The role of additional FDG PET/CT in differential diagnosis of the thyroid incidentaloma

Objectives Thyroid incidentaloma found on the FDG PET/CT performed in patients with malignancy from other origin are common. The purpose of this study was to investigate whether the additional FDG PET/CT, scanned 2 hour after injection, could predict the malignancy of thyroid incidentaloma.

Methods FDG PET/CT between November 2007 and February 2009 were reviewed on 29 patients with known malignancy not associated with thyroid but shown abnormal thyroid FDG uptake. To give a differential diagnosis,additional PET/CT were also performed and the maximum SUVs of the thyroid lesions were reviewed. All the patients had undergone Fine needle aspiration and/or operation if there were any definite nodule by palpation or ultrasonography. SUVmax of initial image (SUV1), SUVmax of delayed image (SUV2) and SUVmax change ratio (dSUV) were compared between the patient groups with benign and malignant thyroid nodules. Predictability of FDG PET-CT for malignancy was assessed by ROC curves.

Results Pathologic report revealed that 9 out of 29 patients had papillary thyroid carcinoma, 16 were benign lesion and 4 cases were suspected to be metastatic carcinoma from the primary origin. SUV2 of the 9 PTC (mean 16.7) was significantly higher than that of 16 benign lesions (mean 10.4)(p<0.05). There was also significant difference in dSUV between the patients group with malignant(mean 17.1%)and benign(mean 10.2%) nodules (p<0.05). When SUV2 of 3.9 was used as cut-off threshold, malignancy on the pathology could be predicted with sensitivity 87.5% and specificity 75%. Thyroid lesion that shows dSUV greater than 12.5% could be expected to be malignant (sensitivity 87.5%, specificity 56.3%). SUV1 was not statistically significant between two groups (15.6 vs.10.9)(p<0.15).

Conclusions Glucose metabolism of thyroid incidentaloma measured by FDG PET/CT combined with 2-hour delayed image could be an indirect criteria in differential diagnosis benign lesion from malignant. The optimal cut-off thresholds of SUV on the delay image was 3.9 and SUV change ratio was 12.5%