Can the pathology of a thyroid nodule be determined by positron emission tomography uptake?

To determine if standardized uptake values (SUV) on positron emission tomography (PET) are predictive of thyroid pathology and the significance of serial SUV measurements of thyroid nodules over time.

Case series with chart review.

Academic health center.

In total, 23,384 PET and PET/computed tomography (CT) scans were performed between December 2001 and April 2011.

Patients with incidental thyroid uptake were identified. SUV(max), age, sex, size of thyroid lesion, indication for PET scan, and cytology/pathology were collected.

Incidental thyroid uptake was noted in 1309 PET scans (5.60%), focal uptake in 690 (2.95%), and diffuse uptake in 619 (2.65%). Complete data were available for 359 PET scans from 103 patients. Malignancy was identified in 28 patients (27%). Twenty-five of the 28 lesions (89%) were primary thyroid malignancies. A significant difference between malignant SUV(max) and benign SUV(max) was found (mean ± SD, 7.04 ± 7.88 for malignancies vs 3.85 ± 3.06 for benign tumors, P = .0292). Receiver operating characteristics curves were constructed on patients with PET data within 3 months of diagnosis and indicated that a SUV(max) of 4.2 differentiated maximally between benign and malignant lesions. Serial SUV uptake had no significant change over time.

All thyroid nodules with focal uptake on (18)F-fluorodeoxyglucose-PET/CT should be considered at higher risk of malignancy than those discovered incidentally by other imaging modalities. Higher SUV(max) values are more indicative of malignant lesions. All lesions should be evaluated with ultrasonography ± fine-needle aspiration if no clinical contraindications exist. Size of the primary nodule does not influence SUV(max) uptake.