Neuron-Specific Enolase

Neuron specific enolase (NSE, gamma gamma enolase) is a glycolytic enzyme of 78 kDa found in neurons and neuroendocrine cells, and also in red cells and platelets. It is released into the blood and cerebrospinal fluid when these cells are damaged or when tumours of neuroendocrine origin turn over rapidly. It is measured in serum or plasma by immunoassay, in microg/L (ng/mL).

The test is a marker of small cell lung cancer (SCLC) and neuroblastoma, where it supports diagnosis, prognosis and monitoring. It is also raised in other neuroendocrine tumours and, in cardiac arrest, high levels at 48 to 72 hours predict poor neurological outcome.

Small cell lung cancer accounts for about 13 to 15% of lung cancers, is closely linked to smoking, and has a median survival of 2 to 4 months untreated and 10 to 16 months with treatment. Neuroblastoma is the commonest extracranial solid tumour in children, with an incidence of about 10 per million children under 15 years each year, and most cases present in children under 5 years.

Indications. Suspected or proven small cell lung cancer, to support the diagnosis when histology is limited, and to monitor treatment, neuroblastoma in children, neuroendocrine tumours, and prognosis after cardiac arrest or severe brain injury.

Specimen. Serum separated within 1 hour, since red cells and platelets contain NSE and haemolysis falsely raises the result. Plasma or cerebrospinal fluid can also be used. Avoid haemolysed samples and delay at room temperature.

Reference range. Serum under 16.3 microg/L in most assays (ranges 12.5 to 17 microg/L). Children have somewhat higher values, up to 25 microg/L in infants.

Small cell lung cancer.

• Sensitivity is about 60 to 80% overall, with raised values in 70% of extensive disease and about 40% of limited disease. Specificity is 80 to 90% against benign lung disease.
• Values over 100 microg/L strongly suggest SCLC in a patient with a lung mass, and values over 25 microg/L raise the possibility.
• A fall of over 50% within 1 to 2 cycles of chemotherapy indicates response, and a rise, or an early rise after an initial fall, suggests relapse and may precede imaging by weeks.
• Baseline levels correlate with tumour burden and extensive stage, and with shorter survival.

Neuroblastoma. Raised in most patients with advanced disease (over 100 microg/L in extensive disease), and higher values are associated with poor prognosis. It is used with urinary catecholamine metabolites (vanillylmandelic acid and homovanillic acid), ferritin and LDH. Levels are prognostic in stage 4 disease.

Other tumours. Carcinoid and other neuroendocrine tumours, phaeochromocytoma, medullary thyroid carcinoma, melanoma, seminoma, Merkel cell carcinoma, pancreatic neuroendocrine tumours, and some non-small cell lung cancers (large cell neuroendocrine carcinoma).

Brain injury and cardiac arrest. NSE over 60 microg/L at 48 to 72 hours after cardiac arrest, especially if rising, predicts poor outcome, with a false positive rate under 5% in many studies. It is interpreted with examination, EEG, imaging and evoked potentials, and never alone.

Causes of raised NSE other than cancer.

• Haemolysis (in vivo or in vitro), which falsely raises results by several fold.
• Benign lung disease, pneumonia, interstitial lung disease, and tuberculosis.
• Stroke, traumatic brain injury, seizures, encephalitis, Creutzfeldt-Jakob disease, and neuronal death.
• Kidney failure and dialysis (modest rise).

Critical values and findings needing immediate action.

• NSE over 100 microg/L with a lung mass or mediastinal widening: urgent staging, since SCLC doubles in size in weeks.
• Superior vena cava obstruction, spinal cord compression, hyponatraemia (SIADH), or Cushing syndrome from ectopic ACTH in a patient with a raised NSE.
• A rapid rise in a child with an abdominal mass and bone pain.
• Rising NSE during or soon after treatment in SCLC.

Interferences and pitfalls.

• Haemolysis is the commonest source of error, and a haemolysed sample is rejected.
• NSE is not a screening test. It is normal in about 30% of SCLC and cannot exclude it.
• Assays differ in range, and serial testing uses the same method.
• Poor sample handling (delayed separation) gives falsely high values.
• Heterophile antibodies and biotin can interfere.
• NSE does not distinguish SCLC from other neuroendocrine tumours, and tissue diagnosis is needed.
• Pro-gastrin-releasing peptide (ProGRP) is an additional marker with less haemolysis interference.
• After cardiac arrest, targeted temperature management changes timing, and the values must be taken at the right hours.

Special groups.

• Children: higher reference ranges, and the test is part of neuroblastoma assessment.
• Smokers and older adults: higher baseline in lung disease.
• Dialysis patients: modest elevation.
• Neonates: higher values from birth injury.
• Regions: where tissue diagnosis is limited, NSE may support a presumptive diagnosis of SCLC.

Follow up tests. ProGRP, LDH, sodium, calcium, cortisol and ACTH when Cushing syndrome is suspected, CT of chest and abdomen, brain MRI, bone scan or PET-CT, bronchoscopy with biopsy and immunohistochemistry (synaptophysin, chromogranin, CD56, TTF-1), bone marrow examination, urinary catecholamine metabolites and MYCN testing in neuroblastoma, and EEG and imaging after cardiac arrest.