Early EEG and serum biomarkers for prognostication after cardiac arrest

Abstract

Rationale: Early electroencephalography (EEG), initiated within 12 h after cardiac arrest, allows reliable prognostication in approximately 50 % of comatose patients. Serum biomarkers may complement early EEG-based outcome prediction, particularly in indeterminate cases. We evaluated the potential additive prognostic value of the serum biomarkers neuron-specific enolase (NSE), neurofilament light chain (NfL), S100 calcium binding protein β (S100B), and phosphorylated tau (p-Tau 181) when combined with early EEG.

Methods: In this pilot study, we analysed serum biomarker concentrations at multiple time points (<12 h, 24 h, 72 h, 7 days) post-CA in comatose out-of-hospital CA patients included in the ghrelin in coma (GRECO) trial. EEG recordings were visually evaluated at 12 and 24 h post-arrest. Neurological outcomes were assessed using the Cerebral Performance Category (CPC) score at 6 months, dichotomised into good (CPC 1-2) or poor (CPC 3-5).

Results: A total of 49 patients were included; 24 (49 %) had a poor neurological outcome at 6 months. Serum biomarker concentrations were significantly higher in poor-outcome patients within 24 h post-CA. NfL achieved an AUC of 0.90, followed by p-Tau181 (AUC 0.81), S100B (AUC 0.80) and NSE (AUC 0.76) for poor neurological outcome at 24 h post-CA. In 40 patients with EEG, NfL > 100 pg/mL reclassified indeterminate cases as likely poor outcome, with >128 pg/mL at any time achieving 100 % specificity, unlike other biomarkers.

Significance: NfL demonstrated superior prognostic performance compared to other serum biomarkers. Our preliminary findings suggest added prognostic value of NfL when combined with early EEG (12-24 h), particularly in patients with indeterminate EEG findings.

Keywords: Biomarkers; Cardiac arrest; EEG; Neurological outcome; Prognostication.

Fig. 1

Flowchart of patient inclusion. Day 0 is defined as ≤12 h post-cardiac arrest (CA). CPC, Cerebral Performance Category; EEG, electroencephalography; OHCA, out-of-hospital cardiac arrest.

Fig. 2

Serum biomarker concentrations between neurological outcome groups across multiple time points. (A–D) Box-and-whisker plots showing serum concentrations stratified by 6-month neurological outcome. Biomarker levels were significantly elevated in patients with poor neurological outcomes across multiple time points, with the most pronounced differences observed within 24 h. (E–F) Receiver operating characteristic (ROC) curves for predicting poor outcome. Number of patient samples per time point: ≤12 h, n = 49; 24 h, n = 45; 72 h, n = 39; Day 7, n = 16. ns, not significant; *: p < 0.05, **: p < 0.01, ***: p < 0.001, Mann-Whitney U test with Bonferroni correction.

Fig. 3

Peak biomarker concentrations present at unfavourable, indeterminate and favourable EEG patterns recorded within 24 h after cardiac arrest and the overall neurological outcome at 6 months in 40 cardiac arrest patients. Dots represent patients with good and crosses those with poor neurological outcomes. In the group with indeterminate EEG, all patients with a NfL > 100 pg/mL within 24 h had a poor neurological outcome. The one patient with low biomarker concentrations and a favourable EEG died from a non-neurological cause.