Background: Chromogranin A and B (CgA, CgB) are widely used as markers of neuroendocrine tumour (NET) disease. However, the sensitivity of the test is variable and the value of CgB measurement is unclear. We thus sought to review 3 years of CgA and CgB requests made in our hospital.
Methods: CgA and CgB were measured at the Hammersmith Hospital. Case notes were reviewed in each of these patients.
Results: There were 520 tests performed in 152 patients (70 male, 82 female) between 2004 and 2006. Requests were made mainly by endocrinologists (56%) or pancreaticobiliary surgeons (26%) with a smaller number made by gastroenterologists (5%) or others. Most requests were made in confirmed/suspected NETs (73%) but a substantial number were also made in MEN1 (16%), with smaller numbers in phaeochromocytoma (5%) and unexplained diarrhoea (6%). CgA levels were elevated in 42% of subjects, while CgB was elevated in only 14 (9%), all of whom had markedly elevated CgA. Using a cut-off of >70 pmol/l, specificity of CgA for NET was 87%, increasing to 97% with a cut-off of >100 pmol/l. CgA levels were highest in metastatic mid-gut NETs (median 397, range 71->1000 pmol/l), where they correlated with disease stage (73  localised versus 417 [71>1000] pmol/l distant metastases, P<0.001). CgA was raised in only 6/23 subjects with MEN1; 4 had pancreatic NETs on MRI compared to only 2 of 17 with negative CgA. Sensitivity of CgA was 33% in the small number of phaeochromocytomas where it was measured.
Conclusions: CgA measurement is generally being requested appropriately but has little value in the investigation of diarrhoea alone. Its role in the investigation of phaeochromocytoma and MEN1 requires further clarification. CgB has no added value to CgA in the management of NETs.