Longitudinal dynamics of serum neurofilament light chain and prognostic value for functional outcome after acute stroke: a prospective cohort study.

Serum NfL rose significantly from within 24 h of onset (D1) to day 7 (D7) in patients with acute stroke.

• The abstract reports that sNfL increased significantly from D1 to D7 (p < 0.001).
• The cohort comprised 104 acute stroke patients: ICH (n = 28), ischemic stroke with intravenous thrombolysis (IS+IVT, n = 33), and ischemic stroke without thrombolysis (IS-IVT, n = 43).
• sNfL was measured at D1 (within 24 h of onset) and at D7, and analyzed with linear mixed-effects models.

Patients with stroke had higher sNfL than healthy controls at D1.

• The comparison was between 104 patients with acute stroke and 86 healthy controls (HCs).
• The difference at D1 was reported as p < 0.001.
• Enrollment was consecutive between January 2023 and December 2024.

Longitudinal sNfL trajectories did not differ significantly among the ICH, IS+IVT, and IS-IVT groups.

• In the linear mixed-effects model adjusted for age and admission NIHSS score, the time × group interaction was p = 0.225.
• The authors note that the predominantly mild-to-moderate severity profile of the cohort may have limited detection of trajectory differences.

Higher D1 sNfL was independently associated with higher odds of poor functional outcome at 3 months (primary endpoint).

• Each 10-pg/mL increase in D1 sNfL was associated with 5% higher odds of poor outcome (OR = 1.05; 95% CI, 1.00-1.12; p = 0.047).
• Poor functional outcome was assessed by the modified Rankin Scale at 3 months.
• Estimates came from multivariable logistic regression; the lower confidence limit is at 1.00, so the association is borderline.

Higher D1 sNfL was independently associated with higher odds of poor functional outcome at 6 months (secondary endpoint).

• Each 10-pg/mL increase in D1 sNfL was associated with 7% higher odds (OR = 1.07; 95% CI, 1.01-1.14; p = 0.025).
• Outcome was assessed by mRS at 6 months using multivariable logistic regression.

D7 sNfL was not independently associated with poor functional outcome after adjustment.

• This was reported for the multivariable logistic regression models.
• In contrast, D1 sNfL showed significant associations at both 3 and 6 months.

Exploratory restricted cubic spline analyses found no evidence of nonlinearity for D1 sNfL but suggested a nonlinear association between D7 sNfL and 3-month poor outcome.

• RCS and interaction analyses were prespecified as exploratory.
• For D1 sNfL, there was no statistical evidence of nonlinearity.
• For D7 sNfL, the relationship with 3-month poor functional outcome appeared nonlinear.