Clinical Radiology xxx (2017) 1e8

Contents lists available at ScienceDirect

Clinical Radiology
journal homepage: www.clinicalradiologyonline.net

Initial radiofrequency ablation failure for

hepatocellular carcinoma: repeated
radiofrequency ablation versus transarterial
chemoembolisation
S.S. Kim a, T.W. Kang a, *, M. Kim a, M.W. Lee a, S.K. Cho a, Y.H. Paik b, c,
M.-J. Kim d
a
Department of Radiology and Center for Imaging Science, Samsung Medical Center, Sungkyunkwan University
School of Medicine, Seoul, Republic of Korea
b
Division of Hepatology, Department of Medicine, Samsung Medical Center, Sungkyunkwan University School of
Medicine, Seoul, Republic of Korea
c
Department of Health Sciences and Technology, SAIHST, Sungkyunkwan University, Seoul, Republic of Korea
d
Biostatics and Clinical Epidemiology Center, Samsung Biomedical Research Institute, Seoul, Republic of Korea

art icl e i nformat ion AIM: To compare the long-term therapeutic outcomes of repeated radiofrequency ablation
(RFA) with that of transarterial chemoembolisation (TACE) in patients with local tumour
Article history: progression (LTP) after initial RFA treatment for hepatocellular carcinoma (HCC).
Received 12 May 2017 MATERIALS AND METHODS: This retrospective study was approved by the institutional
Received in revised form review board and the requirement for informed consent was waived. Between July 2006 and
13 July 2017 February 2012, 713 patients underwent RFA for single HCC as a first-line treatment. Fifty-eight
Accepted 24 July 2017 patients who showed LTP as initial tumour recurrence post-RFA treatment were included.
Patients were treated with either repeated RFA (n¼33) or TACE (n¼25). TACE was performed
as an alternative therapeutic option when repeated RFA was not feasible based on the plan-
ning ultrasonography. Recurrence-free and overall survival rates were estimated using the
KaplaneMeier method. Prognostic factors for outcomes were evaluated using the Cox pro-
portional hazards model.
RESULTS: Both groups did not show significant differences in terms of baseline character-
istics, with the exception being the proportion of subphrenic tumours (p¼0.031). The RFA and
TACE groups did not differ significantly in their 5-year recurrence-free and overall survival
rates (17% versus 10.7% and 72.7% versus 51.9%, respectively, with all p-values >0.05). In
addition, multivariate analyses revealed that type of treatment was not associated with
recurrence-free or overall survival in patients with post-RFA LTP.
CONCLUSION: TACE is an effective treatment, comparable to repeated RFA, in patients with
LTP after initial RFA when repeated RFA is not feasible.
Ó 2017 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.

* Guarantor and correspondent: T. W. Kang, Department of Radiology and Center for Imaging Science, Samsung Medical Center, Sungkyunkwan University
School of Medicine and Department of Health Sciences and Technology, SAIHST, Sungkyunkwan University, Gangnam-gu, Irwon-ro 81, Seoul 135-710, Republic
of Korea. Tel.: þ822 3410 0518; fax: þ822 3410 2559.
E-mail addresses: garamond@hanamil.net, kaienes.kang@samsung.com (T.W. Kang).

http://dx.doi.org/10.1016/j.crad.2017.07.020
0009-9260/Ó 2017 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Kim SS, et al., Initial radiofrequency ablation failure for hepatocellular carcinoma: repeated radiofrequency
ablation versus transarterial chemoembolisation, Clinical Radiology (2017), http://dx.doi.org/10.1016/j.crad.2017.07.020
2 S.S. Kim et al. / Clinical Radiology xxx (2017) 1e8
Introduction
Radiofrequency ablation (RFA) has been added to the
international hepatocellular carcinoma (HCC) management
guidelines as a curative treatment modality for very early or
early-stage HCCs1e4; however, despite recent advances in
RFA technology,5 the risk of local tumour progression (LTP)
following RFA treatment for HCC is still problematic. For
patients subjected to RFA treatment, tumour recurrence at 5
year is 4e27%.6e8 This is due to the inherent limitations of
the local thermal ablation method: it does not allow for the
systemic removal of a hepatic segment to eliminate minute
satellites around the main tumour,9,10 and it is difficult to
evaluate the exact tumour-free margin during the RFA
procedure.11
Recently, the utilisation of RFA as a treatment of HCC has
markedly increased over time.12 Therefore the number of
patients with post-treatment LTP would have increased. In
clinical practice, LTP in patients who have undergone RFA as
the first-line treatment for HCC due to difficulty of surgical
treatment are usually treated with local-regional therapies
such as repeated RFA or transarterial chemoembolisation
(TACE) due to the same reason; however, treating LTP with
RFA is difficult as LTP after RFA tends to occur in high-risk
locations, such as subphrenic or perivascular areas,
because such sites are often associated with technical dif-
ficulties during the initial RFA procedure. Therefore, the
high-risk location would hamper the achievement of com-
plete ablation with sufficient ablative margins.13,14 In these
cases, treating LTP with RFA is often difficult and TACE could
be an alternative treatment; however, few data exist for
outlining an optimal protocol for the treatment of LTP after
initial RFA.15 In addition, there have been no studies
comparing the therapeutic efficacy of these two modalities
for LTP treatment after initial RFA.
Therefore, the aim of the present study was to evaluate
retrospectively the recurrence-free and overall survival of
repeated RFA and TACE in LTP patients who had undergone
RFA for HCC as a first-line treatment. A second aim was to
analyse the prognostic factors for these outcomes.
Materials and methods
The institutional review board approved this retrospec-
tive study and waived the requirement for informed con-
sent. The terminology used for the description of the RFA
and TACE procedures followed the proposed guidelines.16,17
Patients
The inclusion criteria of RFA for patients with HCCs were
identical to those described in previous studies6,14: (a) very
early or early stage HCCs defined as a single tumour 5 cm
in diameter, or 2e3 nodules 3 cm in diameter without
extra-hepatic metastasis or any macro-vascular invasion
(Barcelona Clinic Liver Cancer [BCLC] stage 0 or A); (b)
ChildePugh class A or B; and (c) a prothrombin time within
the normal range and platelet count 50,000 cells/ml3.
Please cite this article in press as: Kim SS, et al., Initial radiofrequency ablation failure for hepatocellular carcinoma: repeated radiofrequency
ablation versus transarterial chemoembolisation, Clinical Radiology (2017), http://dx.doi.org/10.1016/j.crad.2017.07.020
Between July 2006 and February 2012, 713 consecutive
patients underwent RFA for single HCC as a first-line
treatment at Samsung Medical Center, Sunkyunkwan Uni-
versity, Seoul, Korea. First-line treatment was defined as the
absence of prior treatment during HCC diagnosis. Among
them, 126 patients (17.7%) exhibited LTPs during follow-up
after RFA treatment.
The LTP diagnosis was based on imaging results and
indicated by a newly developed lesion, showing arterial
enhancement and delayed wash-out around the previously
treated zone.16,18 From these patients, 69 patients who
exhibited LTPs as the initial tumour recurrence were
selected to exclude patients with possible intrahepatic
metastasis from other untreated tumours. Among these
patients, 11 were excluded from the study for the following
reasons: (a) patients who had been subjected to other
treatments besides repeated RFA or TACE (n¼6); (b) pa-
tients with tumour thrombus along with the LTP (n¼2); (c)
loss to follow-up at 1 month after treatment (n¼1); and (d)
the technical failure of repeated RFA or TACE to treat the LTP
(n¼2). The remaining 58 patients (37 men and 21 women;
age range, 38e78 years) comprised the study cohort
(Electronic Supplementary Material Fig. S1). TACE was
performed as an alternative therapeutic option for LTP
when repeated RFA was not feasible on planning ultraso-
nography (US) for the following reasons: poor sonic win-
dow or RF electrode path (n¼19) and poor lesion
conspicuity (n¼6). Treatment technical failure was defined
as the presence of an enhancing tumour on 1-month
follow-up computed tomography (CT) after the initial
treatment. Two patients (3.3%, 2/61) with incomplete
treated tumours after initial treatment underwent surgical
resection or RFA for tumour control and they were excluded
from recurrence analysis.
Repeated RFA
The planning US for RFA was performed at an outpatient
clinic to assess the feasibility of repeated RFA.19 When
considered feasible, patients were treated with repeated
RFA on an inpatient basis by four different radiologists
(H.K.L., H.R., Y.S.K. and M.W.L.), each with at least 7 years of
clinical experience in RFA before the starting point of this
study.
All procedures were performed percutaneously under US
guidance with local anaesthesia and conscious sedation.
They were performed by using one of the following US
systems: iU22 (Philips Medical Systems, Bothell, WA, USA),
LOGIQ, E9 (GE Healthcare, Waukesha, WI, USA), and Acuson
Sequoia 512 (Siemens Medical Solutions, Mountain View,
CA, USA). The fusion imaging technique (Volume Naviga-
tion, GE Healthcare) was routinely used after July 2010.20
Commercially available internally cooled electrode sys-
tems with generators were used (VIVA RF System,
STARmed, Goyang, Korea; Cool-tip RF System, Covidien,
Mansfield, MA, USA). The type of electrode used and the
length of active tip were dependent on equipment avail-
ability and tumour size. The methodology used to carry out
RFA was identical to those performed in previous studies.6,14
 S.S. Kim et al. / Clinical Radiology xxx (2017) 1e8
The aim of the treatment was to achieve a hyperechoic area
with ablative margins of at least 0.5 cm beyond the tumour
boundary on US, with the exception of subcapsular14 and
perivascular tumours.21 If residual unablated tumour was
detected on CT performed immediately after the treatment,
additional RFA was performed. If the tumour was located at
a high-risk location,13 hydrodissection, using 5% dextrose in
a water solution was used to either prevent collateral
thermal injury or to enhance the sonic window during the
procedure.22
TACE
All TACE procedures were performed on an inpatient
basis by two radiologists (S.W.S. and S.K.C.), each with at
least 9 years of experience in vascular intervention. Angi-
ographies were performed, using a 5-F catheter (Yashiro,
Terumo, Tokyo, Japan or RH, Cook, Bloomington, IN, USA), to
identify hepatic artery anatomy, tumour staining, and the
tumour-feeding artery. By using a coaxial microcatheter
(Progreat, Terumo, Tokyo, Japan or Microferret, Cook,
Bloomington, IN, USA), TACE was performed as selectively
as possible via the lobar, segmental, or subsegmental ar-
teries, based on the patient’s hepatic function and tumour
distribution. After positioning the microcatheter into or as
close as possible to the tumour feeding artery, an emulsion
of doxorubicin hydrochloride (Adriamycin, Dong-A Pharm,
Seoul, Korea) and ethiodised oil (Lipiodol, Guerbet, Aulnay-
sous-Bois, France) was infused until presence of arterial
flow stasis and/or ethiodised oil appeared in the portal
branches. The dose of Adriamycin and Lipiodol was
dependent on the tumour size and vascularity, with a
maximum of 40 mg Adriamycin and 10 ml Lipiodol per
session. Then, embolisation of the feeding artery with 1- to
2-mm-diameter gelatin sponge pledgets (Cutanplast, Mas-
cia Brunelli, Milan, Italy) was performed in the presence of
arterioportal shunts or large tumours. At the end of TACE,
cone-beam CT was performed to evaluate the extent of
Lipiodol uptake.
Follow-up
For the assessment of therapeutic outcomes and poten-
tial complications, all patients were followed up after a
month had passed since their initial discharge, followed by
an assessment every 3 months during the first 2 years, and
every 4e6 months thereafter. During every visit, chest
radiography, multiphase CT, and laboratory tests, including
serum a-fetoprotein, were performed. If tumour recurrence
was detected during the follow-up period, appropriate
treatment such as RFA, TACE, radiation treatment, or liver
transplantation was determined by a multidisciplinary
tumour team meeting.
Study outcomes
The outcomes of the present study included recurrence-
free and overall survival. Recurrence-free survival was
defined as the time during the follow-up period at which
the patient did not experience any event, including second
Please cite this article in press as: Kim SS, et al., Initial radiofrequency ablation failure for hepatocellular carcinoma: repeated radiofrequency
ablation versus transarterial chemoembolisation, Clinical Radiology (2017), http://dx.doi.org/10.1016/j.crad.2017.07.020
3
LTP, intrahepatic distant recurrence (IDR), and extrahepatic
recurrence (ER). The second LTP was defined as newly
developed LTP after the treatment of the initial LTP. The
observation time for survival analysis was defined as the
interval between the repeated RFA or TACE treatments and
either the occurrence of an event (tumour recurrence or
death) or the last visit to the outpatient clinic before 31
December 2015. Major complications were defined, ac-
cording to the Society of Interventional Radiology guide-
lines, as clinical events that require additional treatment,
hospitalisation or result in permanent adverse sequelae or
death due to treatment-related complications.17
The following variables were used for the prognostic
factor analyses for each treatment outcome: treatment type
(repeated RFA versus TACE), age, sex, cause of liver disease,
ChildePugh class, presence of liver cirrhosis, use of antiviral
treatment during follow-up, size of the LTP, serum a-feto-
protein concentration prior to treatment, and perivascular
and subphrenic location of the LTP. The perivascular LTP was
defined as a recurrent tumour with any type of contact with
the first- or second degree branches of a portal or hepatic
vein that was 3 mm or greater in axial diameter.21 In
addition, when LTP was located within 1 cm from the dia-
phragm, it was defined as a subphrenic recurrent tumour.
All these analyses for tumour locations were assessed using
axial and coronal images from the picture archiving and
communication system (Centricity, GE healthcare, Chicago,
IL, USA). If a patient had multiple LTPs, the largest tumour
was used for the evaluation of imaging findings.
Statistical analysis
Comparison of baseline and clinical variables between the
two groups was performed using a Wilcoxon rank sum test
or a t-test for continuous variables, according to their
normality, and by Fischer’s exact test for categorical vari-
ables. Recurrence-free and overall survival rates between the
two groups were plotted using the KaplaneMeier method.
These outcomes were compared with using a log-rank test.
To identify the independent prognostic factor for
recurrence-free and overall survival, a Cox proportional
hazards regression model was performed for univariate and
multivariate analyses. For multivariate analysis, variables
with p-values <0.2 from the univariate analyses were
included in the final model. For these analyses, the estimated
hazard ratio (HR), 95% confidence interval (CI), and p-values
were provided. A p-value <0.05 was considered an indicator
of statistical significance. All analyses were performed using
SAS software version 9.4 (SAS Institute, Cary, NC, USA).
Results
Baseline characteristics
The overall incidence of patients with LTP after RFA for
HCC as a first-line treatment was 17.8% (126/713). The
median follow-up period of the patients was 51 months
(range, 2.2e107.2 months) in the repeated RFA group
(n¼33) and 40 months (range, 9.3e87.6 months) in the
4 S.S. Kim et al. / Clinical Radiology xxx (2017) 1e8

TACE group (n¼25), with no significant difference between
the two groups (p¼0.265; Figs 1 and 2). The baseline
characteristics of the two groups, before the initial RFA for
HCC as the first-line treatment, were not significantly
different (Table 1).
LTP after the initial RFA
of 25 (92%) patients in the TACE group. Recurrence-free
survival at 1-, 3-, and 5-years were estimated to be 66.5%
(95% CI¼47.7e79.9%), 20.4% (8.4e36.1%), and 17%
(6.3e32.1%) in the repeated RFA group, and 52% (95%
CI¼31.3e69.2%), 16% (5e32.5%), and 10.7% (2.2e27.1%) in
the TACE group, respectively. The differences between these
curves were not statistically significant (p¼0.191; Fig 3a).

The LTP characteristics for both groups are described in
Table 2. The sizes of the LTP were not significantly different
between the two groups (median, 1.4 cm; range, 1e2.7 cm
in the repeated RFA group versus median, 1.3 cm; range,
1e3.5 cm in TACE group; p¼0.215). The median time to LTP
diagnosis after the initial RFA treatment was 14.8 months
(range, 2.1e39.3 months) in the repeated RFA group and
11.2 months (range, 3.2e42 months) in the TACE group
(p¼0.239). All patients had LTPs at BCLC 0 or A stage. Be-
sides the subphrenic location of the LTP (p¼0.030), other
characteristics did not show significant differences between
the two groups (p>0.05).
Comparison of therapeutic outcomes
Recurrence-free survival
During follow-up, recurrent tumours were identified in
26 of 33 (78.8%) patients in the repeated RFA group and 23
Overall survival
Nine patients (27.3%) in the repeated RFA group and 11
patients (44%) in the TACE group had died. The overall
survival rates at 3- and 5-years were estimated to be 84.3%
(95% CI¼66.3e93.2%) and 72.7% (52.2e85.6%) in the
repeated RFA group, and 66.5% (95% CI¼43.9e81.7%) and
51.9% (29.7e70.2%) in the TACE group, respectively. There
were no significant statistical differences in overall survival
curves between the two groups (p¼0.179; Fig 3b).
Major complications
There were no treatment-related deaths in either group.
In addition, no patient had immediate major complications
associated with either treatment. Although a patient in the
repeated RFA group had peritoneal tumour seeding as a
delayed major complication, no statistically significant dif-
ference between the two groups could be observed (1/33,
3% versus 0/25, 0%; p¼1.000).

Figure 1 A 75-year-old-man treated with repeated RFA for LTP. (a) An axial MRI image obtained during the hepatobiliary phase shows a 1.5-cm
HCC (arrow) in segment VI before initial RFA. (b) An axial CT image obtained 1 month after RFA shows no residual tumour around the ablation
zone (A). (c) During follow-up, LTP (arrow) was detected around the upper margin of the previous ablation zone 16 months after RFA. Repeated
RFA was performed using a 3-cm active tip RFA electrode with artificial ascites assistance to decrease collateral thermal injury of the adjacent
colon. (d) An axial CT image obtained 1 month after repeated RFA shows no residual tumour around the ablation zone (A). Secondary LTP did not
occur until the last follow-up.

Please cite this article in press as: Kim SS, et al., Initial radiofrequency ablation failure for hepatocellular carcinoma: repeated radiofrequency
ablation versus transarterial chemoembolisation, Clinical Radiology (2017), http://dx.doi.org/10.1016/j.crad.2017.07.020
 S.S. Kim et al. / Clinical Radiology xxx (2017) 1e8 5

Figure 2 A 69-year-old-man treated with TACE for LTP. (a) An axial CT image obtained during the hepatic arterial phase shows a 1.8-cm HCC
(arrow) in segment VIII before RFA. (b) During follow-up, LTP (arrow) was detected around the lateral margin of the previous ablation zone 17
months after RFA. The index tumour was not visualised on planning US due to the subphrenic location of the tumour. (c) During TACE, angi-
ography shows tumour staining (arrow) in segment VIII of the liver. For this recurrent tumour, 30 mg Adriamycin and 6 ml Lipiodol were used.
(d) An axial CT image obtained 1 month after TACE shows dense Lipiodol uptake (L) within the LTP lesion without viable tumours. Secondary LTP
was not encountered until the last follow-up.

Prognostic factor analysis for therapeutic outcomes
Recurrence-free survival
The a-fetoprotein level prior to re-treatment was the
only significant independent factor associated with
recurrence-free survival after LTP treatment in both the
univariate and multivariate analysis (HR¼1.74; 95% CI¼1.17,
2.57; p¼0.006 and HR¼1.81; 95% CI¼1.17, 2.81; p¼0.008,
respectively; Table 3).
Overall survival
On univariate analysis, the a-fetoprotein level prior to re-
treatment (HR¼2.31; 95% CI¼1.27, 4.21; p¼0.006) and the
subphrenic location of the LTP (HR¼2.55; 95% CI¼1.01, 6.47;
p¼0.048) were significant prognostic factors for overall
survival. After adjustment of the other variables, the pres-
ence of hepatitis C virus (HR¼3.50; 95% CI¼1.13, 10.85;
p¼0.027) was found to be a significant independent prog-
nostic factor for overall survival (Table 3).
Discussion
The treatment approach for HCC recurrence after cura-
tive treatments varies from centre to centre. It has been
proposed that an aggressive treatment approach for recur-
rence after surgical resection for HCC confers good
treatment outcomes23,24; however, there is no standard
guideline available on current practices for approaching LTP
after RFA. In the present study, it was found that in patients
initially managed with RFA, treating LTP with TACE when
repeated RFA was not feasible and showed no significant
differences with repeated RFA in terms of long-term ther-
apeutic outcomes.
The present results showed that patients in the TACE
group tended to have subphrenically located LTPs compared
to those in the repeated RFA group. This discrepancy in
tumour location for the two groups may be due to the
following reasons: although artificial ascites or pleural
effusion was used to decrease collateral thermal damage to
the diaphragm or to enhance the sonic window when
repeated RFA for LTPs were performed in subphrenic loca-
tions, the subphrenic location constitutes one of the blind
spots at US, as that section of the liver is surrounded by the
lung and is considered to be one of the deepest areas in
US.25 In line with the present results, a previous study also
reported that the proximity of the tumour to the diaphragm
was a significant factor affecting the technical feasibility of
RFA for HCC treatment.26
According to the current BCLC system for HCC manage-
ment, TACE is the standard of care for multiple nodular
HCCs at an intermediate stage (BCLC stage B). In contrast to
RFA, it has been considered to be a non-curative treatment
for HCC due to the lower rate of initial complete response

Please cite this article in press as: Kim SS, et al., Initial radiofrequency ablation failure for hepatocellular carcinoma: repeated radiofrequency
ablation versus transarterial chemoembolisation, Clinical Radiology (2017), http://dx.doi.org/10.1016/j.crad.2017.07.020
6 S.S. Kim et al. / Clinical Radiology xxx (2017) 1e8

Table 1 difference in the overall survival rates between RFA and

Baseline characteristics of patients in both groups before initial radio- TACE in patients with small HCCs (<2 cm, BCLC stage 0).28
frequency ablation (RFA) for hepatocellular carcinoma (HCC).
Furthermore, Yang et al. demonstrated that TACE was an
Characteristics Repeated RFA TACE p-value effective treatment that was capable of achieving long-term
group (n ¼33) group survival rates, comparable to those seen with surgical
(n ¼25)
resection or RFA in patients with small single nodular HCCs
Age (year) 61.3  10.1 59.4  10.9 0.479
(< 3 cm, BCLC stage 0 or A).29
Male 21 (63.6) 16 (64) 0.977
Aetiology of chronic liver disease 0.931 In line with a previous study,24 which reported that RFA
HBV-related 20 (60.6) 16 (64) and TACE may be equally effective for intrahepatic recur-
HCV-related 8 (24.2) 5 (20) rence after hepatectomy, the present study found no sta-
Other 5 (15.2) 4 (16) tistically significant differences in long-term therapeutic
ChildePugh classification 0.082
outcomes between repeated RFA and TACE for patients with
A 30 (90.9) 18 (72)
B 3 (9.1) 7 (28) LTPs after initial RFA treatment. Most of the LTPs in the
Liver cirrhosis 0.742 present study were single and the median tumour size was
Present 26 (78.8) 21 (84) <1.5 cm. This is because all of the LTPs were detected early
Absent 7 (21.2) 4 (16)
due to periodic surveillance after initial RFA treatment for
Antiviral treatment during follow-up 0.912
Present 18 (54.5) 14 (56)
HCC. These characteristics of recurrent tumours, due to
Absent 15 (45.5) 11 (44) early detection, may have enabled the present TACE cohort
Size of initial HCC (cm) 2.1  0.7 2.3  0.6 0.269 to achieve a very high complete response rate for LTPs
BCLC staging for 0.778 (92.6%, 25/27) at the 1-month follow-up CT examination
initial HCC
and may have also favourably affected the therapeutic
0 11 (33.3) 7 (28)
A 22 (66.7) 18 (72) outcomes in the present TACE group. Similarly, a previous
a-FP prior to 16.8 (2.1e1834.1) 14.9 (3.5e618.1) 0.677 study29 also showed that complete response rates after the
initial RFA (ng/ml) initial treatment for HCC independently affected overall
Continuous variables were described by means  standard deviations or survival.
median (range) according to the normality. Categorical variables were In the present study, the serum a-fetoprotein level prior
described by number of patients (percentage). to LTP treatment was independently associated with
TACE, transarterial chemoembolisation; HBV, hepatitis B virus; HCV, hepa-
recurrence-free survival. In general, high serum a-fetopro-
titis C virus; BCLC, Barcelona Clinic Liver Cancer; aeFP, a efetoprotein.
tein level is usually observed in HCC with poorly differen-
tiated histological grade,30 which has been known to be a
and the higher risk of recurrence after treatment2; however, predictive factor for HCC recurrence and poor prognosis for
in patients with very early or early HCCs (BCLC stage 0 or A) both RFA and TACE treatment.31,32 Based on the present
at subphrenic locations or in tumours adjacent to the main data, overall survival was worse in patients with hepatitis C
bile duct, major vessels, gallbladder, or bowel loop, TACE virus than in patients with hepatitis B virus. A previous
may be an alternative treatment option as it does not cause study33 reported that hepatitis C virus-related tumours
collateral thermal damage to adjacent structures.27 have higher recurrence rates after curative resection than
Recently, Kim et al. reported that there was no significant hepatitis B virus-related HCCs. Additionally, it is known that
cirrhosis almost always accompanies hepatitis C virus-
related HCCs.34 These differences may have affected the
Table 2 overall survival rates in the present patients with hepatitis C
Characteristics of local tumour progression (LTP) after initial radiofrequency
ablation (RFA) in both groups.
virus.
There are several limitations to the present study. First,
Characteristics Repeated RFA TACE p-value the study was retrospectively designed. Thus, it was
group (n ¼33) group (n ¼25)
inherently flawed by selection and indication bias; how-
Median time to LTP (month) 14.8 (2.1e39.3) 11.2 (3.2e42) 0.239
ever, head-to-head prospective comparison study between
Median size of LTP (cm) 1.4 (1e2.7) 1.3 (1e3.5) 0.215
Median number of LTP 1 (1e2) 1 (1e3) 0.402 RFA and TACE was problematic ethically due to lack of ev-
BCLC staging for LTP 0.390 idence. Second, all LTPs were diagnosed using the current
0 25 (75.8) 16 (64) guidelines based on imaging findings.16 Therefore, the
A 8 (24.2) 9 (36) possibility of false-positive cases cannot be completely
Perivascular location 0.440
excluded. Third, CT guidance was not used for RFA. Although
Yes 11 (33.3) 6 (24)
No 22 (66.7) 19 (76) CT guidance can access tumours in subphrenic locations, CT
Subphrenic location 0.030 has radiation issues and cannot be used repeatedly. Finally,
Yes 4 (12.1) 9 (36) although a large retrospective cohort composed of patients
No 29 (87.9) 16 (64)
who had been treated with RFA for HCC was studied, a
a-FP prior to treatment 7 (1.3e475) 10.2 (2e2233) 0.206
for LTP (ng/ml)
relatively small number of patients were ultimately
included in the study. Thus, the present study may warrant
Continuous variables were described by median (range). Categorical vari-
ables were described by number of patients (percentage).
a larger multicentre study to reassess the role of locore-
TACE, transarterial chemoembolisation; HBV, hepatitis B virus; HCV, hepa- gional treatments in patients with LTPs after RFA treatment.
titis C virus; aeFP, a -fetoprotein. Despite these drawbacks, the results of the present study

Please cite this article in press as: Kim SS, et al., Initial radiofrequency ablation failure for hepatocellular carcinoma: repeated radiofrequency
ablation versus transarterial chemoembolisation, Clinical Radiology (2017), http://dx.doi.org/10.1016/j.crad.2017.07.020
 S.S. Kim et al. / Clinical Radiology xxx (2017) 1e8 7

Figure 3 Recurrence-free and overall survival rate curves for both groups. (a) Recurrence-free survival according to the treatment group. (b)
Overall survival according to the treatment group.

Table 3
Analyses of prognostic factors for effects on recurrence-free and overall survival.

Variables Recurrence-free survival Overall survival

Univariate analysis Multivariate analysis Univariate Analysis Multivariate analysis

HR (95% CI) p-Value HR (95% CI) p-Value HR (95% CI) p-Value HR (95% CI) p-Value
TACE (repeated RFA) 1.45 (0.83, 2.55) 0.194 1.50 (0.83, 2.70) 0.182 1.82 (0.75, 4.39) 0.185 1.72 (0.67, 4.45) 0.263
Age 1.00 (0.97, 1.02) 0.777 - - 0.99 (0.95, 1.04) 0.802 - -
Sex (male) 1.10 (0.62, 1.95) 0.756 - - 0.97 (0.40, 2.38) 0.948 - -
Aetiology of chronic liver disease 0.083 0.946 0.097 0.040
HBV 1 [reference] 1 [reference] 1 [reference] 1 [reference]
HCV 2.09 (0.97, 4.51) 0.062 1.13 (0.39, 3.29) 1.000 2.66 (0.92, 7.20) 0.079 3.50 (1.13, 10.85) 0.027
Other 1.61 (0.64, 4.08) 0.498 0.99 (0.28, 3.47) 1.000 0.95 (0.17, 5.38) 1.000 1.11 (0.19, 6.51) 1.000
ChildePugh B class (A) 0.93 (0.45, 1.93) 0.846 - - 1.31 (0.38, 4.48) 0.664 - -
Liver cirrhosis (absent) 1.36 (0.64, 2.92) 0.424 - - 5.54 (0.74, 41.44) 0.095 4.52 (0.55, 37.31) 0.161
Antiviral treatment (absent) 0.61 (0.34, 1.07) 0.085 0.49 (0.20, 1.20) 0.117- 0.59 (0.24, 1.41) 0.234 - -
Size of LTP 1.16 (0.74, 1.81) 0.530 - - 1.01 (0.46, 2.22) 0.981 - -
a-FP (mg/ml)a 1.74 (1.17, 2.57) 0.006 1.81 (1.17, 2.81) 0.008 2.31 (1.27, 4.21) 0.006 2.02 (0.99, 4.12) 0.053
Perivascular (absent) 1.43. (0.78, 2.65) 0.248 - - 1.42 (0.56, 3.55) 0.458 - -
Subphrenic (absent) 1.76 (0.91, 3.41) 0.094 1.80 (0.89, 3.64) 0.103 2.55 (1.01, 6.47) 0.048 2.42 (0.90, 6.49) 0.079

The reference category for each variable is in the square brackets in the first column.
TACE, transarterial chemoembolisation; HBV, hepatitis B virus; HCV, hepatitis C virus; a-FP, a-fetoprotein prior to treatment for LTP.
a
Log transformation was used for model fitting. Cox proportional hazards model was used for statistical analysis.

Please cite this article in press as: Kim SS, et al., Initial radiofrequency ablation failure for hepatocellular carcinoma: repeated radiofrequency
ablation versus transarterial chemoembolisation, Clinical Radiology (2017), http://dx.doi.org/10.1016/j.crad.2017.07.020
8 S.S. Kim et al. / Clinical Radiology xxx (2017) 1e8
are useful given the current lack of reliable data on how to
manage LTPs after RFA treatment and the long-term ther-
apeutic effectiveness of repeated RFA and TACE in such
circumstances. Based on the present results, rather than
performing repeated RFA for LTPs in high-risk locations,
which has the possibility of thermal injury and has to be
done under a poor sonic window, TACE could be considered
as an alternative treatment option.
In conclusion, the results of the present study show that
in patients with LTPs that have arisen after RFA treatment,
TACE could be an effective and safe treatment when
repeated RFA is not feasible. TACE is capable of achieving
recurrence-free survival, overall survival, and major
complication rates comparable to those seen in repeated
RFA.
Appendix A. Supplementary data
Supplementary data related to this article can be found at
http://dx.doi.org/10.1016/j.crad.2017.07.020.
References
1. EASL-EORTC clinical practice guidelines: management of hepatocellular
carcinoma. J Hepatol 2012;56:908e43.
2. Forner A, Llovet JM, Bruix J. Hepatocellular carcinoma. Lancet
2012;379:1245e55.
3. Lee JM, Park JW, Choi BI. 2014 KLCSG-NCC Korea practice guidelines for
the management of hepatocellular carcinoma: HCC diagnostic algo-
rithm. Dig Dis 2014;32:764e77.
4. Kudo M, Izumi N, Kokudo N, et al. Management of hepatocellular car-
cinoma in Japan: consensus-based clinical practice guidelines proposed
by the Japan Society of Hepatology (JSH) 2010 updated version. Dig Dis
2011;29:339e64.
5. Kang TW, Rhim H. Recent advances in tumor ablation for hepatocellular
carcinoma. Liver Cancer 2015;4:176e87.
6. 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;58:89e97.
7. Lee DH, Lee JM, Lee JY, et al. Radiofrequency ablation of hepatocellular
carcinoma as first-line treatment: long-term results and prognostic
factors in 162 patients with cirrhosis. Radiology 2014;270:900e9.
8. Shiina S, Tateishi R, Arano T, et al. Radiofrequency ablation for hepato-
cellular carcinoma: 10-year outcome and prognostic factors. Am J Gas-
troenterol 2012;107:569e77. quiz 78.
9. Kang TW, Kim JM, Rhim H, et al. Small hepatocellular carcinoma: radi-
ofrequency ablation versus nonanatomic resectiondpropensity score
analyses of long-term outcomes. Radiology 2015;275:908e19.
10. Hasegawa K, Kokudo N, Makuuchi M, et al. Comparison of resection and
ablation for hepatocellular carcinoma: a cohort study based on a Japa-
nese nationwide survey. J Hepatol 2013;58:724e9.
11. Kim YS, Lee WJ, Rhim H, et al. The minimal ablative margin of radio-
frequency 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;195:758e65.
12. Massarweh NN, Park JO, Farjah F, et al. Trends in the utilization and
impact of radiofrequency ablation for hepatocellular carcinoma. J Am
Coll Surg 2010;210:441e8.
13. Teratani T, Yoshida H, Shiina S, et al. Radiofrequency ablation for he-
patocellular carcinoma in so-called high-risk locations. Hepatology
2006;43:1101e8.
14. Kang TW, Lim HK, Lee MW, et al. Long-term therapeutic outcomes of
radiofrequency ablation for subcapsular versus nonsubcapsular
Please cite this article in press as: Kim SS, et al., Initial radiofrequency ablation failure for hepatocellular carcinoma: repeated radiofrequency
ablation versus transarterial chemoembolisation, Clinical Radiology (2017), http://dx.doi.org/10.1016/j.crad.2017.07.020
hepatocellular carcinoma: a propensity score matched study. Radiology
2016:151243.
15. Xie X, Jiang C, Peng Z, et al. Local recurrence after radiofrequency
ablation of hepatocellular carcinoma: treatment choice and outcome. J
Gastrointest Surg 2015;19:1466e75.
16. Ahmed M, Solbiati L, Brace CL, et al. Image-guided tumor ablation:
standardization of terminology and reporting criteriada 10-year up-
date. Radiology 2014;273:241e60.
17. 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:S293e5.
18. Bruix J, Sherman M. American Association for the Study of Liver D.
Management of hepatocellular carcinoma: an update. Hepatology
2011;53:1020e2.
19. Rhim H, Choi D, Kim YS, et al. Ultrasonography-guided percutaneous
radiofrequency ablation of hepatocellular carcinomas: a feasibility
scoring system for planning sonography. Eur J Radiol 2010;75:253e8.
20. Lee MW, Rhim H, Cha DI, et al. Percutaneous radiofrequency ablation of
hepatocellular carcinoma: fusion imaging guidance for management of
lesions with poor conspicuity at conventional sonography. AJR Am J
Roentgenol 2012;198:1438e44.
21. Kang TW, Lim HK, Lee MW, et al. Perivascular versus nonperivascular
small HCC treated with percutaneous RF ablation: retrospective com-
parison of long-term therapeutic outcomes. Radiology 2014;270:888e99.
22. Rhim H, Lim HK. Radiofrequency ablation for hepatocellular carcinoma
abutting the diaphragm: the value of artificial ascites. Abdom Imaging
2009;34:371e80.
23. Tranchart H, Chirica M, Sepulveda A, et al. Long-term outcomes
following aggressive management of recurrent hepatocellular carci-
noma after upfront liver resection. World J Surg 2012;36:2684e91.
24. Koh PS, Chan AC, Cheung TT, et al. Efficacy of radiofrequency ablation
compared with transarterial chemoembolization for the treatment of
recurrent hepatocellular carcinoma: a comparative survival analysis.
HPB (Oxford) 2016;18:72e8.
25. Lee MW, Kim YJ, Park HS, et al. Targeted sonography for small hepato-
cellular carcinoma discovered by CT or MRI: factors affecting sono-
graphic detection. AJR Am J Roentgenol 2010;194:W396e400.
26. Kim PN, Choi D, Rhim H, et al. Planning ultrasound for percutaneous
radiofrequency ablation to treat small (</¼3 cm) hepatocellular carci-
nomas detected on computed tomography or magnetic resonance im-
aging: a multicenter prospective study to assess factors affecting
ultrasound visibility. J Vasc Interv Radiol 2012;23:627e34.
27. Han K, Kim JH. Transarterial chemoembolization in hepatocellular car-
cinoma treatment: Barcelona Clinic Liver Cancer Staging System. World J
Gastroenterol 2015;21:10327e35.
28. Kim JW, Kim JH, Sung KB, et al. Transarterial chemoembolization versus
radiofrequency ablation for the treatment of single hepatocellular car-
cinoma 2 cm or smaller. Am J Gastroenterol 2014;109:1234e40.
29. Yang HJ, Lee JH, Lee DH, et al. Small single-nodule hepatocellular car-
cinoma: comparison of transarterial chemoembolization, radio-
frequency ablation, and hepatic resection by using inverse probability
weighting. Radiology 2014;271:909e18.
30. Sasaki K, Matsuda M, Ohkura Y, et al. The influence of histological dif-
ferentiation grade on the outcome of liver resection for hepatocellular
carcinomas 2 cm or smaller in size. World J Surg 2015;39:1134e41.
31. N’Kontchou G, Mahamoudi A, Aout M, et al. Radiofrequency ablation of
hepatocellular carcinoma: long-term results and prognostic factors in
235 western patients with cirrhosis. Hepatology 2009;50:1475e83.
32. Wang Y, Chen Y, Ge N, et al. Prognostic significance of alpha-fetoprotein
status in the outcome of hepatocellular carcinoma after treatment of
transarterial chemoembolization. Ann Surg Oncol 2012;19:3540e6.
33. Huang YH, Wu JC, Chen CH, et al. Comparison of recurrence after hepatic
resection in patients with hepatitis B versus hepatitis C-related small
hepatocellular carcinoma in hepatitis B virus endemic area. Liver Int
2005;25:236e41.
34. Yeh MM, Daniel HD, Torbenson M. Hepatitis C-associated hepatocellular
carcinomas in non-cirrhotic livers. Mod Pathol 2010;23:276e83.