GDNF infusion trials (Amgen, Bristol)

The research record

This is the project's file for this landmark, as written and checked. Sources are listed at the end; anything not yet confirmed is marked "Not yet verified".

Who

Anthony Lang, Steven Gill and a multicentre team (Annals of Neurology 2006; some authors were from Amgen); Alan Whone, Steven Gill and colleagues at Bristol (Brain 2019).

What happened

Open-label results in five people (landmark 39) sparked two controlled trials.

• Amgen Phase 2 (2006). 34 people randomised 1:1 to liatermin (GDNF) 15 micrograms per putamen per day, infused continuously through catheters, or placebo. The primary measure was the change in off-medication motor score at 6 months: −10.0% versus −4.5%, difference not significant (95% CI −23.0 to 12.0, p=0.53). PET dopamine uptake (18F-dopa) was 32.5% higher on GDNF (p=0.019). Problems: serious device-related events (catheters repositioned in two people, devices removed in one) and neutralising antibodies to the drug in three.
• Bristol trial (2019). A 6-person pilot, then 35 people randomised 1:1 to GDNF 120 micrograms per putamen or placebo, every 4 weeks for 40 weeks, delivered by a skull-mounted port with convection-enhanced delivery (pressure pushes the liquid through the tissue). Main measure: percentage change in off-medication motor score at 40 weeks. It fell 17.3% on GDNF and 11.8% on placebo (difference −4.9%, 95% CI −16.9 to 7.1, p=0.41); secondary measures also did not differ. PET dopamine uptake rose in the putamen only on GDNF (25% to 100% by region). A post hoc analysis found nine of 21 (43%) GDNF patients and no placebo patients with a large (10 points or more) improvement. No drug-related serious adverse events were reported. An open-label extension gave everyone GDNF for another 40 weeks: no group difference at week 80, and the authors said future trials probably need 80 weeks and/or a higher dose.
• Intraventricular GDNF (Nutt et al., Neurology 2003). The earliest controlled test, with GDNF given into the brain's fluid spaces (ventricles) through an implanted catheter instead of into the putamen. 50 people with advanced Parkinson's in a multicentre, double-blind, placebo-controlled study: monthly doses of 25 to 4,000 micrograms of GDNF (12 on placebo, 7 or 8 per dose group) for 8 months, then an open-label extension in 16 people for up to 20 more months. Motor and total UPDRS scores did not improve at any dose. Side effects were nausea, vomiting, weight loss, electric-shock sensations and low blood sodium (over half of those on 75 micrograms or more). The authors suggested GDNF did not reach the putamen or the substantia nigra.

Why it moved us toward a cure

It is the key cautionary pair for growth factors. Better PET scans did not mean better patients, so scan changes cannot stand in for clinical benefit. Authors suggested the dose may have been too low, treatment too short, or patients too far advanced. These questions shaped later GDNF gene therapy (AB-1005, landmark 39) and a related neurturin trial (landmark 147).

Outcome

Dead end for GDNF protein infusions. The growth-factor idea is still tested by gene therapy, but with the same warning.

Sources

• Lang et al., Ann Neurol 2006;59:459–466, DOI 10.1002/ana.20737, PMID 16429411 (abstract read via Europe PMC)
• Whone et al., Brain 2019;142:512–525, PMID 30808022 (abstract read via Europe PMC; citation also via MDedge summary) · Cure Parkinson's · University of Bristol, Feb 2019
• Nutt et al., "Randomized, double-blind trial of glial cell line-derived neurotrophic factor (GDNF) in PD," Neurology 2003;60(1):69-73, DOI 10.1212/WNL.60.1.69, PMID 12525720 (OHSU record; abstract read via Europe PMC)
• Whone et al., open-label extension, PMID 30829619 (abstract via Europe PMC).
• Not yet verified: who sponsored the 2006 trial and why Amgen stopped; the Bristol DOI; the full papers (abstracts only).