Clinical diagnostic criteria for MSA were first proposed by Quinn in 1989 (11) and later slightly modified in 1994 (12). According to this schema, patients are classified as either striatonigral degeneration (SND) or olivopontocerebellar atrophy (OPCA) type MSA depending on the predominance of parkinsonism or cerebellar ataxia. There are three levels of diagnostic probability: possible, probable, and definite. Patients with sporadic adult-onset poorly levodopa-responsive parkinsonism fulfill criteria for possible SND. The presence of other atypical features such as severe autonomic failure, cerebellar or pyramidal signs, or a pathological sphincter electromyogram (EMG) is required for a diagnosis of probable SND. Patients with sporadic late-onset predominant cerebellar ataxia with additional mild parkinsonism or pyramidal signs are considered possible OPCA-type MSA. This may result in confusion since some patients with possible OPCA may also qualify for probable SND provided predominant cerebellar ataxia is accompanied by parkinsonian features. A diagnosis of probable OPCA-type MSA requires the additional presence of severe autonomic failure or a pathological sphincter EMG. A definite diagnosis rests on neuropathological confirmation. Predominant SND- or OPCA-type presentations may be distinguished from pure types on the basis of associated cerebellar (predominant SND) or parkinsonian features (predominant OPCA). Since some degree of autonomic failure is present in almost all SND- and OPCA-type MSA patients (13,14) a further "autonomic" subtype (Shy-Drager syndrome) was not considered useful (15). A number of exclusion criteria were also proposed: onset should be age 30 yr or more, and in order to exclude inherited adult-onset ataxias there should be no family history of MSA.
The validity of Quinn's criteria was evaluated in a clinicopathologic study by Litvan and coworkers. This study revealed the criteria for the diagnosis of MSA proposed by Quinn present a suboptimal specificity (79% for possible MSA and 97% for probable MSA, at the first visit), a low sensitivity (53% for possible MSA and 44% for probable MSA, at the first visit), and a predictive value of 30% for possible MSA and 68% for probable MSA (16). Because of the suboptimal diagnostic accuracy of Quinn's criteria, in 1998 an International Consensus Conference promoted by the American Academy of Neurology was convened to develop new and optimized criteria for a clinical diagnosis of MSA (Table 1) (8). The Consensus criteria are now widely used for a clinical diagnosis of MSA. These criteria specify three diagnostic categories of increasing certainty: possible, probable, and definite. The diagnosis of possible and probable MSA are based on the presence of clinical features listed in Table 1. In addition, exclusion criteria have to be considered. A definite diagnosis requires a typical neuropathological lesion pattern as well as deposition of a-synuclein-positive glial cytoplas-mic inclusions.
A subsequent study analyzed the agreement between Quinn's criteria and Consensus criteria in a clinical series of 45 MSA patients. Concordance was moderate for possible MSA and substantial for probable MSA (17). Moreover, four cases with probable (n = 2) or possible (n = 2) MSA according to Quinn's criteria were unclassifiable according to the Consensus criteria.
A recent retrospective evaluation of the Consensus criteria on pathologically proven cases showed excellent positive predictive values (PPVs) for both possible (93%) and probable MSA (100%) at the first clinic visit; however, sensitivity for probable MSA was poor especially in early stages of the disease (16% at the first clinic visit) (18). Interestingly, the Consensus criteria and Quinn's criteria had similar PPVs. Whether the Consensus criteria will improve recognition of MSA patients especially in early disease stages needs to be investigated by prospective surveys with neuropathological confirmation in as many cases as possible.
Was this article helpful?