GBA mutations: commonest genetic risk factor

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

Ellen Sidransky (National Human Genome Research Institute, NIH, first author), Mike Nalls, Jan Aasly and colleagues at 16 centres in the Americas, Europe, Israel and Asia. Europe PMC lists NIH funding: intramural programmes and NINDS grants.

What happened

Each centre genotyped a common DNA panel. The study compared 5,691 Parkinson's patients (780 Ashkenazi Jewish) with 4,898 controls (387 Ashkenazi Jewish). Among Ashkenazi Jews, the L444P or N370S variants were found in 15% of patients and 3% of controls; in non-Ashkenazi people, 3% and under 1%. Full sequencing of 1,883 non-Ashkenazi patients found GBA mutations in 7%. The pooled odds ratio was 5.43. Carriers tended to have earlier onset and more atypical features. NEJM 361:1651–1661, 1 Oct 2009.

Why it moved us toward a cure

It tied Parkinson's to lysosomal function (glucocerebrosidase, the enzyme deficient in Gaucher disease) and defined a genetic subgroup for targeted trials: GCase-enhancing drugs, ambroxol, venglustat and others.

Outcome

Advanced. The trials so far are mostly negative (venglustat 2023, BIA 28-6156 in 2026; see dead ends). The ambroxol Phase 3 (ASPro-PD, 330 people, 1,260 mg a day for 104 weeks, £5.5 million, run by Cure Parkinson's with the Van Andel Institute) is recruiting with no results. The association itself has held up: the 2014 meta-analysis of 13,708 cases found a GBA signal with a second independent variant (landmark 237). No correction, erratum or retraction is listed for the paper in Europe PMC.

Sources

• Sidransky et al., NEJM 2009, DOI 10.1056/NEJMoa0901281
• Abstract numbers via an NTNU repository record
• ASPro-PD, UCL Comprehensive Clinical Trials Unit (status and design) · Europe PMC record
• Discrepancy: a news piece (MDedge) gave 980 Ashkenazi patients; the abstract says 780, which is used here.
• Not yet verified: the confidence interval for the odds ratio (not in the abstract; full text not read). A search summary of the PMC full text gave 1,642 sequenced patients and 6.9%, against 1,883 and 7% in the abstract; the abstract figures are used.