Neurofilament light

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

Brit Mollenhauer (University Medical Center Göttingen; first author) and Danielle Graham (Biogen; last author), with Claudia Trenkwalder, Henrik Zetterberg, Tanya Simuni, Ken Marek and others. They used the DeNoPa cohort and the Parkinson's Progression Markers Initiative (PPMI; landmark 24). PPMI is sponsored by the Michael J. Fox Foundation and co-funded by a list of drug companies, including Biogen; the paper's acknowledgements say the study was funded by the Foundation and its partners.

What happened

Neurofilament light chain (NfL) is a structural protein of nerve fibres that leaks into spinal fluid and blood when neurons are injured. The team first checked it in spinal fluid in DeNoPa (176 people, including drug-naive newly diagnosed patients), then in blood serum from PPMI (1,190 people with scheduled blood draws up to 60 months, including patients, healthy controls, people with other neurodegenerative diseases, prodromal people and gene carriers). Baseline serum levels were slightly higher in Parkinson's than in controls (13 ± 7.2 versus 12 ± 6.7 pg/mL, P = 0.0336) and rose over time (P < 0.0001 for the Parkinson's-by-time interaction). Each doubling of NfL went with about 3.45 more points on the MDS-UPDRS total (landmark 114). Spinal-fluid and serum values were strongly correlated in a separate cross-sectional "bridging" sample (figure in the paper; I did not confirm the coefficient). Age and sex matter: levels rose about 3.35% per year of age and were 6.79% higher in women. Movement Disorders 35(11):1999-2008.

Why it moved us toward a cure

Slow-moving disease needs a fast, objective gauge of whether a drug protects neurons. A blood draw is cheaper and less invasive than a spinal tap or scan, and can be repeated. The authors call it "the first blood-based biomarker candidate" able to track clinical progression in Parkinson's, a claim I could not test against earlier papers.

Outcome

Open. The authors stress it is not specific to Parkinson's (diagnostic accuracy for Parkinson's versus controls was low, area under the curve about 0.6, as the paper itself says) and that its use to judge response to a neuroprotective treatment is still untested. It also does not identify the biology, which the seed assay does (landmark 9). Prodromal people and unaffected gene carriers showed no significant rise during follow-up. I found no later qualification of NfL by a regulator and no Parkinson's drug trial using blood NfL as an endpoint; a 2023 review (Buhmann et al.) still called its role in Parkinson's unclear. The paper has no correction or retraction (Crossref, Europe PMC).

Sources

• Mollenhauer et al., Mov Disord 2020;35:1999-2008, DOI 10.1002/mds.28206, PMID 32798333 · open full text, PMC8017468 (online 15 Aug 2020; numbers above checked against it)
• Not yet verified: earlier NfL-in-Parkinson's papers (2017-19), so the "first" claim is the authors' own; the bridging-cohort size and correlation coefficient; whether any trial has used blood NfL as a secondary endpoint (a search found none).