Long- versus short-interval follow-up of cytologically benign thyroid nodules: a prospective cohort study

Thyroid nodules are common, and most are benign. Given the risk of false-negative cytology (i.e. malignancy), follow-up is recommended after 1-2 years, though this recommendation is based solely on expert opinion. Sonographic appearance may assist with planning, but is limited by large inter-observer variability. We therefore compared the safety and efficacy of long- versus short-interval follow-up after a benign initial aspiration, regardless of sonographic appearance.

This study evaluated all patients referred to the Brigham and Women's Hospital Thyroid Nodule Clinic, between 1999 and 2010, with a cytologically benign nodule >1 cm and who had returned for follow-up sonographic evaluation. Despite standard clinical recommendations, variation in patient compliance resulted in variable follow-up intervals from time of initial aspiration to the first repeat evaluation. Main outcome measures included nodule growth, repeat fine needle aspiration (FNA), thyroidectomy, malignancy, and disease-specific mortality.

We evaluated 1,254 patients with 1,819 cytologically benign nodules, with a median time to first follow-up of 1.4 years (range, 0.5-14.1 years). The longer the follow-up interval, the more nodules grew and the more repeat FNAs were performed (P <0.001). The most clinical meaningful endpoints of malignancy or mortality, however, did not differ between the various follow-up intervals. The risk of a thyroidectomy (usually because of compressive symptoms) increased when time to first follow-up exceeded >3 years (4.9% vs. 1.2%, P = 0.0001), though no difference in malignancy risk was identified (0.2-0.8%, P = 0.77). No (0%) thyroid cancer-specific deaths were identified in either cohort.

While expert opinion currently recommends repeat evaluation of a cytologically benign nodule at 1-2 years, these are the first data to demonstrate that this interval can be safely extended to 3 years without increased mortality or patient harm. Nodule growth can be expected, though detection of malignancies is unchanged. While replication of these data in large prospective multicenter studies is needed, this extension in follow-up interval would reduce unnecessary visits and medical interventions for millions of affected patients worldwide, leading to healthcare savings. Please see related commentary article: http://dx.doi.org/10.1186/s12916-016-0559-9 and research article: http://dx.doi.org/10.1186/s12916-015-0419-z .