ELLDOPA: levodopa and progression

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

The Parkinson Study Group (Stanley Fahn, lead author). Published in the New England Journal of Medicine, December 2004.

What happened

Levodopa, the drug that replaces lost dopamine (landmark 1), had been used for decades. Some doctors feared it might speed up nerve-cell loss. ELLDOPA tested this directly.

• Design. Double-blind, placebo-controlled. 361 people with early Parkinson's took carbidopa-levodopa at 150, 300 or 600 mg of levodopa a day, or placebo, for 40 weeks. Then all treatment stopped for 2 weeks (a washout) so that the drug's short-term symptom relief could wear off.
• Clinical result. The main measure was the change in total UPDRS (a standard severity score, where higher is worse) from baseline to week 42. The mean rise was 7.8 points on placebo, 1.9 on 150 mg, 1.9 on 300 mg, and fell 1.4 on 600 mg (P<0.001). The higher doses also caused more dyskinesia (involuntary movements), nausea and headache.
• Imaging result. 142 people had brain scans at baseline and week 40 (β-CIT SPECT, which measures dopamine transporters on nerve endings). In the substudy of 116, uptake fell 1.4% on placebo versus 6% (150 mg), 4% (300 mg) and 7.2% (600 mg) on levodopa (P=0.036). Nineteen people with a normal baseline scan were excluded.
• The authors' conclusion. The clinical data fit either slowed disease or a long-lasting symptom effect. The scans fit either faster loss of nerve endings or a direct effect of the drug on the transporter itself. They called the long-term effect uncertain.

Why it moved us toward a cure

It showed that a two-week washout is too short to separate symptom relief from protection, and that a scan marker and a clinical score can point in opposite directions. That pushed the field toward the delayed-start design used in landmark 162 and later in PROUD and LEAP. It also followed DATATOP, where the same symptom-versus-protection question had been left open.

It asked whether levodopa, the standard treatment since the 1960s (landmark 285), speeds or slows the disease.

Outcome

Superseded. The delayed-start trial LEAP (landmark 166), 445 people on levodopa for 80 weeks, found no difference in UPDRS change from baseline at week 80 between early and delayed levodopa (1.0 point, 95% CI −1.5 to 3.5; P=0.44) and concluded that levodopa has no disease-modifying effect.

Sources

• Fahn et al., NEJM 2004;351:2498–2508, doi:10.1056/NEJMoa033447 (abstract read via Europe PMC, PMID 15590952)
• Verschuur et al. (LEAP), NEJM 2019;380:315–324, doi:10.1056/NEJMoa1809983
• ClinicalTrials.gov NCT00134784: this is the long-term imaging follow-up (142 people, Phase 2, Institute for Neurodegenerative Disorders, completed), not the main trial. A registry ID for the main trial was not found.
• Not yet verified: the number of sites and the sponsor of the main trial; why 142 scanned minus 19 excluded gives 116 (the abstract states both figures); the full-text washout results.