Benign Symptomatic Thyroid Nodules—A Place for a Randomised Trial of Sham Procedure Versus Radiofrequency Ablation?

Thyroid nodules are frequently an incidental finding on imaging, being present in up to 68% of adults on ultrasonography, while being palpable in only 3%–7% of the population [1, 2]. The number of Medicare items processed for ultrasonography of the neck in Australia has increased tenfold over the last three decades, from 42 691 services in the financial year 1994–1995 to 425 565 services in 2023–2024 [3]. Radiofrequency ablation (RFA) for benign non-functioning thyroid nodules was first described in 2006 [4]. The procedure subsequently saw wide adoption in Asia and Europe, though its use has been more gradual in the United States (US). Only a small number of practitioners are performing RFA in Australia currently, and to date there is no Medicare item reimbursement for its use. Society guidelines and consensus statements are consistent in recommending RFA only be performed for nodules causing local pressure symptoms or cosmetic concerns with benign cytology on at least two fine-needle aspirations. The minimum size for consideration varies significantly between guidelines, ranging from at least 20 to 30 mm in maximal diameter with continuous growth, to at least 20 mL in volume [5, 6]. The predominant symptomatic indication for RFA of benign thyroid nodules in studies has been dysphagia. Dysphagia, however, is commonly reported, surveys in Australia and the US each finding a prevalence of 16% in the general adult population [7, 8]. Ninety-two percent of the US study who reported dysphagia experienced symptoms within the previous week, 16.3% of whom described their symptoms as ‘quite a bit’ or ‘very severe’ [7]. There is a paucity of surgical literature regarding symptoms due to thyroid nodules. Eng et al. [9] found the mean largest nodule size in patients with compressive symptoms to be 38 versus 22 mm in asymptomatic patients. Of those who underwent surgery for compressive symptoms, 92.7% had improvement in their symptoms postoperatively. Most patients in this study, however, had multinodular goitre, with only a very small proportion having solitary benign nodules. It is not clear from the study what proportion of the patients were evaluated because of symptoms, compared with those who were identified because of the incidental finding of thyroid nodules on ultrasound, and asked about symptoms after they were aware of the finding of nodular disease. Pharmaceutical medications are routinely subjected to randomised double-blind placebo-controlled studies regarding efficacy and adverse effects. Sham-controlled trials of surgery and other procedures are rare due to ethical and practical concerns. Many sources of bias in trials of surgery and procedures exist. In particular, the Hawthorne effect describes changes in patient behaviour related to participation in the study rather than the intervention itself [10]. Notably, of 15 sham-controlled studies of procedures and surgical interventions, none demonstrated a benefit of the actual intervention over the sham intervention [10]. RFA is effective in reducing thyroid nodule volume and 88% report an improvement in compressive symptoms where these were present pre-procedure [4]. More than one treatment is commonly required for nodules with an initial volume of at least 20 mL. The rate of early complications is low—approximately 4.5%. Long-term complications include the risk of regrowth and subsequent detection of thyroid cancer. In a study of 421 patients treated with RFA for benign thyroid nodules with a median follow-up period of 90 months post-procedure, regrowth was noted in 12% of nodules [11]. The median time to regrowth was 4 years, with a mean initial volume of 29 ± 27 mL. Thyroid surgery was performed in 26 patients (6%), with thyroid cancer in 21 patients, six of whom showed metastatic disease. Ten of the 21 newly diagnosed thyroid cancers arose within ablated nodules. as made regarding surgical and pathological challenges post-RFA procedures. A study of subsequent thyroidectomy in the setting of previous thermal ablation compared with thyroidectomy in a non-ablation group as controls found the previous ablation group had longer operative times (99.5 vs. 69.5 min), higher difficulty scale, more severe macroscopic adhesion scores, and ablation-induced tissue alterations that complicated histologic evaluation [12]. A final consideration is the costs of RFA. The listed RFA probe cost in Australia is $2200. Individuals undergoing RFA require at least two fine-needle aspiration samples pre-procedure, many will require more than one RFA procedure, as well as lifelong ultrasound examinations to monitor nodule regrowth and the potential development of thyroid malignancy. Further, the costs of the radiofrequency generator and training practitioners warrant consideration. A prospective, single-blind, randomised, controlled trial of sham versus actual RFA may provide evidence as to the true symptomatic benefit of the intervention. Inclusion criteria would be non-pregnant individuals of age at least 18 years of age with symptomatic thyroid nodules at least 20 mm in diameter, proven to be benign on fine-needle aspiration cytology on at least two occasions, who did not have an allergy to local anaesthetic, and who provided signed informed consent. Individuals would be randomly assigned in a 1:1 ratio to undergo trans-isthmus moving-shot sham treatment or RFA, using identical procedures and equipment under local anaesthesia. It would be essential to emphasise both verbally and in written instructions to participants that the procedure is pain-free in 97%–98% of individuals. Additionally, the radiofrequency generator and needle would have to be modified so that the popping/crackling noises associated with RFA ablation occurred at a similar frequency and time with sham treatment. Follow-up would be for 6 months post-procedure. Primary efficacy endpoints would be participant-reported outcomes measurement for thyroid disease compression score performed at baseline, and 2-weeks, 6-weeks, 3-months and 6-months post-intervention. Participants would be instructed not to undergo thyroid imaging until nodule size was measured by ultrasound following the outcomes measurement at the 6-month follow-up visit. All randomised participants would be asked to guess their treatment assignments prior to discharge from the treatment centre and at the 6-month follow-up visit, and a blinding index method calculated. Barriers to such a sham-controlled study would include ethical concerns as to whether the risk of sham control justifies the value of the knowledge to be gained, whether the administration of sham control is adequately disclosed and accepted by the participant during the informed consent process, costs of RFA electrodes, and the demeanour of the interventional radiologist being unchanged for participants receiving sham-control or RFA treatments. In addition to providing information regarding the symptomatic benefit of RFA, a sham-controlled trial may additionally inform as to whether RFA treatment justifies the costs, requirement for long-term monitoring, and risks of nodule regrowth and thyroid cancer. Adam Morton: conceptualization, writing – original draft, writing – review and editing. Open access publishing facilitated by The University of Queensland, as part of the Wiley - The University of Queensland agreement via the Council of Australian University Librarians.