Objective: The purpose of our study was to determine the prevalence of malignancy in incidental abnormalities of the thyroid gland detected on CT and to determine the relative accuracy of characterizing these abnormalities on CT as compared with sonography and pathology.
Materials and methods: We searched our department's computerized clinical database for all thoracic and cervical CT scans in which a new abnormality was incidentally identified in the thyroid gland from 1998-2001. Two hundred thirty patients with abnormal findings in the gland on CT subsequently underwent thyroid sonography, and 118 of the 230 patients underwent a diagnostic biopsy or resection. CT and sonographic images were directly reviewed to identify imaging features of each thyroid abnormality, including the location, size, appearance, and presence or absence of calcifications. Associations were evaluated using Fisher's exact test of significance and the Student's t test. The overall rate of malignant and potentially malignant lesions among these incidental abnormalities of the thyroid gland was calculated.
Results: CT findings matched the sonographic characterization in 122 patients (53.0%), correctly identified the dominant nodule but missed multinodularity in 69 (30.0%) patients, and underestimated the number of nodules in 24 (10.4%) patients. CT overestimated the number of nodules in 5 (2.2%) patients and was false-positive for lesions in 10 patients (4.3%). Ninety-one patients with a single or dominant nodule on CT had pathologic correlation: 7 nodules were malignant, 17 showed malignant potential, and 67 were benign. Of 27 patients with multinodular or enlarged thyroid glands on CT and histopathologic correlation, 2 lesions were malignant and 25 benign. The presence of punctate calcifications on CT significantly correlated to the presence of microcalcifications on sonography (p < 0.02). Benign nodules were significantly smaller (mean, 2.16 +/- 1.01 cm; range, 0.6-4.5 cm) than malignant and potentially malignant nodules (mean, 2.79 +/- 0.99 cm; range, 0.7-4.6 cm) (p = 0.01). Patients 35 years or younger who had a thyroid lesion on CT were more likely to have malignancy (p < 0.01). Overall, among incidentally detected lesions of the thyroid gland, there was at least a 3.9% rate of malignancy (95% CI: 1.8-7.3%) and 7.4% rate of malignant potential (95% CI: 4.4-11.6%).
Conclusion: There is at least an 11.3% prevalence of malignant or potentially malignant lesions among incidental thyroid abnormalities detected on CT. Patients 35 years or younger who have incidental abnormalities have a significantly greater rate of malignancy. No CT feature reliably distinguishes benign from malignant lesions in the thyroid gland. CT underestimates the number of nodules relative to sonography, which suggests that sonography is a useful adjunctive test after the incidental detection of a thyroid abnormality on CT.