Primary Progressive Aphasia
Primary progressive aphasia (PPA) is a clinical syndrome of progressive language decline caused by neurodegeneration of language-dominant brain regions, most commonly within the frontotemporal lobar degeneration (FTLD) spectrum. Gorno-Tempini and colleagues published consensus classification criteria in 2011 that define three canonical variants — semantic variant PPA (svPPA), nonfluent/agrammatic variant PPA (nfvPPA), and logopenic variant PPA (lvPPA) — each with distinct linguistic profiles, neuroimaging signatures, and underlying pathologic substrates. The mapping from clinical presentation to histopathology is imperfect but clinically consequential: svPPA is strongly associated with FTLD-TDP type C, nfvPPA with FTLD-tau (including progressive supranuclear palsy and corticobasal degeneration pathology), and lvPPA is most commonly caused by Alzheimer disease (AD). Correct classification guides prognostication, genetic counseling, and emerging therapeutic trial eligibility.
Bottom Line
- Classification: The 2011 Gorno-Tempini criteria define three PPA variants based on language profile — semantic (svPPA), nonfluent/agrammatic (nfvPPA), and logopenic (lvPPA); each requires progressive, isolated language decline as the principal cause of functional impairment for at least 2 years
- svPPA: Loss of word meaning (single-word comprehension deficit), impaired confrontation naming, surface dyslexia, and eventual prosopagnosia; associated with anterior temporal lobe atrophy (left-predominant) and FTLD-TDP type C pathology in >80% of cases
- nfvPPA: Effortful, halting speech with agrammatism and/or apraxia of speech; spared single-word comprehension; associated with left posterior fronto-insular atrophy and FTLD-tau pathology (PSP, CBD, Pick disease)
- lvPPA: Word-finding pauses in spontaneous speech, impaired sentence repetition (length-dependent), relatively spared single-word comprehension and motor speech; usually caused by AD pathology — making it the AD-associated PPA variant
- Bedside classification: Assess word comprehension and grammar first (per Mesulam); three brief tests — word-picture matching, syntactic knowledge (Northwestern Anagram Test), and sentence repetition — can classify most PPA patients; validated batteries include the Sydney Language Battery (80% accuracy) and Mini Linguistic State Examination (91% accuracy)
- Progression: Each variant evolves to overlap with other neurodegenerative syndromes — svPPA with behavioral variant FTD, nfvPPA with corticobasal syndrome or PSP, and lvPPA with typical AD; FTD is more heritable than AD, with pathogenic variants in C9orf72, GRN, or MAPT identified in >10% of FTD cases
PPA Classification Framework
All three PPA variants share a common overarching criterion: progressive language decline must be the most prominent clinical feature at symptom onset and during the initial phases of the disease, and language impairment must be the principal cause of limitations in daily living activities. PPA is distinguished from acute aphasias (stroke, trauma) by its insidious onset and gradual progression over years. The differential diagnosis includes recurrent strokes mimicking progressive disease, acquired epileptic aphasia (in children), viral or autoimmune encephalitis (more rapid course, not restricted to language), and rarely progressive leukoencephalopathy.
Mesulam's Diagnostic Approach to PPA
- Start with word comprehension and grammar rather than fluency, as fluency varies significantly with emotional state and conversation topic
- Impaired word comprehension + intact grammar → semantic variant PPA (lower right quadrant)
- Intact word comprehension + impaired grammar → nonfluent/agrammatic variant PPA (upper left quadrant)
- Both impaired → "mixed" PPA
- Neither impaired → likely logopenic variant PPA (word-finding pauses and sentence repetition deficit)
- Three bedside tests can classify most patients: word-picture matching, a syntactic construction test (e.g., Northwestern Anagram Test), and word/phrase/sentence repetition with progressively lengthier items
Semantic Variant PPA (svPPA)
Semantic variant PPA is characterized by a progressive loss of word meaning, such that the sound pattern of a word is correctly perceived but is no longer associated with its meaning. Patients may bluntly ask about common words (e.g., "What is a hamburger?"). Speech remains fluent — sometimes so fluent that no language disorder is initially apparent — and patients conceal word-finding deficits through circumlocution, producing lengthy but comprehensible utterances. Both core features (impaired confrontation naming and impaired single-word comprehension) must be present, along with at least three supportive features.
Key Clinical Features of svPPA
- Impaired confrontation naming: Patients give no response or say "I don't know" when asked to name objects, especially low-frequency items
- Impaired single-word comprehension: Unable to point to objects in response to their names; may ask, "What do you mean by 'whistle'?"
- Surface dyslexia/dysgraphia: Difficulty reading words with irregular spellings (e.g., yacht, island, colonel); in Japanese, difficulty reading kanji with preserved kana reading
- Prosopagnosia: Loss of face recognition — reflects spread of semantic deficits beyond language into nonverbal domains
- Spared repetition: Patients readily repeat words they cannot comprehend
- Spared speech production: Fluent, lengthy replies free of grammatical or motor speech errors
Semantic Dementia vs Semantic Variant PPA
- When semantic deficits extend to nonverbal semantics (object knowledge, face recognition, emotional recognition), the broader term semantic dementia is used
- Right temporal predominance produces prosopagnosia, loss of empathy, impaired facial affect recognition, and behavioral changes — termed semantic behavioral variant FTD
- Left temporal predominance produces the classic verbal semantic deficit (svPPA)
- A specific deficit in knowledge about animals has been described in right-predominant cases
- Many patients with mixed semantic-behavioral features eventually meet criteria for behavioral variant FTD
Nonfluent/Agrammatic Variant PPA (nfvPPA)
Nonfluent/agrammatic variant PPA is defined by effortful, halting speech and/or agrammatism. At least one core feature must be present. Agrammatism refers to the omission of grammatical affixes or function words (e.g., "She go mall"), affecting both language output and comprehension of syntactically complex sentences (e.g., passive voice, object-relative clauses). Apraxia of speech — an impaired ability to program articulators for correct syllable sequences — contributes to the nonfluency. Writing may be relatively preserved, especially when apraxia of speech is the dominant contributor to impaired fluency. Some patients present with progressive apraxia of speech alone, without true aphasia, representing an overlap between primary progressive apraxia of speech and nfvPPA.
Key Clinical Features of nfvPPA
- Agrammatism: Omission of suffixes and function words; "telegraphic" speech; grammatical simplification in writing
- Apraxia of speech: Inconsistent speech sound errors and distortions; inability to repeat polysyllabic words (e.g., "catastrophe") consistently
- Impaired complex syntax comprehension: Fails to assign thematic roles correctly in sentences with object-relative clauses
- Spared single-word comprehension: Performs perfectly on spoken word-picture matching
- Spared object knowledge: Correctly answers questions about common objects and famous people
Logopenic Variant PPA (lvPPA)
Logopenic variant PPA is the AD-associated PPA syndrome, characterized by prominent word-finding pauses in spontaneous conversation and deficits of auditory verbal working memory that result in impaired sentence repetition in a length-dependent manner. Unlike nfvPPA, motor speech is normal and grammar is intact. Unlike svPPA, single-word comprehension is preserved. The core deficit is at the level of phonologic retrieval — the patient knows the word's meaning but cannot efficiently access its sound form. Most patients with lvPPA are found to have AD pathology at autopsy, and AD biomarkers (amyloid PET, CSF amyloid/tau) are typically positive.
Comparison of PPA Variants
| Feature | Semantic (svPPA) | Nonfluent/Agrammatic (nfvPPA) | Logopenic (lvPPA) |
|---|---|---|---|
| Speech fluency | Fluent (sometimes excessively so) | Effortful, halting, reduced phrase length | Slow with word-finding pauses; otherwise fluent |
| Core deficit | Loss of word meaning (semantic knowledge) | Agrammatism ± apraxia of speech | Impaired phonologic retrieval and verbal working memory |
| Confrontation naming | Severely impaired | Variably impaired | Impaired (anomia with phonemic paraphasias) |
| Single-word comprehension | Impaired (defining feature) | Spared | Spared |
| Sentence comprehension | Spared (for grammar) | Impaired for complex syntax | Impaired for long sentences |
| Repetition | Spared | Variably impaired | Impaired (length-dependent; sentences > words) |
| Motor speech | Normal | Apraxia of speech common | Normal |
| Reading | Surface dyslexia (irregular words) | Impaired reading aloud (articulatory) | Relatively preserved |
| Object knowledge | Impaired (especially low-frequency items) | Spared | Spared |
Neuroimaging Patterns
The atrophy patterns in PPA variants recapitulate distinct functional brain networks, likely because pathologic changes propagate among neurons that interact with one another through transneuronal degeneration. Neuroimaging is required for imaging-supported diagnosis in each variant.
| PPA Variant | Atrophy Pattern (MRI) | Metabolic Pattern (FDG-PET/SPECT) | Network Affected |
|---|---|---|---|
| svPPA | Anterior temporal lobes, markedly asymmetric (left > right); expansion of temporal horn | Anterior temporal hypometabolism/hypoperfusion | Semantic memory network (amodal hubs in perirhinal cortex) |
| nfvPPA | Left posterior fronto-insular region; left pars opercularis; anterior insula | Left posterior fronto-insular hypometabolism/hypoperfusion | Left perisylvian speech-motor network |
| lvPPA | Left posterior temporal and inferior parietal regions (temporoparietal junction) | Left temporoparietal hypometabolism | Phonologic loop / auditory-verbal working memory network |
Imaging Pearls for PPA Diagnosis
- svPPA: "Knife-edge" anterior temporal atrophy is characteristic; right-predominant atrophy suggests semantic behavioral variant FTD with prosopagnosia and behavioral changes rather than classic verbal semantic loss
- nfvPPA: Agrammatism and apraxia of speech map to adjacent but distinct regions in the left inferior frontal lobe; expressive vs receptive agrammatism localizes to left frontal vs left temporal lobes, respectively
- lvPPA: The temporoparietal junction atrophy pattern overlaps with the typical AD distribution, consistent with the underlying AD pathology in most cases
- FDG-PET differentiates FTD from AD with high accuracy and is commonly employed in ambiguous cases; patterns of hypometabolism differ starkly between the two conditions
- Amyloid PET: Positive in lvPPA (AD pathology) but negative in svPPA and nfvPPA (FTLD pathology), providing a useful discriminator
Neuropathology
Nearly all cases of the three core FTD syndromes can be classified into one of three neuropathologic categories based on the primary molecular defect: TDP-43 (~50% of all FTLD cases), MAPT (tau) (~45%), and FUS (~5%). The PPA variants have different pathologic associations, and clinical features can provide clues to the underlying molecular defect.
| PPA Variant | Most Common Pathology | Other Pathologies | Pathologic Clues |
|---|---|---|---|
| svPPA | FTLD-TDP type C (>80%): long dystrophic neurites in superficial cortical layers; neuronal cytoplasmic inclusions in hippocampal dentate layer | Pick disease (3R tau); argyrophilic grain disease | Highly predictable pathology; semantic variant is the most pathologically homogeneous PPA |
| nfvPPA | FTLD-tau (PSP, CBD, Pick disease): both tau and TDP-43 common | FTLD-TDP (types A, B); rarely AD | Apraxia of speech → more likely tau; anomia/agrammatism without apraxia → more likely TDP-43 |
| lvPPA | Alzheimer disease pathology (amyloid plaques and neurofibrillary tangles) | Occasionally FTLD-TDP or FTLD-tau | Positive AD biomarkers (amyloid PET, CSF Aβ42/tau ratio) strongly support AD substrate |
Clinicopathologic Pitfalls
- AD can mimic FTD: AD infrequently presents with a behavioral variant FTD or nfvPPA syndrome that cannot be distinguished from FTLD on clinical grounds alone; AD biomarkers are essential for differentiation
- GRN mutations may produce atypical PPA: Patients with GRN (progranulin) variants can present with logopenic-type aphasia or mixed semantic-grammatical features that defy canonical classification
- FUS pathology is NOT associated with PPA: FUS is exclusively linked to behavioral variant FTD, often with very young onset (reported as young as 22 years)
- Tau PET limitations: The only currently approved tau tracer (flortaucipir) exhibits off-target binding in patients with suspected TDP-43 pathology; PET tracers specific for FTLD tau isoforms and TDP-43 are still needed
- Cholinesterase inhibitors are NOT recommended in FTD/PPA with FTLD pathology and may worsen cognition and behavior; however, they may be appropriate if lvPPA with underlying AD is confirmed
Genetics of PPA
FTD is more heritable than AD, with up to 43% of patients having a family history of dementia, ALS, or Parkinson disease. Approximately 10%–20% carry a known pathogenic variant. The three major genes — C9orf72, GRN, and MAPT — account for the majority of genetic FTD cases and have distinct associations with PPA subtypes.
| Gene | Protein / Mechanism | Mean Onset Age | PPA Association | Pathology | Imaging Pattern |
|---|---|---|---|---|---|
| C9orf72 | GGGGCC hexanucleotide repeat expansion (noncoding) | 61.3 years | Primarily bvFTD and ALS; PPA uncommon but can occur with nfvPPA or svPPA features | TDP-43 type A or B | Relatively symmetric; frontal > temporal atrophy |
| GRN | Progranulin haploinsufficiency | 58.2 years | bvFTD most common; nfvPPA second; may produce logopenic or mixed PPA features | TDP-43 type A | Markedly asymmetric; frontal, temporal, and parietal; side varies within families |
| MAPT | Tau protein missense/splice-site variants | 49.5 years | Semantic impairment common; anomia detectable in presymptomatic carriers; more disinhibition than apathy | Tau (filamentous inclusions; may resemble Pick bodies) | Relatively symmetric; temporal > frontal atrophy |
When to Pursue Genetic Testing in PPA
- Family history of dementia, ALS, or parkinsonism in first-degree relatives strongly supports genetic evaluation; note that phenotypic variability within families is common (e.g., a PPA patient whose parent had parkinsonism may share a GRN variant)
- Young onset (<55 years) — especially if <50, consider MAPT (mean onset 49.5 years)
- Concomitant motor neuron disease features strongly suggest C9orf72 expansion with TDP-43 pathology
- Psychotic features (delusions, hallucinations) in FTD raise suspicion for C9orf72
- Genetic counselor involvement is essential before testing — laws, testing parameters, and cultural attitudes vary widely
- Presymptomatic biomarker changes: Plasma neurofilament light chain deviates from normal ~30 years before symptoms in C9orf72 carriers and ~15 years before symptoms in GRN carriers; MRI may detect volume loss 1–6 years before symptom onset depending on the gene
Progression and Overlap with Other Syndromes
PPA variants do not remain isolated language disorders indefinitely. As neurodegeneration spreads beyond the initial epicenter, patients develop features of other neurodegenerative syndromes. The pattern of evolution provides additional clues to the underlying pathology.
- svPPA → behavioral variant FTD: Semantic deficits extend to nonverbal domains (prosopagnosia, loss of empathy, impaired emotion recognition); many patients eventually meet criteria for bvFTD; right temporal predominance accelerates behavioral overlap
- nfvPPA → corticobasal syndrome or PSP: Asymmetric rigidity, dystonia, myoclonus, alien limb, and/or supranuclear gaze palsy may emerge; corticobasal syndrome frequently begins with or develops nonfluent aphasia
- nfvPPA → FTD-ALS: Motor neuron disease features (fasciculations, weakness, bulbar dysfunction) may develop, most commonly with TDP-43 pathology; occurs in ~15% of FTD patients and ~30% of ALS patients
- lvPPA → typical AD: Progressive amnesia, visuospatial dysfunction, and global cognitive decline emerge as the disease advances beyond the language network
Speech and Language Assessment
A structured approach to language testing is essential for PPA classification. The key domains to assess at the bedside include fluency, naming, comprehension, repetition, reading, and writing. Standardized batteries provide the most reliable classification.
| Assessment Tool | What It Tests | Classification Accuracy |
|---|---|---|
| Sydney Language Battery | Naming, word comprehension, repetition, semantic association | ~80% |
| Mini Linguistic State Examination | Comprehensive brief PPA classification battery | ~91% |
| Northwestern Anagram Test | Syntactic construction ability (written, motor-free) | Identifies agrammatism specifically |
| Word-picture matching | Single-word comprehension (semantic knowledge) | Core test for svPPA |
| Sentence repetition (graded length) | Auditory verbal working memory; phonologic loop | Distinguishes lvPPA (length-dependent failure) |
| Irregular word reading | Surface dyslexia (semantic contribution to reading) | Supports svPPA diagnosis |
Treatment Considerations
No FDA-approved disease-modifying treatments exist for FTLD-associated PPA. Management is primarily supportive, focused on speech-language therapy, behavioral interventions, caregiver education, and off-label pharmacotherapy for comorbid behavioral symptoms. Clinical trials are underway, including some targeting presymptomatic genetic variant carriers.
- Speech therapy: Script training — an approach commonly used in stroke aphasia — may offer sustained benefit for patients with nfvPPA; gains from other speech therapy approaches may be transient and not generalize to untrained material
- Behavioral symptoms: SSRIs (sertraline, citalopram, paroxetine) for disinhibition, irritability, compulsions, and depression; trazodone for irritability, agitation, and disordered eating; DICE model (Describe-Investigate-Create-Evaluate) for tailored behavioral interventions
- Avoid in FTLD: Cholinesterase inhibitors (may worsen cognition and behavior); typical antipsychotics (increased extrapyramidal sensitivity); memantine (well tolerated but no demonstrated benefit)
- lvPPA exception: If AD pathology is confirmed by biomarkers, cholinesterase inhibitors and memantine may be appropriate per standard AD treatment guidelines
- Caregiver support: FTD imposes greater caregiver burden than AD; education about the disease trajectory and nonpharmacologic interventions can reduce caregiver stress
Medication Pitfalls in PPA and FTD
- Cholinesterase inhibitors: Cholinergic projections are largely intact in FTLD; donepezil, rivastigmine, and galantamine are NOT recommended and may cause cognitive and behavioral worsening
- Antipsychotics: Patients with FTLD pathology are more sensitive to extrapyramidal side effects; if needed, use atypical agents at the lowest effective dose and discontinue as soon as possible
- Dopaminergic agents: May help apathy in some cases (e.g., selegiline, amantadine), but evidence is limited
- Confirm the underlying pathology before choosing therapy: lvPPA (AD pathology) and svPPA/nfvPPA (FTLD pathology) require fundamentally different pharmacologic approaches; AD biomarkers should be obtained in all PPA patients when available
References
- Clark DG. Frontotemporal dementia. Continuum (Minneap Minn). 2024;30(6):1642-1672.
- Gorno-Tempini ML, Hillis AE, Weintraub S, et al. Classification of primary progressive aphasia and its variants. Neurology. 2011;76(11):1006-1014.
- Rascovsky K, Hodges JR, Knopman D, et al. Sensitivity of revised diagnostic criteria for the behavioural variant of frontotemporal dementia. Brain. 2011;134(Pt 9):2456-2477.
- Mesulam MM, Coventry C, Bigio EH, et al. Nosology of primary progressive aphasia and the neuropathology of language. Adv Exp Med Biol. 2021;1281:33-49.
- Marshall CR, Hardy CJD, Volkmer A, et al. Primary progressive aphasia: a clinical approach. J Neurol. 2018;265(6):1474-1490.
- Gorno-Tempini ML, Dronkers NF, Rankin KP, et al. Cognition and anatomy in three variants of primary progressive aphasia. Ann Neurol. 2004;55(3):335-346.
- Savage S, Hsieh S, Leslie F, et al. Distinguishing subtypes in primary progressive aphasia: application of the Sydney language battery. Dement Geriatr Cogn Disord. 2013;35(3-4):208-218.
- Patel N, Peterson KA, Ingram RU, et al. A "mini linguistic state examination" to classify primary progressive aphasia. Brain Commun. 2022;4(2):fcab299.
- Lorca-Puls DL, Gajardo-Vidal A, Mandelli ML, et al. Neural basis of speech and grammar symptoms in non-fluent variant primary progressive aphasia spectrum. Brain. 2024;147(2):607-626.
- Ralph MAL, Jefferies E, Patterson K, Rogers TT. The neural and computational bases of semantic cognition. Nat Rev Neurosci. 2017;18(1):42-55.
- Mesulam MM, Coventry CA, Bigio EH, et al. Neuropathological fingerprints of survival, atrophy and language in primary progressive aphasia. Brain. 2022;145(6):2133-2148.
- Roytman M, Chiang GC, Gordon ML, Franceschi AM. Multimodality imaging in primary progressive aphasia. AJNR Am J Neuroradiol. 2022;43(9):1230-1243.
- Greaves CV, Rohrer JD. An update on genetic frontotemporal dementia. J Neurol. 2019;266(8):2075-2086.
- Staffaroni AM, Quintana M, Wendelberger B, et al. Temporal order of clinical and biomarker changes in familial frontotemporal dementia. Nat Med. 2022;28(10):2194-2206.
- Botha H, Josephs KA. Primary progressive aphasias and apraxia of speech. Continuum (Minneap Minn). 2019;25(1):101-127.