Note: This copy is for your personal non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, contact us at www.rsna.org/rsnarights.
Purpose:
To compare radiofrequency (RF) ablation with nonanatomic resection (NAR) as first-line treatment in patients with a single Barcelona Clinic Liver Cancer (BCLC) stage 0 or A hepatocellular carcinoma (HCC) and to evaluate the long-term outcomes of both therapies.
Materials and Methods:
This retrospective study was approved by the institutional review board. The requirement for informed consent was waived. Data were reviewed from 580 patients with HCCs measuring 3 cm or smaller (BCLC stage 0 or A) who underwent ultrasonographically (US) guided percutaneous RF ablation (n = 438) or NAR (n = 142) as a first-line treatment. Local tumor progression, intrahepatic distant recurrence, disease-free survival, and overall survival rates were analyzed by using propensity score matching to compare therapeutic efficacy. In addition, major complications and length of postoperative hospital stay were compared.
Results:
Before propensity score matching (n = 580), the 5-year cumulative rates of local tumor progression for RF ablation and NAR (20.9% vs 12.7%, respectively; P = .093) and overall survival rates (85.5% vs 90.9%, respectively; P = .194) were comparable, while the 5-year cumulative intrahepatic distant recurrence rates (62.7% vs 36.6%, respectively; P , .001) and disease-free survival rates (31.7% vs 61.1%, respectively; P , .001) in the NAR group were significantly better than those in the RF ablation group. After matching (n = 198), there were no significant differences in therapeutic outcomes between the RF ablation and NAR groups, including 5-year cumulative intrahepatic distant recurrence (47.0% vs 40.2%, respectively; P = .240) and disease-free survival rates (48.9% vs 54.4%, respectively; P = .201). RF ablation was superior to NAR for major complication rates and length of postoperative hospital stay (P , .001).
Conclusion:
In patients with one BCLC stage 0 or A ( 3 cm) HCC who received RF ablation or NAR as first-line treatment, there were no significant differences in long-term therapeutic outcomes; however, RF ablation was associated with fewer major complications and a shorter hospital stay after treatment.
1
From the Department of Radiology and Center for Imaging Science (T.W.K., H.R., M.W.L., Y.S.K., H.K.L., D.C., K.D.S.), Department of Surgery (J.M.K., C.H.D.K., J.W.J.), Division of Hepatology, Department of Medicine (S.W.P., Y.H.P.), and Biostatics Unit, Samsung Biomedical Research Institute (J.H.A.), Samsung Medical Center, Sungkyunkwan University School of Medicine, 50 Irwon-Dong, Gangnam-gu, Seoul 135-710, Korea. Received June 24, 2014; revision requested August 11; revision received October 23; accepted October 31; final version accepted November 30. Supported by Samsung Medical Center (grant GFO1130071). Address correspondence to H.R. (e-mail: [email protected]).
RSNA, 2015
q
RSNA, 2015
q
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
1
and Interventional Radiology
Tae Wook Kang, MD Jong Man Kim, MD Hyunchul Rhim, MD Min Woo Lee, MD Young-sun Kim, MD Hyo Keun Lim, MD Dongil Choi, MD Kyoung Doo Song, MD Choon Hyuck David Kwon, MD Jae-Won Joh, MD Seung Woon Paik, MD Yong Han Paik, MD Joong Hyun Ahn, PhD
Original Research n Vascular
Small Hepatocellular Carcinoma: Radiofrequency Ablation versus Nonanatomic Resection—Propensity Score Analyses of Long-term Outcomes1
VASCULAR AND INTERVENTIONAL RADIOLOGY: Radiofrequency Ablation versus Resection for Small Hepatocellular Carcinoma
A
ccording to recent clinical guidelines, liver transplantation, surgical resection, and radiofrequency (RF) ablation are considered to be curative treatment modalities for hepatocellular carcinoma (HCC) (1). Although liver transplantation has merit compared with the other two modalities because it involves removal of undetectable tumor and preneoplastic cirrhotic tissue, the shortage of liver donors limits its availability for most patients. Therefore, RF ablation and surgical resection have been the mainstays of curative treatment for HCC. Two major types of surgical resection are performed: anatomic resection, ranging from segmentectomy to extended hemihepatectomy, and nonanatomic resection (NAR), such as tumor enucleation and wedge resection (2). Of the two techniques, NAR is more similar to RF ablation because it does not allow systemic removal of the corresponding hepatic segment fed by tumor-bearing portal tributaries
Advances in Knowledge nn For first-line treatment of a single Barcelona Clinic Liver Cancer (BCLC) stage 0 or A ( 3 cm) hepatocellular carcinoma (HCC), patients in the radiofrequency (RF) ablation group were significantly older and more likely to be classified as having poor liver function compared with those in the nonanatomic resection (NAR) group in the pooled cohort (n = 580).
but allows preservation of nontumorous parenchyma, especially in patients with cirrhosis who have small HCCs (3). NAR and RF ablation have similar advantages, such as the preservation of liver tissue to maintain adequate hepatic reserve, and disadvantages, such as limited resection (4) or ablative margins (5), which are associated with risk of marginal recurrence (6) and transportal tumor spread (7). The type of surgical resection is a reasonable consideration in decision making between RF ablation and resection, especially for treatment of small HCCs. However, to our knowledge, there have been no studies in which authors compared therapeutic outcomes in patients treated with RF ablation with those of patients treated with NAR. In addition, authors of retrospective studies (8,9) have shown a high propensity toward enrolling younger patients with better liver function in surgical resection groups. Authors of a comparative study (10) used the highly selective criterion of a Child-Pugh score of 5, which is a well-known prognostic factor for overall survival after RF ablation or surgical resection (11,12). To help clarify these issues, we aimed to compare retrospectively the long-term therapeutic outcomes of RF ablation and NAR in patients with a small ( 3 cm) Barcelona Clinic Liver Cancer (BCLC) stage of 0 or A HCC by using propensity score analysis.
Materials and Methods Study Design We conducted a comparative study by using a database of records of patients
nn NAR yielded better cumulative intrahepatic distant recurrence and disease-free survival rates than did RF ablation in the pooled cohort (n = 580); however, the results for NAR and RF ablation were similar after propensity score adjustment in the matched cohort (n = 198).
nn RF ablation as the first-line treatment for a single BCLC stage 0 or A ( 3 cm) HCC can provide effective long-term therapeutic outcomes comparable to those with NAR; in addition, RF ablation offers a lower rate of major complications and shorter posttreatment hospital stay.
2
nn Cumulative local tumor progression and overall survival rates were not significantly different between the RF ablation and NAR groups before and after propensity score matching.
Implication for Patient Care
Kang et al
with HCC at a single tertiary referral center. Our institutional review board approved this study and waived the requirement for informed consent. However, written informed consent for RF ablation and NAR was obtained from all patients before treatment, in accordance with clinical protocols. To minimize heterogeneous baseline covariates in this retrospective observational study, we evaluated the clinical variables affecting therapeutic outcomes and performed propensity score analysis. The selection of these covariates was based on clinical relevance and the results of previous studies (12–17). We used standardized terminology and reporting criteria for surgical resection and treatment outcomes for RF ablation (2,18).
Patients Between January 2006 and October 2010, 5981 patients received a diagnosis of HCC at the Samsung Medical Center (Sungkyunkwan University, Seoul, Korea). Among them, 580 consecutive patients (444 men, 136 women; mean age, 56.66 years; range, 30–79 years) underwent either RF ablation (n = 438) or NAR (n = 142) as
Published online before print 10.1148/radiol.15141483 Content code: Radiology 2015; 000:1–12 Abbreviations: BCLC = Barcelona Clinic Liver Cancer CI = confidence interval HCC = hepatocellular carcinoma HR = hazard ratio NAR = nonanatomic resection RF = radiofrequency Author contributions: Guarantors of integrity of entire study, T.W.K., J.M.K., H.R., J.W.J., S.W.P., Y.H.P.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; agrees to ensure any questions related to the work are appropriately resolved, all authors; literature research, T.W.K., J.M.K., H.R., H.K.L., S.W.P.; clinical studies, T.W.K., J.M.K., H.R., M.W.L., Y.S.K., H.K.L., D.C., C.H.K., J.W.J., S.W.P., Y.H.P.; experimental studies, H.R., S.W.P.; statistical analysis, T.W.K., J.M.K., H.R., S.W.P., J.H.A.; and manuscript editing, T.W.K., J.M.K., H.R., M.W.L., Y.S.K., H.K.L., D.C., K.D.S., S.W.P., Y.H.P. Conflicts of interest are listed at the end of this article.
radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
VASCULAR AND INTERVENTIONAL RADIOLOGY: Radiofrequency Ablation versus Resection for Small Hepatocellular Carcinoma
Kang et al
Figure 1
Figure 1: Flow diagram for our study. Anatomic resection was performed as first-line treatment in 226 of 378 (59.8%) patients who underwent surgical resection for small HCCs. Among them, major hepatectomy (resection of three or more Couinaud segments) was performed in 45 (11.9%) patients. Technical failure occurred in 13 of 510 patients treated with percutaneous RF ablation (RFA) for a single nodular HCC. ECOG = Eastern Cooperative Oncology Group, PEI = percutaneous ethanol injection, TACE = transcatheter arterial chemoembolization.
first-line treatment and were included in this study according to the following criteria: (a) presence of a small ( 3 cm) nodular HCC; (b) Child-Pugh class A or B with Eastern Cooperative Oncology Group performance status of 0 (BCLC stage 0 or A HCC); (c) absence of vascular invasion and extrahepatic metastasis at the time of diagnosis; (d) percutaneous approach with ultrasonographic (US) guidance in the RF ablation group and technical success after initial treatment; and (d) completion of at least 6 months of follow-up care. First-line treatment was defined as no prior treatment at the time of
diagnosis of HCC. Our flowchart of inclusion criteria for study patients is shown in Figure 1. Our diagnostic criteria for HCC were based on the clinical guidelines that were most current at the time of treatment (19,20). HCC was confirmed histologically in 23 (5.3%) patients in the RF ablation group by means of percutaneous biopsy before treatment, and histologic diagnoses were made in all patients in the NAR group after treatment. Reasons for undergoing RF ablation rather than NAR included preference for noninvasive treatment, refusal of general
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
anesthesia, insufficient postoperative hepatic reserve, high risk for complications of surgery associated with old age, and surgical resection deemed unfeasible or hazardous because of lesion location (21). The clinical variables used for outcome analysis included cirrhosis confirmed at histologic or clinical examination, including laboratory and US findings, and portal hypertension diagnosed according to the presence of one or more of the following parameters: formation of gastroesophageal varices, ascites, and splenomegaly with decreased platelet count (100,000/mm3) (14). 3
VASCULAR AND INTERVENTIONAL RADIOLOGY: Radiofrequency Ablation versus Resection for Small Hepatocellular Carcinoma
RF Ablation Procedure All RF ablation procedures were performed by one of five radiologists (M.W.L., D.C., H.R., H.K.L., and Y.S.K., each with at least 4 years of experience in RF ablation before the start of this study). All patients scheduled for RF ablation underwent outpatient US to assess the feasibility of USguided percutaneous RF ablation (22). The method and therapeutic strategy of RF ablation were the same as those described in previous studies (13,23). All procedures were performed percutaneously with US guidance and commercially available internally cooled electrode systems with generators (Cool-tip RF System, Covidien, Mansfield, Mass; or VIVA RFA System, STARmed, Goyang, Korea). Moderate sedation with intravenous injection of pethidine hydrochloride (Samsung Pharmaceuticals, Seoul, Korea) and fentanyl citrate (GUJU Pharma, Seoul, Korea) was applied for each treatment. Our therapeutic goal for RF ablation was to achieve an ablative margin of at least 0.5 cm in the normal liver tissue surrounding the tumor, with the exception of subcapsular and perivascular tumors. If residual unablated tumor was detected at follow-up computed tomography (CT), which was performed immediately after RF ablation, an additional RF ablation session was attempted, if possible, on the following day. Incomplete tumor ablation after an additional treatment session was considered to be a technical failure, and these cases were excluded from our study for recurrence and survival analyses. Surgical Procedure All surgical procedures were performed by one of three experienced surgeons (J.W.J., C.H.K., and J.M.K., each with more than 7 years of experience in hepatobiliary surgery). The type of hepatic resection was defined according to the proposed guidelines (2). Anatomic resection was defined as the complete removal of at least one Couinaud segment containing the tumor and the corresponding hepatic territory. NAR refers to removal of the tumor with a minimal tumor-free margin without regard to 4
segmental anatomy, such as wedge resection or tumor enucleation (24,25). The surgical approach was chosen on the basis of the hepatic functional reserve, tumor location, and preference of the operator. In general, peripherally located tumors and patients with cirrhosis were treated with NAR rather than with anatomic resection. All surgical procedures were performed by using standard surgical techniques for hepatectomy (26). Intraoperative US was used to clarify tumor margins and locations to help determine the optimal dissection plane.
Posttreatment Assessment For assessment of therapeutic outcomes and complications, all patients in the RF ablation group underwent contrast material–enhanced multiphasic CT immediately after RF ablation to determine the technical success of the procedure, while resection margins and status (R0 vs R1 resection) in the NAR group were evaluated according to the absence of microscopic tumor invasion at the resection margin. In addition, patients in the NAR group underwent CT imaging with the same protocol if complications were clinically suspected during the postoperative hospital stay. Patients in both groups underwent multiphasic CT, chest radiography, and laboratory tests including serum a-fetoprotein analysis 1 month after initial discharge, every 3 months during the first 2 years, and every 4–6 months thereafter. If there was a possibility of extrahepatic recurrence based on clinical symptoms or unexplained elevation of a-fetoprotein, chest CT, whole-body bone scintigraphy, and brain MR imaging also were performed. When local tumor progression (defined as the appearance of enhancing tumor around the ablation zone or resection margin), intrahepatic distant recurrence, or extrahepatic recurrence developed during the follow-up period, second-line treatment such as RF ablation, transcatheter arterial chemoembolization, resection, or administration of sorafenib was initiated according to the recommendations of a multidisciplinary tumor board regarding liver function and general condition of the patient and the characteristics of tumor recurrence.
Kang et al
Comparison of Long-term Therapeutic Outcomes For comparison of long-term therapeutic outcomes between the two groups, local tumor progression, intrahepatic distant recurrence, disease-free survival, and overall survival were analyzed before and after propensity score matching. Disease-free survival was defined as the time during the follow-up period during which the patient did not experience local tumor progression, intrahepatic distant recurrence, extrahepatic recurrence, or death. The overall survival rate was calculated from the date of the first treatment to either the date of death or the last visit to the outpatient clinic before May 30, 2013. Liver transplantation was considered to be censored at the time of surgery. In addition, the length of posttreatment hospital stay, major complications, treatment-related mortality, and in-hospital mortality were compared. Major complications were defined as clinical events that led to additional therapeutic interventions or prolonged hospitalization (27). In-hospital mortality was defined as death occurring during hospitalization. Statistical Analysis Continuous data were evaluated with the assumption of normality by using twosample t tests or Mann-Whitney tests. Categorical variables were analyzed by using x2 tests or Fisher exact tests. To control for overt bias introduced by observed confounding factors between the groups, we used one-to-one nearest neighbor individual matching without replacement by using propensity scores generated from logistic regression (28). We used a nonparsimonious logistic regression model to generate a propensity score. The logistic regression was estimated to include age at enrollment, sex, cause of chronic liver disease, presence of liver cirrhosis, Child-Pugh classification, presence of portal hypertension, tumor size, proportion of BCLC stage 0 HCC, serum a-fetoprotein concentration, antiviral treatment, platelet count, total bilirubin level, albumin, and prothrombin time. Before matching, variables potentially influencing disease-free and overall
radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
VASCULAR AND INTERVENTIONAL RADIOLOGY: Radiofrequency Ablation versus Resection for Small Hepatocellular Carcinoma
survival were evaluated with a Cox proportional hazards model. After matching, all paired statistical analyses were performed as follows: (a) Continuous data were evaluated by using one-sample t tests or Wilcoxon signed rank tests according to the normality assumption. (b) For categorical variables, the McNemar test was used. (c) To evaluate the covariate of balance irrelevant to the sample size, effect size was reported as a standardized mean difference before and after matching: Values less than or equal to 0.10 indicated very small differences, 0.1–0.3 indicated small differences, 0.3– 0.5 indicated moderate differences, and greater than 0.5 indicated large differences (29). (d) Local tumor progression, intrahepatic distant recurrence, and disease-free and overall survival rates were estimated by using the KaplanMeier method and were compared with Cox proportional hazards models before and after matching. Competing risk regression was performed as part of the recurrence analysis, because death was considered to be a competing risk. By using the sample size after matching, we calculated the magnitude of the difference of 5-year local tumor progression in both groups that would have been detectable with 80% power and a significance level of 5%. All statistical analyses were performed by using software (SAS version 9.3; SAS Institute, Cary, NC, and R 3.0.2.; R Foundation for Statistical Computing, Vienna, Austria). A P value less than .05 was considered to indicate a significant difference.
Results Patients This study included a total of 580 patients, 438 of whom were included in the RF ablation group and 142 of whom were included in the NAR group. The median follow-up periods were 50.9 months (range, 12.6–92.8 months) in the RF ablation group and 59.7 months (range, 10.6–92.5 months) in the NAR group. Patients in the RF ablation group were significantly older and more likely to be classified as having poor liver function compared with those in the
Kang et al
Table 1 Demographic and Clinical Characteristics of Enrolled Patients Variable Age at enrollment (y)* No. of men Hepatitis B positive Antihepatitis C positive Other cause of liver disease Liver cirrhosis Child-Pugh class B Portal hypertension Tumor size* BCLC stage 0 HCC (, 2 cm) a-fetoprotein concentration (µg/L)*† Antiviral treatment Platelet count (x 109/L)* Total bilirubin (mg/dL)*‡ Albumin (g/dL)*§ Prothrombin time (INR)*
RF Ablation Group (n = 438) 58 (30–80) 337 (76.9) 324 (74.0) 64 (14.6) 31 (7.1) 355 (81.1) 71 (16.2) 226 (51.6) 1.9 (1.1–3.0) 256 (58.5) 15.4 (1.0–3772.5) 144 (58.5) 102.5 (20–395) 0.7 (0.2–3.1) 3.8 (2.3–4.8) 1.2 (0.9–1.8)
NAR Group (n = 142)
P Value
Effect Size
53 (28–74) 107 (75.4) 120 (84.5) 10 (7.1) 7 (4.9) 91 (64.1) 7 (4.9) 34 (23.9) 2 (1.1–3.0) 76 (53.5) 22.2 (1.0–5517.3)
,.001 .698 .012 .020 .439 ,.001 .001 ,.001 .327 .303 .628
0.581 0.037 0.295 0.099 0.044 0.352 0.519 0.571 0.103 0.161 0.184
.025 ,.001 .214 ,.001 ,.001
0.231 0.823 0.312 1.004 1.433
102 (41.5) 142.5 (48–309) 0.7 (0.3–1.7) 4.1 (2.9–5.1) 1.1 (0.9–1.3)
Note.—Unless otherwise indicated, data are numbers of patients, with percentages in parentheses. Log transformation was used for analysis of a-fetoprotein concentration and platelet count. Effect size values less than 0.10 indicate very small differences, 0.1–0.3 indicate small differences, 0.3–0.5 indicate moderate differences, and greater than 0.5 indicate large differences. INR = international normalized ratio. * Data are medians, with the range in parentheses. †
To convert to Système International (SI) units (micrograms per liter), multiply by 1.
‡
To convert to SI units (micromoles per liter), multiply by 17.104.
§
To convert to SI units (grams per liter), multiply by 10.
NAR group, with the exception of total bilirubin level, for which there was no significant difference between the groups. However, tumor profile, index tumor size, proportion of tumors that were BCLC stage 0 HCC (, 2 cm), and a-fetoprotein concentration before treatment were not significantly different between the groups. All 142 patients in the NAR group had undergone R0 resection and histopathologic analysis (mean resection margin, 1.0 cm; range, 0.1–2.3 cm). For seven (4.9%) patients, NAR was performed by using a laparoscopic approach. Other baseline characteristics of patients included in both groups are shown in Table 1.
Comparison of Therapeutic Outcomes for All Study Patients before Propensity Score Matching Tumor recurrence.—At the time of censoring, local tumor progression had developed in 84 of 438 (19.2%)
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
patients in the RF ablation group and in 19 of 142 (13.4%) patients in the NAR group. The cumulative local tumor progression rates at 1, 3, and 5 years were 6.6%, 16.4%, and 20.9%, respectively, for the RF ablation group and 7.0%, 10.0%, and 12.7%, respectively, for the NAR group. The differences between the groups were not significant (P = .093). Intrahepatic distant recurrence was identified in 250 of 438 (57.1%) patients in the RF ablation group and in 50 of 142 (35.2%) patients in the NAR group. Unlike rates of local tumor progression, cumulative intrahepatic distant recurrence rates were significantly higher in the RF ablation group than those in the NAR group (16.0% vs 13.4% at 1 year, 49.7% vs 27.2% at 3 years, and 62.7% vs 36.6% at 5 years, respectively; P , .001) (Fig 2). Extrahepatic recurrence occurred in 34 of 438 (7.8%) patients in the RF ablation group and nine of 5
VASCULAR AND INTERVENTIONAL RADIOLOGY: Radiofrequency Ablation versus Resection for Small Hepatocellular Carcinoma
Kang et al
Figure 2
Figure 2: Graphs show types of recurrence and survival curves in patients with small HCCs who underwent RF ablation (RFA) and NAR. Between the two groups of study patients (n = 580), (a) cumulative local tumor progression and (d) overall survival rates were not significantly different. However, (b) cumulative intrahepatic distant recurrence and (c) disease-free survival rates were significantly better in NAR group than in RF ablation group. IDR = intrahepatic distant recurrence, LTP = local tumor progression.
142 (6.3%) patients in the NAR group (P = .480). Regarding the second-line treatment modality for initial tumor 6
recurrence, there were no significant differences between the groups (P . .05) (Table 2).
Disease-free and overall survival.— During the follow-up period, 51 of 438 (11.6%) patients in the RF ablation
radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
VASCULAR AND INTERVENTIONAL RADIOLOGY: Radiofrequency Ablation versus Resection for Small Hepatocellular Carcinoma
Kang et al
Table 2 Second-Line Treatment Modalities for Patients with the Initial Tumor Recurrence in Both Groups Type of Tumor Recurrence Intrahepatic Distant Recurrence (n = 300)†
Local Tumor Progression (n = 103)* Second-Line Treatment RF ablation TACE RF ablation and TACE Surgical resection Liver transplantation Sorafenib Not available‡
RF Ablation Group (n = 84)
NAR Group (n = 19)
RF Ablation Group (n = 250)
NAR Group (n = 50)
36 (42.8) 41 (48.8) 5 (6.0) 1 (1.2) 0 0 1 (1.2)
6 (31.5) 11 (57.9) 1 (5.3) 1 (5.3) 0 0 0
86 (34.3) 132 (52.8) 18 (7.3) 4 (1.6) 3 (1.2) 2 (0.8) 5 (2.0)
19 (38) 28 (56) 1 (2) 1 (2) 1 (2) 0 0
Note.—Data are number of patients, with percentages in parentheses. Sorafenib was used for advanced-stage HCCs according to the BCLC clinical guidelines and temporal availability. TACE = transcather arterial chemoembolization * P = .568. †
P = .740.
‡
These patients did not receive any treatment for recurrent HCC at the time of censoring.
group and 12 of 142 (8.5%) patients in the NAR group died. The diseasefree survival rates at 3 and 5 years were estimated to be 43.8% and 31.7%, respectively, in the RF ablation group and 71.5% and 61.1%, respectively, in the NAR group. The difference was significant (P , .001). However, both groups had similar overall survival rates of 92.4% and 85.5% in the RF ablation group and 93.4% and 90.9% in the NAR group at 3 and 5 years, respectively (P = .194) (Fig 2). Evaluation of variables affecting disease-free and overall survival.—At multivariate analysis of all study patients (n = 580), NAR treatment (P = .002; hazard ratio [HR], 0.60; 95% confidence interval [CI]: 0.43, 0.83), tumor size (P = .027; HR, 1.29; 95% CI: 1.03, 1.62), hepatitis C viral infection (P = .034; HR, 1.59; 95% CI: 1.04, 2.45), and serum albumin level (P = .045; HR, 0.73; 95% CI: 0.54, 0.99) were independent prognostic factors for disease-free survival. For overall survival, tumor size (P = .015; HR, 1.82; 95% CI: 1.12, 2.96), serum albumin level (P = .003; HR, 0.35; 95% CI: 0.18, 0.70), and hepatitis C viral infection (P = .005; HR, 2.30; 95% CI: 1.28, 4.13) were significant independent prognostic factors (Table 3).
Comparison of Therapeutic Outcomes after Propensity Score Matching To balance the clinical variables and control for confounding bias, a total of 198 patients (99 from each group) were matched by applying one-toone propensity score matching. This sample size after matching allowed detection of a difference of at least 18% in the 5-year local tumor progression rate, the primary outcome in our study, with 80% power and a significance level of 5%. Matched pairs showed similar baseline characteristics (Table 4). In the matched cohort, the cumulative local tumor progression rates in the RF ablation group at 1, 3, and 5 years were 5.1%, 14.5%, and 19.4%, respectively, and 8.1%, 12.3%, and 18.6%, respectively, in the NAR group (P = .860). The 3- and 5-year overall survival rates were 95.7% and 87.2%, respectively, in the RF ablation group, and 92.7% and 90.2%, respectively, in the NAR group, a difference that was not significant (P = .594). In addition, the cumulative intrahepatic distant recurrence and disease-free survival rates did not differ between the two groups, because these intergroup differences were markedly decreased after propensity score matching: cumulative intrahepatic distant recurrence rates
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
(14.1% vs 16.2% at 1 year; 38.4% vs 30.9% at 3 years; and 47.0% vs 40.2% at 5 years); disease-free survival rates (53.3% vs 67.1% at 3 years; and 48.9% vs 54.4% at 5 years), and RF ablation versus NAR (P = .240 and P = .201, respectively) (Fig 3).
Major Complications and Posttreatment Hospital Stay There were no treatment-related or inhospital deaths in either group (Table 5). Posthepatectomy liver failure was not observed in the NAR group according to the Dindo-Clavien classification. Major complications were significantly more common in the NAR group (eight of 142 [5.6%] patients) than in the RF ablation group (nine of 438 [2.1%]; P = .016). In addition, the length of stay after treatment was significantly longer in the NAR group (median, 8.5 days; range, 5–22 days) than in the RF ablation group (median, 1.0 day; range, 1–3 days; P , .001). Discussion In this retrospective study, we found that the long-term therapeutic efficacy of percutaneous RF ablation was similar to that of NAR as first-line treatment of small ( 3 cm) HCCs. Unlike anatomic resection, RF ablation and NAR are 7
VASCULAR AND INTERVENTIONAL RADIOLOGY: Radiofrequency Ablation versus Resection for Small Hepatocellular Carcinoma
Kang et al
Table 3 Evaluation of Variables for Effects on Disease-free and Overall Survival Disease-free Survival Univariate Analysis Variable Treatment type Age at enrollment Men Hepatitis B positive Antihepatitis C positive Other cause of liver disease Presence of liver cirrhosis Child-Pugh class B Portal hypertension Tumor size a-fetoprotein concentration Antiviral treatment Platelet count Total bilirubin Albumin Prothrombin time
Overall Survival
Multivariate Analysis
Univariate Analysis
Multivariate Analysis
HR
P Value
HR
P Value
HR
P Value
HR
P Value
0.45 (0.33, 0.60) 1.02 (1.01, 1.04) 1.18 (0.91, 1.54) 0.71 (0.56, 0.89) 1.94 (1.50, 2.50) 0.88 (0.56, 1.36) 1.40 (1.07, 1.83) 1.52 (1.12, 2.05) 1.48 (1.20, 1.84) 1.27 (1.03, 1.57) 0.99 (0.99, 1.00) 0.93 (0.75, 1.16) 0.99 (0.99, 1.00) 1.21 (0.97, 1.52) 0.52 (0.42, 0.64) 6.27 (2.73, 14.43)
,.001 ,.001 .210 .004 ,.001 .550 .014 .007 ,.001 .029 .380 .520 ,.001 .096 ,.001 ,.001
0.60 (0.43, 0.83) 1.01 (1.00, 1.03) ... 1.13 (0.77, 1.66) 1.59 (1.04, 2.45) ... 1.06 (0.81, 1.40) 0.80 (0.51, 1.24) 1.06 (0.80, 1.42) 1.29 (1.03, 1.62) ... ... 0.95 (0.68, 1.34) 1.01 (0.76, 1.35) 0.73 (0.54, 0.99) 2.43 (0.61, 9.71)
.002 .057 ... .520 .034 ... .670 .310 .680 .027 ... ... .770 .940 .045 .210
0.66 (0.35, 1.24) 1.01 (0.98, 1.03) 1.01 (0.56, 1.80) 2.09 (0.29, 15.21) 5.83 (0.78, 43.81) 1.29 (0.12, 14.19) 2.04 (0.97, 4.29) 1.24 (0.63, 2.44) 1.81 (1.10, 2.97) 1.91 (1.19, 3.06) 0.99 (0.99, 1.00) 0.72 (0.44, 1.20) 0.99 (0.99, 1.00) 1.11 (0.66, 1.85) 0.33 (0.21, 0.52) 7.58 (1.55, 36.95)
.194 .614 .981 .465 .087 .837 .059 .534 .020 .007 .504 .209 .061 .697 ,.001 .012
1.10 (0.55, 2.19) ... ... ... 2.30 (1.28, 4.13) ... 1.54 (0.70, 3.39) ... 1.24 (0.65, 2.38) 1.82 (1.12, 2.96) ... ... 1.10 (0.49, 2.44) ... 0.35 (0.18, 0.70) 0.36 (0.02, 5.46)
.798 ... ... ... .006 ... .285 ... .516 .016 ... ... .822 ... .003 .460
Note.—Data in parentheses are 95% CIs. Log transformation was used for analysis of a-fetoprotein concentration and platelet count. Each variable with a P value of less than or equal to .20 at univariate analysis was entered into the final model. All statistical analyses for prognostic factors were conducted with Cox proportional hazards models.
Table 4 Demographic and Clinical Characteristics of the RF Ablation and NAR Groups after Propensity Analysis Variable Age at enrollment (y)* No. of men Hepatitis B positive Anti-hepatitis C positive Other cause of liver disease Liver cirrhosis Child-Pugh class B Portal hypertension Tumor size* BCLC stage 0 HCC (, 2 cm) a-fetoprotein concentration (ng/mL)*† Antiviral treatment Platelet count (x 109/L)* Total bilirubin (mg/dL)*‡ Albumin (g/dL)*§ Prothrombin time (INR)*
RF Ablation Group (n = 99)
NAR Group (n = 99)
P Value
Effect Size
55 (32–80) 77 (77.8) 83 (83.8) 8 (8.1) 4 (4.0) 71 (71.7) 4 (4.0) 30 (30.3) 1.9 (1.1–3.0) 50 (50.5) 25.6 (1.0–1873) 71 (71.7) 131 (50–395) 0.7 (0.2–1.9) 4.1 (2.3–4.8) 1.1 (0.9–1.3)
54 (31–74) 77 (77.8) 83 (83.8) 8 (8.1) 4 (4.0) 71 (71.7) 4 (4.0) 29 (29.3) 2 (1.1–3.0) 46 (46.5) 15.2 (1.0–3412.2) 68 (68.7) 132 (50–254) 0.7 (0.3–1.7) 4.1 (2.9–5.1) 1.1 (0.9–1.3)
.506 ..999 ..999 ..999 ..999 ..999 ..999 .866 .728 .688 .747 .742 .710 .806 .603 .806
0.071 0 0 0 0 0 0 0.024 0.078 0.082 0.044 0.067 0.005 0.003 0.059 0.023
Note.—Unless otherwise indicated, data are numbers of patients, with percentages in parentheses. One-to-one nearest-neighbor individual matching without replacement with propensity scoring was used. Effect size values less than 0.10 indicate very small differences, 0.1–0.3 indicate small differences, 0.3–0.5 indicate moderate differences, and greater than 0.5 indicate large differences. Paired comparisons were used for all statistical analyses. INR = international normalized ratio. * Data are medians, with the range in parentheses.
8
†
To convert to Système International (SI) units (micrograms per liter), multiply by 1.
‡
To convert to SI units (micromoles per liter), multiply by 17.104.
§
To convert to SI units (grams per liter), multiply by 10.
radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
VASCULAR AND INTERVENTIONAL RADIOLOGY: Radiofrequency Ablation versus Resection for Small Hepatocellular Carcinoma
Kang et al
Figure 3
Figure 3: Graphs show types of recurrence and survival curves after propensity score matching for patients with small HCCs who underwent RF ablation (RFA) or NAR. After propensity score matching (n = 198), there were no significant differences in any of the measured therapeutic outcomes between groups. IDR = intrahepatic distant recurrence, LTP = local tumor progression.
important curative treatment options that allow preservation of functioning hepatic parenchyma, especially in patients with cirrhosis. These techniques
also improve the ability to perform additional multimodality treatments including re-resection, RF ablation, and transcatheter arterial chemoembolization in
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
cases of recurrent HCCs and prevent posthepatectomy liver failure (6). In our study, cumulative local tumor progression rates were not significantly 9
VASCULAR AND INTERVENTIONAL RADIOLOGY: Radiofrequency Ablation versus Resection for Small Hepatocellular Carcinoma
Table 5 Major Complications and Hospital Stay after Each Treatment Outcome In-hospital death Treatment-related death Posttreatment hospital stay (d) Major complications Tumor seeding Intra-abdominal abscess Biloma Wound problem Hepatic abscess Abdominal bleeding Bile duct stricture Intestinal obstruction Hepatic infarction
RF Ablation Group (n = 438)
NAR Group (n = 142)
0 0 1.0 (1–3)* 9 (2.1)† 3
0 0 8.5 (5–22)* 8 (5.6)† 3
2 3 2 1 1 1 1
Note. Unless otherwise indicated, data are numbers of patients with percentages in parentheses. * Data are medians, with range in parentheses. P , .001 for comparison of posttreatment hospital stay for RF and NAR groups with Wilcoxon rank sum test. †
P = .016 for comparison of major complications between RF and NAR groups. After matching, major complications in RF and NAR groups estimated at 1% (1 of 99) and 7% (7 of 99), respectively; P = .034, McNemar test.
different between the RF ablation and NAR groups before and after propensity score matching. This equivalent treatment outcome for local tumor control might have been associated with small tumor size in our study patients because microsatellite nodules are less frequently identified in patients with small ( 3 cm) HCCs than in those with large HCCs (30). In clinical practice, achieving an ideal circumferential ablative margin that includes these peritumoral microsatellite nodules is difficult in patients with large tumors due to an insufficient ablation zone, even with the use of cluster or multiple single electrodes (5,31). In addition, NAR with a relatively narrow resection margin (mean; 1.0 cm), as was used in our surgical resection group, could have led to unfavorable local tumor control compared with anatomic resection (6,25,32). Inclusion of only small HCCs and the use of NAR in our study patients may have offset the disadvantage of using RF ablation for local tumor control. NAR yielded better cumulative intrahepatic distant recurrence and diseasefree survival rates than did RF ablation in the pooled cohort (n = 580). However the results were similar after propensity 10
score adjustment in the matched cohort (n = 198). These changes may be explained by differences in demographic and clinical characteristics between the groups. Poor liver condition in the RF ablation group could have resulted in higher intrahepatic distant recurrence rates because of multistep or de novo carcinogenesis that may have arisen from preneoplastic liver tissue associated with lower serum albumin levels or cirrhosis (33,34). In addition, patients in the RF ablation group were more likely to have hepatitis C viral infection. Unlike hepatitis B, hepatitis C is associated with a higher tumor recurrence rate after RF ablation (12) or hepatic resection (35). Intrahepatic distant recurrence was important to the difference in diseasefree survival between the groups before propensity score adjustment compared with local tumor progression and extrahepatic recurrence. In contrast to our results, authors of a recent study who used RF ablation and surgical resection for tumors of similar sizes reported that poor local tumor control (local tumor progression) rather than intrahepatic distant recurrence in the RF ablation group was the main contributor
Kang et al
to disease-free survival (14). Differences between their study and ours in the number of patients who underwent NAR in the resection group could explain these findings, because NAR can lead to poor local tumor control and early tumor recurrence due to less effective clearance of micrometastases and microvascular invasion around the tumor (6,25,36). We suspect that the differences in the proportion of patients who underwent NAR in the surgical resection groups might have contributed to our results conflicting with those of previous randomized clinical trials (37– 39) and two recent nationwide surveys (8,14), in which the authors compared RF ablation and surgical resection for the treatment of HCCs. With respect to overall survival, there were no significant differences between the two groups before or after propensity score matching (P = .194 and 0.594, respectively). These results are consistent with those of previous studies on the treatment of small HCCs with RF ablation and surgical resection (11,14). The potential effects of first line-treatment may be smaller in HCCs than those observed in other primary malignant tumors, because the function of the remaining liver tissue considerably affects tumor recurrence and overall survival during the follow-up period as well as the efficacy of initial tumor treatment (8,13,38). In addition, the HCC recurrence rate was estimated to be 70% at 5 years after initial treatment, reflecting either intrahepatic metastases (true recurrences) or the development of de novo tumors (15). As a result, second-line treatment is important to improving overall survival. Nearly half of the patients in our matched cohort experienced tumor recurrence during the scheduled follow-up period, despite having small HCCs with BCLC stages 0 and A. In our study, we found that RF ablation was associated with fewer major complications and shorter hospital stays after treatment, which indicates that RF ablation is a relatively safe and noninvasive modality for treating HCCs. These results corresponded with those of a previous meta-analysis in which
radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
VASCULAR AND INTERVENTIONAL RADIOLOGY: Radiofrequency Ablation versus Resection for Small Hepatocellular Carcinoma
RF ablation and hepatic resection were compared (40). Our study had several limitations. First, our retrospective study design may have resulted in selection bias. Although we attempted to simulate randomization by using propensity score analysis (28), there remained a possibility of uncontrolled confounding factors between the groups. Second, the small number of biopsy-proven HCCs in the RF ablation group may have influenced therapeutic outcomes. However, all patients without a histopathologic diagnosis met the diagnostic criteria for HCC according to the guidelines in place at the time of initial diagnosis (19,20). Third, because this was a single-center study and there was no universal consensus on the indications for NAR for the treatment of small HCCs, our results might not be consistently reproducible in other settings. Despite these limitations, our findings may help to validate the current Barcelona treatment strategy by contributing additional clinical evidence that compares RF ablation with NAR through propensity score analysis of a large retrospective cohort with long-term follow-up data. To our knowledge, previous investigators have not taken the type of surgical resection into consideration when designing studies comparing RF ablation and surgical resection. Additional randomized controlled trials are needed to define the best treatment for small HCCs in patients with chronic liver disease. There may be differences between RF ablation and major hepatectomy (resection of three or more Couinaud segments) regarding the remnant liver reservoir, which is important for liver function in patients with chronic liver disease, for potential intrahepatic distant recurrence associated with preneoplastic cirrhotic liver tissue, and for determination of second-line treatment in cases of recurrence. In our population, major hepatectomy was used in a considerable proportion of the patients who underwent surgical resection for small HCCs (45 of 378, 11.9%). In summary, there were no significant differences in long-term therapeutic outcomes including local tumor
progression, intrahepatic distant recurrence, disease-free survival, and overall survival between RF ablation and NAR as first-line treatment in patients with small BCLC stage 0 or A HCCs. In addition, RF ablation resulted in a lower rate of major complications and shorter posttreatment hospital stay. Disclosures of Conflicts of Interest: T.W.K. disclosed no relevant relationships. J.M.K. disclosed no relevant relationships. H.R. disclosed no relevant relationships. M.W.L. disclosed no relevant relationships. Y.S.K. disclosed no relevant relationships. H.K.L. disclosed no relevant relationships. D.C. disclosed no relevant relationships. K.D.S. disclosed no relevant relationships. C.H.D.K. disclosed no relevant relationships. J.W.J. disclosed no relevant relationships. S.W.P. disclosed no relevant relationships. S.H.P. disclosed no relevant relationships. J.H.A. disclosed no relevant relationships.
References 1. Forner A, Llovet JM, Bruix J. Hepatocellular carcinoma. Lancet 2012;379(9822):1245– 1255. 2. Strasberg SM. Nomenclature of hepatic anatomy and resections: a review of the Brisbane 2000 system. J Hepatobiliary Pancreat Surg 2005;12(5):351–355. 3. Cucchetti A, Cescon M, Trevisani F, Pinna AD. Current concepts in hepatic resection for hepatocellular carcinoma in cirrhotic patients. World J Gastroenterol 2012;18(44): 6398–6408. 4. Regimbeau JM, Kianmanesh R, Farges O, Dondero F, Sauvanet A, Belghiti J. Extent of liver resection influences the outcome in patients with cirrhosis and small hepatocellular carcinoma. Surgery 2002;131(3):311– 317. 5. Kim YS, Lee WJ, Rhim H, Lim HK, Choi D, Lee JY. The minimal ablative margin of radiofrequency ablation of hepatocellular carcinoma (. 2 and , 5 cm) needed to prevent local tumor progression: 3D quantitative assessment using CT image fusion. AJR Am J Roentgenol 2010;195(3):758–765. 6. Shi M, Guo RP, Lin XJ, et al. Partial hepatectomy with wide versus narrow resection margin for solitary hepatocellular carcinoma: a prospective randomized trial. Ann Surg 2007;245(1):36–43. 7. Mori Y, Tamai H, Shingaki N, et al. Diffuse intrahepatic recurrence after percutaneous radiofrequency ablation for solitary and small hepatocellular carcinoma. Hepatol Int 2009;3(3):509–515.
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
Kang et al
8. Hasegawa K, Kokudo N, Makuuchi M, et al. Comparison of resection and ablation for hepatocellular carcinoma: a cohort study based on a Japanese nationwide survey. J Hepatol 2013;58(4):724–729. 9. Yun WK, Choi MS, Choi D, et al. Superior long-term outcomes after surgery in childpugh class a patients with single small hepatocellular carcinoma compared to radiofrequency ablation. Hepatol Int 2011;5(2): 722–729. 10. Hong SN, Lee SY, Choi MS, et al. Comparing the outcomes of radiofrequency ablation and surgery in patients with a single small hepatocellular carcinoma and wellpreserved hepatic function. J Clin Gastroenterol 2005;39(3):247–252. 11. Nishikawa H, Inuzuka T, Takeda H, et al. Comparison of percutaneous radiofrequency thermal ablation and surgical resection for small hepatocellular carcinoma. BMC Gastroenterol 2011;11:143. 12. Shiina S, Tateishi R, Arano T, et al. Radiofrequency ablation for hepatocellular carcinoma: 10-year outcome and prognostic factors. Am J Gastroenterol 2012;107(4): 569–577; quiz 578. 13. Kim YS, Lim HK, Rhim H, et al. Ten-year outcomes of percutaneous radiofrequency ablation as first-line therapy of early hepatocellular carcinoma: analysis of prognostic factors. J Hepatol 2013;58(1):89–97. 14. Pompili M, Saviano A, de Matthaeis N, et al. Long-term effectiveness of resection and radiofrequency ablation for single hepatocellular carcinoma 3 cm. Results of a multicenter Italian survey. J Hepatol 2013; 59(1):89–97. 15. European Association For The Study Of The Liver; European Organisation For Research And Treatment Of Cancer. EASL-EORTC clinical practice guidelines: management of hepatocellular carcinoma. J Hepatol 2012;56(4):908–943. 16. Giannini EG, Savarino V, Farinati F, et al. Influence of clinically significant portal hypertension on survival after hepatic resection for hepatocellular carcinoma in cirrhotic patients. Liver Int 2013;33(10):1594– 1600. 17. Shimozawa N, Hanazaki K. Longterm prognosis after hepatic resection for small hepatocellular carcinoma. J Am Coll Surg 2004;198(3):356–365. 18. Goldberg SN, Charboneau JW, Dodd GD 3rd, et al. Image-guided tumor ablation: proposal for standardization of terms and reporting criteria. Radiology 2003;228(2): 335–345.
11
VASCULAR AND INTERVENTIONAL RADIOLOGY: Radiofrequency Ablation versus Resection for Small Hepatocellular Carcinoma
19. Bruix J, Sherman M; Practice Guidelines Committee, American Association for the Study of Liver Diseases. Management of hepatocellular carcinoma. Hepatology 2005;42(5): 1208–1236.
26. Kim JM, Kwon CH, Joh JW, et al. Differences between hepatocellular carcinoma and hepatitis B virus infection in patients with and without cirrhosis. Ann Surg Oncol 2014;21(2):458–465.
20. Bruix J, Sherman M, Llovet JM, et al. Clinical management of hepatocellular carcinoma. Conclusions of the Barcelona-2000 EASL conference. European Association for the Study of the Liver. J Hepatol 2001;35(3): 421–430.
27. Omary RA, Bettmann MA, Cardella JF, et al. Quality improvement guidelines for the reporting and archiving of interventional radiology procedures. J Vasc Interv Radiol 2003;14 (9 Pt 2):S293–S295.
21. Peng ZW, Lin XJ, Zhang YJ, et al. Radiofrequency ablation versus hepatic resection for the treatment of hepatocellular carcinomas 2 cm or smaller: a retrospective comparative study. Radiology 2012;262(3):1022–1033. 22. Rhim H, Choi D, Kim YS, Lim HK, Choe BK. Ultrasonography-guided percutaneous radiofrequency ablation of hepatocellular carcinomas: a feasibility scoring system for planning sonography. Eur J Radiol 2010;75(2): 253–258. 23. Kang TW, Lim HK, Lee MW, Kim YS, Choi D, Rhim H. Perivascular versus nonperivascular small HCC treated with percutaneous RF ablation: retrospective comparison of long-term therapeutic outcomes. Radiology 2014;270(3):888–899. 24. Kang CM, Choi GH, Kim DH, et al. Revisiting the role of nonanatomic resection of small (, or = 4 cm) and single hepatocellular carcinoma in patients with well-preserved liver function. J Surg Res 2010;160(1):81–89. 25. Cucchetti A, Qiao GL, Cescon M, et al. Anatomic versus nonanatomic resection in cirrhotic patients with early hepatocellular carcinoma. Surgery 2014;155(3):512–521.
12
28. McDonald RJ, McDonald JS, Kallmes DF, Carter RE. Behind the numbers: propensity score analysis-a primer for the diagnostic radiologist. Radiology 2013;269(3):640–645. 29. Burnand B, Kernan WN, Feinstein AR. Indexes and boundaries for “quantitative significance” in statistical decisions. J Clin Epidemiol 1990;43(12):1273–1284. 30. Shi M, Zhang CQ, Zhang YQ, Liang XM, Li JQ. Micrometastases of solitary hepatocellular carcinoma and appropriate resection margin. World J Surg 2004;28(4):376–381. 31. Livraghi T, Goldberg SN, Lazzaroni S, et al. Hepatocellular carcinoma: radio-frequency ablation of medium and large lesions. Radiology 2000;214(3):761–768. 32. Hasegawa K, Kokudo N, Imamura H, et al. Prognostic impact of anatomic resection for hepatocellular carcinoma. Ann Surg 2005; 242(2):252–259. 33. Park YN. Update on precursor and early lesions of hepatocellular carcinomas. Arch Pathol Lab Med 2011;135(6):704–715.
Kang et al
carcinoma: extending the indication for resection? Surgery 2012;152(5):809–820. 35. Huang YH, Wu JC, Chen CH, et al. Comparison of recurrence after hepatic resection in patients with hepatitis B vs. hepatitis C-related small hepatocellular carcinoma in hepatitis B virus endemic area. Liver Int 2005;25(2):236–241. 36. Ueno S, Kubo F, Sakoda M, et al. Efficacy of anatomic resection vs nonanatomic resection for small nodular hepatocellular carcinoma based on gross classification. J Hepatobiliary Pancreat Surg 2008;15(5):493–500. 37. Huang J, Yan L, Cheng Z, et al. A randomized trial comparing radiofrequency ablation and surgical resection for HCC conforming to the Milan criteria. Ann Surg 2010;252(6): 903–912. 38. Feng K, Yan J, Li X, et al. A randomized controlled trial of radiofrequency ablation and surgical resection in the treatment of small hepatocellular carcinoma. J Hepatol 2012; 57(4):794–802. 39. Chen MS, Li JQ, Zheng Y, et al. A prospective randomized trial comparing percutaneous local ablative therapy and partial hepatectomy for small hepatocellular carcinoma. Ann Surg 2006;243(3):321–328. 40. Duan C, Liu M, Zhang Z, Ma K, Bie P. Radiofrequency ablation versus hepatic resection for the treatment of early-stage hepatocellular carcinoma meeting Milan criteria: a systematic review and meta-analysis. World J Surg Oncol 2013;11(1):190.
34. Chang WT, Kao WY, Chau GY, et al. Hepatic resection can provide long-term survival of patients with non-early-stage hepatocellular
radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
Purpose:
To compare radiofrequency (RF) ablation with nonanatomic resection (NAR) as first-line treatment in patients with a single Barcelona Clinic Liver Cancer (BCLC) stage 0 or A hepatocellular carcinoma (HCC) and to evaluate the long-term outcomes of both therapies.
Materials and Methods:
This retrospective study was approved by the institutional review board. The requirement for informed consent was waived. Data were reviewed from 580 patients with HCCs measuring 3 cm or smaller (BCLC stage 0 or A) who underwent ultrasonographically (US) guided percutaneous RF ablation (n = 438) or NAR (n = 142) as a first-line treatment. Local tumor progression, intrahepatic distant recurrence, disease-free survival, and overall survival rates were analyzed by using propensity score matching to compare therapeutic efficacy. In addition, major complications and length of postoperative hospital stay were compared.
Results:
Before propensity score matching (n = 580), the 5-year cumulative rates of local tumor progression for RF ablation and NAR (20.9% vs 12.7%, respectively; P = .093) and overall survival rates (85.5% vs 90.9%, respectively; P = .194) were comparable, while the 5-year cumulative intrahepatic distant recurrence rates (62.7% vs 36.6%, respectively; P , .001) and disease-free survival rates (31.7% vs 61.1%, respectively; P , .001) in the NAR group were significantly better than those in the RF ablation group. After matching (n = 198), there were no significant differences in therapeutic outcomes between the RF ablation and NAR groups, including 5-year cumulative intrahepatic distant recurrence (47.0% vs 40.2%, respectively; P = .240) and disease-free survival rates (48.9% vs 54.4%, respectively; P = .201). RF ablation was superior to NAR for major complication rates and length of postoperative hospital stay (P , .001).
Conclusion:
In patients with one BCLC stage 0 or A ( 3 cm) HCC who received RF ablation or NAR as first-line treatment, there were no significant differences in long-term therapeutic outcomes; however, RF ablation was associated with fewer major complications and a shorter hospital stay after treatment.
1
From the Department of Radiology and Center for Imaging Science (T.W.K., H.R., M.W.L., Y.S.K., H.K.L., D.C., K.D.S.), Department of Surgery (J.M.K., C.H.D.K., J.W.J.), Division of Hepatology, Department of Medicine (S.W.P., Y.H.P.), and Biostatics Unit, Samsung Biomedical Research Institute (J.H.A.), Samsung Medical Center, Sungkyunkwan University School of Medicine, 50 Irwon-Dong, Gangnam-gu, Seoul 135-710, Korea. Received June 24, 2014; revision requested August 11; revision received October 23; accepted October 31; final version accepted November 30. Supported by Samsung Medical Center (grant GFO1130071). Address correspondence to H.R. (e-mail: [email protected]).
RSNA, 2015
q
RSNA, 2015
q
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
1
and Interventional Radiology
Tae Wook Kang, MD Jong Man Kim, MD Hyunchul Rhim, MD Min Woo Lee, MD Young-sun Kim, MD Hyo Keun Lim, MD Dongil Choi, MD Kyoung Doo Song, MD Choon Hyuck David Kwon, MD Jae-Won Joh, MD Seung Woon Paik, MD Yong Han Paik, MD Joong Hyun Ahn, PhD
Original Research n Vascular
Small Hepatocellular Carcinoma: Radiofrequency Ablation versus Nonanatomic Resection—Propensity Score Analyses of Long-term Outcomes1
VASCULAR AND INTERVENTIONAL RADIOLOGY: Radiofrequency Ablation versus Resection for Small Hepatocellular Carcinoma
A
ccording to recent clinical guidelines, liver transplantation, surgical resection, and radiofrequency (RF) ablation are considered to be curative treatment modalities for hepatocellular carcinoma (HCC) (1). Although liver transplantation has merit compared with the other two modalities because it involves removal of undetectable tumor and preneoplastic cirrhotic tissue, the shortage of liver donors limits its availability for most patients. Therefore, RF ablation and surgical resection have been the mainstays of curative treatment for HCC. Two major types of surgical resection are performed: anatomic resection, ranging from segmentectomy to extended hemihepatectomy, and nonanatomic resection (NAR), such as tumor enucleation and wedge resection (2). Of the two techniques, NAR is more similar to RF ablation because it does not allow systemic removal of the corresponding hepatic segment fed by tumor-bearing portal tributaries
Advances in Knowledge nn For first-line treatment of a single Barcelona Clinic Liver Cancer (BCLC) stage 0 or A ( 3 cm) hepatocellular carcinoma (HCC), patients in the radiofrequency (RF) ablation group were significantly older and more likely to be classified as having poor liver function compared with those in the nonanatomic resection (NAR) group in the pooled cohort (n = 580).
but allows preservation of nontumorous parenchyma, especially in patients with cirrhosis who have small HCCs (3). NAR and RF ablation have similar advantages, such as the preservation of liver tissue to maintain adequate hepatic reserve, and disadvantages, such as limited resection (4) or ablative margins (5), which are associated with risk of marginal recurrence (6) and transportal tumor spread (7). The type of surgical resection is a reasonable consideration in decision making between RF ablation and resection, especially for treatment of small HCCs. However, to our knowledge, there have been no studies in which authors compared therapeutic outcomes in patients treated with RF ablation with those of patients treated with NAR. In addition, authors of retrospective studies (8,9) have shown a high propensity toward enrolling younger patients with better liver function in surgical resection groups. Authors of a comparative study (10) used the highly selective criterion of a Child-Pugh score of 5, which is a well-known prognostic factor for overall survival after RF ablation or surgical resection (11,12). To help clarify these issues, we aimed to compare retrospectively the long-term therapeutic outcomes of RF ablation and NAR in patients with a small ( 3 cm) Barcelona Clinic Liver Cancer (BCLC) stage of 0 or A HCC by using propensity score analysis.
Materials and Methods Study Design We conducted a comparative study by using a database of records of patients
nn NAR yielded better cumulative intrahepatic distant recurrence and disease-free survival rates than did RF ablation in the pooled cohort (n = 580); however, the results for NAR and RF ablation were similar after propensity score adjustment in the matched cohort (n = 198).
nn RF ablation as the first-line treatment for a single BCLC stage 0 or A ( 3 cm) HCC can provide effective long-term therapeutic outcomes comparable to those with NAR; in addition, RF ablation offers a lower rate of major complications and shorter posttreatment hospital stay.
2
nn Cumulative local tumor progression and overall survival rates were not significantly different between the RF ablation and NAR groups before and after propensity score matching.
Implication for Patient Care
Kang et al
with HCC at a single tertiary referral center. Our institutional review board approved this study and waived the requirement for informed consent. However, written informed consent for RF ablation and NAR was obtained from all patients before treatment, in accordance with clinical protocols. To minimize heterogeneous baseline covariates in this retrospective observational study, we evaluated the clinical variables affecting therapeutic outcomes and performed propensity score analysis. The selection of these covariates was based on clinical relevance and the results of previous studies (12–17). We used standardized terminology and reporting criteria for surgical resection and treatment outcomes for RF ablation (2,18).
Patients Between January 2006 and October 2010, 5981 patients received a diagnosis of HCC at the Samsung Medical Center (Sungkyunkwan University, Seoul, Korea). Among them, 580 consecutive patients (444 men, 136 women; mean age, 56.66 years; range, 30–79 years) underwent either RF ablation (n = 438) or NAR (n = 142) as
Published online before print 10.1148/radiol.15141483 Content code: Radiology 2015; 000:1–12 Abbreviations: BCLC = Barcelona Clinic Liver Cancer CI = confidence interval HCC = hepatocellular carcinoma HR = hazard ratio NAR = nonanatomic resection RF = radiofrequency Author contributions: Guarantors of integrity of entire study, T.W.K., J.M.K., H.R., J.W.J., S.W.P., Y.H.P.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; agrees to ensure any questions related to the work are appropriately resolved, all authors; literature research, T.W.K., J.M.K., H.R., H.K.L., S.W.P.; clinical studies, T.W.K., J.M.K., H.R., M.W.L., Y.S.K., H.K.L., D.C., C.H.K., J.W.J., S.W.P., Y.H.P.; experimental studies, H.R., S.W.P.; statistical analysis, T.W.K., J.M.K., H.R., S.W.P., J.H.A.; and manuscript editing, T.W.K., J.M.K., H.R., M.W.L., Y.S.K., H.K.L., D.C., K.D.S., S.W.P., Y.H.P. Conflicts of interest are listed at the end of this article.
radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
VASCULAR AND INTERVENTIONAL RADIOLOGY: Radiofrequency Ablation versus Resection for Small Hepatocellular Carcinoma
Kang et al
Figure 1
Figure 1: Flow diagram for our study. Anatomic resection was performed as first-line treatment in 226 of 378 (59.8%) patients who underwent surgical resection for small HCCs. Among them, major hepatectomy (resection of three or more Couinaud segments) was performed in 45 (11.9%) patients. Technical failure occurred in 13 of 510 patients treated with percutaneous RF ablation (RFA) for a single nodular HCC. ECOG = Eastern Cooperative Oncology Group, PEI = percutaneous ethanol injection, TACE = transcatheter arterial chemoembolization.
first-line treatment and were included in this study according to the following criteria: (a) presence of a small ( 3 cm) nodular HCC; (b) Child-Pugh class A or B with Eastern Cooperative Oncology Group performance status of 0 (BCLC stage 0 or A HCC); (c) absence of vascular invasion and extrahepatic metastasis at the time of diagnosis; (d) percutaneous approach with ultrasonographic (US) guidance in the RF ablation group and technical success after initial treatment; and (d) completion of at least 6 months of follow-up care. First-line treatment was defined as no prior treatment at the time of
diagnosis of HCC. Our flowchart of inclusion criteria for study patients is shown in Figure 1. Our diagnostic criteria for HCC were based on the clinical guidelines that were most current at the time of treatment (19,20). HCC was confirmed histologically in 23 (5.3%) patients in the RF ablation group by means of percutaneous biopsy before treatment, and histologic diagnoses were made in all patients in the NAR group after treatment. Reasons for undergoing RF ablation rather than NAR included preference for noninvasive treatment, refusal of general
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
anesthesia, insufficient postoperative hepatic reserve, high risk for complications of surgery associated with old age, and surgical resection deemed unfeasible or hazardous because of lesion location (21). The clinical variables used for outcome analysis included cirrhosis confirmed at histologic or clinical examination, including laboratory and US findings, and portal hypertension diagnosed according to the presence of one or more of the following parameters: formation of gastroesophageal varices, ascites, and splenomegaly with decreased platelet count (100,000/mm3) (14). 3
VASCULAR AND INTERVENTIONAL RADIOLOGY: Radiofrequency Ablation versus Resection for Small Hepatocellular Carcinoma
RF Ablation Procedure All RF ablation procedures were performed by one of five radiologists (M.W.L., D.C., H.R., H.K.L., and Y.S.K., each with at least 4 years of experience in RF ablation before the start of this study). All patients scheduled for RF ablation underwent outpatient US to assess the feasibility of USguided percutaneous RF ablation (22). The method and therapeutic strategy of RF ablation were the same as those described in previous studies (13,23). All procedures were performed percutaneously with US guidance and commercially available internally cooled electrode systems with generators (Cool-tip RF System, Covidien, Mansfield, Mass; or VIVA RFA System, STARmed, Goyang, Korea). Moderate sedation with intravenous injection of pethidine hydrochloride (Samsung Pharmaceuticals, Seoul, Korea) and fentanyl citrate (GUJU Pharma, Seoul, Korea) was applied for each treatment. Our therapeutic goal for RF ablation was to achieve an ablative margin of at least 0.5 cm in the normal liver tissue surrounding the tumor, with the exception of subcapsular and perivascular tumors. If residual unablated tumor was detected at follow-up computed tomography (CT), which was performed immediately after RF ablation, an additional RF ablation session was attempted, if possible, on the following day. Incomplete tumor ablation after an additional treatment session was considered to be a technical failure, and these cases were excluded from our study for recurrence and survival analyses. Surgical Procedure All surgical procedures were performed by one of three experienced surgeons (J.W.J., C.H.K., and J.M.K., each with more than 7 years of experience in hepatobiliary surgery). The type of hepatic resection was defined according to the proposed guidelines (2). Anatomic resection was defined as the complete removal of at least one Couinaud segment containing the tumor and the corresponding hepatic territory. NAR refers to removal of the tumor with a minimal tumor-free margin without regard to 4
segmental anatomy, such as wedge resection or tumor enucleation (24,25). The surgical approach was chosen on the basis of the hepatic functional reserve, tumor location, and preference of the operator. In general, peripherally located tumors and patients with cirrhosis were treated with NAR rather than with anatomic resection. All surgical procedures were performed by using standard surgical techniques for hepatectomy (26). Intraoperative US was used to clarify tumor margins and locations to help determine the optimal dissection plane.
Posttreatment Assessment For assessment of therapeutic outcomes and complications, all patients in the RF ablation group underwent contrast material–enhanced multiphasic CT immediately after RF ablation to determine the technical success of the procedure, while resection margins and status (R0 vs R1 resection) in the NAR group were evaluated according to the absence of microscopic tumor invasion at the resection margin. In addition, patients in the NAR group underwent CT imaging with the same protocol if complications were clinically suspected during the postoperative hospital stay. Patients in both groups underwent multiphasic CT, chest radiography, and laboratory tests including serum a-fetoprotein analysis 1 month after initial discharge, every 3 months during the first 2 years, and every 4–6 months thereafter. If there was a possibility of extrahepatic recurrence based on clinical symptoms or unexplained elevation of a-fetoprotein, chest CT, whole-body bone scintigraphy, and brain MR imaging also were performed. When local tumor progression (defined as the appearance of enhancing tumor around the ablation zone or resection margin), intrahepatic distant recurrence, or extrahepatic recurrence developed during the follow-up period, second-line treatment such as RF ablation, transcatheter arterial chemoembolization, resection, or administration of sorafenib was initiated according to the recommendations of a multidisciplinary tumor board regarding liver function and general condition of the patient and the characteristics of tumor recurrence.
Kang et al
Comparison of Long-term Therapeutic Outcomes For comparison of long-term therapeutic outcomes between the two groups, local tumor progression, intrahepatic distant recurrence, disease-free survival, and overall survival were analyzed before and after propensity score matching. Disease-free survival was defined as the time during the follow-up period during which the patient did not experience local tumor progression, intrahepatic distant recurrence, extrahepatic recurrence, or death. The overall survival rate was calculated from the date of the first treatment to either the date of death or the last visit to the outpatient clinic before May 30, 2013. Liver transplantation was considered to be censored at the time of surgery. In addition, the length of posttreatment hospital stay, major complications, treatment-related mortality, and in-hospital mortality were compared. Major complications were defined as clinical events that led to additional therapeutic interventions or prolonged hospitalization (27). In-hospital mortality was defined as death occurring during hospitalization. Statistical Analysis Continuous data were evaluated with the assumption of normality by using twosample t tests or Mann-Whitney tests. Categorical variables were analyzed by using x2 tests or Fisher exact tests. To control for overt bias introduced by observed confounding factors between the groups, we used one-to-one nearest neighbor individual matching without replacement by using propensity scores generated from logistic regression (28). We used a nonparsimonious logistic regression model to generate a propensity score. The logistic regression was estimated to include age at enrollment, sex, cause of chronic liver disease, presence of liver cirrhosis, Child-Pugh classification, presence of portal hypertension, tumor size, proportion of BCLC stage 0 HCC, serum a-fetoprotein concentration, antiviral treatment, platelet count, total bilirubin level, albumin, and prothrombin time. Before matching, variables potentially influencing disease-free and overall
radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
VASCULAR AND INTERVENTIONAL RADIOLOGY: Radiofrequency Ablation versus Resection for Small Hepatocellular Carcinoma
survival were evaluated with a Cox proportional hazards model. After matching, all paired statistical analyses were performed as follows: (a) Continuous data were evaluated by using one-sample t tests or Wilcoxon signed rank tests according to the normality assumption. (b) For categorical variables, the McNemar test was used. (c) To evaluate the covariate of balance irrelevant to the sample size, effect size was reported as a standardized mean difference before and after matching: Values less than or equal to 0.10 indicated very small differences, 0.1–0.3 indicated small differences, 0.3– 0.5 indicated moderate differences, and greater than 0.5 indicated large differences (29). (d) Local tumor progression, intrahepatic distant recurrence, and disease-free and overall survival rates were estimated by using the KaplanMeier method and were compared with Cox proportional hazards models before and after matching. Competing risk regression was performed as part of the recurrence analysis, because death was considered to be a competing risk. By using the sample size after matching, we calculated the magnitude of the difference of 5-year local tumor progression in both groups that would have been detectable with 80% power and a significance level of 5%. All statistical analyses were performed by using software (SAS version 9.3; SAS Institute, Cary, NC, and R 3.0.2.; R Foundation for Statistical Computing, Vienna, Austria). A P value less than .05 was considered to indicate a significant difference.
Results Patients This study included a total of 580 patients, 438 of whom were included in the RF ablation group and 142 of whom were included in the NAR group. The median follow-up periods were 50.9 months (range, 12.6–92.8 months) in the RF ablation group and 59.7 months (range, 10.6–92.5 months) in the NAR group. Patients in the RF ablation group were significantly older and more likely to be classified as having poor liver function compared with those in the
Kang et al
Table 1 Demographic and Clinical Characteristics of Enrolled Patients Variable Age at enrollment (y)* No. of men Hepatitis B positive Antihepatitis C positive Other cause of liver disease Liver cirrhosis Child-Pugh class B Portal hypertension Tumor size* BCLC stage 0 HCC (, 2 cm) a-fetoprotein concentration (µg/L)*† Antiviral treatment Platelet count (x 109/L)* Total bilirubin (mg/dL)*‡ Albumin (g/dL)*§ Prothrombin time (INR)*
RF Ablation Group (n = 438) 58 (30–80) 337 (76.9) 324 (74.0) 64 (14.6) 31 (7.1) 355 (81.1) 71 (16.2) 226 (51.6) 1.9 (1.1–3.0) 256 (58.5) 15.4 (1.0–3772.5) 144 (58.5) 102.5 (20–395) 0.7 (0.2–3.1) 3.8 (2.3–4.8) 1.2 (0.9–1.8)
NAR Group (n = 142)
P Value
Effect Size
53 (28–74) 107 (75.4) 120 (84.5) 10 (7.1) 7 (4.9) 91 (64.1) 7 (4.9) 34 (23.9) 2 (1.1–3.0) 76 (53.5) 22.2 (1.0–5517.3)
,.001 .698 .012 .020 .439 ,.001 .001 ,.001 .327 .303 .628
0.581 0.037 0.295 0.099 0.044 0.352 0.519 0.571 0.103 0.161 0.184
.025 ,.001 .214 ,.001 ,.001
0.231 0.823 0.312 1.004 1.433
102 (41.5) 142.5 (48–309) 0.7 (0.3–1.7) 4.1 (2.9–5.1) 1.1 (0.9–1.3)
Note.—Unless otherwise indicated, data are numbers of patients, with percentages in parentheses. Log transformation was used for analysis of a-fetoprotein concentration and platelet count. Effect size values less than 0.10 indicate very small differences, 0.1–0.3 indicate small differences, 0.3–0.5 indicate moderate differences, and greater than 0.5 indicate large differences. INR = international normalized ratio. * Data are medians, with the range in parentheses. †
To convert to Système International (SI) units (micrograms per liter), multiply by 1.
‡
To convert to SI units (micromoles per liter), multiply by 17.104.
§
To convert to SI units (grams per liter), multiply by 10.
NAR group, with the exception of total bilirubin level, for which there was no significant difference between the groups. However, tumor profile, index tumor size, proportion of tumors that were BCLC stage 0 HCC (, 2 cm), and a-fetoprotein concentration before treatment were not significantly different between the groups. All 142 patients in the NAR group had undergone R0 resection and histopathologic analysis (mean resection margin, 1.0 cm; range, 0.1–2.3 cm). For seven (4.9%) patients, NAR was performed by using a laparoscopic approach. Other baseline characteristics of patients included in both groups are shown in Table 1.
Comparison of Therapeutic Outcomes for All Study Patients before Propensity Score Matching Tumor recurrence.—At the time of censoring, local tumor progression had developed in 84 of 438 (19.2%)
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
patients in the RF ablation group and in 19 of 142 (13.4%) patients in the NAR group. The cumulative local tumor progression rates at 1, 3, and 5 years were 6.6%, 16.4%, and 20.9%, respectively, for the RF ablation group and 7.0%, 10.0%, and 12.7%, respectively, for the NAR group. The differences between the groups were not significant (P = .093). Intrahepatic distant recurrence was identified in 250 of 438 (57.1%) patients in the RF ablation group and in 50 of 142 (35.2%) patients in the NAR group. Unlike rates of local tumor progression, cumulative intrahepatic distant recurrence rates were significantly higher in the RF ablation group than those in the NAR group (16.0% vs 13.4% at 1 year, 49.7% vs 27.2% at 3 years, and 62.7% vs 36.6% at 5 years, respectively; P , .001) (Fig 2). Extrahepatic recurrence occurred in 34 of 438 (7.8%) patients in the RF ablation group and nine of 5
VASCULAR AND INTERVENTIONAL RADIOLOGY: Radiofrequency Ablation versus Resection for Small Hepatocellular Carcinoma
Kang et al
Figure 2
Figure 2: Graphs show types of recurrence and survival curves in patients with small HCCs who underwent RF ablation (RFA) and NAR. Between the two groups of study patients (n = 580), (a) cumulative local tumor progression and (d) overall survival rates were not significantly different. However, (b) cumulative intrahepatic distant recurrence and (c) disease-free survival rates were significantly better in NAR group than in RF ablation group. IDR = intrahepatic distant recurrence, LTP = local tumor progression.
142 (6.3%) patients in the NAR group (P = .480). Regarding the second-line treatment modality for initial tumor 6
recurrence, there were no significant differences between the groups (P . .05) (Table 2).
Disease-free and overall survival.— During the follow-up period, 51 of 438 (11.6%) patients in the RF ablation
radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
VASCULAR AND INTERVENTIONAL RADIOLOGY: Radiofrequency Ablation versus Resection for Small Hepatocellular Carcinoma
Kang et al
Table 2 Second-Line Treatment Modalities for Patients with the Initial Tumor Recurrence in Both Groups Type of Tumor Recurrence Intrahepatic Distant Recurrence (n = 300)†
Local Tumor Progression (n = 103)* Second-Line Treatment RF ablation TACE RF ablation and TACE Surgical resection Liver transplantation Sorafenib Not available‡
RF Ablation Group (n = 84)
NAR Group (n = 19)
RF Ablation Group (n = 250)
NAR Group (n = 50)
36 (42.8) 41 (48.8) 5 (6.0) 1 (1.2) 0 0 1 (1.2)
6 (31.5) 11 (57.9) 1 (5.3) 1 (5.3) 0 0 0
86 (34.3) 132 (52.8) 18 (7.3) 4 (1.6) 3 (1.2) 2 (0.8) 5 (2.0)
19 (38) 28 (56) 1 (2) 1 (2) 1 (2) 0 0
Note.—Data are number of patients, with percentages in parentheses. Sorafenib was used for advanced-stage HCCs according to the BCLC clinical guidelines and temporal availability. TACE = transcather arterial chemoembolization * P = .568. †
P = .740.
‡
These patients did not receive any treatment for recurrent HCC at the time of censoring.
group and 12 of 142 (8.5%) patients in the NAR group died. The diseasefree survival rates at 3 and 5 years were estimated to be 43.8% and 31.7%, respectively, in the RF ablation group and 71.5% and 61.1%, respectively, in the NAR group. The difference was significant (P , .001). However, both groups had similar overall survival rates of 92.4% and 85.5% in the RF ablation group and 93.4% and 90.9% in the NAR group at 3 and 5 years, respectively (P = .194) (Fig 2). Evaluation of variables affecting disease-free and overall survival.—At multivariate analysis of all study patients (n = 580), NAR treatment (P = .002; hazard ratio [HR], 0.60; 95% confidence interval [CI]: 0.43, 0.83), tumor size (P = .027; HR, 1.29; 95% CI: 1.03, 1.62), hepatitis C viral infection (P = .034; HR, 1.59; 95% CI: 1.04, 2.45), and serum albumin level (P = .045; HR, 0.73; 95% CI: 0.54, 0.99) were independent prognostic factors for disease-free survival. For overall survival, tumor size (P = .015; HR, 1.82; 95% CI: 1.12, 2.96), serum albumin level (P = .003; HR, 0.35; 95% CI: 0.18, 0.70), and hepatitis C viral infection (P = .005; HR, 2.30; 95% CI: 1.28, 4.13) were significant independent prognostic factors (Table 3).
Comparison of Therapeutic Outcomes after Propensity Score Matching To balance the clinical variables and control for confounding bias, a total of 198 patients (99 from each group) were matched by applying one-toone propensity score matching. This sample size after matching allowed detection of a difference of at least 18% in the 5-year local tumor progression rate, the primary outcome in our study, with 80% power and a significance level of 5%. Matched pairs showed similar baseline characteristics (Table 4). In the matched cohort, the cumulative local tumor progression rates in the RF ablation group at 1, 3, and 5 years were 5.1%, 14.5%, and 19.4%, respectively, and 8.1%, 12.3%, and 18.6%, respectively, in the NAR group (P = .860). The 3- and 5-year overall survival rates were 95.7% and 87.2%, respectively, in the RF ablation group, and 92.7% and 90.2%, respectively, in the NAR group, a difference that was not significant (P = .594). In addition, the cumulative intrahepatic distant recurrence and disease-free survival rates did not differ between the two groups, because these intergroup differences were markedly decreased after propensity score matching: cumulative intrahepatic distant recurrence rates
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
(14.1% vs 16.2% at 1 year; 38.4% vs 30.9% at 3 years; and 47.0% vs 40.2% at 5 years); disease-free survival rates (53.3% vs 67.1% at 3 years; and 48.9% vs 54.4% at 5 years), and RF ablation versus NAR (P = .240 and P = .201, respectively) (Fig 3).
Major Complications and Posttreatment Hospital Stay There were no treatment-related or inhospital deaths in either group (Table 5). Posthepatectomy liver failure was not observed in the NAR group according to the Dindo-Clavien classification. Major complications were significantly more common in the NAR group (eight of 142 [5.6%] patients) than in the RF ablation group (nine of 438 [2.1%]; P = .016). In addition, the length of stay after treatment was significantly longer in the NAR group (median, 8.5 days; range, 5–22 days) than in the RF ablation group (median, 1.0 day; range, 1–3 days; P , .001). Discussion In this retrospective study, we found that the long-term therapeutic efficacy of percutaneous RF ablation was similar to that of NAR as first-line treatment of small ( 3 cm) HCCs. Unlike anatomic resection, RF ablation and NAR are 7
VASCULAR AND INTERVENTIONAL RADIOLOGY: Radiofrequency Ablation versus Resection for Small Hepatocellular Carcinoma
Kang et al
Table 3 Evaluation of Variables for Effects on Disease-free and Overall Survival Disease-free Survival Univariate Analysis Variable Treatment type Age at enrollment Men Hepatitis B positive Antihepatitis C positive Other cause of liver disease Presence of liver cirrhosis Child-Pugh class B Portal hypertension Tumor size a-fetoprotein concentration Antiviral treatment Platelet count Total bilirubin Albumin Prothrombin time
Overall Survival
Multivariate Analysis
Univariate Analysis
Multivariate Analysis
HR
P Value
HR
P Value
HR
P Value
HR
P Value
0.45 (0.33, 0.60) 1.02 (1.01, 1.04) 1.18 (0.91, 1.54) 0.71 (0.56, 0.89) 1.94 (1.50, 2.50) 0.88 (0.56, 1.36) 1.40 (1.07, 1.83) 1.52 (1.12, 2.05) 1.48 (1.20, 1.84) 1.27 (1.03, 1.57) 0.99 (0.99, 1.00) 0.93 (0.75, 1.16) 0.99 (0.99, 1.00) 1.21 (0.97, 1.52) 0.52 (0.42, 0.64) 6.27 (2.73, 14.43)
,.001 ,.001 .210 .004 ,.001 .550 .014 .007 ,.001 .029 .380 .520 ,.001 .096 ,.001 ,.001
0.60 (0.43, 0.83) 1.01 (1.00, 1.03) ... 1.13 (0.77, 1.66) 1.59 (1.04, 2.45) ... 1.06 (0.81, 1.40) 0.80 (0.51, 1.24) 1.06 (0.80, 1.42) 1.29 (1.03, 1.62) ... ... 0.95 (0.68, 1.34) 1.01 (0.76, 1.35) 0.73 (0.54, 0.99) 2.43 (0.61, 9.71)
.002 .057 ... .520 .034 ... .670 .310 .680 .027 ... ... .770 .940 .045 .210
0.66 (0.35, 1.24) 1.01 (0.98, 1.03) 1.01 (0.56, 1.80) 2.09 (0.29, 15.21) 5.83 (0.78, 43.81) 1.29 (0.12, 14.19) 2.04 (0.97, 4.29) 1.24 (0.63, 2.44) 1.81 (1.10, 2.97) 1.91 (1.19, 3.06) 0.99 (0.99, 1.00) 0.72 (0.44, 1.20) 0.99 (0.99, 1.00) 1.11 (0.66, 1.85) 0.33 (0.21, 0.52) 7.58 (1.55, 36.95)
.194 .614 .981 .465 .087 .837 .059 .534 .020 .007 .504 .209 .061 .697 ,.001 .012
1.10 (0.55, 2.19) ... ... ... 2.30 (1.28, 4.13) ... 1.54 (0.70, 3.39) ... 1.24 (0.65, 2.38) 1.82 (1.12, 2.96) ... ... 1.10 (0.49, 2.44) ... 0.35 (0.18, 0.70) 0.36 (0.02, 5.46)
.798 ... ... ... .006 ... .285 ... .516 .016 ... ... .822 ... .003 .460
Note.—Data in parentheses are 95% CIs. Log transformation was used for analysis of a-fetoprotein concentration and platelet count. Each variable with a P value of less than or equal to .20 at univariate analysis was entered into the final model. All statistical analyses for prognostic factors were conducted with Cox proportional hazards models.
Table 4 Demographic and Clinical Characteristics of the RF Ablation and NAR Groups after Propensity Analysis Variable Age at enrollment (y)* No. of men Hepatitis B positive Anti-hepatitis C positive Other cause of liver disease Liver cirrhosis Child-Pugh class B Portal hypertension Tumor size* BCLC stage 0 HCC (, 2 cm) a-fetoprotein concentration (ng/mL)*† Antiviral treatment Platelet count (x 109/L)* Total bilirubin (mg/dL)*‡ Albumin (g/dL)*§ Prothrombin time (INR)*
RF Ablation Group (n = 99)
NAR Group (n = 99)
P Value
Effect Size
55 (32–80) 77 (77.8) 83 (83.8) 8 (8.1) 4 (4.0) 71 (71.7) 4 (4.0) 30 (30.3) 1.9 (1.1–3.0) 50 (50.5) 25.6 (1.0–1873) 71 (71.7) 131 (50–395) 0.7 (0.2–1.9) 4.1 (2.3–4.8) 1.1 (0.9–1.3)
54 (31–74) 77 (77.8) 83 (83.8) 8 (8.1) 4 (4.0) 71 (71.7) 4 (4.0) 29 (29.3) 2 (1.1–3.0) 46 (46.5) 15.2 (1.0–3412.2) 68 (68.7) 132 (50–254) 0.7 (0.3–1.7) 4.1 (2.9–5.1) 1.1 (0.9–1.3)
.506 ..999 ..999 ..999 ..999 ..999 ..999 .866 .728 .688 .747 .742 .710 .806 .603 .806
0.071 0 0 0 0 0 0 0.024 0.078 0.082 0.044 0.067 0.005 0.003 0.059 0.023
Note.—Unless otherwise indicated, data are numbers of patients, with percentages in parentheses. One-to-one nearest-neighbor individual matching without replacement with propensity scoring was used. Effect size values less than 0.10 indicate very small differences, 0.1–0.3 indicate small differences, 0.3–0.5 indicate moderate differences, and greater than 0.5 indicate large differences. Paired comparisons were used for all statistical analyses. INR = international normalized ratio. * Data are medians, with the range in parentheses.
8
†
To convert to Système International (SI) units (micrograms per liter), multiply by 1.
‡
To convert to SI units (micromoles per liter), multiply by 17.104.
§
To convert to SI units (grams per liter), multiply by 10.
radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
VASCULAR AND INTERVENTIONAL RADIOLOGY: Radiofrequency Ablation versus Resection for Small Hepatocellular Carcinoma
Kang et al
Figure 3
Figure 3: Graphs show types of recurrence and survival curves after propensity score matching for patients with small HCCs who underwent RF ablation (RFA) or NAR. After propensity score matching (n = 198), there were no significant differences in any of the measured therapeutic outcomes between groups. IDR = intrahepatic distant recurrence, LTP = local tumor progression.
important curative treatment options that allow preservation of functioning hepatic parenchyma, especially in patients with cirrhosis. These techniques
also improve the ability to perform additional multimodality treatments including re-resection, RF ablation, and transcatheter arterial chemoembolization in
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
cases of recurrent HCCs and prevent posthepatectomy liver failure (6). In our study, cumulative local tumor progression rates were not significantly 9
VASCULAR AND INTERVENTIONAL RADIOLOGY: Radiofrequency Ablation versus Resection for Small Hepatocellular Carcinoma
Table 5 Major Complications and Hospital Stay after Each Treatment Outcome In-hospital death Treatment-related death Posttreatment hospital stay (d) Major complications Tumor seeding Intra-abdominal abscess Biloma Wound problem Hepatic abscess Abdominal bleeding Bile duct stricture Intestinal obstruction Hepatic infarction
RF Ablation Group (n = 438)
NAR Group (n = 142)
0 0 1.0 (1–3)* 9 (2.1)† 3
0 0 8.5 (5–22)* 8 (5.6)† 3
2 3 2 1 1 1 1
Note. Unless otherwise indicated, data are numbers of patients with percentages in parentheses. * Data are medians, with range in parentheses. P , .001 for comparison of posttreatment hospital stay for RF and NAR groups with Wilcoxon rank sum test. †
P = .016 for comparison of major complications between RF and NAR groups. After matching, major complications in RF and NAR groups estimated at 1% (1 of 99) and 7% (7 of 99), respectively; P = .034, McNemar test.
different between the RF ablation and NAR groups before and after propensity score matching. This equivalent treatment outcome for local tumor control might have been associated with small tumor size in our study patients because microsatellite nodules are less frequently identified in patients with small ( 3 cm) HCCs than in those with large HCCs (30). In clinical practice, achieving an ideal circumferential ablative margin that includes these peritumoral microsatellite nodules is difficult in patients with large tumors due to an insufficient ablation zone, even with the use of cluster or multiple single electrodes (5,31). In addition, NAR with a relatively narrow resection margin (mean; 1.0 cm), as was used in our surgical resection group, could have led to unfavorable local tumor control compared with anatomic resection (6,25,32). Inclusion of only small HCCs and the use of NAR in our study patients may have offset the disadvantage of using RF ablation for local tumor control. NAR yielded better cumulative intrahepatic distant recurrence and diseasefree survival rates than did RF ablation in the pooled cohort (n = 580). However the results were similar after propensity 10
score adjustment in the matched cohort (n = 198). These changes may be explained by differences in demographic and clinical characteristics between the groups. Poor liver condition in the RF ablation group could have resulted in higher intrahepatic distant recurrence rates because of multistep or de novo carcinogenesis that may have arisen from preneoplastic liver tissue associated with lower serum albumin levels or cirrhosis (33,34). In addition, patients in the RF ablation group were more likely to have hepatitis C viral infection. Unlike hepatitis B, hepatitis C is associated with a higher tumor recurrence rate after RF ablation (12) or hepatic resection (35). Intrahepatic distant recurrence was important to the difference in diseasefree survival between the groups before propensity score adjustment compared with local tumor progression and extrahepatic recurrence. In contrast to our results, authors of a recent study who used RF ablation and surgical resection for tumors of similar sizes reported that poor local tumor control (local tumor progression) rather than intrahepatic distant recurrence in the RF ablation group was the main contributor
Kang et al
to disease-free survival (14). Differences between their study and ours in the number of patients who underwent NAR in the resection group could explain these findings, because NAR can lead to poor local tumor control and early tumor recurrence due to less effective clearance of micrometastases and microvascular invasion around the tumor (6,25,36). We suspect that the differences in the proportion of patients who underwent NAR in the surgical resection groups might have contributed to our results conflicting with those of previous randomized clinical trials (37– 39) and two recent nationwide surveys (8,14), in which the authors compared RF ablation and surgical resection for the treatment of HCCs. With respect to overall survival, there were no significant differences between the two groups before or after propensity score matching (P = .194 and 0.594, respectively). These results are consistent with those of previous studies on the treatment of small HCCs with RF ablation and surgical resection (11,14). The potential effects of first line-treatment may be smaller in HCCs than those observed in other primary malignant tumors, because the function of the remaining liver tissue considerably affects tumor recurrence and overall survival during the follow-up period as well as the efficacy of initial tumor treatment (8,13,38). In addition, the HCC recurrence rate was estimated to be 70% at 5 years after initial treatment, reflecting either intrahepatic metastases (true recurrences) or the development of de novo tumors (15). As a result, second-line treatment is important to improving overall survival. Nearly half of the patients in our matched cohort experienced tumor recurrence during the scheduled follow-up period, despite having small HCCs with BCLC stages 0 and A. In our study, we found that RF ablation was associated with fewer major complications and shorter hospital stays after treatment, which indicates that RF ablation is a relatively safe and noninvasive modality for treating HCCs. These results corresponded with those of a previous meta-analysis in which
radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
VASCULAR AND INTERVENTIONAL RADIOLOGY: Radiofrequency Ablation versus Resection for Small Hepatocellular Carcinoma
RF ablation and hepatic resection were compared (40). Our study had several limitations. First, our retrospective study design may have resulted in selection bias. Although we attempted to simulate randomization by using propensity score analysis (28), there remained a possibility of uncontrolled confounding factors between the groups. Second, the small number of biopsy-proven HCCs in the RF ablation group may have influenced therapeutic outcomes. However, all patients without a histopathologic diagnosis met the diagnostic criteria for HCC according to the guidelines in place at the time of initial diagnosis (19,20). Third, because this was a single-center study and there was no universal consensus on the indications for NAR for the treatment of small HCCs, our results might not be consistently reproducible in other settings. Despite these limitations, our findings may help to validate the current Barcelona treatment strategy by contributing additional clinical evidence that compares RF ablation with NAR through propensity score analysis of a large retrospective cohort with long-term follow-up data. To our knowledge, previous investigators have not taken the type of surgical resection into consideration when designing studies comparing RF ablation and surgical resection. Additional randomized controlled trials are needed to define the best treatment for small HCCs in patients with chronic liver disease. There may be differences between RF ablation and major hepatectomy (resection of three or more Couinaud segments) regarding the remnant liver reservoir, which is important for liver function in patients with chronic liver disease, for potential intrahepatic distant recurrence associated with preneoplastic cirrhotic liver tissue, and for determination of second-line treatment in cases of recurrence. In our population, major hepatectomy was used in a considerable proportion of the patients who underwent surgical resection for small HCCs (45 of 378, 11.9%). In summary, there were no significant differences in long-term therapeutic outcomes including local tumor
progression, intrahepatic distant recurrence, disease-free survival, and overall survival between RF ablation and NAR as first-line treatment in patients with small BCLC stage 0 or A HCCs. In addition, RF ablation resulted in a lower rate of major complications and shorter posttreatment hospital stay. Disclosures of Conflicts of Interest: T.W.K. disclosed no relevant relationships. J.M.K. disclosed no relevant relationships. H.R. disclosed no relevant relationships. M.W.L. disclosed no relevant relationships. Y.S.K. disclosed no relevant relationships. H.K.L. disclosed no relevant relationships. D.C. disclosed no relevant relationships. K.D.S. disclosed no relevant relationships. C.H.D.K. disclosed no relevant relationships. J.W.J. disclosed no relevant relationships. S.W.P. disclosed no relevant relationships. S.H.P. disclosed no relevant relationships. J.H.A. disclosed no relevant relationships.
References 1. Forner A, Llovet JM, Bruix J. Hepatocellular carcinoma. Lancet 2012;379(9822):1245– 1255. 2. Strasberg SM. Nomenclature of hepatic anatomy and resections: a review of the Brisbane 2000 system. J Hepatobiliary Pancreat Surg 2005;12(5):351–355. 3. Cucchetti A, Cescon M, Trevisani F, Pinna AD. Current concepts in hepatic resection for hepatocellular carcinoma in cirrhotic patients. World J Gastroenterol 2012;18(44): 6398–6408. 4. Regimbeau JM, Kianmanesh R, Farges O, Dondero F, Sauvanet A, Belghiti J. Extent of liver resection influences the outcome in patients with cirrhosis and small hepatocellular carcinoma. Surgery 2002;131(3):311– 317. 5. Kim YS, Lee WJ, Rhim H, Lim HK, Choi D, Lee JY. The minimal ablative margin of radiofrequency ablation of hepatocellular carcinoma (. 2 and , 5 cm) needed to prevent local tumor progression: 3D quantitative assessment using CT image fusion. AJR Am J Roentgenol 2010;195(3):758–765. 6. Shi M, Guo RP, Lin XJ, et al. Partial hepatectomy with wide versus narrow resection margin for solitary hepatocellular carcinoma: a prospective randomized trial. Ann Surg 2007;245(1):36–43. 7. Mori Y, Tamai H, Shingaki N, et al. Diffuse intrahepatic recurrence after percutaneous radiofrequency ablation for solitary and small hepatocellular carcinoma. Hepatol Int 2009;3(3):509–515.
Radiology: Volume 000: Number 0— 2015 n radiology.rsna.org
Kang et al
8. Hasegawa K, Kokudo N, Makuuchi M, et al. Comparison of resection and ablation for hepatocellular carcinoma: a cohort study based on a Japanese nationwide survey. J Hepatol 2013;58(4):724–729. 9. Yun WK, Choi MS, Choi D, et al. Superior long-term outcomes after surgery in childpugh class a patients with single small hepatocellular carcinoma compared to radiofrequency ablation. Hepatol Int 2011;5(2): 722–729. 10. Hong SN, Lee SY, Choi MS, et al. Comparing the outcomes of radiofrequency ablation and surgery in patients with a single small hepatocellular carcinoma and wellpreserved hepatic function. J Clin Gastroenterol 2005;39(3):247–252. 11. Nishikawa H, Inuzuka T, Takeda H, et al. Comparison of percutaneous radiofrequency thermal ablation and surgical resection for small hepatocellular carcinoma. BMC Gastroenterol 2011;11:143. 12. Shiina S, Tateishi R, Arano T, et al. Radiofrequency ablation for hepatocellular carcinoma: 10-year outcome and prognostic factors. Am J Gastroenterol 2012;107(4): 569–577; quiz 578. 13. Kim YS, Lim HK, Rhim H, et al. Ten-year outcomes of percutaneous radiofrequency ablation as first-line therapy of early hepatocellular carcinoma: analysis of prognostic factors. J Hepatol 2013;58(1):89–97. 14. Pompili M, Saviano A, de Matthaeis N, et al. Long-term effectiveness of resection and radiofrequency ablation for single hepatocellular carcinoma 3 cm. Results of a multicenter Italian survey. J Hepatol 2013; 59(1):89–97. 15. European Association For The Study Of The Liver; European Organisation For Research And Treatment Of Cancer. EASL-EORTC clinical practice guidelines: management of hepatocellular carcinoma. J Hepatol 2012;56(4):908–943. 16. Giannini EG, Savarino V, Farinati F, et al. Influence of clinically significant portal hypertension on survival after hepatic resection for hepatocellular carcinoma in cirrhotic patients. Liver Int 2013;33(10):1594– 1600. 17. Shimozawa N, Hanazaki K. Longterm prognosis after hepatic resection for small hepatocellular carcinoma. J Am Coll Surg 2004;198(3):356–365. 18. Goldberg SN, Charboneau JW, Dodd GD 3rd, et al. Image-guided tumor ablation: proposal for standardization of terms and reporting criteria. Radiology 2003;228(2): 335–345.
11
VASCULAR AND INTERVENTIONAL RADIOLOGY: Radiofrequency Ablation versus Resection for Small Hepatocellular Carcinoma
19. Bruix J, Sherman M; Practice Guidelines Committee, American Association for the Study of Liver Diseases. Management of hepatocellular carcinoma. Hepatology 2005;42(5): 1208–1236.
26. Kim JM, Kwon CH, Joh JW, et al. Differences between hepatocellular carcinoma and hepatitis B virus infection in patients with and without cirrhosis. Ann Surg Oncol 2014;21(2):458–465.
20. Bruix J, Sherman M, Llovet JM, et al. Clinical management of hepatocellular carcinoma. Conclusions of the Barcelona-2000 EASL conference. European Association for the Study of the Liver. J Hepatol 2001;35(3): 421–430.
27. Omary RA, Bettmann MA, Cardella JF, et al. Quality improvement guidelines for the reporting and archiving of interventional radiology procedures. J Vasc Interv Radiol 2003;14 (9 Pt 2):S293–S295.
21. Peng ZW, Lin XJ, Zhang YJ, et al. Radiofrequency ablation versus hepatic resection for the treatment of hepatocellular carcinomas 2 cm or smaller: a retrospective comparative study. Radiology 2012;262(3):1022–1033. 22. Rhim H, Choi D, Kim YS, Lim HK, Choe BK. Ultrasonography-guided percutaneous radiofrequency ablation of hepatocellular carcinomas: a feasibility scoring system for planning sonography. Eur J Radiol 2010;75(2): 253–258. 23. Kang TW, Lim HK, Lee MW, Kim YS, Choi D, Rhim H. Perivascular versus nonperivascular small HCC treated with percutaneous RF ablation: retrospective comparison of long-term therapeutic outcomes. Radiology 2014;270(3):888–899. 24. Kang CM, Choi GH, Kim DH, et al. Revisiting the role of nonanatomic resection of small (, or = 4 cm) and single hepatocellular carcinoma in patients with well-preserved liver function. J Surg Res 2010;160(1):81–89. 25. Cucchetti A, Qiao GL, Cescon M, et al. Anatomic versus nonanatomic resection in cirrhotic patients with early hepatocellular carcinoma. Surgery 2014;155(3):512–521.
12
28. McDonald RJ, McDonald JS, Kallmes DF, Carter RE. Behind the numbers: propensity score analysis-a primer for the diagnostic radiologist. Radiology 2013;269(3):640–645. 29. Burnand B, Kernan WN, Feinstein AR. Indexes and boundaries for “quantitative significance” in statistical decisions. J Clin Epidemiol 1990;43(12):1273–1284. 30. Shi M, Zhang CQ, Zhang YQ, Liang XM, Li JQ. Micrometastases of solitary hepatocellular carcinoma and appropriate resection margin. World J Surg 2004;28(4):376–381. 31. Livraghi T, Goldberg SN, Lazzaroni S, et al. Hepatocellular carcinoma: radio-frequency ablation of medium and large lesions. Radiology 2000;214(3):761–768. 32. Hasegawa K, Kokudo N, Imamura H, et al. Prognostic impact of anatomic resection for hepatocellular carcinoma. Ann Surg 2005; 242(2):252–259. 33. Park YN. Update on precursor and early lesions of hepatocellular carcinomas. Arch Pathol Lab Med 2011;135(6):704–715.
Kang et al
carcinoma: extending the indication for resection? Surgery 2012;152(5):809–820. 35. Huang YH, Wu JC, Chen CH, et al. Comparison of recurrence after hepatic resection in patients with hepatitis B vs. hepatitis C-related small hepatocellular carcinoma in hepatitis B virus endemic area. Liver Int 2005;25(2):236–241. 36. Ueno S, Kubo F, Sakoda M, et al. Efficacy of anatomic resection vs nonanatomic resection for small nodular hepatocellular carcinoma based on gross classification. J Hepatobiliary Pancreat Surg 2008;15(5):493–500. 37. Huang J, Yan L, Cheng Z, et al. A randomized trial comparing radiofrequency ablation and surgical resection for HCC conforming to the Milan criteria. Ann Surg 2010;252(6): 903–912. 38. Feng K, Yan J, Li X, et al. A randomized controlled trial of radiofrequency ablation and surgical resection in the treatment of small hepatocellular carcinoma. J Hepatol 2012; 57(4):794–802. 39. Chen MS, Li JQ, Zheng Y, et al. A prospective randomized trial comparing percutaneous local ablative therapy and partial hepatectomy for small hepatocellular carcinoma. Ann Surg 2006;243(3):321–328. 40. Duan C, Liu M, Zhang Z, Ma K, Bie P. Radiofrequency ablation versus hepatic resection for the treatment of early-stage hepatocellular carcinoma meeting Milan criteria: a systematic review and meta-analysis. World J Surg Oncol 2013;11(1):190.
34. Chang WT, Kao WY, Chau GY, et al. Hepatic resection can provide long-term survival of patients with non-early-stage hepatocellular
radiology.rsna.org n Radiology: Volume 000: Number 0— 2015
