Objective: Serum calcitonin (CT) level is used to detect medullary thyroid carcinoma (MTC), but the cut-off level is unclear. We aimed at identifying the optimal cut-off value of basal serum CT levels for detecting MTC.
Design and patients: We retrospectively enrolled patients with hypercalcitoninemia (≥2·9 pmol/l) who had undergone thyroid ultrasonography (US) and subsequent work-up between 2001 and 2013 at Asan Medical Center. We divided patients into four groups: proven MTC (group 1, n = 93), pathologically proven non-MTC after surgery (group 2, n = 57), benign single nodule by cytology (group 3, n = 68) and patients without nodules on US (group 4, n = 24).
Measurement: Basal serum CT levels were evaluated.
Results: The median CT level of group 1 (119·5 pmol/l) was significantly higher than those of other groups (4·0, 3·8 and 3·8 pmol/l, P < 0·001). When we adopted 19·0 pmol/l of CT level as a cut-off value, the sensitivity, specificity, and positive and negative predictive values were 77·4%, 98·7%, 97·3% and 87·8%, respectively. When we compared 29·2 pmol/l (100 pg/ml) and 19·0 pmol/l (65 pg/ml) as cut-off values, 19·0 pmol/l was more sensitive and accurate than 29·2 pmol/l. Factors associated with hypercalcitoninemia in non-MTC groups were autoimmune thyroiditis, chronic kidney disease, proton pump inhibitors and other malignancies. Serum CT levels tended to decrease spontaneously in non-MTC groups.
Conclusion: Basal serum CT levels higher than 19·0 pmol/l can be a useful cut-off value for detecting macroscopic MTC, even though values below 19·0 pmol/l cannot exclude the presence of MTC like small volume MTC or premalignant C-cell hyperplasia.
© 2014 John Wiley & Sons Ltd.