GM1 ganglioside

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

Jay Schneider and colleagues (Stephen Gollomp, Stephanie Sendek, Amy Colcher, Franca Cambi, Wei Du) at Thomas Jefferson University, Philadelphia. The registry lists Thomas Jefferson University as sponsor with the US National Institute of Neurological Disorders and Stroke as collaborator, under NIH grant R01NS038681.

What happened

GM1 is a ganglioside, a lipid in nerve-cell membranes. The authors suggest it may protect dopamine neurons and partly restore their function; the mechanism is not established. The trial (NCT00037830, Phase 2, enrolled from November 1999; registry completion June 2010) was single-centre, double-blind, and used a "delayed-start" design, meant to separate symptom effects from slowing. 77 people already on Parkinson's treatment were randomised: one group took GM1 for 120 weeks (early start), the other took placebo for 24 weeks and then GM1 for 96 weeks (delayed start). A third, non-randomised group of 17 people on standard care was followed for comparison, and washout evaluations took place one and two years after treatment ended. The primary outcome was the change in UPDRS motor score (lower is better).

At week 24, the early-start group improved (least-squares mean −3.57, standard error 0.50) while the delayed-start group worsened (+1.85, SE 0.50; P < 0.0001, registry results, 37 and 38 people analysed). At week 120 the early-start group was at −1.80 (SE 1.12) against +1.53 (SE 1.09) for delayed start (P = 0.037). Both groups worsened significantly during washout. The authors concluded that 24 weeks of GM1 beat placebo for motor symptoms and that longer use gave a lower-than-expected rate of progression. Published as Schneider et al., J Neurol Sci 2013;324(1–2):140–148 (online November 2012).

A 2015 pilot imaging substudy (40 people from the trial; [11C]methylphenidate PET for the dopamine transporter) reported slower loss of binding in several striatal regions in GM1-treated people, and said the result needed confirming in more people (J Neurol Sci 2015;356:118–123).

Why it moved us toward a cure

The delayed-start design is one of the few ways to test slowing without a long washout, and the persistent early-start gap fitted a protective effect. The authors called it a small proof-of-concept study. Schneider has since argued the case in a 2023 review (Int J Mol Sci 24:9183, DOI 10.3390/ijms24119183) and reported neuroprotection by GM1 in a rat synuclein model (Sci Rep 2019).

Outcome

Open, unreplicated. The comparison group was not randomised and was not compared statistically with the treated groups, the imaging substudy was small (40 people), and worsening after withdrawal suggests that at least part of the benefit was symptomatic. A registry search on 19 Oct 2026 found no Phase 3 of GM1 for slowing. A separate Phase 2 by Qilu Pharmaceutical (NCT07585565, 276 planned, not yet recruiting) tests a GM1 injection for "on" time in people with motor fluctuations, which is a symptom trial.

Sources

• Schneider et al., J Neurol Sci 2013;324:140–148, DOI 10.1016/j.jns.2012.10.024, PMID 23199590 (abstract via Europe PMC)
• ClinicalTrials.gov NCT00037830 (API, 19 Oct 2026, results posted)
• Schneider et al., imaging pilot, J Neurol Sci 2015, DOI 10.1016/j.jns.2015.06.028, PMID 26099170
• Schneider, GM1 review, Int J Mol Sci 2023 · Schneider et al., GM1 in a rat synuclein model, Sci Rep 2019 · NCT07585565
• Not yet verified: the paper's confidence intervals and P values beyond the registry results (full text not read); the 2023 review's own account of later trials.