DOI: 10.1210/clinem/dgaf058 ISSN: 0021-972X

Radiofrequency ablation for thyroid nodules (RATED study) - analysis of a learning curve and predictors of success

Manon M D van der Meeren, Tim Boers, Pim de Graaf, Katya M Duvivier, Koen M A Dreijerink, Laura N Deden, Peter Veendrick, Paul Cernohorsky, Frank B M Joosten, Angelique B M C Savelberg, Sicco J Braak, Sean H P P Roerink, Michel Versluis, Srirang Manohar, Wim J G Oyen

Abstract

Context

Radiofrequency ablation (RFA) is used as treatment for symptomatic thyroid nodules. Factors influencing the volume reduction ratio (VRR) at 12 months are not yet fully understood.

Objective

The primary objective was evaluating the VRR at 12 months after RFA. Secondary objectives were the assessment of a learning curve and factors influencing the VRR at 12 months.

Design

A retrospective observational cohort study.

Setting

Three Dutch referral hospitals.

Patients and intervention

Patients who underwent RFA for symptomatic thyroid nodules with available ultrasound follow-up.

Main outcome measures

Ultrasound based VRR at 12 months and chronologically numbered RFA procedures. All patients’ baseline, treatment, and early follow-up factors were assessed for correlation with VRR at 12 months.

Results

A total of 337 patients with 356 nodules were included in the learning curve analysis. VRR at 12 months increases for the first 20 treatments per center and stabilizes thereafter, indicating a plateau phase after a learning curve. These initial cases were removed from further analysis. In the remaining 299 nodules, median VRR at 3, 6 and 12 months was 57.1, 65.6 and 70.8%. Baseline nodule volume negatively correlated with VRR at 12 months but VRR was high for every volume category. Energy delivered per volume did not correlate to VRR.

Conclusions

In RFA for thyroid nodules a stable treatment efficacy is achieved after 20 treatments, with a median VRR of 70.8%. Baseline nodule volume, energy delivered and prolonged follow-up 6 months after treatment may not be clinically relevant to predict treatment success.

More from our Archive