Landmark 170 · Trial in people
TCH346 (GAPDH binder)
Made no difference in 301 people with early Parkinson's. It had protected cells in animals.
In short
TCH346 (CGP 3466), a drug that binds the enzyme GAPDH to stop a cell-death pathway, did not differ from placebo in 301 people with early Parkinson's.
- What it is
- Trial in people
- Outcome
- Did not work
- When
- 2002–2006
- Size
- 301 people · Phase 2
- Approach
- Keep vulnerable cells alive
- Research lane
- Cell survival & anti-apoptosis
- Trial registry
- NCT00407212
- Last checked
- 19 Oct 2026
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
Novartis; lead author C. Warren Olanow with an international group at 45 clinics. The Lancet Neurology 2006;5:1013–1020. NCT00407212 (registry phase: 1/2; the trial ran January 2002 to November 2003 per the registry, and the paper came out in 2006).
What happened
TCH346 is related to the MAO-B drug deprenyl from landmark 61, but does not block MAO-B. It sticks to GAPDH and in animals prevented nerve-cell death. In rhesus monkeys given MPTP (landmark 2) it prevented worsening of symptoms and loss of striatal FDOPA uptake (Andringa et al. 2003).
- Design. 301 people with early untreated Parkinson's, randomised to 0.5 mg (n=78), 2.5 mg (n=79), 10 mg (n=73) or placebo (n=71) daily for 12–18 months, then a 4-week washout. The main measure was time until disability needed dopamine treatment.
- Result. 255 completed. Treatment was needed in 34%, 38%, 33% on the three doses and 32% on placebo. UPDRS and quality of life did not differ. Few stopped for adverse events.
- Interpretation. The authors suggested the animal models may not reflect Parkinson's, the doses may have been low, or the endpoints too insensitive.
Why it moved us toward a cure
Together with landmark 169, it closed the anti-apoptotic route as a stand-alone neuroprotection idea in the 2000s. It shows how little preclinical protection in MPTP models predicts human benefit.
Outcome
Dead end. The paper found no sign of a neuroprotective effect. No source read says what Novartis did next.
Sources
- Olanow et al., Lancet Neurol 2006;5:1013, doi:10.1016/S1474-4422(06)70602-0 (abstract via Europe PMC) · ClinicalTrials.gov NCT00407212
- Andringa et al., Neurobiol Dis 2003;14:205, PMID 14572443 (monkey MPTP study)
- Not yet verified: that the programme was dropped by Novartis; the link to deprenyl (stated in a reference-chemical listing, not a primary paper).
How it connects
Built on
- tests MPTP: a chemical that causes Parkinson's · 1983
- builds on DATATOP: deprenyl · 1987
Led to
None recorded.