Primary progressive aphasia: New insights paving the way toward clinical research tools.

Neurology 2010;75:582–583 When a patient presents with acute aphasia, today’s neurologist takes swift action to characterize the clinical syndrome and use neuroimaging and other tools to identify its pathophysiologic basis. One goal of these urgent efforts is to determine whether the patient has an ischemic process that merits thrombolytic intervention. We can only hope that tomorrow’s neurologist is as well-equipped with clinical and pathobiologic assessment tools when faced with a patient with progressive aphasia. Three articles in the current issue move us in this direction. Patients with primary progressive aphasia (PPA) may exhibit impairments spanning the spectrum of speech and language function,1 with 3 major subtypes currently recognized and ongoing efforts to reach consensus regarding diagnostic criteria.2 Although the classification of patients with aphasia as fluent or nonfluent may be difficult because of the multifaceted nature of fluent speech,3 it has been obvious for many years that some patients with PPA mostly exhibit difficulties understanding the meaning of words but are generally able to speak fluently— this is semantic dementia (SD),4 or the PPA-semantic variant. Other patients may be able to understand word meaning well but have difficulty with the rate, rhythm, grammatic structure, or articulation of speech. Originally termed nonfluent or progressive nonfluent aphasia (PNFA), many clinicians are now classifying these patients as agrammatic or logopenic5 based on their language characteristics. Patients with logopenic PPA may be variably fluent depending on the type of conversational exchange or test being undertaken.1 Partly because of this variability, research groups have differed in their approach to classifying this type of PPA. Two of the present studies employ quantitative psycholinguistic analysis of speech samples to investigate further the underlying mechanisms of impaired speech fluency in patients with progressive aphasia. In their 16 patients with PNFA, Gunawardena et al.6 found that the nearly 70% reduction in words per minute compared to controls was related to the level of grammatic impairment and not to the level of speech-sound errors or executive dysfunction. This finding supports the concept that, at least in this patient sample using these measures, nonfluent speech is related to agrammatism. It was somewhat surprising that speech-sound errors did not make some contribution given that 9 of the patients had a motor speech disorder. The study by Rohrer et al.7 provides complementary data by including all 3 major PPA subtypes as well as a larger variety of psycholinguistic measures. Their patients, whose MRI data were recently published elsewhere, were classified into 1 of 4 groups based on whether or not there was agrammatism or apraxia of speech (AOS) in speech samples (semantic patients were classified separately). Using these categories, they then examined psycholinguistic features, which generally fit well with the characteristics proposed as being central to the diagnostic subtypes. Another group of 3 patients—all with mutations in the progranulin gene—had different characteristics that will be important to investigate in other patient cohorts. As for the quantitative speech analysis, rate of speech in the 2 groups that fit with PNFA was reduced to a similar level as that in the study by Gunawardena et al. (!70% lower than controls), while that in the logopenic group was reduced by 50%. The quantitative analysis of speech samples is a cumbersome activity that requires significant skill, and it will be of great interest to see these datasets further explored for additional measures, such as mean and variance in length of utterance, the latter potentially capturing a core feature of logopenic PPA (LPA). These data are convergent with other recent studies aiming to quantify psycholinguistic deficits in PPA subtypes, and highlight the importance of assessments of grammar,8 motor speech,9 and other language features in differentiating forms of PPA. The initial success of these studies provides further support for the efforts of the working group to refine the clinical diagnostic crite-