Longitudinal flortaucipir, metabolism and volume differ between phonetic and prosodic speech apraxia

Abstract

Progressive apraxia of speech (PAOS) is a neurodegenerative motor-speech disorder that most commonly arises from a four-repeat tauopathy. Recent studies have established that progressive apraxia of speech is not a homogenous disease but rather there are distinct subtypes: the phonetic subtype is characterized by distorted sound substitutions, the prosodic subtype by slow and segmented speech and the mixed subtype by a combination of both but lack of predominance of either. There is some evidence that cross-sectional patterns of neurodegeneration differ across subtypes, although it is unknown whether longitudinal patterns of neurodegeneration differ. We examined longitudinal patterns of atrophy on MRI, hypometabolism on 18F-fluorodeoxyglucose-PET and tau uptake on flortaucipir-PET in a large cohort of subjects with PAOS that had been followed for many years. Ninety-one subjects with PAOS (51 phonetic, 40 prosodic) were recruited by the Neurodegenerative Research Group. Of these, 54 (27 phonetic, 27 prosodic) returned for annual follow-up, with up to seven longitudinal visits (total visits analysed = 217). Volumes, metabolism and flortaucipir uptake were measured for subcortical and cortical regions, for all scans. Bayesian hierarchical models were used to model longitudinal change across imaging modalities with PAOS subtypes being compared at baseline, 4 years from baseline, and in terms of rates of change. The phonetic group showed smaller volumes and worse metabolism in Broca's area and the striatum at baseline and after 4 years, and faster rates of change in these regions, compared with the prosodic group. There was also evidence of faster spread of hypometabolism and flortaucipir uptake into the temporal and parietal lobes in the phonetic group. In contrast, the prosodic group showed smaller cerebellar dentate, midbrain, substantia nigra and thalamus volumes at baseline and after 4 years, as well as faster rates of atrophy, than the phonetic group. Greater hypometabolism and flortaucipir uptake were also observed in the cerebellar dentate and substantia nigra in the prosodic group. Mixed findings were observed in the supplementary motor area and precentral cortex, with no clear differences observed across phonetic and prosodic groups. These findings support different patterns of disease spread in PAOS subtypes, with corticostriatal patterns in the phonetic subtype and brainstem and thalamic patterns in the prosodic subtype, providing insight into the pathophysiology and heterogeneity of PAOS.

Keywords: FDG-PET; MRI; apraxia of speech; flortaucipir; neurodegeneration.

Figure 1

MRI volume differences between phonetic and prosodic progressive apraxia of speech at baseline and after 4 years. Data are shown as phonetic group’s mean volume minus the prosodic group’s mean volume along the x-axis: if the thin-line is to the right of the dotted line, this shows strong evidence that the prosodic group had lower volume for that region (i.e. increased atrophy), and if the data are to the left of the dotted line, the phonetic group had lower volume for that region. Bilat. = bilateral; L = left; R = right; Vol = volume.

Figure 2

¹⁸F-fluorodeoxyglucose-PET differences in metabolism between phonetic and prosodic progressive apraxia of speech at baseline and after 4 years. Data are shown as phonetic group’s mean standardized uptake value ratio (SUVR) minus the prosodic group’s mean SUVR along the x-axis: if the thin-line is to the left of the dotted line, this indicates that the phonetic group had lower ¹⁸F-fluorodeoxyglucose (FDG) metabolism for that region (i.e. increased hypometabolism), and if the data are to the right of the dotted line, the prosodic group had lower FDG metabolism. L = left; R = right.

Figure 3

Flortaucipir-PET differences between phonetic and prosodic progressive apraxia of speech at baseline and after 4 years. Data are shown as phonetic group’s mean standardized uptake value ratio (SUVR) minus the prosodic group’s mean SUVR along the x-axis: if the thin line is to the left of the dotted line, this indicates that the prosodic group had more tau deposition (as measured by flortaucipir uptake) for that region, and if the data are to the right of the dotted line, the phonetic group had more tau deposition. Bilat. = bilateral; L = left; R = right.

Figure 4

Differences in annualized rate of volume loss and change in metabolism (¹⁸F-fluorodeoxyglucose) and flortaucipir (tau) uptake between phonetic and prosodic progressive apraxia of speech. Data are shown as phonetic group’s mean annualized rate of change minus the prosodic group’s mean annualized rate of change along the x-axis. If the thin line is left of the dotted line, the prosodic group had a faster rate of decline, and if data are right of the dotted line, the phonetic group had a faster rate of decline. Bilat. = bilateral; FDG = ¹⁸F-fluorodeoxyglucose; L = left; R = right; Vol = volume.