Dashboard

Disease indication

Parkinson's Disease

Why the animal models predict symptom relief — and almost nothing else

Therapeutic goal

Disease-modifying: Slow or halt the nigrostriatal degeneration itself. Every model put to this test has so far failed to predict the clinic: strong preclinical neuroprotection, then a null result in patients — ten times over.

Approved disease-modifying therapies
0

No therapy has ever been shown to slow the progression of Parkinson's disease.

high confidence
Reached Phase 2+ and succeeded
0 of 10

Every disease-modifying programme below with a completed readout. Unlike the symptomatic count, this one has a real denominator — the entries were not selected on their outcome. One further programme is still running.

high confidence
Typical pivotal duration
2–5 years

A delayed-start or withdrawal design is required to separate symptom masking from true modification — the single largest cost driver.

high confidence
Typical pivotal N
~600–1,500

Larger and longer than symptomatic trials because the effect being measured is a change in slope.

moderate confidence

What a trial has to prove

The quietest reason this disease is hard. Toggle between the two goals and watch the standing column: symptom relief has measures everyone accepts, and disease modification has none.

No endpoint is accepted as establishing disease-modifying benefit in Parkinson's Disease. Every measure below is either contested or exploratory. A trial can therefore succeed on its own terms and still not support the claim.

Endpoints used in Parkinson's Disease trials for the Disease-modifying goal, how each is measured, how long it takes to read, and whether it is accepted as evidence for the claim. Select one for detail.
EndpointTypeRead atStanding
MDS-UPDRS progression slopeComposite2-5 yearsNone accepted
Delayed-start designTrial design2-5 yearsContested
DaTscan dopamine transporter imagingImaging2-4 yearsContested
Alpha-synuclein seed amplification assayLaboratoryBaselineExploratory
Accepted
Regulators accept this as evidence for the claim.
Surrogate
A stand-in for clinical benefit, accepted or under negotiation for accelerated approval — the benefit itself must still be confirmed.
Contested
Used in trials, but not accepted as establishing the claim on its own.
Exploratory
Informative for development — target engagement, enrichment — but never the basis of an approval.
None accepted
No endpoint has ever been accepted as establishing this claim in this disease.

Preclinical model validity

Three standard axes. Toggle the therapeutic goal above and watch only the third column change — that asymmetry is the whole story.

Validity of Parkinson's Disease preclinical models across face, construct, and predictive validity. Predictive validity is shown for the Slow progression goal.
ModelFaceDoes it look like the disease?ConstructDoes it arise the same way?PredictiveSlow progression
ValidityNoneLowModerateHighNo evidencenever tested — not a low score

Translation ledger

Every programme below cleared its animal model. None has yet changed the course of the disease in a patient.

Interventions that showed benefit in a Parkinson's Disease preclinical model, and what happened when they reached the clinic.
InterventionRested onPrimary endpointReachedOutcome
Exenatide
UCL / NIHR · GLP-1 receptor agonist
6-OHDAMPTP mouse
MDS-UPDRS Part III off medication at 96 weeks
Phase 3
Exenatide-PD3 · 2025
Failed
BIIB122 / DNL151
Denali / Biogen · LRRK2 kinase inhibition
LRRK2 G2019S
Change in MDS-UPDRS Parts II and III (ongoing)
Phase 2b
LUMA · 2025
Ongoing
Prasinezumab (PRX002)
Roche / Prothena · Anti-alpha-synuclein monoclonal antibody
α-syn PFFAAV α-syn
PASADENA: MDS-UPDRS total change at 52 weeks. PADOVA: time to confirmed motor progression
Phase 2b
PASADENA / PADOVA · 2024
Missed primary
Cinpanemab (BIIB054)
Biogen · Anti-alpha-synuclein monoclonal antibody
α-syn PFF
Change in MDS-UPDRS total score
Phase 2
SPARK · 2021
Failed
Inosine
NINDS / Parkinson Study Group · Urate elevation / antioxidant
6-OHDAMPTP mouse
Rate of change in MDS-UPDRS total over 2 years
Phase 3
SURE-PD3 · 2021
Failed
Nilotinib
Georgetown University · c-Abl inhibition promoting alpha-synuclein clearance
AAV α-synα-syn PFF
Safety and tolerability, with MDS-UPDRS and CSF measures as secondary
Phase 2
NILO-PD · 2021
Failed
Isradipine
NINDS / Parkinson Study Group · Cav1.3 calcium-channel blocker
MPTP mouse6-OHDA
Change in MDS-UPDRS total (on medication) over 36 months
Phase 3
STEADY-PD III · 2020
Failed
Creatine
NINDS / NET-PD · Mitochondrial bioenergetic buffering
MPTP mouse
Global statistical test across five clinical outcomes over 5 years
Phase 3
NET-PD LS-1 · 2015
Failed
Coenzyme Q10
NINDS · Mitochondrial complex-I support / antioxidant
MPTP mouse
Change in MDS-UPDRS total to end of study
Phase 3
QE3 · 2014
Failed
CERE-120 (AAV2-neurturin)
Ceregene · Neurotrophic factor gene therapy
6-OHDAMPTP primate
UPDRS motor score off medication at 12 months
Phase 2b
CERE-120 Phase 2b · 2013
Failed
GDNF (intraputaminal)
Amgen · GDNF/RET neurotrophic signalling
6-OHDAMPTP primate
UPDRS motor score at 6 months
Phase 2
Intraputaminal GDNF · 2006
Failed
Endpoint · Composite

MDS-UPDRS progression slope

Rate of change in total score across motor and non-motor domains

None acceptedRead at 2-5 years
Scale
Rate of change in an ordinal sum spanning 0-260 across all four parts.
How it is administered
The full MDS-UPDRS administered repeatedly over years. The analysis is fitted to the trajectory of scores rather than to any single visit, which is what makes it a progression measure rather than a snapshot.
What it contains
  • Part I - non-motor experiences of daily living, 13 items
  • Part II - motor experiences of daily living, self-reported, 13 items
  • Part III - the motor examination, 33 items
  • Part IV - motor complications, 6 items
How a score is produced
Either total change from baseline, or the fitted slope of change per year, compared between arms over two to five years.
What counts as success
None accepted, and this is the crux. Any drug with even a modest symptomatic effect shifts this score downward without altering the rate of neuronal loss underneath, so a between-group difference is compatible with both disease modification and simple symptom relief. No analysis has separated the two to regulators' satisfaction.
Why it matters here

The central problem on this page. Any drug with a symptomatic effect improves this score without touching progression, so a difference cannot be read as disease modification. No trial using it has been accepted as establishing the claim.

STEADY-PD III, SURE-PD3, NET-PD LS-1high confidence

What the clinic costs

No per-trial cost figure specific to Parkinson's exists. Rather than quote one that doesn't, this computes cost from cited trial designs and a single labelled assumption — and shows the arithmetic.

The endpoint is a change in the slope of progression, which cannot be seen quickly and cannot be distinguished from symptom masking without a delayed-start or withdrawal design. Both facts inflate exposure enormously.

13×
the exposure
6.6×
the cost

6,685 patient-years against 515 for a symptomatic programme. Cost grows more slowly than exposure because the fixed enrolment cost of each participant — screening, baseline imaging — is paid once however long they are then followed. The goal, not the molecule, still sets the budget.

Citedenrolment, duration and study counts, from named trials.Assumptionthe two cost parameters below.Computedeverything else.

Clinical programme by phase: enrolment, duration, exposure in patient-years, and modelled cost.
PhaseDesignN × studiesDurationPatient-yearsModelled cost
Phase 1
SAD / MAD, healthy volunteers
Reads: Safety, tolerability and pharmacokinetics
Identical in kind to the symptomatic programme. Nothing has diverged yet.
Conventional first-in-human designmoderate confidence
60
9 mo
45
$1.6M–$2.7M
Phase 2
Parallel-group with delayed-start extension
Reads: MDS-UPDRS change, with imaging and fluid biomarkers
Sized and timed after PASADENA (n=316) and SPARK (n=357), both of which ran a one-year blinded period followed by an extension.
PASADENA (Pagano et al., NEJM 2022); SPARK (Lang et al., NEJM 2022)high confidence
320
2 yr
640
$17M–$28M
Phase 3
Delayed-start or long-term parallel-group
Reads: MDS-UPDRS progression slope under a delayed-start design
Between STEADY-PD III (n=336 over 36 months) and NET-PD LS-1 (n=1,741 over 5 years). A registrational modification claim needs both the scale and the years — this is the single largest driver of the gap.
STEADY-PD III (Ann Intern Med 2020); NET-PD LS-1 (JAMA 2015)moderate confidence
1,000 × 2
3 yr
6,000
$144M–$246M
Total5.8 yr of clinical development, treating phases as sequential6,685$162M–$277M
AssumptionThe two cost parameters
Fixed, per participant
$12,000–$18,000
Running, per year on study
$20,000–$35,000

Model assumptions, not citations — the only free parameters here. The fixed component covers screening, baseline imaging and randomisation, incurred once per participant; the running component covers visits, monitoring and site fees for each year on study. Splitting them matters: a single rate per patient-year would price short symptomatic trials far below what pivotal trials actually cost.

Calibration: priced this way a single pivotal trial in this programme comes to $9.0M–$14M, against published pivotal trial costs of $1233M (median $19M) across all indications. A 20-week motor-scale study belongs toward the bottom of that spread, against multi-year outcome trials — which is where it lands. Moore et al., JAMA Internal Medicine 2018.

ComputedRisk-adjusted cost
$2.7B$4.6B

Programme cost divided by a 6% likelihood of approval from Phase 1 in neurology — what one success costs once the failures are paid for. BIO / Informa / QLS, Clinical Development Success Rates — neurology is among the lowest of any therapeutic area.

Read this one carefully: That rate is area-level and does not separate symptomatic from disease-modifying — the very distinction this page is about. Applied unchanged to both, it flatters disease modification, where the observed record is 0 of 10.

moderate confidence
For scale — industry-wide, per approved drug
$985M
Median R&D cost per approval
Wouters et al., JAMA 2020 (capitalised; estimates range $314M–$2.8B)
$2.6B
Capitalised cost per approval
DiMasi et al., J Health Econ 2016

These are portfolio figures spanning all of R&D and already carry the cost of failure, so they are not comparable with the per-programme clinical costs above — they are the order of magnitude those costs roll up into.

What is changing

The predictive-validity problem is being attacked at its root — by defining the disease biologically rather than clinically.

Parkinson's can now be defined biologically

Seed amplification assays detect misfolded alpha-synuclein in cerebrospinal fluid with roughly 88% sensitivity in sporadic PD. For the first time, trials can enrol biologically confirmed patients rather than clinically defined ones — plausibly the reason earlier disease-modifying trials were doomed before they began.

Siderowf et al., Lancet Neurology 2023 (PPMI)

Staging by biology rather than symptoms

The Neuronal Alpha-Synuclein Disease Integrated Staging System proposes staging from biomarkers instead of motor signs, opening a route to intervening years before the motor threshold — by which point most nigral neurons are already lost.

Simuni et al., Lancet Neurology 2024

Targets anchored in human genetics, not toxins

LRRK2 and GBA1 programmes derive from human genetics rather than from a toxin model, sidestepping the predictive-validity problem at its root. The trade-off is that the corresponding mice barely degenerate, so efficacy rests on biomarkers until a pivotal trial reads out.

LUMA (BIIB122) programme disclosures

Educational. Validity ratings are a coded reading of the published literature, not a consensus standard. Ledger outcomes and dates are drawn from trial reports and sponsor announcements; every figure carries an explicit confidence flag. Not medical, regulatory, or investment advice.