The course of MS is more severe in African Americans than white persons,11 but it is not known whether NMO outcomes are likewise more severe.
The therapeutic experience with NMO is limited to small retrospective case studies.1,4 A small prospective study suggested benefit from long-term oral corticosteroids in combination with azathioprine therapy for 18 months.4 However, Wingerchuk et al1 reported less benefit with azathioprine, with only 3 of 9 patients showing a reduction in relapses. The therapeutic benefit related to azathioprine is usually delayed up to 6 months, which may explain the insufficient response in some patients. Intravenous methotrexate (50 mg weekly) in combination with oral prednisone was reported to be effective in a small case series of NMO.5 Plasmapheresis is helpful in the recovery from acute attacks but does not prevent further relapses of NMO.1 Rituximab, an anti-CD20 monoclonal antibody that targets B cells, was shown to decrease the relapse rate and improve EDSS scores in 8 patients with NMO.6
Anecdotal experience suggests that interferon beta therapy, proven beneficial for relapsing MS, is ineffective for NMO.12 However, a randomized study of interferon beta therapy in Japanese patients with MS reported that the therapeutic benefit of high-dose interferon beta therapy did not differ between opticospinal and classic MS subsets.13 The results require cautious interpretation because the trial was not powered to reliably evaluate effects according to MS phenotype.12
Our small study suggests a beneficial effect of mitoxantrone for the treatment of RNMO. Close monitoring for preventing mitoxantrone adverse effects such as cardiac toxic reactions by performing regular cardiac evaluation before each mitoxantrone treatment may increase the therapeutic safety profile of this therapy. A larger randomized prospective study of mitoxantrone for NMO seems warranted.
Correspondence: Bianca Weinstock-Guttman, MD, The William C. Baird Multiple Sclerosis Center, Jacobs Neurological Institute, 100 High St, Buffalo, NY 14203 (BGuttman@TheJNI.org).
Accepted for Publication: March 15, 2006.
Author Contributions:Study concept and design: Weinstock-Guttman. Acquisition of data: Weinstock-Guttman, Lincoff, Napoli, Feichter, and Bakshi. Analysis and interpretation of data: Weinstock-Guttman, Ramanathan, Napoli, Sharma, and Bakshi. Drafting of the manuscript: Weinstock-Guttman, Ramanathan, Sharma, and Feichter. Critical revision of the manuscript for important intellectual content: Ramanathan, Lincoff, Napoli, and Bakshi. Statistical analysis: Weinstock-Guttman and Ramanathan. Obtained funding: Weinstock-Guttman. Administrative, technical, and material support: Weinstock-Guttman, Napoli, Sharma, Feichter, and Bakshi. Study supervision: Weinstock-Guttman, Lincoff, Napoli, and Bakshi.
Financial Disclosure: Dr Weinstock-Guttman has received honoraria for speaking and consulting from Teva Neuroscience and Biogen Idec. Dr Napoli has received honoraria for speaking and consulting from Biogen Idec, Serono, and Teva Neuroscience. Dr Bakshi has received honoraria for speaking, consulting fee, or research support from Berlex, Biogen Idec, Serono, and Teva Neuroscience.
Funding/Support: This study was supported in part by Serono and by The William C. Baird Multiple Sclerosis Center and a Sylvia Lawry grant from the National Multiple Sclerosis Society (Dr Napoli).