CHARACTERIZATION OF THE PRIMARY PROGRESSIVE APHASIAS SUBTYPES AFTER THE 2011 RECOMMENDATION: A PRELIMINARY META-ANALYSIS

Abstract

To improve the etiological diagnosis of PPA, the 2011 classification system was set to distinguish three subtypes: semantic variant (svPPA), nonfluent/agrammatic variant (nfvPPA), and the previously under recognized logopenic variant (lvPPA). Our aim is to describe the frequency of primary progressive aphasias (PPA) subtypes using the 2011 Gorno-Tempini classification system. Search of MEDLINE was undertaken in June 2016 for observational cohort studies reporting PPA subtypes that used the novel criteria. We excluded studies reporting A) case control, case report/series, review/meta-analysis, or letter/editorial, B) on only 1 or 2 specifically selected PPA subtypes, single family mutation, C) used the previous FTD diagnostic criteria (Neary 1998), D) combining subtypes (e.g., IvPPA and svPPA vs. nfvPPA). For each study, the number of cases of PPA subtypes and sex ratio was recorded and subsequently analyzed for frequency (event rate). Duplicate data published by same groups were controlled at the level of analysis. The search yielded 509 studies. After examining for selection criteria, the aggregate event rates were: nfv 32 % (95%CI: 28–36%; 49% male; Studies=40; n=644/2087), sv 25 % (95%CI: 21–29%; 53% male; Studies=40; n=548/2087), Iv 34% (95%CI: 29–39%; 48% male; Studies=39; n=695/2061), and unclassified 16% (95%CI: 12–21%; 45% male; Studies=21; n=200/1126). After controlling for duplicated data in multiple papers from same groups, the event rates were [nfvPPA: 33% (27–39%; Studies=22; n=424/1326); lvPPA: 31% (25–37%; Studies=22; n=399/1326); svPPA: 27% (22–32%; Studies=22; n=389/1326); and unclassified: 14% (8–22%; Studies=11; n=114/655), and no sex differences]. The most frequently reported PPA subtype was nfvPPA, followed by lvPPA, and by svPPA. No significant gender differences were found. Remarkably, nearly 1/8 of the PPA cases were unclassifiable. Thus recognizing the three subtypes does not enclose the full spectrum of PPA phenotypes, which warrants enhancement of the recommendations. Future studies should better characterize the unclassifiable group, and determine if adding another subtype is necessary.