Background
New-onset refractory status epilepticus, or NORSE, is a medical emergency in which a person with no clear prior epilepsy history develops continuous or repeatedly recurring seizures that do not stop with first-line treatments. A cryptogenic case, often called cNORSE, means that no cause is identified despite extensive testing. Because ongoing seizures can injure the brain, clinicians urgently need markers that show whether neuronal damage is actively happening and how severe it is.
This study focused on two blood and spinal fluid biomarkers: neurofilament light chain, or NfL, which rises when nerve cell axons are damaged, and S100-beta, or S100B, a protein associated with glial injury and blood-brain barrier disruption. The researchers asked whether these markers could distinguish cNORSE from status epilepticus with a defined cause, and from people with chronic epilepsy or no status epilepticus at all.
Study design
This was an international cross-sectional study conducted from 2013 to 2025 across multiple hospitals in the United States, Canada, Italy, France, and Belgium. The investigators included patients with cNORSE and patients with etiology-defined status epilepticus, or eSE, in whom blood or cerebrospinal fluid samples were collected while seizures were still ongoing. Comparison groups included patients with chronic epilepsy and healthy participants. No enrolled participants were excluded from the analysis.
The main outcomes were the levels of NfL and S100B in serum and cerebrospinal fluid, and whether these levels were associated with short-term functional outcome at discharge. Functional status was assessed using the Glasgow Outcome Scale extended, where lower scores indicate worse recovery.
Key findings
The study included 78 patients with cNORSE and two independent eSE cohorts of 211 and 73 patients. Patients with cNORSE had markedly higher NfL levels than the eSE groups. In cerebrospinal fluid, NfL levels were about 10 times higher in cNORSE; in serum, they were about 4 times higher. The differences were substantial and statistically significant.
Compared with people who had chronic epilepsy and healthy controls, serum NfL in cNORSE was nearly 20 times higher. This suggests that acute neuroaxonal injury in cNORSE is not just present, but severe.
Serum and cerebrospinal fluid NfL levels tracked closely together, meaning a blood test may reflect what is happening in the central nervous system. NfL levels also rose rapidly over time after seizure onset. They increased from week 1 to week 2, and again by week 3, showing a clear early escalation in injury signal during the acute phase of illness.
By contrast, S100B did not show meaningful differences between groups and did not follow a consistent time pattern. In this study, S100B was not as useful as NfL for identifying acute injury in NORSE.
Clinical meaning
The findings support the idea that cNORSE is associated with major acute injury to nerve axons. This matters because the degree of injury may influence both prognosis and treatment urgency. If neuronal damage rises quickly during the first weeks, then there may be a narrow window in which aggressive treatment can prevent irreversible harm.
NfL also helped distinguish cNORSE from eSE, with moderate to good accuracy, and from patients without status epilepticus, with excellent accuracy. That suggests NfL could eventually become a practical biomarker for helping clinicians recognize especially severe seizure-related brain injury.
Higher serum NfL was also independently linked to poorer functional outcome at discharge. In other words, patients with more injury signal in the blood were more likely to leave the hospital with worse neurologic function. This does not prove causation, but it strengthens the case that NfL reflects clinically meaningful brain damage.
What this means for treatment
Although this study was not a treatment trial, it has important implications for management. NORSE often requires rapid escalation beyond standard anti-seizure medications, including anesthetic agents, immune-directed therapy when autoimmune causes are suspected, and intensive care support. The rapid rise in NfL suggests that waiting too long may allow substantial injury to accumulate before seizures are controlled.
In practical terms, the study supports the need for fast diagnostic workup, early consideration of etiologies that can be treated, and timely seizure suppression. It also raises the possibility that future therapies in cNORSE may need to be not only anti-seizure but also neuroprotective.
Limitations
This was a cross-sectional study, so it can show association but not prove that high NfL causes poor outcome. The timing of sample collection varied among patients, and the eSE groups came from independent cohorts, which can introduce differences in age, underlying illness, and care practices. Also, while NfL appears promising, it is not yet a routine bedside test for NORSE in most hospitals.
Even with these limitations, the consistency of the findings across serum and cerebrospinal fluid, and the strong separation between cNORSE and control groups, make the results compelling.
Bottom line
This study shows that cNORSE is associated with severe and rapidly developing acute brain injury, best captured by elevated neurofilament light chain levels. S100B was not helpful in the same way. The results point to a short therapeutic window early after seizure onset, reinforcing the need for urgent, effective, and potentially neuroprotective treatment in patients with cNORSE.