Alpha-Synuclein Target Engagement: What CSF and Imaging Can’t Show You
Alpha-Synuclein Target Engagement
Two alpha-synuclein-targeting antibodies reached Phase 2 with strong cerebrospinal fluid (CSF) target engagement data and came up short on the clinical endpoints that mattered. A third program showed detectable CSF antibody titers in fewer than half of dosed patients at the timepoint that counted most. None of this is a failure of the alpha-synuclein hypothesis. It’s a measurement problem — and it’s one that CSF and imaging, on their own, were never built to solve.
Key Takeaways
- CSF alpha-synuclein seed amplification assays (SAA) report a categorical result — positive, negative, or inconclusive — making them well suited to diagnosis but poorly suited to tracking dose-dependent, longitudinal changes across a treatment period.
- DAT-SPECT and other imaging endpoints move slowly and showed no significant treatment effect in either the PASADENA (prasinezumab) or SPARK (cinpanemab) Phase 2 trials, despite documented target engagement in blood and CSF [5,6].
- Quantitative systems pharmacology modeling shows that target engagement measured in CSF can overstate the antibody’s actual reach into the synaptic cleft by more than an order of magnitude [8].
- A neuron-derived extracellular vesicle (NDEV)-enriched plasma alpha-synuclein readout adds a repeatable, blood-based signal that can be sampled far more frequently than CSF or imaging — without replacing either.
- Sponsors running alpha-synuclein-targeting immunotherapies, ASOs, or small molecules can use this peripheral layer to close the gap between a categorical diagnostic readout and a continuous pharmacodynamic one.
Why Isn’t CSF Alpha-Synuclein Enough to Track Target Engagement in a Trial?
CSF seed amplification assays confirm the presence of misfolded alpha-synuclein and support patient selection, but they were designed as a diagnostic classifier, not a longitudinal pharmacodynamic tool. A positive or negative SAA result doesn’t scale with dose, doesn’t move predictably over a treatment period, and can’t be sampled often enough to build a dose-response curve without repeated lumbar punctures — which is exactly the lumbar puncture alternative biopharma teams are trying to avoid in the first place.
The Alpha-Synuclein Drug Pipeline Has a Biomarker Blind Spot
Alpha-synuclein-targeting therapeutics — active immunotherapies, monoclonal antibodies, antisense oligonucleotides, small-molecule aggregation inhibitors — are one of the most active areas of Parkinson’s disease (PD) and broader synucleinopathy drug development. Nearly every program in this space has converged on the same two biomarker anchors: a CSF alpha-synuclein seed amplification assay to confirm biological status, and dopamine transporter (DAT) imaging to track dopaminergic integrity. Both are legitimate, well-validated tools. Neither one was built to answer the question a translational medicine team actually needs answered mid-trial: is the drug engaging its target, and is that engagement tracking with dose?
The Parkinson’s Progression Markers Initiative (PPMI) helped establish CSF SAA as the closest thing the field has to a biological definition of alpha-synuclein disease, and the assay’s diagnostic accuracy across large PPMI and multi-cohort datasets is genuinely strong [1,2]. Refinements to the underlying detection chemistry — the seed amplification method itself, sometimes still referred to under its earlier names of real-time quaking-induced conversion (RT-QuIC) or protein misfolding cyclic amplification — have pushed sensitivity and specificity for PD diagnosis into the 85–98% range in published protocols [3]. Independent comparisons of CSF SAA platforms across academic and commercial labs have shown these assays converge well on diagnostic classification [4].
None of that changes what the assay is designed to output: a categorical call. It tells a program whether misfolded alpha-synuclein is present. It does not, on its own, give a drug developer a quantitative, repeatable readout of how much of that pathological signal has changed since the last dose — which is a fundamentally different measurement problem, and one that matters enormously once a therapeutic actually starts working on its target.
When Target Engagement Data Doesn’t Predict the Clinical Outcome
The two most closely watched alpha-synuclein antibody programs of the past several years illustrate exactly where this gap bites. In the SPARK trial, cinpanemab produced no significant change in the Movement Disorder Society-sponsored Unified Parkinson’s Disease Rating Scale (MDS-UPDRS) total score at 52 or 72 weeks, and no meaningful shift on DAT-SPECT imaging in any dose group compared with placebo [5]. In the parallel PASADENA trial, prasinezumab also missed its primary MDS-UPDRS endpoint, though a prespecified exploratory analysis of patients with faster motor progression showed a numerical trend favoring treatment on the motor subscale — a signal that has since carried into an ongoing Phase 2b/3 program [6,7].
Both trials had access to the two standard tools: CSF-adjacent target engagement measures and DAT imaging. Neither tool explained, in real time, why apparent target engagement wasn’t translating into the expected clinical or imaging benefit. A quantitative systems pharmacology (QSP) analysis published after these trial readouts modeled exactly this disconnect. Using a physiologically based pharmacokinetic framework calibrated against clinical CSF and interstitial fluid (ISF) data, the model found that while several anti-alpha-synuclein and anti-tau antibodies achieved substantial target engagement in the experimentally accessible CSF compartment, engagement of the actual pathological protein pool in the synaptic cleft — the site presumed relevant to disease-modifying mechanism of action — was dramatically lower, in some cases falling from near-complete CSF engagement to single-digit percentages once diffusion into brain interstitial space and the synaptic cleft was accounted for [8]. CSF target engagement, in other words, can systematically overstate what’s actually happening where the drug needs to act.
This is precisely the kind of finding that reframes what a peripheral, non-CSF biomarker layer is actually for. It isn’t a replacement for CSF or imaging — it’s a way to generate the sampling density needed to catch a dose-response relationship, or the absence of one, well before a Phase 2 primary endpoint readout.
Beyond Antibodies: ASOs, Small Molecules, and the Same Blind Spot
Monoclonal antibodies aren’t the only modality running into this measurement gap. Antisense oligonucleotides (ASOs) designed to reduce alpha-synuclein expression at the transcript level face an even sharper version of the same problem: an ASO’s mechanism of action is upstream of protein aggregation altogether, which means a CSF SAA result — designed to detect misfolded, aggregated protein — may lag well behind the biological effect a knockdown program is actually producing. A sponsor running an alpha-synuclein-lowering ASO needs a pharmacodynamic signal that can move earlier and more frequently than a categorical aggregation assay, ideally one sensitive enough to register a change in the pool of neuron-derived alpha-synuclein reaching the periphery before aggregate load has shifted enough to flip a CSF SAA call.
Small-molecule aggregation inhibitors and stabilizers sit somewhere between the two. Like antibodies, they’re often designed to act directly on misfolded species; like ASOs, many programs want to demonstrate a dose-dependent pharmacodynamic effect well before a Phase 2 clinical readout, ideally from Phase 1 healthy-volunteer or early patient cohorts where lumbar puncture burden is hardest to justify and imaging is rarely powered to detect anything. Across all three modalities — immunotherapy, ASO, and small molecule — the underlying need is identical: a CNS target engagement blood-based measurement that can be collected on the same visit schedule as PK and safety labs, doesn’t require an additional invasive procedure, and produces a quantitative result rather than a categorical one.
This is also where the case extends past Parkinson’s disease specifically. Alpha-synuclein pathology is the defining feature not just of PD but of the broader synucleinopathy spectrum — dementia with Lewy bodies (DLB), multiple system atrophy (MSA), and pure autonomic failure among them — and drug programs targeting any of these face the same CSF-and-imaging default, often with even thinner natural history data to fall back on. A peripheral, NDEV-enriched alpha-synuclein biomarker platform built once can, in principle, support pharmacodynamic monitoring across that entire indication set rather than requiring a bespoke blood test for each disease-specific program.
Figure 1. Where CSF, Imaging, and Peripheral NDEV Biomarkers Each Fit
| Attribute | CSF Alpha-Synuclein SAA | DAT-SPECT / MRI Imaging | Plasma NDEV Alpha-Synuclein |
|---|---|---|---|
| Output type | Categorical (positive/negative/inconclusive) | Continuous but slow-moving | Continuous, quantitative |
| Sampling burden | Lumbar puncture; typically 1–3 timepoints per patient | Scan visit; typically annual or less frequent | Venous blood draw; compatible with visit-level frequency |
| Time to detect change | Not designed for short-interval change detection | Months to years | Weeks, in principle, given assay reproducibility |
| Primary use case | Diagnosis, patient selection, biological staging | Structural/functional disease progression | Pharmacodynamic / target engagement monitoring |
| Site-of-action relevance | Reflects CSF compartment, not synaptic cleft | Reflects nigrostriatal dopaminergic integrity | Reflects neuron-derived cargo circulating peripherally |
What a Peripheral NDEV Alpha-Synuclein Readout Actually Adds
Whole-plasma alpha-synuclein has never been a workable blood test on its own. The overwhelming majority of alpha-synuclein circulating in blood originates from red blood cells and other peripheral sources, not neurons, which floods any unfractionated plasma measurement with signal that has nothing to do with CNS pathology [9]. This is the specific limitation that plasma p-tau and other successful blood-based Alzheimer’s biomarkers didn’t have to contend with in the same way, and it’s a large part of why blood-based alpha-synuclein lagged years behind blood-based tau and neurofilament light (NfL) as a usable modality.
Neuron-derived extracellular vesicles change that equation. By enriching for the sub-population of circulating EVs that originate from neurons before quantifying alpha-synuclein cargo, published assays have repeatedly shown that neuron-derived exosomal alpha-synuclein tracks with PD status and severity in ways that unfractionated plasma alpha-synuclein does not, across cohorts spanning several hundred to over 700 subjects internationally [10]. That enrichment step — isolating the neuron-derived fraction before measurement — is the mechanism that turns a noisy peripheral tissue signal into a biomarker candidate worth building a pharmacodynamic program around.
For a drug program, the practical value isn’t diagnostic. It’s operational. A peripheral, blood-based alpha-synuclein readout can be collected at ordinary visit cadence, alongside routine PK sampling, without adding a procedure burden most patients and sites already push back on when it comes to repeat lumbar punctures. That sampling density is what turns a single CSF timepoint and an annual scan into an actual longitudinal curve — the kind of curve a translational team needs to see whether a dose is doing anything before the trial reaches its primary readout.
NeuroDex’s own recent work applying this NDEV-enrichment approach specifically to alpha-synuclein has extended earlier synaptic-protein NDEV methodology into the synucleinopathy space, evaluating neuron-derived vesicle alpha-synuclein as a blood-based biomarker across Parkinson’s disease, dementia with Lewy bodies, and related conditions [11,12]. The consistent thread across this and the broader NDEV literature is that enrichment — pulling the neuronal signal out of a peripheral tissue dominated by non-neuronal noise — is what separates a usable assay from an unusable one.
It’s worth being direct about how this compares to ultrasensitive digital immunoassay platforms like Simoa, which have driven most of the recent progress in blood-based NfL and p-tau181 measurement. Simoa’s advantage is raw analytical sensitivity — it can detect proteins at femtomolar concentrations directly from whole plasma or serum. That advantage matters less for alpha-synuclein specifically, because the limiting factor isn’t detecting the protein — plasma alpha-synuclein concentrations are actually comparatively high — it’s that almost all of what’s detected is peripherally derived and biologically irrelevant to CNS pathology. Running whole plasma through an ultrasensitive digital assay doesn’t fix that; it just measures the peripheral noise more precisely. The enrichment step has to happen before quantification, not instead of it, which is why NDEV isolation and immunoassay quantification are complementary steps rather than competing platforms. A well-designed program pairs neuron-derived EV enrichment with a sufficiently sensitive downstream readout, rather than treating platform sensitivity alone as the solution to a sample-composition problem.
Reproducibility is the other question every translational team asks before committing a Phase 1/2 biomarker plan to a new modality, and it’s a fair one — NDEV isolation as a category has drawn legitimate scrutiny over inter-lab variability and pre-analytical handling. Multi-cohort validation work spanning hundreds of subjects across independent sites has shown that neuron-derived exosomal alpha-synuclein measurements can converge on consistent diagnostic cutoffs when isolation protocols are standardized and pre-analytical variables are controlled [10]. That’s the bar any NDEV-based pharmacodynamic biomarker needs to clear before a sponsor builds a go/no-go decision on top of it, and it’s the reason NeuroDex’s own validation work has focused on cross-cohort consistency as a precondition for use in a drug-program setting rather than a discovery-stage nicety.
Figure 2. The Target Engagement Gradient Problem
Illustrative visualization based on the QSP modeling data reported for anti-alpha-synuclein antibodies [8]. Bar chart: X-axis = compartment (CSF, ISF, synaptic cleft); Y-axis = modeled percent target engagement at clinical dose. Takeaway: engagement measured in the experimentally accessible CSF compartment substantially overstates engagement at the presumed site of action.
Target engagement (%) by compartment — illustrative model output
CSF ████████████████████████████████████████ 80–99%
ISF ██████ 4–6%
Synaptic cleft █ 1–2%
Building Peripheral Alpha-Synuclein Into a Phase I/II Biomarker Plan
None of this argues for dropping CSF SAA or imaging from an alpha-synuclein program’s biomarker plan — both remain the field’s accepted tools for diagnosis, patient stratification, and regulatory-facing disease staging, and neither is going away as the field’s biological anchor [2]. The argument is narrower: a peripheral, NDEV-enriched alpha-synuclein readout fills the specific gap those two tools leave open, which is frequent, quantitative, dose-linked pharmacodynamic tracking between the diagnostic timepoints.
A practical framework looks like this. Use CSF SAA at screening to confirm biological eligibility. Use DAT imaging on its existing annual or biannual cadence to track the disease-progression endpoint the field already trusts. Layer in peripheral NDEV alpha-synuclein at every dosing visit — the same blood draw already scheduled for PK and safety labs — to build a within-patient pharmacodynamic curve that shows whether target engagement is tracking with dose, and whether it’s doing so early enough to inform a go/no-go decision before the trial reaches its powered clinical endpoint. This is the same logic increasingly applied across other CNS drug classes for pharmacodynamic monitoring in neurodegeneration trials generally, not just in synucleinopathies — a peripheral, sampling-dense layer that complements, rather than competes with, the field’s existing CSF and imaging endpoints.
For companion diagnostic and natural history programs running in parallel with a therapeutic pipeline, the same NDEV alpha-synuclein readout also supports patient stratification, letting a sponsor separate rapid progressors from stable patients earlier than an imaging-based read would allow, and feed that stratification back into trial design for the next cohort.
There’s also a patient-burden argument that translational teams tend to underweight until a site or IRB pushes back on it directly. The multicenter Systemic Synuclein Sampling Study — designed specifically to test whether repeated tissue and biofluid collection for alpha-synuclein research, including CSF, was feasible and safe across sites — was itself built around the recognition that multi-timepoint biofluid collection in this population needed dedicated feasibility testing before it could be assumed to work at scale [13]. Every additional CSF timepoint a protocol adds is a separate negotiation with sites, a separate informed-consent conversation with patients, and, in a multi-year Phase 2/3 program, a real driver of dropout risk. A peripheral biomarker that piggybacks on an already-scheduled blood draw sidesteps that negotiation entirely, which is as much an operational and enrollment argument as it is a scientific one.
None of this is a small-biotech-only story, either. Sponsors running biomarker qualification programs aimed at eventual FDA or EMA acceptance of a novel blood-based CNS biomarker — a multi-year, often multi-sponsor undertaking — need exactly this kind of cross-program, reproducible peripheral dataset to build the case. A single Phase 2 program generating NDEV alpha-synuclein data in isolation is useful to that program; a biomarker platform generating comparable data across multiple alpha-synuclein-targeting programs, with consistent isolation and quantification methodology, is what eventually supports a qualification package regulators can act on.
Figure 3. Sampling Density Across a 52-Week Trial
Illustrative timeline comparing how often each biomarker modality can realistically be collected across a one-year dosing period, based on typical trial designs referenced in [5,6].
| Modality | Typical timepoints in a 52-week trial | Added procedure burden |
|---|---|---|
| CSF SAA / CSF antibody titer | 1–3 (baseline, occasionally mid-study, end-of-study) | Lumbar puncture per timepoint |
| DAT-SPECT imaging | 1–2 (baseline, week 52) | Full imaging visit |
| Plasma NDEV alpha-synuclein | Every dosing visit (routine blood draw) | None beyond existing PK/safety draw |
What Comes Next for Alpha-Synuclein Biomarker Strategy
The alpha-synuclein antibody field is not abandoning the target — it’s refining the trial designs and patient populations around it, and the exploratory subgroup signal from PASADENA’s fast-progressing cohort is a clear example of a program using better patient stratification to salvage a target hypothesis that a blunt, single-endpoint design nearly buried [6,7]. As more programs adopt enrichment strategies for both patient selection and pharmacodynamic monitoring, the biomarker plans that succeed will likely be the ones that stop asking CSF and imaging to do a job they were never designed for, and instead add a peripheral layer built specifically to answer the dose-response question in real time.
References
[1] Siderowf A, Concha-Marambio L, Lafontant DE, et al. Assessment of heterogeneity among participants in the Parkinson’s Progression Markers Initiative cohort using α-synuclein seed amplification: a cross-sectional study. Lancet Neurol. 2023;22(5):407-417. https://doi.org/10.1016/S1474-4422(23)00405-2
[2] Simuni T, Chahine LM, Poston K, et al. A biological definition of neuronal α-synuclein disease: towards an integrated staging system for research. Lancet Neurol. 2024;23(2):178-190. https://doi.org/10.1016/S1474-4422(23)00404-0
[3] Concha-Marambio L, Pritzkow S, Shahnawaz M, Farris CM, Soto C. Seed amplification assay for the detection of pathologic alpha-synuclein aggregates in cerebrospinal fluid. Nat Protoc. 2023;18(4):1179-1196. https://doi.org/10.1038/s41596-022-00787-3
[4] Kang UJ, Boehme AK, Fairfoul G, et al. Comparative study of cerebrospinal fluid α-synuclein seeding aggregation assays for diagnosis of Parkinson’s disease. Mov Disord. 2019;34(4):536-544. https://doi.org/10.1002/mds.27646
[5] Lang AE, Siderowf AD, Macklin EA, et al; SPARK Investigators. Trial of cinpanemab in early Parkinson’s disease. N Engl J Med. 2022;387(5):408-420. https://doi.org/10.1056/NEJMoa2203395
[6] Pagano G, Taylor KI, Anzures-Cabrera J, et al; PASADENA Investigators and Prasinezumab Study Group. Trial of prasinezumab in early-stage Parkinson’s disease. N Engl J Med. 2022;387(5):421-432. https://doi.org/10.1056/NEJMoa2202867
[7] Pagano G, Taylor KI, Anzures Cabrera J, et al. Prasinezumab slows motor progression in rapidly progressing early-stage Parkinson’s disease. Nat Med. 2024;30(4):1096-1103. https://doi.org/10.1038/s41591-024-02886-y
[8] Geerts H, van der Graaf PH, Courade JP, et al. Analysis of clinical failure of anti-tau and anti-synuclein antibodies in neurodegeneration using a quantitative systems pharmacology model. Sci Rep. 2023;13:14342. https://doi.org/10.1038/s41598-023-41382-0
[9] Eijsvogel P, Misra P, Concha-Marambio L, et al. Target engagement and immunogenicity of an active immunotherapeutic targeting pathological α-synuclein: a phase 1 placebo-controlled trial. Nat Med. 2024;30(9):2631-2640. https://doi.org/10.1038/s41591-024-03101-8
[10] Jiang C, Hopfner F, Berg D, Hu MT, Pilotto A, Borroni B, Davis JJ, Tofaris GK. Validation of α-synuclein in L1CAM-immunocaptured exosomes as a biomarker for the stratification of Parkinsonian syndromes. Mov Disord.2021;36(11):2663-2669. https://doi.org/10.1002/mds.28591
[11] Eitan E, Thornton-Wells T, Elgart K, Erden E, Gershun E, Levine A, Volpert O, Azadeh M, Smith DG, Kapogiannis D. Synaptic proteins in neuron-derived extracellular vesicles as biomarkers for Alzheimer’s disease: novel methodology and clinical proof of concept. Extracell Vesicles Circ Nucl Acids. 2023;4(1):133-150. https://doi.org/10.20517/evcna.2023.13
[12] Eitan E, Ayupova D, Volpert O, Gekas A. Neuron-derived extracellular vesicles as a blood-based biomarker for alpha-synuclein across synucleinopathies. Alzheimers Dement. 2026;22(Suppl 1) [conference abstract]. https://doi.org/10.1002/alz.70856
[13] Chahine LM, Beach TG, Seedorff N, et al; Systemic Synuclein Sampling Study. Feasibility and safety of multicenter tissue and biofluid sampling for α-synuclein in Parkinson’s disease: the Systemic Synuclein Sampling Study (S4). J Parkinsons Dis. 2018;8(4):517-527. https://doi.org/10.3233/JPD-181434
Leave a Reply