Ethanol ablation uses a high concentration (95–99%) of ethanol. The skin is punctured with a 16–21-gauge needle, and the target nodule is approached through the thyroid isthmus to prevent ethanol leakage during the procedure. The method of ethanol instillation depends on the solid component of the nodule (1516343536). For a cystic or predominantly cystic nodule, the tip of the needle is inserted into the center of the cyst, as much fluid as possible is aspirated, and then ethanol is injected (16). If the cyst contents are viscous, the fluid is aspirated using a large-bore needle, followed by irrigation with normal saline to remove internal debris and colloid material before ethanol instillation. The volume of ethanol injected is usually 50% of the aspirated fluid volume. After 2 minutes of ethanol retention with the needle in place, the injected ethanol is removed completely and the needle is withdrawn. If a predominantly solid nodule contains a cystic component, the latter is punctured, almost completely aspirated, and an appropriate amount of ethanol is instilled (37). The tip of the needle is subsequently inserted into the solid component of the target nodule, followed by infusion of an appropriate volume of ethanol. That volume is based on nodule size and echogenicity of the solid portion. In treating a purely solid nodule, ethanol is injected directly into the nodule (3637). Adequate coverage of the target nodule, as indicated by its echogenicity (called intranodular echo-staining), is achieved by adjusting the injection of ethanol under US guidance. After injection, the needle is rapidly withdrawn.

Radiofrequency ablation is performed using a generator and internally cooled 7-cm electrodes with an 18-gauge active tip measuring 0.5, 0.7, or 1 cm (257). The RF power and the size of the active tip chosen depend on the size and internal characteristics of the targeted nodule. The initial RF power is usually 30–50 W with a 1-cm active tip, but may be 10 W with a 0.5-cm tip or 20 W with a 0.7-cm tip. If a transient hyperechoic zone does not appear at the tip of the electrode within 5–10 seconds, the RF power on a 1 cm tip is increased at 10 W increments to a maximum of 80 W (30 W with a 0.5 cm tip and 50 W with a 0.7 cm tip). Using a trans-isthmic approach, the electrode is inserted from the isthmus to the lateral aspect of a target nodule. The entire length of the electrode should be visualized to minimize possible complications. Benign thyroid nodules can be treated using the moving-shot technique (23839). Given that most nodules are usually ellipsoid in shape, there is little margin between the nodule and the normal thyroid parenchyma. Therefore, a fixed electrode technique, which creates a round ablation zone, is considered unsuitable. For this procedure, the target nodule is divided into multiple small conceptual ablation units and RFA is performed unit-by-unit by moving the electrode. The electrode tip is initially positioned in the deepest and most lateral portion of the nodule, after which it is moved backward to the most superficial and most medial portion, thereby preventing visual disturbances caused by echogenic bubbles. RFA is terminated when all conceptual units of the targeted nodule have been transformed into transient hyperechoic zones.

Ethanol ablation has been recommended as the first-line treatment for benign cystic thyroid nodules (12131516). Retrospective and prospective studies show that the mean reduction in nodule volume after EA for thyroid nodules with large > 90% cystic portions (> 90% of the lesion) ranges from 85 to 98.5% (12131516). A prospective study showed that EA was superior to RFA in volume reduction and cost-effectiveness (16). EA has also been used to treat predominantly cystic thyroid nodules–lesions that are 50–90% cystic. However, after a single EA session, the reduction in volume of these nodules was significantly less than that of the > 90% cystic nodules (78.2% vs. 89.7%) (1840). Moreover, the long-term recurrence rate of predominantly cystic nodules is as high as 38.3%, necessitating additional treatment sessions. Reportedly, 5–25% of predominantly cystic thyroid nodules are refractory to EA (13414243). Factors influencing the efficacy of EA are the solid component, initial nodule volume > 10 mL, vascularity of the solid component, and volume of injected ethanol (131435444546). Solid components are thought to be more resistant to ethanol, a factor that is independently associated with less volume reduction after EA (46). Greater vascularity of the solid component favors drainage of ethanol. Therefore, the therapeutic efficacy of EA is limited (144546).

After EA for a predominantly cystic thyroid nodule, patients whose symptoms are incompletely resolved or who have residual unablated areas with internal vascularity on US may benefit from an additional session of EA or RFA (Table 1). Few studies have assessed the efficacy of repeat EA. Two groups reported efficacies of 20–33.3% after the second session and 10–25% after the third session; thus, the efficacy markedly decreases as the number of sessions increases (4147). By contrast, an additional session of RFA is effective and safe for treating incompletely resolved nodules after initial EA treatment (4546). Additional RFA resulted in a mean reduction in nodule volume of 92% and significant improvements in symptoms and cosmetic appearance after 6 months, with no major complications (e.g., voice change, infection, or esophageal or tracheal injury) (45). These findings were prospectively verified, showing that patients with recurrent or residual thyroid nodules were effectively treated by additional RFA (46). The 1-month recurrence rate after a single EA session was 33%, with a larger solid component in predominantly cystic nodules being an independent factor for recurrence (46). These findings indicate that EA followed by RFA is a safe and effective combination, as shown by volume reduction and resolution of symptoms, especially in treating predominantly cystic nodules with a substantial solid component (Table 1, Fig. 1).

Radiofrequency ablation is an effective alternative to surgery for treating benign solid thyroid nodules (34671718192127384849). The technical success rate is high (100%), and mean reduction in nodule volume is reportedly 33–58% at 1 month and 51–85% at 6 months (123385051). One study showed a mean reduction in nodule volume of 93.4% after 4 years. These findings indicate that RFA is effective in both reducing nodule volume immediately and producing a durable long-term response (6). A single session of RFA is usually sufficient for volume reduction, resolving symptoms, for cosmesis. However, additional RFA sessions may be required for large (> 20 mL) nodules and for those located close to critical structures (e.g., recurrent laryngeal nerve, trachea, and esophagus) (5253) (Table 1). It is technically difficult to completely ablate entire large nodules during a single RFA session. Following shrinkage of previously ablated areas, additional RFA sessions can be performed safely and easily, thus making complete ablation possible. However, even for small nodules located close to critical structures, peripheral residual unablated areas are likely to remain, especially when RFA is performed by less experienced operators. Complete ablation is important for preventing recurrence. Nodules undertreated during initial RFA have shown regrowth at their peripheral portions (6), suggesting the need for additional treatment of residual unablated nodules seen on follow-up US.

Ethanol ablation may be an alternative to additional RFA in such cases, as it cannot thermally damage adjacent structures (37). When performed by a less experienced operator (< 20 RFAs/year), the rate of success (defined as the absence of vascular solid components on follow-up US) of a single RFA session was 55.6%, whereas the success rate after additional EA for residual unablated areas was 62.5% (37). As the efficacy of EA is mainly affected by vascularity and the size of the solid component, EA is recommended in patients after incomplete RFA when the remaining component has a volume of ≤ 5 mL and is not highly vascular (13143537, 444546). Possible serious complications of EA include leakage of ethanol, causing direct damage to adjacent nerves or critical structures. However, precise US-guided injection and meticulous handling of the injection needle may reduce or prevent complications (Table 1). In a preliminary study, no serious complications after EA were observed, the only adverse effects being transient neck pain and diffuse glandular hemorrhage (37). These findings indicate that RFA followed by EA can be performed safely by personnel less experienced with RFA. This combination may be effective in treating peripheral residual unablated areas caused by technical difficulties or patient discomfort.

Although RFA is successful at treating solid tumors in many organs, a large percentage are difficult to treat or recur locally (2930313354555657585960). For example, RFA has had limited success in treating hepatic tumors > 3 cm and those located near large vessels or the hilum. The region of coagulation necrosis is limited or there may be technical difficulties in optimal positioning of the electrode (2930313354555859,60). The effectiveness of RFA is also limited by the presence of nearby vessels ≥ 3 mm in diameter, which dissipate heat away from target tissues, a phenomenon called the heat-sink effect (293031335455585960). The therapeutic effect of RFA may be enhanced by combining it with transcatheter arterial chemoembolization or saline injection. More recently, ethanol injection prior to RFA was introduced for the treatment of hepatocellular carcinomas (3233565960). Compared with RFA alone, this method results in significantly larger areas of coagulation necrosis and a significantly lower energy requirement for coagulation per unit volume of tumor ethanol injection may enhance the therapeutic effects of RFA by destroying tumor vessels and enhancing thermal conduction in tumor tissues, thus reducing the heat-sink effect. Moreover, ethanol heated by RFA may extend the area of tissue necrosis (5561).

In contrast to lesions in a large organ such as the liver, complete ablation of a nodule in the relatively small thyroid gland does not require a large area of coagulation necrosis. Rather, ethanol may be instilled to control internal bleeding in predominantly cystic thyroid nodules when aspiration is performed prior to RFA (Table 1). Because decompressed nodules are easier to treat, it is recommended that internal fluid be aspirated before RFA (1516626364). However, active bleeding can develop during aspiration due to vessel injury by the aspiration needle or venous bleeding in response to reduced internal pressure (65). This bleeding enlarges the nodule volume, causes a heat-sink effect, and limits the therapeutic effect of RFA. If bleeding is caused by needle injury to feeding arteries in and around the nodule, the bleeding site is usually easily identified and hemostasis can be achieved by RFA (65). However, if venous bleeding from the nodule occurs during aspiration, the bleeding site cannot be identified. Ethanol instillation may then help in controlling internal venous oozing by inducing coagulation necrosis of the cystic wall and thrombosis of microvessels (404555) (Table 1). This method has been reported to reduce the total number of treatment sessions to a mean of 1.2 (65). Therefore, concomitant ethanol injection during RFA can effectively control active bleeding that occurs when internal fluid is aspirated, as well as reduce the number of sessions of RFA (Fig. 2).