Chromogranin A

Chromogranin A (CgA) is an acidic 49 kDa glycoprotein stored with hormones in the secretory granules of neuroendocrine cells and released with them. It is raised in most functioning and non-functioning neuroendocrine tumours (NETs) of the gut, pancreas and lung, in phaeochromocytoma and paraganglioma, and in medullary thyroid carcinoma. It is measured by immunoassay in serum or plasma, in microg/L (ng/mL) or U/L, depending on the assay.

The main use is in monitoring a known NET, as an adjunct to imaging and clinical assessment. Its value in diagnosis is limited by poor specificity, since many benign conditions raise it, especially acid suppression.

Neuroendocrine tumours have an incidence of about 5 to 7 per 100 000 people a year, with small intestine and pancreas the common primary sites in the gut. Most are slow growing, and the median survival is years even with metastases. Carcinoid syndrome occurs in about 20 to 30% of small bowel NETs with liver metastases.

Indications. Follow up of a known NET after resection or during treatment, assessment of tumour burden in hepatic metastases, prognosis in metastatic NET, and supportive evidence in a suspected NET with symptoms such as flushing, diarrhoea, wheeze or hypoglycaemia.

Specimen. Serum or EDTA plasma, taken fasting where possible, and frozen if not analysed the same day. Proton pump inhibitors should be stopped for at least 2 weeks before sampling if safe to do so. Record all medicines and kidney function.

Ranges are assay specific and vary up to ten fold between methods. Typical upper limits are under 100 ng/mL for the radioimmunoassay and about 100 to 120 U/L for the older ELISA, and 5 to 6 nmol/L for the newer kits.

In neuroendocrine tumours.

• Sensitivity is 60 to 80% in NETs overall, with higher values in tumours with large volume of disease, midgut NETs with liver metastases, and well differentiated tumours. It is lower in insulinoma, poorly differentiated carcinomas and small tumours.
• Specificity is only 70 to 90% because of false positives.
• Levels correlate with tumour burden, and a rise of over 25% on serial samples suggests progression, while a fall suggests response. Imaging confirms.
• High pre-treatment levels predict poorer prognosis, and a fall after surgery suggests complete removal.
• Specific hormones are more useful than CgA in functioning tumours, namely 5-hydroxyindoleacetic acid in carcinoid syndrome, gastrin in gastrinoma, insulin and C-peptide in insulinoma, and calcitonin in MTC.

Other tumours. Phaeochromocytoma and paraganglioma, medullary thyroid carcinoma, small cell lung cancer, prostate cancer with neuroendocrine differentiation (a marker of castration resistance), and Merkel cell carcinoma.

Causes of false positive results.

• Proton pump inhibitors and H2 blockers (values often two to ten times the upper limit).
• Atrophic gastritis and pernicious anaemia.
• Kidney impairment (reduced clearance, values often elevated with an eGFR under 60 mL/min).
• Hypertension, heart failure and acute coronary syndromes, cardiac arrhythmia.
• Inflammatory bowel disease, irritable bowel syndrome, chronic hepatitis and cirrhosis, pancreatitis.
• Rheumatoid arthritis, obstructive lung disease, glucocorticoid use, and non-neuroendocrine cancers (breast, colon, prostate, pancreatic).

Critical values and findings needing immediate action.

• A rapid rise in a patient with a known NET and new symptoms (pain, jaundice, weight loss, carcinoid syndrome with heart murmur or right heart failure).
• Marked elevation (over 10 times the upper limit) in a patient not on acid suppression and with kidney function normal, which requires imaging.
• Carcinoid crisis during surgery or intervention in a patient with a functioning tumour, which is prevented with octreotide cover.
• Very high CgA in a patient with hypertension and paroxysms, which needs metanephrines before any procedure.

Interferences and pitfalls.

• Not a screening test, and not a stand alone diagnostic test.
• Proton pump inhibitor use is the commonest cause of a false rise, and the effect lasts about 2 weeks after stopping.
• Different assays and different antibodies recognise different fragments and cannot be compared, and serial results use the same method.
• Heterophile antibodies and the hook effect can distort results.
• Kidney impairment causes a persistent rise.
• Somatostatin analogues may lower values without a change in disease.
• A normal CgA does not exclude a NET, and it is often normal in rectal, appendiceal, insulinoma and poorly differentiated tumours.
• Pancreastatin (a CgA fragment) may be more specific in some centres.

Special groups.

• Older adults: higher baseline from comorbid disease and medicines.
• Patients with kidney failure: the test has little value.
• Children: rarely used, except in neuroblastoma and paraganglioma syndromes.
• Pregnancy: no major change.
• Regions: the test may be unavailable, and hormonal markers with imaging are used.

Follow up tests. Repeat after stopping acid suppression, urinary 5-HIAA (24-hour or spot), gastrin, insulin and C-peptide, calcitonin, plasma metanephrines, pancreatic polypeptide, somatostatin receptor PET-CT, contrast CT or MRI, endoscopic ultrasound, biopsy with Ki-67 index and grade, echocardiography in carcinoid syndrome, and genetic testing (MEN1, MEN2, VHL, NF1, SDHx) in familial disease.