Sung initially reported a superiority of PEI vs. RFA in cystic nodules, but in their first experience, internal fluid was not aspirated before RFA [41]. In a more recent prospective trial of the same authors, at six-month follow-up volume reduction was 96.9 % in the PEI group and 93.3 % in the RFA group [42].

PEI is less effective for treating thyroid nodules with a larger solid component [18], so further thermal ablation could be required to achieve significant shrinkage [43, 44].

In summary, no clear superiority of a technique over the other in cystic nodules has been demonstrated, but PEI is by far less expensive than thermal ablation; therefore it is definitely preferable. Only large mixed nodules with a considerable solid portion could be sequentially treated by PEI and thermal ablation to obtain a more complete volume reduction.

In summary, the effects of thermal ablation generally last over time, and a regrowth of the treated nodules occurs in only 5–10 % of cases. The regrowing nodules can be further treated by thermal ablation or surgically; in few of our patients who were operated on because of a poor response to RFA, no scars were found in the extranodular tissue which could impair the outcome of the surgery. This clinical observation is in agreement with the findings of Holmer et al. who in experimental animals showed a clear demarcation between areas of radiofrequency induced necrosis and the surrounding vital tissue [28].

Hu HY enrolled 30 patients with a nodule volume of about 13 mL, randomly assigned to receive one or two treatments; at six months volume reduction was 70.2 +/− 13.2 % in the group assigned to one treatment, while in the patients who received two treatment sessions one month apart, the volume reduction was 78.3 %. Thus, a second treatment produced a slight but not significant further benefit [46]. We can speculate that the second session was performed too early after the first treatment.

In a series of 77 patients with very large nodules (mean volume about 55 ml) treated by two or three LTA cycles (each consisting of three sessions 1 month apart), volume reduction was 43 % after one cycle, 72 % after two cycles, and 81 % after three cycles [48].

In summary, a second or third procedure is able to obtain a further volume reduction of 15–20 %, which could lead to a global reduction of about 80 % or more; in our opinion more treatment sessions should be considered for larger nodules, when a significant portion of untreated tissue remains after the first procedure and/or when the patient complains of incomplete resolution of the symptoms. However, in these cases, a careful analysis of the cost–effectiveness should be performed, taking into account the cost of the many devices required, the clinical condition of the patients, and the aim of the treatment.

As regards the timing of a second treatment, up until now no studies have specifically investigated this issue, but, as the maximum volume reduction is obtained between 6 and 12 months after the first thermal ablation, it would be reasonable to plan it after this period of follow–up.

Amabile et al. treated large (>40 mL) hyperfunctioning nodules by 1 to 3 cycles of LTA (each cycle consisted of three sessions performed at an interval of 1 month); normal serum TSH was obtained in 87 % of the patients after three cycles of LTA [48].

Some evidence suggests that the efficacy of RFA in hot nodules depends on the extent of hyperfunction, pretoxic nodules showing a normalization of their function more frequently than toxic nodules [26].

In summary, thermal ablation should not be the first choice for the treatment of hyperfunctioning thyroid nodules, due to the difficulty of obtaining complete ablation of the whole nodular tissue, so the peripheral untreated areas could maintain hyperactivity [53]. On the other hand, radioiodine treatment is more effective in nodules < 4 cm and is by far less expensive, while surgery is the best choice for very large nodules following a course of antithyroid drugs to normalize thyroid function, providing there are no serious contraindications to surgery. We argue that thermoablation may be considered in two opposite conditions, the first being small autonomous nodules in young patients who refuse surgery or radioiodine (though, this is not a frequent condition), because a complete ablation can be achieved, and the second being very large nodules in patients at high surgical risk, in combination with radioiodine.

In the series with the longest follow-up, no malignancies are reported [29, 37, 38, 54, 57]; in our experience no cases of malignancy were found, and in those few patients who underwent subsequent surgery due to poor response to thermal ablation, the histological examination confirmed the benignancy.

If we examine this problem from the opposite point of view, we cannot ignore that in patients undergoing surgery for benign thyroid nodules, microcarcinomas are occasionally found on histological examination. This aspect is highlighted in the study of Bernardi et al., who compared RFA and surgery; in their series of 74 patients operated on, 8 microcarcinomas were found on histology [55]. However, this percentage seems quite elevated; we believe that there is a strong bias related to the retrospective analysis.

In summary, the risk of malignancies is almost inexistent, providing that an accurate cytological evaluation has been carried out before treatment, possibly by performing two FNABs. Furthermore, we can speculate that small foci of cancer which are occasionally present inside the nodule could be ablated by heat.

Both our group and Baek et al. did not report any case of hyperthyroidism in patients with nodules treated by RFA, neither cold nor autonomous [25, 27, 52].

A transient TSH decrease and fT4 increase were seen by Valcavi et al. at day 1 and week 1, respectively, after LTA; they returned to baseline levels within one month. Serum fT3 did not change. Two patients developed overt hyperthyroidism one month after the procedure, associated with a transient TRAb peak, and were treated with methimazole and 131I, respectively [38]. No changes in thyroid function or antithyroid antibodies were reported in the recent trial of Papini et al. [37].

This issue has been accurately and extensively addressed by Bernardi et al., who demonstrated that a single RFA session is cheaper than hemithyroidectomy (1,661.50 euros vs. 4,556.30) [55]. According to the analysis performed by the aforementioned authors, RFA could still be advantageous in patients with very large nodules who require further treatment sessions; indeed, we have to note that such patients are generally poor candidates for surgery due to their age or clinical conditions. Therefore, we should also take into account this aspect in the overall evaluation of the cost/benefit ratio in these cases.

The same considerations could be applied to the LTA, bearing in mind that the cost of the latter procedure is dependent on the number of fibers which are used in any single session (about 300–400 euros/fiber). We can argue that the cost of LTA is inferior not only to that of the surgery but even to that of radiofrequency when small nodules are treated by one or two fibers, while it could be similar to or greater than that of RFA when three or four fibers are required to treat large nodules. These aspects should be taken into consideration in the choice of the technique to be used for individual patients in centers, such as our institution, where both laser and radiofrequency are adopted.

LTA is reported to be able to induce a reduction of the volume of benign thyroid nodules ranging between 36 and 60 %, depending on baseline volume, number of treatment sessions, and the morphology of the nodules [33–37, 65].

As well as LTA, the volume reduction obtained by RFA varies depending on the devices used, the baseline volume, and the other characteristics of the nodule. The reported reduction ranges between 50 and 84 % [25, 27, 30, 31, 40] and thyroid function is not compromised, even in patients who had previously undergone lobectomy [6].