Dedicated Neck 18F-FDG PET/CT: An Additional Tool for Risk Assessment in Thyroid Nodules at Ultrasound Intermediate Risk

Pierpaolo Trimboli; Arnoldo Piccardo; Maria Alevizaki; Camilla Virili; Mehrdad Naseri; Simona Sola; Gaetano Paone; Gilles Russ; Luca Giovanella


Clin Endocrinol. 2019;90(5):737-743. 

In This Article

Abstract and Introduction


Background: Several ultrasound (US) risk stratification systems have been proposed for the assessment of thyroid nodules, and their performance was shown as good. However, the rate of nodules assessed at intermediate risk is not negligible and whether they should be submitted or not to further examination is still under debate. The present study aimed to evaluate the reliability of 18F-FDG PET/CT in stratifying the risk of malignancy in these lesions.

Methods: Two institutions participated to this retrospective study in which a dedicated 18F-FDG PET/CT was proposed to patients having a thyroid nodule with US assessment of EU-TIRADS 4 or 5. 18F-FDG PET/CT did not influence the diagnostic and therapeutic decision. Histology was the gold standard for all patients.

Results: Ninety-three patients were included for the study with 48 EU-TIRADS 4 and 45 EU-TIRADS 5 nodules. Of these, 26 underwent thyroidectomy following FNAC suspicious for or consistent with malignancy, 38 for inconclusive cytology, 27 because of large goitre and 2 for high-risk lesion at US. At histology, 35 carcinomas and 58 benign lesions were found. Cancer prevalence was 16.7% in EU-TIRADS 4 and 60% in EU-TIRADS 5. Overall, 18F-FDG PET/CT was positive in 33/35 cancers (94.5% sensitivity) and negative in 31/58 benign lesions (53.4% specificity). When considering only EU-TIRADS 4, 18F-FDG PET/CT was positive in 7/8 cancers and negative in 20/40 benign lesions; among these, there were 36 cases with FNAC indication according to dimensional cut-off (ie >1.5 cm), and 18F-FDG PET/CT showed 85.7% sensitivity and 41.4% specificity.

Conclusions: 18F-FDG PET/CT may have a role in stratifying the cancer risk of thyroid nodules with an intermediate ultrasound assessment. More specifically, thyroid lesions classified as EU-TIRADS 4 and with no 18F-FDG uptake could be ruled out from further examination, similar to other anamnestic and clinical suspicious factors of patients. Further prospective and cost-effectiveness studies are needed.


Ultrasound (US) is pivotal in the initial assessment of a thyroid nodule. In practice, according to the US characteristics a nodule will either undergo further investigation or be referred to clinical follow-up. This approach is supported by a significant number of papers published in the last two decades.[1,2] Nevertheless, some limitations of US, such as low reproducibility and operator-depending performance, may reduce its value. Consequently, in the last years several US risk stratification systems have been proposed to cover these limitations and provide to thyroidologists an easy-to-use tool guide for their clinical practice.[3–8] As a proof of concept, these systems have shown a good performance in identifying suspicious thyroid nodules to be submitted to fine needle aspiration cytology (FNAC).[9] However, the rate of nodules assessed at intermediate risk is not negligible and whether they should be submitted to FNAC or not is still under debate; in fact, there is a huge step separating their risk from that of high suspicious nodules (ie 6%-17% vs 26%-87% for EU-TIRADS,[7] 10%-20% and 70%-90% in ATA,[3] respectively). Specifically, to further improve the selection of nodules for FNAC some guidelines suggest adopting a size cut-off, such as 1.5 cm in EU-TIRADS[7] and 1 cm in ATA.[3] These thresholds were arbitrarily proposed, and their applicability during the clinical practice may be poor.

The use of fluorine-18-fluorodeoxy-glucose (18F-FDG) positron emission tomography/computed tomography (PET/CT) has significantly spread during the last years. In differentiated thyroid cancer (DTC), 18F-FDG PET/CT has reached a role for the assessment of patients with DTC suspicious for relapse and negative 131I-whole body scan.[3–10] Also, in the field of thyroid nodule, several papers showed that FDG avid thyroid lesions incidentally detected during PET/CT acquisitions for oncologic purpose carry high risk of malignancy.[11–13] In addition, molecular markers[14] hold a potential role in discriminating high from low-risk nodules among thyroid indeterminate lesions (ie Thy 3).[15,16] However, whether 18F-FDG PET/CT could also have a role in the risk assessment of thyroid nodules clinically detected in the routine practice remains unclear. Particularly, to the best of our knowledge no data are available about the ability of 18F-FDG PET/CT, specifically performed for thyroid purpose, in stratifying the risk of malignancy of thyroid nodules with intermediate-suspicious US features.

The present study was designed to evaluate the potential role of 18F-FDG PET/CT in the specific category of EU-TIRADS 4 nodules at US intermediate risk of malignancy.[7] Thus, initially we analysed the performance of 18F-FDG PET/CT in discriminating cancers from benign lesions among nodules classified as EU-TIRADS 4 in terms of sensitivity, specificity, positive and negative predictive value, also when considering the above dimensional cut-off. In addition, to better assess the reliability of 18F-FDG PET/CT, the results recorded in EU-TIRADS 4 were compared with those obtained in a series of nodules at high risk (ie EU-TIRADS 5).[7]