Microwave ablation compared with radiofrequency ablation for the treatment of liver cancer: a systematic review and meta-analysis

Over the past 30 years, several ablative methods have been developed for the treatment of hepatic cancer as an alternative to surgical resection and liver transplantation in patients with unresectable cancer or in selected patients with resectable disease. Recent guidelines recommend radiofrequency ablation (RFA) and microwave ablation (MWA) as the ablative methods with the highest efficacy in the treatment of liver cancer.1, 2

Tumor location near to the main biliary tree, abdominal organs, or diaphragm is a relative contraindication for RFA because of the risk of severe complications.2 RFA is prone to heat-sink effect, which reduces further the efficacy of the treatment.1, 2 On the other hand, MWA, which is a more recent thermal ablation technique, is associated with higher intratumoral temperatures, resulting in faster, larger, and more homogenous ablation compared to RFA.2 Furthermore, MWA is less prone to heat-sink effect and can be utilized in tumors adjacent to vessels.

According to the European Association for the Study of the Liver (EASL) recommendations, MWA showed promising results for local tumor control in patients with hepatocellular cancer (HCC).1 The guidelines by the American Association for the Study of Liver Diseases (AASLD) reported that MWA has potential advantages over RFA; however, further studies are required to provide safety and efficacy data.2 The Cochrane meta-analysis conducted in 2013 failed to provide evidence regarding the role of ablative methods in the treatment of HCC since only one randomized clinical trial (RCT) with high risk of bias was available.3 The last conducted meta-analysis in 2019 reported beneficial outcomes in favor of MWA.4 However, low quality randomized and observational studies, which were affected by confounding bias were included in this meta-analysis, which could influence the reliability of the outcomes.

Despite the promising results of MWA in the treatment of liver cancer, efficacy and safety of MWA compared to RFA is unclear. Aim of this meta-analysis is to compare RFA and MWA in the treatment of HCC and liver metastases. Our hypothesis is that the beneficial characteristics of MWA are translated into better oncological outcomes compared to RFA.

RFA is currently one of the most widely used thermal ablation modalities. On the other hand, utilization of MWA has been increased the last years as a result of significant advancements in technology of new generation devices. These advancements are translated into higher temperatures and faster heating compared to RFA, large ablation volumes, and less heat sink effect.29 However, MWA has not been adequately compared with RFA and selection of appropriate treatment is not based on high level of evidence.30 On the basis of these considerations, we conducted the present meta-analysis to evaluate the role of MWA in the treatment of liver cancer.

Meta-analysis of CA rates, which included more than 2,500 tumor lesions, demonstrated no significant differences between MWA and RFA. In the subgroup analysis of RCTs with 438 tumors, similar rates of CA were found between the two methods. Analysis of all included studies revealed no significant difference in LTP between MWA and

RFA. Since increased heterogeneity was detected among the studies, subgroup analysis of RCTs was conducted to decrease heterogeneity and to evaluate the influence of observational studies on the outcomes. The RCTs by Abdelaziz et al. and Vietti Violi et al. included 255 patients with HCC and up to three lesions with less than 5 cm and 4 cm tumor size, respectively.15,16 Furthermore, new generation MWA devices with 2,450 MHz generators were utilized. Meta-analysis of the two RCTs demonstrated statistically decreased rates of LTP in the MWA group. Specifically, LTP was reported in 5.2% and 12.2% of tumor lesions treated with MWA and RFA, respectively.

The finding of the RCTs is consistent with the physics and characteristics of radiofrequency and microwave energies. MWA is associated with higher temperatures, faster heating, larger ablation volumes, and less heat sink effect compared to RFA, which are translated into better oncological outcomes in terms of LTP in the present meta-analysis. On the other hand, meta-analysis of retrospective studies failed to demonstrate superiority of MWA over RFA, which is attributed to the significant inter-study heterogeneity.

Consequently, though CA was comparable between the two procedures, LTP was beneficial in favor of MWA. These conflicting results are not surprising given the limitations associated with measurement and evaluation of complete ablation response. Imaging modalities cannot detect with 100% accuracy whether neoplastic cells have been sufficiently ablated. For that reason, ablation response cannot be considered as the most reliable indicator of treatment effectiveness. On the other hand, follow-up imaging examinations and LTP have been considered of great importance in detecting treatment failure. LTP is the most reliable indicator of treatment effectiveness and can be utilized as assessment tool of treatment efficacy.

IDR was comparable between the two ablative methods. Subgroup analysis of two RCTs demonstrated similar rates of IDR between MWA and RFA. The RCT by Kamal et al. reported IDR rates of 18.2% at 12-month follow-up14, while the survey by Abdelaziz et al. reported rates between 13.6% and 22.22% at 27-month follow-up.16 The beneficial outcomes in LTP were not associated with a decreased incidence of intrahepatic recurrence in the MWA group. This result is attributed to a variety of factors, which are associated with cancer disease, underlying liver disease, and indications of treatment. Patients were oft assigned to treatment based on tumor proximity to blood vessels or biliary tract. These tumors are characterized by increased incidence of local metastases, which in the majority of cases cannot be prevented with an effective ablation therapy. Furthermore, an underlying hepatic disease in patients with HCC or an advanced primary tumor in patients with hepatic metastases are predisposing factors for tumor recurrence, which cannot be eliminated with an ablation procedure.

The risk of complications was not significantly different between the groups and both procedures presented a limited number of adverse events. This finding is important since larger ablation zones, which are achieved through MWA, could be perceived to cause more perioperative complications and damage to liver function compared to RFA. This assumption was refuted with the results of our meta-analysis.

CA and LTP were compared separately among patients with HCC and CRLM. As mentioned above, results derived from the two RCTs in HCC patients showed statistically decreased rates of LTP following MWA compared to RFA.15, 16 In the present meta-analysis, only three retrospective studies compared the two methods in patients with CRLM; consequently, reliable conclusions cannot be drawn, though results showed no significant difference.

In accordance with our results, previous studies reported similar rates of CA between RFA and MWA.4,31, 32, 33, 34 Glassberg et al. reported statistically decreased rates of LTP in the MWA group compared to RFA. Systematic reviews and meta-analyses conducted before 2015 reported comparable rates of LTP between RFA and MWA.31, 32, 33, 34 However, results were derived from studies that in many cases utilized first generation MWA devices. In our meta-analysis, studies published before 2010 were excluded to eliminate this factor. Since the majority of surveys in our analysis utilized new generation devices, which provide controlled and enhanced ablation, beneficial results of MWA over RFA can be attributed to this factor.

Subgroup analysis showed no difference between RFA and MWA for tumor size less or larger than 3 cm. Similar to our findings, Luo et al. concluded that CA and LTP were comparable between RFA and MWA in tumors with diameter larger than 3 cm.34 In contrast to our results, Facciorusso et al. reported significantly decreased incidence of LTP in the MWA group compared to RFA when meta-analysis was restricted to studies with high tumor burden.32 However, the authors failed to define the size of lesions with high tumor burden. This subgroup analysis was performed without clear criteria and results should be evaluated with caution.

In contrast to our results, Glassberg et al. found that LTP in patients with tumor sizes > 2.5 cm was statistically reduced in MWA group compared to RFA.4 However, authors did not report the studies that were included in this subgroup analysis. For that reason, level and quality of evidence cannot be assessed. At this point, we should mention that Glassberg et al. included observational studies with low quality, which were excluded from our meta-analysis, since were associated with high risk of confounding bias and insufficient comparison of baseline characteristics.35, 36, 37, 38, 39, 40 Furthermore, studies that compared RFA or MWA combined with TACE were included in the meta-analysis by Glassberg et al., which could influence the results of the ablation methods.

Contrary to our findings, the meta-analysis conducted by Glassberg et al. reported that distant recurrence was significantly reduced by 15% with MWA compared to RFA when only RCTs were included in the subgroup meta-analysis.4 These results were derived from the RCTs conducted by Abdelaziz et al. as well as by Yu et al.13,16 The second RCT was assessed as high risk of bias in all domains during full-text screening in our study. Consequently, results from a high risk study cannot be assessed as reliable and interpretation should be performed with caution.

The findings in the present meta-analysis should be interpreted in view of certain limitations. First, observational studies without randomization were included in the analysis, which is associated with potential confounding, selection, measurement, and reporting bias. In order to eliminate bias attributed to observational studies, only surveys with low or moderate overall risk of bias were included. Second, significant inter-study heterogeneity was observed for certain outcomes. In these cases, influence of retrospective studies on the results and sources of heterogeneity were examined with subgroup analysis of RCTs and retrospective studies separately. Third, different MWA and RFA devices were utilized across the surveys, which could influence the results of our analysis. Since various devices were used, a subgroup analysis based on the type of devices was not possible. Fourth, limited number of studies included patients with liver metastases or CRLM. Consequently, further RCTs are required to compare MWA with RFA in patients with hepatic metastases.

In addition, in the present study, the proved superiority of MWA over RFA in terms of LTP cannot be translated into better long-term oncological outcomes, since survival outcomes were not evaluated. Overall survival and disease-free survival were not included in our analysis, since limited data can be drawn from the available studies. The majority of surveys were retrospective in design and have included patients with no 100% matching in oncological characteristics. Furthermore, some patients underwent simultaneously surgical resection and ablation. Survival of these patients is multifactor in etiology and causality. Regarding patients with liver metastases, neoadjuvant or adjuvant treatment and tumor stage were not 100% similar between the two groups. For that reason, survival after ablation is associated with several parameters, which could not be attributed only to the effectiveness of the ablative procedures. In fact, LTP and CA are generally considered the best indicators of treatment effectiveness for ablative methods rather than overall survival or disease-free survival.

The meta-analysis is strengthened by its broad inclusion of 15 studies with a total of 2,169 patients. In contrast to other meta-analyses, low quality studies were excluded. Consequently, results were derived from high or moderate quality studies. Taking into consideration the results of the present meta-analysis, we suggest that MWA should be the ablation method of choice in the treatment of HCC. Finally, since the majority of studies included patients with HCC, further RCTs are required to evaluate the role of ablation treatments in patients with liver metastases.