THYROID NODULE GROWTH AS A PREDICTOR OF MALIGNANCY.

To assess which measure of thyroid nodule growth on serial neck ultrasound, if any, is associated with malignancy. Retrospective exploratory chart review of malignant thyroid nodules assessed at Kingston Health Sciences Centre (2006-2016) and benign thyroid nodules (2016), at least 1 cm in diameter and with 2 ultrasounds completed at least 30 days apart. Groups were compared using independent samples Student's test, chi-square test, or Mann-Whitney test as appropriate, as well as multivariable logistic and linear regression modelling to adjust for age and baseline volume. One hundred and seventy-eight nodules were included in the study. When growth was defined as >20% increase in 2 dimensions (minimum 2 mm), malignant nodules (MNs) underwent significantly more growth than benign nodules (BNs) (16.8% BN versus 29.8% MN [ = .026]; odds ratio = 2.49; 95% confidence interval = 1.12 to 5.56). There was no significant difference between the groups when growth was defined as >2 mm/year or ≥50% volume growth. Nodules shrank >2 mm/year in each group and the difference was not statistically significant (24.2% BN versus 20.7% MN [ = .449]). The median doubling time for the nodules that grew was 1022.1 days in the BN group and 463.2 days in the MN group ( = .036). The median doubling time for all nodules was 456.5 days in the BN group and 244.2 days in the MN group ( = .015). Thyroid nodule growth defined as >20% increase in 2 dimensions (minimum 2 mm) is associated with risk of malignancy. Nodule shrinkage did not distinguish between BNs and MNs. = benign nodule; = confidence interval; = fine needle aspiration; = Kingston Health Science Centre; = malignant nodule; = odds ratio; = receiver operating characteristic.