Background: The extension of the compartment-oriented neck dissection at primary surgery in medullary thyroid carcinoma (MTC) is controversial. Because a <50 % decrease in intraoperative calcitonin levels (IO-CT) after total thyroidectomy plus central neck dissection (TT-CND) has been associated with residual disease, IO-CT monitoring has been proposed to predict the completeness of surgery. The goal of the present prospective study was to verify the accuracy of IO-CT monitoring.
Methods: All patients scheduled for primary surgery for suspected or proven MTC between November 2010 and January 2013 were included. Calcitonin was measured pre-incision (basal level), after tumor manipulation, at the time TT-CND was accomplished (ablation level), 10 and 30 min after ablation. A decrease >50 % with respect to the highest IO-CT level 30 min after ablation was considered predictive of cure.
Results: Twenty-six patients were included, and IO-CT monitoring identified 18 of 23 cured patients (true negative results) and 2 of 3 patients with persistent disease (true positive result). In 5 patients with normal basal and stimulated postoperative calcitonin levels, a decrease <50 % was observed (false positive results). In one of three patients with persistent disease a >50 % decrease in IO-CT was observed (false negative results). Specificity, sensitivity, and accuracy of IO-CT were 78.2, 66.6, and 76.9 %, respectively.
Conclusions: Intraoperative calcitonin monitoring is not highly accurate in predicting the completeness of surgical resection. In the present series, relying on IO-CT would result in limited resection in about one third of the patients with residual neck disease and in unnecessary lateral neck dissection in about 20 % of the cured patients.