Minocycline 1-year therapy in multiple-system-atrophy: Effect on clinical symptoms and [11C] (R)-PK11195 PET (MEMSA-trial)

Authors


  • Potential conflict of interest: The authors report no conflicts of interest and have no financial disclosures to make in respect to this article.

Abstract

The aim of the study was to investigate the efficacy of the antibiotic minocycline as a drug treatment in patients with Multiple-System-Atrophy Parkinson-type (MSA-P). Sixty-three patients were randomized to minocycline 200 mg/d (n = 32) or a matching placebo (n = 31). The primary outcome variable was the change in the value of the motor score of the Unified Multiple-System-Atrophy Rating-Scale (UMSARSII) from baseline to 48 weeks. Secondary outcome variables included subscores and individual Parkinsonian symptoms as determined by the UMSARS and the Unified-Parkinson's-Disease Rating-Scale (UPDRS). Health-related quality of life (HrQoL) was assessed using the EQ-5D and SF-12. “Progression rate” was assumed to be reflected in the change in motor function over 48 weeks. At 24 weeks and 48 weeks of follow-up, there was a significant deterioration in motor scores in both groups, but neither the change in UMSARSII nor in UPDRSIII differed significantly between treatment groups, i.e. “progression rate” was considered to be similar in both treatment arms. HrQoL did not differ among the two treatment arms. In a small subgroup of patients (n = 8; minocycline = 3, placebo = 5)[11C](R)-PK11195-PET was performed. The three patients in the minocycline group had an attenuated mean increase in microglial activation as compared to the placebo group (P = 0.07) and in two of them individually showed decreased [11C](R)-PK11195 binding actually decreased. These preliminary PET-data suggest that minocycline may interfere with microglial activation. The relevance of this observation requires further investigation. This prospective, 48 week, randomized, double-blind, multinational study failed to show a clinical effect of minocycline on symptom severity as assessed by clinical motor function. © 2009 Movement Disorder Society

Multiple system atrophy (MSA) is a sporadic neurodegenerative disorder that usually manifests in the early fifties and progresses continuously with a mean survival of 6 to 9 years. Clinically, the primary features of the disease include autonomic failure, a Parkinsonian syndrome, cerebellar ataxia, and pyramidal signs.1 Parkinsonian symptoms predominate in 80% of patients (MSA-P) and cerebellar ataxia is the main feature in approximately 20% of the patients (MSA-C). Pathologically, MSA is characterized by a neuronal multisystem degeneration and abnormal glial cytoplasmic inclusions (GCI) containing alpha-synuclein aggregates.2–4 Recently, it was reported that the distribution of microglial activation in MSA is consistent with the known pattern of neurodegeneration and is correlated with the GCI burden in affected areas.2 Microglia have direct toxic effects through the secretion of inflammatory cytokines, complement proteins or free radical-related toxic molecules.5, 6 Indirect toxic effects, e.g. mediated through NO release by astrocytes, may also contribute to the propagation of neuronal cell death.7 Whether these effects are relevant to MSA is far from being understood. Experimental evidence, however, suggests that inhibition of microglial activation may interfere with the progression of neurodegeneration and of the clinical symptoms.2, 8 Therefore, therapeutic trials with agents that interfere with microglial activation and their downstream events are warranted in MSA-patients.

Minocycline is a semisynthetic second-generation tetracycline that exerts anti-inflammatory effects separate and distinct from its antimicrobial action.9,10 Minocycline passes the blood brain barrier and it has been shown to have neuroprotective effects in several animal models of neurodegenerative disorders and ischemia.11–13 Minocycline induced reduction in infarct size and increased survival of hippocampal neurons after ischemia. Cell death was significantly reduced in the MPTP-induced degeneration of dopaminergic neurons and treatment with minocycline delayed mortality in the R6/2-mouse model of Huntington's disease.12, 14 We and others have shown that minocycline inhibits microglial activity and their downstream event such as secretion of cytokines.12, 15–19

Although some experimental studies have questioned the neuroprotective effect of minocycline,20, 21 the evidence that has accumulated supports the need to investigate minocycline in a clinical trial.22–27 Long-term clinical trials have reported the safety of minocycline even during long treatment periods.10

We now report the results of a prospective, placebo-controlled, randomised double-blind multicenter clinical trial involving the oral administration of 200 mg/d of minocycline for 48 weeks in patients with MSA-P (MEMSA-trial: minocycline's efficacy for MSA patients). In addition, a subgroup of patients was investigated with [11C](R)-PK11195-PET, a marker of microglial activation.

MATERIALS AND METHODS

Patient Recruitment and Assessment

Patients were recruited from specialized Movement Disorder Centres. Inclusion criteria consisted of a clinical diagnosis of clinically probable MSA-P.28 The exclusion criteria were severe dementia, clinically relevant hepatic, renal, or cardiac dysfunction, a diagnosis of epilepsy, a history of seizures as an adult, stroke or TIA within the last year, significant skin hypersensitivity or hypersensitivity to tetracyclines, pregnancy, or nursing. After the baseline evaluation, adjustments in medication were made only if considered necessary to maintain an optimal clinical response.

Ethics committee approval was obtained, the study was announced to the competent authorities and informed consent was provided by each patient. Permission to administer radiotracers was granted by the ARSAC, UK.

Study Design

The study followed a multicentric, randomized, two-arm, prospective, double-blind, placebo-controlled design. The ratio of those randomized to minocycline versus those given placebo treatment was 1:1. Randomization lists with permutated blocks of randomly varying size were prepared before the start of the trial for each center. Randomization was carried out centrally by the Coordinating Center for Clinical Trials, Marburg.

Interventions

Minocycline was obtained from ratiopharm (Germany). Several studies have reported variable in vitro effects of minocycline, mainly due to the light sensitivity of this drug. We tested several batches of minocycline in an in vivo animal model of ischemia to assure the activity of minocycline (data not shown). The same active batch of minocycline was used throughout the study. The capsules were provided in light opaque boxes.

Patients were randomly assigned to orally receive either 2 × 50 mg minocycline or the matching placebo twice daily for a period of 48 weeks. Visits took place at baseline, 4, 8, 16, 24, 32, 40, 48, and 56 weeks. Drug intake was closely monitored in a drug accountability log.

Efficacy and Safety Evaluations

The primary efficacy endpoint was the change from baseline to 48 weeks in part II of the Unified Multiple-System-Atrophy Rating-Scale (UMSARS), for which higher scores indicate worse mobility.29 Secondary efficacy endpoints included the evolution over time (baseline, 24 weeks, 48 weeks) of further disease-specific scales including the UMSARS (I, III)29 and the Unified Parkinson's Disease Rating Scale (UPDRSII-III).30 Health-related quality of life (HrQoL) was measured by the SF-12 as well as the EQ-5D.31

Safety was assessed in terms of the frequency and severity of reported adverse events from randomization until the final visit on week 56. Any new symptoms or worsening of preexisting symptoms were classified as adverse events.

Imaging

Eight patients underwent [11C](R)-PK11195-PET at baseline and after 24 weeks of treatment. [11C](R)-PK11195-PET is a marker of peripheral benzodiazepine sites, which are selectively expressed by activated microglia.32

PET was performed and analysed as previously described.32 Briefly, datasets were acquired using an ECAT EXACT HR++(CTI/Siemens 966) camera. Thirty seconds after the start of the emission scan, 305.6 MBq ± 13.6 MBq (mean ± SD) of [11C](R)-PK11195 in 5 mL saline were infused intravenously over 10 seconds. Three-dimensional sinograms of emission data were then acquired over 60 minutes as 18 time frames.

Binding potentials (BP), measures of specific binding of the tracer (Bmax/Kd, the available concentration of binding sites/receptor dissociation constant), were calculated on a voxel by voxel basis using a simplified reference tissue model.33, 34 Because MSA is associated with a widespread distribution of pathological changes, cluster analysis was used to extract and identify a normal brain grey matter reference input function for individual MSA cases as previously described.8, 35

MRI

Volumetric T1-weighted MR-images were acquired at baseline for the purpose of coregistration with PET in order to define anatomical regions of interest (1.0 Tesla Picker HPQ scanner; Picker).

The volumes of interest were then applied to the corresponding [11C](R)-PK11195 binding potential maps (at baseline and follow up) using ANALYZE software36 and automated three-dimensional co-registration.37 The quality of co-registration was checked by visual inspection.

All analyses were performed by the same investigator, who was blinded to the treatment groups.

The differences between baseline and follow-up BP values were weighted by region size (number of voxels in each region) and then used as a global measure of microglial activity.

Statistical Analysis

The original protocol required that UPDRSIII be used as a primary endpoint and UMSARSII as a secondary endpoint. In order to adjust for baseline, the difference over the period from baseline to 48 weeks was used. The sample size was set to 30 evaluable patients in each group in order to obtain 80% power in a two-sided Wilcoxon-Mann-Whitney test with a significance level of 0.05 and based on assumptions of slowing down the clinical progression of the disease by 2.4 points with a SD of 4 (data were obtained from a prospective natural history study on progression in MSA patients; assumed UPDRSIII deterioration of 9 ± 4 points in the placebo group and 6.6 ± 4 in the minocycline group). Based on emerging data on the favourable validity of UMSARS,30 the primary endpoint was changed to the difference from baseline to 48 weeks in UMSARSII by an amendment to the protocol four months (i.e. 16 weeks) after randomisation of the first patient. Due to the similarities between the items of the UMSARSII and UPDRSIII no change in planned effect sizes was expected and hence the planned sample size was kept.

The confirmatory analysis of the primary endpoint was performed in the intention-to-treat population, consisting of all patients who were randomized regardless of the treatment they actually received. The changes in UMSARSII from baseline to 48 weeks were compared between groups using a two-sided Wilcoxon-Mann-Whitney test at a significance level of 0.05. Missing differences were ranked at the end of the list, i.e. imputed by worst values. Given the high drop out rate (see Results) we performed additional sensitivity analyses best-value and last-observation-carried-forward imputations, a complete-case analysis as well as corresponding analyses of the per-protocol population, which was defined as patients with a treatment period of at least 24 weeks.

Analysis of secondary endpoints was merely descriptive. Linear mixed models with therapy and time as fixed effects and patient-specific intercepts were fitted for longitudinal data. All tests were performed as two-sided and at a significance-level of α = 0.05. Statistical analyses were performed with the SAS statistical package (9.1).

The study was registered online (http://www.clinicaltrials.gov/):NCT00146809.

RESULTS

Characteristics of the Patients

A total of 63 patients underwent randomization (FIG. 1). Among these, 32 patients were assigned to receive minocycline (50.8%), and 31 to receive placebo (49.2%) (Table 1).

Figure 1.

Study assignment and outcome.

Table 1. Baseline Characteristics of MSA-P patients enrolled in the MEMSA-trial by randomization assignmenta
 Minocycline group (n = 32)Placebo group (n = 31)Difference and 95% CIa
  • a

    Absolute number, percentage and risk difference for categorical variables, mean ± standard deviation, range and difference in means for continuous variables.

  • CI, confidence interval.

Gender   
 Male17 (53%)12 (39%)14.5 (−10.0 to 39.0)
 Female15 (47%)19 (61%)−14.5 (−39.0 to 10.0)
Age (yr)63 ± 7 (40 to 72)61 ± 8 (45 to 75)1.2 (−2.5 to 4.9)
Time since diagnosis (mo)19 ± 23 (0.3 to 111)14 ± 13 (0.5 to 50)4.5 (−4.9 to 14.0)
Levodopa-equivalent dose (mg)587 ± 425 (0 to 1588)495 ± 465 (0 to 1400)92 (−133 to 316)

Most patients had coexisting medical conditions at baseline. The most common were cardiovascular (30%), genitourinary (29%), endocrine (29%), and musculoskeletal disorders (24%). All randomized patients received at least one dose of the study drug except for one patient in the minocycline group who withdrew his consent.

The mean daily equivalent levodopa dose38 at baseline in the minocycline group was 587 ± 425 mg and in the placebo group was 495 ± 465 mg. By week 4, this had increased to 684 ± 657 mg in the minocycline group and 581 ± 511 mg in the placebo group. The increase over time was statistically significant (P = 0.010), but did not differ between groups (P = 0.463).

A total of 63 patients were included in the efficacy analysis (Fig. 1). Twenty-three patients in the minocycline group and ten in the placebo group discontinued treatment prematurely. Adverse events were the primary reason for discontinuation, resulting in the withdrawal of eight patients in the minocycline group and six patients in the placebo group. Further reasons for premature discontinuation in the minocycline group were insufficient therapeutical effect (n = 4), withdrawal of consent (n = 3), and other (n = 3). Three patients in the minocycline group and two in the placebo group died.

Outcome

The UMSARSII score increased significantly from baseline to 48 weeks in both the minocycline (8.2 ± 6.3, 95% CI: 4.9–11.4) and the placebo group (7.0 ± 7.1, 95% CI: 3.7–10.2). This change in UMSARSII was not significantly different between the groups (P = 0.1750) (Table 2). Likewise, the UPDRSIII increased significantly within each group (minocycline: 9.1 ± 7.6, 95% CI: 5.2–13.0; placebo: 7.2 ± 6.5, 95% CI: 4.2–10.1), but this change did not differ significantly between groups (P = 0.8260). Results of the analysis of the secondary efficacy variables are depicted in Table 2. Similar results to those of the primary efficacy variables were found in the remaining subscores of the UPDRS and UMSARS.

Table 2. Results of the efficacy variables, health-related quality of life as measured by EQ-5D and SF-12 in the treatment and placebo group
Efficacy variables Minocycline groupPlacebo group 
  • a

    Wilcoxon-Mann-Whitney-Test of worst-value imputed differences from baseline to week 48 (confirmatory analysis of primary endpoint).

  • b

    Wald-Test of therapy-time interaction after fitting a linear mixed model with therapy and time as fixed effects and patient-specific intercepts using all available time points (to test whether there is a difference in the rate of progression between groups).

  • RRs, systolic blood pressure difference standing vs. supine; RRD, diastolic blood pressure difference standing vs. supine.

  BaselineW24W48BaselineW24W48P value
UMSARS II (primary efficacy variable)n3225173127210.1750a
mean ± SD21.9 ± 7.526.6 ± 7.529.5 ± 9.223.7 ± 6.227.4 ± 7.929.6 ± 8.6 
median24.027.030.024.027.030.0 
range8.0 to 34.09.0 to 42.010.0 to 44.013.0 to 35.015.0 to 42.015.0 to 43.0 
UMSARS In3225173127210.6211b
mean ± SD20.4 ± 6.623.9 ± 7.728.2 ± 9.320.0 ± 6.822.7 ± 6.425.0 ± 6.9 
median19.027.029.022.024.026.0 
range6.0 to 30.012.0 to 36.09.0 to 43.08.0 to 30.09.0 to 33.011.0 to 36.0 
UMSARS III RRSn3121122921180.2854b
mean ± SD−19.2 ± 24.0−14.7 ± 21.1−14.3 ± 17.3−17.4 ± 26.7−25.4 ± 23.5−11.4 ± 22.0 
median−11.0−10.0−10.0−20.0−20.0−16.0 
range−80.0 to 20.0−70.0 to 20.0−50.0 to 10.0−70.0 to 40.0−73.0 to 10.0−45.0 to 26.0 
UMSARS III RRDn2220122421180.b
mean ± SD−8.5 ± 15.8−7.3 ± 14.3−5.9 ± 10.5−7.8 ± 9.1−7.0 ± 26.1−1.9 ± 12.9 
median−10.0−6.0−2.5−10.0−10.00.0 
range−45.0 to 20.0−44.0 to 11.0−21.0 to 10.0−23.0 to 10.0−40.0 to 90.0−25.0 to 20.0 
UMSARS III orthostatic symptomsn312412312220 
yes (%)6 (19.4%)4 (16.7%)3 (25.0%)9 (29.0%)4 (18.2%)4 (20.0%) 
UPDRS IIn3225173127210.6359b
mean ± SD19.8 ± 7.622.9 ± 8.627.6 ± 8.820.4 ± 7.222.5 ± 7.425.7 ± 7.9 
median18.024.025.021.024.028.0 
range6.0 to 34.010.0 to 34.017.0 to 41.08.0 to 35.05.0 to 31.011.0 to 44.0 
UPDRS IIIn3225173127210.8260b
mean ± SD25.0 ± 8.029.3 ± 8.233.6 ± 9.326.8 ± 6.930.4 ± 8.233.0 ± 9.2 
median27.029.035.027.030.035.0 
range9.0 to 36.09.0 to 42.011.0 to 46.013.0 to 38.012.0 to 44.016.0 to 45.0 
EQ-5D scoren3121103018200.1352b
mean ± SD55.4 ± 17.040.0 ± 20.443.6 ± 23.650.2 ± 20.239.4 ± 16.341.0 ± 19.6 
median59.929.143.452.132.932.9 
range16.6 to 97.719.9 to 97.713.2 to 84.116.6 to 97.717.2 to 68.113.2 to 76.0 
EQ-5D VASn3020103017200.3064b
mean ± SD46.6±20.543.2±21.349.1±21.248.1±14.145.0±18.545.3±14.6 
median41.048.054.050.040.042.5 
range0.0 to 90.08.0 to 90.018.0 to 75.020.0 to 70.020.0 to 70.025.0 to 80.0 
SF-12 Physical Component Summaryn291992817170.2911b
mean ± SD39.4 ± 7.839.4 ± 9.638.1 ± 6.636.5 ± 7.839.3 ± 6.537.6 ± 7.0 
median40.638.636.034.938.537.8 
range26.3 to 59.324.6 to 57.031.7 to 53.626.0 to 53.229.7 to 52.527.4 to 57.4 
SF-12 Mental Component Summaryn291992817170.2040b
mean ± SD42.0 ± 6.642.3 ± 6.641.4 ± 9.740.4 ± 8.640.4 ± 6.945.0 ± 7.2 
median42.841.936.842.039.844.6 
range27.7 to 55.632.3 to 57.926.5 to 54.121.9 to 57.026.7 to 53.930.3 to 55.1 

Treatment with minocycline was not associated with a better HrQoL outcome (Table 2). Patients deteriorated from 0.55 to 0.44 in the EQ-5D index-score in the minocycline group and from 0.50 to 0.41 in the placebo group. No significant change was found in the EQ-5D-VAS nor in the SF-12 at 48 weeks as compared with baseline in both treatment groups.

To explore the effect of imputation on UMSARSII analyses, we performed sensitivity analyses that lead to results consistent with findings reported above (see additional online material). Additional analyses of the effect of centre did not reveal patterns that altered the interpretation of the results.

Imaging

In a previous study8 by us MSA patients showed increased [11C](R)-PK11195 binding especially in the basal ganglia, midbrain, and pons compared to controls. When comparing binding at baseline and follow-up, we observed individual increases in binding potential in the majority of placebo group patients. One of the three patients in the active treatment group also showed a small increase of microglial activation but the two others showed decreased [11C](R)-PK11195 binding (Wilcoxon-test P = 0.07; FIG. 2). The small number of patients in the PET study precluded a detailed statistical analysis, but we did not detect an association between [11C](R)-PK11195 binding and clinical markers such as duration of disease, UMSARSII or UPDRSIII, age, gender, levodopa dose in either the treatment or the placebo group at baseline, follow-up or in the incremental difference between these.

Figure 2.

Transverse sections of [11C](R)-PK11195 binding potential maps coregistered to the individual MRI in patient M-3 at baseline (a) and after 24 weeks of treatment with minocycline (b). At baseline increased [11C](R)-PK11195 binding is evident especially in the putamen, thalamus, and frontal cortex bilaterally (arrows). Following treatment with minocycline microglial activation is reduced in most regions. The colour bar denotes binding potential values from 0 to 1. The table provides patient demographics and clinical variables as well as the difference in mean regional binding potential values in the minocycline and placebo groups at baseline and follow-up after 6 months.

Tolerability and Safety

The prevalence of adverse events are summarized in Table 3. To address whether adverse events occurred more frequently in one or the other study arm, we calculated the difference in absolute risks and 95% confidence intervals. We only found a significant difference between the placebo treated and the minocycline treated group in eye disorders and psychiatric disorders (Fig. 1) However, two patients of the minocycline group reported worsening of Parkinsonian features during the treatment with minocycline, while no such reports were documented in the placebo-group. One patient in the minocycline group ended therapy prematurely due to skin-hyperpigmentation and weight-loss.

Table 3. Number of patients with Adverse Events by MedDRA System Organ Classa in both arms of the study
 Minocycline group%Placebo group%Difference in absolute risks with 95% CI
  • a

    Sixty-two patients, 31 in the minocycline and 31 in the placebo arm, were evaluable for safety analysis. The table shows the number of patients with at least one adverse event in a given MedDRA System Organ Class. Positive risk differences indicate higher risk in the minocycline group.

Blood and lymphatic system disorders13.200.03.2 (−3.0 to 9.5)
Cardiac disorders39.739.70.0 (−14.7 to 14.7)
Congenital, familial and genetic disorders00.013.2−3.2 (−9.5 to 3.0)
Ear and labyrinth disorders13.213.20.0 (−8.8 to 8.8)
Endocrine disorders00.013.2−3.2 (−9.5 to 3.0)
Eye disorders00.0412.9−12.9 (−24.7 to −1.1)
Gastrointestinal disorders1238.71032.36.5 (−17.3 to 30.2)
General disorders and administration site conditions1135.51032.33.2 (−20.3 to 26.8)
Immune system disorders13.213.20.0 (−8.8 to 8.8)
Infections and infestations1445.21754.8−9.7 (−34.5 to 15.1)
Injury, poisoning and procedural complications929.0929.00.0 (−22.6 to 22.6)
Metabolic and nutrition disorders13.213.20.0 (−8.8 to 8.8)
Musculoskeletal and connective tissue disorders1445.2825.819.4 (−4.0 to 42.7)
Neoplasms benign, malignant and unspecified (incl. cysts and polyps)13.226.5−3.2 (−13.9 to 7.4)
Nervous system disorders929.01135.5−6.5 (−29.7 to 16.7)
Psychiatric disorders412.91238.7−25.8 (−46.6 to −5.0)
Renal and urinary disorders412.9722.6−9.7 (−28.5 to 9.2)
Reproductive system and breast disorders13.213.20.0 (−8.8 to 8.8)
Respiratory, thoracic, and mediastinal disorders39.7412.9−3.2 (−19.0 to 12.5)
Skin and subcutaneous tissue disorders1238.7722.616.1 (−6.5 to 38.7)
Surgical and medical procedures39.739.70.0 (−14.7 to 14.7)
Vascular disorders516.1929.0−12.9 (−33.5 to 7.7)

DISCUSSION

This randomized double blind placebo controlled study investigated the therapeutic efficacy of minocycline in MSA-P. We failed to show any beneficial effect of minocycline, either symptomatic or neuroprotective. Treatment with minocycline failed to improve global ratings of motor function, such as the UMSARSII (primary endpoint) and UPDRSIII (secondary endpoint) over the 48 week observation period. The differences in the UMSARSII-scores between baseline and 48 weeks were 8.2 and 7.0 in the minocycline and placebo group, respectively. This decline in motor function within each group from baseline to 48 weeks was significant and thus the observation period of this trial was determined to be adequate. However, our clinical trial did not support the positive findings from laboratory studies.

Secondary efficacy variables included subscores of the UMSARS (I,III), the UPDRS and more broadly defined outcomes, such as the health-related quality of life. With these scales, similarly, we also failed to find differences between the treatment groups.

The profile of adverse events observed was compatible with those in earlier reports on long-term treatment with minocycline. Toxic reactions similar to those documented in an earlier study were not found.10 More than half (n = 33; 72% of the minocycline group vs. 32% of the placebo group, P = 0.002) of our patients withdrew from the trial and five patients died during the study period (three receiving minocycline, two receiving placebo). This high discontinuation rate points to feasibility problems of long-term minocycline treatment in this population. In another recent trial, growth hormone was tested in MSA patients and an equally high discontinuation rate was recorded (54% in the GH group and 29% in the placebo group).39 Similarly, in a futility trial in Parkinson's disease patients 21% discontinued treatment in the minocycline group compared to 6% in the calibration placebo arm, in most cases due to gastrointestinal discomfort.40, 41 Though the rate of discontinuation in the present study is unexpectedly high and differs from earlier observations of non-neurological patient populations, it may reflect the advanced stage of MSA in these patients when they entered the study and illustrates the difficulties that long term studies of rapidly progressive neurodegenerative disorders such as MSA may face.

Due to the relatively small sample size and the high discontinuation rate, we performed additional sensitivity analyses of the primary endpoint. Although these analyses were in line with the (negative) result of the main predefined analysis, we cannot exclude, that minocycline may have a deleterious effect on the progression of MSA - as was described recently in patients with amyotrophic lateral sclerosis25 - if a larger number of patients had been included into the trial. On the other hand, the patients we ascertained may have been too advanced to respond to any mild neuroprotective effect of minocycline. Further analyses using latent class analysis are currently underway to uncover subpopulations that might respond differently to the intervention.

With [11C](R)-PK11195-PET, an established marker of microglial activation we found a reduction in microglial activation over 24 weeks8, 42, 43 in two out of three patients receiving minocycline. No patients in the placebo group experienced such a reduction. In contrast, a mean increase of 22% in PK11195 binding was observed in the placebo group at 24 weeks. None of the clinical ratings, however, showed a correlation with PK11195 binding. This could reflect the lower power of the trial, however, other interpretations are possible:

  • 1.Assuming that microglial activation is positively associated with neurodegeneration, it may be possible that [11C](R)-PK11195 binding does not selectively label those microglial cells that are associated with deleterious effects on neuronal survival.
  • 2.Although a strong correlation has been found between the main pathways of neurodegeneration in MSA-brains and microglial activation in histopathological sections,2 it may be that microglial activation is a secondary response and not a primary source for neurodegeneration. Thus, inhibiting microglial activation may not help in adapting or inhibiting events leading to neurodegeneration.
  • 3.As suggested by several preclinical studies, minocycline has a strong effect on microglial activation, an assertion that is supported in part by our in vivo [11C](R)-PK11195-PET data.11, 16–19, 44 The effect of minocycline in our study, however, may be too short in duration or the effect may not be strong enough to inhibit the pathways downstream of microglial activation.

Several additional interpretations can be envisioned, however, more detailed studies are necessary.

In conclusion, minocycline failed to positively influence the clinical progression of advanced MSA-P over an observation period of 48 weeks. This result is similar to other prospective trials that investigated the potentially protective role of anti-inflammatory drugs (e.g. NSAIDS) in neurodegenerative disorders. Minocycline (200 mg/d) may induce an effect on microglial activation in the brain of MSA-P patients as demonstrated by [11C](R)-PK11195-PET. The meaning, however, remains unclear.

Although contradictory results have arisen from several clinical studies,40, 41 the evidence available at present does not support a clinically relevant effect of minocycline in neurodegenerative disorders.

Acknowledgements

We gratefully acknowledge the significant contributions of our patients with Multiple System Atrophy to the completion of this study. This study received support from the German Ministry of Education and Research through award number BMBF Nr. 01GI9901, 01GI0201, 01GI040, by the European MSA Study Group (www.emsa-sg.org), by the Willy Robert Pitzer foundation (16/03; K.E.), and by the PDS UK MAP 02/04 (A.G.).

Author Roles:

Silke Böttger, data acquisition, revising of the text; David Brooks: conception and design, data acquisition and analysis, editing of the text; Christine Daniels: data acquisition, revising of the text; Günther Deuschl: conception and design, data acquisition, revising of the text; Richard Dodel: conception and design, data acquisition and analysis, drafting, editing of the text; Yansheng Du: data acquisition and analysis, revising of the text; Karla Eggert: conception and design, data acquisition, revising of the text; Thomas Gasser: conception and design, data acquisition, revising of the text; Alexander Gerhard: conception and design, data acquisition or analysis, drafting, editing of the text; Felix Geser: data acquisition, revising of the text; Doreen Gruber: data acquisition, revising of the text; Birgit Herting: data acquisition, revising of the text; Christoph Kamm: data acquisition, revising of the text; Thomas Klockgether: conception and design, revising of the text; Manja Kloss: data acquisition, revising of the text; Martin Köllensperger: data acquisition, revising of the text; Martin Krause: data acquisition, revising of the text; Andreas Kupsch: data acquisition, revising of the text; Axel Lipp: data acquisition, revising of the text; Markus Naumann: data acquisition, revising of the text; Wolfgang Oertel: conception and design, data acquisition and analysis, drafting, revising of the text; Werner Poewe: data acquisition, design, revising of the text; Silvia Reinecker: conception and design, data analysis, revising of the text; Alexander Reuss: conception and design, data analysis, drafting, editing of the text; Martin Sawires: data acquisition, revising of the text; Schade-Brittinger Carmen: conception and design, data acquisition or analysis, revising of the text; Nicole Schimke: data acquisition, revising of the text; Klaus Seppi: conception and design, data acquisition revising of the text; Friederike Sixel-Döring: data acquisition, revising of the text; Annika Spottke: conception and design, data acquisition or analysis, drafting, editing of the text; Claudia Trenkwalder: conception and design, data acquisition, revising of the text; Federico Turkheimer: conception and design, data acquisition or analysis, revising of the text; Gregor Wenning: conception and design, data acquisition revising of the text.

Full Financial Disclosure for the Previous 12 Months: Böttger Silke: nothing to declare; Brooks David: Consultancies and Honoraria with/from Takeda, Orion Pharma, Safra Foundation, Merck Serono, Synosia, Teva, Eisai, Pfizer, GSK, Novartis, Elan; Daniels Christine nothing to declare; Deuschl Günther: Honoraria from Medtronic, Orion, Lundbeck, Teva. Grants from German Research Council, German Ministry of Education and Research, Medtronic; Dodel Richard: Consultancies with Rentschler, Lilly. Honoraria from Baxter, Bayer, Bohringer Ingelheim, GSK, Lundbeck, Merz, Novartis, Octapharma, Pfizer, Solvay. Grants from M.J. Fox, Rentschler, ZLB Behring; Du Yansheng: funding unrelated to this research; Eggert Karla: Consultancies with Orion Pharma, Schwarz Pharma Neuroscience (UCB), Solvay Pharmaceuticals, Valeant Pharmaceuticals International, Desitin. Advisory Boards for Orion Pharma, Schwarz Pharma Neuroscience (UCB), Valeant Pharmaceuticals International. Honoraria from Orion Pharma, Schwarz Pharma Neuroscience (UCB), Solvay Pharmaceuticals, Valeant Pharmaceuticals International, Desitin, Boehringer Ingelheim, GlaxoSmithKline. Grants from the German Ministry of Education and Health, from the Pitzer foundation and from the Rhön foundation; Gasser Thomas: Honoraria from Novartis, Merck Serono, Schwarz Pharma, Boehringer Ingelheim, Valeant Pharma. Grants from Novartis Pharma, German Research Ministery (BMBF, NGFN plus), German Research Ministery BMBF ERANET Neuron, Helmholtz Association: Helmholtz Alliance for Health in an Ageing Society (HELMA). Consultancies with Cefalon Pharma, Merck Serono; Gerhard Alexander: funding by GE Health; Geser Felix: nothing to declare; Gruber Doreen: nothing to declare; Herting Birgit: nothing to declare; Kamm Christoph: nothing to declare; Klockgether Thomas: Grants from German Research Council, German Ministry of Education and Research EU; Kloss Manja nothing to declare; Köllensperger Martin: Grants from the Medizinische Forschungsfond Innsbruck; Krause Martin: Grants from Parkinson Society Australia and the NHMRC; Kupsch Andreas: Honoraria from Medtronic, Boehringer-Ingelheim. Grants from German Research Council, German Ministry of Education and Research, Medtronic; Lipp Axel: nothing to declare; Naumann Markus: nothing to declare; Oertel Wolfgang: Stock ownership Roche 100, Consultancies with Proteosys, Novartis, Orion Pharma, Schwarz Pharma Neuroscience (UCB), Solvay Pharmaceuticals, Desitin, Synosia. Advisory Boards with Bayer-Schering, Bioprojet, Boehringer Ingelheim, Meda Pharmaceuticals International, Merck Serono, Neurosearch, Novartis, Orion Pharma, Schering Plough, Schwarz Pharma Neuroscience (UCB), Teva. Honoraria from Novartis, Orion Pharma, Schwarz Pharma Neuroscience (UCB), Solvay Pharmaceuticals, Meda Pharmaceuticals International, Desitin, Boehringer Ingelheim, GSK, Teva. Grants from German Ministry of Education and Health; Poewe Werner: lecture fees and consulting honoraria from Boehringer Ingelheim, GlaxoSmithKline, Lundbeck, Merck Serono, Novartis, Solvay, Teva, and UCB; Reinecker Silvia: Grants from German Research Council, German Ministry of Education and Research; Reuss Alexander: Grants from German Research Council, German Ministry of Education and Research; Sawires Martin: nothing to declare; Schade-Brittinger Carmen: Grants from German Research Council, German Ministry of Education and Research; Schimke Nicole: nothing to declare; Seppi Klaus: Advisory Boards for GlaxoSmithKline. Honoraria from Lundbeck; UCB Pharma; Novartis; Boehringer Ingelheim. Grants from Medical University Innsbruck; Sixel-Döring Friederike: Honoraria from Boehringer Ingelheim, Cephalon, GlaxoSmithKline, Medtronic, Orion Pharma, Roche Pharma, Schwarz Pharma. Advisory board for Orion Pharma and for Medtronic; Spottke Annika: nothing to declare; Trenkwalder Claudia: Consultancies for Novartis, Orion Pharm, Mundipharm. Advisory Boards for UCB, Boehringer Ingelheim. Honoraria from UCB, Boehringer Ingelheim, GSK, Teva, Grants from MJF Foundation. Expert Testimony for Mundipharm; Turkheimer Federico: Grants from International Joint Project The Royal Society, CRUK-EPSRC Imaging Centre; Wenning Gregor: Advisory board for Schwarz Pharma. Consultancies for Teva/Lundbeck. Honoraria from GSK, Teva/Lundbeck.

Ancillary