Thyroid Nodule Doubling Time is Not a Reliable Indicator of Benign or Malignant Nature

Thyroid nodules are often followed up with serial ultrasound imaging. Doubling time is well established in the evaluation of lung nodules. We examined whether benign and malignant thyroid nodules exhibit differences in doubling time. This retrospective, IRB-approved study included patients with nodules aspirated between January and June 2012 (benign), and January 2012 to December 2014 (suspicious or malignant), no interval thyroidectomy, and two ultrasound examinations longer than 180 days apart. Diameters and morphology were assessed by a single observer, demographics recorded, and doubling time calculated. Area under the receiver operating characteristic curve was derived. Society of Radiologists in Ultrasound criteria were used to determine aspiration appropriateness initially and after follow-up. 59 patients with 61 nodules were included. Statistically significant between-group differences exist for sex, solidity, echogenicity, and microcalcifications, with no significant differences in doubling time, age, days between studies, vascularity, or mean diameter. Benign nodules' doubling time was 340 to 7134 days (mean, 2196; median, 1593), with 9 nodules that decreased in size. Malignant nodules' doubling time was 451 to 17182 days (mean, 3940; median, 2137), with 7 nodules that decreased in size. Using a threshold of 1100 days, sensitivity and specificity of doubling time to predict malignancy are 19.0% and 86.7%, respectively. Area under the receiver operating characteristic curve is 0.39. Follow-up imaging resulted in 6 additional benign and 0 additional malignant nodules meeting criteria. There is no significant difference in benign or malignant nodules' doubling times, and a decrease in nodule size is nonspecific. These findings challenge the underlying rationale for routine imaging follow-up of thyroid nodules.