Real-world EUROCRINE registry data challenge the reliability of Bethesda cytopathology for thyroid surgery indication.

Objectives: Fine-needle aspiration cytology (FNAC) is recommended by international guidelines for the preoperative evaluation of suspicious thyroid nodules >1 cm. Despite robust evidence from endocrine centers demonstrating the key role of FNAC results for the indication of surgery, the method is not routinely used in European clinics. The database EUROCRINE, which was introduced in 2015 with the scope of registering operations of the endocrine system, allows for a large-scale analysis of the current service reality in Europe concerning FNAC use and associated accuracy.

Methods: Operations performed to "exclude malignancy", registered from January 2015 to December 2018 in EUROCRINE, were analyzed. Parameters of accuracy were calculated for FNAC. FNAC results were considered "test positive" in the case of Bethesda category IV, V, and VI, since these categories usually prompt surgical interventions in European centers for thyroid surgery. Bethesda category II and III were considered "test negative".

Results: Of 8,791 operations, 5,780 had preoperative FNAC (65.7%). The overall malignancy rate was 28.3% (2,488/8,791). Malignancy rates were 68.8% for Bethesda VI, 69.9% for Bethesda V, 32.6% for Bethesda IV, 28.2% for III, 20.2% for Bethesda II, and 24.5% for Bethesda I. After exclusion of papillary microcarcinomas (PTMCs), the sensitivity of FNAC was 71.7% and specificity 43.5%, the positive predictive value was 29.1% and the negative predictive value 82.7%.

Conclusions: Although the indication to "exclude malignancy" was the predominant reason that prompted thyroid resection in the present cohort, FNAC was only used in about 65.7% of cases. When performed, FNAC was associated with unexpectedly low accuracy. Interestingly, in Bethesda II, 20.2% of malignant entities were present (13.3% after the exclusion of PTMCs).