Prevalence of amyloid-β pathology in distinct variants of primary progressive aphasia

Abstract

Objective: To estimate the prevalence of amyloid positivity, defined by positron emission tomography (PET)/cerebrospinal fluid (CSF) biomarkers and/or neuropathological examination, in primary progressive aphasia (PPA) variants.

Methods: We conducted a meta-analysis with individual participant data from 1,251 patients diagnosed with PPA (including logopenic [lvPPA, n = 443], nonfluent [nfvPPA, n = 333], semantic [svPPA, n = 401], and mixed/unclassifiable [n = 74] variants of PPA) from 36 centers, with a measure of amyloid-β pathology (CSF [n = 600], PET [n = 366], and/or autopsy [n = 378]) available. The estimated prevalence of amyloid positivity according to PPA variant, age, and apolipoprotein E (ApoE) ε4 status was determined using generalized estimating equation models.

Results: Amyloid-β positivity was more prevalent in lvPPA (86%) than in nfvPPA (20%) or svPPA (16%; p < 0.001). Prevalence of amyloid-β positivity increased with age in nfvPPA (from 10% at age 50 years to 27% at age 80 years, p < 0.01) and svPPA (from 6% at age 50 years to 32% at age 80 years, p < 0.001), but not in lvPPA (p = 0.94). Across PPA variants, ApoE ε4 carriers were more often amyloid-β positive (58.0%) than noncarriers (35.0%, p < 0.001). Autopsy data revealed Alzheimer disease pathology as the most common pathologic diagnosis in lvPPA (76%), frontotemporal lobar degeneration-TDP-43 in svPPA (80%), and frontotemporal lobar degeneration-TDP-43/tau in nfvPPA (64%).

Interpretation: This study shows that the current PPA classification system helps to predict underlying pathology across different cohorts and clinical settings, and suggests that age and ApoE genotype should be considered when interpreting amyloid-β biomarkers in PPA patients. Ann Neurol 2018;84:737-748.

FIGURE 1:

Prevalence of amyloid-β pathology in primary progressive aphasia (PPA) variants by modality. Ninety-three patients had multiple measures of Aβ pathology available (62 positron emission tomography [PET] + cerebrospinal fluid [CSF], 19 PET + autopsy, 12 CSF + autopsy), yielding 92% concordance between modalities. CSF = cerebrospinal fluid; lvPPA = logopenic variant of PPA; nfvPPA = nonfluent variant of PPA; PET = positron emission tomography; PPA-M/U = mixed/unclassified PPA; svPPA = semantic variant of PPA.

FIGURE 2:

Prevalence of amyloid-β positivity in primary progressive aphasia (PPA) variants. Prevalence estimate of amyloid-β positivity is based on generalized estimating equations analyses. Data for normal controls and typical Alzheimer disease (AD) dementia come from the Amyloid PET Study Group.^,

FIGURE 3:

Autopsy results. (A) Pie charts showing the respective prevalence of amyloid, tau, TAR DNA-binding protein 43 (TDP), and other pathologies in primary progressive aphasia (PPA). (B) Breakdown of the different pathologies for each PPA variant. Aβ = amyloid-β; AGD = argyrophilic grain disease; CBD = corticobasal degeneration; CJD = Creutzfeldt–Jakob disease; DLB = dementia with Lewy bodies; FTLD = frontotemporolobar degeneration; FUS = fused in sarcoma; GGT = globular glial tauopathy; lvPPA = logopenic variant of PPA; nfvPPA = nonfluent variant of PPA; PiD = Pick disease; PPA-M/U = mixed/unclassified PPA; PSP = progressive supranuclear palsy; svPPA = semantic variant of PPA; VaD = vascular dementia.