Meta-analysis of cryoablation versus radiofrequency ablation in the treatment of malignant liver tumors

Abstract

Objective

A meta-analysis was conducted to assess the efficacy and safety of cryoablation (CRA) compared with radiofrequency ablation (RFA).

Methods

A systematic search of PubMed, EMBASE, Cochrane Library, Wanfang, CNKI, and VIP databases was conducted to identify clinical controlled studies comparing CRA versus RFA for hepatic malignancies up to July 2022. The meta-analysis was performed using RevMan 5.3.

Results

A comprehensive analysis was conducted on 8 clinical controlled studies involving a total of 943 patients. There were no significant differences in the incidence of complications, complete ablation of lesions, local recurrence, and 1-year survival between the CRA and RFA groups (OR = 0.98, 95%CI: 0.61–1.55, p = 0.92; OR = 1.08, 95%CI: 0.62–1.90, p = 0.78; OR = 1.28, 95%CI: 0.49–3.36, p = 0.61; and OR = 1.14, 95%CI: 0.63–2.06, p = 0.66, respectively).

Conclusion

The efficacy and safety profile of CRA was comparable to that of RFA in the context of ablation therapy for hepatic malignancies. These findings suggested that CRA may be a valuable alternative to RFA in the treatment of hepatic malignancies.

Keywords:

• Malignant liver tumor
• cryoablation
• radiofrequency ablation
• meta-analysis

Introduction

Liver cancer is the fifth most common cancer and the second leading cause of cancer-related death worldwide, with hepatocellular carcinoma (HCC) accounting for 90% of primary liver cancers [1]. In addition, the liver is a major primary site for the development of metastatic disease and the leading cause of death from gastrointestinal malignancies such as colon, gastric, and pancreatic cancer as well as melanoma, breast, and sarcoma [2]. Surgical resection of HCC and liver metastases has been shown to provide a substantial long-term survival benefit for patients [1,3]. However, for those with primary and metastatic liver malignancies that are not suitable for surgical resection, it is essential to investigate novel therapeutic options to manage and potentially cure the disease. The rapid advancement in minimally invasive interventional therapy and imaging medicine has led to the widespread adoption of percutaneous ablation technology in the treatment of liver malignant tumors [4,5].

At present, percutaneous radiofrequency ablation (RFA) is currently the most widely used ablation therapy for HCC (According to EASL Clinical Practice Guidelines: evidence high; recommendation strong)^[1]. At the same time, percutaneous cryoablation (CRA) is a promising local ablation technique. Compared with RFA, CRA can provide accurate intraoperative monitoring of lesions by a variety of imaging techniques, including computed tomography (CT), magnetic resonance imaging (MRI), and ultrasound (US) [6,7]. Currently, there is no consensus in the academic community regarding whether CRA or RFA is the preferred local treatment modality for malignant liver tumors. In light of this ambiguity, the objective of our research is to employ a comprehensive meta-analysis method to investigate and compare the differences in efficacy and safety between CRA and RFA for treating malignant liver tumors. This endeavor is to provide a more robust basis for evidence-based medicine.

Materials and methods

Literature search

PubMed, EMBASE, Cochrane Library, Wanfang, CNKI, and VIP databases were searched for relevant Chinese and English articles. Keywords: ‘radiofrequency ablation’, ‘RFA’, ‘cryoablation’, ‘cryoablation ablation’, ‘cryosurgery’, ‘cryotherapy’, ‘CRA’, ‘thermal ablation’, ‘hepatic carcinoma’, ‘HCC’, ‘hepatoma’, ‘liver carcinoma’, ‘liver neoplasms’, ‘liver tumors’, ‘liver malignancies’, ‘liver cancer’ and ‘cancer of liver’. At the same time, the references of relevant articles were also searched manually. The search time was from the establishment of the database to July 31, 2022.

Inclusion and exclusion criteria

Inclusion Criteria: The following criteria were used to select studies for this study: (1) the full text of the articles was written in English or Chinese; (2) the study design was limited to RCT and cohort study; (3) the study population was patients with liver malignancies; (4) studies comparing percutaneous CRA versus percutaneous RFA were included, with outcome measures including rates of complications, complete ablation, local recurrence, and survival. Exclusion Criteria: The present analysis excluded the following types of literature: (1) case reports, re-published articles, conferences, comments, letters, abstracts, and review articles; (2) studies that were incomplete or lacked the necessary data for the study; (3) studies that utilized multiple ablation techniques or combined them with other treatments simultaneously.

Quality evaluation

The quality of RCTs was evaluated utilizing the Cochrane Collaborative Network Bias Risk Assessment Tool. The assessment considered the potential sources of bias including the method of random assignment, allocation concealment, blinding status, completeness of outcome data, selective reporting of study results, and other sources of bias [8]. Newcastle–Ottawa Scale (NOS) was used to evaluate the quality of cohort studies, and a score greater than or equal to 5 points was considered a high-quality study [9].

Data extraction

Data extraction from the included studies was performed independently by two investigators. In the event of any discrepancies, the consensus was reached through consultation. The extracted data comprised the following information: authors, year of publication, type of study design, location (country or region), patient age, number of patients, number of tumors, tumor size (the length of the tumor), major complications, complete ablation rate, local recurrence rate, and survival rate. Major complications were defined as events that prolonged the patient’s hospital stay or required a higher level of care, including subcapsular/intrahepatic hematoma, biliary tract system injury, hemothorax, pleural effusion, hepatic infarction, portal thrombosis, and abscess, etc. The complete ablation rate was determined as the absence of any enhancement within the enhanced CT/MR ablation area within four weeks post-surgery. Local recurrence was defined as the presence of tumor enhancement near the enhanced CT/MR ablation area during the follow-up period.

Statistical analysis

Data analysis for the meta-analysis was conducted using the Rev Man 5.3 statistical software. The level of statistical heterogeneity among the included studies was assessed using the I² test (where I² values greater than 50% indicated heterogeneity) and the chi-square test (where P values less than 0.10 indicated heterogeneity). If the results were homogenous, a fixed-effects model was utilized for analysis. Conversely, if heterogeneity was present, a random-effects model was employed. The odds ratio (OR) and 95% confidence interval (CI) were used to evaluate the incidence of complications, complete ablation rate, local recurrence rate, and 1-year survival rate. A P value less than 0.05 was considered statistically significant. A sensitivity analysis was performed by removing each included study sequentially. Publication bias was tested using Egger’s method.

Result

Search result

A comprehensive search yielded 1519 studies. According to the inclusion and exclusion criteria, one [10] randomized controlled trial (RCT) study and seven [11–17] cohort studies were selected for analysis. (Figure 1). The studies were further categorized into those focused solely on HCC (n = 6) and those addressing liver malignancies, including both HCC and liver metastases (n = 2). A total of 943 patients were included in the study, consisting of 466 in the CRA group, 477 in the RFA group. The characteristics of the included studies were summarized in Tables 1 and 2.

Figure 1. Flowchart of study inclusion.

Table 1. Characteristics of every study included in the meta-analysis.

Author	Country/year	Study design	Group	Patient age (years)	Number of Patients	Number of Tumors	Tumor size(cm)
Shen [19]	China/2015	Cohort	CRA	52.3 ± 11.2	52	84	0.7–5.9
			RFA	50.7 ± 12.4	56	100	0.8–6.1
Li [4]	China/2019	Cohort	CRA	53 ± 9	54	63	2.7 ± 1.0
			RFA	56 ± 10	65	78	2.5 ± 0.9
Adam [11]	France/2002	Cohort	CRA	60.1 ± 9.6	31	31	2.22 ± 1.05
			RFA	63.5 ± 9.9	33	33	2.8 ± 1.67
Wang [10]	China/2015	RCT	CRA	53.8 ± 9.587	180	199	≤4
			RFA	53.3 ± 8.905	180	189	≤4
Ko [14]	Korea/2020	Cohort	CRA	60.9 ± 9.7	25	NR	1.5 ± 0.4(1–2.5)
			RFA	59.6 ± 10	31	NR	2 ± 0.4(1.3–2.8)
Dunne [13]	USA/2014	Cohort	CRA	67.8 ± 10.7	25	NR	2.8(1.5–4.9)
			RFA	64.4 ± 10.8	22	NR	2(0.4–6.3)
Cha [12]	Korea/2019	Cohort	CRA	61(44–84)	61	NR	1.3(0.6–2.0)
			RFA	70(44–80)	50	NR	1.3(0.7–2.7)
Pan [17]	China/2018	Cohort	CRA	62 ± 5	38	42	1.1–5.8
			RFA	59 ± 6	40	44	1.0–5.4

CRA: cryoablation; RFA: radiofrequency ablation: NR: not reported.

Table 2. Outcome measures of every study included in the meta-analysis.

Author	Country/year	Study design	Group	Complication	Complete ablation rate	Local recurrence rate	1-year survival rate
Shen [19]	China/2015	Cohort	CRA	0/52	82/84	6/52	92.30%
			RFA	4/56	90/100	8/56	88.10%
Li [4]	China/2019	Cohort	CRA	6/54	89%	NR	87.7%
			RFA	3/65	94%	NR	93.6%
Adam [11]	France/2002	Cohort	CRA	9/31	69%	53%	73%
			RFA	8/33	76%	18%	60%
Wang [10]	China/2015	RCT	CRA	7/180	NR	5.6%	97%
			RFA	6/180	NR	10%	97%
Ko [14]	Korea/2020	Cohort	CRA	0/25	NR	NR	NR
			RFA	2/31	NR	NR	NR
Dunne [13]	USA/2014	Cohort	CRA	7/25	NR	NR	NR
			RFA	3/22	NR	NR	NR
Cha [12]	Korea/2019	Cohort	CRA	8/61	NR	11/61	NR
			RFA	14/50	NR	6/50	NR
Pan [17]	China/2018	Cohort	CRA	3/38	76.2%	NR	NR
			RFA	0/40	75%	NR	NR

CRA: cryoablation; RFA: radiofrequency ablation; NR: not reported.

Literature quality evaluation

One RCT was conducted, in which the method of the random assignment was specified, loss to follow up, and withdrawals were reported. The results of the Cochrane Collaborative Network Bias Risk Assessment Tool for this trial were summarized in Table 3. NOS scores for 7 cohort studies were summarized in Table 4.

Table 3. Assessment of randomized study quality.

Trial, year	Random sequence production	Allocation concealment	Blind method	Withdrawals	Score
Wang 2015 [10]	Adequate	Unclear	Inadequate	Description	4

Table 4. Assessment of nonrandomized study quality.

Trial, year Quality indicators from Newcastle–Ottawa scale
	1	2	3	4	5A	5B	6	7	8	Score
Shen 2015 [19]	Yes	Yes	Yes	NO	Yes	NO	Yes	Yes	Yes	7
Li 2019 [4]	Yes	Yes	NO	Yes	Yes	Yes	Yes	NO	Yes	7
Adam 2002 [11]	Yes	Yes	Yes	Yes	NO	Yes	Yes	Yes	Yes	8
Ko 2020 [14]	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	9
Dunne 2014 [13]	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	9
Cha 2019 [12]	Yes	Yes	Yes	Yes	NO	Yes	Yes	Yes	Yes	8
Pan 2018 [17]	Yes	Yes	Yes	NO	Yes	Yes	Yes	NO	Yes	7

For cohort studies, CRA group was defined as exposed cohort; RFA group was defined as nonexposed cohort. 1 indicates exposed cohort truly representative; 2, nonexposed cohort drawn from the same community; 3, ascertainment of exposure; 4, outcome of interest not present at start; 5 A, cohorts comparable on basis of age; 5B, cohorts comparable on other factor (s); 6, quality of outcome assessment; 7, follow-up long enough for outcomes to occur; and 8, complete accounting for cohorts.

Results of meta-analysis

Complication rate

Eight studies reported complication rates for CRA versus RFA [10–17], and the heterogeneity test revealed no significant difference (I² = 41%, p = 0.10). The analysis was conducted using the fixed-effect model. There was no significant difference in complication rate between the two groups (OR = 0.98, 95% CI: 0.61–1.55, p = 0.92, Figure 2). No significant publication bias was demonstrated in the Egger-s test (bias = 0.15, 95% CI: −2.97 to 3.37, p = 0.882).

Figure 2. Forest plot for complications between CRA and RFA.

Complete ablation rate

Four studies reported the complete ablation rates of CRA and RFA [11,15–17], and the heterogeneity test revealed no significant difference (I²=42%, p = 0.16). The fixed-effect model was employed for the analysis, and no statistically significant difference was observed in the complete ablation rates between the two groups (OR = 1.08, 95%CI: 0.62–1.90, p = 0.78, as shown in Figure 3). No significant publication bias was demonstrated in the Egger-s test (bias = 0.83, 95% CI: −14.32 to 21.15, p = 0.495).

Figure 3. Forest plot for complete ablation rate between CRA and RFA.

Local recurrence rate

Four studies reported the local recurrence rates for CRA versus RFA [10–12,16], and the test for heterogeneity was statistically significant (I²=72%, p = 0.01). The analysis was conducted using the random-effects model. There was no statistically significant difference in the local recurrence rate between the two groups (OR = 1.28, 95% CI: 0.49–3.36, p = 0.61, Figure 4). No significant publication bias was demonstrated in the Egger’s test (bias = 1.49, 95% CI: −14.61 to 30.13, p = 0.274). The sensitivity analyses displayed substantial heterogeneity disappeared (I²=25%, p = 0.27) when Adam et al.’s [9] study was excluded. In this case, a fixed-effect model was employed and the OR was not statistically significant (OR = 0.79; 95% CI 0.46–1.36; p = 0.39).

Figure 4. Forest plot for local recurrence rate between CRA and RFA.

1-year survival rate

Four studies reported postoperative 1-year survival for CRA versus RFA [10,11,15,16], and the test for heterogeneity was not statistically significant (I² = 11%, p = 0.34). The analysis was conducted using the fixed-effects model. There was no statistically significant difference in the postoperative 1-year survival between the two groups (OR = 1.14, 95% CI: 0.63–2.06, p = 0.66, see Figure 5). No significant publication bias was demonstrated in the Egger-s test (bias = −1.03, 95% CI: −33.78 to 20.76, p = 0.412).

Figure 5. Forest plot for 1-year survival rate between CRA and RFA.

Discussion

Primary liver cancer and liver metastases are major causes of morbidity and cancer-related mortality on a global scale [18]. The available treatments for these conditions include surgical resection, transarterial chemoembolization, and ablation [19]. As medical technology and instruments continue to advance, percutaneous ablation technology has gained significant recognition and adoption in clinical practice [20]. Among them, RFA has emerged as the mainstay of treatment for local control of HCC, colorectal metastatic cancer, and other malignancies. This is due to its convenience of operation, short surgery time, lower cost, good controllability of the ablation range, and survival rates comparable to surgical interventions [10,12,13]. However, the available literature has reported that RFA has a relatively high local recurrence rate of up to 50% for tumors with a diameter greater than 3 cm and/or those located adjacent to blood vessels [21]. With the continuous advancement of CRA technology, which includes the development of argon-helium cryo-devices with CRA fine needles, a variety of imaging techniques (CT, MR, US) have been applied to intraoperative monitoring, as well as the accumulation of clinical experience, CRA has become a potential technology for local ablation treatment of liver malignant tumors [7]. According to studies, CRA has been found to cause less pain than other thermal-based ablation techniques [10]. Moreover, several guidance modalities such as ultrasound, CT, and MR imaging can be used to visualization of the ablation zone. This provides advantages such as safer treatment for tumors located close to critical structures [13].

In this study, we conducted a systematic review and meta-analysis of published studies on percutaneous CRA and percutaneous RFA. Our findings indicated that there were no significant differences in the outcome measures of postoperative complications, complete ablation rate, local recurrence rate, and 1-year survival rate between CRA and RFA.

This study’s findings indicate that there was no statistically significant difference in the incidence of complications between the CRA group and the RFA group (OR = 0.98, 95%CI: 0.61-1.55, p = 0.92). These results suggest that CRA and RFA have similar safety profiles in the treatment of malignant liver tumors. However, it is worth noting that the incidence of complications reported in different studies varies. Cha et al. [12] indicate that compared to CRA, RFA is associated with a higher incidence of thrombosis (16.0% vs 9.8%, p = 0.493) and hepatic infarction (12.0% vs 3.3%, p = 0.137). Additionally, Ko et al. [14] found that the RFA group had significantly higher bile duct complications compared to the CRA group (p = 0.007), and RFA was identified as an independent risk factor for bile duct complications. Another study comparing CRA with RFA in patients with cirrhotic HCC showed [13] that the CRA group had higher rates of thrombocytopenia and myoglobulinemia compared to the RFA group (both without statistical significance). The thrombocytopenia following CRA may be related to systemic inflammatory response or platelet aggregation and sequestration in the ablated area, which may increase the risk of bleeding. Myoglobulinemia may increase the risk of renal failure in patients with potential renal insufficiency. RFA has fewer concerns in this respect, and the authors consider RFA to be more suitable for patients with baseline renal insufficiency or thrombocytopenia. Furthermore, another small-scale study suggest that although there is no statistical difference in complications between CRA and RFA, for tumors in high-risk locations, CRA causes less damage to adjacent tissue structures and is recommended as a treatment option [16].

However, current HCC treatment guidelines do not endorse CRA as the standard treatment for HCC. The author speculates that this may be due to the fact that studies on CRA for the treatment of malignant liver tumors are mostly small-scale retrospective studies, lacking large-scale randomized controlled trials. In addition, safety data from two early comparative studies of CRA and RFA based on open surgical procedures (such as laparotomy or laparoscopic approaches) [18,22] have also influenced the application of CRA. Bilchik et al. [22] reported on open surgical ablation, stating that compared to RFA (3.3%), CRA (40.7%) had a higher incidence of complications, with pleural effusion being the most common complication after CRA (80/159, 50%). It is evident that open surgical procedures result in greater trauma compared to percutaneous approaches, and the incidence of trauma is closely associated with the occurrence of pleural effusion. These two studies, due to their open surgical nature, may have overestimated the postoperative complication rates of CRA. To avoid introducing confounding biases from different surgical approaches, only studies comparing percutaneous CRA and RFA were included in this research. Other studies have shown [21,23] that the use of older versions of large-gauge cryoprobes for CRA treatment of large HCCs may lead to severe complications such as cryogenic shock, liver rupture, or excessive bleeding. However, in the past decade, with the increasing application of percutaneous procedures, updates in CRA equipment, proper patient selection, and the growing experience of operators, severe complications from percutaneous CRA have rarely been reported [14]. Littrup et al. [21] found that only 1.2% of patients required intervention for pleural effusion after percutaneous CRA. Another study reported a significant reduction in post-ablation bleeding rates when using small-sized cryoprobes [6]. A study that included both HCC and liver metastases [11] demonstrated a similar rate of complications after percutaneous CRA (29%) compared to percutaneous RFA (24%). These studies collectively indicate that with the continuous development and improvement of percutaneous CRA techniques, the incidence of severe complications has significantly decreased. In this study, all CRAs included were performed via a percutaneous approach, and no severe complications were reported after CRA. Given that the post-treatment complications of percutaneous CRA and RFA for malignant liver tumors showed no statistical difference, percutaneous CRA can be recommended as one of the local treatment modalities for malignant liver tumors. However, further validation is still needed through well-designed, clearly defined large-scale randomized controlled trials.

We revealed no statistically significant difference in the local recurrence rate between CRA and RFA (OR = 1.28, 95% CI: 0.49-3.36, p = 0.61). However, it should be noted that substantial heterogeneity existed among the various studies included in the present study (I²=72%, p = 0.01). Adam et al. [11] indicate that the local recurrence rate after CRA for malignant liver tumors is significantly higher than that of RFA (53% vs 18%, p = 0.003). However, the other studies [12,16] found no statistically significant difference in local recurrence rates between CRA and RFA for malignant liver tumors. Sensitivity analysis also suggests that when excluding Adam et al.'s study [11], the heterogeneity in the outcome measure of local recurrence rate disappears (I²=25%, p = 0.27), indicating that Adam et al.'s study [11] is a major source of heterogeneity. Upon retrospective analysis of the included literature, the author found that Adam’s study [11] included 34 cases (53%) of HCC patients and 30 cases (47%) of liver metastasis patients. Although the local recurrence rate after CRA for malignant liver tumors was significantly higher than that of RFA (53% vs 18%, p = 0.003), further stratified analysis showed that there was no statistically significant difference in local recurrence rates between HCC patients treated with CRA and RFA (38% vs 17%, p = 0.25). However, for liver metastasis patients, the local recurrence rate after CRA was significantly higher than that of RFA (71% vs 19%, p = 0.004), suggesting that different tumor types may also exhibit different treatment responses to CRA and RFA. Additionally, Wang et al. [10] found that for patients with Child-Pugh class A-B liver cirrhosis and HCC lesions ≤4 cm with no more than 2 lesions, CRA and RFA demonstrated similar efficacy and safety. However, for HCC tumors with a diameter between 3 cm and 4 cm, CRA exhibited a lower local recurrence rate compared to RFA (7.7% vs 18.2%, p = 0.041), suggesting that tumors within the 3 cm to 4 cm range may benefit more from CRA [10]. The possible reason is that CT localization and real-time ultrasound guidance during the procedure allow for precise monitoring of iceball formation, enabling dynamic adjustments of iceball shape and size to ensure sufficient tumor coverage. This may contribute to treatment optimization and reduce the chance of tumor recurrence. These controversial results indicated that further studies with larger sample sizes and classifying types of liver malignant tumors are necessary to evaluate recurrence rates related to local treatment of liver malignancies. The results of this study indicated that there was no significant difference in the complete ablation rate (OR = 1.08, 95% CI: 0.62–1.90, p = 0.78) and 1-year survival rates (OR = 1.14, 95% CI: 0.63–2.06, p = 0.66) between CRA and RFA suggesting that CRA and RFA had similar effectiveness and comparable short-term survival.

The present study has some limitations that need to be addressed. Firstly, the current limited direct comparative studies between cryoablation and radiofrequency ablation for malignant liver tumors may have limited the strength of evidence. RCTs are key to elevating the level of evidence. The direction of future research should include more high-quality RCTs to enhance the quality of evidence. Secondly, the study population was patients with liver malignancy, including both HCC and metastases. This may have caused heterogeneity in the response to different ablation therapies between the two groups. A well-designed and well-classified study is needed to further validate the conclusions drawn from this study. Thirdly, there is a lack of consensus in defining complications related to local treatment for malignant liver tumors, and different studies use inconsistent criteria for defining complications. Some studies may categorize mild adverse reactions as complications, which can lead to selective reporting bias and make it challenging to avoid. Therefore, it is necessary to establish standardized definitions for complications associated with local treatment for malignant liver tumors in order to minimize such biases and ensure consistency in research reporting. Overall, while this study provides valuable insights into the effectiveness of ablation therapies for liver malignancy, the above limitations should be considered when interpreting the results. Addressing these limitations in future studies will improve the quality of evidence and increase our understanding of the effectiveness of ablation therapies for liver malignancy.

In conclusion, both CRA and RFA demonstrated similar efficacy and safety in the ablation of malignant liver tumors. Thus, CRA may be a valuable alternative to RFA in the treatment of hepatic malignancies.

Ethical approval

This article does not contain any studies with human participants or animals performed by any of the authors.

Informed consent

For this type of study informed consent is not required.

Consent for publication

For this type of study consent for publication is not required.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

I confirm I understand the terms of the share upon reasonable request data policy, the DOI of our data is 10.6084/m9.figshare.24230857.