Risk stratification of neck lesions detected sonographically during the follow-up of differentiated thyroid cancer

Context: The European Thyroid Association (ETA) has classified post-treatment cervical ultrasound findings in thyroid cancer patients based on their association with disease persistence/recurrence. Objective: To assess this classification's ability to predict the growth and persistence of such lesions during active post-treatment surveillance of patients with differentiated thyroid cancer (DTC). Design: Retrospective, observational study Setting: Thyroid cancer center, large Italian teaching hospital. Patients: Center referrals (2005–2014) were reviewed and patients selected with pathologically confirmed DTC; total thyroidectomy, with or without neck dissection and/or radioiodine remnant ablation; abnormal findings on ≥2 consecutive post-treatment neck sonograms; subsequent follow-up consisting of active surveillance. Baseline ultrasound abnormalities (thyroid bed masses, lymph nodes) were classified according to the ETA system. Patients were divided into group S (those with ≥1 lesion classified as ‘suspicious’) and group I (‘indeterminate’ lesions only). We recorded baseline and follow-up clinical data through 30 June 2015. Main Outcomes: Patients with growth (> 3 mm, largest diameter) of ≥1 lesion during follow-up, patients with ≥1 persistent lesion at the final visit. Results: The cohort included 58 (9%) of the 637 DTC cases screened. A total of 113 lesions were followed (18 thyroid bed masses, 95 lymph nodes). During surveillance (median 3.7 years), group I had significantly lower rates than group S of lesion growth (8% vs. 36%, p=0.01) and persistence (64% vs. 97%, p=0.014). Median time to scan normalization: 2.9 years. Conclusions: The ETA's evidence-based classification of sonographically detected neck abnormalities can help identify PTC patients eligible for more relaxed follow-up.