Primary progressive aphasia: six questions in search of an answer

Abstract

Here, we review recent progress in the diagnosis and management of primary progressive aphasia-the language-led dementias. We pose six key unanswered questions that challenge current assumptions and highlight the unresolved difficulties that surround these diseases. How many syndromes of primary progressive aphasia are there-and is syndromic diagnosis even useful? Are these truly 'language-led' dementias? How can we diagnose (and track) primary progressive aphasia better? Can brain pathology be predicted in these diseases? What is their core pathophysiology? In addition, how can primary progressive aphasia best be treated? We propose that pathophysiological mechanisms linking proteinopathies to phenotypes may help resolve the clinical complexity of primary progressive aphasia, and may suggest novel diagnostic tools and markers and guide the deployment of effective therapies.

Keywords: Alzheimer’s disease; Frontotemporal dementia; Logopenic aphasia; Primary progressive aphasia; Semantic dementia.

Fig. 1

A ‘roadmap’ for making a syndromic diagnosis of primary progressive aphasia (PPA) at the bedside, in patients presenting with progressive speech and/or language impairment as leading and dominant symptoms. On the left of the Figure, we list key clinical features that are most discriminating for major variant syndromes of primary progressive aphasia (see also Table 1), according to current consensus diagnostic criteria [7] (see Table S1). Speech production impairment (apraxia and/or grammatical errors in speaking or writing) points to the nonfluent/agrammatic variant (nfvPPA), impaired single-word comprehension to the semantic variant (svPPA), and impaired repetition of phrases (disproportionate to single words) to the logopenic variant (lvPPA). Speech apraxia, agrammatism and word comprehension impairment are elicited on history and examination; impaired phrasal repetition must be confirmed on examination. Note that clinical features often seen in PPA but less useful in differentiating syndromes have not been included here (e.g. anomia is prominent in both svPPA and lvPPA). Cases may not conform to a single canonical syndrome (Atypical PPA); this may be due to relatively circumscribed language impairments that lack additional features (e.g. dynamic aphasia and progressive pure anomia), more complex mixed language phenotypes, or the presence of prominent non-language features (see text). Further investigations are indicated following the clinical syndromic diagnosis (see Fig. 2), to substantiate the bedside impression and fully characterise the syndrome (neuropsychometry where available; brain MRI in all cases), and to identify the underlying proteinopathy with a view to symptomatic treatment (Alzheimer’s disease biomarkers) or genetic counselling (where clinically appropriate). Adapted under a CC-BY 4.0 license from: Marshall et al., J Neurol 2018; 265: 1474–1490

Fig. 2

Next steps after bedside diagnosis in syndromes of primary progressive aphasia: semantic variant primary progressive aphasia (svPPA), nonfluent/agrammatic variant (nfv)PPA, logopenic variant (lv)PPA and atypical or ‘mixed’ PPA. Where available, assessment by a neuropsychologist is very valuable in fully defining and quantifying the cognitive phenotype, over linguistic as well as non-linguistic domains. The ‘target diagrams’ (top panels) show typical profiles of neuropsychological test performance for each syndrome; concentric circles indicate percentile scores relative to a healthy age-matched population and distance along the radial dimension represents level of functioning in the cognitive domains assessed (exec, executive skills; lit, literacy skills (spelling, arithmetic); name, naming; nv mem, nonverbal memory; rep ph, repetition of phrases; rep w, repetition of single words; sent, sentence processing (construction and comprehension); vis, visuo-spatial; v mem, verbal memory; vocab, vocabulary (single-word comprehension)). Brain imaging (wherever possible, MRI) is an essential part of the diagnostic workup of any patient with suspected PPA; coronal T1-weighted brain MRI sections representing characteristic atrophy profiles in each syndrome are shown (middle panels; left hemisphere presented on the right). In svPPA, the profile of asymmetric (predominantly left-sided) anterior, mesial and inferior temporal lobe atrophy is highly consistent, whereas atrophy profiles in nfvPPA (predominantly left inferior frontal, insular and anterior superior temporal gyrus atrophy) and lvPPA (predominantly involving left posterior superior temporal and inferior parietal cortices) are much more variable between individual patients. In patients with lvPPA, mixed PPA and nfvPPA, we have a low threshold for trialling a symptomatic therapy for Alzheimer’s disease (AD), such as donepezil. In younger patients, assessing AD biomarkers in CSF or brain amyloid PET is likely to provide diagnostically relevant information, and genetic testing for a mutation in the frontotemporal dementia spectrum is also a consideration (see text), particularly where there is a suggestive family history, atypical clinical features or a strikingly asymmetric atrophy profile (as in the patient with mixed PPA here, who had a pathogenic progranulin gene mutation). The possibility of conjoint pathologies should be kept in mind [30]. Adapted under a CC-BY 4.0 license from: Marshall et al., J Neurol 2018; 265: 1474–1490

Fig. 3

Proposed pathophysiology of primary progressive aphasia. The figure diagrams a lateral view of the left cerebral hemisphere, overlaid with core neural mechanisms that we propose underpin each of the major variant syndromes of primary progressive aphasia (PPA): nonfluent/agrammatic (nfvPPA), semantic (svPPA) and logopenic (lvPPA). Oblongs signify major processing ‘hubs’ within the language network: each instantiates a key neural ‘template-matching’ operation in which incoming data (represented by black hatching) is iteratively reconciled with stored predictions and transformed into an output (‘predictive coding’; see text). The bidirectional arrows represent the reciprocal exchange of data and predictions between core processing modules. Processing modules are organised hierarchically, in that representations of incoming sensory data in temporo-parietal junctional cortex (blue) are transformed into increasingly abstract conceptual representations in anterior temporal cortex (green) and may ultimately be used in generating a motor output via anterior peri-Sylvian mechanisms (red). The putative core physiological mechanism targeted in each syndrome gives rise to the essential features of the syndrome. In lvPPA, this mechanism is proposed to be the transformation of sensory data into phonological codes; if these codes are defective, they do not enter working memory normally and cannot be used to access other components of the language network (e.g. during naming). The semantic appraisal network, targeted in svPPA, stores multimodal representations corresponding to words, objects and concepts, based on the computation of higher-order regularities in sensory data: erosion of these semantic representations leads initially to loss of vocabulary and ultimately, a pan-modal impairment of semantic memory. In nfvPPA, the core pathophysiological mechanism may be impaired transformation of rule-based command sequences (such as those governing grammar and articulation) into behavioural routines (for example, because excessively rigid predictions lead to delayed error resolution [169]); Reproduced under a CC-BY 4.0 license from: Ruksenaite et al., Curr Neurol Neurosci Rep 2021; 21: 7