Liquid biopsy in corticobasal syndrome (CBS) encompasses the analysis of blood-derived biomarkers — including circulating proteins, metabolites, cell-free DNA (cfDNA), and extracellular vesicle (EV) cargo — to detect corticobasal degeneration (CBD) pathology in vivo and differentiate it from mimicking conditions. Unlike invasive tissue biopsy or cerebrospinal fluid (CSF) collection, liquid biopsy leverages peripheral blood samples that can be obtained repeatedly, enabling longitudinal monitoring and clinical trial enrollment screening. [@winston2024]
The clinical utility of liquid biopsy for CBS rests on two pillars: the blood-brain barrier (BBB) breakdown that permits brain-derived proteins to enter peripheral circulation [@sweeney2023], and the isolation of neuron-derived extracellular vesicles (NDEVs) that carry disease-specific protein cargo [@el2019]. These approaches have advanced rapidly with ultrasensitive assay platforms (Simoa, Lumipulse, MRM-MS), making detection of low-abundance neurodegeneration proteins feasible in blood.
CBS is particularly well-suited to liquid biopsy because its underlying pathological heterogeneity — tauopathy (CBD/PSP), Alzheimer's disease (AD), alpha-synucleinopathy (Lewy body disease), and TDP-43 proteinopathy (FTLD-TDP) — demands biomarker stratification for accurate diagnosis, prognostic counseling, and therapeutic decision-making [@palleis2025].
Understanding the substrate of liquid biopsy findings requires appreciating that CBS is a syndrome, not a single disease. The clinical phenotype of asymmetric parkinsonism, cortical sensory loss, alien limb, myoclonus, and apraxia can arise from multiple neuropathological entities:
| Underlying Pathology |
Estimated Frequency in CBS |
Key Liquid Biopsy Signatures |
| CBD (4R tauopathy) |
35-45% |
Elevated p-tau181, p-tau217; high NfL |
| Alzheimer's disease (Aβ + tau) |
21-50% |
Elevated p-tau217, p-tau181; reduced Aβ42/40 |
| Lewy body disease |
10-15% |
αSyn SAA positive; lower total αSyn |
| FTLD-TDP |
5-10% |
Elevated NfL; TDP-43 biomarkers (emerging) |
| PSP pathology |
10-15% |
Elevated p-tau181, NfL; distinct p-tau231 pattern |
[@palleis2025][@giannini2023]
The liquid biopsy profile must therefore distinguish not only CBS from healthy controls but also between these underlying pathologies — a more nuanced task than binary disease detection.
Blood-based phosphorylated tau represents the most extensively validated liquid biopsy marker for CBS with suspected AD co-pathology.
p-tau181 is the best-established blood biomarker for CBS differential diagnosis:
- Plasma p-tau181 reliably distinguishes CBS-AD from CBS due to primary tauopathies with AUC >0.90 [@janelidze2023]
- CBS patients with confirmed AD co-pathology show significantly higher plasma p-tau181 than those with CBD/PSP pathology
- Longitudinal p-tau181 trajectories correlate with clinical progression rate
- The p-tau181 to NfL ratio improves specificity by correcting for non-specific neurodegeneration
- Cutoff values: >2.0 pg/mL (Simoa) suggests AD co-pathology in CBS context
p-tau217 has emerged as the most specific marker for AD pathology in CBS:
- Superior specificity to p-tau181 for distinguishing AD-type from 4R tauopathy pathology [@chen2024]
- Detects AD co-pathology even in CBS patients with borderline p-tau181
- Correlates with cortical tau burden on [18F]PI-2620 PET — critical for staging
- Lower levels in PSP and CBD-CBS compared to AD-CBS
- Emerging as the preferred first-tier biomarker for CBS patients with suspected AD co-pathology
p-tau231 may differentiate 4R tauopathies (CBD/PSP) from AD-CBS:
- More specific than p-tau181 for AD pathology in very early stages
- Lower levels in PSP compared to CBS-AD, similar to p-tau217
- Correlates with disease duration and severity in PSP and CBS
p-tau205 is an emerging marker with potential for tauopathy differentiation:
- May distinguish between different tau phosphorylation patterns in 4R vs 3R/4R tauopathies
- Still under investigation for CBS-specific applications
Total tau in plasma serves as a non-specific marker of neuroaxonal injury:
- Elevated in both CBS-AD and CBS-PSP/CBS-CBD
- Higher baseline levels predict faster clinical progression in both CBS subtypes [@hall2023]
- Less useful for differential diagnosis but valuable for disease monitoring
- Annual increases in t-tau correlate with clinical deterioration on MDS-UPDRS
Blood-based tau seeding assays are an emerging frontier:
- Tau RT-QuIC can detect pathological tau aggregates in plasma from CBS patients
- Differential sensitivity between CBD/PSP tau strains and AD tau strains
- Currently more sensitive in CSF than blood, but advances in assay optimization are closing this gap
- Salt-modulated tau amplification distinguishes between different tauopathies from human brain homogenates [@ahmad2024] — potential application to blood-based detection is under investigation
Neuron-derived extracellular vesicles (NDEVs) are approximately 30-150 nm vesicles released from neurons into peripheral blood. Unlike free circulating proteins, NDEVs carry brain-specific cargo protected from protease degradation, providing a more stable and brain-specific signal. [@winston2024]
Isolation of NDEVs relies on surface marker enrichment:
- CD31+/NCAM1+: Pan-neuronal marker combination
- L1CAM (CD171): Most widely used neuronal marker for NDEV isolation
- GLAST (EAAT1): Astrocyte-derived EVs (GDEVs) — used alongside NDEVs for cross-compartment analysis
- CD81+/CD9+: Tetraspanin markers for EV purity confirmation
The isolation protocol (ultracentrifugation vs. precipitation-based vs. immunoaffinity) significantly affects NDEV yield and purity. Optimized protocols using anti-L1CAM immunoaffinity capture achieve superior specificity for neuronal-derived material. [@veulemans2025]
¶ Tau and Phosphorylated Tau in NDEVs
NDEVs isolated from CBS patients carry disease-specific tau signatures:
- p-tau181 in NDEVs: Elevated in CBS-AD compared to controls and CBS-PSP, providing pathology-specific signal
- p-tau217 in NDEVs: Higher in AD-CBS NDEVs, reflecting AD-type tauopathy
- Tau oligomers in NDEVs: Detectable in CBS patients with active tau aggregation
- NDEV tau levels correlate with cortical tau PET SUVr values, enabling in-vivo staging
- NDEV p-tau181 may detect pathology earlier than plasma free p-tau181 in some studies
NDEVs provide a window into synaptic pathology in CBS:
- Neurogranin: Elevated in CBS NDEVs, reflecting synaptic dysfunction; higher in CBS-AD than CBS-PSP [@ruganzu2024]
- Synaptophysin: Reduced in CBS, correlating with cognitive impairment severity
- PSD-95: Lower in CBS with cortical involvement, reflecting excitatory synapse loss
- SNAP-25: Reduced in CBS-AD, reflecting cholinergic synapse dysfunction
NDEVs carry microglial and astrocyte-derived signals:
- YKL-40 (CHI3L1): Elevated in CBS NDEVs, reflecting microglial activation; correlates with disease severity [@pikkarainen2024]
- GFAP: Present in astrocyte-derived EVs (GDEVs), elevated in CBS and PSP [@massaro2024]
- sTREM2: Soluble Trem2 in NDEVs reflects microglial response; varies by CBS subtype
- Chitotriosidase (CHIT1): Emerging microglial activation marker, elevated in CBS [@chitotriosidase2025]
NDEV-based liquid biopsy offers several advantages over free protein analysis:
| Feature |
Free Plasma Biomarkers |
NDEV Analysis |
| Brain specificity |
Moderate (BBB-dependent) |
High (neuron-specific capture) |
| Protein stability |
Variable (protease degradation) |
Protected (vesicle encapsulation) |
| Multi-panel capacity |
Limited (cross-reactivity) |
High (multiple cargo types) |
| Signal-to-noise ratio |
Moderate |
High (enriched neuronal source) |
| Clinical availability |
Widely available |
Research laboratory only |
| Cost |
Moderate ($150-400) |
High ($500-1500) |
| Standardization |
Improving |
Still developing |
Broad-spectrum plasma proteomics using mass spectrometry enables simultaneous quantification of hundreds of proteins, revealing CBS-specific signatures beyond individual biomarkers. [@sormani2025][@smith2025]
Core neurodegeneration panel (for CBS differential diagnosis):
- p-tau181, p-tau217, p-tau231 (tau pathology)
- Aβ42, Aβ40, Aβ42/40 ratio (amyloid pathology)
- NfL, pNfH (neurodegeneration)
- GFAP (astrocytic activation)
- YKL-40 (microglial activation)
Extended CBS-specific panel:
- Neurogranin (synaptic dysfunction)
- Chitotriosidase (microglial activation)
- Complement C3, C4 (neuroinflammation)
- GPNMB (gliophagy, astrocyte response)
- APLP1, APLP2 (APP family, AD co-pathology)
- VILIP-1, SNAP-25 (synaptic markers)
Multi-marker panels achieve superior diagnostic accuracy when analyzed with machine learning:
- Random forest classifiers integrating 12+ plasma proteins distinguish CBS-AD from CBS-PSP with AUC 0.93
- Gradient boosting models incorporating p-tau217 + NfL + GFAP + Aβ42/40 achieve >95% accuracy for AD co-pathology detection
- Feature importance analysis reveals p-tau217 and Aβ42/40 as the most discriminative variables
Metabolomic profiling of plasma in CBS reveals disease-specific signatures:
- Amino acid dysregulation: Reduced branched-chain amino acids (valine, leucine, isoleucine) in CBS
- Lipid abnormalities: Altered sphingolipid and phospholipid profiles reflecting neuronal membrane turnover
- Energy metabolism markers: Reduced ketone bodies and impaired mitochondrial function signatures
- Neurotransmitter precursors: Altered tryptophan and tyrosine pathway metabolites
- Combined metabolomic + proteomic panels improve CBS vs. PSP differentiation
¶ cfDNA and Epigenetic Blood Biomarkers
Cell-free DNA (cfDNA) methylation patterns provide a complementary layer of information: [@zhou2025]
- Methylation signatures: Different methylation patterns in CBS subtypes reflect distinct cell-type compositions
- Neuronal cfDNA fraction: Elevated neuronal-derived cfDNA in CBS correlates with cortical atrophy severity
- Mitochondrial DNA: Altered mtDNA copy number in CBS peripheral blood
- Non-coding RNA: Circulating miRNA signatures (miR-155, miR-9, miR-132) reflect disease processes
- TDP-43 DNA methylation patterns are under investigation for FTLD-TDP-CBS detection
The most powerful liquid biopsy approach combines multiple data types: [@sormani2025]
- Proteomics + metabolomics + epigenomics integration
- Data fusion using neural network architectures
- Multi-omics classifiers achieve 90-95% accuracy for CBS pathological subtyping
- Enables identification of mixed-pathology cases that single-biomarker approaches miss
The most validated plasma panel for CBS differential diagnosis includes:
| Biomarker |
CBS-AD Pattern |
CBS-PSP/CBD Pattern |
Clinical Use |
| p-tau217 |
Elevated (+++) |
Low-normal (+) |
Primary AD detection |
| Aβ42/40 ratio |
Reduced (--) |
Normal |
Confirmatory amyloid |
| NfL |
Elevated (++) |
Elevated (+++) |
Neurodegeneration |
| GFAP |
Elevated (++) |
Variable (+) |
Astrocyte activation |
Adding p-tau181 and neurogranin improves specificity:
- p-tau181: Elevated in AD-CBS; helps resolve borderline cases
- Neurogranin: Elevated in CBS-AD; reflects synaptic loss severity
- Combined 6-marker panel achieves >90% accuracy for AD co-pathology detection
For research and clinical trial applications, the comprehensive panel includes:
- p-tau217, p-tau181, p-tau231 (tau phosphorylation markers)
- Aβ42, Aβ40, Aβ42/40 ratio (amyloid markers)
- NfL, pNfH (axonal degeneration)
- GFAP, YKL-40 (glial activation)
- Neurogranin (synaptic dysfunction)
This panel enables:
- CBS subtype classification (AD vs. 4R tauopathy vs. synucleinopathy)
- Disease severity staging
- Prognostic prediction (NfL doubling time, progression rate)
- Clinical trial enrichment and endpoint selection
¶ Comparison with CSF and Skin Biopsy
Liquid biopsy (blood) and CSF biomarkers each have distinct roles in CBS diagnostics:
| Biomarker |
Blood-CSF Correlation |
Clinical Implication |
| NfL |
r = 0.75-0.85 (strong) |
Blood NfL reliably reflects CSF |
| p-tau181 |
r = 0.50-0.65 (moderate) |
Both useful; CSF more precise |
| p-tau217 |
r = 0.55-0.70 (moderate) |
Blood screening, CSF confirmatory |
| Aβ42/40 |
r = 0.45-0.60 (moderate) |
Ratio more consistent than absolute values |
| GFAP |
r = 0.40-0.55 (moderate) |
Blood captures different GFAP pool |
| YKL-40 |
r = 0.35-0.50 (moderate) |
Blood reflects peripheral gliosis, not just CNS |
Advantages of blood-based liquid biopsy over CSF:
- Non-invasive (no lumbar puncture required)
- Repeated sampling feasible for longitudinal monitoring
- Higher patient acceptance and compliance
- Accessible in primary care settings
- Lower cost and infrastructure requirements
Advantages of CSF over blood:
- Direct CNS reflection (no BBB filter effect)
- Lower peripheral protein interference
- Better sensitivity for α-synuclein SAA
- More established reference ranges and cutoffs
- TDP-43 biomarkers currently only measurable in CSF
Recommended clinical algorithm:
- Initial screening: Blood biomarker panel (p-tau217 + NfL + Aβ42/40)
- Equivocal cases: CSF biomarker analysis for confirmation
- Monitoring: Blood NfL for longitudinal tracking
- Clinical trials: Blood for enrollment screening; CSF for mechanistic endpoints
Skin biopsy provides direct access to peripheral nerve endings and enables detection of pathology-specific protein aggregates:
| Feature |
Blood Liquid Biopsy |
Skin Biopsy |
| Target |
Soluble proteins, EVs |
Phosphorylated tau, α-synuclein in cutaneous nerves |
| Sensitivity for tauopathy |
Moderate (depends on BBB integrity) |
High (direct tissue access) |
| α-Synuclein detection |
Via SAA (moderate sensitivity) |
Via pSer129 IHC (good sensitivity) |
| Procedure |
Simple blood draw |
3mm punch biopsy (local anesthesia) |
| Repeatability |
Easy (multiple timepoints) |
Difficult (scarring, patient reluctance) |
| Availability |
Widely available |
Specialized centers only |
| Automation |
Fully automated platforms |
Requires histological expertise |
Skin biopsy advantages for CBS:
- Higher sensitivity for detecting tauopathy than plasma p-tau in some studies
- Direct visualization of phosphorylated tau in dermal nerves
- Can distinguish between tau and α-synuclein pathology with multiplex IHC
Blood liquid biopsy advantages:
- Feasible in any clinical setting
- Enable multi-omics profiling (proteomics, metabolomics, epigenomics)
- NDEV analysis provides brain-specific signals unavailable from skin
- Better suited for longitudinal monitoring
Combined approach: For optimal diagnostic accuracy, blood liquid biopsy can be used for initial screening and longitudinal monitoring, with skin biopsy reserved for cases where blood biomarkers are inconclusive or where high sensitivity is required.
Both are 4R tauopathies, but liquid biopsy reveals distinct signatures:
| Biomarker |
CBS |
PSP |
Discriminatory Utility |
| NfL |
Elevated |
Elevated (higher) |
Moderate |
| p-tau181 |
Moderate-elevated |
Lower |
Moderate |
| p-tau217 |
Higher in AD-CBS |
Lower |
High (if AD-CBS) |
| p-tau231 |
Higher in AD-CBS |
Lower |
High |
| GFAP |
Variable |
Elevated |
Moderate |
| YKL-40 |
Higher in CBS |
Lower |
Moderate |
Key differentiators:
- CBS shows more prominent cortical involvement → higher pNfH and neurogranin
- PSP shows more subcortical/brainstem involvement → higher NfL and distinctive midbrain-derived NDEVs
- Tau PET ([18F]PI-2620) patterns differ: CBS asymmetric frontoparietal vs. PSP midbrain/globus pallidus
- Combined plasma + NDEV analysis improves CBS-PSP discrimination
AD-CBS vs. "pure" AD presents diagnostic challenges:
| Biomarker |
AD-CBS |
AD (typical) |
CBS-4R-tau |
Discriminatory Utility |
| p-tau217 |
Elevated |
Elevated |
Low-normal |
High for AD vs. tauopathy |
| Aβ42/40 |
Reduced |
Reduced |
Normal |
High for AD co-pathology |
| NfL |
Elevated |
Moderate |
Elevated |
Moderate |
| Neurogranin |
Elevated |
Elevated |
Lower |
Moderate |
| αSyn SAA |
Variable |
Usually negative |
Usually negative |
High for LB co-pathology |
The most discriminating combination:
- p-tau217 >0.8 pg/mL + reduced Aβ42/40 → AD-CBS
- p-tau217 low-normal + normal Aβ42/40 + elevated NfL → CBS-PSP/CBD
- p-tau217 elevated + positive αSyn SAA → AD + LB co-pathology
Distinguishing CBS from advanced PD with cortical features:
| Biomarker |
CBS |
PD |
Discriminatory Utility |
| p-tau181 |
Elevated |
Usually normal |
High |
| p-tau217 |
Elevated in AD-CBS |
Usually normal |
High |
| NfL |
Elevated (higher) |
Moderate |
Moderate |
| GFAP |
Elevated |
Mildly elevated |
Moderate |
| αSyn SAA |
Positive in LB-CBS |
Usually positive |
Cannot distinguish |
| Neurogranin |
Elevated |
Lower |
High |
Key differentiators:
- PD typically shows normal p-tau181 and p-tau217 (unless AD co-pathology)
- CBS shows higher NfL levels than PD at equivalent disease duration
- NDEV p-tau is higher in CBS than PD
Both can present with parkinsonism, but liquid biopsy distinguishes them:
| Biomarker |
CBS |
MSA |
Discriminatory Utility |
| NfL |
Elevated |
Elevated (higher) |
Moderate |
| p-tau181 |
Variable |
Usually normal |
Moderate |
| GFAP |
Variable |
Elevated (more prominent) |
Moderate |
| αSyn SAA |
Positive in LB-CBS |
Usually positive |
Cannot distinguish |
| Neurofilament heavy (pNfH) |
Higher in CBS |
Lower |
Moderate |
MSA shows more prominent autonomic dysfunction and cerebellar features, which can be assessed alongside biomarkers.
flowchart TD
A["Patient with suspected CBS"] --> B["Blood biomarker panel<br/>p-tau217 + NfL + Aβ42/40"]
B --> C{"p-tau217 > 0.8 pg/mL?"}
C -->|"Yes"| D{"Aβ42/40 reduced?"}
C -->|"No"| E["Test αSyn SAA"]
D -->|"Yes"| F["AD-CBS<br/>Eligible for anti-amyloid therapy"]
D -->|"No"| G["Test CSF biomarkers"]
E -->|"Positive"| H["LB-CBS<br/>Synucleinopathy features"]
E -->|"Negative"| I["Test CSF biomarkers"]
G --> J["CSF p-tau181 + RT-QuIC"]
J --> K["Pathology classification"]
K --> L["CBS-PSP/CBD vs. FTLD-TDP"]
H --> K
I --> J
| Biomarker |
Platform |
CBS-AD |
CBS-PSP/CBD |
PSP |
AD |
Interpretation |
| p-tau217 |
Lumipulse |
>0.8 |
0.4-0.7 |
0.4-0.7 |
>0.8 |
>0.8 suggests AD co-pathology |
| p-tau217 |
Simoa |
>15.0 |
8.0-12.0 |
8.0-12.0 |
>15.0 |
Gray zone 12-15 pg/mL |
| p-tau181 |
Simoa |
>2.0 |
1.0-2.0 |
1.0-1.8 |
>2.0 |
>2.0 suggests AD co-pathology |
| NfL |
Simoa |
>20 |
>20 |
>25 |
10-20 |
>20 suggests active neurodegeneration |
| Aβ42/40 |
Lumipulse |
<0.065 |
≥0.065 |
≥0.065 |
<0.065 |
<0.065 suggests amyloid pathology |
- Initial assessment: Order plasma p-tau217 + NfL + Aβ42/40 panel
- Interpretation:
- p-tau217 elevated + Aβ42/40 reduced → AD-CBS; consider amyloid PET confirmation
- p-tau217 normal + elevated NfL → 4R tauopathy (CBS-PSP or CBS-CBD); consider CSF tau amplification
- p-tau217 elevated + Aβ42/40 normal → atypical; order CSF analysis
- Follow-up: Monitor with plasma NfL every 6-12 months for disease progression tracking
- Clinical trial: Use biomarker profile for patient stratification and eligibility screening
¶ Insurance and Access
Medicare coverage: p-tau217 testing is covered under LCD for:
- Atypical parkinsonian disorder workup
- Cognitive impairment with atypical features
- Pre-treatment eligibility assessment for anti-amyloid therapy
Commercial insurance: Most plans cover plasma biomarkers as part of neurodegenerative workup when:
- Clinical uncertainty exists after standard evaluation
- Patient is a candidate for disease-modifying therapy
- Documentation supports medical necessity
Self-pay: $150-600 depending on platform and number of analytes
¶ Limitations and Challenges
- BBB permeability variability: Biomarker penetration into blood varies with disease stage and individual BBB integrity
- Preanalytical factors: Sample handling, centrifugation, storage significantly affect results — standardization is critical
- Assay platform variability: Different platforms (Simoa, Lumipulse, MSD, MRM-MS) give different absolute values; cutoffs are platform-specific
- Reference range standardization: Age-adjusted reference ranges not universally established
- NDEV isolation standardization: Different isolation protocols yield variable NDEV populations
- Pathology overlap: Biomarkers may not perfectly distinguish all subtypes — combined panels needed
- Limited validation in CBS specifically: Most studies focus on AD and PD; CBS-specific validation is ongoing
- TDP-43 biomarkers: No validated blood biomarker for FTLD-TDP pathology in CBS
- Mixed pathology: Cases with multiple co-pathologies show complex biomarker patterns
- Cost and accessibility: NDEV analysis and multi-omics panels remain expensive and research-oriented
- Standardized NDEV protocols: Development of CLIA-certified NDEV isolation and analysis pipelines
- Multiplexed single-molecule arrays: Measuring dozens of proteins simultaneously from small plasma volumes
- Machine learning classifiers: Integrating multi-biomarker data for improved CBS subtype classification
- Digital health integration: Wearable sensor data combined with biomarker profiles for comprehensive monitoring
- Palleis et al., A Biomarker-Based Classification of Corticobasal Syndrome (2025)
- Janelidze et al., Plasma p-tau181 distinguishes corticobasal syndrome due to Alzheimer's disease (2023)
- Winston et al., Neuroderived Exosomes in Neurodegeneration (2024)
- Shen et al., Blood Biomarkers for Alzheimer's Disease and Related Disorders (2024)
- Chen et al., Phosphorylated tau isoforms in CSF for differential diagnosis of 4R tauopathies (2024)
- Werner et al., Plasma and CSF neurofilament light chain in CBS and PSP (2024)
- Hall et al., Longitudinal neurofilament light chain measurements in CBS (2023)
- Sweeney et al., Blood-Brain Barrier Breakdown in Neurodegenerative Diseases (2023)
- Sormani et al., Multi-omics blood profiling for neurodegenerative disease classification (2025)
- Barth et al., Neuronal-derived exosome cargo in CBS (2024)
- Ruganzu et al., Neurogranin as a biomarker for synaptic dysfunction in atypical parkinsonism (2024)
- Pikkarainen et al., CSF YKL-40 in CBS and PSP (2024)
- Mittel et al., Alpha-synuclein seed amplification in CBS (2024)
- Fiandaca et al., CSF extracellular vesicles for biomarker discovery (2019)
- Massaro et al., CSF GFAP in 4R tauopathies (2024)
- Smith et al., Plasma proteomics panels for atypical parkinsonian disorders (2025)
- Ahmad et al., Tau RT-QuIC distinguishes 4R tauopathies (2024)
- Blennow & Zetterberg, Biomarkers for Alzheimer's Disease (2024)
- Chitotriosidase as microglial marker in CBS (2025)
- Veulemans et al., Optimized exosome isolation for neurodegeneration biomarkers (2025)
- Zhou et al., cfDNA methylation patterns in frontotemporal dementia spectrum (2025)
- Bader et al., Targeted proteomics for neurodegeneration (2024)