Conflicts of Interest
Clinical outcomes of radiofrequency ablation and surgical resection for small hepatocellular carcinoma: A meta-analysis
Version of Record online: 21 DEC 2011
© 2011 Journal of Gastroenterology and Hepatology Foundation and Blackwell Publishing Asia Pty Ltd
Journal of Gastroenterology and Hepatology
Volume 27, Issue 1, pages 51–58, January 2012
How to Cite
Li, L., Zhang, J., Liu, X., Li, X., Jiao, B. and Kang, T. (2012), Clinical outcomes of radiofrequency ablation and surgical resection for small hepatocellular carcinoma: A meta-analysis. Journal of Gastroenterology and Hepatology, 27: 51–58. doi: 10.1111/j.1440-1746.2011.06947.x
The authors disclose no conflicts
- Issue online: 21 DEC 2011
- Version of Record online: 21 DEC 2011
- Accepted manuscript online: 17 OCT 2011 10:44AM EST
- Accepted for publication 8 September 2011.
- hepatocellular carcinoma;
- radiofrequency ablation;
Background and Aim: To evaluate the evidence comparing radiofrequency ablation (RFA) and surgical resection (RES) on the treatment of hepatocellular carcinoma (HCC) using meta-analytical techniques.
Methods: Literature search was undertaken until March 2011 to identify comparative studies evaluating survival rates, recurrence rates, and complications. Pooled odds ratios (OR) and 95% confidence intervals (95% CI) were calculated with either the fixed or random effect model.
Results: These studies included a total of 877 patients: 441 treated with RFA and 436 treated with RES. The overall survival was significantly higher in patients treated with RES than RFA at 1, 3 and 5 years (respectively: OR: 0.50, 95% CI: 0.29–0.86; OR: 0.51, 95% CI: 0.28–0.94; OR: 0.62, 95% CI: 0.45–0.84). In the RES group the 1, 3, and 5 years recurrence-free survival rates were significantly higher than the RFA group (respectively: OR: 0.65, 95% CI: 0.44–0.97; OR: 0.65, 95% CI: 0.47–0.89; OR: 0.52, 95% CI: 0.35–0.77). RFA had a higher rate of local recurrence (OR: 4.08, 95% CI: 2.03–8.20). For tumors ≤ 3 cm RES was better than RFA in the 3-year overall survival rates (OR: 0.38, 95% CI: 0.16–0.89).
Conclusions: Surgical resection was superior to RFA in the treatment of HCC. However, the findings have to be carefully interpreted due to the lower level of evidence.
Hepatocellular carcinoma (HCC) is the sixth most common malignant tumor and the third most common cause of cancer-related death on earth, with an estimated 500 000 deaths per year.1–4 It is prevalent in Asia and Africa, and increasing in the United States and Europe.5,6 The incidence of HCC ranges from < 10 cases per 100 000 population in North America and Western Europe as well as in Iran, Iraq and India to 50–150 cases per 100 000 population in parts of Africa and Asia.7 HCC can be curable by liver transplantation, hepatic resection, or local ablation therapies. Conforming to the Milan criteria (single HCC ≤ 5 cm or up to 3 nodules < 3 cm), theoretically the best treatment is liver transplantation. But the scarcity of donors limits this treatment.8,9 Thus, surgical resection (RES) is still considered trustworthy.10,11 Nevertheless, only 9% to 29% of patients with HCC are candidates for surgery owing to either poor hepatic reserve resulting from underlying chronic liver disease or a multifocal distribution of tumor nodules.12–20 Therefore, many nonsurgical ablation methods have been developed, such as percutaneous ethanol injection (PEI), percutaneous acetic acid injection (PAI), radiofrequency ablation (RFA), percutaneous microwave coagulation therapy (PMCT), laser-induced interstitial thermotherapy (LITT), cryoablation and transcatheter arterial chemoembolization (TACE).
In Japan: “Evidence-based clinical guidelines for the diagnosis and treatment of HCC” published in 2005 and revised in 2009, and “Consensus-based clinical practice manual for HCC” edited by the Japan Society of Hepatology (JSH), which recommends: (i) hepatectomy for a single tumor regardless of the tumor size, but local treatment may be selected for a 2-cm or smaller tumor in Child–Pugh B patients; (ii) hepatectomy or local treatment when the number of tumors is two or three and the tumor size is within 3 cm. In Europe and North America, the algorithm established by the American Association of the Study of the Liver Disease (AASLD) recommends local treatment for three or fewer 3-cm or smaller early-stage HCCs and 2-cm or smaller very-early-stage HCCs with complications, such as portal hypertension. Although the typical treatment algorithms in Japan, North America, and Europe are slightly different, basically, RFA is recommended for three or fewer 3-cm or smaller HCCs.21
However, there are some controversies in clinic studies. Huang et al.,22 Molinari et al.23 and Takayama et al.24 reputed RES had more advantages (survival and recurrence rates) regardless of tumor size larger or smaller than 3 cm, even for smaller than 2 cm. But Chen et al.,25 Hong et al.,26 Vivarellim et al.,27 and Montorsim et al.28 concluded that RFA was as effective as RES in the treatment of solitary and small HCC. What is more, Livraghi et al.29 regarded RFA as the first-line treatment for small, operable HCCs (≤ 2 cm).
There have been some reports published in the medical literature comparing the efficacy of RES with RFA for HCC. The purpose of this study was to retrospectively evaluate the long-term results of RFA as compared with RES for the treatment of HCC.
Electronic searches were performed of the MEDLINE, Cochrane Controlled Trial Register (CENTRAL) and EMBASE databases until March 2011. The following Mesh search headings were used: (radiofrequency, radio-frequency or radio frequency) (surgical resection or hepatectomy) and (liver or hepatic or hepatocellular carcinoma) in English (Fig. 1).
Data extraction and quality assessment
Data were extracted by three independent observers using standardized forms. The recorded data included the number of patients, overall survival and recurrence-free survival rates, local recurrence rates and complications. The quality of all selected articles was ranked in accordance with the score of the non-randomized controlled clinical trial quality evaluation standard.
Study selection criteria
Inclusion criteria for this study were as follows: (i) no extrahepatic metastasis; (ii) no radiologic evidence of invasion into the major portal/hepatic vein branches; (iii) no previous treatment of HCC; (iv) a single HCC of less than or equal to 5 cm in diameter or up to three HCCs that were each less than or equal to 3 cm in diameter (v) good liver function with Child–Pugh Class A or B, with no history of encephalopathy, ascites, refractory to diuretics or variceal bleeding; (vi) no previous or simultaneous malignancies; (vii) patients should be suitable for treatment with either RES or RFA, and (viii) the baseline characteristics of patients should be similar.
Criteria for exclusion
Abstracts, letters, editorials and expert opinions, reviews without original data, case reports and studies lacking control groups were excluded. The following studies were also excluded: (i) those dealing with unresectable HCC or HCC recurrence after hepatectomy; (ii) those with no clearly reported outcomes of interest; (iii) those evaluating patients with cholangiocellular carcinomas or liver metastases.
The meta-analysis was performed using the software package RevMan 5.1.0. odds ratio (OR) or mean difference with 95% confidence intervals (95%CI) were calculated for dichotomous outcomes and continuous outcomes, respectively. A random-effects model and a fixed-effect model were used using “intention-to-treat” analysis. If results were not different between the two models, the random-effects model was reported, as this model was used for the indirect comparisons. If results differed between the two models, both results were reported. Heterogeneity was explored by χ2 and I2. I2 < 25% and I2 > 50% reflect small and large inconsistency, respectively. P < 0.05 was considered significant.
A subgroup analysis was performed considering HCC ≤ 3 cm.
A funnel plot was used to explore bias. Asymmetry in the funnel plot of trial size against treatment effect was used to assess the risk of bias.
Description of studies
Of 877 patients in six studies, 441 were allocated to RFA group, whereas 436 to RES group to evaluate their therapeutic effects of HCC. Patient characteristics and evaluation index are shown in Table 1.
|Author||Design||Country or distinct||Treatment||No. patients||M/F||Mean age (year)||Mean tumor size (cm)||Mean AFP (ng/mL)||Child–Pugh A/B||Mean follow-up (months)|
|Chen25 2006||RCT||China||RFA||71||56/15||51.9 ± 11.2||—||> 200(n = 31)||71/0||27.9 ± 10.6|
|RES||90||75/15||49.4 ± 10.9||—||> 200(n = 30)||90/0||29.2 ± 11.9|
|Lupo30 2007||NRCT||Italy||RFA||60||47/13||68†||3.65†||> 200(n = 3)||44/16||27 ± 18.7|
|RES||42||33/9||67†||4†||> 200(n = 5)||28/14||31.3 ± 24.3|
|Hiraoka31 2008||NRCT||Japan||RFA||105||76/29||69.4 ± 9.1||1.98 ± 0.52||114.5 ± 319.5||79/26||847.4 ± 700.3‡|
|RES||59||44/15||62.4 ± 10.6||2.27 ± 0.55||427.8 ± 1317.6||54/5||927.1 ± 698.4‡|
|Santambrogio33 2009||NRCT||Italy||RFA||74||59/15||68 ± 7||2.63 ± 1.07||9(2–4282)||74/0||38.2 ± 28.4|
|RES||78||55/23||68 ± 8||2.87 ± 1.21||8(1–28 558)||78/0||36.2 ± 23.5|
|Huang22 2010||RCT||China||RFA||115||79/36||56.57 ± 14||—||> 400(n = 29)||110/5||37.2†|
|RES||115||85/30||55.91 ± 12||—||> 400(n = 30)||106/9||46.4†|
Overall survival rates
One-year survival rates: The meta-analysis (all trials reported this data) showed that there was significant difference and favorable to RES group (OR, 0.50 [95% CI, 0.29–0.86]; P = 0.01), with certain heterogeneity (Fig. 2). (Table 2)
|Variables||No. studies furnishing data||Results||OR (95%CI)||P-value||I2|
|1-year||6 (22,25,30–33)||90.93%||94.95%||0.50 (0.29–0.86)||= 0.01||42%|
|3-year||5 (22,25,30,31,33)||72.24%||81.09%||0.51 (0.28–0.94)||= 0.03||64%|
|5-year||5 (22,30–33)||50.54%||60.41%||0.62 (0.45–0.84)||= 0.002||29%|
|1-year||5 (22,25,30–32)||79.02%||83.80%||0.65 (0.44–0.97)||= 0.03||26%|
|3-year||4 (22,25,30,31)||50.75%||58.95%||0.65 (0.47–0.89)||= 0.008||0%|
|5-year||4 (22,30–32)||22.30%||33.58%||0.52 (0.35–0.77)||= 0.001||60%|
|Overall survival in HCCs ≤ 3 cm|
|1-year||2 (22,31)||93.83%||99.04%||0.21 (0.04–1.15)||= 0.07||0%|
|3-year||2 (22,31)||83.95%||93.27%||0.38 (0.16–0.89)||= 0.03||55%|
|5-year||2 (22,31)||59.88%||69.23%||0.69 (0.41–1.16)||= 0.16||70%|
|Local recurrence||3 (22,32,33)||17.04%||4.85%||4.08 (2.03–8.20)||< 0.0001||27%|
|Complication||4 (22,25,30–32)||6.58%||28.21%||0.29 (0.08–1.10)||= 0.07||84%|
Three-year survival rates: The meta-analysis (five trials reported this data) showed that the group of RFA were significantly lower in the group of RES (OR, 0.51 [95% CI, 0.28–0.94]; P = 0.03), with certain heterogeneity (Fig. 3).
Five-year survival rates: The meta-analysis (five trials reported this data) showed that the experimental group of RFA was significantly lower than the group of RES (OR, 0.62 [95% CI, 0.45–0.84]; P = 0.002), with no evidence of significant heterogeneity (Fig. 4).
Recurrence-free survival rates
One-year recurrence-free survival rates: The meta-analysis (five trials reported this data) showed there was significant difference and favorable to RES group (OR, 0.65 [95% CI, 0.44–0.97]; P = 0.03), with no evidence of significant heterogeneity.
Three-year recurrence-free survival rates: The meta-analysis (four trials reported this data) showed that there was significant difference and favorable to RES group (OR, 0.65 [95% CI, 0.47–0.89]; P = 0.008) and there was no heterogeneity.
Five-year recurrence-free survival rates: The meta-analysis (four trials reported this data) showed there was significant difference and favorable to RES group (OR, 0.52 [95% CI, 0.35–0.77]; P < 0.000 01), with certain heterogeneity.
Local recurrence rates
The meta-analysis (three trials reported this data) showed that the experimental group of RFA was significantly higher than the control group of RES (OR, 4.08 [95% CI, 2.03–8.20]; P < 0.0001), with no evidence of significant heterogeneity.
The meta-analysis (four trials reported this data) showed that there was no statistics (OR, 0.29 [95%CI, 0.08–1.10]; P = 0.07). But in the group of RFA (6.58%) the incidence of complications was significantly less than the group of RES (28.21%).
Comparison overall survival rates between the two groups in HCCs ≤ 3 cm
One-year overall survival rates: The meta-analysis (two trials reported this data) showed that there were no statistics (OR, 0.21 [95% CI, 0.04–1.15]; P = 0.07).
Three-year overall survival rates: The meta-analysis (two trials reported this data) showed that there was significant difference and favorable to RES group (OR, 0.38 [95% CI, 0.16–0.89]; P = 0.03), with certain heterogeneity.
Five-year overall survival rates: The meta-analysis (two trials reported this data) showed that there were no statistics (OR, 0.69 [95% CI, 0.41–1.16]; P = 0.16).
Sensitivity analysis and publication bias
Publication bias may exist when no significant findings remain unpublished, thus artificially inflating the apparent magnitude of an effect.
Complication, survival and recurrence rates following RFA or RES for the treatment of small HCC were calculated by the fixed-effect model and random-effect model, respectively. The results were similar and the combined results were highly reliable.
Funnel plots of the study results were shown in Figure 5. The funnel plots on one-year overall survival rates following RFA or RES for the treatment of small HCC showed asymmetry, which suggested there was some publication bias.
Meta-analysis, a quantitative technique for therapeutic evaluation, may be used when controversy persists after several trials. We acknowledge that a very limited number of quality RCTs were available and included in this meta-analysis. The main reason for this is that it remains a challenge to conduct clinical research trials with randomization support and double blinding, both of which are effective means of preventing bias and improving the objectivity of clinical evidence for both the efficacy and the safety of any approved medical product or procedure or device. Although meta-analysis has traditionally been applied and is best confined to RCT, meta-analytical techniques using non-randomized controlled trials (NRCT) might be a valid method in some clinical settings in which either the number or the sample size of RCT is insufficient.34
This meta-analysis suggested that the incidence of complications after RFA for the treatment of HCC were less than the RES group, which obviously embodied the RFA's microinvasive characterization. RFA is a minimally invasive, target-selective technique that has been used in clinical studies in the 1990s. As a relatively new therapy, RFA destroys tumors by heat, using one needle, which is placed through the skin and then into the tumor. This procedure could be performed under conscious sedation and most patients can leave on the same day without hospitalization.
Surgical resection had significantly better survival rates in terms of overall survival rates at 1, 3, and 5 years, and recurrence-free survival rates at 1, 3, and 5 years. This could be partly explained by the increased understanding of liver segmental anatomy, and the improvements in surgical techniques, which have led to a dramatic decrease in operative mortality and an improvement in surgical outcome.21 Furthermore, owing to the new arisen technique, some physicians haven't skillfully grasped it, and weren't able to detect minor indication of complication and recurrence, which is responsible for this result. The delay of surveillance in effective treatment using RFA also adversely impacted the survival of patients.35–37
High rate of recurrence after RFA or RES is the main cause of late death of patients with HCC.7,38–42 The risk factors for tumor recurrence after treatment include tumor size, insufficient safety margin, multi-nodular tumor, tumor location.43 Recurrences arise also because of pre-existing microscopic tumor foci that are undetected by imaging modalities or malignant cells that have been disseminated during manipulation.44–47 Other risk factors for local recurrence are a high pre-treatment α-fetoprotein level, tumor α-fetoprotein production, hepatitis C infection48,49 and a lower platelet count (≤ 100 000/mL). In the current study, local recurrence was found to be more frequent after RFA than RES. This may be a result of the safety margin of RFA being narrower than that of RES, as RES usually removes the entire Couinaud segment containing tumors, so the clearance of tumors and any potential sites of microscopic disease will be more complete in these patients.50,51 Local recurrences after RFA may be attributable to insufficient ablation of the primary tumor and/or the presence of tumor venous invasion in the adjacent liver. However, RES could remove the primary tumor and venous tumor thrombus.52–54
The subgroup analysis evaluating tumor size showed no obvious difference in the survival estimates between RFA and RES for HCC ≤ 3 cm and overall survival. In 3-year overall survival rates, the group of RES had marked advantages. Therefore, we consider that RFA is significantly worse than RES in the long-term. However, the limitation of these analyses is that only a few studies were included and that the sample size was small. We also separately analyzed the two RCT trials, but all aspects of the survival are not statistically significant.
Limitations of the study
The conclusions of this meta-analysis are limited by various factors. First, the number of studies that are included in this review is few. This may lead to false positive or false negative conclusions (risk of random errors). Second, the majority of data in the present study comes from NRCTs; therefore the overall level of clinical evidence is low. Even the randomization procedure was unclear or inadequate in the trials.55
Funnel plots can be suggestive of publication bias with a lack of negative small RCTs. However, a firm conclusion about bias is difficult to reach as the asymmetry of the funnel plot is minimal. In addition, funnel plots can show asymmetry for reasons other than publication bias. Therefore, our pooled OR might be an overestimate of the true effect. Due to data constraints, this meta-analysis could not analyze the quality of life score and was unable to carry out stratified analyses of other possible confounding factors. If the method is to be more effective, then larger samples and randomized controlled studies with longer follow-up are required.
Further, for staging of HCC, several systems have been proposed that address the extent and the prognosis of the disease: the “Okuda staging system,” the “TNM classification” and its modification by the “Union International Centre Cancer” (UICC), the “Barcelona Clinic Liver Cancer” (BCLC) classification, the “Cancer of the Liver Italian Program” (CLIP) score, the “Japan Integrated Staging” (JIS) score, the “Groupe d'Etude de Traitement du Carcinoma Hepatocellulaire”(GRETCH) score and the “Chinese University Prognostic Index” (CUPI). Simultaneously, several different RFA systems were used in the treatment centers, such as RITA Medical System (Mountain View, CA, USA), Radionics (Burlington, MA, USA) and Valleylab (Boulder, CO, USA). Different approaches and RFA systems would also impact on the pooling of data and interpretation of results. However, we failed to find a conversion between these systems because of the limited data of these trials.
Surgical resection was superior to RFA in the treatment of patients with small HCC eligible for surgical treatments. However, the findings have to be carefully interpreted due to the lower level of evidence (systematic error and random error). Further RCTs are warranted to clarify the exact value of RES and RFA for small HCC.
- 11National Comprehensive Cancer Network. NCCN clinical practice guidelines in oncology. Hepatobiliary Cancers Version. 1. 2011. Cited 4 Sept 2011. Available from URL: http://www.nccn.org/professionals/physician_gls/pdf/hepatobiliary.pdf