Key words: stereotactic body radia-
tion therapy, huge hepatocellular car-
Correspondence to: Mi-Sook Kim,
MD, Department of Radiation Oncolo-
gy, Korea Cancer Center Hospital, Ko-
rea Institute of Radiological and Med-
ical Sciences, 215-4 Gongneung-
dong, Nowon-gu, Seoul 139-706, Re-
public of Korea.
Received March 31, 2009;
accepted June 30, 2009.
Pilot study of stereotactic body radiotherapy
for huge hepatocellular carcinoma unsuitable
for other therapies
Young-Joo Shin1, Mi-Sook Kim1, Seong Yul Yoo1, Chul Koo Cho1,
Young Seok Seo1, Jin-kyu Kang1, Su Cheol Park2, Chul Ju Han2,
Sang Beom Kim3, Byong Hee Lee4, and Dong Han Lee5
1Department of Radiation Oncology,2Department of Internal Medicine,3Department of Surgery,
4Department of Radiology, and5Cyberknife Center, Korea Cancer Center Hospital, Korea Institute
of Radiological and Medical Sciences, Seoul, Korea
Aims. To determine the feasibility and efficacy of stereotactic body radiotherapy
(SBRT) for huge hepatocellular carcinoma unsuitable for other therapies.
Methods. Six patients with very large hepatocellular carcinomas (>10 cm) unsuitable
for surgical resection or that failed to respond to transcatheter arterial chemoem-
bolization (TACE) were treated by SBRT. Doses ranged from 32 Gy to 40 Gy in four
fractions. Survival, response, and toxicities were evaluated.
Results. After a median follow-up of 25.9 months (range 8.1-56 months), three pa-
tients had died and three were alive. Overall, treatment was well tolerated and no
dose-limiting toxicity or radiation-induced liver disease was observed. The median
survival was 10 months (range 3-56 months) and the median progression-free dura-
tion was 6 months (range, 2-21 months). Partial response was achieved by four pa-
tients, stable disease by one, and one patient had disease progression. One patient
with a partial response who underwent lobectomy after SBRT was alive 56 months
Conclusion. This study suggests that SBRT can be delivered safely at 32-40 Gy in four
other modalities such as surgery orTACE might prolong survival. Free full text avail-
able at www.tumorionline.it
Hepatocellular carcinoma (HCC) is the third most common cause of cancer death1.
In patients with a small HCC, hepatic resection and nonsurgical treatment modalities
have contributed to good survival. However, the treatment of patients with a HCC of
10 cm or larger in diameter (so-called huge HCC) is a challenge. Hepatic resection ap-
pears to be the most reasonable treatment for huge HCC, but the resection of such
large tumors is technically difficult and usually requires major hepatic resection,
which is associated with a higher operative mortality risk2,3. In patients with inopera-
ble huge HCC, transcatheter arterial chemoembolization (TACE) is considered a sec-
ond choice, but response rates are generally poor for large tumors and the 5-year sur-
vival rate is less than 10%4. Furthermore, after failure of TACE no standard treatment
is available, and various nonsurgical treatments such as hormonal therapy, im-
munotherapy, systemic and intra-arterial chemotherapy have been assessed in pilot
trials, but to date survival benefits have been marginal.
Several studies have demonstrated the efficacy of stereotactic body radiotherapy
(SBRT) for the treatment of liver tumors. In the majority of these studies, SBRT has been
shown to achieve a high rate of local control with low toxicity5-7. However, previous re-
ports have predominantly addressed HCCs of less than 7
cm and no information is available regarding the use of
the use of internal markers (fiducials), SBRT allows more
accurate application by reducing the error margin. Nor-
ing motion. Furthermore, fractionated SBRT may have 3
operable huge HCC (>1000 mL) and no other treatment
options, good liver function reserve, and good perform-
ease progression or enabling surgery or another local
modality. We focused on its potential to treat huge HCC.
Furthermore, although the number of cases involved was
small, we attempted to determine survival and response,
and to assess further treatment possibilities after SBRT.
Materials and methods
Patients and tumor characteristics
Patients were diagnosed based on histological or radi-
ological findings described by the Barcelona criteria8.
The inclusion criteria were as follows: 1) a single HCC
with/without a daughter nodule; 2) a closely attached
mass and daughter nodule treatable in one session; 3) a
longest diameter >12 cm or a target volume >1000 mL;
4) a clear tumor margin on computed tomography (CT);
5) an inoperable tumor; 6) failure to respond toTACE; 7)
an ECOG performance status of 0 or 1; 8) Child-Pugh
class A liver function. The exclusion criteria were as fol-
lows: 1) presence of extrahepatic metastasis; 2) tumor
attached to the esophagus, stomach, or colon as visual-
ized by CT; 3) Child-Pugh Class B or C; 4) presence of
uncontrolled ascites; 5) tumor occupying the whole liv-
er; 6) regional lymph node metastases detected by com-
puted tomography (CT) or positron emission tomogra-
phy/computed tomography (PET/CT).
The study protocol was approved by the gastrointesti-
nal multidisciplinary tumor board at the Korea Institute
of Radiological and Medical Sciences. Patients and/or
guardians provided written consent after receiving an
explanation about the benefits and possible complica-
tions of SBRT. Six patients treated between March 2003
and March 2008 were enrolled in this study. Mandatory
baseline examinations were intravenous contrast-en-
hanced CT of the abdomen and pelvis and laboratory
tests including complete blood count, liver function
test, and alpha-fetoprotein (AFP). If tumors diminished
in size after SBRT, salvage treatments such as surgery or
TACE were considered.
Gold fiducials (4 mm long and 0.8 mm in diameter)
were used as landmarks. SBRT was performed using a
CyberKnife (Accuray Inc., Sunnyvale, CA, USA). Six fidu-
cials were placed percutaneously using an 18-gauge
spinal needle under fluoroscopy at the transverse
processes of the vertebral bodies nearest the lesions.
Custom-made devices consisting of an Alpha Cradle
(Smithers Medical Products, North Canton, OH, USA)
and 4 belts were prepared to restrict respiratory motion
5-7 days after fiducial placement. The immobilization
approach consists of abdominal compression that lim-
its diaphragmatic excursion and thereby the respiratory
motion of the tumor. A previous study conducted at our
institution showed that this device reduces diaphragm
motion by 32-80% as compared with a previously re-
ported range9. CT images were obtained using a 2-mm
slice thickness. To better delineate tumor volumes,
magnetic resonance imaging (MRI) or PET/CT were
used to provide reference images. The images included
target movements due to respiration, so the tumor vol-
umes utilized for planning were larger than the gross tu-
mor volumes (GTV). Tumor volumes as delineated by
the CyberKnife planning system refer to internal target
volumes (ITV); planning target volumes (PTV) were
equal to ITV plus setup margins. These margins ranged
from 0 to 2 mm, because our previous study indicated
that this adequately compensated for respiratory move-
ment9.However, in some cases the margins were modi-
fied due to normal tissue constraints of spinal cord,
stomach, and intestine. Radiation doses were pre-
scribed to the isodose line (66-72% of maximum dose)
that covered at least 94% of the PTV (Figure 1). Treat-
ments involved hundreds of pencil beams shaped using
a single 20-mm-, 25-mm-, 30-mm-, 35-mm-, or 40-mm-
diameter circular collimator.
The median tumor volume in our 6 patients was
1287.5 mL (range, 1008.2-1574.4 mL). Table 1 summa-
rizes the SBRT dosage details.
Individualized prescribed doses, doses per fraction,
and numbers of fractions were determined based on tu-
mor size, location, and organs at risk. SBRT doses were
escalated from 32 Gy to 40 Gy; they started at 8 Gy per
fraction and were increased by 1 Gy per fraction. The
SBRT doses were converted to the normalized total dose
at a fraction size of 2 Gy (NTD2Gy) using a linear quad-
ratic equation (BED = total dose x (1 + dose per frac-
tion/α/β), α/β = 10 for early responding tissue, α/β = 3 for
late responding tissue). NTD2Gy (α/β = 10) of SBRT
ranged from 57.6 Gy to 80 Gy in our study.
The applied constraint was that the maximum dose
delivered to any point within the esophagus could not
exceed 30 Gy for all 4 fractions. Uninvolved normal liver
(V<17Gy) tissues administered less than 17 Gy, and maxi-
mum doses to any point in the stomach, intestine, and
right kidney were calculated (Table 1).
66 Y-J SHIN, M-S KIM, S-YYOO ET AL
Follow-up,response,and toxicity assessments
Tumor response was assessed as described in the Re-
sponse Evaluation and Criteria in Solid Tumors (RE-
Local failure was defined as an increase in tumor size or
the development of a new lesion in the radiation field.
Regional failure was defined as the development of a
new lesion in non-targeted liver. Distant metastasis was
defined as recurrence beyond the liver.
Acute and late toxicities were defined based on labora-
described in the Common Terminology Criteria for Ad-
verse Events v3.0 (CTCAE)11. Radiation-induced liver dis-
ease (RILD) was defined as the presence of nonmalignant
ascites and an elevation of alkaline phosphatase levels to
progression7. Dose-limiting toxicity was defined as any
any grade 4 toxicity attributable to SBRT.
Survival was measured from the date of SBRT com-
mencement. Median survival times and progression-
free survival times were recorded.
STEREOTACTIC BODY RADIOTHERAPY FOR HUGE HEPATOCELLULAR CARCINOMA 67
Figure 1 - Planning of stereotactic body radiotherapy (SBRT) in huge HCC. Internal target volume (ITV) was defined as the visualized liver mass.
Radiation doses (32-40 Gy in 4 fractions) were prescribed to the 66-72% isodose line of the maximum dose in order to cover the ITV.
Six patients with huge HCCs consented to treatment
from March 2003 to March 2008. Patient ages ranged
from 44 to 55 years (median 48.5 years). Underlying
liver diseases included hepatitis B infection in 3, cir-
rhotic liver in 2, and portal vein thrombosis in 2 pa-
All 6 patients had undergone TACE (number of ses-
sions: 1-9) but this proved ineffective. After completing
SBRT, 1 patient who achieved an operable status under-
went lobectomy, 4 patients underwent TACE, and 1 pa-
tient was not administered further treatment because of
disease progression and poor general condition.
Patients were assessed approximately 3 months post-
SBRT. One patient experienced progression from Child-
Pugh Classification A to B over 3 months post-SBRT. He
had progressive disease at 2 months and died at 3
months. However, another patient who achieved a par-
tial response and who underwent lobectomy after SBRT
was alive at 56 months.
Three patients were alive at the time of analysis (me-
patients died within 18 months of SBRT (Table 2).
Overall, treatment was well tolerated and no dose-
limiting toxicity or RILD was observed. However, in 1
patient alkaline phosphatase (ALP) and liver enzymes
increased to more than 2-fold the pre-SBRT values due
to disease progression. All patients had an elevated
baseline level of AFP (median 6312.5 IU/L; range 22.8-
627,200 IU/L), but AFP levels diminished in 5 patients
over the first 3 months post-SBRT (median 27.5 IU/L;
range 15.4-4129 IU/L).
During treatment, 5 patients had mild nausea (G1),
but no case of grade 2-4 nausea occurred. In addition, 4
patients developed transient asymptomatic right-sided
pleural effusions at 3 months post-SBRT (Table 3).
Response and survival
Partial response was defined as a >30% decrease in
the sum of the longest diameters of target lesions versus
baseline, as detailed in RECIST. Partial response was ob-
tained in 4 patients, stable disease in 1, while 1 patient
had progressive disease. Three patients developed local
failure and 4 patients regional failure. Four of the 6 pa-
tients developed lung metastasis. First progression usu-
ally occurred beyond the treated volumes. The median
survival was 10 months (range 3-56 months) and the
median progression-free survival 6 months (range 2-21
Hepatic resection is the only treatment that offers any
However, tumor sizes exceeding 10 cm indicate that the
disease is advanced and the possibility of tumor spread,
including satellite nodules and microvascular inva-
sion12,13. Large tumors also have a higher risk of recur-
rence, even after radical resection, so the benefits of he-
patic resection are marginal14. Lee et al.15reported that
patients with a target tumor volume of over 1000 mL
treated surgically have a 1-year recurrence rate of 66%.
Stereotactically guided radiation therapy enables lo-
cally curative treatment approaches to be adopted in
HCC. However, only a few centers have published SBRT
results, and little information is available concerning
the use of hypofractionated radiotherapy in HCC. Her-
farth et al.16applied single-fraction SBRT to primary
and metastatic liver lesions, and safely escalated the
dose applied from 14 to 26 Gy.Wulf et al.17applied frac-
tionated SBRT (30 Gy in 3 fractions) to 23 patients with
solitary liver lesions and observed no grade 3 or high
acute or late toxicity related to treatment. Tse et al.6ap-
plied an individualized dose allocation approach to
HCC. In this study, radiation doses ranged from 24 Gy to
54 Gy in 6 fractions depending on the volume of affect-
ed liver (median tumor volume 173 mL, range, 9-1913
68 Y-J SHIN, M-S KIM, S-YYOO ET AL
Table 1 - Volumes and doses delivered (in four fractions)
Maximum dose (Gy)
(Gy) (mL)StomachIntestine Right kidney
NTD2Gy, normalized total dose at a fraction size of 2 Gy; Vol<17 Gy, uninvolved liver volume receiving less than 17 Gy.
The maximum doses to any point within the spinal cord must not exceed 18 Gy for all 4 fractions and the maximum dose delivered to any
point within the esophagus must not exceed 30 Gy for all 4 fractions.
mL). No radiation-induced liver disease or treatment-
related grade 4/5 toxicity occurred 3 months post-SBRT.
However, 8 of 31 patients developed grade 3 liver en-
zymes and 3 patients developed transient asympto-
matic right-sided pleural effusions within 3 months.
They concluded that individualized 6-fraction SBRT is
tions achieved local control in the majority of patients
with a median survival of 11.7 months.
Our treatment design involves the use of multiple
fractions rather than a single-fraction regimen. Because
of the huge tumor mass in our patients, SBRT doses
ranged from 32 Gy to 40 Gy in 4 fractions (NTD2Gy 57.6-
80 Gy). Four of our 6 patients obtained partial respons-
es, and 1 patient with a partial response who underwent
lobectomy after SBRT was alive 56 months post-SBRT.
Our experience suggests that SBRT is not a curative
treatment for huge HCC, but that SBRT can lead to an
operable disease state.
A simplified form of the “critical volume” model, ini-
tially proposed by Yaes and Kalend18, was applied to
construct the normal liver dose restriction in the Uni-
versity of Colorado trial19. They recommended that at
least 700 mL of normal liver had to receive at total dose
of less than 15 Gy in 3 fractions, the maximum dose to
any point within the spinal cord must not exceed 18 Gy,
and the maximum point dose administered to stomach
or small intestine must not exceed 30 Gy.
The median volume of uninvolved liver (normal liver
minus GTV) was 1313 mL (range 892-1591 mL). As we
used 4 fractions, the median volume of uninvolved liver
that received a cumulative dose of <17 Gy in 4 fractions
was 781.5 mL (range 567-1223 mL). Although in 2 pa-
tients the uninvolved normal liver volume of <17Gywas
less than 700 mL, treatment was well tolerated with no
dose-limiting toxicity. Furthermore, although the maxi-
mum dose to any point in the stomach and intestine
was >30 Gy in some patients, no grade 3-4 toxicity was
In summary, this study suggests that SBRT can be de-
livered safely at 32-40 Gy in 4 fractions to huge inopera-
ble HCC. However, our study is limited by the small
number of patients and the short follow-up. A long-
1. Parkin DM, Bray F, Ferlay J, Pisani P: Global cancer statis-
tics, 2002. CA Cancer J Clin, 55: 74-108, 2005.
2. Nagasue N, Kohno H, Chang YC, Taniura H, Yamanoi A,
Uchida M, Kimoto T, Takemoto Y, Nakamura T, Yukaya H:
Liver resection for hepatocellular carcinoma. Results of
229 consecutive patients during 11 years. Ann Surg, 217:
3. Abdel-Wahab M, Sultan A, el-Ghawalby A, Fathy O, el-
Ebidy G, Abo-Zeid M, Aboel-Enin A, Abdallah T, Fouad A,
STEREOTACTIC BODY RADIOTHERAPY FOR HUGE HEPATOCELLULAR CARCINOMA 69
Table 2 - Treatment and results
15.8 18152 PRTACE (1)
+ + (3 mo)Lung12 AWD
12.510082 PR+--9 AWD
+ (21 mo)
+ (4 mo)
+ (2 mo)
6 1415741 PR- + (4 mo)Lung8D
TACE, transcatheter arterial chemoembolization; SBRT, stereotactic body radiotherapy; AWD, alive with disease;
NED, no evidence of disease; D, dead; PR, partial response; SD, stable disease; PD, progressive disease; mo, months.
Tabe 3 - Toxicity three months after stereotactic body radiotherapy
Liver enzymes, gradeALP, grade Bilirubin, grade Nausea, gradePleural effusion, grade
CTCAE v3.0, Common Terminology Criteria for Adverse Events version 3.0; ALP, alkaline phosphatase.