Brain Damage in Refractory Status Epilepticus Is 10x Worse Than Other Seizure Disorders
NfL biomarker levels in NORSE patients were 10-fold higher in CSF and 4-fold higher in serum than other seizure disorders, signaling massive neuroaxonal injury.
Summary
A large international study measured brain injury biomarkers in patients with new-onset refractory status epilepticus (NORSE), a devastating seizure disorder with no known cause. Researchers found that neurofilament light chain (NfL), a protein released when neurons are damaged, was roughly 10 times higher in spinal fluid and 4 times higher in blood compared to patients with other types of status epilepticus. Levels rose sharply week by week, nearly 20-fold above healthy controls. Importantly, higher NfL levels predicted worse functional outcomes at hospital discharge. These findings suggest a narrow window for aggressive treatment and position NfL as a clinically useful biomarker for guiding management of this life-threatening condition.
Detailed Summary
New-onset refractory status epilepticus (NORSE) is one of the most severe neurological emergencies, characterized by prolonged, treatment-resistant seizures with no identifiable structural or metabolic cause. Despite its devastating outcomes — including death, severe disability, and chronic epilepsy — the degree of acute brain injury it inflicts has never been systematically quantified using validated biomarkers. This international cross-sectional study, published in JAMA Neurology, set out to fill that gap by measuring neurofilament light chain (NfL) and S100-beta (S100B) proteins in serum and cerebrospinal fluid (CSF) across multiple patient groups.
The study enrolled 78 patients with cryptogenic NORSE (cNORSE), drawn from a biorepository spanning 36 US hospitals plus sites in Canada, Italy, France, and Belgium, with samples collected between 2013 and 2025. Two independent comparison cohorts included 211 and 73 patients with etiology-defined status epilepticus (eSE), along with control groups of patients with chronic epilepsy and healthy individuals. NfL is a structural protein of neuronal axons released into biofluids upon neuroaxonal injury, while S100B is an astrocyte-derived marker of glial damage. Both are well-validated in traumatic brain injury and neurodegenerative disease but had not been rigorously studied in NORSE.
The results were striking. CSF NfL in cNORSE patients reached a median of 6,408 pg/mL (IQR 1,503–22,963), compared to 694 pg/mL (IQR 219–2,389) in eSE cohorts — approximately a 10-fold difference (P<0.001). Serum NfL was similarly elevated: 231 pg/mL (IQR 99–855) in cNORSE versus 55 pg/mL (IQR 20–135) in eSE (P<0.001), and nearly 20-fold higher than in patients with chronic epilepsy (median 11 pg/mL) or healthy controls (median 7 pg/mL). Serum and CSF NfL were strongly correlated (Spearman ρ=0.75, P<0.001), suggesting blood-based testing could substitute for lumbar puncture in monitoring injury.
A critical temporal analysis revealed that NfL levels rose sharply and progressively: median serum NfL was 101 pg/mL in week 1, 197 pg/mL in week 2, and 598 pg/mL by week 3 after seizure onset (P<0.001). This rapid accumulation implies that neuronal destruction is ongoing and accelerating during the first weeks of NORSE, defining a narrow therapeutic window. NfL discriminated cNORSE from eSE with an AUROC of 0.79 (95% CI 0.68–0.90) and from non-SE controls with an AUROC of 0.99 (95% CI 0.78–1.00). Higher serum NfL was independently associated with poor functional outcome at discharge (Glasgow Outcome Scale extended score 1–4; OR 1.01, 95% CI 1.00–1.03, P=0.03).
In contrast, S100B — the astroglial injury marker — showed no significant differences between groups and no consistent temporal pattern, suggesting that in NORSE, neuroaxonal damage dramatically outpaces glial injury. This dissociation is clinically meaningful: it implies the primary pathological process targets neurons and axons rather than supporting glial cells. The authors argue these findings support early, aggressive, and potentially neuroprotective interventions in NORSE, and that serial serum NfL monitoring could serve as a practical bedside tool to gauge treatment response and guide escalation decisions.
Key Findings
- CSF NfL was ~10-fold higher in cNORSE (median 6,408 pg/mL) vs etiology-defined SE cohorts (median 694 pg/mL; P<0.001)
- Serum NfL was ~4-fold higher in cNORSE (median 231 pg/mL) vs eSE (median 55 pg/mL; P<0.001) and ~20-fold higher than healthy controls (median 7 pg/mL)
- Serum NfL rose sharply over time: 101 pg/mL at week 1, 197 pg/mL at week 2, 598 pg/mL at week 3 (P<0.001), indicating accelerating neuronal destruction
- Serum and CSF NfL were strongly correlated (Spearman ρ=0.75, P<0.001), supporting blood-based monitoring as a surrogate for CSF
- NfL discriminated cNORSE from eSE with AUROC 0.79 (95% CI 0.68–0.90) and from non-SE controls with AUROC 0.99 (95% CI 0.78–1.00)
- Higher serum NfL independently predicted poor functional outcome at discharge (OR 1.01, 95% CI 1.00–1.03; P=0.03)
- S100B (astroglial marker) showed no significant differences between groups, indicating disproportionate neuroaxonal vs glial injury in NORSE
Methodology
This international cross-sectional study enrolled 78 cNORSE patients from a multi-site biorepository (36 US hospitals plus Canada, Italy, France, Belgium; 2013–2025) and two independent eSE cohorts (n=211 and n=73), with chronic epilepsy and healthy control comparison groups. NfL and S100B were measured in serum and CSF collected during active seizure activity. Statistical analyses included Spearman correlations, receiver operating characteristic curve analysis, and logistic regression for outcome prediction, with P<0.001 thresholds for primary comparisons.
Study Limitations
The cross-sectional design limits causal inference, and the NORSE cohort was relatively small (n=78) given the rarity of the condition. Samples were collected at variable time points across patients, introducing heterogeneity in temporal comparisons. Multiple authors disclosed financial relationships with pharmaceutical companies including UCB, Jazz, Eisai, Marinus, and Ceribell, though funders had no role in study design or analysis.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.