Fine‐needle aspiration in all thyroid incidentalomas at 18 F‐ FDG PET / CT : Can EU ‐ TIRADS revise the dogma?

Summary Objective Focal thyroid incidentalomas ( TI s) are observed in 2% of 18 F‐ FDG PET / CT representing malignancy in one‐third of cases. Currently, due to the lack of evidence on their optimal management, guidelines suggest fine‐needle aspiration cytology ( FNAC ). The study aim was to evaluate the role of ultrasound evaluation according to EU ‐ TIRADS to assess the risk of TI s and inform FNAC prescriptions. Design We retrospectively reviewed 18 F‐ FDG PET / CT TI s recorded during the period 2014‐2017. Enrolled were TI s with histological outcome and autonomous nodules. Cases with uncertain matching between 18 F‐ FDG PET / CT , ultrasound and histology were excluded. Results According to the selection criteria, 75 TI s, being 13 (17.3%) malignant and 62 (82.7%) benign, were included. Cancers had significantly higher SUV max and SUV max ratio (Mann‐Whitney P  < 0.01) than benign, and the most accurate cut‐offs were >7.1 and >3.65, respectively. At ultrasound, the cancer rate was 0% in EU ‐ TIRADS 2, 2.9% in EU ‐ TIRADS 3, 4.2% in EU ‐ TIRADS 4% and 78.6% in EU ‐ TIRADS 5 (chi‐squared P  < 0.001). Sensitivity, specificity, PPV , NPV and accuracy for malignancy were 92%, 64%, 35%, 98% and 69% for SUV max; 85%, 68%, 36%, 96% and 71% for SUV max ratio; and 85%, 95%, 79%, 97% and 93% for EU ‐ TIRADS , respectively. The absence of all these three features reached a specificity of 97.1%. Conclusions EU ‐ TIRADS , within a clinical careful approach, can discriminate with significant accuracy lesions at high risk of malignancy from those at low risk among TI s at 18 F‐ FDG PET / CT . Additionally, a centre‐based threshold for SUV parameters should be useful for the initial assessment of these lesions during PET / CT reading and reporting.