Article Text

Download PDFPDF

Mitoxantrone prior to interferon beta-1b in aggressive relapsing multiple sclerosis: a 3-year randomised trial
  1. G Edan1,2,3,
  2. G Comi4,5,
  3. E Le Page1,2,
  4. E Leray2,3,
  5. M A Rocca5,6,
  6. M Filippi5,6,
  7. for The French–Italian Mitoxantrone Interferon-beta-1b Trial Group*
  1. 1Department of Neurology, University Hôpital Pontchaillou, Centre Hospitalier Universitaire, Service de Neurologie, Rennes, France
  2. 2Institut National de la Santé et de la Recherche Médicale, CIC 0203, Hôpital Pontchaillou, Rennes, France
  3. 3Université Européenne de Bretagne, Université de Rennes 1, Faculté de médecine, Rennes, France
  4. 4Department of Neurology, University Hospital San Raffaele, Milan, Italy
  5. 5Division of Neuroscience, University Vita-Salute, Institute of Experimental Neurology, Milan, Italy
  6. 6Neuroimaging Research Unit, University Hospital San Raffaele, Milan, Italy
  1. Correspondence to Professor Gilles Edan, Service de Neurologie, CHU Pontchaillou, Rennes cedex 35033, France; gilles.edan{at}


Objectives The long-term impact of interferon-beta-1b (IFN) might be improved by short-term immunosuppression with mitoxantrone (MITOX) in aggressive relapsing-remitting multiple sclerosis (ARMS) patients.

Methods In this 3-year clinical and MRI study, 109 ARMS patients (two or more relapses in the previous 12 months and one or more gadolinium (Gd)-enhancing MRI lesion) were randomised into two groups: 54 patients received MITOX monthly (12 mg/m2; maximum 20 mg) combined with 1 g of methylprednisolone (MP) for 6 months followed by IFN for the last 27 months, and 55 patients received IFN for 3 years combined with 1 g of MP monthly for the first 6 months. The primary endpoint was the time to worsen by at least one Expanded Disability Status Scale point confirmed at 3 months.

Results The time to worsen by at least one Expanded Disability Status Scale point confirmed at 3 months was delayed by 18 months in the MITOX group compared with the IFN group (p<0.012). The 3-year risk of worsening disability was reduced by 65% in the MITOX group relative to the IFN group (11.8% vs 33.6%). MITOX patients had a reduced relapse rate by 61.7%, a reduced number of Gd-enhancing lesions at month 9 and a slower accumulation of new T2 lesions at each time point.

Conclusions Although there were limitations in this investigator–academic-driven study, the data do suggest that mitoxantrone induction therapy prior to INF beta-1b may have a role in aggressive disease.

  • Aggressive relapsing-remitting multiple sclerosis
  • mitoxantrone
  • interferon beta-1b
  • disability
  • MRI
  • multiple sclerosis
View Full Text

Statistics from


Two types of disease-modifying drugs, immunomodulatory and immunosuppressive, have been approved to treat multiple sclerosis (MS). The immunomodulatory drugs include interferon-beta (IFN)1–3 and glatiramer acetate.4 The immunosuppressive drugs include mitoxantrone (MITOX)5–10 and natalizumab.7 Modern therapeutic strategies may be based on: (1) escalating treatment,8 the most commonly recommended, in which the treatment starts with an immunomodulatory agent followed by a more potent, but also more toxic, drug in non-responding patients; and (2) induction treatment aimed at halting focal inflammation more efficiently, followed by an immunomodulatory agent to prevent further inflammatory events.

In recent years, the concept of induction treatment followed by a long-term maintenance treatment has been considered for treating autoimmune diseases, such as MS.9 In a randomised study11 of 40 relapsing MS patients, 3-monthly infusion of mitoxantrone followed by 12 months of daily glatiramer was more effective on MRI activity than daily glatiramer. This approach seems particularly appropriate for patients with aggressive disease, defined as at least two relapses with incomplete recovery in the previous 12 months, associated with gadolinium (Gd)-enhancing lesions on MRI scans.6 Although the promise of such an induction-maintenance regimen has not been fully investigated in large, randomised, controlled trials, several observational studies have suggested that they can control disease activity in patients with a particularly aggressive disease.12 13 These data led us to design this randomised trial to assess the value of such an induction strategy in aggressive relapsing-remitting MS (ARMS) patients.


Study design

This study was a randomised, two-arm controlled study conducted in 21 centres in France and Italy. Patients were randomly assigned by the trial randomisation code in the study coordination centre in a 1/1 ratio to receive MITOX (at a dose of 12 mg/m2 with a maximum of 20 mg) combined with methylprednisolone (1 g) intravenously (as in our pivotal study6) monthly for 6 months, followed 3 months later by IFN-beta-1b (250 μg subcutaneously every other day) for 27 months (MITOX group), or IFN-beta-1b (250 μg every other day) for 36 months, combined with methylprednisolone (1 g) intravenously, monthly during the first 6 months (IFN group) (figure 1). Enrolment was limited to clinically definite MS patients,14 18 to 45 years of age, and an ARMS (ie, two or more relapses in the 12 preceding months or Expanded Disability Status Scale (EDSS) increased by two or more points (unconfirmed at 3 months, but assessed outside a relapse) and one or more Gd-enhancing lesion on MRI assessed by the local radiologist. Patients also had to have a significant disability, defined as an EDSS between 2.5 and 5.5. Exclusion criteria were: pregnancy and breastfeeding, non-use of effective contraception, previous global immunosuppressive treatment with MITOX, cyclophosphamide or total lymphoid irradiation, azathioprine during the preceding 3 months (previous treatment with IFN-beta-1a or 1b and by glatiramer acetate allowed), relapse or treatment with high doses of corticosteroids in the 30 days preceding inclusion, presence of severe psychiatric disorders not controlled by an appropriate treatment, cardiac disease (echocardiogram had to be normal with a left ventricular ejection fraction (LVEF) higher than 50% at screening) and history of a malignant disease.

Figure 1

Study design. EDSS, Expanded Disability Status Scale; Gd, gadolinium; IFN, interferon-beta-1b; MITOX, mitoxantrone; MP, methylprednisolone.

Standard protocol approvals, registrations and patient consents

The study protocol was approved by the regional Institutional Review Board (CCPRB) and the Ethics Committees of each country. Written informed consent was obtained from each patient participating in the study. Participants and patients did not receive any special remuneration for this study.

The present study was registered to the Clinical Trials registry ( with the number NCT00219908.

Study outcomes and safety assessment

Primary outcomes

EDSS scores were evaluated quarterly from randomisation to month 36, in a blinded fashion. Since the infusion-related adverse events associated with MITOX and the flu-like syndrome associated with IFN precluded the possibility of double blinding, in each centre we used a local evaluating neurologist who was blind to treatment assignments for clinical outcomes. The primary measure of efficacy was the time to sustained accumulation of disability equal to or greater than one EDSS point over the 3 years of the study. The progression had to be confirmed 3 months later and to be sustained up to the last evaluation of the 3-year period of study and not to be associated with a relapse

Secondary outcomes

Secondary outcomes were the frequency of clinical relapses and the number of focal MRI lesions. Relapses were defined as new or worsening symptoms that lasted for at least 48 h,14 as confirmed by the blinded neurologist. The frequency of relapses was assessed using the percentage of relapse-free patients, the mean annualised relapse rate per patient (number of relapses divided by the time under the study drug for each patient), the mean annualised relapse rate per group (number of relapses observed in each group divided by the cumulative time under the study drug in each group) and the time to the first relapse after inclusion.

Among the 21 centres participating in the study, 13 agreed to an MRI follow-up evaluation with MRI scans at screening and at 9, 24 and 36 months. MRI scans included conventional or fast spin echo sequences to obtain proton-density and T2-weighted images and conventional spin echo T1-weighted images with the same scan geometry, 5 min after the injection of 0.1 mmol/kg of Gd. The slices were positioned to run parallel to a line that joins the most inferoanterior and inferoposterior parts of the corpus callosum. All MRI scanners were operating at a field strength of 1.5 T. On follow-up scans, patients were carefully repositioned according to published guidelines.15

MRI outcomes included the number of new T2 lesions, number of Gd-enhancing lesions and number of new T1-hypointense lesions (black holes) at each of the time points, interpreted at the MRI reference centre (Neuroimaging Research Unit, Milan, Italy), where the observer was blind to the study group. Safety was assessed quarterly by the treating neurologist who was aware of the study-group assignment.

Safety assessment

An echocardiogram was performed at screening and at months 9, 24 and 36. Blood tests were performed 10 days after and just before each MITOX infusion in order to adjust the dose. At the end of the 6-month course, white-blood-cell counts were measured quarterly up to the end of the study.

Statistical analysis

The study was designed to achieve an 80% probability (β=0.20) of demonstrating a difference (one-way test; α=0.05) expressed by a 40% to 20% reduction in disease progression of at least one EDSS point in the MITOX group compared with the IFN group. A total of 124 patients needed to be included (ie, 62 patients in each arm) to reach this goal.

A statistical analysis was performed among all patients who received the study drug.

The time to sustained disability was compared between the two treatment groups using a discrete time survival analysis with a complementary log–log function, to take into account the interval-censored data in the study. Patients who discontinued the trial before reaching the primary endpoint were considered as right-censored data and were included during their follow-up period, that is from the baseline visit to the last visit. The number of patients who would need to be treated with MITOX instead of IFN to prevent one patient from having sustained a disability was calculated according to the proportion of patients who did not have this outcome at the end of the study.16

The significance level was 0.05, and a statistical analysis was performed using SAS version 9.1.


One hundred and twenty-three patients were initially randomised, and of these, 14 patients withdrew consent prior to treatment. The first patient was randomised in June 1999, and the last follow-up period ended in July 2006. A total of 109 patients were enrolled (92 completed year 1, 70 year completed 2, and 61 year completed 3; figure 2). Five out the 21 centres had fewer than three patients entered into the trial. In the IFN group, 26/54 (47%) patients discontinued the study. In the MITOX group, all patients completed the induction phase. However, 22/55 (40%) patients discontinued the study (during IFN treatment). The mean study drug exposure was 29 (±10) months for the MITOX group and 24 (±13) months for the IFN group (p<0.034).

Figure 2

Flow diagram. IFN, interferon-beta-1b; MITOX, mitoxantrone.

Baseline demographic and clinical characteristics were similar in the two treatment groups (table 1).

Table 1

Baseline characteristics of the two study groups

Twenty-one patients (38%) in the MITOX group and 13 patients (24%) in the IFN group were treated previously by disease-modifying treatment (DMT) during at least 6 months within the 12 prior months (p=0.11). The prior DMT were in the MITOX group 17 IFN (12 beta-1a and 5 beta-1b), three glatiramer acetate, one azathioprine received outside the 3-month preperiod and in the IFN group 13 IFN (11 beta-1a and 2 beta-1b). Baseline characteristics were similar in naïve patients and patients under DMT prior to the randomisation (data not shown). The mean time between the onset of the last relapse and starting of study treatment was 3.4 (SD=2.2) months in the MITOX group and 3.7 (SD=2.2) months in the IFN group (p=0.38). Apart from the screening MRI, 54 patients from 13 centres had an MRI follow-up assessment at month 9, 42 patients at month 24 and 34 patients at month 36. The median number of Gd-enhancing lesions at screening was five and six in the MITOX and IFN groups, respectively (p=0.96).


Primary endpoint: disability

Compared with the IFN group, in the MITOX group the time to ≥1 EDSS point worsening was delayed (p<0.012; figure 3A).The 3-year risk of sustained disability was reduced by 65% (12% vs 34%). MITOX delayed the 12% probability of confirmed worsening disability by 18 months compared with the INF β group. The crude percentage of sustained disability was also reduced (5/55=9.1% vs 14/54=25.9%) (table 2A). The number of patients needed to be treated with MITOX before IFN to avoid one sustained disability event during the 36 months was six. The mean EDSS improved by 0.45 point (±1.19) at the last observation in the MITOX group (from 4.1±1.1 to 3.6±1.8; p<0.007), and remained unchanged in the IFN group (from 3.8±0.9 to 3.7±1.7; p=0.771) (table 2A). Among the MS patients who dropped out in both groups, the mean EDSS at drop-out was 4.25 in the INF group and 3.86 in the Mitox group (p=0.4).

Figure 3

(A) Time to worsen by at least one Expanded Disability Status Scale (EDSS) point (primary endpoint) according to the treatment. (B) Influence of baseline disability level (EDSS≤ or >4) on the primary endpoint. (C) Influence of prior disease-modifying treatment on the primary endpoint. DMT, disease-modifying treatment; IFN, interferon-beta-1b; MITOX, mitoxantrone.

Table 2

Measures of disability, relapse and MRI outcomes during the trial

Secondary endpoints


The proportion of patients who remained relapse-free was increased in the MITOX group (p<0.008) (table 2A). The number of patients needed to be treated with MITOX before IFN to prevent one patient from having any relapse over 36 months (ie, to be relapse free over 36 months) was four. From baseline to the last observation in the MITOX group, the annualised relapse rate per patient was lower (p<0.03) as well as the annualised relapse rate per group (p<0.001). The time to first relapse after treatment institution was delayed in the MITOX group by 21 months compared with the INF β group (p<0.001).


The mean number of new T2 lesions was significantly lower in the MITOX group at each of the time points (table 2B). The mean cumulative number of new T2 lesions over 36 months was lower in the patients of the MITOX group compared with those from the IFN group (p<0.041). Compared with the IFN group, the mean number of Gd-enhancing lesions was lower in the MITOX group at month 9 (p<0.012), and the percentage of patients without any Gd-enhancing lesion was also higher in the MITOX group (p<0.010). There was no significant difference between the two groups in the mean number of Gd-enhancing lesions or in the percentage of inactive MRI scans at months 24 and 36. The mean number of new T1-hypointense lesions (black holes) was not different between the groups (cumulative and at each time point; data not shown).

Posthoc analyses

With the experiences accumulated since the start of this study, it was deemed necessary to examine two parameters carefully: the severity of disability at screening and previous DMT before the start of the study. When considering the level of disability at screening (EDSS≤4 vs EDSS>4) in the primary endpoint assessment, the impact of MITOX remained significant on the one-point EDSS progression (p<0.006). The 3-year risks of sustained worsening were respectively 3.9%, 30.8%, 22.3% and 40.3% in patients with baseline EDSS≤4 in the MITOX and IFN groups, and in patients with EDSS>4 in the MITOX and IFN groups (figure 3B). The crude percentages were respectively 3.1% (1/32), 23.7% (9/38), 17.4% (4/23) and 31.2% (5/16) in those four subgroups.

When considering the two subgroups of patients having or not having previous DMT at screening in the primary endpoint assessment, the impact of MITOX remained significant on the one-point EDSS progression (p<0.015). The 3-year risks of sustained worsening were respectively 0%, 51.7%, 21.0% and 27.9% in patients who were non-responders to DMT before inclusion in the MITOX and IFN groups, and in DMT-naïve patients in the MITOX and IFN groups (figure 3C). The crude percentages were respectively 0% (0/21), 38.5% (5/13), 14.7% (5/34) and 22.0% (9/41) in those four subgroups.

The overall impact of MITOX on the time to worsen by one-point EDSS was significant (p<0.009) after adjusting for disability level at baseline (p=0.09) and prior DMT (p=0.80).


Both groups received IFN and had the common adverse effects related to this drug (flu-like syndrome, injection-site reaction, etc). There were several differences in the safety profiles of the two groups, which were related to MITOX (table 3). In the MITOX group, a higher number of patients had upper-respiratory-tract infections (25 vs 13; p<0.019) (with two patients having fever and severe neutropenia). Mild or moderate leucopenia was observed in 20 patients in the MITOX group compared with nine patients in the IFN group (p<0.020). No clinically symptomatic cardiac event was observed in any of the patients, but two patients in the MITOX group had a transient LVEF<50%. Two out of 36 women had a persistent amenorrhoea; they received MITOX at the age of 35 and 37 years. Finally, mild alopecia was observed in five patients.

Table 3

Patients treated with either mitoxantrone or interferon who experienced adverse events


The justification for including two variables (methylprednisolone and mitoxantrone) instead of ideally one (mitoxantrone alone) came from our pivotal trial.6 In this trial, we compared 6-month mitox+MEP against 6 month MEP alone, allowing the clinical and MRI impact of mitox to be isolated for this treatment period. Demonstrating a dramatic MRI and clinical impact with this combination, we thought it reasonable to keep this well-tolerated combination when using mitoxantrone as an induction therapy. In the present study concerning ARMS patients, the combination of MITOX and methylprednisolone monthly for 6 months followed by IFN-beta-1b 3 months later significantly reduced the risk of sustained accumulation of disability by 64.9% and the annualised relapse rate per patient by 61.7%, as compared with IFN-beta-1b combined for the first 6 months with methylprednisolone. Our study suggests that this induction strategy with MITOX may be more effective in patients with a lower disability: the risk of sustained disability was reduced by 86.9% in patients treated with a low baseline EDSS (≤4). In patients with EDSS>4, some might have already entered into the secondary progressive phase, during which the impact of DMT drugs is likely to have been reduced, in line with the concept of a two-stage disability progression in MS.17 Compared with naïve patients, as expected, the impact of MITOX was stronger in patients who were not responders to DMT (mainly IFN-beta).

Our results are similar to those reported with Alemtuzumab compared with IFN-beta-1a (Rebif 44 μg) over 2 years in RRMS.18 Although comparisons were limited by missing MRI data and the high rate of patient withdrawal from the study when receiving IFN-beta-1b, the impact of MITOX on the focal inflammatory process was evident by the significant reduction in the number of Gd-enhancing lesions at month 9 (82.9% reduction) and the significant reduction in accumulation of new T2 lesions at each time point. However, it is not possible to make any judgement on any additive effect or synergy between the two agents. Indeed, it is not possible to distinguish specifically what might be the consequence of prolonged effect of MITOX (ie, at least 1 year after MITOX stop10 15 19) from what was the therapeutic impact of IFN. A design utilising an arm randomised to MITOX only would be needed to assess the added value of IFN in this combination. Although neutralising INF antibodies were not tested in this study, we could expect only a marginal clinical impact of the 9-month difference of INF treatment duration between the two groups.

Inclusion criteria made the selection difficult; it was the first clinical trial on such aggressive MS patients. Before starting the study drug, 14 patients withdrew consent for various reasons, and the degree of withdrawal was higher in the IFN group than in the MITOX group. Indeed, natalizumab was not available in France at the time of this trial, and putting patients who had already deteriorated on DMT back on this medication may raise some concern. However, the DMT used in this trial was betaferon, which was not the most commonly used drug in the patients previously treated using DMT.

During the 3-year follow-up, we did not observe drop-out during MITOX treatment, but only during IFN treatment. The overall drop-out rate (48/109=44.0%) was similar to the drop-out rate observed in the IFN group of the Alemtuzumab trial18 (41/111=36.9%). This high drop-out rate in both studies might be explained not only by prior IFN treatment before inclusion (one-third of patients in our study) but also because IFN was an approved and available drug outside the study.

For this academic study, there was no financial support for MRI. That is why the MRI part was limited to the centres that used MRI in their routine follow-up. The small number of patients lowered the power of the results, but we thought it was informative to look at the impact of the study drug on MRI outcomes. Unlike the analysis at month 9, there was no significant difference between the two groups in the mean number of Gd-enhancing lesions or in the percentage of inactive MRI scans at months 24 and 36. This might be due to the high drop-out rate, excluding the non-responders to the treatment strategy, indicating that the patients who dropped out from the study moved to other treatments and thus were not evaluated in the present study.

Although there were no major safety concerns among the patients included in this trial, the potential toxicity of MITOX in MS includes sporadic cases of heart disease20–22 and acute leukaemia.23–28 A long-term prospective follow-up survey on the safety profile of MITOX was performed in 802 MS patients from 12 French centres.29 In this French cohort followed for at least 5 years after MITOX institution and treated with a similar regimen to that used in this study, acute heart failure was observed in 0.1% of patients, as well as a decrease in the LVEF<50% in 4%, transient in 3%, persistent in 1%, therapy-related acute leukaemia in 0.25% and age-dependent persistent amenorrhoea in 17.3%. These long-term clinical data are not satisfactory compared with data from more recent drugs that have a shorter follow-up safety profile evaluation (fingolimod, cladribine and even natalizumab), although it is now established that in the patients who received natalizumab for more than 36 months, the cumulative risk of PML reached about 2.8/1000, that is, close to the risk of acute leukaemia with mitox.29 30

In conclusion, this study suggests that mitoxantrone may have a role as an induction therapy in the management of aggressive disease. The key remaining question is whether mitoxantrone offers any real advantages over therapies that have more specific, targeted effects on the immune system. It is still too early to claim that, for safety reasons, long-term use of these new drugs (Natalizumab, Cladribine, Fingolimod) offers a better benefit/RR over mitox when used as an induction treatment followed by an immunomodulatory drug. In this line, a new 3-year randomised trial comparing natalizumab versus mitoxantrone followed by an immunomodulator, funded by the French Health Ministry, has just started.


The authors are grateful to the steering committee members: O Sabouraud, P Hartung, L Kappos and J Oger. The authors also thank D Goodkin and P Narayana, for their fruitful discussions and comments on the manuscript, B Laviolle and S Hamonic, for their help in the statistical analysis, and Biotrial (CRO), for supervising the data entry.

The French-Italian Mitoxantrone-Interferon-beta-1b trial group

The following co-investigators were local site investigators: Bari, Italy (M Trojano, D Paolicelli, M D'Onghia); Besançon, France (L Rumbach); Clermont-Ferrand, France (P Clavelou, D Aufauvre); Dijon, France (T Moreau); Florence, Italy (MP Amato, E Portaccio); Gallarate, Italy (A Ghezzi); Genova, Italy (A Mancardi); Lille, France (P Vermersch, P Hautecoeur, J De Sèze); Limoges, France (L Magy, JM Vallat); Lyon, France (C Confavreux, S Vukusic, I Ionescu, S Blanc); Marseille, France (J Pelletier, I Malikova-Klemina, JP Ranjeva); Milano, Italy (G Comi, MA Rocca, M Filippi); Nancy, France (M Debouverie, S Pittion); Nice, France (C Lebrun); Hôpital Tenon, Paris, France (E Roullet, O Heinzlef); Fondation Rotschild, Paris, France (O Gout); La Pitié-Salpétrière, Paris, France (C Lubetzki, B Stankoff, A Tourbah); Rennes, France (G Edan, E Le Page, D Veillard); Strasbourg, France (JM Warter, C Tranchant); Toulouse, France (I Berry, D Brassat, M Clanet); Torino, Italy (L Durelli, M Clerico).


View Abstract


  • * For the list of co-investigators in The French—Italian Mitoxantrone Interferon-beta-1b Trial Group see the end of the paper

  • Funding Supported by an academic grant (Programme Hospitalier de Recherche Clinique, from the French Health Ministry), with additional grants from industry (Bayer-Schering, Lederle and Immunex). GE has received research support and compensation as a speaker from Biogen Idec, Serono and Sanofi-Aventis, Bayer Schering Pharma AG, LFB, and has acted as a consultant for Teva Pharmaceuticals, Merck-Serono, Bayer-Schering, Biogenidec, and LFB. GC has received consulting fees from Novartis, Teva Pharmaceuticals, Sanofi-Aventis, Merck Serono and Bayer Schering, and lecture fees from Novartis, Teva Pharmaceuticals, Sanofi-Aventis, Merck Serono, Biogen-Dompe and Bayer Schering. ELP has received compensation as a speaker from Biogen Idec, Serono, Sanofi-Aventis and Bayer Schering Pharma AG. MF has received research support and compensation as a speaker from Teva Pharmaceuticals, Merck-Serono, Bayer-Schering, Biogen-Dompe and Genmab, and has acted as a consultant for Teva Pharmaceuticals, Merck-Serono, Bayer-Schering, Biogen-Dompe and Genmab.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Ethics approval was provided by the Centre de Protection des Personnes pour la Recherche Biomédicale Centre Hospitalier Universitaire Rennes.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.