Mycophenolate mofetil in multiple sclerosis: a multicentre retrospective study on 344 patients
- L Michel1,2,
- S Vukusic3,4,5,6,
- J De Seze7,
- F Ducray3,4,5,6,
- J C Ongagna7,
- F Lefrère2,
- M Jacq-Foucher2,
- C Confavreux3,4,5,6,
- S Wiertlewski1,2,
- D A Laplaud1,2
- 1Service de Neurologie, CHU Nantes, Nantes, France
- 2INSERM CIC004, CHU Nantes, Nantes, France
- 3Service de Neurologie A and Eugène Devic EDMUS Foundation, Hôpital Neurologique Pierre-Wertheimer, Hospices Civils de Lyon, Bron cedex, France
- 4INSERM U1028, Centre des Neurosciences de Lyon, Equipe Neuro-oncologie et Neuro-inflammation, Lyon, France
- 5CNRS UMR5292, Centre des Neurosciences de Lyon, Equipe Neuro-oncologie et Neuro-inflammation, Lyon, France
- 6Université Lyon I, Lyon, France.
- 7Service de Neurologie, University Hospital of Strasbourg, Strasbourg, France
- Correspondence to Dr L Michel, Service de Neurologie, Bld Jacques Monod, Saint Herblain, CHU Nantes, Nantes 44800, France;
- Received 27 February 2013
- Revised 19 April 2013
- Accepted 25 April 2013
- Published Online First 23 May 2013
Objectives Mycophenolate mofetil (MMF) is an immunosuppressive agent, sometimes used as a disease-modifying therapy for multiple sclerosis (MS). Several studies have reported the relative safety of this treatment but, to date, its efficacy has rarely been described. We performed a retrospective study to assess the safety and efficacy of MMF in patients with MS.
Methodology Three French MS centres included all of their patients treated by MMF. The main outcome criterion was annualised relapse rate (ARR) in the 1 year period after onset of MMF compared with the 1 year control period. Treatment with another immunosuppressive drug, such as mitoxantrone or cyclophosphamide, in the 2 years preceding initiation of MMF was included in a subgroup analysis. MMF safety and progression of the Expanded Disability Status Scale (EDSS) score were also assessed.
Results 344 patients were included; 149 patients were previously treated with another immunosuppressant (IS group). Mean MMF treatment duration was 25.3±1.1 months. During the 1 year control period, ARR was 1.11±0.08, and for the 1 year treatment period, ARR was reduced significantly to 0.35±0.05 (p<0.0001, Wilcoxon paired test). Adverse events (occurring in 11% of patients) were mainly digestive disorders, benign infections, asthenia and transitory lymphopenia. Concerning the progression of disability, in the subgroup of patients without previous immunosuppressant treatment, EDSS remained stable between initiation and 1 year after the beginning of MMF.
Interpretation Our results suggest that MMF can improve or stabilise MS patients and can be used as an alternative therapy.
Major medical advances have recently been made in the development of new therapies for multiple sclerosis (MS), an inflammatory and demyelinating disease of the CNS. Immunomodulatory drugs (such as interferon β or glatiramer acetate) have been demonstrated to be safe, but have mild efficacy and poor compliance due to injection related side effects. Immunosuppressive agents, such as mitoxantrone, natalizumab or fingolimod, have better efficacy and convenience but with worse risk profiles, leading to limitation of their prescription in more aggressive MS as secondline therapies (only in European countries for fingolimod).
Mycophenolate mofetil (MMF) is an inhibitor of purine synthesis in lymphocytes. It exerts broad effects on T cells, B cells and macrophages, such as suppression of lymphocyte proliferation,1 inhibition of interferon γ and interleukin 6 secretions,2 and inhibition of blood–brain barrier crossing.3 It is recommended for preventing allograft rejection4 but there is a growing body of evidence suggesting its beneficial effect for the treatment of several inflammatory autoimmune diseases.5–11
MMF has also been used for the treatment of MS in several case studies and in small short term clinical trials.12–16 However, the duration of these pilot studies was 1 year or less and included less than 20 patients per arm. Consequently, no data are currently available on the clinical efficacy of MMF used as a single treatment or on its long term safety. In France, MMF was widely used before the authorisation of oral drugs for MS. We therefore decided to perform a retrospective analysis of MMF efficacy and safety in three French MS centres.
Material and methods
Patient selection and data collection
Three French MS centres (Lyon, Strasbourg and Nantes university hospitals) participated in this study. These centres systematically collect data prospectively (including Expanded Disability Status Scale (EDSS) scores, and date and type of relapses) on all of their MS patients using specific software, the European Database for Multiple Sclerosis (EDMUS).17 Individual MS case reports include identification and demographic data, medical history, key episodes in the MS course (relapses, onset of the progressive phase), neurological examination at each visit and patient treatment at each visit. Data confidentiality and safety were ensured according to the recommendations of the French National Ethics Committee (CNIL-Commission Nationale Informatique et Libertés), which provided approval for the EDMUS database in each centre. At the time of the study, a cohort of 7498 patients had been included in the three MS centres (Nantes n=762 patients, Strasbourg n=1460 patients, Lyon n=5276 patients). Data concerning patients having received at least one MMF treatment were automatically extracted from these databases. In this cohort, all patients received MMF as the only disease modifying therapy. This treatment was proposed for several reasons, mainly adverse events with firstline therapies (flu-like symptoms, cutaneous intolerance), injection constraints refused by patients or lack of therapy with proven efficacy in progressive forms. At the time of the study (censoring date November 2008), recent treatments such as fingolimod or natalizumab were not or only rarely available. All types of clinical courses were included in the study. MS diagnosis was classified according to the 2005 revised Mc Donald criteria.18 An MS relapse was defined as the appearance, reappearance or worsening of symptoms of neurological dysfunction, lasting more than 24 h. Fatigue alone was not considered as a relapse. Disability was assessed with Kurtzke's EDSS.
Patients were considered lost to follow-up when no information was available in the year preceding the censoring date (1 November 2008). As treatment with another immunosuppressive (IS) drug, such as mitoxantrone or cyclophosphamide, in the 2 years preceding initiation of MMF could be considered a confounding factor because of the possible long term effect on relapses, this variable was included in subgroup analyses.
Efficacy of MMF was assessed by comparing the annualised relapse rate (ARR) in the 1 year period after onset of MMF with the 1 year period just before the beginning of MMF and a 1 year control period (between 24 months and 12 months before the beginning of treatment) (see online supplementary figure S1). The control period was used to avoid regression to the mean phenomenon because one may hypothesise that switching treatment was performed during a high risk period of relapse.
Other outcomes were: (i) sensitivity analysis comparing ARR for other periods of time (6 months) immediately before and after the beginning of MMF and during a similar control period 1 year before the beginning of MMF; and (ii) comparison of EDSS scores 1 year before initiation of MMF, at baseline and 1 year after MMF.
A safety profile was determined by analysing spontaneous reporting of adverse effects in the patients’ medical files. Each patient was systematically questioned about eventual side effects at each consultation.
Biological concerns were also taken into account as haemogram and liver function were regularly analysed (at least once a month) in the three MS centres at initiation and during treatment follow-up.
Statistical analysis was performed using Wilcoxon paired tests to compare ARR and EDSS before and after MMF. All results are presented as mean±SEM. Prism software V.4 was used for the Wilcoxon paired test. Results were considered significant when p<0.05.
A total of 344 patients with MS were included in the study, representing 4.6% of the whole MS population followed at that time in the three MS centres (n=7498). Their clinical and demographic characteristics are summarised in table 1. At the time of MMF initiation, clinical course was relapsing–remitting in 138 patients (40.1%), secondary progressive in 129 (37.5%), primary progressive in 61 (17.7%) and 16 patients (4.7%) had a clinically isolated syndrome. At the time of MMF initiation, mean disease duration was 10.9±0.5 years and mean EDSS score was 5.0±0.1. In this cohort, 149 (43%) patients had been treated with another IS drug (IS group) in the previous 2 years. In the IS group, mean time interval between stopping the IS drug and beginning MMF was 3.9±5.4 months. In the whole cohort, mean treatment duration with MMF was 25.3±1.1 months; in the IS and non-IS group, durations were 27.7±1.7 and 23.5±1.5 months, respectively. MMF was stopped in 122 (35%) patients because of side effects (n=26, 7.5%), treatment failure (n=24, 7%), patient convenience (n=20, 5.8%), scheduled termination (n=7, 2%), cancer (n=1, 0.2%), death (n=2, 0.4%) or for unknown reasons (n=42, 12.2%) (figure 1). Termination of MMF was equally observed in patients with a relapsing–remitting type of disease (50/155 patients, 32%) and in those with a chronic progressive form (72/189, 38%).
Efficacy of MMF on relapses
For the whole cohort of patients treated with MMF for 1 year periods, we observed a significant decrease in ARR from 1.11±0.08 for the control period to 0.35±0.05 during the treatment period (p<0.0001, Wilcoxon test) (figure 2A). This decrease was also significant compared with the ARR 1 year just before treatment (ARR=1.32±0.08, p<0.0001, Wilcoxon test) (figure 2A).
For the subgroup analysis, we observed a significant decrease in ARR in the non-IS group (patients not treated with IS drugs at least 2 years before the beginning of MMF) to 0.38±0.07 for the 1 year period of treatment compared with the 1 year control period (ARR 0.81±0.08, p<0.0001, Wilcoxon test) (figure 2B). This decrease was also statistically significant compared with the ARR 1 year before treatment (1.47±0.12, p<0.0001, Wilcoxon tests) (figure 2B). As expected, ARR in the year preceding the beginning of MMF was higher compared with the control period, explaining the reason for the start of disease modifying therapy.
For the IS group (patients treated with IS drugs in the 2 years before the beginning of MMF), we also observed a significant decrease in ARR during the 1 year period under treatment (0.32±0.06) compared with the ARR during the control period (1.42±0.14) and the ARR during the 1 year period before the start of treatment (1.14±0.11) (p<0.0001 for both comparisons, Wilcoxon tests) (figure 2C).
Regarding sensitivity analysis, for periods shorter than 6 months, the same results were found, with a significant decrease in ARR during the period of treatment compared with the control period or with the period just before the start of MMF (ARR in the 6 month period before MMF start=1.20±0.10, ARR during the control period=1.20±0.11 and ARR during the 6 month period after MMF start=0.33±0.05, p<0.0001 for both comparisons, Wilcoxon test) (figure 2A). For the non-IS group, we also found a significant decrease in ARR under treatment compared with the control period of 6 months or the 6 month period just before treatment (ARR in the 6 month period before MMF initiation 1.66±0.15, ARR during the control period 0.81±0.11 and ARR during the 6 month period after MMF start 0.33±0.07; p<0.0001, Wilcoxon tests for both comparisons) (figure 2B). For the IS group, we observed the same decrease in ARR under MMF (0.33±0.08) compared with the control period 1 year before the beginning of MMF (1.64±0.2, p<0.0001, Wilcoxon test) (figure 2C) or compared with the 6 month period just before MMF (0.67±0.11, p<0.05, Wilcoxon test) (figure 2C).
When taking into account only patients with a relapsing–remitting form of disease in the whole cohort or in the IS and non-IS subgroups, very similar results were found, with a significant decrease in ARR after introducing MMF (see online supplementary figure S2).
Efficacy of MMF on disability
For the whole cohort of patients treated with MMF, we found a slight increase in EDSS score from 4.96±0.13, 1 year before MMF start, to 5.00±0.10 at baseline (p=0.08, Wilcoxon test) (figure 3A). Then, 1 year after the start of MMF, EDSS score stabilised at 4.96±0.13 (figure 3A).
Regarding subgroup analysis, in the non-IS group, we observed a significant increase in EDSS score from 4.28±0.22 to 4.68±0.15 during the year preceding treatment initiation (p=0.03, Wilcoxon paired test). One year after beginning MMF, the EDSS score slightly decreased to 4.57±0.19 compared with 4.68±0.15 at the beginning of MMF (p=0.85, Wilcoxon paired test), suggesting stabilisation of disability progression by MMF (figure 3B). For the IS group, EDSS remained stable at the three different periods, with a score of 5.42±0.14, 1 year before MMF, 5.41±0.13 at initiation of MMF and 5.37±0.16, 1 year after the beginning of MMF (NS, Wilcoxon paired tests) (figure 3C).
Tolerance and safety during MMF treatment
An adverse event was mentioned in 11% of patient medical files (n=38), leading to definitive treatment discontinuation in 26 cases (7.5%) (table 2). Most of the adverse events were gastrointestinal symptoms (n=13, 3.7%), asthenia (n=11, 3.2%) and benign upper respiratory tract infections (n=6, 1.7%). Transient lymphopenia was also reported in six cases (1.7%), leading to definitive treatment discontinuation in half of the cases. Cutaneous eruption and insomnia were reported in two and three cases, respectively. Oedema, pruritus, vagal malaise and alopecia were observed in one case each. No severe adverse event was notified.
One patient was diagnosed with breast cancer. Three patients died at home during the treatment, with no clear explanation. Two of them had been treated with MMF for 24 and 43 months, respectively, at the time of death. For the third patient, MMF treatment was stopped 42 months before he died.
MMF is an immunosuppressive drug, commonly used for preventing graft rejection. This treatment has been suggested to be beneficial in patients with MS. However, because of the small cohorts and short duration of follow-up, it remains difficult to reach a conclusion concerning the efficacy and tolerance of MMF in MS patients. Our retrospective study is currently the largest analysing the efficacy and safety of MMF in MS patients, with a mean treatment duration of more than 2 years. In our study, MMF treatment was proposed for different reasons: refusal of injections, side effects with immunomodulatory drugs, failure of firstline therapies and lack of alternative therapies for progressive forms.
Concerning the main outcome, we observed a significant reduction in ARR after 6 months and 1 year of treatment compared with the period just before, but also in relation to a distant control period. This period was defined to minimise a potential regression to the mean effect when comparing with the period immediately preceding the introduction of MMF, in the whole cohort of patients as well as in predefined subgroups. We observed a significant reduction of 68.1% in ARR after 1 year of treatment in the whole group compared with the control period. This beneficial therapeutic effect on relapses confirms and strengthens some results already published in case studies or small clinical trials. Indeed, in 2001, Ahrens et al reported on the improvement or stability of disability progression in five of seven patients suffering from chronic progressive or relapsing MS and treated with MMF.12 More recently, Frohman et al showed similar radiological efficiency with good tolerability of MMF versus interferon β1a in a randomised 6 month study.15
None of these studies, probably because of their short durations, reported on the progression of disability. In our study, MS patients in the whole cohort presented with stabilisation of their EDSS score between baseline and after 1 year of MMF treatment and a slight increase in EDSS score between the control period and after 1 year of treatment using some Wilcoxon paired tests. However, if the beneficial effect of MMF on relapses seems to be clear, stabilisation of EDSS could be due to the significant decrease in ARR under treatment. Moreover, statistical analysis of disability progression has been performed in only 40 patients in the non-IS group and will have to be confirmed in a prospective study.
This study allowed us to confirm the good tolerance of MMF in MS. Indeed, adverse events presented by patients were minor and included mainly digestive disorders and asthenia. Only 7.5% of patients had to stop treatment because of these adverse events. In our cohort, we observed one case of breast cancer but its relation to MMF treatment is unlikely as it was started only 4 months before. Two patients died during MMF treatment, the cause of their deaths remaining unknown, but these patients were significantly disabled and no relation to treatment could be demonstrated. Among the whole cohort, 42 patients stopped treatment for unknown reasons. On the whole, considering that 122 patients (35%) stopped treatment, this could limit the use of MMF in patients with MS. However, such a discontinuation rate of therapy is also observed with interferon b in patients with MS.19 Our results concerning safety are in agreement with data found in previous studies: in 2004, a study of 79 MS patients treated with MMF suggested a good safety profile with potential therapeutic effects.13 Two other preliminary studies reported on the good tolerability and safety of a combination of MMF with interferon, but both were in small groups (24 and 30 MS patients, respectively) with a short duration of follow-up (1 year and 6 months, respectively).14 ,16
Our study has several limitations. We performed a retrospective analysis and consequently 42 patients were lost to follow-up during the study. Another limitation in the interpretation of the data may arise from the lack of a control group. We made the choice to study several control periods in the same cohort of patients rather than studying another cohort as a control group in order to decrease interindividual variability but also the mean regression phenomenon.
In conclusion, our results suggest a clear effect of MMF on MS relapses. Moreover, this study confirms that MMF as monotherapy is well tolerated in MS.
We thank Yohann Foucher from the Department of Biostatistics (Nantes University) for his help.
SW and DAL are co-last authors.
Contributors LM wrote the article. LM, DAL, SW, SV, CC, JdS, FD, JCO, FL and MJ-F participated in the collection of the data and revised the article critically. DAL, SW, JdS and SV helped in the design of the study. LM, DAL, JdS, SW, CC and SV helped in the analysis of the results. All authors gave their approval of the manuscript to be published.
Funding This work was supported by a public grant from the French Agence Nationale de la Recherche within the context of the ‘investments for the future program’ (ANR-10-COHO-002).
Competing interests JdS received lecture fees, consulting and research support from Biogen, Merck, Bayer, LFB, Sanofi-Aventis, Teva, Genzyme, Alergan UCB, AB Sciences and Almiral. SV received consulting and lecture fees, travel grants and research support from Bayer-Schering, Biogen Idec, Genzyme, Novartis, Merck Serono, Sanofi, Aventis and Teva Pharma. CC received consulting fees from Biogen Idec, Genzyme, Novartis, Merck Serono, Sanofi Aventis, Teva Pharma and UCB Pharma; lecture fees from Bayer-Schering, Biogen Idec, LFB, Merck Serono, Sanofi Aventis and Teva Pharma; and research support from Bayer-Schering, Biogen Idec, Genzyme, Merck Serono, Novartis, Sanofi-Aventis and Teva Pharma. DAL received honoraria for consulting and lectures from Novartis, Biogen-Idec, Genzyme and Teva. SW received honoraria for consulting and lectures from Novartis, Biogen-Idec, Sanofi Aventis and Teva.
Ethics approval Data confidentiality and safety were ensured according to the recommendations of the French National Ethics Committee (CNIL-Commission Nationale Informatique et Libertés), which provided approval for the European Database for Multiple Sclerosis (EDMUS) in each centre.
Provenance and peer review Not commissioned; externally peer reviewed.