Radiofrequency Ablation of Primary or Metastatic Liver Tumors - CAM 70191HB

Description
Radiofrequency ablation (RFA) is a procedure in which a probe is inserted into the center of a tumor and heated locally by a high-frequency, alternating current that flows from electrodes. The local heat treats the tissue adjacent to the probe, resulting in a 3 to 5 cm sphere of dead tissue. The cells killed by RFA are not removed but are gradually replaced by fibrosis and scar tissue. If there is a local recurrence, it occurs at the edge of the treated tissue and, in some cases, is retreated. Radiofrequency ablation may be performed percutaneously, laparoscopically, or as an open procedure.

Summary of Evidence
For individuals who have primary, operable hepatocellular carcinoma (HCC) who receive radiofrequency ablation (RFA), the evidence includes meta-analyses of randomized controlled trials (RCTs) and/or retrospective observational studies and additional observational studies. Relevant outcomes are overall survival (OS), disease-specific survival, change in disease status, and morbid events. The majority of data found that patients undergoing surgical resection experienced longer survival outcomes and lower recurrence rates than patients receiving RFA, though complication rates were higher with surgical resection. Some meta-analyses of specifically selected populations (e.g., small tumor sizes or Child-Pugh Class A liver function or HCC within the Milan criteria) found that OS and disease-free survival (DFS) rates were not significantly different between RFA and surgical resection. Results from observational studies have suggested that RFA alone or RFA plus percutaneous ethanol injection (PEI) could be as effective as a resection for small HCC tumors as OS and DFS rates were not significantly different between RFA and surgical resection. An exact tumor cutoff size has not been established. Some studies found that OS was similar in patients receiving RFA or resection when tumor size was 3 cm or less; however, OS was significantly longer in patients undergoing resection if the tumor size was between 3.1 cm and 5 cm. Further study in a multicenter RCT would permit greater certainty whether RFA, with or without other ablative or arterial-directed therapies, is as effective as surgical resection in treating HCC tumors 3 cm or smaller. The evidence is insufficient to determine that the technology results in an improvement in the net health outcome.

For individuals who have inoperable HCC who receive RFA, the evidence includes RCTs and several systematic reviews and meta-analyses. Relevant outcomes are OS, disease-specific survival, change in disease status, and morbid events. When resection is not an option, nonsurgical options include RFA, PEI, transarterial chemoembolization (TACE), cryoablation, microwave ablation, and systemic therapy. Meta-analyses comparing RFA to other local ablative therapies have found that RFA and microwave ablation are similarly effective, that RFA is more effective than PEI, and that RFA may be better than cryoablation. The evidence comparing RFA with TACE is limited, and no conclusions can be drawn. RFA has also been shown to improve survival in patients with unresectable HCC as an adjunct to chemotherapy. The evidence is sufficient to determine that the technology results in an improvement in the net health outcome.

For individuals who have inoperable HCC awaiting liver transplant who receive RFA, the evidence includes small case series. Relevant outcomes are OS, disease-specific survival, and change in disease status. A number of approaches are used in this patient population, including RFA and other locoregional therapies, particularly TACE. Locoregional therapy has reduced the dropout rate of patients with HCC awaiting a liver transplant. The evidence is sufficient to determine that the technology results in an improvement in the net health outcome.

For individuals who have inoperable hepatic metastases of colorectal origin who receive RFA, the evidence includes an RCT, systematic reviews and meta-analyses, prospective cohort series, and retrospective case series. Relevant outcomes are OS, disease-specific survival, symptoms, change in disease status, morbid events, quality of life, and treatment-related morbidity. There are no RCTs comparing RFA with alternative treatments for patients who have unresectable colorectal liver metastases. However, an RCT assessing RFA plus chemotherapy found improved survival at 8 years compared with chemotherapy alone. In addition, prospective studies have demonstrated that OS following RFA is at least equivalent to and likely better than currently accepted systemic chemotherapy in well-matched patients with unresectable hepatic metastatic colorectal cancer (CRC) who do not have extrahepatic disease. Results from a number of uncontrolled case series also have suggested RFA of hepatic CRC metastases produces long-term survival that is at a minimum equivalent to but likely superior to historical outcomes achieved with systemic chemotherapy. Evidence from a comparative study has indicated RFA has fewer deleterious effects on quality of life than chemotherapy and that RFA patients recover their quality of life significantly faster than chemotherapy recipients. It should be noted that patients treated with RFA in different series might have had better prognoses than those who had chemotherapy, suggesting patient selection bias might at least partially explain the better outcomes observed following RFA. The evidence is sufficient to determine that the technology results in an improvement in the net health outcome.

For individuals who have inoperable hepatic metastases of neuroendocrine origin who receive RFA, the evidence includes case series and a systematic review of case series. Relevant outcomes are OS, disease-specific survival, symptoms, change in disease status, morbid events, quality of life, and treatment-related morbidity. Most reports of RFA treatment for neuroendocrine liver metastases have assessed small numbers of patients or subsets of patients in reports of multiple ablative methods or very small subsets of larger case series of patients with various diagnoses. The available evidence has indicated that durable tumor and symptom control of neuroendocrine liver metastases can be achieved using RFA in individuals whose symptoms are not controlled by systemic therapy or who are ineligible for resection. The evidence is sufficient to determine that the technology results in an improvement in the net health outcome.

For individuals who have hepatic metastases, not of colorectal or neuroendocrine origin who receive RFA, the evidence includes a systematic review, small, nonrandomized comparative studies and small case series. Relevant outcomes are OS, disease-specific survival, symptoms, change in disease status, morbid events, quality of life, and treatment-related morbidity. Similar to primary HCC, resection appears to have the most favorable outcomes. For patients who are ineligible for resection, RFA may provide a survival benefit. However, the evidence is limited by study designs with a high-risk of bias and small sample sizes. The evidence is insufficient to determine that the technology results in an improvement in the net health outcome.

Additional Information
Not applicable.

Background
Hepatic and Neuroendocrine Tumors
Hepatic tumors can arise as primary liver cancer (hepatocellular cancer) or by metastasis to the liver from other tissues. Local therapy for hepatic metastasis may be indicated when there is no extrahepatic disease, which rarely occurs for patients with primary cancers other than colorectal carcinoma or certain neuroendocrine malignancies. A study from 2016 determined that the incidence of liver cancer was higher among White individuals, Black individuals, and Hispanic individuals born after 1938.¹ The incidence of hepatocellular carcinoma was twice as high for U.S.-born Hispanic men compared to Hispanic men born outside of the US. This may be due to the increased risk of smoking, hepatitis B or C infection, and diabetes among U.S.-born Hispanic individuals.

Neuroendocrine tumors are tumors of cells that possess secretory granules and originate from the neuroectoderm. Neuroendocrine cells have roles both in the endocrine system and in the nervous system. They produce and secrete a variety of regulatory hormones, or neuropeptides, which include neurotransmitters and growth factors. Overproduction of the specific neuropeptides produced by the cancerous cells causes various symptoms, depending on the hormone produced. They are rare, with an incidence of 2 to 4 per 100,000 per year.

Treatment
Treatment options for hepatocellular carcinoma (HCC) range from potentially curative treatments, such as resection or liver transplantation, to nonsurgical options, which include ablative therapies (radiofrequency ablation [RFA], cryoablation, microwave ablation, percutaneous ethanol, or acetic acid injection), transarterial chemoembolization, radiation therapy, and systemic therapy. Choice of therapy depends on the severity of the underlying liver disease, size, and distribution of tumors, vascular supply, and patient overall health. Treatment of liver metastases is undertaken to prolong survival and reduce endocrine-related symptoms and hepatic mass-related symptoms.

At present, surgical resection with adequate margins or liver transplantation constitutes the only treatments available with demonstrated curative potential for hepatic tumors. However, most hepatic tumors are unresectable at diagnosis, due either to their anatomic location, size, number of lesions, or underlying liver reserve. Comorbid conditions may also make patients unqualified for surgical resection.

Radiofrequency Ablation
Radiofrequency ablation is a procedure in which a needle electrode is inserted into a tumor either percutaneously, through a laparoscope, or through an open incision. The electrode is heated by a high-frequency, alternating current, which destroys tissue in a 3 to 5 cm sphere of the electrode. The cells killed by RFA are not removed but are gradually replaced by fibrosis and scar tissue. If there is a local recurrence, it occurs at the edge of the treated tissue and, in some cases, is retreated. Radiofrequency ablation has been investigated as a treatment for unresectable hepatic tumors, both as a primary intervention and as a bridge to a liver transplant. In the latter setting, RFA is being tested to determine whether it can reduce the incidence of tumor progression in patients awaiting transplantation and thus maintain patients' candidacy for liver ablation, transhepatic arterial chemoembolization, microwave coagulation, percutaneous ethanol injection, and radioembolization (yttrium-90 microspheres).

Note that RFA of extrahepatic tumors is addressed in evidence review 70195.

Regulatory Status
RFA devices have been cleared for marketing by the U.S. Food and Drug Administration through the 510(k) process. Food and Drug Administration product code GEI.

Related Policies
70175 Cryosurgical Ablation of Primary or Metastatic Liver Tumors
70195 Radiofrequency Ablation of Miscellaneous Solid Tumors Excluding Liver Tumors
80111 Transcatheter Arterial Chemoembolization (TACE) to Treat Primary or Metastatic Liver Malignancies
80143 Radioembolization for Primary and Metastatic Tumors of the Liver

Policy
Radiofrequency ablation of primary, inoperable (e.g., due to location of lesion[s] and/or comorbid conditions), hepatocellular carcinoma may be considered MEDICALLY NECESSARY under the following conditions:

As a primary treatment of hepatocellular carcinoma meeting the Milan criteria (a single tumor of ≤ 5 cm or up to 3 nodules < 3 cm)
As a bridge to transplant, where the intent is to prevent further tumor growth and to maintain a patient’s candidacy for liver transplant

Radiofrequency ablation as a primary treatment of inoperable hepatic metastases may be considered MEDICALLY NECESSARY under the following conditions:

Metastases are of colorectal origin and meet the Milan criteria (a single tumor of ≤ 5 cm or up to 3 nodules < 3 cm)
Metastases are of neuroendocrine in origin and systemic therapy has failed to control symptoms

Radiofrequency ablation of primary, inoperable, hepatocellular carcinoma is investigational and/ or unproven and therefore considered NOT MEDICALLY NECESSARY under the following conditions:

When there are more than 3 nodules or when not all sites of tumor foci can be adequately treated
When used to downstage (downsize) hepatocellular carcinoma in patients being considered for liver transplant

Radiofrequency ablation of primary, operable hepatocellular carcinoma is investigational and/ or unproven and therefore considered NOT MEDICALLY NECESSARY.

Radiofrequency ablation for hepatic metastasis is investigational and/or unproven and therefore considered NOT MEDICALLY NECESSARY for:

Hepatic metastases from colorectal cancer or neuroendocrine tumors that do not meet the criteria above.
Hepatic metastases from other types of cancer except colorectal cancer or neuroendocrine tumors.

Policy Guidelines
None

Rationale
Evidence reviews assess the clinical evidence to determine whether the use of technology improves the net health outcome. Broadly defined, health outcomes are the length of life, quality of life, and ability to function including benefits and harms. Every clinical condition has specific outcomes that are important to patients and managing the course of that condition. Validated outcome measures are necessary to ascertain whether a condition improves or worsens; and whether the magnitude of that change is clinically significant. The net health outcome is a balance of benefits and harms.

To assess whether the evidence is sufficient to draw conclusions about the net health outcome of technology, 2 domains are examined: the relevance and the quality and credibility. To be relevant, studies must represent 1 or more intended clinical use of the technology in the intended population and compare an effective and appropriate alternative at a comparable intensity. For some conditions, the alternative will be supportive care or surveillance. The quality and credibility of the evidence depend on study design and conduct, minimizing bias and confounding that can generate incorrect findings. The randomized controlled trial (RCT) is preferred to assess efficacy; however, in some circumstances, nonrandomized studies may be adequate. Randomized controlled trials are rarely large enough or long enough to capture less common adverse events and long-term effects. Other types of studies can be used for these purposes and to assess generalizability to broader clinical populations and settings of clinical practice.

Radiofrequency Ablation to Treat Primary, Operable Hepatocellular Carcinoma
The evidence is evaluated separately for operable and inoperable tumors. If data are available, separate analyses by tumor size are evaluated.

Clinical Context and Therapy Purpose
The purpose of radiofrequency (RFA) is to provide a treatment option that is an alternative to or an improvement on existing therapies, such as surgical resection, in individuals with primary, operable hepatocellular carcinoma (HCC).

The following PICO was used to select literature to inform this review.

Populations
The relevant population of interest is individuals with primary, operable HCC.

Interventions
The therapy being considered is RFA.

Comparators
Comparators of interest include surgical resection. Surgical resection is a potentially curative therapy for individuals with HCC with adequate/preserved liver functional reserve (Child-Pugh Class A or Class B in certain circumstances). Some staging systems can be used to direct treatment or predict survival after therapeutic intervention. Two notable systems include the Barcelona Clinic Liver Cancer (BCLC) staging system and Milan criteria. The BCLC system is currently the standard classification system for the clinical management of individuals with HCC. Hepatic resection is proposed for early-stage HCC (BCLC-0/A). Milan criteria can aid in determining eligibility for transplantation. Milan criteria include: single tumor <5 cm, no more than 3 foci with each not exceeding 3 cm, absence of angioinvasion, and absence of extrahepatic involvement. Individuals with resectable HCC are also potentially eligible for a liver transplant. However, the availability of liver donors limits its use.

Outcomes
The general outcomes of interest are overall survival (OS), disease-specific survival, change in disease status, and morbid events (Table 1).

Table 1. Outcomes of Interest for Individuals with Primary, Operable Hepatocellular Carcinoma

Outcomes	Details
Overall survival	Survival rate or proportion dead [30 days to 10 years]
Disease-specific survival	Disease/recurrence-free survival [1 year to 10 years]
Morbid events	Complications, adverse events [peri- or post-procedure]

Study Selection Criteria
Methodologically credible studies were selected using the following principles:

To assess efficacy outcomes, comparative controlled prospective trials were sought, with a preference for RCTs;
In the absence of such trials, comparative observational studies were sought, with a preference for prospective studies.
To assess long-term outcomes and adverse events, single-arm studies that capture longer periods of follow-up and/or larger populations were sought.
Consistent with a 'best available evidence approach,' within each category of study design, studies with larger sample sizes and longer durations were sought.
Studies with duplicative or overlapping populations were excluded.

Review of Evidence
The evidence on RFA as a treatment of resectable HCC includes RCTs, meta-analyses, and observational studies that combined RFA with transhepatic arterial chemoembolization (TACE) or other locally ablative procedures.

Systematic Reviews
Several systematic reviews are available comparing health outcomes between RFA, with or without other locally ablative procedures, and surgical resection. The most recent evaluations in patients with early HCC who are suitable candidates for either RFA or surgical resection are summarized below and in Tables 2, 3, and 4. The vast majority of trials included in available systematic reviews were conducted in China, Japan, and Korea.

Yang et al (2025) compared the efficacy of liver resection to RFA in patients with single HCC tumors measuring ≤3 cm.² The analysis included 39 trials (2 RCTs, 37 nonrandomized), which included 6356 patients with tumors ≤3 cm and 5829 patients with tumors ≤2 cm. There was no difference in OS when only the RCTs were analyzed. When only the nonrandomized studies were included, there was an improvement in OS (hazard ratio [HR], 0.80; 95% confidence interval [CI], 0.68 to 0.93; p=.005) with resection compared to RFA. Similarly, DFS was better with resection compared to RFA (HR, 0.63; 95% CI, 0.49 to 0.81). When tumor size was considered, resection led to improved OS (HR, 0.73; 95% CI, 0.54 to 0.97) and DFS (HR, 0.74; 95% CI, 0.57 to 0.90) in patients with tumors ≤2 cm, compared to RFA.

Zhang et al (2022) compared the efficacy of liver resection, RFA alone, and RFA plus TACE in patients with very early or early stage HCC.³ Randomized trials (n=10) and propensity score-matched cohort analyses (n=15) were included. In a network meta-analysis, 1-year OS was similar between resection and RFA alone, but 3-year and 5-year OS favored resection ( HR , 0.74 ; 95% CI , 0.56 to 0.96 and HR, 0.73; 95% CI, 0.55 to 0.94, respectively). Recurrence-free survival at 1, 3, and 5 years was also significantly higher with resection compared to RFA alone. There were no significant differences in survival outcomes at any time point between resection and RFA plus TACE.

Jia et al (2021) performed a meta-analysis to compare clinical efficacy between RFA and surgical resection in patients with HCC meeting Milan criteria.⁴ The analysis included RCTs, accounting for 8 trials (N=1177). There were no significant differences found between RFA and surgical resection in OS and disease-free survival (DFS) rates. In a subgroup analysis stratifying by tumor size, there was still no significant difference between the 2 therapies for both tumors ≤4 cm and >4 cm. Limitations of the analysis include the inclusion of clinical trials with small sample sizes and a lack of double-blinding.

Shin et al (2021) compared the efficacy of surgical resection with local ablative therapies for HCC meeting Milan criteria.⁵ The analysis included 7 RCTs and 18 non-randomized trials (N=5629) that compared surgical resection with either RFA, microwave ablation, or RFA plus TACE. Four of the RCTs were judged to be at high risk of bias overall, due to either lack of reported randomization method or missing data. All non-randomized trials were classified as having a high risk for bias due to the missing data. There was no significant difference between surgical resection and RFA alone when the RCTs were analyzed; the 3- and 5-year OS favored surgical resection in the analysis of the non-randomized trials. A multiple treatment meta-analysis using a frequentist framework random effect model found that 5-year recurrence-free survival was highest with surgical resection (HR, 0.64; 95% CI, 0.56 to 0.74 vs. RFA), followed by RFA plus TACE (HR, 0.70; 95% CI, 0.53 to 0.92 vs. RFA); no difference was found between microwave ablation and RFA (HR, 0.93; 95% CI, 0.63 to 1.37). However, the latter comparisons were limited by the number of trials evaluating RFA plus TACE (5 studies) and microwave ablation (2 studies).

Li et al (2020) also evaluated the comparative efficacy of RFA and surgical resection in patients with HCC meeting Milan criteria with liver function Child-Pugh scores of grade A or B.⁶ One RCT and 15 retrospective observational studies were included in their analysis. Surgical resection was associated with significantly improved OS and DFS rates. In a subgroup analysis stratified by tumor size, 5-year OS rates were significantly improved in patients receiving surgical resection in patients with tumors ≤3 cm and >3 cm. The authors noted that the observational studies, which comprised most of the data, had significant heterogeneity and were prone to potential selection biases.

The network meta-analysis by Zhu et al (2018) compared the safety and effectiveness of several treatments for small HCC, including RFA, percutaneous ethanol injection (PEI), percutaneous acetic acid injection, and surgical resection.⁷ The authors identified 12 RCTs and 2 quasi-RCTs with a mean follow-up period of 22 months for most trials. The directed meta-analysis assessed mortality, local recurrence, and adverse events. It showed that PEI had a higher risk of proportion dead than RFA, and RFA had a higher risk of proportion dead than surgical resection; a single study found that percutaneous acetic acid injection had a higher risk of proportion dead than RFA (Table 2). For local recurrence, PEI had a higher recurrence than RFA, RFA had a higher recurrence than surgical resection, and percutaneous acetic acid injection had a higher recurrence than RFA. Adverse events were fewer with RFA than with surgical resection (odds ratio [OR], 0.11; 95% CI, 0.03 to 0.34), but there were no significant effects in reducing adverse events between PEI versus RFA and percutaneous acetic acid injection versus RFA. The authors used GRADE (Grading of Recommendations Assessment, Development, and Evaluation) to rate the quality of evidence for primary outcomes and found it to be very low for most comparisons. Further interpretation of results is limited due to the heterogeneity of the data, as well as the small sample sizes in the included studies.

Jia et al (2017) evaluated the comparative efficacy of RFA and surgical resection in patients with HCC and Child-Pugh Class A liver function.⁸ Two RCTs and 13 retrospective observational studies were selected for inclusion. In the overall population, patients receiving surgical resection had increased odds for 3-year and 5-year survival compared to RFA. In studies that were limited to patients with solitary tumors or those with tumors ≤3 cm, the OS and DFS rates were not significantly different between RFA and surgical resection. Limitations of the meta-analysis are similar to others including the use of observational data, which increased heterogeneity and potentially compares groups that may not have equivalent baseline characteristics.

Feng et al (2015) compared RFA to surgical resection in patients with small HCC.⁹ Three RCTs and 20 retrospective observational studies were included in the analysis. Rates of OS and recurrence-free survival with surgical resection were significantly higher than RFA. However, complication rates were higher in the surgical resection group compared to RFA (OR, 0.37; 95% CI, 0.24 to 0.58).

Table 2. Comparison of Meta-Analyses of Radiofrequency Ablation for Primary, Operable Hepatocellular Carcinomaa

Study	Study type	Country	Feng et al (2015)⁹	Jia et al (2017)⁸	Zhu et al (2018)⁷	Li et al (2020)⁶	Jia et al (2021)⁴	Shin et al (2021)⁵	Zhang et al (2022)³	Yang et al (2025)²
Kang et al (2023)	NRT	Korea								⚫
Ivanics et al (2022)	NRT	Canada								⚫
Ko et al (2022)	NRT	Korea								⚫
Lee et al (2022)	NRT	Korea								⚫
Conticchio et al (2022)	NRT	Italy								⚫
Zhang et al (2022)	NRT	China								⚫
Delvecchio et al (2021)	NRT	Italy								⚫
Lee et al (2021)	NRT	Korea							⚫	⚫
Li et al (2021)	NRT	China							⚫	⚫
Takayama et al (2021)	RCT	Japan								⚫
Wu et al (2021)	NRT	China								⚫
Zhang et al (2021)	RCT	China							⚫
Cha et al (2020)	NRT	Korea								⚫
Lin et al (2020)	NRT	China								⚫
Pan et al (2020)	NRT	China						⚫	⚫
Zheng et al (2020)	NRT	China								⚫
Chong et al (2019)	NRT	China						⚫	⚫
Chu et al (2019)	NRT	Korea						⚫		⚫
Kim et al (2019)	NRT	Korea						⚫	⚫	⚫
Lee et al (2019)	NRT	Korea							⚫
Wang et al (2019)	NRT	China								⚫
Yi et al (2019)	NRT	China							⚫
Lee et al (2018)	RCT	Korea					⚫	⚫	⚫	⚫
Takayasu et al (2018)	NRT	Japan								⚫
Bholee et al (2017)	NRT	China							⚫
Lee et al (2017)	NRT	Korea							⚫
Ng et al (2017)	RCT	Japan					⚫	⚫	⚫
Kang et al (2016)	NRT	Korea						⚫	⚫
Kato et al (2018)	NRT	Japan						⚫
Kim et al (2016)	NRT	Korea						⚫		⚫
Liu et al (2016)	NRT	Taiwan						⚫	⚫
Liu et al (2016)	RCT	China							⚫	⚫
Santambrogio et al (2016)	NRT	italy								⚫
Song et al (2016)	NRT	China								⚫
Vitali et al (2016)	NRT	France								⚫
Jiang et al (2015)	NRT	China						⚫	⚫
Kang et al (2015)	NRT	Korea								⚫
Lee et al (2015)	NRT	Taiwan		⚫
Song et al (2015)	NRT	China						⚫	⚫
Fang et al (2014)	RCT	China			⚫		⚫	⚫	⚫
Kim et al (2014)	NRT	Korea		⚫						⚫
Yang et al (2014)	NRT	Korea								⚫
Desiderio et al (2013)	NRT	Italy	⚫	⚫
Guo et al (2013)	NRT	China	⚫
Hasegawa et al (2013)	NRT	Japan	⚫
Imai et al (2013)	NRT	Japan	⚫							⚫
Pompili et al (2013)	NRT	Italy	⚫	⚫				⚫	⚫	⚫
Takuma et al (2013)	NRT	Japan							⚫
Tohme et al (2013)	NRT	United States	⚫	⚫
Wong et al (2013)	NRT	Taiwan	⚫	⚫						⚫
Feng et al (2012)	RCT	China	⚫		⚫		⚫	⚫	⚫
Peng et al (2012)	NRT	China	⚫	⚫						⚫
Wang et al (2012)	NRT	Taiwan	⚫							⚫
Giorgio et al (2011)	RCT	Italy			⚫
Huang et al (2011)	NRT	China				⚫
Hung et al (2011)	NRT	Taiwan	⚫			⚫		⚫		⚫
Ikeda et al (2011)	NRT	Japan	⚫
Kong et al (2011)	NRT	China	⚫
Nishikawa et al (2011)	NRT	Japan	⚫			⚫				⚫
Tashiro et al (2011)	NRT	Japan				⚫
Yun et al (2011)	NRT	Korea	⚫							⚫
Huang et al (2010)	RCT	China	⚫	⚫	⚫		⚫	⚫	⚫	⚫
Morimoto et al (2010)	RCT	Japan							⚫
Nanashima et al (2010)	NRT	Japan		⚫
Takayama et al (2010)	NRT	Japan								⚫
Santambrogio et al (2009)	NRT	Italy		⚫		⚫
Shibata et al (2009)	RCT	Japan							⚫
Ueno et al (2009)	NRT	Japan	⚫
Abu-Hilal et al (2008)	NRT	United Kingdom	⚫			⚫
Brunello et al (2008)	RCT	Italy			⚫
Guglielmi et al (2008)	NRT	Italy	⚫			⚫
Hiraoka et al (2008)	NRT	Japan	⚫			⚫				⚫
Ueno et al (2008)	NRT	Japan				⚫
Lupo et al (2007)	NRT	Italy	⚫			⚫
Chen et al (2006)	RCT	China	⚫	⚫	⚫	⚫	⚫	⚫	⚫	⚫
Lu et al (2006)	RCT	China					⚫
Wakai et al (2006)	NRT	Japan				⚫
Chen et al (2005)	RCT	China					⚫
Cho et al (2005)	NRT	Korea		⚫		⚫
Hong et al (2005)	NRT	Korea		⚫		⚫
Lin et al (2005)	RCT	Taiwan			⚫
Montorsi et al (2005)	NRT	Italy				⚫
Ogihara et al (2005)	NRT	United States				⚫
Shiina et al (2005)	NRT	Japan			⚫
Sung et al (2005)	NRT	Korea	⚫
Lin et al (2004)	RCT	Taiwan			⚫
Vivarelli et al (2004)	NRT	Italy		⚫		⚫
Guglielmi et al (2003)	NRT	Italy		⚫
Lencioni et al (2003)	RCT	Italy			⚫
Livraghi et al (1999)	RCT	Italy			⚫

NRT: non-randomized trial; RCT: randomized controlled trial; RFA: radiofrequency ablation.
^aFor meta-analyses that evaluated more than 1 ablative therapy, only trials that evaluated RFA are listed in the table.

Table 3. Characteristics of Meta-Analyses of Radiofrequency Ablation for Primary, Operable Hepatocellular Carcinoma

Study	Dates	Trials	Participants	N (Range)	Design	Duration
Yang et al (2025)²	2006-2023	39	Pts with solitary HCC lesions ≤3 cm	N=6356 (73 to 2550)	RCTs and NRTs	NR
Zhang et al (2022)³	2006- 2021	25	Pts with HCC.	N=4249 (19 to 354)	RCTs and NRTs	Mean follow-up, range 24.2 to 93 months
Jia et al (2021)⁴	2005-2019	8	Pts with primary HCC meeting Milan criteria^a; liver function Child-Pugh class A or B; suitable candidates for surgical resection and/or RFA.	N=1177 (63 to 230)	RCTs	Mean follow-up range, 27.9 to 92.4 months
Shin et al (2021)⁵	2006-2020	25	Pts with primary HCC meeting Milan criteria.^a	N=5629 (52 to 1208)	RCTs and NRTs	NR
Li et al (2020)⁶	2000-2018	25	Pts with primary HCC meeting Milan criteria^a; liver function Child-Pugh class A or B; suitable candidates for surgical resection and/or RFA.	N=13,147 (NR)	RCT and observational comparative studies	1 to 5 years
Zhu et al (2018)⁷	1998-2013	14	Pts diagnosed with small HCC meeting Milan criteria.	N=2096 (29 to 143)	RCTs and quasi-RCTs	Mean, 22 months
Jia et al (2017)⁸	2003-2015	15	Pts with early-stage HCC; liver function Child-Pugh class A; suitable candidates for surgical resection and/or RFA.	N=3627 (67 to 1061)	RCTs and observational comparative studies	1 to 5 years
Feng et al (2015)⁹	2005-2013	23	Pts with small HCC not previously treated with RFA or surgical resection; suitable candidates for surgical resection and/or RFA.	N=15,482 (63 to 10,909)	RCTs and NRTs	1 to 5 years

Table 4. Results of Meta-Analyses of Radiofrequency Ablation for Primary, Operable Hepatocellular Carcinoma

Study	Overall Survival OR or HR (95% CI)			Disease-free Survival OR or HR (95% CI)
	1 yr	2/3 yr	5 yr	1 yr	2/3 yr	4/5 yr
Yang et al (2025)²
N	NR			NR
SR vs. RFA (HR)	Survival duration not specified RCTs: 0.73 (0.26 to 2.05) NRTs: 0.80 (0.68 to 0.93)			Survival duration not specified RCTs: not analyzed NRTs: 0.63 (0.49 to 0.81)
I² (p)	RCTs: 0% (.55) NRTs: 58% (.005)			RCTs: NA NRTs: 81% (<.00001)
Feng et al (2015)⁹
N	4199	15,414 (3-yr)	14,686	3544	3389 (3-yr)	2984 (5-yr)
RFA vs. SR (OR)	0.71 (0.52 to 0.96)	0.62 (0.49 to 0.78)	0.55 (0.47 to 0.66)	0.58 (0.45 to 0.76)	0.52 (0.40 to 0.68)	0.50 (0.34 to 0.76)
I² (p)	30% (.10)	NR (<.001)	NR (.02)	53% (.004)	NR (<.001)	NR (.00)
Jia et al (2017)⁸
N	NR (14 studies)	NR (15 studies; 3-yr)	NR (9 studies)	NR (9 studies)	NR (9 studies; 3-yr)	NR (6 studies; 5-yr)
RFA vs. SR (OR)	1.095 (0.636 to 1.885)	1.753 (1.197 to 2.567)	1.552 (1.026 to 2.348)	1.209 (0.935 to 1.563)	1.517 (1.076 to 2.140)	1.810 (1.071 to 3.058)
I² (p)	49% (.02)	74.2% (.000)	72.6% (.000)	20.4% (.261)	68.3% (.001)	68.5% (.007)
Zhu et al (2018)^a⁷
PEI vs. RFA (OR)	-	1.66 (1.13 to 2.44)	-	-	2.74 (1.42 to 5.29)	-
PAI vs. RFA (OR)	-	1.63 (0.67 to 3.96)	-	-	2.79 (1.19 to 6.54)	-
RFA vs. SR (OR)	-	1.21 (0.62 to 2.35)	-	-	2.02 (1.01 to 4.02)	-
Li et al (2020)⁶
N	3921	4053 (3-yr)	3397	3394	3326 (3-yr)	3076 (5-yr)
RFA vs. SR (OR)	0.757 (0.578 to 0.989)	0.530 (0.401 to 0.700)	0.566 (0.423 to 0.758)	0.569 (0.456 to 0.711)	0.418 (0.267 to 0.653)	0.374 (0.231 to 0.606)
I² (p)	0% (.55)	61% (.0005)	71% (<.0001)	42% (.06)	70% (.0001)	57% (.01)
Jia et al (2021)⁴
N	1177	947 (3-yr)	281	1114	1072 (3-yr)	-
RFA vs. SR (OR)	0.91 (0.45 to 1.83)	0.82 (0.56 to 1.19)	1.03 (0.61 to 1.73)	0.87 (0.63 to 1.21)	0.79 (0.58 to 1.07)	-
I² (p)	37% (.13)	23% (.25)	0% (.80)	0% (.76)	31% (.19)	-
Shin et al (2021)^b⁵				Recurrence-free survival	Recurrence-free survival (3-yr)	Recurrence-free survival (5-yr)
N (RCTs)	916	916 (3-yr)	691	978	978	690
SR vs. RFA (HR)	0.76 (0.31 to 1.83)	0.72 (0.45 to 1.14)	0.85 (0.55 to 1.29)	0.86 (0.64 to 1.15)	0.83 (0.65 to 1.06)	0.75 (0.62 to 0.92)
I² (p)	53% (.08)	61% (.04)	56% (.08)	2% (.40)	46% (.10)	10% (.35)
N (NRT)	1750	3412 (3-yr)	2928	3012	3012	2658
SR vs. RFA (HR)	1.91 (0.76 to 4.80)	0.75 (0.59 to 0.95)	0.72 (0.58 to 0.89)	0.54 (0.42 to 0.70)	0.61 (0.53 to 0.70)	0.61 (0.52 to 0.72)
I² (p)	44% (.08)	18% (.27)	33% (.15)	50% (.03)	31% (.16)	52% (.03)
Zhang et al (2022)³
N	2734	2995	1785	2738	2999	1785
SR vs. RFA (OR)	0.93 (0.59 to 1.47)	0.75 (0.58 to 0.97)	0.71 (0.55 to 0.92)	0.66 (0.51 to 0.84)	0.69 (0.58 to 0.82)	0.61 (0.48 to 0.78)
I² (p)	28% (.76)	53% (.03)	53% (.009)	40% (.006)	60% (<.0001)	70% (<.0001)

CI: confidence interval; HR: hazard ratio; NA: not applicable; NR: not reported; NRT: non-randomized trial; OR: odds ratio; PAI: percutaneous acetic acid injection; PEI: percutaneous ethanol injection; RCT: randomized controlled trial; RFA: radiofrequency ablation; SR: surgical resection.
^aZhu et al (2018) reported proportion dead vs overall survival and local recurrence vs disease-free survival.
^bShin et al (2021) conducted separate meta-analyses for RCTs and NRTs.

Zhang et al (2022) conducted a meta-analysis of 16 studies (1 RCT, 15 nonrandomized) that compared surgical resection and RFA in patients with HCC and cirrhosis.¹⁰ Most measures of survival were better with resection than RFA, including 3-year OS (OR, 0.48; 95% CI, 0.35 to 0.67), 5-year OS (OR, 0.49; 95% CI, 0.38 to 0.63), 1-year DFS (OR, 0.42; 95% CI, 0.32 to 0.54), 3-year DFS (OR, 0.36; 95% CI, 0.24 to 0.53), and local recurrence. RFA had better postoperative complication rates and operative times. Most analyses had significant heterogeneity. The authors concluded that high quality multicenter prospective studies are needed to identify patient subgroups that would benefit most from each treatment. This analysis is not included in Tables 2, 3, and 4 since the studied population (ie, with cirrhosis) does not match the populations in the other analyses.

Randomized Controlled Trials
Song et al (2024) published results from a single-center, unblinded RCT in China comparing resection to RFA for treatment of HCC.¹¹ Patients with HCC were eligible if they had a single nodule no larger than 5 cm, or up to 3 nodules of 3 cm or smaller. Patients were randomized to receive either liver resection or RFA (N=150). The primary outcome of OS did not differ between groups. Similarly, the secondary outcome of recurrence-free survival did not differ between groups. The 1-, 3-, and 5- year OS rates with laparoscopic resection were 94.7%, 80%, and 74.7%, respectively, and with RFA were 93.3%, 78.7%, and 67.9%, respectively. Tables 5 and 6 describe trial characteristics and results, respectively. The incidence of postoperative complications was higher in the resection group compared to the RFA group (22 [29.3%] vs. 8 [10.7%] adverse events; p=.004). Results are limited by the small sample size and single-center design.

Table 5. Summary of Key RCT Characteristics

Study	Countries	Sites	Dates	Participants	Interventions
					Active	Comparator
Song et al (2024)¹¹	China	1	2014-2015	Adults aged less than 70 years with small HCC (1 nodule no larger than 5 cm, or up to 3 nodules of 3 cm or smaller); liver function Child-Pugh class A or B; suitable candidates for surgical resection and/or RFA.	RFA (n=73)	SR (n=77)

HCC: hepatocellular carcinoma; RCT: randomized controlled trial; RFA: radiofrequency ablation; SR: surgical resection.

Table 6. Summary of Key RCT Results

Study	OS	1 year OS, n (%)	3 years OS, n (%)	5 year OS, n (%)	RFS	1 year RFS, n (%)	3 year RFS, n (%)	5 year RFS, n (%)
Song et al (2024)¹¹								N
RFA	n=73	70 (93.3%)	59 (78.7%)	49 (67.9%)	n=73	52 (69.3%)	40 (53.3%)	29 (41.0%)
SR	n=77	71 (94.7%)	60 (80.0%)	53 (74.7%)	n=77	59 (78.7%)	46 (61.3%)	36 (51.6%)
HR (95% CI)	1.26 (0.69 to 2.30)				1.34 (0.86 to 2.08)
p-value	.451				.189

CI: confidence interval; HR: hazard ratio; OS: overall survival; RCT: randomized controlled trial; RFA: radiofrequency ablation; RFS: recurrence-free survival; SR: surgical resection.

The purpose of the study limitations tables (see Tables 7 and 8) is to display notable limitations identified in each study.

Table 7. Study Relevance Limitations

Study	Population^a	Intervention^b	Comparator^c	Outcomes^d	Duration of Follow-up^e
Song et al (2024)¹¹	4. single-center in China; main cause of HCC was HBV infection, which may differ from causes of HCC in non-Asian regions

HBV: hepatitis B virus; HCC: hepatocellular carcinoma. The study limitations stated in this table are those notable in the current review; this is not a comprehensive gaps assessment. ^aPopulation key: 1. Intended use population unclear; 2. Study population is unclear; 3. Study population not representative of intended use; 4. Enrolled populations do not reflect relevant diversity; 5. Other. ^bIntervention key: 1. Not clearly defined; 2. Version used unclear; 3. Delivery not similar intensity as comparator; 4. Not the intervention of interest (e.g., proposed as an adjunct but not tested as such); 5: Other. ^cComparator key: 1. Not clearly defined; 2. Not standard or optimal; 3. Delivery not similar intensity as intervention; 4. Not delivered effectively; 5. Other. ^dOutcomes key: 1. Key health outcomes not addressed; 2. Physiologic measures, not validated surrogates; 3. Incomplete reporting of harms; 4. Not establish and validated measurements; 5. Clinically significant difference not prespecified; 6. Clinically significant difference not supported; 7. Other. ^eFollow-Up key: 1. Not sufficient duration for benefit; 2. Not sufficient duration for harms; 3. Other.

Table 8. Study Design and Conduct Limitations

Study	Allocation^a	Blinding^b	Selective Reporting^c	Data Completeness^d	Power^e	Statistical^f
Song et al (2024)¹¹		1. not blinded, would be unable to blind due to surgical procedure

The study limitations stated in this table are those notable in the current review; this is not a comprehensive gaps assessment. ^aAllocation key: 1. Participants not randomly allocated; 2. Allocation not concealed; 3. Allocation concealment unclear; 4. Inadequate control for selection bias; 5. Other. ^bBlinding key: 1. Participants or study staff not blinded; 2. Outcome assessors not blinded; 3. Outcome assessed by treating physician; 4. Other. ^cSelective Reporting key: 1. Not registered; 2. Evidence of selective reporting; 3. Evidence of selective publication; 4. Other. ^dData Completeness key: 1. High loss to follow-up or missing data; 2. Inadequate handling of missing data; 3. High number of crossovers; 4. Inadequate handling of crossovers; 5. Inappropriate exclusions; 6. Not intent to treat analysis (per protocol for noninferiority trials); 7. Other. ^ePower key: 1. Power calculations not reported; 2. Power not calculated for primary outcome; 3. Power not based on clinically important difference; 4. Other. ^fStatistical key: 1. Analysis is not appropriate for outcome type: (a) continuous; (b) binary; (c) time to event; 2. Analysis is not appropriate for multiple observations per patient; 3. Confidence intervals and/or p values not reported; 4. Comparative treatment effects not calculated; 5. Other.

Observational Studies
Chen et al (2018) retrospectively analyzed data from 2 hospitals and compared a combination of RFA plus PEI (n=141) with surgical resection (n=130) in patients with HCC.¹² The study included patients with tumors 2.1 to 5 cm in size. The race and ethnicity of included patients were not described. Overall, patients receiving RFA plus PEI experienced significantly better OS and relapse-free survival than patients undergoing resection. However, subgroup analysis by tumor size showed that significant improvements in OS and relapse-free survival were only experienced by patients with tumors 2.1 to 3 cm (see Table 9).

Table 9. Survival Following Surgical Resection or Radiofrequency Ablation Plus Percutaneous Ethanol Injection for Resectable Hepatocellular Carcinoma

Outcomes	1 Year, %	3 Years, %	5 Years, %	p-value
Overall survival
2.1 to 3.0 cm
RFA plus PEI, n=77	98.0	82.3	74.2
Surgical resection, n=70	89.4	65.1	61.9	.02
3.1 to 5.0 cm
RFA plus PEI, n=64	86.4	65.1	55.4
Surgical resection, n=60	88.9	64.5	49.6	.13
Recurrence-free survival
2.1 to 3.0 cm
RFA plus PEI	79.6	54.7	45.1
Surgical resection	57.6	43.9	31.7	.02
3.1 to 5.0 cm
RFA plus PEI	53.5	29.4	24.0
Surgical resection	42.2	26.6	21.9	.71

Adapted from Chen et al (2018).¹²
PEI: percutaneous ethanol injection; RFA: radiofrequency ablation.

Zhao et al (2019) compared outcomes for RFA, resection, or transplantation in patients from the Surveillance, Epidemiology, and End Results database.¹³ A total of 7664 patients treated between 2004 and 2015 with a single HCC tumor measuring up to 50 mm met study criteria. Outcomes for the 3 treatment arms were evaluated for both the unadjusted population and a propensity score-adjusted population to account for differences in baseline characteristics between patients. Median follow-up for the whole cohort was 55 months for OS. In the overall cohort, liver transplantation was associated with an improved OS (5-year OS, 66%) compared to RFA and resection in both unadjusted and adjusted populations (5-year OS [adjusted], 66% vs. 53% vs. 52%, respectively), but no significant difference was found between RFA and resection. Stratification by tumor size generally showed more survival benefits with resection compared to RFA. Further analysis by prognostic factors found that RFA may be the preferred treatment strategy for patients with low tumor risk (eg, tumor size <20 mm, tumor grade 0, fibrosis score/F0) and good general health condition (Table 10).

Table 10. Overall Survival Probability for Overall Cohort and Stratified by Lesion Size

	Overall Survival, HR (95% CI)
Group Analyzed^a	SR vs. RFA	LT vs. RFA	LT vs. SR
Total Cohort	1.0 (0.9 to 1.1)	0.6 (0.6 to 0.7)	0.7 (0.6 to 0.7)
Tumor Size
<20 mm	0.7 (0.6 to 0.8)	0.3 (0.2 to 0.4)	0.8 (0.6 to 1.2)
21 to 30 mm	1.1 (0.1 to 9.5)	0.5 (0.1 to 3.7)	0.9 (0.6 to 1.2)
31 to 35 mm	0.2 (0.0 to 2.1)	0.1 (0.0 to 1.2)	0.9 (0.6 to 1.2)
31 to 50 mm	0.8 (0.7 to 0.9)	0.1 (0.0 to 0.2)	0.5 (0.3 to 0.6)

^aResults for inverse of probability treatment-weighted adjusted population shown.
CI: confidence interval; HR: hazard ratio; LT: liver transplantation; RFA: radiofrequency ablation; SR: surgical resection.
Adapted from Zhao et al (2019)¹³

Additional observational studies published since the systematic reviews have reported inconsistent results, with some finding no difference in survival outcomes between RFA and resection^14,15 and some finding resection to be superior to RFA, particularly in cases with tumor sizes measuring between 3 and 5 cm, though some studies favored resection in smaller tumors as well.^{16,17,18,19,20,21}

Reviewer: Kim et al 2025 found that resection significantly increased OS compared to RFA for patients with single tumors <3 cm and single tumors >3 cm, but not with multiple tumors. Due to the inconsistent results and a smaller population than the Zhao et al 2019 study, I am including here rather than summarizing the results in its own paragraph/table.

Section Summary: Radiofrequency Ablation to Treat Primary, Operable Hepatocellular Carcinoma
The evidence on RFA as a primary treatment for primary, operable HCC includes meta-analyses of RCTs and/or retrospective observational studies, an RCT, and additional observational studies. Numerous meta-analyses have shown that patients undergoing surgical resection experienced longer survival outcomes and lower recurrence rates than patients receiving RFA, though complication rates were higher with surgical resection. Some meta-analyses of specifically selected populations (eg, small tumor sizes or Child-Pugh Class A liver function or HCC within the Milan criteria) found that OS and DFS rates were not significantly different between RFA and surgical resection. A 2024 RCT found similar results in a similarly specifically selected population, but was limited by its sample size and single-center design. Generally results from meta-analyses were limited by heterogeneous populations and a lack of randomization leading to potential selection bias. Results from observational studies have suggested that RFA alone or RFA plus PEI could be as effective as a resection for small HCC tumors. However, other studies have found resection to be superior to RFA for survival outcomes regardless of tumor size. An exact tumor cutoff size has not been established; however, some studies have shown that survival outcomes following RFA and resection for tumors 3 cm or smaller may be similar while survival outcomes for tumors 3.1 to 5 cm may favor resection.

Radiofrequency Ablation as a Primary Treatment of Inoperable Hepatocellular Carcinoma
Clinical Context and Therapy Purpose
The purpose of RFA is to provide a treatment option that is an alternative to or an improvement on existing therapies, such as systemic therapy and other locally ablative techniques, in individuals with inoperable HCC.

The following PICO was used to select literature to inform this review.

Populations
The relevant population of interest is individuals with inoperable HCC. Examples of individuals not eligible for hepatic resection include those with inadequate liver function, presence of major vascular invasion, and presence of extrahepatic metastases.

Interventions
The therapy being considered is RFA.

Comparators
Comparators of interest include systemic therapy and other locally ablative techniques. For individuals with liver-confined disease, locoregional therapies are the preferred treatment option (eg, PEI, cryoablation, TACE, external beam radiation therapy). Systemic therapy is considered for those with advanced disease, especially if an individual has progressed after receiving locoregional therapies or if they have extrahepatic metastases. Potential first-line systemic options include sorafenib, lenvatinib, and FOLFOX (folinic acid, fluorouracil, and oxaliplatin).

Outcomes
The general outcomes of interest are OS, disease-specific survival, change in disease status, and morbid events (Table 11).

Table 11. Outcomes of Interest for Individuals with Inoperable Hepatocellular Carcinoma

Outcomes	Details
Overall survival	Survival or mortality rate [Timing: 6 months to 3 years]
Change in disease status	Local/tumor recurrence [Timing: 1 year to 3 years] Tumor progression [Timing: 1 year to 3 years]
Morbid events	Complications [Timing: peri- or post-procedure]

Study Selection Criteria
Methodologically credible studies were selected using the following principles:

To assess efficacy outcomes, comparative controlled prospective trials were sought, with a preference for RCTs;
In the absence of such trials, comparative observational studies were sought, with a preference for prospective studies.
To assess long-term outcomes and adverse events, single-arm studies that capture longer periods of follow-up and/or larger populations were sought.
Consistent with a 'best available evidence approach,' within each category of study design, studies with larger sample sizes and longer durations were sought.
Studies with duplicative or overlapping populations were excluded.

Review of Evidence
The evidence on the use of RFA as a primary treatment option for inoperable HCC includes RCTs comparing RFA with other nonsurgical interventions, RFA as an adjunct to chemotherapy, and systematic reviews of the RCTs.

Systematic Reviews
A TEC Assessment (2003) addressed RFA for the treatment of unresectable primary or metastatic liver tumors.²² Since that report, many systematic reviews and meta-analyses have assessed RFA for HCC. Several are discussed below.

Cheng et al (2023) performed a systematic review and meta-analysis of 26 studies with locally ablative therapies in patients with inoperable HCC (RFA, microwave ablation, stereotactic ablative radiotherapy, and particle radiotherapy).²³ For the primary outcome of local control, microwave ablation and particle radiotherapy showed improved outcomes compared to RFA (both p<.001). Regional progression was also significantly better with microwave ablation (p=.002) and particle radiotherapy (p=.036) compared to RFA. Distant progression was better with stereotactic ablative radiotherapy and particle radiotherapy compared to RFA (p<.001 and p=.002, respectively). The highest overall survival at 2, 3, and 4 years was with RFA, which was statistically similar to microwave ablation but superior to the other 2 therapies.

Yu et al (2021) performed a meta-analysis of RCTs comparing RFA with microwave ablation for the treatment of localized, very early- or early-stage HCC.²⁴ Five RCTs comparing RFA (n=413) and microwave ablation (n=431) were identified. The OS between microwave ablation and RFA was not significantly different at 1 year (OR, 0.705; 95% CI, 0.382 to 1.301) or 3 years (OR, 0.972; 95% CI, 0.615 to 1.538). Similarly, there was no difference observed in recurrence-free survival between microwave ablation and RFA at 1 year (OR, 1.167; 95% CI, 0.568 to 2.396) and 3 years (OR, 0.981; 95% CI, 0.616 to 1.562). Among the procedure-related complications evaluated, there were no statistically significant differences between the 2 groups.

Han et al (2020) also evaluated RFA compared with microwave ablation for early-stage HCC in a meta-analysis, but included both RCT and observational trial data.²⁵ There were 5 RCTs, 1 prospective cohort, and 20 retrospective cohorts included in the analysis, providing data for 2393 patients treated with RFA and 2003 treated with microwave ablation. The median 1-year, 3-year, and 5-year OS rates were 93.3%, 71.3%, and 57.4%, respectively, in the microwave ablation group compared with 89.5%, 68.1%, and 55.5%, respectively, in the RFA group. Pooled HR for OS did not show any difference between microwave ablation versus RFA (HR, 0.891; 95% CI, 0.740 to 1.072). There was also no difference observed between groups for DFS (HR, 1.014; 95% CI, 0.811 to 1.209).

Majumdar et al (2017) published a Cochrane review and network meta-analysis on the management of early and very early-stage HCC.²⁶ Reviewers included 14 RCTs (N=2533 patients with unresectable HCC) of nonsurgical treatments compared with each other, sham, or no intervention in patients. The quality of the evidence was rated as low or very low for all outcomes. Follow-ups ranged from 6 to 37 months. Compared with RFA, mortality was higher for percutaneous acetic acid injection (HR, 1.8; 95% CI, 1.1 to 2.8; 1 trial; n=125) and PEI (HR, 1.49; 95% CI, 1.2 to 1.9; 5 trials; n=882). No trials reported health-related quality of life.

Shen et al (2013) conducted a systematic review of 4 RCTs and quasi-RCTs (N=766 patients), comparing RFA with PEI for the treatment of HCC nodules up to 3 cm.²⁷ Overall survival was significantly longer for RFA than for PEI at 3 years (HR, 0.66; 95% CI, 0.48 to 0.90; p=.009), and local recurrence risk was lower with RFA (HR, 0.38; 95% CI, 0.15 to 0.96; p=.040). However, there was no difference in distant intrahepatic recurrence, and RFA resulted in more complications.

Tiong and Maddern (2011) conducted a systematic review of the literature from 2000 to 2010 and a meta-analysis of survival and disease recurrence after RFA for HCC.²⁸ Studies reporting on patients with HCC who were treated with RFA, either in comparison to or in combination with other interventions (eg, surgery, PEI), were eligible for inclusion. Outcomes were OS, DFS, and disease recurrence rates. Only RCTs, quasi-RCTs, and nonrandomized comparative studies with more than 12 months of follow-up were included. Forty-three articles, including 12 RCTs, were selected for review. Most articles reported on the use of RFA for unresectable HCC, often in combination with other treatments (eg, PEI, TACE, surgery). A meta-analysis of 5 RCTs showed that RFA was better than PEI, with higher OS and DFS rates. Data comparing RFA with microwave ablation were inconclusive. Reviewers concluded that RFA could achieve good clinical outcomes for unresectable HCC.

In a meta-analysis comparing RFA with cryoablation for HCC, Huang et al (2013) evaluated 3 prospective studies and 1 retrospective study.²⁹ Included in the studies were 180 RFA and 253 cryoablation patients. RFA was significantly superior to cryoablation in complication rates (OR, 2.80; 95% CI, 1.54 to 5.09), local recurrence rates (OR, 4.02; 95% CI, 1.93 to 8.39), and local tumor recurrence rates (OR, 1.96; 95% CI, 1.12 to 3.42). However, mortality rates did not differ significantly (OR, 2.21; 95% CI, 0.45 to 10.8) between groups.

Randomized Controlled Trials
Sugimoto et al (2024) conducted an RCT that compared RFA to microwave ablation in patients (N=236) with HCC tumors up to 4 cm who were not surgical candidates.³⁰ Local progression after 2 years of follow-up was lower with microwave ablation compared to RFA (16.4% vs. 30.4%; p=.007). There was no difference between groups in OS or recurrence-free survival.

Giorgio et al (2016) conducted an RCT comparing RFA plus chemotherapy with chemotherapy alone in 99 patients who had unresectable HCC invading the portal vein.³¹ The HCC nodules ranged in size from 2.1 to 6.5 cm. The primary outcome was OS at 3 years. The OS rates at 1, 2, and 3 years were 60%, 35%, and 26% in the combined therapy group and 37% and 0% at 1 and 2 years in the chemotherapy-alone arm (HR, 2.87; 95% CI, 1.61 to 5.39), respectively.

Section Summary: Radiofrequency Ablation as a Primary Treatment of Inoperable Hepatocellular Carcinoma
Randomized and nonrandomized trials have compared RFA with alternative treatments for HCC in individuals ineligible for surgery. Meta-analyses comparing RFA to other local ablative therapies have found that RFA and microwave ablation are similarly effective, that RFA is more effective than PEI, and that RFA may be better than cryoablation. The evidence comparing RFA with TACE is limited, and no conclusions can be drawn. Radiofrequency ablation has also been shown to improve survival in patients with unresectable HCC as an adjunct to chemotherapy. Overall, the evidence supports the use of RFA in patients who are inoperable.

Radiofrequency Ablation for Inoperable Hepatocellular Carcinoma as a Bridge to Liver Transplant
Clinical Context and Therapy Purpose
The purpose of RFA is to provide a treatment option that is an alternative to or an improvement on existing therapies, such as other locoregional therapies, in individuals with inoperable HCC awaiting a liver transplant.

The following PICO was used to select literature to inform this review.

Populations
The relevant population of interest is individuals with inoperable HCC awaiting a liver transplant.

Interventions
The therapy being considered is RFA.

Comparators
Comparators of interest include other locoregional therapies. Potential locoregional therapies include ablative strategies (eg, PEI, cryoablation), arterially directed therapies (eg, TACE), and radiation therapy (eg, external beam radiation therapy).

Outcomes
The general outcomes of interest are OS, disease-specific survival, and change in disease status (Table 12). The goal of receiving bridge therapy is to reduce tumor progression and the dropout rate while waiting for liver transplantation.

Table 12. Outcomes of Interest for Individuals with Inoperable Hepatocellular Carcinoma Awaiting Liver Transplant

Outcomes	Details
Overall survival	Survival rate [Timing: ≤10 years]
Disease-specific survival	Posttransplant relapse-free survival [Timing: ≤5 years]
Change in disease status	Tumor progression/de-listed rated [Timing: 3 months to 4 years] Tumor downgrading rate Posttransplant tumor recurrence Waitlist dropout rate

Study Selection Criteria
Methodologically credible studies were selected using the following principles:

To assess efficacy outcomes, comparative controlled prospective trials were sought, with a preference for RCTs;
In the absence of such trials, comparative observational studies were sought, with a preference for prospective studies.
To assess long-term outcomes and adverse events, single-arm studies that capture longer periods of follow-up and/or larger populations were sought.
Consistent with a 'best available evidence approach,' within each category of study design, studies with larger sample sizes and longer durations were sought.
Studies with duplicative or overlapping populations were excluded.

Review of Evidence
In 2002, the United Network for Organ Sharing (UNOS) introduced a new liver allocation system-Model for End-stage Liver Disease (MELD)-for adults awaiting a liver transplant; MELD was most recently reaffirmed in 2024.³² In considering how to allocate donor organs, UNOS sought to balance the risk of death on the waiting list against the risk of tumor recurrence after transplant. Under UNOS criteria, patients with T1 lesions (1 nodule ≤1.9 cm) are considered at low-risk of death while on the waiting list, and those with T3 lesions (1 nodule >5 cm, or 2 or 3 nodules with at least 1 nodule >3 cm) are at high-risk of posttransplant recurrence. Patients with T2 tumors (1 nodule 2 to 5 cm, or 2 or 3 nodules 1 to 3 cm) are more likely to die while on the waiting list than those with T1 lesions and carry an acceptable risk of post-transplant tumor recurrence. Therefore, UNOS criteria prioritize T2 HCC and makes a standardized MELD exception if the patient has an alpha-fetoprotein level >1000 ng/mL at any time or ≤1000 ng/mL and meets Milan criteria. The definition of T2 lesions is also referred to as the Milan criteria.³³Liver transplants for patients with T3 HCC are not prohibited but these patients do not receive priority on the waiting list. All patients with HCC awaiting transplantation are reassessed at 3-month intervals. Those whose tumors have progressed and are no longer T2 tumors lose allocation points.

The UNOS allocation system incentivizes the use of locoregional therapies for 2 purposes: (1) to prevent the progression of T2 tumors while on the waiting list and (2) to downsize T3 tumors to T2 status to meet the UNOS criteria for additional allocation points.

Pomfret et al (2010) summarized findings and recommendations from a national conference on outcomes of liver transplantation for patients with HCC.³⁴ The workgroup on locoregional therapy found compelling evidence that pretransplant locoregional therapy decreases waitlist dropout, especially for patients who wait more than 3 to 6 months for a transplant. The group noted that "there is a paucity of data comparing RFA with transarterial therapies for the treatment of HCC prior to liver transplant and most single-center trials have a mixture of [locoregional therapies] included in the study population" and that, while early studies have suggested a high rate of tumor seeding with percutaneous RFA, it is rare in larger series from experienced centers. The workgroup considering evidence to support the expansion of MELD criteria for patients with HCC reported wide regional variation in the risk of death for patients without HCC. The "MELD score of the non-HCC patients was quite low in some regions. Posttransplant survival in HCC patients ranged from 25% in regions with few non-HCC patients with high MELD scores to greater than 70% in regions in which there was a greater need for liver transplant (higher MELD scores) in the non-HCC population." The workgroup observed that there is extreme variability in the time to transplantation of patients with HCC in the United States, suggesting that management of patients on the waitlist and outcomes may vary. Additionally, "[c]oncern has been raised that short times to liver transplant may lead to an increase in posttransplant recurrence because the tumor biology [aggressiveness] has not had enough time to be expressed. The lack of national data on recurrence rates limits one's ability to study this national experiment of nature based on the divergent waiting times for transplantation for HCC." There was a consensus for the development of a calculated continuous HCC priority score for ranking HCC candidates on the list that would incorporate the calculated MELD score, α-fetoprotein, tumor size, and rate of tumor growth. Only candidates with at least stage T2 tumors would receive additional HCC priority points. Pomfret et al (2010) also discussed pretransplant locoregional therapy to allow patients to maintain transplant candidacy and to downstage tumors to meet MELD criteria.

Observational Studies
Radiofrequency Ablation to Prevent Tumor Progression
Several studies have reported dropout rates of waitlisted patients treated with locoregional therapy. However, lacking controlled data, it is difficult to assess the contributions of locoregional therapy to time on the waiting list. Additionally, in 2002, as previously discussed, UNOS revised its liver allocation policy, such that wait times for patients with HCC meeting the Milan criteria have now declined. Given these limitations, the following case series and cohort studies have been reported.

Lee et al (2017) reported on a 10-year intention-to-treat analysis of RFA to prevent progression and reduce the chance of posttransplant HCC.³⁵ Patients were selected for analysis if they had cirrhosis with treatment-naive HCC, were on the transplant waiting list, and had RFA as a stand-alone treatment. Only tumors that could safely be treated with a 5 mm margin received RFA. Of 1016 patients who had HCC and were on the transplant waiting list, 121 were treated with RFA and were included in this analysis. Patients returned for follow-up imaging every 3 to 6 months. The outcomes of interest were the dropout rate from the waitlist, posttransplant recurrence, and OS at 10 years. The mean time on the waiting list was 10.2 months (range, 0.3 to 38 months). At the end of follow-up, 89 (73.6%) patients had undergone a liver transplant, 16 (13.2%) were delisted, 14 (11.6%) died, and 2 (1.7%) remained on the waitlist. The number of patients delisted due to the tumor was 9 (7.4%). Intention-to-treat analysis of all patients estimated 8-year OS at 60.0% and disease-specific survival at 89.5%.

Mazzaferro et al (2004) presented 50 patients with HCC who underwent RFA while awaiting transplantation; no patient had to be removed from the waiting list due to tumor progression over a mean wait time of 9.5 months.³⁶ The median tumor size was 3 cm, and 80% of patients met the Milan criteria. Similarly, Lu et al (2005) reported on 52 patients who underwent RFA as a bridge to transplantation, 42 of whom met the Milan criteria.³⁷ After a mean of 12 months, 5.8% had dropped off the waiting list due to tumor progression.

Porrett et al (2006) retrospectively compared 31 patients treated using RFA with 33 untreated controls.³⁸ Study endpoints included OS and DFS, tumor recurrence, explant tumor viability, and the ability of magnetic resonance imaging to detect viable tumors after therapy. Both cohorts had similar demographic, radiographic, and pathologic characteristics, although untreated patients waited longer for transplantation (119 days [untreated] vs. 54 days [RFA] after MELD assignment; p=.05). Only 20% of treated tumors demonstrated complete ablation (necrosis) as defined by histologic examination of the entire lesion. Only 55% of lesions with histologic viable tumors were detected by magnetic resonance imaging after pretransplant therapy. After 36 months of follow-up, there was no difference between the treated and the untreated groups in OS (84% vs. 91%), DFS (74% vs. 85%), cancer recurrence (23% vs. 12%), or mortality from cancer recurrence (57% vs. 25%) rates, all respectively p>.1. The authors concluded that viable tumor frequently persists after pretransplant locoregional therapy, and neoadjuvant treatment does not appear to improve posttransplant outcomes in the current MELD era.

Radiofrequency Ablation to Downgrade Hepatocellular Carcinoma
Yao et al (2008) analyzed longer-term outcomes data on HCC downstaging in a cohort of 61 patients with tumor stage exceeding T2 criteria enrolled between 2002 and 2007.³⁹ Eligibility criteria for downstaging included the following: (1) 1 lesion between 5 and 8 cm; (2) 2 to 3 lesions with at least 1 lesion between 3 and 5 cm, with total tumor diameter up to 8 cm; or (3) 4 to 5 lesions with none greater than 3 cm, with total tumor diameter up to 8 cm. The main methods used were TACE and laparoscopic RFA either alone or in combination as follows: 11 patients received laparoscopic RFA alone, 14 received TACE and laparoscopic RFA, and 9 received TACE and percutaneous RFA. A minimum observation period of 3 months after downstaging was required before liver transplant. Tumor downstaging was successful in 43 patients (70.5%). Thirty-five (57.4%) patients received a liver transplant, including 2 with live-donor liver transplantation. Treatment failure was observed in 18 (29.5%) patients, primarily due to tumor progression. In the explant of 35 patients who underwent a transplant, 13 had complete tumor necrosis, 17 met T2 criteria, and 5 exceeded T2 criteria. The Kaplan-Meier intention-to-treat survival rates at 1 and 4 years after downstaging were 87.5% and 69.3%, respectively. The 1- and 4-year posttransplantation survival rates were 96.2% and 92.1%, respectively. No patient had HCC recurrence after a median posttransplantation follow-up of 25 months. The only factor predicting treatment failure was pretreatment α-fetoprotein level greater than 1000 ng/mL. From this small series, the authors concluded that successful downstaging could be achieved with excellent posttransplant outcomes.

Yao et al (2005) also reported on a case series of 30 patients with HCC who underwent locoregional therapy specifically to downstage tumors to meet the University of California San Francisco (UCSF) criteria (see below for brief discussion of the UCSF criteria).⁴⁰ Eligibility for locoregional therapy seeking to downstage patients included either (1) 1 nodule between 5 and 8 cm in diameter; (2) 2 or 3 nodules with at least 1 between 3 and 5 cm in diameter, with a sum of diameters no greater than 8 cm; or (3) 4 or 5 nodules all 3 cm or less, with a sum of diameters less than 8 cm. Among the 30 patients, 21 (70%) met the criteria for locoregional therapy and 16 of them were successfully downstaged and underwent transplantation. No tumors recurred at a median follow-up of 16 months. The authors concluded that downstaging could be successfully achieved in most patients but that data on tumor recurrence required longer follow-up.

Radiofrequency Ablation to Reduce Risk of Recurrence
An additional indication for locoregional therapies has focused on their use to reduce the incidence of recurrence posttransplant. If the incidence of recurrence can be reduced, then advocates have argued that the UNOS allocation criteria should not discriminate against patients with larger tumors.^{41,42,43,44,45} Some patients with T3 lesions are cured with a liver transplant, although most experience tumor recurrence. For example, in the seminal study, Mazzaferro et al (1996)³³ reported that 4-year recurrence-free survival was 92% in those who met the Milan criteria compared with 59% in those who did not; additional studies have confirmed this difference in the recurrence-free survival rate.⁴⁰ However, other institutions have reported similar outcomes with expanded criteria. For example, Yao et al (2002) reported similar recurrence-free survival rates after transplant in patients with T2 tumors and a subset of those with T3 tumors.⁴³ This T3 subset was defined as a single lesion 6.5 cm or less or 3 or fewer lesions with none greater than 3 cm and with a sum of tumor diameters of 8 cm or less. These expanded criteria are known as the UCSF criteria.

The question is whether locoregional therapies (including both RFA and chemoembolization) decrease the recurrence rate in patients meeting the UCSF criteria. The authors also compared the recurrence-free survival rates of those who did and did not receive locoregional therapy. For those with T2 lesions, recurrence rates were similar whether or not the patient received locoregional therapy. However, for T3 lesions (including both T3A and T3B), the 5-year recurrence-free survival rate was 85.9% for those who received locoregional therapy compared with 51.4% for those who did not. When data for T2 and T3 lesions were pooled, the 5-year recurrence-free survival rate was 93.8% for those who received locoregional therapy and 80.6% for those who did not. The authors concluded that preoperative locoregional therapy might confer a survival benefit in those with T2 or T3 lesions.

The authors noted several study limitations, including the retrospective nature of the data and the marginal statistical significance of the improved survival, given the small numbers of patients in each subgroup. For example, only 19 patients were in the T3A (ie, UCSF expanded criteria) subgroup. Additionally, no protocol specified which type of locoregional therapy to offer different patients. These therapies are only offered to patients with adequate liver reserve; such patients may have an improved outcome regardless of the preoperative management.

In the 2017 study by Lee et al (2017; described above), of 89 patients with HCC who received RFA before the liver transplant, 5 (5.6%) had HCC recurrence.³⁵

Section Summary: Radiofrequency Ablation for Inoperable Hepatocellular Carcinoma as a Bridge to Liver Transplant
Evidence on the use of RFA for HCC in patients awaiting transplant consists of case series and uncontrolled trials. There is sufficient evidence to conclude that locoregional therapy with RFA or alternatives decreases the dropout rate from the transplant list. This is especially true if patients wait more than 3 to 6 months for a transplant. Therefore, outcomes are improved for this group.

For other uses of RFA in patients awaiting transplant, such as to downgrade tumors for eligibility for transplant, and/or to prevent disease recurrence, the evidence is insufficient to make conclusions.

Radiofrequency Ablation for Inoperable Hepatic Metastases of Colorectal Origin
Clinical Context and Therapy Purpose
The purpose of RFA is to provide a treatment option that is an alternative to or an improvement on existing therapies, such as chemotherapy, other locally ablative techniques, and the best supportive care, in individuals with inoperable hepatic metastases of colorectal origin.

The following PICO was used to select literature to inform this review.

Populations
The relevant population of interest is individuals with inoperable hepatic metastases of colorectal origin.

Interventions
The therapy being considered is RFA.

Comparators
Comparators of interest include chemotherapy, other locally ablative techniques (eg, microwave ablation, cryoablation, or elecro-coagulation), and the best supportive care.

Outcomes
The general outcomes of interest are OS, disease-specific survival, symptoms, change in disease status, morbid events, quality of life, and treatment-related morbidity (Table 13).

Table 13. Outcomes of Interest for Individuals with Inoperable Hepatic Metastases of Colorectal Origin

Outcomes	Details
Overall survival	Survival or mortality rate [Timing: 30 days to 9.7 years]
Disease-specific survival	Disease-free survival [Timing: 30 days to 5 years]
Change in disease status	Progression-free survival [Timing: ≤5 years] Recurrence rate [Timing: ≤5 years]

Study Selection Criteria
Methodologically credible studies were selected using the following principles:

To assess efficacy outcomes, comparative controlled prospective trials were sought, with a preference for RCTs;
In the absence of such trials, comparative observational studies were sought, with a preference for prospective studies.
To assess long-term outcomes and adverse events, single-arm studies that capture longer periods of follow-up and/or larger populations were sought.
Consistent with a 'best available evidence approach,' within each category of study design, studies with larger sample sizes and longer durations were sought.
Studies with duplicative or overlapping populations were excluded.

Review of Evidence
More than half of patients with colorectal cancer (CRC) will develop liver metastases, generally with a poor prognosis.⁴⁶ A median survival of 21 months has been observed in patients with a single CRC liver metastasis; those with several unilobar lesions have a median survival of 15 months, and those with disseminated metastases have a median survival of less than 1 year. A number of first-line systemic chemotherapy regimens have been used to treat metastatic CRC, with a 2-year survival rate of 25% for those treated with 5-fluorouracil or 5-fluorouracil plus leucovorin.⁴⁶ With the introduction of newer agents (eg, irinotecan, oxaliplatin) and targeted drugs (eg, cetuximab, bevacizumab), 2-year survival rates have increased to between 30% and 39%, with marked improvement in OS. Because the liver is often the only site of metastases from CRC, locoregional therapies have been investigated. Surgical resection is considered the criterion standard for treatment of CRC liver metastases, with 5-year OS rates that historically range from 28% to 38%, but may reach 58% in appropriately selected, resectable patients without the widely disseminated disease^.47,48 However, only 10% to 25% of patients with CRC metastases are eligible for surgical resection because of the extent and location of the lesions within the liver or because of the presence of comorbid conditions or disseminated disease. Unresectable cases or cases in which surgery is contraindicated typically are treated with systemic chemotherapy, with poor results and considerable adverse events. Alternatively, RFA has been proposed to treat metastatic CRC in the liver.

Systematic Reviews
A meta-analysis by Meijerink et al (2018) compares RFA and microwave ablation to systemic chemotherapy and to partial hepatectomy (PH) for the treatment of colorectal liver metastases.⁴⁹ Forty-eight articles were identified, most of which were observational studies and case series, although 2 RCTs and 8 systematic reviews were included. The authors found 18 observational studies of very low quality that looked at RFA alone compared to PH alone or PH plus RFA. For OS, their analysis concluded that PH alone was superior to RFA alone (HR, 1.78; 95% CI, 1.35 to 2.33). The meta-analysis for 30-day mortality comparing RFA alone to PH alone showed no difference between the 2 interventions (risk ratio [RR], 0.64; 95% CI, 0.21 to 1.95). Disease-free survival was higher for PH alone over RFA alone (HR, 1.49; 95% CI, 1.23 to 1.81), as well as local progression-free survival (HR, 5.36; 95% CI, 1.64 to 17.52). However, complication rates were lower for RFA alone than for PH alone (RR, 0.47; 95% CI, 0.28 to 0.78). One limitation of this review is that the included observational studies were all confounded by indication because RFA was only performed on unresectable lesions. Observational studies are also at increased risk for publication bias.

In a Health Technology Assessment, Loveman et al (2014) found insufficient evidence to draw conclusions on the clinical effectiveness of ablative therapies, including RFA, for liver metastases.⁵⁰

Weng et al (2012) reported on a meta-analysis comparing RFA with liver resection for the treatment of CRC liver metastases.⁵¹ One prospective study and 12 retrospective studies were included in the analysis. Overall survival at 3 and 5 years was significantly longer after liver resection than after RFA (RR, 1.38; 95% CI, 1.25 to 1.52 vs. RR, 1.47; 95% CI, 1.28 to 1.69, respectively). Disease-free survival was also significantly longer after liver resection than after RFA at 3 and 5 years (RR, 1.73; 95% CI, 1.48 to 2.03; RR, 2.23; 95% CI, 1.82 to 2.72, respectively). While postoperative morbidity with liver resection was significantly higher than with RFA (RR, 2.49; 95% CI, 1.88 to 3.31), mortality did not differ significantly between treatments. Liver resection also produced significantly better outcomes than RFA when data were analyzed in 3 subgroups: tumors less than 3 cm, solitary tumor, and open or laparoscopic approach. However, hospital stays were significantly shorter (9.2 days vs. 3.9 days; p<.01) and rates of complications lower (18.3% vs. 3.9%; p<.01) with RFA than with liver resection. Interpretation of the meta-analysis was limited by the retrospective design of most studies.

A systematic review by Pathak et al (2011) assessed the long-term outcome and complication rates of various ablative therapies used in the management of colorectal liver metastases.³²The literature search was from 1994 to 2010, and inclusion criteria were a minimum of 1-year follow-up and a sample size greater than 10 patients. In all, 75 met inclusion criteria. Most studies were single-arm, single-center, and retrospective or prospective. There was wide variability in patient groups, adjuvant therapies, and management approaches within individual studies. Several studies combined results for colorectal and non-colorectal metastases, often reporting combined outcomes. The endpoints were not reported uniformly, with varying definitions of survival time, recurrence time, and complication rates. Cryotherapy (26 studies) had local recurrence rates ranging from 12% to 39%, with mean 1-, 3-, and 5-year survival rates of 84%, 37%, and 17%, respectively. Major complication rates ranged from 7% to 66%. Microwave ablation (13 studies) had local recurrence rates ranging from 5% to 13%, with mean 1-, 3-, and 5-year survival rates of 73%, 30%, and 16%, respectively, and major complication rates ranging from 3% to 16%. Radiofrequency ablation (36 studies) had local recurrence rates ranging from 10% to 31%, with mean 1-, 3-, and 5-year survival rates of 85%, 36%, and 24%, respectively, and major complication rates ranging from 0% to 33%. Reviewers concluded that ablative therapies offer significantly improved survival compared with palliative chemotherapy alone, with 5-year survival rates ranging from 17% to 24%, and that complication rates of commonly used techniques are low.

A review by Guenette and Dupuy (2010) summarized the literature on the use of RFA for colorectal hepatic metastases.⁵² Seventeen studies with more than 50 patients treated with RFA for colorectal hepatic metastases reported survival. Average tumor size, reported in 15 studies, ranged from 2.1 to 4.2 cm. Five-year OS rates, reported in 12 studies, ranged from 2% to 55.3% (mean, 24.5%). The largest study series (Lencioni et al [2004]⁴⁸) included in the review consisted of 423 patients, with average tumor size of 2.7 cm, 4 or fewer metastases, each 5 cm or less at greatest dimension, and no extrahepatic disease. Overall survival rates in that study at 1, 3, and 5 years were 86%, 47%, and 24%, respectively. Guenette and Dupuy concluded that 5-year survival rates following RFA were similar to those following resection, but that long-term data associated with RFA and colorectal hepatic metastases were sparse, as randomized trials had failed recruitment, and patients with the resectable disease should undergo resection if possible. However, given the efficacy of RFA compared with chemotherapy alone, they noted that RFA should be considered a primary treatment option for patients with unresectable disease.

Randomized Controlled Trials
Ruers et al (2012, 2017) published the results of a multicenter RCT that compared RFA plus systemic treatment with systemic treatment alone for unresectable colorectal liver metastases.^53,54This RCT, originally designed as a phase 3 study, was completed as a phase 2 study due to slow accrual (N=119). To be included in the trial, patients had to have nonresectable liver metastases with fewer than 10 nodes and without extrahepatic disease. In the experimental arm, RFA, with or without additional resection, was given in combination with systemic therapy. The primary endpoint was a 30-month survival greater than 38% in the experimental arm based on intention-to-treat analysis. At 3 years, OS did not differ significantly between groups (see Table 14). However, there was a significant improvement in progression-free survival (HR, 0.74; 95% CI, 0.42 to 0.95; p=.03) at 3 years, with 10.6% in the systemic therapy arm and 27.6% in the combined treatment arm. At a median follow-up of 9.7 years, 39 (65%) of 60 patients in the combined treatment arm had died compared with 53 (89.8%) of 59 in the systemic treatment arm (HR, 0.58; 95% CI, 0.38 to 0.88; p=.01).

Table 14. Percent Overall Survival at 3, 5, and 8 Years

Treatment	3 Years (95% CI), %	5 Years (95% CI), %	8 Years (95% CI), %
Combined treatment	56.9 (43.3 to 68.5)	43.1 (30.3 to 55.3)	35.9 (23.8 to 48.2)
Systemic alone	55.2 (41.6 to 66.9)	30.3 (19.0 to 42.4)	8.9 (3.3 to 18.1)

Ruers et al (2017).⁵⁴
CI: confidence interval.

Nonrandomized Comparative Studies
Nonrandomized studies have compared RFA with resection or systemic chemotherapy in patients with localized CRC metastases and no evidence of additional metastatic disease.

Kong et al (2025) conducted a retrospective evaluation of 157 patients with inoperable CRC metastases who received chemotherapy with or without RFA.⁵⁵ Median follow-up was 38 months in the RFA + chemotherapy group, and 23.9 months in the chemotherapy only group. Both OS (34.9 vs. 21.2 months; p<.0001) and progression-free survival (17.16 vs. 8.35 months; p=.00064) were improved in the group that received RFA compared to the group that received chemotherapy only (after propensity score matching). Results without propensity score matching were consistent (OS, p<.0001; progression-free survival, p<.0001).

Sarioglu et al (2024) conducted a single center, retrospective study of RFA versus microwave ablation for patients (N=242) with CRC liver metastases.⁵⁶ Local recurrence among patients with at least 1 year of follow-up was 29% in the RFA group and 13% in the microwave ablation group (p<.001). Survival analysis among a matched cohort found that local recurrence-free survival was higher with microwave ablation than with RFA (HR, 1.87; 95% CI, 1.30 to 2.68; p=.0005).

Hof et al (2016) analyzed data from 431 patients in an institutional database.⁵⁷ All patients underwent locoregional treatment for hepatic metastases from CRC. Initial treatment was either hepatic resection (n=261), open RFA (n=26), percutaneous RFA (n=75), or a combination of resection plus RFA (n=69). Mean follow-up was 38.6 months. The overall recurrence rate was 83.5% (152/182) in patients treated with RFA compared with 66.6% (201/302) in patients treated with hepatic resection (p<.001). The 5-year OS estimate by Kaplan-Meier analysis was 51.9% for RFA and 53% for hepatic resection (p=.98).

Abdalla et al (2004) examined recurrence and survival rates for clinically similar patients treated with hepatic resection only (n=190), resection plus RFA (n=101), RFA only (n=57), open laparotomy with biopsy or systemic chemotherapy alone (n=70).⁵⁸ In the key relevant comparison, RFA versus chemotherapy in chemotherapy-naive patients with nonresectable CRC metastases (median, 1 lesion per patient; range, 1 to 8; median tumor size, 2.5 cm), OS at 4 years was 22% in the RFA group and 10% in the chemotherapy group (p=.005). Median survival was estimated at 25 months in the RFA group and 17 months in the chemotherapy group (p-value not reported). Recurrence at a median follow-up of 21 months was 44% in the RFA group and 11% in the resection-only group (p<.001), although the proportion of patients with distant recurrence as a component of failure was similar (41% resection vs. 40% RFA, p-value not significant).

A consecutive series by Ruers et al (2007) of well-defined, previously untreated patients (N=201) without extrahepatic disease underwent laparotomy to determine the therapeutic approach.⁵⁹ Three groups were identified: patients amenable to hepatic resection (n=117); patients amenable to resection plus local ablation (RFA, n=27; cryoablation, n=18); and patients deemed unresectable and ineligible for local ablation (n=39) who received systemic chemotherapy. Median OS was 61 months (95% CI, 41 to 81 months) in resected patients (median, 1 tumor per patient; range, 1 to 9; median diameter, 3.8 cm), 31 months (95% CI, 20 to 42 months) in locally ablated patients (median, 4 tumors per patient; range, 1 to 19; median diameter, 3 cm), and 26 months (95% CI, 17 to 35 months) in the chemotherapy patients (median, 4 tumors per patient; range, 1 to 17; median diameter, 4 cm; p=.052, ablated vs. chemotherapy). Results from 2 validated quality of life instruments (EuroQol-5D, European Organization for Research and Treatment of Cancer core questionnaire [EORTC QLQ C-30]) showed that patients treated with local ablation returned to baseline values within 3 months, whereas those treated with chemotherapy remained significantly lower (ie, worse quality of life) than the baseline over 12 months posttreatment (p<.05).

Van Tilborg et al (2011) reported on long-term results for 100 patients with unresectable colorectal liver metastases who underwent a total of 126 RFA sessions (237 lesions).⁶⁰ Lesion size ranged from 0.2 to 8.3 cm (mean, 2.4 cm). Mean follow-up was 29 months (range, 6 to 93 months). Major complications (including abscess, hemorrhage, grounding pad burns, and diaphragm perforation) occurred in 8 patients. Factors that determined procedural success included lesion size and the number and location of the lesions. Local tumor site recurrence was 5.6% for tumors less than 3 cm, 19.5% for tumors 3 to 5 cm, and 41.2% for those greater than 5 cm. Centrally located lesions recurred more often than peripheral (21.4% vs. 6.5%, respectively; p=.009). Mean survival from the time of RFA was 56 months (95% CI, 45 to 67 months).

Section Summary: Radiofrequency Ablation for Inoperable Hepatic Metastases of Colorectal Origin
There are no RCTs comparing RFA with alternative treatments for patients with unresectable colorectal liver metastases. However, an RCT of RFA combined with chemotherapy found improved survival at 8 years compared with chemotherapy alone. Additionally, prospective studies have demonstrated that OS following RFA is at least equivalent and likely better than that obtained with currently accepted systemic chemotherapy in well-matched patients with unresectable hepatic metastatic CRC who do not have extrahepatic disease. Results from a number of case series have also suggested RFA of hepatic CRC metastases produces long-term survival that is at least equivalent and likely superior to systemic chemotherapy, compared with historical outcomes. Evidence from a comparative study has suggested RFA has fewer deleterious effects on quality of life than chemotherapy and that RFA patients recover their quality of life significantly faster than chemotherapy patients. Patient selection bias may partially explain the better outcomes in the case series because patients chosen to receive RFA might have had better prognoses than patients given chemotherapy.

Radiofrequency Ablation for Inoperable Hepatic Metastases of Neuroendocrine Origin
Clinical Context and Therapy Purpose
The purpose of RFA is to provide a treatment option that is an alternative to or an improvement on existing therapies, such as chemotherapy, other locally ablative techniques, and the best supportive care, in individuals with inoperable hepatic metastases of neuroendocrine origin.

The following PICO was used to select literature to inform this review.

Populations
The relevant population of interest is individuals with inoperable hepatic metastases of neuroendocrine origin.

Interventions
The therapy being considered is RFA.

Comparators
Comparators of interest include chemotherapy, other locally ablative techniques (eg, cryoablation), and the best supportive care.

Outcomes
The general outcomes of interest are OS, disease-specific survival, symptoms, change in disease status, morbid events, quality of life, and treatment-related morbidity (Table 15).

Procedure and diagnosis codes on Medical Policy documents are included only as a general reference tool for each policy. They may not be all-inclusive.

This medical policy was developed through consideration of peer-reviewed medical literature generally recognized by the relevant medical community, U.S. FDA approval status, nationally accepted standards of medical practice and accepted standards of medical practice in this community, Blue Cross Blue Shield Association technology assessment program (TEC) and other nonaffiliated technology evaluation centers, reference to federal regulations, other plan medical policies, and accredited national guidelines.

History From 2024 Forward

08/01/2025	Annual review, updating rationale and references.
08/23/2024	Annual review, no change to policy intent.
01/01/2024	New Policy.

Policy List

Edit Policy

{{ navigationConstituentPage.title || { 'medicalpolicyhb': 'Healthy Blue Medical Policies', 'medicalpolicyih': 'InStil Health Medical Policies' }[navigationSitePage.urlName] || 'Medical Policies' }}

Radiofrequency Ablation of Primary or Metastatic Liver Tumors - CAM 70191HB