18F-FDG-Avid Thyroid Incidentalomas: The Importance of Contextual Interpretation.

ABSTRACT

18F-FDG-avid thyroid incidentaloma (TI) is seen in approximately 2.5% of patients imaged for staging or response assessment of malignancy and represents thyroid cancer in approximately 35% of cases. Consequently, the 2015 American Thyroid Association guidelines strongly recommend investigation of all 18F-FDG-avid nodules 1 cm or larger with ultrasound and fine-needle aspiration cytology (FNA). This study aimed to assess the overall and thyroid cancer-specific survival in a large cohort of patients with 18F-FDG-avid TI with long-term follow-up to assess the validity of this approach. Methods: Retrospective review of 45,680 PET/CT scans performed at a comprehensive cancer center from January 2007 to January 2015 identified 2,588 18F-FDG PET/CT reports referring to the thyroid. After exclusion of nonavid thyroid nodules, diffuse 18F-FDG uptake, known thyroid cancer, abnormalities adjacent to the thyroid, and repeat studies, 500 patients (1.1%) with TI were identified, of whom 362 had confirmed death or more than 12 mo of clinical follow-up. Variables including age, sex, primary malignancy, overall survival, thyroid cancer-specific survival, FNA, and histopathology were collected until January 2016. Multivariate logistic regression and survival analysis were performed. Results: The 362 analyzed patients (65% female) had a median age of 65 y (range, 19-96 y) and follow-up of 24 mo (range, 1-103 mo). Lymphoid, lung, and colorectal malignancy were the most common staging indications. Median overall survival was 20 mo (interquartile range, 9.5-39 mo). Most of the 180 observed deaths were due to the primary malignancy under investigation (92.2%) or to causes not related to cancer (7.2%); one patient (0.6%) died from incidentally detected medullary thyroid cancer. 18F-FDG avidity in the index malignancy, an advanced stage for that malignancy, and a clinician decision not to investigate 18F-FDG-avid TI were all predictors of mortality, with hazard ratios of 8.5, 3.0, and 3.3, respectively, and 95% confidence intervals of 4.6-15.8, 2.3-3.9, and 2.0-5.0, respectively (P < 0.001). Of 131 patients suitable for cytologic or histopathologic evaluation, 47 (36%) had incidental thyroid cancer (24 papillary, 11 malignant FNA, 5 oncocytic/Hürthle cell, 2 medullary, 1 follicular, and 4 metastases from underlying malignancy). Conclusion: Overall survival with 18F-FDG-avid TI was poor because of the prognosis associated with underlying malignancy, which must be considered before investigation of 18F-FDG-avid TI and certainly before aggressive treatment. Active surveillance should be considered in this group of patients.