The main finding of this randomized trial was that local control with MWA and RFA is similar for medium-sized liver tumors (1.5 to 4 cm in diameter) with no difference in oncological outcomes in either group. Although the definition of medium-sized tumors differs among authors 31, it is clear that local tumor progression in tumors under 2 cm, even with classical RFA devices, is less than 1% 37,38, while liver tumors over 4 or 5 cm may lead to an unacceptable rate of local tumor progression, sometimes over 60% in classical references 37,39. In any case, these medium-sized tumors are the everyday challenge of any physician performing thermal ablations, which is why we focused our study on them.
Several retrospective 6,30,40,41 and randomized clinical trials 19,22,24,42 compared microwave and radiofrequency ablation systems dealing with primary or secondary tumors or both. Some of these, however, were carried out with first line thermoablative systems, as in the RCT by Shibata et al 42, which treated patients with HCC nodules smaller than 4 cm in diameter and found equivalent results between both techniques with a 24% overall proportion of local tumor progression in the MWA Group. The RCT conducted by Abdelaziz et al 22 could have been limited by its short follow-up (not specified) and a high lost to follow-up (52%) 43. The RCT recently conducted by Vietti Violi et al 24 compared MWA and RFA in treating HCC in up to three lesions of 4 cm or less should be highlighted for its well established design. In this last study the authors observed 6% and 12% LTP at 2 years for MWA and RFA with no significant differences between groups (median follow-up of 26 months). However, the lesion size in this study was smaller than in our study (1.8 cm vs. 2.4 cm) due to the different inclusion criteria. In the present study, after excluding tumors with less than 1.5 cm with a mean follow-up of 25 months (complete duration of the study: 58 months and no patients lost to follow-up), we observed 21% and 12% LTP at 2 years for MWA and RFA, respectively, also with no differences between groups. It also gave an insight into the long-term outcomes not only in HCC but also in liver metastases with an inherently different natural history.
MWA was initially considered to be less safe than RFA because the ablation zone shape was considered less predictable 17,44,45. For example, Van Tilborg et al 44, in a retrospective study with 774 colorectal metastases treated by MWA or RFA in 243 patients, observed that biliary complications (biloma/biliary leakage, biliary obstruction and bilio-pleural fistula) were especially common after peribiliary-MWA (57%) vs. peribiliary-RFA (3%) and these significant differences did not decrease with operator experience. However, in a multicenter Italian study based on 1037 MWA in 736 patients, major complications were found in only 2.9% of the patients -similar to historical RFA procedures- 20. Nor were there any significant differences in device-related complications found in any of the available RCTs comparing MWA and RFA 19,22,24,42. In our study major complications (grade 3–4) were found in 5 cases in the MWA Group (11%) (including one delayed bilio-bronchial fistula and one delayed biliary fistula) vs. 2 cases in the RFA Group (4%), with no significant differences between the groups and no procedure-related deaths reported in either group.
Even though both MWA and RFA destroy tissue via thermally coagulative necrosis, they are very different because of their inherently different energy deposition mechanisms, which may explain their many differences in terms of ablation zones 10,14. Many improvements have been seen in recent years in thermal deposition during ablation, particularly in MWA, which makes it difficult to compare different devices based on the same technology 14. Even with the latest generation of devices, ablation zone shapes created by MWA tend to be larger, more elongated and maybe less affected by the heat-sink effect 10, 16–18,33, but usually have less control of heat propagation than RFA, possibly because MWA is highly dependent on tissue properties (water content) 10,14,15. In our study, after a careful 3D reconstruction of each ablation zone, we confirmed that MWA created larger ablation zones mainly because of their more elongated shape (greater long diameter) but were not significantly less predictable taking sphericity ratio, coefficient of variability and surface area into account.
This study had several limitations: first, the trial was performed in a single third referral center with a special dedication to these procedures and a special focus on medium-sized tumors, which are our daily challenge. Although our results are in line with previous RCT results, it is conceivable that some results may differ with different patient inclusion criteria. Second, to reduce the risk of device-related bias we chose to use only one advanced device in each group, even though there are numerous commercially available devices with similar characteristics. In any case, we have to acknowledge that accurate comparisons between MWA and RFA systems are difficult. Third, as in other RCT in this field 21,24, per protocol analysis was performed to correctly evaluate the true efficacy of each technology (particularly for evaluating ablation zone shape) even though this could be deemed a sub-optimal approach in general terms. Fourth, although a single ablation per tumor is preferable to achieve a reproducible evaluation of ablation zone shape, we performed a mean of more than two ablations per tumor in a pragmatic approach. Although no differences were found between the groups in this last variable and there were no significant differences in coagulation shape among those groups in single ablation per tumor cases (results not shown), some overlapping ablations could mask slight differences in coagulation shape.