Autonomic failure and reduced survival in multiple system atrophy

Multiple system atrophy (MSA) is a progressive neurodegenerative disease that affects 3–5 people out of every 100 000 in the general population (Gilman et al., 2008). It is characterized by a varying combination of parkinsonian, cerebellar, pyramidal and autonomic symptoms. The latter include orthostatic hypotension and bladder, bowel, sexual and breathing disorders. The clinical abnormalities are due to a complex pathophysiology that reflects neurodegeneration of cortical areas, the basal ganglia, brainstem and cerebellum (Trojanowski and Revesz, 2007; Gilman et al., 2008; Bologna et al., 2014; Suppa et al., 2014). Widespread neuronal degeneration results in a short median survival time, ranging across studies from 6 to 9 years (Wenning et al., 2004). With no known biomarkers for disease diagnosis or progression, clinicians and scientists have long recognized the importance of classifying MSA into distinct subtypes to be able to identify features that influence the rate of progression and response to interventions. Two main disease subtypes are recognized: (i) patients with predominantly parkinsonian features are said to have MSAParkinsonism (MSA-P); and (ii) patients with predominantly cerebellar ataxia are said to have MSACerebellar (MSA-C). Clinical features of autonomic involvement are present in both cases, although their severity varies. Autonomic failure in MSA is caused by pathology of the medullar and spinal autonomic nuclei, which invariably display neuronal loss and the presence of -synuclein immunoreactive glial and neuronal cytoplasmic inclusions. With regard to the natural history of the disease, whether there are any differences between the MSA-P and MSA-C subtypes remains a matter of debate (Wenning et al., 2004). There is also controversy over the role of autonomic disorders as predictors of a more aggressive form of MSA, with some studies showing that early or initial autonomic symptoms are associated with worse survival (Wenning et al., 2013; Low et al., 2015), while others have been unable to replicate this association (Roncevic et al., 2014). In this issue of Brain, Coon and co-workers investigate the role of autonomic testing as a means of evaluating survival in patients with MSA and conclude that the results of such tests can indeed serve as a prognostic marker (Coon et al., 2015). Between January 1998 and December 2012, Coon and colleagues evaluated a large number of patients (n = 685) and then retrospectively reviewed them for the current study; 594 met the consensus criteria for probable MSA, and the remaining 91 for possible MSA (Gilman et al., 2008). MSA-P was the predominant subtype, found in 430 patients (63%). Median disease duration from symptom onset to death was 7.5 years. Although the retrospective nature of this study to some extent limits interpretation of the results, the study does comprise the largest published cohort of patients examined and diagnosed at a single referral centre. In addition, the validity of the study was further strengthened by the fact that the diagnosis of MSA was neuropathologically confirmed in all 36 patients who were examined post-mortem. Lastly, all patients underwent comprehensive and standardized autonomic testing, including the sudomotor axon reflex, heart rate response to deep breathing, blood pressure responses to the Valsalva manoeuvre and head-up tilting. Coon and co-workers used multivariate analysis to identify clinical predictors of shortened survival; six variables were included in the final model as independent predictors of shortened survival, the strongest being falls occurring within 3 years of symptom onset. However, four of the other five independent predictors were related to autonomic symptoms (orthostatic intolerance, bladder symptoms and early catheterization) or the general degree of autonomic impairment evaluated by means of a specific and validated scale, the Composite Autonomic Severity Score (Suarez et al., 1999). Later age of onset was the remaining negative BRAIN 2015: 138; 3466–3475 | 3466