Conceived and designed the experiments: RD MTH TM SZ WF ACL. Performed the experiments: RD MTH TM SZ JK WF FS EL MO TG AH ASW UB RB BS CW JG ACL. Analyzed the data: LD JK MO JD ACL. Wrote the paper: LD.
¶ Membership of the GERP ALS study group is provided in the Acknowledgments.
Takeda Pharma GmbH partly funded this study. The trial used Pioglitazone which is a Takeda product. There are no other patents, products in development or marketed products to disclose. This does not alter the authors’ adherence to all the PLOS ONE policies on sharing data and materials.
Pioglitazone, an oral anti-diabetic that stimulates the PPAR-gamma transcription factor, increased survival of mice with amyotrophic lateral sclerosis (ALS).
We performed a phase II, double blind, multicentre, placebo controlled trial of pioglitazone in ALS patients under riluzole. 219 patients were randomly assigned to receive 45 mg/day of pioglitazone or placebo (one: one allocation ratio). The primary endpoint was survival. Secondary endpoints included incidence of non-invasive ventilation and tracheotomy, and slopes of ALS-FRS, slow vital capacity, and quality of life as assessed using EUROQoL EQ-5D. The study was conducted under a two-stage group sequential test, allowing to stop for futility or superiority after interim analysis. Shortly after interim analysis, 30 patients under pioglitazone and 24 patients under placebo had died. The trial was stopped for futility; the hazard ratio for primary endpoint was 1.21 (95% CI: 0.71–2.07, p = 0.48). Secondary endpoints were not modified by pioglitazone treatment. Pioglitazone was well tolerated.
Pioglitazone has no beneficial effects on the survival of ALS patients as add-on therapy to riluzole.
Clinicaltrials.gov
Amyotrophic lateral sclerosis (ALS) is a lethal neurological disease with limited therapeutic options. Only riluzole demonstrated efficacy in prolonging survival of ALS patients
Pioglitazone is an oral anti-diabetic that stimulates the transcriptional activity of peroxisome proliferation activated receptors (PPARs), most notably of the PPAR subtype. Pioglitazone possesses anti-inflammatory properties that might be beneficial in ALS patients
In all, pioglitazone represented a candidate drug able to act through multiple protective mechanisms. Here, we sought to test whether pioglitazone is beneficial in ALS by performing a phase II, multicentre, stratified, parallel-group, placebo-controlled trial of pioglitazone in ALS to assess the potential efficacy of pioglitazone as an add-on therapy to riluzole on survival as a primary endpoint. Incidence of non-invasive ventilation (NIV) and tracheotomy, and slopes of revised ALS functional rating scale (ALS-FRS-R), slow vital capacity (SVC), and quality of life (using EUROQoL EQ-5D; http:
This phase II clinical trial was designed as a multicentre, stratified, parallel-group, placebo-controlled trial of pioglitazone in patients with ALS as an add-on therapy to riluzole. The trial protocol can be accessed at
Patients with possible, probable (clinically or laboratory-supported) or definite ALS according to the revised version of the El Escorial World Federation of Neurology criteria were considered for enrolment into the study. Included patients displayed onset of progressive weakness within 36 months prior to study and had a disease duration of more than six months and less than three years (inclusive) with disease onset defined as date of first muscle weakness, excluding fasciculation and cramps. They reached a best-sitting SVC between 50% and 95% of predicted normal. They were capable of thoroughly understanding the information provided and giving full informed consent. Included women of childbearing age were non-lactating, and surgically sterile, or used a highly effective method of birth control, and had a negative pregnancy test. All included patients had been treated with 100 mg riluzole daily for at least three months prior to inclusion.
Exclusion criteria were: Participation in another clinical study within the preceding 12 weeks; tracheotomy or assisted ventilation during the preceding three months; gastrostomy; any medical condition known to have an association with motor neuron dysfunction which might confound the diagnosis of ALS; presence of any life-threatening disease or impairment likely to interfere with functional assessment; confirmed hepatic insufficiency or abnormal liver function (ASAT and/or ALAT >1.5 upper limit of normal); renal insufficiency (serum creatinine >2.26 mg/dL); evidence of major psychiatric disorder or clinically evident dementia; known hypersensitivity to the study drugs; patient likely to be not cooperative, not comply with the trial requirements (as assessed by the investigator), or unable to be reached in emergency; use of other antidiabetic drugs; heart failure, or history of heart failure (NYHA I to IV); history of macular oedema; treatment with thiazolidinediones within three months prior to screening; known or suspected history of alcohol and/or drug abuse; treatment with gemfibrozil within three months prior to screening.
All patients gave written informed consent. The ethics committee of the University of Ulm approved the study protocol.
The two treatment groups were 100 mg riluzole plus 45 mg pioglitazone (pioglitazone group), and 100 mg riluzole plus placebo (placebo group).
After inclusion, patients underwent a four-week screening phase, and a treatment phase (18 months). The dosage was increased stepwise: 15 mg once daily (od) during the first two weeks, 30 mg od during week three and four and 45 mg od from week five. Clinical and physical examinations, blood sampling, and drug compliance were recorded at on-site visits (1, 2, 6, 12 and 18 months after baseline visit). Body weight, and functional status (including NIV and ALS-FRS-R) were also recorded at 9 and 15 months after baseline visit through telephone contacts. The investigator observed patients for adverse events (AEs), and instructed patients to report any events. A period of 14 days for follow-up of AEs after the patient had routinely or prematurely terminated the study was performed additionally.
We sought to test whether pioglitazone 45 mg daily is beneficial in ALS patients under riluzole.
The primary endpoint was survival time,
Sample size was calculated according to a previously published clinical trial in ALS
Pioglitazone was originally developed as oral antidiabetic drug and clinical trials showed that pioglitazone was well tolerated. At present, it has been prescribed to over 3.5 million patients. Thus, the use of a one-sided statistical test for primary efficacy analysis was considered as acceptable. Assuming an accrual time of six months, a treatment period of 18 months, and a constant accrual rate, a total of 176 subjects (88 subjects per group) was expected to yield the necessary number of deaths. Under these assumptions, the time points of analyses were expected to be 12.3 month’s observation time (interim analysis) and 24 months observation time (final analysis). Considering a dropout rate of 10%, inclusion of 196 patients was planned.
At the randomisation visit, each patient eligible for study participation was randomised to one of the two treatment groups, and received the next consecutive randomisation/patient number according to her/his stratum from a block of randomisation numbers per site. The code was broken only for interim analysis, and at the end of study.
Haupt Pharma Brackenheim GmbH (Brackenheim, Germany) generated the randomisation list. Block-randomisation with a block size of four was performed. A pseudo-random number generator ensured that the resulting treatment sequence was reproducible. Patients were assigned to their stratum according to the location of the earliest ALS symptom. Patients with simultaneous spinal and bulbar onset were defined as bulbar. Cervical and respiratory onsets were stratified to the spinal-onset stratum.
The trial was double-blinded. Study medication was packed and blinded by Haupt Pharma Brackenheim.
The study population was analysed according to the intention-to-treat principle. All randomized patients who received at least one intake of study medication were analysed for efficacy and safety.
To investigate efficacy, the unstratified log rank test was used to compare groups. The statistical hypotheses in terms of the hazard ratio were the following:
H0: λ2/λ1> = 1 (i.e. identical or poorer survival in the pioglitazone group).
H1: λ2/λ1<1 (i.e. better survival in the pioglitazone group),
Where λ2/λ1 is the hazard ratio, λ1 denotes the hazard in the placebo group; λ1 denotes the hazard in the pioglitazone group.
The overall type-I-error rate was set at α = 0.025 (one-sided).
For comparison of baseline characteristics, we performed group comparisons using Mann-Whitney test for continous items and χ2 test for binary outcomes. Binary secondary outcomes (incidence of tracheotomy or NIV) were calculated both as total number and as percentage including the 95% confidence interval and group comparisons were performed using the χ2 test. For continuous secondary outcomes (ALSFRS-R, EUROQoL EQ-5D) group comparisons were performed using mixed effects regression model analysis. All secondary endpoints were analysed descriptively. All statistical tests performed were two-sided at a significance level of 5%. All results from analysis of the secondary endpoints were regarded as hypothesis generating only, and not as proof of efficacy.
An independent statistician performed the interim analysis at July 17th 2009 at the Institute of Epidemiology and Medical Biometrics at the University of Ulm. The interim analysis was performed with data from all available patients on June 10th, 2009 (197 patients, pioglitazone n = 99, placebo n = 98). Interim analysis was initially planned after observation of 31 deaths. However, because of the unexpected low number of deaths from start of trial until March 2009, the protocol was amended to perform the interim analysis either after observation of a total of 31 deaths or in June 2009, depending on what condition happens first. Statistical analyses were performed at the Institute of Epidemiology and Medical Biometrics at the University of Ulm using the statistical software package SAS Version 9.2 under Windows.
Between the 29th of May 2008 and the 14th of August 2009, 219 ALS patients were enrolled, and randomly allocated to either placebo (n = 110; bulbar: n = 33, spinal: n = 77), or pioglitazone (n = 109; bulbar: n = 32, spinal: n = 77) treatment after stratification based on site of onset (bulbar or spinal). One patient in the placebo group did not take any dose of study medication and was therefore excluded from the study.
The figure presents the numbers of participants who were randomly assigned, received pioglitazone or placebo, and were analysed for the primary outcome.
Baseline characteristics of both groups were similar (
Pioglitazone (n = 109) | Placebo (n = 109) | ||
Sex ratio (M/F) | 63/46 | 71/38 | 0.33 |
Age at enrolment (years) | 58.9 (10.6) | 59.0 (10.4) | 0.94 |
Site of onset (bulbar/spinal) | 31/77 | 33/76 | 0.88 |
ALSFRS-R | 37.5 (6.0) | 37.0 (5.6) | 0.51 |
SVC (%) | 76.2 (16.6) | 73.5 (15.4) | 0.21 |
Body weight (kg) | 72.8 (12.7) | 74.1 (13.1) | 0.46 |
BMI (kg/m2) | 24.7 (3.8) | 24.8 (3.7) | 0.84 |
Glycaemia (mg/dL) | 95.9 (13.3) | 94.3 (14.8) | 0.40 |
ASAT (U/L) | 35.8 (12.1) | 32.7 (11.3) | 0.06 |
ALAT (U/L) | 40.1 (17.9) | 37.8 (21.3) | 0.39 |
Blood pressure (systolic/diastolic, mm Hg) | 133.2 (16.9)/83.0 (9.5) | 138.5 (19.7)/87.0 (11.3) | 0.07/0.02 |
Time from symptom onset to diagnosis (months) | 9.6 (5.6) | 8.7 (5.7) | 0.25 |
Time from symptom onset to baseline screening (months) | 18.9 (8.6) | 18.6 (9.0) | 0.81 |
Values are indicated as mean (SD).
In interim analysis, which was performed about one year after start of recruiting patients, we observed death of 24 patients, 14 in the pioglitazone group and 10 in the placebo group. There was no difference between pioglitazone group and placebo group. Since the number of deaths was unexpectedly low, the data monitoring and safety committee (DMSC) requested continuation of the trial. Number of deaths and death rate were provided monthly and the DMSC recommended to stop the trial for futility on the 28th of April 2010 due to the increased death rate in the pioglitazone group (pioglitazone, 30 deaths; placebo, 24 deaths). In final statistical analysis, the hazard ratio was 1.21 (95%CI: 0.71–2.07,
Kaplan-Meier plot for survival, the primary endpoint of the trial. Placebo-treated patients are in blue and pioglitazone treated patients are in red. Ticks represent censored patients. The shaded box indicates the first month of treatment during which a stepwise increase in pioglitazone dosage was performed. Numbers below the X axis indicate the number of patients still alive (“at risk”,
Secondary variables for efficacy were also non-significantly affected by pioglitazone treatment. ALS-FRS-R score and slope was not affected by pioglitazone (
Total changes in ALS-FRS-R score after initiation of the treatment. Placebo-treated patients are in blue and pioglitazone treated patients are in red. The table below indicates the number of patients (n) per time point. Error bars are standard errors.
Pioglitazone was well tolerated and most adverse events were due to ALS disease progression,
Adverse event | Pioglitazone (n = 109) | Placebo (n = 109) | |
Dysphagia | 32 (29.4%) | 34 (31.2%) | 0.88 |
Dyspnoea | 35 (32.1%) | 31 (28.4%) | 0.66 |
Respiratory failure | 33 (30.3%) | 31 (28.4%) | 0.88 |
Oedema | 26 (23.9%) | 18 (16.5%) | 0.24 |
Depression | 14 (12.8%) | 13 (11.9%) | 1.00 |
Weight loss | 18 (16.5%) | 9 (8.3%) | 0.10 |
Constipation | 15 (13.8%) | 11 (10.1%) | 0.53 |
Fatigue | 12 (11.0%) | 14 (12.8%) | 0.83 |
Nasopharyngitis | 13 (11.9%) | 12 (11.0%) | 0.83 |
Fall | 11 (10.1%) | 11 (10.1%) | 1.00 |
Back pain | 9 (8.3%) | 10 (9.2%) | 1.00 |
Musculoskeletal pain | 9 (8.3%) | 7 (6.4%) | 0.80 |
Headache | 7 (6.4%) | 7 (6.4%) | 1.00 |
Muscle spasms | 7 (6.4%) | 7 (6.4%) | 1.00 |
Salivary hyper secretion | 6 (5.5%) | 8 (7.3%) | 0.78 |
Sleep disorder | 6 (5.5%) | 8 (7.3%) | 0.78 |
Bronchitis | 7 (6.4%) | 5 (4.6%) | 0.77 |
Dizziness | 8 (7.3%) | 4 (3.7%) | 0.37 |
Pain in extremity | 10 (9.2%) | 2 (1.8%) | 0.03 |
Pneumonia | 7 (6.4%) | 5 (4.6%) | 0.77 |
Urinary tract infection | 9 (8.3%) | 3 (2.8%) | 0.13 |
Nausea | 6 (5.5%) | 5 (4.6%) | 1.00 |
Pain | 6 (5.5%) | 5 (4.6%) | 1.00 |
In this randomized clinical trial, we could not show that pioglitazone is a valid therapeutic option for ALS patients under riluzole.
A number of clinical trials have been performed in ALS since the positive trial of riluzole (
In the specific case of pioglitazone, three groups independently reported protection by pioglitazone in mutant SOD1 mice
Summarizing, we could not find any protective effect of pioglitazone in ALS patients under riluzole. The cause for this failed clinical translation remains elusive.
(DOC)
(DOC)
(PDF)
We thank the patients and families.
Investigators are listed by alphabetical order of centre and investigator:
Nadja Borisow; Theresa Holm; Andre Maier; Thomas Meyer
Paula Budde; Torsten Grehl; Anne-Katrin Guettsches
Malte Bewersdorff; Michael Heneka
Andreas Hermann; Alexander Storch
Tobias Frank; Bettina Göricke; Jochen Weishaupt
Katharina Eger; Frank Hanisch; Stephan Zierz
Anna-Lena Cordes; Reinhard Dengler; Sonja Koerner; Katja Kollewe; Susanne Petri
Julian Grosskreutz; Albrecht Kunze; Tino Prell; Thomas Ringer; Jan Zinke
Johanna Anneser; Gian Domenico Borasio; Christine Chahli; Andrea S. Winkler
Matthias Boentert; Bianca Stubbe-Draeger; Peter Young
Ulrich Bogdahn; Steffen Franz; Verena Haringer; Norbert Weidner
Reiner Benecke; Stefanie Meister; Johannes Prudlo; Matthias Wittstock
Johannes Dorst; Corinna Hendrich; Albert C. Ludolph; Anne-Dorte Sperfeld; Ulrike Weiland
Sabine Neidhardt; Berthold Schrank
Marcus Beck; Peter Kraft; Klaus Toyka; Jochen Ulzheimer; Carsten Wessig