The coordinator gets the referral on a Tuesday: a potential participant flagged through polysomnogram-confirmed REM sleep behavior disorder, no motor symptoms, no Parkinson’s diagnosis. The eligibility checklist says CSF alpha-synuclein seed amplification assay required. The site has never run a presymptomatic neurodegenerative cohort. The question on the coordinator’s desk: how do you screen for a disease someone does not yet have?
That is the operational problem the Parkinson’s Precision Medicine Initiative (PPMI) has been working through at scale, and its data now offers something practical for clinical operations teams managing early-detection neurology trials. PPMI enrolled participants using two parallel strategies: polysomnogram-confirmed REM sleep behavior disorder (RBD) and centralized identification of hyposmia via smell testing. The biomarker yield difference between those strategies is not marginal. Among participants with both RBD and hyposmia, 94% were positive for CSF alpha-synuclein seed amplification assay (CSFaSynSAA). Among participants with hyposmia combined with self-reported dream enactment behavior, the positivity rate was 85%. Among participants with RBD alone without hyposmia, it dropped to 40%.
Those numbers are the enrollment calculus your startup team needs to work backward from before a single screening visit is scheduled.
The Screen Failure Math Nobody Budgets For
Neurology trials already carry a screen failure problem. Screen failure rates in neurological trials average 57%, compared to 36.3% across all therapeutic areas. A presymptomatic cohort compounds this further: you are recruiting people who feel fine, identifying them through sleep studies and smell tests, and then asking them to return for lumbar puncture and dopamine transporter imaging. Every step in that chain is a potential dropout, and most site budgets are built as if the referral funnel and the enrollment funnel are the same thing.
PPMI’s data separates those funnels explicitly. The RBD-without-hyposmia group carried a 40% CSFaSynSAA positivity rate. That means for every ten participants you screen through that pathway, six will not meet biomarker criteria. If your protocol requires CSFaSynSAA positivity for randomization, those six participants generate screening costs, coordinator hours, CSF collection visits, and lab processing fees with no path to enrollment. The sponsors I see underprice this most consistently are the ones who treat RBD as a monolithic enrichment criterion rather than stratifying by co-occurring hyposmia at the referral stage.
The operational fix is earlier stratification, not more screening. PPMI’s approach runs smell testing centrally as a pre-screening filter. For multi-site sponsors, that means building a centralized pre-qualification step, smell testing administered before any site visit is triggered, so sites receive referrals already stratified by the biomarker risk profile. PREDICT-PD uses a comparable logic: a two-stage approach beginning with remote screening questionnaires before escalating to in-person phenotyping. Sites absorb cost at in-person phenotyping; every step that happens upstream of the site visit is a screen failure cost the site does not carry.
This changes what site selection criteria should look like. A site that can perform polysomnography, that has sleep medicine referral relationships, and that has staff trained on standardized olfactory testing is a fundamentally different operational asset than a site with neurology experience but no sleep lab access. Site feasibility questionnaires for early Parkinson’s trials may not routinely ask about polysomnography capability. They should.
What the Biomarker Letter of Support Actually Requires Operationally
The CSFaSynSAA is not an approved diagnostic. The FDA issued a Letter of Support for its use as a biomarker in clinical trials, which means every site running it is operating under a protocol-specific assay framework, not an off-the-shelf lab order. That distinction matters the moment a coordinator opens the lab manual.
A Letter of Support positions the assay as fit-for-purpose in a research context, but it does not standardize collection handling, shipping windows, or freeze-thaw requirements across sites the way a cleared diagnostic does. In practice, sites running CSFaSynSAA in multi-site studies need central lab vendor alignment baked into the protocol before SIV. The questions that surface during site initiation visits, about collection tube type, processing time post-collection, temperature stability for transit, and the chain of custody documentation, are not administrative details. They are the variables that determine whether a sample is evaluable or wasted. A wasted CSF sample from a participant who underwent lumbar puncture is not recoverable. The participant is unlikely to consent to a repeat collection, and that becomes a screen failure with an elevated ethical cost.
Across sites managing early neurodegenerative cohorts, the pattern I see most often is central lab technical documents arriving after site initiation activities have already concluded. That sequencing inverts the operational logic. CSFaSynSAA sample handling requirements need to be in the hands of the site pharmacist and laboratory coordinator during the SIV, not during the first monitoring visit when a sample handling deviation has already been entered.
Building the Cohort Before You Need It
PPMI had 4,649 participants enrolled as of May 2026, a number that reflects years of referral network development through sleep clinics, movement disorder practices, and community smell-testing campaigns. That infrastructure did not exist when PPMI began enrolling. It was built participant by participant through relationships sites cultivated before enrollment pressure hit.
The conversion timeline is the operational reality sponsors need to plan around. Studies of patients with idiopathic REM sleep behavior disorder show that phenoconversion to Parkinson’s disease or related dementia occurs over years, with risk accumulating progressively the longer the RBD diagnosis persists. Participants who are CSFaSynSAA-positive and presymptomatic today represent a finite, identifiable window. That window closes when motor symptoms emerge and the participant no longer meets the trial’s presymptomatic inclusion criteria. Sites that have not developed sleep clinic referral pipelines before a trial opens will spend months trying to build them under enrollment pressure, and that is the version of site activation that produces twelve-month delays to first-patient-in.
The FDA’s guidance on enrichment strategies for clinical investigations defines enrichment as the prospective use of any patient characteristic to select a study population in which the detection of a drug effect, if present, is more likely than in an unselected population. PPMI’s data operationalizes that definition at the biomarker level: CSFaSynSAA positivity combined with RBD and hyposmia gives sponsors a cohort where 94% of participants carry the biological substrate the trial is designed to study. The enrichment yield from stratified co-occurring risk factors is meaningfully higher than from any single criterion alone. Protocols that build dual-criterion eligibility, requiring both a sleep-based and an olfactory-based qualifier, will need to account for the referral source relationships and the pre-screening infrastructure to make those criteria executable at the site level.
For site directors activating on an early Parkinson’s protocol in the next twelve months: the pre-screening workflow review belongs in the site feasibility assessment, not the enrollment plan. If the smell testing procedure, the polysomnography referral pathway, and the CSFaSynSAA sample handling SOP are not resolved before the regulatory green light arrives, you will be resolving them during the enrollment period. And participants your referral network would have captured early in the enrollment period may have moved past the presymptomatic window by the time infrastructure is in place.
References
- medRxiv, “Strategies to establish a biologically defined Parkinson’s disease cohort prior to the onset of clinically significant motor symptoms”
- Boston University Parkinson’s Disease & Movement Disorders Center, “PPMI Study Annual Meeting 2026: 4,649 participants enrolled”
- NeurologyLive, “FDA Issues Letter of Support for Alpha-Synuclein Seed Amplification Assay as Biomarker in Clinical Trials”
- NeurologyLive, “REM Sleep Behavior Disorder: Predicting Conversion to Synucleinopathy”
- ResearchGate, “PREDICT-PD: A Two-Stage Approach to Early Identification of Parkinson’s Disease”
- Alliance Clinical Network, “Tackling High Screen Failure Rates and Boosting Diversity in CNS Clinical Trials”
- FDA, “Guidance for Industry: Enrichment Strategies for Clinical Trials to Support Approval of Human Drugs and Biological Products”
