Efficacy and safety of oral fumarate in patients with relapsing-remitting multiple sclerosis: a multicentre, randomised, double-blind, placebo-controlled phase IIb study

Articles

Efficacy and safety of oral fumarate in patients with relapsing-remitting multiple sclerosis: a multicentre, randomised, double-blind, placebo-controlled phase IIb study Ludwig Kappos, Ralf Gold, David H Miller, David G MacManus, Eva Havrdova, Volker Limmroth, Chris H Polman, Klaus Schmierer, Tarek A Yousry, Minhua Yang, Mefkûre Eraksoy, Eva Meluzinova, Ivan Rektor, Katherine T Dawson, Alfred W Sandrock, Gilmore N O’Neill, for the BG-12 Phase IIb Study Investigators*

Summary Background Oral fumarate (BG00012) might have dual anti-inflammatory and neuroprotective effects. Our aim was to assess the efficacy and safety of BG00012 in patients with relapsing-remitting multiple sclerosis.

Lancet 2008; 372: 1463–72 See Comment page 1447 *Members listed at end of paper

Methods 257 patients, aged 18–55 years, with relapsing-remitting multiple sclerosis were randomly assigned to receive 120 mg once daily (n=64), 120 mg three times daily (n=64), or 240 mg three times daily (n=64) BG00012, or placebo (n=65) for 24 weeks. During an extension period of 24 weeks for safety assessment, patients treated with placebo received BG00012 240 mg three times daily. The primary endpoint was total number of new gadolinium enhancing (GdE) lesions on brain MRI scans at weeks 12, 16, 20, and 24. Additional endpoints included cumulative number of new GdE lesions (weeks 4–24), new or enlarging T2-hyperintense lesions, new T1-hypointense lesions at week 24, and annualised relapse rate. Analysis was done on the efficacy-evaluable population. Safety and tolerability were also assessed. This study is registered with ClinicalTrials.gov, number NCT00168701. Findings Treatment with BG00012 240 mg three times daily reduced by 69% the mean total number of new GdE lesions from week 12 to 24 compared with placebo (1·4 vs 4·5, p<0·0001). It also reduced number of new or enlarging T2-hyperintense (p=0·0006) and new T1-hypointense (p=0·014) lesions compared with placebo. BG00012 reduced annualised relapse rate by 32% (0·44 vs 0·65 for placebo; p=0·272). Adverse events more common in patients given BG00012 than in those given placebo included abdominal pain, flushing, and hot flush. Dose-related adverse events in patients on BG00012 were headache, fatigue, and feeling hot. Interpretation The anti-inflammatory effects and favourable safety profile of BG00012 warrant further long-term phase III studies in large patient groups. Funding Biogen Idec, Inc.

Introduction BG00012, an oral formulation of dimethyl fumarate, might have novel and complex effects on the pathobiology of multiple sclerosis. Preclinical experiments have shown that dimethyl fumarate and its primary metabolite monomethyl fumarate can activate the nuclear-factorE2-related factor-2 (Nrf2) transcriptional pathway,1 which controls phase-2 detoxifying enzyme gene expression, and is crucial for oxidative stress response and immune homoeostasis.2–4 Activation of the Nrf2 pathway defends against oxidative-stress-induced neuronal death,5–8 protects the blood–brain barrier,9 and supports maintenance of myelin integrity10 in the CNS. Dimethyl fumarate induces expression of phase-2 detoxification enzymes in astroglial and microglial cells.11 It also inhibits expression of cytokines and adhesion molecules implicated in the inflammatory response in vitro.11–13 These data suggest that BG00012 could have dual neuroprotective and antiinflammatory effects. An oral formulation of fumaric acid (Fumaderm, Biogen Idec Gmbh, Ismaning, Germany) showed effectiveness in patients with chronic plaque psoriasis, a disorder associated www.thelancet.com Vol 372 October 25, 2008

with immune dysfunction.14 In a pilot study in patients with relapsing-remitting multiple sclerosis, this formulation also reduced the number and volume of gadolinium enhancing (GdE) lesions on brain MRI scans compared with baseline.15 On the basis of these preliminary findings, we have tested the efficacy and safety of three doses of BG00012 versus placebo in a multicentre, randomised, double-blind, placebo-controlled, dose-ranging, phase IIb study in patients with relapsing-remitting multiple sclerosis.

University Hospital Basel, Basel, Switzerland (Prof L Kappos MD); University Clinic Bochum at St Josef Hospital, Bochum, Germany (Prof R Gold MD); Institute of Neurology, University College London, London, UK (Prof D H Miller MD, D G MacManus MSc, K Schmierer PhD, T A Yousry MD); General Teaching Hospital, Prague, Czech Republic (E Havrdova MD); City Hospital of Cologne, Cologne, Germany (V Limmroth MD); VU Medical Centre, Amsterdam, the Netherlands (Prof C H Polman MD); Biogen Idec, Cambridge, MA, USA (M Yang MS, K T Dawson MD, A W Sandrock MD, G N O’Neill MB); University of Istanbul, Istanbul, Turkey (Prof M Eraksoy MD); Motol Hospital, Charles University, second Medical School, Prague, Czech Republic (E Meluzinova MD); and Masaryk University, Brno, Czech Republic (I Rektor MD) Correspondence to: Prof Ludwig Kappos, University Hospital Basel, Neurology and Department of Biomedicine, Petersgraben 4, CH 4031, Basel, Switzerland [email protected]

Methods Patients 257 patients were recruited from 43 centres in the Czech Republic, Germany, Hungary, Netherlands, Poland, Russia, Sweden, Switzerland, Turkey, and UK, between Nov 24, 2004, and March 31, 2005. Participants were aged 18–55 years with a diagnosis of relapsing-remitting multiple sclerosis by McDonald criteria,16 a baseline Expanded Disability Status Scale (EDSS) score between 0 and 5,17 and either at least one relapse within 12 months of randomisation and a previous cranial MRI scan showing 1463

Articles

lesions consistent with multiple sclerosis, or GdE lesions on MRI scans done within 6 weeks of randomisation. Women with child-bearing potential had to use appropriate birth control, and were excluded from the study if they were pregnant or breastfeeding. Patients were excluded if they had progressive forms of multiple sclerosis or serious medical disorders, either concurrently (ie, HIV or hepatitis) or by history (ie, malignant disease or anaphylactic reactions, drug or alcohol abuse within 2 years of randomisation, or a relapse of multiple sclerosis within 50 days of randomisation, or no stabilisation from a previous relapse at the time of randomisation, or both). Patients were excluded if they had been previously treated with fumaric acid, FAG-201, BG00012, cladribine, T-cell or T-cell-receptor vaccination, total lymphoid irradiation, or therapeutic monoclonal antibodies apart from natalizumab. Patients were also excluded if they received mitoxantrone or cyclophosphamide within 1 year; cyclosporine, azathioprine, methotrexate, natalizumab, intravenous immunoglobulin, plasmapheresis, or other investigational drugs within 6 months; glatiramer acetate or interferon beta within 3 months; or corticosteroid (oral or intravenous), 4-aminopyridine, or related products within 30 days of randomisation.

Study design and outcome measures This randomised, double-blind, placebo-controlled, parallelgroup, dose-ranging study was a 24-week, blinded, placebocontrolled treatment period (part 1), followed by 24 weeks for dose-blinded safety assessment (part 2) (figure 1). Recruitment started Nov 24, 2004, and the trial ended May 22, 2006. During the first 24 weeks, patients were randomly assigned to receive oral BG00012 120 mg once daily, 120 mg or 240 mg three times daily, or placebo. During the following 24 weeks, patients who received BG00012 in the first 24 weeks were maintained on the same BG00012 dose, and patients who received placebo in the first 24 weeks were given BG00012 240 mg three times daily. Patients who were randomly assigned to BG00012 240 mg three times daily received 120 mg three times daily for the first week and 240 mg three times daily from week 2.

Screening

Blinded placebo-controlled treatment period (part 1)

Dose-blinded safety extension period (part 2)

Placebo

BG00012 240 mg three times daily

BG00012 120 mg once daily (120 mg per day) Randomisation (1:1:1:1) BG00012 120 mg three times daily (360 mg per day)

BG00012 240 mg three times daily (720 mg per day) 24 weeks

Figure 1: Study design

1464

24 weeks

BG00012 and placebo were administered as entericcoated microtablets in gelatin capsules, which had identical appearance and taste. Daily medication was given in blister packs of six tablets, with different numbers of tablets containing placebo or the active drug to preserve the blinding. In the extension phase, all patients received a new set of medication. During both study periods, reduction to one capsule three times daily for 1 month was allowed for patients unable to tolerate higher doses. Moreover, dosing interruptions were needed for elevated liver or renal function tests, or decreased white blood cell counts. After dose interruption, patients were examined every 2 weeks and allowed to resume dosing when laboratory values returned to normal. Patients with abnormal values for more than 4 consecutive weeks permanently discontinued the study drug. The study protocol was approved by independent ethics committees, and the study was done in accordance with the Declaration of Helsinki, International Conference of Harmonisation and Good Clinical Practice guidelines, and local regulations. Enrolled patients provided written informed consent.

Study procedures and endpoints To prevent unblinding of treatment assignment, separate study personnel were assigned to treat patients and to assess drug efficacy. A treating neurologist was responsible for routine neurological care, assessing and treating adverse events, and analysing laboratory test results. Neurologists not otherwise involved in the care of study participants assessed patients at scheduled and unscheduled relapse examinations. These examining neurologists were trained and certified in EDSS examination after attending special training courses, viewing a DVD-ROM,18 and successfully passing a multiple choice test. Patients attended clinics every 4 weeks during both study periods. Patients were instructed not to take their medication within 4 h before clinic visits because of the possibility for study unblinding due to flushing. Brain MRI scans were done at baseline and at weeks 4, 8, 12, 16, 20, and 24. Participating imaging sites were equipped with MRI systems operating at 1·0 or 1·5 Tesla. For all patients, 46 contiguous 3-mm-thick axial images of the brain were acquired with a dual echo fast (turbo) spin echo (FSE) sequence (repetition time [TR] 2500–3300 ms, echo time 1 [TE1] 10–40 ms, echo time 2 [TE2] 80–100 ms) to provide proton density and T2-weighted images. Furthermore, a conventional spin echo sequence (TR 500–700 ms, TE 10–20 ms) was undertaken before and after injection of gadolinium-based contrast medium to provide T1-weighted images in the corresponding spatial location to the FSE scans. We used a field of view of 250 mm together with a reconstructed image matrix of 256×256, resulting in an in-plane spatial resolution of 0·97×0·97 for all images. All scans were analysed from hard-copy film by two trained observers (KS and DGM) under the supervision of an experienced neuroradiologist (TAY) at a www.thelancet.com Vol 372 October 25, 2008

Articles

309 patients assessed for eligibility

257 patients enrolled and randomly assigned to study groups in placebo-controlled treatment period (part 1)

65 assigned to receive placebo

64 assigned to receive BG00012 120 mg once daily

64 assigned to receive BG00012 120 mg three times daily

64 assigned to receive BG00012 240 mg three times daily 1 did not receive BG00012

6 withdrew from study 3 voluntary withdrawal 1 lack of tolerance 1 non-compliance 1 lost to follow-up

6 discontinued placebo but completed follow-up

59 completed part 1

4 withdrew from study 1 adverse event 1 lack of tolerance 1 voluntary withdrawal 1 non-compliance

6 discontinued BG00012 but completed follow-up

60 completed part 1

5 withdrew from study 2 adverse events 2 voluntary withdrawal 1 lack of tolerance

8 discontinued BG00012 but completed follow-up

59 completed part 1

6 withdrew from study 4 adverse events 1 lack of tolerance 1 voluntary withdrawal

10 discontinued BG00012 but completed follow-up

57 completed part 1

225 patients enrolled in dose-blinded safety-extension phase (part 2)

59 transitioned from placebo to BG00012 240 mg three times daily

3 withdrew from study 3 adverse events

6 discontinued BG00012 but completed follow-up

56 completed part 2

58 continued BG00012 120 mg once daily

2 withdrew from study 2 voluntary withdrawals

56 continued BG00012 120 mg three times daily

2 discontinued BG00012 but completed follow-up

56 completed part 2

1 withdrew from study 1 voluntary withdrawal

3 discontinued BG00012 but completed follow-up

55 completed part 2

52 continued BG00012 240 mg three times daily

0 withdrew from study

1 discontinued BG00012 but completed follow-up

52 completed part 2

Figure 2: Trial profile

central reading centre. New GdE lesions, compared with the previous scan, were identified and marked every month on postcontrast T1-weighted sequences. For new or enlarging T2 lesions, the week-24 scan was compared with the baseline (week-0) scan. The proton-density weighted scan was marked, although the second echo, a more heavily T2-weighted sequence, was also reviewed to confirm the presence of a new lesion. Enlarging T2 lesions were defined as those that appeared larger on two contiguous slices or showed at least twice the diameter compared with the baseline scan. New T1-hypointense black-hole lesions were identified and marked on the unenhanced T1-weighted scans at week 24 compared with the baseline scans. EDSS scores were assessed at baseline and at weeks 12, 24, 36, and 48. Patients were monitored for relapses at all clinic visits, and suspected relapses were assessed on www.thelancet.com Vol 372 October 25, 2008

unscheduled visits. The treating neurologist was contacted within 48 h of symptom onset, and the assessment was done within 72 h; EDSS scores were reviewed during unscheduled visits. Relapses were defined as new or recurrent neurological symptoms lasting for 24 h or longer, not associated with fever or infection, and accompanied by new objective neurological findings on examination. At the neurologists’ discretion, patients with relapses were treated per protocol with intravenous methylprednisolone 1000 mg per day for 3 or 5 days. Haematology, blood chemistry, and urinalysis were done every 4 weeks. Electrocardiographs were done at screening and at weeks 12, 24, 36, and 48. All adverse events were documented throughout the study, regardless of severity or relation to study drug. Patients discontinuing BG00012 during the first or the second part of the study were encouraged to remain in the study until week 24 1465

Articles

Placebo, N=65

BG00012 120 mg once daily, N=64

BG00012 120 mg BG00012 240 mg three times daily, N=64 three times daily, N=63

Age (years)

35·6 (8·2)

34·8 (10·2)

36·3 (9·5)

37·3 (9·1)

Sex (female)

36 (55%)

42 (66%)

44 (69%)

42 (67%)

Ethnic origin (white)

64 (98%)

62 (97%)

64 (100%)

60 (95%)

Median time since symptom onset (year) (IQR)

6 (4·0–11·0)

7 (4·0–11·5)

5 (3·0–9·0)

6 (4·0–12·0)

Median number of relapses Previous 1 year

1

1

1

1

Previous 3 years

2

2

2

3

2·67 (1·23)

2·52 (1·11)

2·51 (1·02)

2·87 (1·33)

EDSS score Number of GdE lesions 0

38 (58%)

30 (47%)

30 (47%)

41 (65%)

1

10 (15%)

19 (30%)

7 (11%)

10 (16%)

2

10 (15%)

3 (5%)

10 (16%)

4 (6%)

3

3 (5%)

3 (5%)

3 (5%)

2 (3%)

≥4

4 (6%)

9 (14%)

14 (22%)

6 (10%)

Mean

1·6 (6·6)

1·4 (2·1)

2·5 (4·2)†

1·3 (3·4)

Data are n (%) or mean (SD), unless otherwise stated. EDSS=Expanded Disability Status Scale. GdE=gadolinium enhancing. IQR=interquartile range. *Intention-to-treat population. †p=0·045 versus placebo; p=not significant versus BG00012 120 mg once daily; p=0·019 versus BG00012 240 mg three times daily based on logit regression.

Table 1: Demographics and baseline clinical characteristics*

or 48, respectively, and to continue protocol-scheduled assessments. The primary endpoint was the total number of new GdE lesions over four scans at weeks 12, 16, 20, and 24 (calculated as the sum of the scans): this time was selected on the basis of an apparent latency to clinical effect from the pilot study of multiple sclerosis.15 Secondary MRI endpoints included the cumulative number of new GdE lesions from week 4 to 24, and the numbers of new or enlarging T2-hyperintense lesions and new T1-hyperintense lesions at week 24. The efficacy of BG00012 on relapses and disability progression, and its safety and tolerability, were also assessed.

Statistical analysis Mean and variability estimates to calculate sample size were derived from summary data from placebo-treated patients with relapsing-remitting multiple sclerosis in other clinical studies (mean=8·6, SD=12·7).19 The standard deviation for the group treated with the active compound was calculated by assuming a negative binomial distribution of the number of MRI lesions on the basis of the mean in the calculation. A sample size of 65 patients per group had 80% power to detect a treatment effect of 55% reduction or more on the total number of new GdE lesions, when comparing the mean of the placebo group with that of the two groups treated with the two highest doses of active drug combined. This sample size calculation assumed that the total lesion number was from four consecutive scans and was based on a two-group Satterthwaite t test of equal means and unequal variances. Additionally, we assumed an alpha value of 0·05 with no adjustment for multiple comparisons. The calculation was done with Nquery (version 5.0) statistical software (Statistical Solutions, Saugus, MA, USA). 1466

MRI endpoints were analysed with the non-parametric Wilcoxon rank sum test. Annualised relapse rate was analysed with a Poisson regression model adjusted for the number of relapses in the 12 months before study entry, and proportion of relapse-free patients was analysed with Fisher exact test. No formal statistics were done on EDSS scores. MRI analyses are reported for the efficacy-evaluable population (ie, patients who had no missing MRI data). MRI scans must have been done before or at least 24 days after steroid treatment. Clinical efficacy analyses are reported for the intention-to-treat population, defined as all patients who were randomised and received at least one dose of drug. All statistical tests were two-sided, with an alpha value of 0·05; no adjustment (type I error rate) for multiple comparisons was made. The safety population included all patients who received at least one BG00012 dose. Differences in adverse-event reporting between all BG00012 groups combined versus placebo group were calculated with the Fisher exact test.

Role of the funding source The employees of Biogen Idec Inc listed as authors participated with the other authors in the study design and statistical analysis of the data. These authors reviewed and approved the manuscript. The corresponding author had full access to all study data and had final responsibility for submission of the publication.

Results 257 patients were randomly assigned to receive BG00012 120 mg once daily, or 120 mg or 240 mg three times daily, or placebo. One patient, who was randomly assigned to the highest dose of BG00012, withdrew before receiving the drug (figure 2). Patients were well-matched for www.thelancet.com Vol 372 October 25, 2008

Articles

demographic and baseline disease characteristics (table 1). The mean number of baseline GdE lesions was higher for the BG00012 120 mg three times daily group than for all other treatment groups (table 1). During the first 24 weeks, more patients receiving BG00012 120 mg and 240 mg three times daily discontinued study drug than patients receiving BG00012 120 mg once daily or placebo (figure 2). Overall, 21 of 256 patients who received the drug withdrew from the first part of the study. An additional ten patients who completed the first part of the study did not enter in the second part. During the second 24 weeks, more patients changing from placebo to BG00012 240 mg three times daily prematurely discontinued study drug than patients maintaining their original treatment. Overall, six patients withdrew from this part of the study. 235 patients completed their study visits during the first 24 weeks and 219 completed the second 24 weeks. Patients receiving BG00012 240 mg three times daily had fewer total new GdE lesions (69% reduction) at weeks 12, 16, 20, and 24 combined compared with those receiving placebo (1·4 vs 4·5, p<0·0001) (figure 3). Similarly, a sensitivity analysis of the intention-to-treat population on this endpoint showed that patients receiving this dose of BG00012 had fewer new GdE lesions compared with those receiving placebo (1·3 vs 4·8, p<0·0001). Although treatment with BG00012 led to fewer total new GdE lesions in patients receiving 120 mg once daily (3·3) and three times daily (3·1) than placebo (4·5), these treatment group differences were not significant. From week 4 to 24, patients receiving BG00012 240 mg three times daily had fewer cumulative new GdE lesions (44% reduction) than those receiving placebo (table 2). BG00012 240 mg also reduced the mean number of new or enlarging T2-hyperintense lesions (48% reduction) during the first 24 weeks compared with placebo. During this period, no new T2 lesions developed in 63% of patients receiving this dose of BG00012 compared with

Total number of GdE lesions

6

*p<0·0001 vs placebo

5 4 3 *

2

69% reduction, BG00012 240 mg three times daily vs placebo

1 n=54

0 Placebo

BG00012 BG00012 BG00012 120 mg once 120 mg three 240 mg three daily times daily times daily Treatment group

Figure 3: Mean total number of GdE lesions from scans at weeks 12, 16, 20, and 24 combined Vertical bars=SE.

26% of those receiving placebo. Furthermore, the mean number of T1-hypointense lesions after 24 weeks of treatment was lower (53% reduction) in the BG00012 240 mg three times daily dose group compared with the placebo group. No significant differences were seen between the lower BG00012 dose groups (ie, 120 mg once daily and 120 mg three times daily) and the placebo group on any of the MRI endpoints assessed. During the first part of the study, the annualised relapse rate for patients who received BG00012 240 mg three times daily decreased by 32% (table 2). The proportion of relapse-free patients was close between the two groups. A similar percentage of patients in every treatment group was treated with intravenous methylprednisolone per protocol for relapses (n=11 [17%] in the placebo group; n=9 [14%] in the BG00012 120 mg once daily group; n=11 [17%] in the BG00012 120 mg three times daily group; and n=6 [10%] in the BG00012 240 mg three times daily group). During the second part of the study, the

Placebo*

BG00012 120 mg once daily

BG00012 120 mg three times daily

BG00012 240 mg three times daily

54

59

56

54

MRI† Number of new GdE lesions (weeks 12−24) Mean

4·5 (7·4)

3·3 (5·1)

3·1 (5·9)

1·4 (3·8)

Median (IQR)

2·0 (0–5·0)

1·0 (0–4·0)

1·0 (0–3·5)

0·0 (0–1·0)

p value vs placebo Number of new GdE lesions (weeks 4−24)

·· 53

0·266 55

0·068 54

<0·0001 52

Mean

6·6 (11·4)

6·2 (8·9)

6·7 (10·9)

3·7 (11·2)

Median (IQR)

3·0 (0–7·0)

3·0 (0–7·0)

2·0 (0–7·0)

0·0 (0–3·0)

p value vs placebo Number of new GdE lesions per patient, per scan (weeks 12−24)

·· 54

0·943 59

0·801 56

0·002 54

Mean

1·13 (1·84)

0·83 (1·29)

0·78 (1·48)

0·35 (0·94)

Median (IQR)

0·50 (0–1·25)

0·25 (0–1·00)

0·25 (0–0·88)

0·00 (0–0·25)

p value‡

··

0·266

0·068

<0·0001 (Continues on next page)

www.thelancet.com Vol 372 October 25, 2008

1467

Articles

Placebo*

BG00012 120 mg once daily

BG00012 120 mg three times daily

BG00012 240 mg three times daily

54

59

56

54

0

14 (26%)

16 (27%)

15 (27%)

34 (63%)

1

10 (19%)

6 (10%)

11 (20%)

6 (11%)

2

8 (15%)

10 (17%)

8 (14%)

0 (0%)

3

1 (2%)

7 (12%)

2 (4%)

2 (4%)

≥4

21 (39%)

20 (34%)

20 (36%)

12 (22%)

Mean

4·2 (5·4)

3·8 (4·7)

4·1 (5·7)

2·2 (5·4)

Median (IQR)

2·0 (0·0–6·0)

2·0 (0·0–5·0)

2·0 (0·0–4·5)

0·0 (0·0–3·0)

(Continued from previous page) Number of new or enlarging T2-hyperintense lesions (week 24)

p value vs placebo Number of new T1-hypointense lesions (week 24)

··

0·965

0·839

0·0006

54

59

56

54

0

25 (46%)

26 (44%)

21 (38%)

35 (65%)

1

9 (17%)

15 (25%)

17 (30%)

11 (20%)

2

7 (13%)

7 (12%)

8 (14%)

4 (7%)

3

3 (6%)

5 (8%)

1 (2%)

3 (6%)

≥4

10 (19%)

6 (10%)

9 (16%)

1 (2%)

Mean

1·7 (2·5)

1·3 (1·8)

1·5 (2·0)

0·8 (2·0)

Median (IQR)

1·0 (0–2·0)

1·0 (0–2·0)

1·0 (0–2·0)

0·0 (0–1·0)

0·732

0·836

0·014

p value vs placebo Clinical§ Weeks 0−24

65

64

64

63

Annualised relapse rate (95% CI)¶

0·65 (0·43–1·01)

0·42 (0·24–0·71)

0·78 (0·52–1·16)

p value vs placebo

··

0·196

0·572

0·272

53 (83%)

44 (69%)

51 (81%)

Relapse-free p value vs placebo|| Weeks 25−48 Annualised relapse rate (95% CI)¶ Relapse-free Weeks 0−48 Annualised relapse rate (95% CI)¶ Relapse-free

49 (75%) ·· 59 0·26 (0·13–0·53)

0·387 58

0·437

0·524

56

0·24 (0·11–0·50)

0·44 (0·26–0·76)

52

0·47 (0·27–0·82)

0·16 (0·07–0·41)

52 (75%)

50 (86%)

43 (77%)

48 (92%)

59

58

56

52

0·41 (0·26–0·64) 40 (68%)

0·29 (0·17–0·50) 44 (76%)

0·60 (0·41–0·88) 32 (57%)

0·28 (0·16–0·50) 40 (77%)

Data are numbers (%) or mean (SD), unless indicated otherwise. GdE=gadolinium enhancing. IQR=interquartile range. *Placebo-treated patients transitioned to BG00012 240 mg three times daily during weeks 25−48. †Results from efficacy population; weeks refer to scans done during that week. ‡Mean number of new GdE lesions per scan, BG00012 vs placebo. §Results from intention-to-treat population. Data for weeks 25−48 and weeks 0−48 were from the intention-to-treat population that entered the extension period. ¶From Poisson regression, adjusted for number of relapses in the 12 months before study entry. Placebo group received BG00012 for 24 weeks, whereas other treatment groups received BG00012 for up to 48 weeks. ||Based on Fisher exact test. Patients who withdrew from the study without experiencing a relapse are considered not to have relapsed.

Table 2: MRI and clinical endpoints

annualised relapse rate for patients who received placebo changing to BG00012 240 mg three times daily, BG00012 120 mg once daily, BG00012 120 mg three times daily, or BG00012 240 mg three times daily was reduced by 60%, 43%, 40%, and 64% compared with annualised relapse rates during the first part of the study. During the first part of the study, 75% of patients treated with placebo and 88% of all patients receiving BG00012 reported at least one adverse event (n=55 [86%] of those treated with BG00012 120 mg once daily; n=59 [92%] of those treated with BG00012 120 mg three times daily; and n=55 [87%] of those treated with BG00012 240 mg three times daily). The most common adverse events included flushing, multiple sclerosis relapse, and 1468

headache (table 3). Adverse events significantly more common in patients receiving BG00012 240 mg three times daily than in those receiving placebo included upper abdominal pain (p=0·029), abdominal pain (p=0·013), hot flush (p=0·013), and flushing (p<0·0001). However, the frequency of flushing events decreased with increasing BG00012 doses. In general, flushing started within 30 min of drug administration and subsided within 90 min. Gastrointestinal adverse events were reported by 25% of patients on placebo, 30% of those on BG00012 120 mg once daily, 39% on BG00012 120 mg three times daily, and 41% on BG00012 240 mg three times daily. The most common gastrointestinal events were nausea, diarrhoea, and upper abdominal pain www.thelancet.com Vol 372 October 25, 2008

Articles

(table 3). The severity of these gastrointestinal events was mild to moderate except for one severe event each in the placebo, BG00012 120 mg once daily, and BG00012 120 mg three times daily groups, and four events in the BG00012 240 mg three times daily group. Adverse events with BG00012 dose-related patterns were headache, fatigue, and feeling hot. Infections were reported in similar proportions of BG00012 and placebo patients (34% in both groups). Only one serious infection, an incident of pelvic inflammatory disease in a patient receiving BG00012 120 mg three times daily, was reported. This patient responded well to standard medical management. Serious adverse events were reported in 12% of patients on placebo and 9% of those on BG00012. The only serious adverse event to occur in more than one patient was multiple sclerosis relapse (table 3). During the first 24 weeks, the frequency of adverse events that led to drug discontinuation increased slightly with increasing doses of BG00012 (8% of patients on BG00012 120 mg once daily, 11% of those on BG00012 120 mg three times daily, and 13% of those on BG00012 240 mg three times daily). Adverse events resulting in drug discontinuation included flushing (one patient on BG00012 120 mg three times daily and one on BG00012 240 mg three times daily), increased alanine aminotransferase concentration (one patient on BG00012 120 mg once daily and one on BG00012 120 mg three times daily), nausea (two patients on BG00012 240 mg three times daily), and vomiting (two patients on BG00012 240 mg three times Placebo, N=65 Flushing

BG00012 120 mg once daily, N=64

daily). Diarrhoea was the only event that led to withdrawal of more than one patient from the first part of the study (two patients on BG00012 120 mg once daily). Only one BG00012-treated patient (120 mg once daily) withdrew because of non-compliance. The profile of adverse events during the second part of the study was broadly similar to that of the first part, and no new safety issues were apparent in patients changing from placebo to BG00012 240 mg (webtable). In addition to nasopharyngitis, the most common infections during the second part of the study included influenza (4%), upper respiratory tract infection (4%), and pharyngitis (4%). These events were reported in similar numbers of patients in every treatment group. For patients receiving BG00012 during both study periods, adverse events such as flushing, headache, nausea, upper abdominal pain, and pruritus took place more frequently during the first 24 weeks than during the second 24 weeks. In all BG00012 groups, the frequency of flushing decreased from 66% (n=127) events during month 1 to 5% (n=9) during month 6, and the frequency of gastrointestinal events decreased from 52% (n=132) during month 1 to 4% (n=10) during month 6 (webfigure). Fewer patients who received BG00012 during the entire study withdrew from the second part of the study than those changing from placebo to BG00012 240 mg three times daily (2% vs 5%). All withdrawals in patients on BG00012 during the entire study were voluntary and in those changing from placebo to BG00012 240 mg three BG00012 120 mg three times daily, N=64

BG00012 240 mg BG00012 total, three times daily, N=63 N=191

6 (9%)

34 (53%)

31 (48%)

25 (40%)

16 (25%)

11 (17%)

20 (31%)

12 (19%)

43 (23%)

7 (11%)

5 (8%)

11 (17%)

13 (21%)

29 (15%)

10 (15%)

7 (11%)

6 (9%)

7 (11%)

20 (10%)

5 (8%)

1 (2%)

9 (14%)

10 (16%)

20 (10%)

Diarrhoea

3 (5%)

6 (9%)

5 (8%)

7 (11%)

18 (9%)

Pruritus

5 (8%)

6 (9%)

5 (8%)

6 (10%)

17 (9%)

Upper abdominal pain

2 (3%)

5 (8%)

4 (6%)

9 (14%)

18 (9%)

Hot flush

0 (0%)

5 (8%)

1 (2%)

6 (10%)

12 (6%)‡

Abdominal pain

0 (0%)

2 (3%)

2 (3%)

6 (10%)

10 (5%) 14 (7%)

Multiple sclerosis relapse Headache Nasopharyngitis Nausea

See Online for webtable

See Online for webfigure

90 (47%)†

Serious adverse events Multiple sclerosis relapse

5 (8%)

3 (5%)

6 (9%)

5 (8%)

Abdominal pain

0 (0%)

0 (0%)

0 (0%)

1 (2%)

Lower limb fracture

1 (2%)

0 (0%)

0 (0%)

0 (0%)

1 (<1%)

Multiple sclerosis

0 (0%)

1 (2%)

0 (0%)

0 (0%)

1 (<1%)

Pelvic inflammatory disease

0 (0%)

0 (0%)

1 (2%)

0 (0%)

1 (<1%)

Phlebitis

0 (0%)

0 (0%)

1 (2%)

0 (0%)

1 (<1%)

Urinary retention

0 (0%)

0 (0%)

0 (0%)

1 (2%)

1 (<1%)

1 (<1%)

Uterine leiomyoma

1 (2%)

0 (0%)

0 (0%)

0 (0%)

1 (<1%)

Vertigo

1 (2%)

0 (0%)

0 (0%)

0 (0%)

1 (<1%)

Data are numbers (%). *Reported in 10% or more of patients in any treatment group, regardless of study drug. †p<0·0001, all BG00012 groups combined versus placebo. ‡p=0·0125, all BG00012 groups combined versus placebo.

Table 3: Adverse events in the placebo-controlled treatment period*

www.thelancet.com Vol 372 October 25, 2008

1469

Articles

times daily were due to adverse events. Drug discontinuations during the second part of the study took place less often in patients receiving BG00012 during the entire study compared with those transitioning from placebo to BG00012 240 mg three times daily (4% vs 15%). The most common reasons for discontinuation were adverse events (2%) or voluntary withdrawal (2%) in patients on BG00012 during the entire study, and lack of tolerance (7%) in those changing from placebo to BG00012 240 mg three times daily. Overall, dose reductions due to tolerability were rare. No clinically meaningful trends were seen for changes from baseline in laboratory indices. Haematology shifts were minor and did not lead to any alteration of drug dosing or correlate with any clinically significant effect, such as opportunistic infection, urticaria, or drug reaction. Clinically significant anaemia or neutropenia did not occur. During the first 24 weeks of the study, the proportion of BG00012-treated patients with raised transaminase concentrations increased in a dose-related manner compared with that of patients on placebo, with the most increases less than twice the upper limit of the normal range. No coincident or delayed increase from baseline bilirubin or other indicator of impaired hepatic function was detected in BG00012-treated patients with an increase in transaminase concentration three times the upper limit of the normal range. No hepatitis-related symptoms occurred coincidentally with transaminase increases. In all cases, reduction of transaminase concentration was seen after BG00012 discontinuation. On rechallenge, some patients tolerated treatment without increases in transaminase concentration. No clinically significant renal or urinary findings were seen. We did not note any substantial finding in vital signs, physical examination, or electrocardiograph results.

Discussion We showed that treatment with BG00012 240 mg three times daily reduces brain MRI activity in patients with relapsing-remitting multiple sclerosis. Treatment with the highest BG00012 dose studied reduced by 69% the mean number of new GdE lesions from scans at weeks 12 to 24 compared with treatment with placebo. Patients in the BG00012 120 mg three times daily group had a higher baseline GdE lesion count compared with other treatment groups, which explained the absence of a more pronounced reduction in new GdE lesions in this group. The efficacy of BG00012 in this study confirms results of a pilot study of ten patients with multiple sclerosis receiving oral fumaric acid ester therapy (1290 mg per day fumaric acid, 720 mg per day of which was dimethyl fumarate) for 18 weeks, which showed reductions in the number (p<0·02) and volume (p<0·018) of GdE lesions on active treatment compared with baseline.15 The presence of GdE lesions is indicative of blood–brain barrier breakdown and continuing inflammatory activity 1470

within the CNS, and has been correlated with acute clinical relapses.20–23 GdE lesion accumulation also is associated with cerebral atrophy, which indicates CNS tissue damage and loss.24 Atrophy was not an outcome of this study because of the short duration of the study. In addition to its effect on GdE lesion formation, treatment with BG00012 240 mg three times daily led to a 48% reduction in new T2-hyperintense lesion development compared with treatment with placebo, indicating a decrease in the overall accumulation of new lesions.21 Treatment with BG00012 240 mg three times daily also reduced by 53% T1-hypointense lesion development compared with treatment with placebo. T1-hypointense lesions have been associated with greater tissue matrix destruction, including axonal damage, than T1-isointense lesions.25 Because new T1-hypointense lesions at week 24 were recorded on pregadolinium scans, they might have included lesions that were still GdE. To address the issue of chronicity, we did a posthoc analysis of the evolution of new GdE lesions at weeks 4, 8, and 12 evolving into T1-hypointense lesions on the postgadolinium week-24 scan, and showed a reduced probability of conversion of GdE lesions into T1-hypointense lesions (odds ratio [OR] 0·51; 95% CI 0·43–0·61; p<0·0001). Although these findings seem to be consistent with in vitro and animal studies that suggest that BG00012 may have neuroprotective and anti-inflammatory effects, additional studies are needed to see whether both these effects are present in human beings and whether they take place independently of one another. The possibility that BG00012 is neuroprotective through its proposed mode of action is of particular interest because several lines of evidence have indicated that oxidative stress contributes to the pathological changes of multiple sclerosis. Studies have shown that macrophages in the CNS generate oxygen and nitrogen free radicals that could contribute to demyelination and axonal injury in experimental autoimmune encephalomyelitis and multiple sclerosis,26,27 and that free radicals can activate transcription factors (eg, nuclear factor-κB) that upregulate expression of many genes that could be implicated in experimental autoimmune encephalomyelitis and multiple sclerosis.28 Evidence also exists of increased free radical activity or deficiencies in important antioxidant enzymes in patients with multiple sclerosis compared with healthy individuals.29–31 Antioxidants are effective in reducing clinical deficits in mice with experimental autoimmune encephalomyelitis.32 BG00012 240 mg three times daily reduced the annualised relapse rate compared with placebo, although the effect was not significant. However, this phase IIb study was not adequately powered for relapse endpoints. Relapse rates in all BG00012 treatment groups—and not only in those who switched from placebo to BG00012 240 mg three times daily—decreased between the first and the second part of the study. This result could indicate a delayed and increasing effect of BG00012 over time. Nevertheless, without a placebo control, regression www.thelancet.com Vol 372 October 25, 2008

Articles

to the mean would also be a plausible explanation. Therefore, further study is needed to explore the meaning of this finding. BG00012 was safe and generally well tolerated. Gastrointestinal adverse events took place mainly in the early phase of treatment and contributed to the higher discontinuation rate of 9·5% after 6 months in the highest BG00012 dose group. Cumulative toxicity was not seen during 48 weeks of treatment. Common adverse events in patients on BG00012 or on placebo included multiple sclerosis relapse and headache. The frequency of infection was low in all treatment groups and not different from that of placebo group. Although the duration of this study was shorter than 1 year, this finding might be consistent with more than a purely immunosuppressive mode of action. Flushing was the most commonly reported adverse event and occurred significantly more often in patients on BG00012 than in those on placebo, but rarely resulted in study drug discontinuation and was not related to dose. Flushing and gastrointestinal-related events decreased during the course of BG00012 use, especially during the first 1–2 months of treatment. After the first 3 months, the frequency of these events was similar to that in the placebo group during the first part of the study. Because patients were instructed not to take their medication within 4 h of clinic visits, and because the frequency of flushing and gastrointestinal events decreased within the first few months of treatment, it is unlikely that the higher frequency in the BG00012 group compared with that in the placebo group compromised clinical assessments in the study. Because flushing and gastrointestinal events did not seem to be related to dose, they would not explain different relapse reporting rates in the three dose groups. Longer-term (phase III) studies of BG00012 in larger patient populations are underway to define its place in the future of relapsing-remitting multiple sclerosis treatment. If these studies show similar relapse rate reductions with BG00012, interferon beta, and glatiramer acetate,33–36 BG00012 could be a suitable initial treatment for relapsing-remitting multiple sclerosis. Because of the convenience of an orally administered product, BG00012 could also be an alternative for patients who cannot tolerate or choose not to initiate injectable therapies because of injection-related effects or anxiety. The potentially unique mode of action of BG00012 could also be valuable as a monotherapy or combination therapy. Contributors As chair of this study’s steering committee, LK was involved in drafting and amending the study protocol and statistical analysis plan, overseeing the conduct of the study, and reviewing statistical analysis. LK was also involved in and contributed to the writing and review of the manuscript. As corresponding author, LK had full access to all data in the study and had final responsibility for the decision to submit for publication. RG participated in patient recruitment, study design, data analysis, and discussion, review, and approval of the manuscript. DHM participated in study design, MRI data analysis, and editing, review, and approval of the manuscript. DGM participated in MRI protocol design, MRI data

www.thelancet.com Vol 372 October 25, 2008

analysis, and review and approval of the manuscript. EH, VL, ME, and AWS participated in study design, data analysis, and review and approval of the manuscript. CHP participated in study design, patient recruitment, data collection and analysis, and review and approval of the manuscript. KS and IR participated in data collection and analysis and review and approval of the manuscript. TAY supervised the MRI analysis and participated in review and approval of the manuscript. MY participated in study design and data analysis, served as study statistician, and reviewed and approved the manuscript. EM participated in data collection and analysis, patient recruitment, and review and approval of the manuscript. KD participated in data analysis and review and approval of the manuscript. GNO participated in study design, patient recruitment, data collection and analysis, safety review, and writing, review, and approval of the manuscript. BG-12 Phase IIb Suty Group Safety Monitoring Committee—C Confavreux, O Lyon-Caen (France); J Ware, G Burton Saibert (USA). Study Group—Czech Republic—Z Ambler, E Ehler, A Kopal, A Novotna, E Havrdová, D Horakova, E Meluzinova, D Krysl, J Pikova, I Rektor, R Talab; Germany—R Gold, R Linker, S Schilling, H-P Hartung, B Keiseier, S Schimrigk, A Haighikia, K Hellwig; Hungary—A Csányi, G Rum, L Csiba, T Csépány, A Lengyel, G Jakab, I Mayer, B Szabo; Netherlands— R Q Hintzen, T Siepman, C Polman, J M Nielsen; Poland—W Drozdowski, R Pogorzelski, W Fryze, J Kochanowicz, R Zimnoch, H Kwiecinski, Z Maciejek, K Selmaj, A Szczudlik, A Wajgt, M Maciejowski; Russia— A N Belova, A Sokolov, I Sokolova, A N Boiko, A P Elchaninov, V A Kozlov, M M Odinak, S B Shvarkov, A A Skoromets, I D Stolyarov, O V Vorobieva, I A Zavalishin; Sweden—J Hillert, E Åkesson-Lindström, V Karrenbauer, J Lycke, T Olsson, L Brundin, F Piehl; Switzerland—L Kappos, L Achtnichts, J Kuhle; Turkey—M Eraksoy, R Karabudak, A Siva, A Altintas, S Saip; United Kingdom—G Giovannoni, R Farrell, C Hawkins, M Sharief, B Sharrack, O Suliman, J Kemp. Conflict of interest statement LK has participated in the past 12 months and is currently participating as principal investigator, member, or chair of planning and steering committees or advisory boards in corporate-sponsored clinical trials in multiple sclerosis. The sponsoring pharmaceutical companies for these trials include Allozyne, Amgen, Bayer HealthCare Pharmaceuticals, Bayer-Schering Pharma, Bayhill, Biogen Idec Inc, Eisai, Elan Pharmaceuticals Inc, Genmab, Genzyme, Merck-Serono, Medicinova, Novartis, Sanofi-Aventis, Shire, Roche, Teva, UCB, and Wyeth. LK has also lectured at medical conferences or in public on various aspects of the diagnosis and management of multiple sclerosis. Often, these talks have been sponsored by non-restricted educational grants from one of the sponsors. Honoraria and other payments for all these activities have been exclusively used for funding of research at LK’s department. Research and the clinical operations (nursing and patient care services) of the Multiple Sclerosis Centre, led by LK, have been supported by non-restricted grants from one or more of the sponsors and by grants from the Swiss MS Society, the Swiss National Research Foundation, the European Union, and the Gianni Rubatto Foundation. RG has received speaker’s honoraria, consultation honoraria, and scientific grants from Biogen Idec Inc. DHM has received research grants (held by University College London) from Biogen Idec Inc, GlaxoSmithKline, Schering AG, and Novartis to do MRI analysis in multiple sclerosis trials. DHM has received honoraria and travel expenses for advisory committee work or as an invited speaker from Biogen Idec Inc, GlaxoSmithKline, Bayer Schering, and the US National Institutes of Health. EH has participated in clinical studies sponsored by Biogen Idec Inc, Schering, Serono, Teva, Octapharma, and Novartis and has given talks at symposia supported by Biogen Idec Inc, Sanofi-Aventis, Schering, Serono, and Teva. VL has received research grants or honoraria for presentations within the past 3 years from Antisense Pharmaceuticals, Bayer, Biogen Idec Inc, GlaxoSmithKline, Janssen Cilag, Johnson & Johnson, Orion, Sanofi-Aventis, Serono, and Teva. CHP has served as a consultant for Biogen Idec Inc, Schering AG, Teva, Serono, Novartis, GlaxoSmithKline, UCB, AstraZeneca, Roche, and Antisense Therapeutics; received lecture fees from Biogen Idec Inc, Schering AG, Novartis, and Teva; and received grant support from Biogen Idec Inc, Schering AG, GlaxoSmithKline, Novartis, Serono, and Teva. KS received research support from Biogen Idec Inc to do the MRI analysis of this study.

1471

Articles

TAY has served as a consultant for Biogen Idec Inc, and has received grants from Biogen Idec Inc and GlaxoSmithKline. KD, MY, GNO, and AWS are employees of and own stock in Biogen Idecc. ME, DGM, EM, and IR declare that they have no conflict of interest. Acknowledgments The authors acknowledge the contributions of J Licata and M Hasson (Scientific Connexions, Newtown, PA, USA) for editorial assistance in preparing this manuscript for submission. Their work was funded by Biogen Idec Inc. References 1 Lukashev M, Zeng W, Ryan S, et al. Activation of Nrf2 and modulation of disease progression in EAE models by BG00012 (dimethyl fumarate) suggests a novel mechanism of action combining anti-inflammatory and neuroprotective modalities. Mult Scler 2007; 13 (suppl 2): S149. 2 Itoh K, Chiba T, Takahashi S, et al. An Nrf2/small Maf heterodimer mediates the induction of phase II detoxifying enzyme genes through antioxidant response elements. Biochem Biophys Res Commun 1997; 236: 313–22. 3 Venugopal R, Jaiswal AK. Nrf2 and Nrf1 in association with Jun proteins regulate antioxidant response element-mediated expression and coordinated induction of genes encoding detoxifying enzymes. Oncogene 1998; 17: 3145–56. 4 Chen XL, Dodd G, Thomas S, et al. Activation of Nrf2/ARE pathway protects endothelial cells from oxidant injury and inhibits inflammatory gene expression. Am J Physiol Heart Circ Physiol 2006; 290: H1862–70. 5 Calabrese V, Ravagna A, Colombrita C, et al. Acetylcarnitine induces heme oxygenase in rat astrocytes and protects against oxidative stress: involvement of the transcription factor Nrf2. J Neurosci Res 2005; 79: 509–21. 6 Li J, Johnson D, Calkins M, Wright L, Svendsen C, Johnson J. Stabilization of Nrf2 by tBHQ confers protection against oxidative stress-induced cell death in human neural stem cells. Toxicol Sci 2005; 83: 313–28. 7 Satoh T, Okamato SI, Cui J, et al. Activation of the Keap1/Nrf2 pathway for neuroprotection by electrophillic phase II inducers. Proc Natl Acad Sci USA 2006; 103: 768–73. 8 Vargas MR, Pehar M, Cassina P, et al. Fibroblast growth factor-1 induces heme oxygenase-1 via nuclear factor erythroid 2-related factor 2 (Nrf2) in spinal cord astrocytes: consequences for motor neuron survival. J Biol Chem 2005; 27: 25571–79. 9 Zhao C, Zawadzka M, Roulois AJ, Bruce CC, Franklin RJ. Promoting remyelination in multiple sclerosis by endogenous adult neural stem/precursor cells: defining cellular targets. J Neurol Sci 2008; 265: 12–16. 10 Hubbs AF, Benkovic SA, Miller DB, et al. Vacuolar leukoencephalopathy with widespread astrogliosis in mice lacking transcription factor Nrf2. Am J Pathol 2007; 170: 2068–76. 11 Wierinckx A, Brevé J, Mercier D, Schultzberg M, Drukarch B, Van Dam AM. Detoxification enzyme inducers modify cytokine production in rat mixed glial cells. J Neuroimmunol 2005; 166: 132–43. 12 Loewe R, Pillinger M, de Martin R, et al. Dimethylfumarate inhibits tumor-necrosis-factor-induced CD62E expression in an NF-κB-dependent manner. J Invest Dermatol 2001; 117: 1363–68. 13 Schilling S, Goelz S, Linker R, Luehder F, Gold R. Fumaric acid esters are effective in chronic experimental autoimmune encephalomyelitis and suppress macrophage infiltration. Clin Exp Immunol 2006; 145: 101–07. 14 Altmeyer PJ, Matthes U, Pawlak F, et al. Antipsoriatic effect of fumaric acid derivatives. Results of a multicenter double-blind study in 100 patients. J Am Acad Dermatol 1994; 30: 977–81. 15 Schimrigk S, Brune N, Hellwig K, et al. Oral fumaric acid esters for the treatment of active multiple sclerosis: an open-label, baseline-controlled pilot study. Eur J Neurol 2006; 13: 604–10. 16 McDonald WI, Compston A, Edan G, et al. Recommended diagnostic criteria for multiple sclerosis: guidelines from the International Panel on the Diagnosis of Multiple Sclerosis. Ann Neurol 2001; 50: 121–27. 17 Kurtzke JF. Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology 1983; 33: 1444–52.

1472

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

Kappos L, Lechner-Scott J, Lienert C. Interactive training DVD-ROM for a standardised, quantified neurological examination and assessment of Kurtzke’s Functional Systems and Expanded Disability Status Scale in multiple sclerosis. Basel, Switzerland: Point de Vue Audiovisuelle Produktionen, 2003. http://www. neurostatus.net/training (accessed Jan 4, 2008). Sormani MP, Rovaris M, Bagnato F, et al. Sample size estimations for MRI-monitored trials of MS comparing new vs standard treatments. Neurology 2001; 57: 1883–85. Katz D, Taubenberger JK, Cannella B, McFarlin DE, Raine CS, McFarland HF. Correlation between magnetic resonance imaging findings and lesion development in chronic, active multiple sclerosis. Ann Neurol 1993; 34: 661–69. Brück W, Bitsch A, Kolenda H, Brück Y, Stiefel M, Lassmann H. Inflammatory central nervous system demyelination: correlation of magnetic resonance imaging findings with lesion pathology. Ann Neurol 1997; 42: 783–93. Kappos L, Moeri D, Radue EW, et al. Predictive value of gadolinium-enhanced magnetic resonance imaging for relapse rate and changes in disability or impairment in multiple sclerosis: a meta-analysis. Gadolinium MRI Meta-analysis Group. Lancet 1999; 353: 964–69. O’Connor P, Miller D, Riester K, et al; International Natalizumab Trial Group. Relapse rates and enhancing lesions in a phase II trial of natalizumab in multiple sclerosis. Mult Scler 2005; 11: 568–72. Richert ND, Howard T, Frank JA, et al. Relationship between inflammatory lesions and cerebral atrophy in multiple sclerosis. Neurology 2006; 66: 551–56. van Walderveen MAA, Barkhof F, Pouwels PJW, van Schijndel RA, Polman CH, Castelijns JA. Neuronal damage in T1-hypointense multiple sclerosis lesions demonstrated in vivo using proton magnetic resonance spectroscopy. Ann Neurol 1999; 46: 79–87. Bo L, Dawson TM, Wesselingh S, et al. Induction of nitric oxide synthase in demyelinating regions of multiple sclerosis brains. Ann Neurol 1994; 36: 778–86. van der Goes A, Brouwer J, Hoekstra K, Roos D, van den Berg TK, Dijkstra CD. Reactive oxygen species are required for the phagocytosis of myelin by macrophages. J Neuroimmunol 1998; 92: 67–75. Barnes J, Karin M. Nuclear factor kappa B—a pivotal transcription factor in chronic inflammatory diseases. N Engl J Med 1997; 336: 1066–71. LeVine SM. The role of reactive oxygen species in the pathogenesis of multiple sclerosis. Med Hypothesis 1992; 39: 271–74. Zagórski T, Dudek I, Berkan L, Mazurek M, Kedziora J, Chmielewski H. Superoxide dismutase (SOD-1) activity in erythrocytes of patients with multiple sclerosis. Neurol Neurochir Pol 1991; 25: 725–30. Langemann H, Kabiersch A, Newcombe J. Measurement of low-molecular-weight antioxidants, uric acid, tyrosine and tryptophan in plaques and white matter from patients with multiple sclerosis. Eur Neurol 1992; 32: 248–52. Liu Y, Zhu B, Wang X, et al. Bilirubin as a potent antioxidant suppresses experimental autoimmune encephalomyelitis: implications for the role of oxidative stress in the development of multiple sclerosis. J Neuroimmunol 2003; 139: 27–35. The IFNB Multiple Sclerosis Study Group. Interferon beta-1b is effective in relapsing-remitting multiple sclerosis. I. Clinical results of a multicenter, randomized, double-blind, placebo-controlled trial. Neurology 1993; 43: 655–61. Johnson KP, Brooks BR, Cohen JA, et al. Copolymer 1 reduces relapse rate and improves disability in relapsing-remitting multiple sclerosis: results of a phase III multicenter, double-blind, placebocontrolled trial. Neurology 1995; 45: 1268–76. Jacobs LD, Cookfair DL, Rudick RA, et al. Intramuscular interferon beta-1a for disease progression in relapsing multiple sclerosis. Ann Neurol 1996; 39: 285–94. PRISMS Study Group. Randomised double-blind placebo-controlled study of interferon β-1a in relapsing-remitting multiple sclerosis. Lancet 1998; 352: 1498–504.

www.thelancet.com Vol 372 October 25, 2008