Circulating levels of neurofilament light chain as a biomarker of infarct and white matter hyperintensity volumes after ischemic stroke.
Lukas HolmegaardChrister JensenAnnie PedersenChristian BlomstrandKaj BlennowHenrik ZetterbergKatarina JoodChristina JernPublished in: Scientific reports (2024)
Serum neurofilament light chain protein (sNfL) shows promise as a biomarker for infarct size in acute ischemic stroke and for monitoring cerebral small vessel disease (cSVD). However, distinguishing the cSVD contribution after stroke may not be possible due to post-stroke sNfL increase. Additionally, it remains unclear if etiologic subtype differences exist. We measured infarct and white matter hyperintensity (WMH) volumes using MRI at the index stroke in ischemic stroke patients (n = 316, mean age 53 years, 65% males) and at 7-year follow-up (n = 187). Serum NfL concentration was measured in the acute phase (n = 235), at 3-months (n = 288), and 7-years (n = 190) post stroke. In multivariable regression, acute and 3-month sNfL concentrations were associated with infarct volume and time since stroke, but not with stroke etiology or infarct location. Seven years post-stroke, sNfL was associated with WMHs and age, but not with stroke etiology. Nonlinear regression estimated that sNfL peaks around 1 month, and declines by 50% at 3 months, and 99% at 9 months. We conclude that sNfL can indicate infarct volume and time since brain injury in the acute and subacute phases after stroke. Due to the significant post-stroke sNfL increase, several months are needed for reliable assessment of cSVD activity.
Keyphrases
- atrial fibrillation
- brain injury
- cerebral ischemia
- acute myocardial infarction
- white matter
- subarachnoid hemorrhage
- liver failure
- respiratory failure
- multiple sclerosis
- magnetic resonance imaging
- percutaneous coronary intervention
- cerebrospinal fluid
- heart failure
- computed tomography
- magnetic resonance
- hepatitis b virus
- left ventricular
- acute coronary syndrome
- artificial intelligence
- contrast enhanced
- deep learning