Multiple system atrophy: current and future approaches to management

Abstract

Multiple system atrophy (MSA) is a rare neurodegenerative disorder without any effective treatment in slowing or stopping disease progression. It is characterized by poor levodopa responsive Parkinsonism, cerebellar ataxia, pyramidal signs and autonomic failure in any combination. Current therapeutic strategies are primarily based on dopamine replacement and improvement of autonomic failure. However, symptomatic management remains disappointing and no curative treatment is yet available. Recent experimental evidence has confirmed the key role of alpha-synuclein aggregation in the pathogenesis of MSA. Referring to this hypothesis, transgenic and toxic animal models have been developed to assess candidate drugs for MSA. The standardization of diagnosis criteria and assessment procedures will allow large multicentre clinical trials to be conducted. In this article we review the available symptomatic treatment, recent results of studies investigating potential neuroprotective drugs, and future approaches for the management in MSA.

Introduction

Multiple system atrophy (MSA) is a progressive neurodegenerative disorder leading to more severe disability and impairment in quality of life than in Parkinson’s disease (PD) [Schrag et al. 2006; Tison et al. 2002]. MSA is characterized by a variable combination of poor levodopa responsive Parkinsonism, cerebellar ataxia and autonomic failure [Gilman et al. 2008]. MSA-Parkinsonism type (MSA-P) predominates in the Western Hemisphere and MSA-cerebellar type (MSA-C) is the major phenotype in the Eastern Hemisphere [Watanabe et al. 2002]. There is no recognized efficient treatment for cerebellar ataxia whereas levodopa may transiently improve Parkinsonism in one third of the patients [Constantinescu et al. 2007; Wenning et al. 1997]. Since autonomic failure strongly correlates with poor quality of life [Schrag et al. 2006], therapeutic management focuses on symptomatic therapy of orthostatic hypotension (OH), constipation, genitourinary and breathing disorders. Mean survival is about 9–10 years after symptom onset [Schrag et al. 2008] with nocturnal sudden death being a major cause of death [Shimohata et al. 2008]. Hitherto, no treatment has been shown to improve survival in MSA, creating a strong need for new therapeutic approaches.

Abnormal aggregation of alpha-synuclein in oligodendrocytes is the pathological hallmark of MSA and may trigger the progressive cell loss in widespread brain areas [Jellinger and Lantos, 2010; Wenning et al. 2008, 1997]. Advances in therapeutics have been achieved during the past 10 years through the development of transgenic animal models and experimental strategies against alpha-synuclein accumulation and for neuroprotection [Stefanova et al. 2009]. This progress may provide a testbed for future candidate drugs discovery and multicentric clinical trials.

Here we review potential neuroprotective drugs, current symptomatic treatment and future approaches in the management of MSA.

Current therapies

A list of the currently symptomatic treatment for MSA is given in Table 1.

Table 1.

Available symptomatic treatment for multiple system atrophy (MSA). Almost all of the therapeutics currently used are based on expert experience and do not meet scientific evidence standards. Double-blind, placebo-controlled trials are indicated in references.

Feature	Current first-line treatment	Alternative treatment
Parkinsonism	Levodopa up to 1 g/day if tolerated, in association with domperidone and physiotherapy	Dopamine agonist, amantadine, paroxetine [Friess et al. 2006]
Cerebellar ataxia	Physiotherapy	Clonazepam, propanolol, baclofen, amantadine, gabapentin, buspirone
Orthostatic hypotension	Nonpharmacological measures (custom-fitted elastic stockings, raising the head of the bed when sleeping, water drinking, small meals). Midodrine from 2.5 to 30 mg per day [Wright et al. 1998; Low et al. 1997; Jankovic et al. 1993]a	Fludrocortisone, pyridostigmine,b droxidopa [Mathias, 2008],c [Kaufmann et al. 2003], antidiuretic hormone desmopressin at bedtime, ephedrine, octreotide
Neurogenic lower urinary tract dysfunction	If postvoid residual >100 ml, clean intermittent catheterization	Botulinum toxin A in the detrusor muscle or the urethral sphincter, surgery options, permanent catheterization
	If postvoid residual <100 ml, anticholinergic agents for detrusor hyperactivity, α-adrenergic antagonists for urethra hypertony
Constipation	High fluid and fibre intake classical laxative therapy, polycarbophil, mosapride citrate, macrogol 3350
Erectile dysfunction	Intracavernosal injection of papaverine or prostaglandin E1, sildenafil [Hussain et al. 2001]	Subcutaneous apomorphine injections
Breathing disorders	Continuous positive air pressure	Tracheostomy (for life-threatening and/or daytime stridor, abnormal vocal cord mobility on laryngoscopy), laryngeal surgery or botulinum toxin injections, adaptive servoventilation
Dystonia	Botulinum toxin injection, levodopa therapy	Anticholinergics, amantadine, dopamine agonists, muscle relaxants, tetrabenazine
Camptocormia	Physiotherapy with specific orthosis, wearing a backpack	Botulinum toxin injections, protirelin tartrate
Rapid eye movement sleep behavior disorder	Clonazepam	Sodium oxybate, temazepam, zopiclone, gabapentin, pramipexole, donepezil, melatonine
Depression	Psychotherapy	Electroconvulsive therapy
	Selective serotonergic reuptake inhibitors, levodopa therapy	Repetitive transcranial magnetic stimulation
Cognitive impairment	Speech therapy
Drooling	Anticholinergic drugs	Injection of botulinum toxin into the salivary glands

^aAmong all of the subjects included in these trials, there was a minority of MSA patients: 40/171 subjects [Low et al. 1997], 7/25 [Wright et al. 1998], and 18/97 [Jankovic et al. 1993]. In available subgroup analysis, beneficial effect on orthostatic hypotension (OH) was reported [Low et al. 1997; Jankovic et al. 1993].

^bPositive effect on OH regarding subgroup analysis in 17 MSA patients of the 58 subjects included, but there was no placebo group control [Singer et al. 2006].

^cPreliminary results.

Parkinsonism

A poor levodopa response is one of the consensus criteria for the diagnosis of MSA [Gilman et al. 2008] and helps to differentiate MSA from PD [Wenning et al. 2000]. Although there are no controlled clinical trials on the efficacy of levodopa in MSA, about one third of patients may experience a benefit, more often in the MSA-P subtype than in MSA-C [Constantinescu et al. 2007]. In series with pathological confirmation, the positive levodopa responsiveness is reported in 28–65% of the patients [Colosimo et al. 1995; Wenning et al. 1997, 1995; Hughes et al. 1992; Fearnley and Lees, 1990; Rajput et al. 1990]. However, this effect persists for several years (as in PD) in only 13% of all patients [Wenning et al. 1994]. Although levodopa induces less delirium and hallucinations in MSA than in PD [Wenning et al. 2000], it can lead to adverse effects such as worsening of OH or pathological hypersexuality [Klos et al. 2005; Wenning et al. 1994]. Moreover, early orofacial levodopa induced dyskinesias may occur in half of MSA patients often in the absence of any motor benefit [Boesch et al. 2002; Wenning et al. 1994]. According to consensus criteria, levodopa unresponsiveness should only be accepted after a treatment with at least 1 g of levodopa per day for at least 3 months [Gilman et al. 2008]. However, some patients report a subjective benefit that may not be apparent in a standardized motor examination, while others convey subsequent motor degradation after levodopa discontinuation [Wenning et al. 1994]. To date, there is no evidence suggesting that such high doses of levodopa accelerate the neurodegenerative process in transgenic animal models of MSA [Stefanova et al. 2007]. As in PD, domperidone can be added to prevent dopaminergic side effects such as nausea and vomiting.

There are no controlled studies that have tested the efficacy of dopamine agonists in MSA so far. In a retrospective case record analysis, only 10% of the patients experienced a benefit with dopamine agonists [Wenning et al. 1994]. Thus, dopamine agonists cannot be recommended as a first-line treatment given their high rate of side effects compared with levodopa, particularly worsening of OH. Amantadine may be an alternative symptomatic treatment. While a double-blind, placebo-controlled crossover trial failed to demonstrate any significant motor benefit in eight MSA patients [Wenning, 2005], a retrospective study disclosed good or excellent responsiveness in 15% [Wenning et al. 1994].

Although serotonergic involvement in MSA remains unclear, postmortem studies disclosed clear evidence for serotonergic depletion in brainstem nuclei [Tada et al. 2009; Benarroch, 2007]. The selective serotonin reuptake inhibitor paroxetine has been assessed for 2 weeks in a small trial including 19 MSA patients [Friess et al. 2006]. In this double-blind, placebo-controlled, randomized study, motor improvement was slight but significant in comparison with placebo. Further investigations are needed before any conclusion can be drawn on the efficacy of selective serotonin reuptake inhibitors in MSA. A double-blind, multicentre trial with fluoxetine is currently underway in France.

Cerebellar ataxia

Physiotherapy remains the best therapeutic option for cerebellar ataxia in MSA [Jain et al. 2004]. When intentional cerebellar tremor predominates, off-label use of low doses of clonazepam may sometimes help in our hands. Off-label use of propanolol, baclofen or amantadine have shown modest and transient efficacy in a retrospective data analysis [Wenning et al. 2003]. Buspirone (off-label) improved upper limb ataxia in an open-label trial including nine MSA-C patients [Heo et al. 2008]. Gabapentin (off-label) may have symptomatic benefits on ataxia, oscillopsia and dysarthria according to a limited report in two MSA-C patients [Gazulla and Benavente, 2005]. Finally, protirelin tartrate (off-label) has been used for ataxic diseases in Japan, but without any apparent effect in MSA [Takei et al. 2009].

Autonomic failure

Orthostatic hypotension

OH is one of the major criteria for MSA diagnosis [Gilman et al. 2008]. Spreading of the neurodegenerative process to the intermediolateral cell column [Wenning et al. 1997] underlies earlier and more severe autonomic failure in MSA compared with PD [Lipp et al. 2009]. First-line treatment should always include nonpharmacological advice: use of custom-fitted elastic stockings, raising the head of the bed by 20–30° when sleeping, performing isotonic exercises, as well as increasing intake of fluid, salt and frequency of small meals [Freeman, 2008]. Drinking more than 350 ml of water a day reduces both OH and orthostatic symptoms but long-term effects and acceptance are unknown [Deguchi et al. 2007; Young and Mathias, 2004].

A broad range of drugs has been assessed in OH, such as fludrocortisone, midodrine, ephedrine or octreotide [Wenning et al. 2003]. Only midodrine has been investigated with a strong, double-blind, placebo-controlled procedure in three studies including few patients with MSA and labelled in the indication of neurogenic OH [Wright et al. 1998; Low et al. 1997; Jankovic et al. 1993]. This α₁-adrenoceptor agonist alleviated moderate to severe OH with oral doses that ranged from 2.5 to 30 mg daily. Blood pressure should be regularly monitored in patients receiving midodrine to detect supine hypertension which may be prevented by taking the last pill more than 4 hours before bedtime [McClellan et al. 1998]. By enhancing sympathetic activity, pyridostigmine modestly but significantly improved OH without supine hypertension in a double-blind, randomized, crossover study [Singer et al. 2006]. Although almost exclusively used in Japan, droxidopa seems to be effective and well tolerated to treat OH in patients with MSA [Kaufmann, 2008; Mathias, 2008; Kaufmann et al. 2003; Mathias et al. 2001; Freeman et al. 1999].

Neurogenic lower urinary tract dysfunction

Urinary urgency, frequency or incomplete bladder emptying, and urinary incontinence remain key features of the consensual diagnosis criteria [Gilman et al. 2008]. Urinary disorders severely affect the quality of life in MSA patients and may lead to lower urinary tract and kidney infection as major causes of morbidity in MSA [Ito et al. 2006]. Clean intermittent self-catheterization (CISC) is recommended as the first-line treatment when the postvoid residual is above 100 ml [Fowler and O’Malley, 2003]. Thus, the residual volume should be regularly monitored, for example with a portable ultrasound device. The critical threshold of 100 ml is reached in a mean of 2 years after disease diagnosis [Ito et al. 2006]. If the postvoid residual is below 100 ml, drugs acting on the detrusor may be used [Papatsoris et al. 2008]. Bladder-oriented anticholinergic agents are often prescribed for detrusor hyperactivity symptoms (e.g. urinary urgency, frequency and urge incontinence) but with a risk of worsening urinary retention. Urethra-orientated α-adrenergic antagonists may reduce the residual urine volume but exacerbation of OH is likely to occur [Sakakibara et al. 2000]. If urge incontinence still persists, third-line treatments may be considered. Synthetic antidiuretic hormone desmopressin (off-label) decreases urinary volumes. Taken just before bedtime, desmopressin reduces nocturia and improves morning hypotension. Regular monitoring of drug safety is necessary since it may induce water intoxication [Sakakibara et al. 2003]. Injection of botulinum toxin A in the detrusor muscle may be an alternative treatment for detrusor overactivity given the positive results in two MSA patients [Giannantoni et al. 2009], as well as urethral sphincter injections for urethral hypertonia when alpha-blockers are not tolerated [Apostolidis et al. 2009]. Botulinum toxin injections (off-label) seem to be safe and effective but further trials with larger samples are needed in MSA. To reduce the risk of infection due to permanent catheterization, surgery such as sphincterotomy or sphincteric wall stenting can be considered as a last option in MSA [Ito et al. 2006; Fowler and O’Malley, 2003].

Constipation

In addition to exercise, high fluid and fibre intake, laxative therapy is often necessary to relieve constipation in MSA. As in PD, classical laxative treatment may be used. Two small open trials have demonstrated the efficacy of polycarbophil [Sakakibara et al. 2007] and mosapride citrate [Liu et al. 2005] to reduce the time of transit in MSA patients. Moreover, macrogol 3350 has been reported to be subjectively efficient in two MSA patients [Eichhorn and Oertel, 2001].

Erectile dysfunction

Intracavernosal injection of papaverine or prostaglandin E1 may be used for erectile dysfunction in MSA [Papatsoris et al. 2008]. Subcutaneous apomorphine administration may also induce erections in PD [O’Sullivan, 2002; O’Sullivan and Hughes, 1998]. Although its use has not been assessed in this indication in MSA patients, apomorphine is easier to handle than papaverine. The use of sildenafil in MSA often faces the problem of its cardiovascular side effects, mainly the risk of a severe blood pressure drop. Efficiency and safety of the latter was assessed in a randomized, double-blind, placebo-controlled, crossover study [Hussain et al. 2001]. Sildenafil enhanced erectile function in the six MSA patients studied, but half of them experienced a severe drop in blood pressure 1 hour after drug ingestion. The pre-existing severity of OH may be predictive for a further decrease in blood pressure and the appearance of clinical signs of orthostatism.

Breathing disorders

MSA may manifest with obstructive and central sleep apnoea, dysrhythmia, dyspnoea, hypoxaemia, inspiratory sighs and laryngeal stridor [Meissner et al. 2010; Gilman et al. 2008; Shimohata et al. 2007; Geser et al. 2005; Vetrugno et al. 2004; Chokroverty et al. 1978]. These breathing disorders reflect the extension of the neurodegenerative process to the pontomedullary respiratory nuclei [Benarroch, 2007]. Since respiratory disorders may manifest in the early stages of MSA [Glass et al. 2006], and may explain sudden death [Tada et al. 2009], they should be considered as a primordial aspect of management in MSA. Stridor is usually the consequence of vocal cord dystonia [Vetrugno et al. 2007; Merlo et al. 2002] but cases with vocal cord abduction muscle denervation have been documented [Hayashi et al. 1997; Bannister et al. 1981]. Its occurrence is known to be associated with poor outcome [Yamaguchi et al. 2003; Silber and Levine, 2000]. Invasive procedures should be considered after the failure of noninvasive ventilation. Tracheostomy may be discussed with the patient for life-threatening and/or daytime stridor, or when abnormal vocal cord mobility is diagnosed on laryngoscopic examination [Silber, 2008; Silber and Levine, 2000]. However, the latter do not seem to prevent sudden death [Shimohata et al. 2008; Jin et al. 2007]. The use of laryngeal surgery or botulinum toxin therapy is sometimes performed with contrasting effects in limited reports [Iranzo, 2005; Merlo et al. 2002].

Concerning obstructive sleep apnoea, treatment with continuous positive air pressure may reduce the apnoea/hypopnoea index and desaturation [Iranzo et al. 2004, 2000], but we found the acceptability to be low when disease progresses [Ghorayeb et al. 2005b]. Moreover, the mean survival does not seem to be modified according to retrospective and/or uncontrolled data [Shimohata et al. 2008; Ghorayeb et al. 2005b]. Finally, recent clinical evidence suggests that adaptive servoventilation may reduce the apnoea index in MSA patients with central sleep apnoea [Suzuki et al. 2010].

Movement disorders

Orofacial dystonia, antecollis and camptocormia are features supporting the consensual diagnosis of MSA and constitute ‘red flags’ for the differential diagnosis with PD [Gilman et al. 2008; Kollensperger et al. 2008].

Dystonia

In a prospective trial including 24 levodopa-naive patients with probable MSA, dystonia occurred in 46% [Boesch et al. 2002]. Botulinum toxin injection is the most commonly used treatment for focal dystonia [Papapetropoulos et al. 2008]. Although no controlled studies with any other drug are yet available, symptomatic relief with anticholinergics, amantadine, dopamine agonists, muscle relaxants or tetrabenazine has been reported in some MSA patients when used off-label [Papapetropoulos et al. 2008; Azher and Jankovic, 2005].

Camptocormia

Camptocormia (CC) is an abnormal forward flexion of the trunk that manifests when standing or walking and relieves in the supine position [Azher and Jankovic, 2005]. Whether CC is a dystonic posture, a consequence of axial rigidity or a focal myopathy is still debated [Margraf et al. 2010; Diederich et al. 2006; Slawek et al. 2006; Azher and Jankovic, 2005]. The levodopa responsiveness is generally poor [Bloch et al. 2006; Diederich et al. 2006; Azher and Jankovic, 2005], but some patients reported fair relief [Song et al. 2008; Slawek et al. 2006]. In a single case report, protirelin tartrate (off-label use) improved CC dramatically, presumably by enhancing motoneuronal excitability of the paraspinal muscles [Takei et al. 2009]. As in focal dystonia, botulinum toxin injections may be effective [Azher and Jankovic, 2005]. Despite the lack of placebo-controlled trials, injections may be more appropriate in rectus abdominus muscles than in lumbar paraspinal muscles because of the risk of transient aggravation of CC. Finally physiotherapy in combination with specific orthosis may improve CC and quality of life [de Seze et al. 2008]. Recent clinical evidence suggests that wearing a backpack may also alleviate CC in PD [Gerton et al. 2010]. Although the efficacy has not been specifically tested in MSA, such noninvasive options should be considered.

Rapid eye movement sleep behaviour disorder

Rapid eye movement sleep behaviour disorder (RBD) may be the presenting symptom of MSA [Tison et al. 1995] and is observed in 69–100% of systematic polysomnography recordings in MSA patients [Vetrugno et al. 2004; Plazzi et al. 1997; Manni et al. 1993]. Although no therapeutic strategy has been validated, the off-label use of clonazepam is consensually and by experience considered as the first-line treatment [Ghorayeb et al. 2005a]. Clonazepam may aggravate existing obstructive sleep apnoea, but alternative treatments are sparse. Sodium oxybate, temazepam, zopiclone, gabapentin and pramipexole were effective when used off-label in limited and uncontrolled single case reports [Anderson and Shneerson, 2009]. Donepezil, a centrally acting reversible acetylcholinesterase inhibitor, is expected to alleviate RBD (off-label use) [Ringman and Simmons, 2000], but its reported clinical efficacy is variable [Boeve et al. 2003]. Melatonin may restore RBD-related desynchronization of the circadian rhythms without major side effects [Boeve et al. 2003], but these data await confirmation.

Nonspecific therapies

Depression

Depression is more prevalent in MSA than in PD [Tison et al. 2006] and correlates with poor quality of life [Schrag et al. 2006]. Selective serotonergic reuptake inhibitors are the most prescribed antidepressive treatment, with less expected risk of inducing OH than tricyclic drugs. Electroconvulsive therapy may be considered in MSA patients with melancholia [Shioda et al. 2006]. Repetitive transcranial magnetic stimulation may also have antidepressant efficacy in PD but the effects in MSA are unknown [Fregni et al. 2004]. Psychological support is necessary, as well as nursing care and family education [Hardy, 2008]. Finally, levodopa may slightly improve mood disorders in MSA [Fetoni et al. 1999].

Cognitive impairment

MSA patients may have impaired visuospatial and constructional function, verbal fluency and executive function compared with control subjects [Kitayama et al. 2009; Kawai et al. 2008]. No specific cognition enhancement treatment has been assessed in MSA.

Medical rehabilitation

For cerebellar ataxia and Parkinsonism, medical rehabilitation seems to improve balance and motor impairment [Landers et al. 2009; Wedge, 2008; Jain et al. 2004] as does the practice of tai chi [Venglar, 2005]. Speech therapy may be necessary to improve communication because of severe dysarthria.

Drooling

Excessive drooling may be a problem in later disease stages. Anticholinergic drugs may be efficient but adverse effects are frequent, including dry mouth, cognitive impairment, constipation, blurred vision or urinary retention. Injection of botulinum toxin type A into the salivary glands is the treatment of choice in excessive drooling as central side effects can be avoided [Mancini et al. 2003]. However, it may aggravate swallowing problems which may require transient nasogastric tube feeding.

Neuroprotective strategies

Although recent advances in basic science have given some clues for neuroprotective strategies in MSA patients, all clinical trials failed to show any disease-modifying properties (Table 2). Experimental evidence in a rodent model of MSA suggested that the antiglutamate drug riluzole may delay neuronal loss [Diguet et al. 2005]. No positive effect was noted in two prospective trials performed in MSA, using validated clinical scales and survival time as outcomes [Bensimon et al. 2009; Seppi et al. 2006].

Table 2.

Neuroprotective drugs for multiple system atrophy (MSA): study failures and ongoing trials.

Putative neuroprotective drugs that have been assessed in MSA	Putative neuroprotective drugs under clinical assessment
Riluzole [Bensimon et al. 2009; Seppi et al. 2006]	Lithium
Minocycline [Dodel et al. 2010]	Rasagiline
Growth hormone [Holmberg et al. 2007]	Intravenous immunoglobulins
Oestrogen [Heo et al. 2008]

To test the hypothesis that minocycline inhibits microglial activation, which is supposed to contribute to the progressive cell death in MSA, a 48-week prospective study was performed in 63 MSA-P patients [Dodel et al. 2010]. Although a subgroup analysis of eight patients revealed a nonsignificant decrease in [11C](R)-PK11195-PET binding, a tracer of glial cells including microglia, astrocytes and infiltrating macrophages, there was no change in clinical measures of motor impairment or health-related quality of life.

Since growth hormone (GH) may also act as a neuronal ‘survival factor’ in MSA, a randomized, double-blind, placebo-controlled trial has been designed with 22 MSA patients receiving recombinant human growth hormone (r-hGH) injection during 1 year and 21 receiving placebo [Holmberg et al. 2007]. Although this study disclosed no significant effect, there was a trend to a smaller increase in Unified Parkinson’s Disease Rating Scale (UPDRS) and Unified Multiple System Atrophy Rating Scale (UMSARS) scores over time as well as in OH and cardiovascular variability. Since the use of r-hGH appears to be safe, further trials with higher doses and more patients are needed.

Finally, in MSA-C patients, the hypothesis that oestrogen could have neuroprotective properties has not been confirmed in a recent trial [Heo et al. 2008].

Deep brain stimulation

Although bilateral subthalamic stimulation may have beneficial effects in a few MSA patients [Visser-Vandewalle et al. 2003], a recent review of the literature highlights the poor efficacy of deep brain simulation (DBS) [Shih and Tarsy, 2007]. Moreover, more than a quarter of patients died within 7 months of surgery. Owing to the limited number of reports, the poor outcome and the possibility of a harmful effect, DBS is currently not recommended in MSA [Wenning and Stefanova, 2009; Lambrecq et al. 2008; Shih and Tarsy, 2007; Santens et al. 2006; Talmant et al. 2006; Tarsy et al. 2003].

Future therapies

Neuroprotective strategies

Although the exact mechanisms of the neurodegenerative process in MSA remain unclear, the aggregation of alpha-synuclein in oligodendrocytes has been identified as a critical step in the pathogenesis [Jellinger and Lantos, 2010; Ubhi et al. 2010; Stefanova et al. 2009; Wenning et al. 2008]. Based on the key role of alpha-synuclein aggregation in MSA, transgenic animal models and genetic strategies have been developed. Transgenic animal models allow the expression of alpha-synuclein in oligodendrocytes under control of specific promoters [Shults et al. 2005; Yazawa et al. 2005; Kahle et al. 2002]. The growing number of MSA animal models [Fernagut et al. 2005; Stefanova et al. 2005] opens up the possibility to create a basis for drug screening in human trials. The efficacy of neuroprotective drugs is assessed in rodent models before translation to clinical trials. Furthermore, transgenic models may be used to understand the alpha-synuclein aggregation process and allow screening for candidate drugs before further assessment in clinical trials [Waxman and Giasson, 2010; Ono et al. 2007].

Lithium is a first-line treatment for bipolar mood disorders [Beaulieu and Caron, 2008]. The set of evidence has grown to suggest that lithium may also have also some neuroprotective properties [Ferrucci et al. 2010; Beaulieu and Caron, 2008; Feng et al. 2008; Fornai et al. 2008]. In human pathology, lithium has been tried in amyotrophic lateral sclerosis [Bedlack et al. 2008; Fornai et al. 2008; Vanacore and Galeotti, 2008]. The safety, tolerability and efficacy of lithium are being assessed in MSA in a prospective phase 2 trial (ClinicalTrials.gov Identifier: NCT00997672).

The results of a recent trial suggest that rasagiline may have ‘disease-modifying’ properties of still uncertain clinical significance in PD patients [Olanow et al. 2009b]. Positive effects of rasagiline on clinical motor scores and neuronal loss have also been suggested in a transgenic model of MSA [Stefanova et al. 2008]. Rasagiline is currently being assessed in MSA-P patients in a randomized, placebo-controlled, multicentre study (ClinicalTrials.gov Identifier: NCT00977665).

Nonsteroidal anti-inflammatory drugs (NSAIDs) reduce alpha-synuclein fibril formation in vitro [Hirohata et al. 2008] and may lower the risk of developing PD [Samii et al. 2009; Chen et al. 2005]. To date, the potential therapeutic benefit of NSAIDs has not been assessed in symptomatic alpha-synucleinopathies such as MSA [Hirsch and Hunot, 2009]. However, the harmful adverse effects of long-term use may limit their evaluation until the development of safer drugs. No clinical trials have yet been conducted or planned in MSA.

An open trial testing intravenous immunoglobulins for 6 months is currently in phase 2 (ClinicalTrials.gov Identifier: NCT00750867). However, a former clinical report does not support any therapeutic effect in MSA [Nanri et al. 2009].

Since rifampicin avoids alpha-synuclein fibril formation in vitro [Ono and Yamada, 2006] and reduces neuronal loss in a rodent model of MSA [Ubhi et al. 2008], this drug may be a candidate neuroprotective agent in MSA. However, to the best of the authors’ knowledge, no human trial has been yet planned.

Neurorestorative approach

Although experimental and open trials may support transplantation of foetal dopaminergic cells into the striatum in PD [Hauser et al. 1999; Kordower et al. 1998; Freeman et al. 1995; Widner et al. 1992; Lindvall et al. 1994, 1990], two double-blind, sham-controlled trials assessing transplantation of foetal nigral cells in PD failed to demonstrate any significant effect of the graft considering the primary outcome [Olanow et al. 2003; Freed et al. 2001]. Moreover, off-medication dyskinesia may occur in half of the transplanted patients [Olanow et al. 2009a], and the set of evidence has grown to suggest that Lewy body pathology may affect graft cells [Apostolidis et al. 2009; Kordower et al. 2008a, 2008b]. Despite these shortcomings, recent evidence in animal models of MSA-P may provide the rationale for human trials in MSA [Kollensperger et al. 2009; Puschban et al. 2005]. The use of autologous stem cells has recently been assessed in an open trial with MSA-C patients during 1 year [Lee et al. 2008]. Parenteral injection of bone marrow mesenchymal stem cells seem to slow UMSARS progression and increase cerebellar and frontal glucose metabolism on positron emission tomography (PET) scans compared with nontreated MSA patients. These preliminary results are still to be taken with caution. The underlying mechanisms of action, in particular how graft cells pass the blood–brain barrier from the arterial or venous peripheral circulation, remain unclear and need more preclinical explanation [Whone and Scolding, 2009; Quinn et al. 2008]. A double-blind, placebo-controlled, randomized clinical trial with initial arterial and subsequent venous administration of autologous mesenchymal stem cells has been launched in MSA patients in May 2009 (ClinicalTrials.gov Identifier: NCT00911365).

At the same time, a stem cell boosting strategy with injection of granulocyte colony stimulating factor (GCSF) has been launched in an open uncontrolled trial in four MSA patients [Pezzoli et al. 2010]. No major side effect was noted except for bone pain. However, given the small sample size, it is currently impossible to draw any conclusions about the efficacy of this approach.

Unexplained sudden death

For Schrag and collaborators, therapeutic management should concentrate on autonomic dysfunction, motor impairment and depression, because these features are associated with poor quality of life in MSA [Schrag et al. 2006]. In addition to symptomatic management and neuroprotective strategies, sudden death does not yet represent a major target for therapeutic intervention and there is no trial designed for such a strategy. Occurrence of sudden death is a common cause of mortality in MSA and may happen in the early stages while disability remains acceptable [Shimohata et al. 2008]. The origin of sudden death in MSA remains unknown, although clinical reports [Shimohata et al. 2007, 2008; Glass et al. 2006; Cormican et al. 2004], experimental [Tsuda et al. 2002] and neuropathological evidence [Tada et al. 2009; Benarroch, 2007] suggest that respiratory dysfunction may be the leading cause. Future preclinical studies should help to determine the underlying mechanisms before testing specific treatment strategies.

Considerations for future clinical trials

MSA is a rare neurodegenerative disease, with an estimated prevalence ranging from 1.9 [Tison et al. 2000] to 5 cases per 100,000 persons [Vanacore et al. 2001; Schrag et al. 1999]. The long-term follow up is limited by the rapid neurodegenerative process leading to reduced life expectancy. Rapid progression may also explain the high rate of drop out in MSA trials [Dodel et al. 2010]. Therefore, prospective trials with large sample sizes are challenging. The development of multicentric network trials in the framework of the European Multiple System Atrophy-Study Group (EMSA-SG) [Geser et al. 2005] or the North American Multiple System Atrophy Study Group (NAMSAS-SG) in the United States [Gilman et al. 2005] may help to overcome this problem. For a better comparison of outcome results, clinical trials should apply consensus criteria for the diagnosis of MSA [Gilman et al. 2008] and assess disease severity by the validated UMSARS [Wenning et al. 2004].

As in other neurodegenerative disorders, a putative neuroprotective drug should be given as early as possible. Therefore, sensitivity and specificity for the diagnosis of MSA have to be improved in early disease stages. In an ideal world, patients at risk of developing MSA should be identified before the onset of core features. However, even when using recent consensual criteria [Gilman et al. 2008] current diagnostic accuracy at the first neurological evaluation remains poor [Osaki et al. 2009]. Early autonomic symptoms or preceding RBD may be useful for early MSA detection [Gaig et al. 2008; Plazzi et al. 1997; Tison et al. 1995], but whether a patient with RBD will develop MSA, PD or no neurodegenerative disorder remains unpredictable [Schenck et al. 1996].

More sensitive progression biomarkers are also needed to assess the small effect of a putative neuroprotective drug in trials. To date, cerebrospinal fluid or blood biomarkers, even in combination, are still not specific enough or need further evidence-based studies before they can be employed in routine use for the clinical diagnosis of MSA [Constantinescu et al. 2010; Mollenhauer and Trenkwalder, 2009]. Promising biomarkers may consist of structural and functional neuroimaging abnormalities [Brooks and Seppi, 2009]. While routine MRI techniques are not sensitive enough to assess disease progression in clinical trials, the use of diffusion-weighted imaging, voxel-based morphometry or scales that rate the progression of brain atrophy may be better quantitative markers for the assessment of disease-modifying properties of a neuroprotective drug [Kollensperger and Wenning, 2009].

Conclusion

Current symptomatic management in MSA should target motor impairment, autonomic failure and depression, as these features are associated with a poor quality of life [Schrag et al. 2006]. Owing to the low number of randomized and controlled trials, practical management is currently based on empirical evidence. Levodopa remains the main treatment for MSA, despite its modest and nonsustained effect. Among the several treatments available for OH, only midodrine meets the criteria of evidence-based medicine [Wright et al. 1998; Low et al. 1997; Jankovic et al. 1993]. Strategies for urinary disorders are well standardized, while other symptoms such as breathing disorders, RBD, depression or dystonia remain out of consensus.

Although there have been major advances in our understanding of the cellular pathology and in performing prospective trials designed for putative neuroprotective drugs, no curative treatment is yet available. The results from clinical trials assessing the disease-modifying potential of rasagiline and lithium will soon be available. Recent advances in genetic models and neurotransplantation may stimulate further clinical trials in MSA. Finally, the growing number of multicentric networks in Europe and North America combining preclinical and clinical research strategies provide some hope for future advances.

Conflict of interest statement

The authors declare that there are no conflicts of interest.