- Email: [email protected]
Three-Dimensional Conformal Radiation Therapy and Intensity-Modulated Radiation Therapy Combined With Transcatheter Arterial Chemoembolization for Locally Advanced Hepatocellular Carcinoma: An Irradiation Dose Escalation Study
Int. J. Radiation Oncology Biol. Phys., Vol. 79, No. 2, pp. 496–502, 2011 Copyright Ó 2011 Elsevier Inc. Printed in the USA. All rights reserved 0360-3016/$–see front matter
doi:10.1016/j.ijrobp.2009.10.070
CLINICAL INVESTIGATION
Liver
THREE-DIMENSIONAL CONFORMAL RADIATION THERAPY AND INTENSITY-MODULATED RADIATION THERAPY COMBINED WITH TRANSCATHETER ARTERIAL CHEMOEMBOLIZATION FOR LOCALLY ADVANCED HEPATOCELLULAR CARCINOMA: AN IRRADIATION DOSE ESCALATION STUDY ZHI-GANG REN, M.D.,* JIAN-DONG ZHAO, M.D.,* KE GU, M.D.,* ZHEN CHEN, M.D.,y JUN-HUA LIN, M.D.,y ZHI-YONG XU, PH.D.,* WEI-GANG HU, PH.D.,* ZHEN-HUA ZHOU, M.D.,y LU-MING LIU, M.D.,y AND GUO-LIANG JIANG, M.D., F.A.C.R. (HON)* Departments of *Radiation Oncology and yIntegrative Oncology, Fudan University Shanghai Cancer Center, 270 Dong An Road, Shanghai, China, 200032 Purpose: To determine the maximum tolerated dose (MTD) of three-dimensional conformal radiation therapy (3DCRT)/intensity-modulated radiation therapy (IMRT) combined with transcatheter arterial chemoembolization for locally advanced hepatocellular carcinoma. Methods and Materials: Patients were assigned to two subgroups based on tumor diameter: Group 1 had tumors <10 cm; Group II had tumors $10 cm. Escalation was achieved by increments of 4.0 Gy for each cohort in both groups. Dose-limiting toxicity (DLT) was defined as a grade of $3 acute liver or gastrointestinal toxicity or any grade 5 acute toxicity in other organs at risk or radiation-induced liver disease. The dose escalation would be terminated when $2 of 8 patients in a cohort experienced DLT. Results: From April 2005 to May 2008, 40 patients were enrolled. In Group I, 11 patients had grade #2 acute treatment-related toxicities, and no patient experienced DLT; and in Group II, 10 patients had grade #2 acute toxicity, and 1 patient in the group receiving 52 Gy developed radiation-induced liver disease. MTD was 62 Gy for Group I and 52 Gy for Group II. In-field progression-free and local progression-free rates were 100% and 69% at 1 year, and 93% and 44% at 2 years, respectively. Distant metastasis rates were 6% at 1 year and 15% at 2 years. Overall survival rates for 1-year and 2-years were 72% and 62%, respectively. Conclusions: The irradiation dose was safely escalated in hepatocellular carcinoma patients by using 3DCRT/ IMRT with an active breathing coordinator. MTD was 62 Gy and 52 Gy for patients with tumor diameters of <10 cm and $10 cm, respectively. Ó 2011 Elsevier Inc. Hepatocellular carcinoma, Three-dimensional conformal radiation therapy, Intensity-modulated radiation therapy, Maximum tolerance dose.
Hepatocellular carcinoma (HCC) is one of the most common cancers in Asia. The treatment of choice for HCC is surgery, with 5-year survival rates of 50% to 90% in carefully selected patients. Unfortunately, only 20% to 25% of HCC patients are amenable to curative surgery at the time of diagnosis because there is no mass screening program for early detection of HCC in China. Nonsurgical treatment modalities, including transcatheter arterial chemoembolization (TACE) and tumor ablations (radiofrequency, high-intensity focused
ultrasound, etc.) have not been demonstrated to be very effective for locally advanced HCC. There was no place for radiotherapy in the management of HCC until the innovation of modern radiation technology, which included three-dimensional conformal radiation therapy (3DCRT) and intensity-modulated radiation therapy (IMRT). Since then, 3DCRT has been tried in the treatment of medically inoperable and technically unresectable HCC cases and has produced encouraging outcomes. Thus, radiation therapy has come to be recognized as having a role in the
Reprint requests to: Guo-Liang Jiang, Department of Radiation Oncology, Fudan University Shanghai Cancer Center, 270 Dong An Road, Shanghai, China, 200032. Tel.: 8621 64175590; Fax: 8621 64439052; E-mail: [email protected] Zhi-Gang Ren and Jian-Dong Zhao share first authorship. This study was supported by a grant from Ministry of Health, China (2004-826) and in part by National Science Foundation of China (grant 30800279).
A part of this study was presented at the 50th Annual Meeting of the American Society of Therapeutic Radiology and Oncology, Boston, MA, Sep 21-Sep 25- Sep 25, 2008. Conflict of interest: none. Acknowledgment—We thank Liu Taifu, M.D., F.A.C.R. (Hon.) for assistance with editing of this article. Received July 25, 2009, and in revised form Oct 28, 2009. Accepted for publication Oct 29, 2009.
INTRODUCTION
496
Radiation dose escalation study for HCC d Z.-G. REN et al.
palliative care of locally advanced HCC or a potentially curative option for early cases (1). Based on recent experience, we realized that HCC was not as radioresistant as we used to think, and its radiosensitivity was similar to that of epidermal carcinomas (2). Also, it was found there was a dose-response relationship. Doses of more than 50 Gy resulted in enhanced tumor response rates (3, 4), and an escalated irradiation dose was independently associated with improved outcome (5, 6). In our previous clinical studies, we also noticed that results for patients treated with higher 3DCRT doses yielded better local control and survival of HCC (7, 8). In order to further investigate increases of dose, we conducted a dose escalation study as a clinical phase I trial to obtain the maximum tolerated dose (MTD) in HCC patients. Moreover, from our dosimetric study, which compared 3DCRT with IMRT, we found that IMRT had more potential to increase tumor dose and simultaneously to spare normal liver (unpublished data). Therefore, in the current trial, we used both 3DCRT and IMRT. METHODS AND MATERIALS Patient eligibility This clinical trial was approved by the Research Ethics Board of Fudan University Shanghai Cancer Center and was registered with ClinicalTrials.gov (NCT 00848094). Inclusion criteria were patient age equal to or older than 18 years; histologically or cytologically confirmed HCC or clinical diagnosis based on clinical criteria proposed by the Chinese Society of Liver Cancer (9); presence of a solitary lesion without intrahepatic and distant metastases; a technically unresectable or medically inoperable lesion or a case in which surgery was declined; liver function level of Child–Pugh A; a Karnofsky performance status of $70; normal renal function and adequate bone marrow reservation; tolerance of an active breathing coordinator (ABC); and dose constraints of organs at risk (OAR) rated satisfactory as seen in 3DCRT/IMRT planning. The exclusion criteria included liver function of Child-Pugh B or C; multiple lesions in the liver or intrahepatic or extrahepatic metastases; indistinct tumor boundary on computed tomography/magnetic resonance imaging (CT/MRI); intolerance of an ABC; and dose constraints of OAR not possible to realize even with IMRT.
Treatment Patients received no more than four courses of TACE, which was followed by irradiation after an interval of 4 weeks. TACE was carried out using the Seldinger technique. The emulsion, including 5-fluorouracil (500-600 mg/m2), cisplatinum (30-40 mg/m2), epirubicin (Epiadriamycin) (40-60 mg/m2), and iodized oil, was injected first, followed by gelatin foam (Gelfoam) embolization. The interval between the TACE courses was 4 to 6 weeks. The methods for patient immobilization and application of ABC have been described previously (10). An ABC was used for all patients. Target volume definitions were as follows: on the simulation CT scan, the gross tumor volume (GTV) was delineated on the axial CT images with the aid of deposited iodinated oil, MRI, or contrastenhanced CT images. Clinical target volume was determined by adding a uniform margin of 5 to 8 mm around the GTV within the liver. The extra margin needed for internal target volume was individually determined depending on the reproducibility of the lesion/ liver position with the ABC. Planning target volume (PTV) was
497
formed by adding a margin of 6 mm around the internal target volume. The tumor dose was escalated from 40 Gy to 62 Gy as prescribed at the isocenter with inhomogeneity correction. The requirements for tumor dose coverage were (1) the PTV should be covered by a 95% isodose volume and (2) inhomogeneity was less than 10%. Dose constraints for OAR were as follows. (1) Liver: mean dose to normal liver (MDTNL) (where normal liver volume = total liver volume minus PTV) was limited to #23 Gy, and the dose-volume histogram (DVH) of the normal liver was within the tolerance area, i.e., V5 of <86%, V10 of <68%, V15 of <59%, V20 of <49%, V25 of <35%, V30 of <28%, V35 of <25%, and V40 of <20% (11); (2) stomach and duodenum: the volume receiving >50 Gy was limited to <1 cm3; (3) spinal cord: the point dose was <50 Gy; and (4) kidney: if one kidney received >20 Gy, >90% of the other kidney would be excluded from the primary beam. Planning with the 3DCRT technique was tried first. If it met the dose requirements and constraints, 3DCRT would be used. Otherwise, various inverse IMRT plans with a step-and-shoot technique were made until all dosimetric parameters were satisfied. When dose constraints of OAR could not be met even after applying IMRT, the patient would be withdrawn from the study. Coplanar and noncoplanar portals were designed so that the normal liver volume would receive as little radiation as possible. For the IMRT plan, we used a simplified IMRT technique, which met the following requirements: the number of portals was #5; the number of subfields in each portal was #5; and the smallest size of subfield was larger than 5 cm 5 cm.
Dose escalation strategy Patients were assigned to two subgroups based on their biggest tumor diameter, i.e., patients with tumors of <10 cm were assigned to Group I, and patients with tumors of $10 cm were assigned to Group II. Total tumor doses of 42 Gy (Group I) and 40 Gy (Group II) in 2 Gy per fraction, five fractions per week, were used as the initial doses. Escalation was carried out in each cohort at increments of 4 Gy tumor dose. The dose escalation started from the lowest dose level, and only one cohort was open for enrollment in a group. Subsequent cohorts received higher doses, up to a predetermined maximum of 62 Gy in Group I and 52 Gy in Group II. In each dose level, 4 patients were enrolled. The dose-limiting toxicity (DLT) was defined as a grade of $3 acute hepatic or gastrointestinal toxicity or any grade 5 treatment-related adverse event during irradiation or radiation-induced liver disease (RILD) within 4 months after irradiation. Toxicity was graded according to the Common Terminology Criteria for Adverse Events version 3.0 (National Cancer Institute). Diagnoses of hepatic injuries would be made only when there was no evidence of intrahepatic progression of HCC. The rule for termination of dose escalation was as follows. After irradiation was completed, patients had to be followed for 4 months to observe late complications, especially RILD. The higher dose level would not be started until 4 patients in the previous dose level demonstrated no DLT. If one patient experienced DLT, an additional 4 patients would be added to that dose-level cohort. When DLT occurred in 1 of 8 (1/8) patients, the dose escalation would be continued, and once DLT occurred in $2 ($2/4 or $2/8) patients, the dose escalation was terminated.
Evaluation and follow-up Physical examinations, complete blood cell counts, and blood chemistry analyses were performed weekly during radiotherapy.
498
I. J. Radiation Oncology d Biology d Physics
After patients completed irradiation, they were followed every 2 months for 4 months and then every 4 months thereafter. CT or MRI scans were taken on the follow-up visit. RECIST criteria were used to assess tumor response and progression. In-field failure was defined as disease progression inside the irradiated volume and out-of-field (intrahepatic) failure as outside the irradiated volume but limited to the liver. In-field progression-free survival (IFPFS) was defined as the time in which the HCC lesion within the irradiated volume remained stable after a decrease in size. The term local progression-free survival (LPFS) was also used, which included in-field and/or out-of-field failures. The other endpoints were median survival time (MST), overall survival (OS) rate, and distant metastasis (DM) rate. All endpoints were estimated using the Kaplan-Meier model, and the observation was started from the date of treatment.
RESULTS Dose escalation status Just before the start of the present study, we found from our previous clinical trial (8), in which TACE and 3DCRT were used and the patients receiving 42 Gy had been followed for over 4 months, no DLT was observed. Thus, it was evident that the total dose of 42 Gy could be tolerated by HCC patients. Therefore, we decided to skip the cohort that received 42 Gy and started with 46 Gy administered to Group I. The dose escalation was successfully carried out to the preplanned highest dose level (62 Gy), and no DLT was recorded. In Group II, no patients experienced DLT until 52 Gy was given, at which point the fourth patient developed DLT (RILD). Thus, an additional 4 patients were enrolled into the 52-Gy cohort, and no further DLT was found. The dose escalation was completed according to the original plan.
Patients From April 2005 to May 2008, a total of 58 HCC patients were enrolled, and 18 patients were excluded due to either intolerance to the ABC or because dose constraints of OAR could not be met. The remaining 40 HCC patients were enrolled in this study, with 20 patients in each group. Thirtytwo patients were diagnosed by histology or cytology and 8 patients by clinical diagnosis. Thirty-six patients were male and 4 patients were female, with a median age of 52.5 years (range, 33-68 years). The median Karnofsky performance status was 80 (range, 70-90). All patient conditions were associated with a liver function of Child-Pugh grade A and a single lesion without intrahepatic, regional node or extrahepatic metastases. According to AJCC staging (2006), there were 20 cases of T2 (stage II), 17 cases of T3 (stage IIIA), and 3 cases of T4 (stage IIIB). Hepatitis B virus infection was detected in 37 patients (93%). An alfa-fetoprotein concentration of >400 IU/ml was found in 17 (43%) patients, 20 IU/ml-400 IU/ml in 10 (25%) patients, and <20 IU/ ml in 13 (33%) patients. The median volume of GTV was 514 cm3 (range, 57-1,336 cm3), and the median maximum lesion diameter was 10.5 cm (range, 5.1-16.4 cm). Patient clinical characteristics are shown in Table 1.
Volume 79, Number 2, 2011
A median of two courses (range, 1-4) of TACE was administered. Irradiation was delivered by 3DCRT in 24 patients and by IMRT in 16 patients. When the total dose was $48 Gy or $58 Gy, IMRT had to be used in 75% and 88% of patients, respectively, for lesions of $10 cm and <10 cm, whereas, none of the patients needed IMRT when lower doses were delivered. For all 40 patients, the median tumor dose was 51 Gy (range, 40-62 Gy), and the MDTNL was 17 Gy (range, 922 Gy). Dosimetric parameters for normal liver, stomach, and duodenum, including the uninvolved normal liver volume, and mean doses to the organs, the maximum dose, and the V5-V40 range of the uninvolved liver are listed in Tables 2 and 3. Adverse events and toxicity The last follow-up was performed in January 2009, with a median follow-up time of 13 months (range, 2-40 months) for the entire group. Acute treatment-related toxicities of grades 1 and 2 were recorded in 11 (55%) patients in Group I and 10 (50%) patients in Group II, and no grade of $3 toxicity was observed (Table 4). The acute hepatic toxicities were mainly demonstrated by abnormalities in serum levels of alkaline phosphatase, glutamic-oxaloacetic transaminase, and glutamic-pyruvic transaminase. Gastrointestinal (GI) toxicities included nausea and diarrhea. All patients recovered from acute toxicities after appropriate treatments, and none of the patients needed to discontinue irradiation. Regarding late complications, one nonclassic case of RILD was observed in the 52-Gy cohort of Group II. The onset of RILD occurred 2 months after completion of irradiation, and the patient died of hepatic failure in 2 weeks. This patient had suffered from hepatitis B infection for over 10 years and was positive for hepatitis surface antigen (HBsAg), chronic HBV infection (HBeAg), and HB core antigen (HBcAb) at the diagnosis of HCC; also, the patient had a history of diabetes mellitus. The patient’s uninvolved normal liver volume was 827 cm3. The dose to normal liver was limited to a tolerable level (MDTNL of 13.09 Gy and DVH V5 of 43%, V10 of 35%,V15 of 30%, V20 of 28%, V25 of 24%, V30 of 19%, V35 of 16%, and a V40 of 12%). Local control, distant metastasis, and survival At the last follow-up visit, 24 patients were still alive (13 patients in Group I and 11 patients in Group II). Among the survivors there were 14 patients with local control and without evidence of intrahepatic and distant metastases (7 patients in Group I and 7 patients in Group II) and 10 patients whose outcomes were associated with failures, including in-field failures alone (2 patients in Group II), out-of-field failures alone (4 patients in Group I and 1 patients in Group II), both in-field and out-of-field failures (2 patients in Group I), and distant metastasis (1 patients in Group II). Sixteen patients died of out-of-field failure (4 patients in Group I and 7 patients in Group II), distant metastases (1 patients in Group II), both out-of-field and distant metastasis (3 patients in
Radiation dose escalation study for HCC d Z.-G. REN et al.
499
Table 1. Clinical characteristics in 40 hepatocellular carcinomas Patient characteristics Sex Female Male Age (yr) Median KPS 70 80 90 Diagnosis by Cytology Histology Clinical Serum AFP (ng/ml) #20 20-400 $400 HBV infection Positive Negative TACE Median course Median maximum tumor diameter (cm) GTV (cc) Median AJCC stage T1N0M0 (I) T2N0M0 (II) T3N0M0 (IIIA) T4N0M0 (IIIB)
Group I (<10-cm tumors) (range)
Group II ($10-cm tumors) (range)
Total
0 20
4 16
4 36
51 (36–68)
55.5 (33–68)
3 16 1
2 15 3
5 31 4
1 18 1
3 10 7
4 28 8
7 5 8
6 5 9
13 10 17
19 1
18 2
37 3
2 (1–3) 7.5 (5.1–9.5)
2 (1–4) 13.6 (10.5–16.4)
2 (1–4) 10.5 (5.1–16.4)
209 (57–474)
819 (506–1336)
514 (57–1336)
0 15 5 0
0 5 12 3
0 20 17 3
52.5 (33–68)
Abbreviations: AFP = alfa-fetoprotein; KPS = Karnofsky performance status; HBV = hepatitis B virus.
Group I), or RILD in 1 patient. Overall, there were 4 patients with in-field failures, 21 out-of-field failures, and 5 distant metastases (Table 5). For the entire 40 patients, IFPFS was 100% at 1 year and 93% at 2 years, and 1-year and 2-year LPFS rates were 69% and 44%, respectively. DM rates were 6% and 15% at 1 year and 2 years, respectively. MST was not reached because more than half of the patients were still alive in Group I and at 22 months (95% confidence interval, 11-33 months) in Group II. The 1-year and 2-year OS rates were 72% and 62%, respectively (Fig 1). DISCUSSION The fractionation and total doses of radiation therapy for HCC varied considerably in a series of publications. Dawson et al. (6) treated 43 patients with intrahepatic hepatobiliary cancer and liver metastases by 3DCRT with a median total dose of 58.5 Gy at 1.5 Gy per fraction twice daily and found that the escalated RT dose was independently associated with improved outcome. Also, we learned from a previous study of HCC that a higher dose yielded a better outcome, with an MST of 38 months and 12 months and a 3-year OS rate of 45% and 17%, respectively, for patients receiving $45 Gy and <45 Gy (8). Thus, we believed that higher irradiation
doses would be needed to produce better local controls. However, a high dose of hepatic irradiation would injure the liver, leading to RILD, which was almost a fatal complication (11). Therefore, we conducted the current clinical phase I trial to escalate irradiation dose and to determine what would be the MTD when irradiation was administered after TACE for HCC patients. When designing the protocol, we had to answer the following questions: What was the dose fractionation, hypofractionation or conventional? What was the preplanned highest dose level? Which irradiation technique should be used, 3DCRT or IMRT? With regard to fractionation, based on our previous study (7), hypofractionation by 5 Gy/fraction given every other day really produced a 3-year OS rate of 33%, but in 19 (19/128 [15%]) patients, RILD occurred, and 16 of 19 patients died as a result. Based on radiobiology studies, the liver is considered a late-responding normal organ with a low a/ b value (12, 13). Theoretically, conventional fractionation would lessen hepatic injury compared to the large fraction size. Therefore, we decided to use conventional fractionation in the current dose escalation study. The preplanned highest doses for escalation were 52 Gy and 62 Gy, respectively, for tumor sizes of <10 cm and $10 cm, which was determined by our experience with
I. J. Radiation Oncology d Biology d Physics
500
Table 2. Median dosimetric parameters of normal liver Parameter Uninvolved normal liver volume (cm3)* MDTNL (Gy)y V5 (%)z V10 (%) V15 (%) V20 (%) V25 (%) V30 (%) V35 (%) V40 (%) Probability of RILD (%)x
Group I (<10 cm)
Group II ($10 cm)
3
3
1,248 cm 1,042 cm (range){ (range)** 16 (9–22) 18 (15–20) 55 (27–80) 58 (41–75) 45 (21–66) 50 (35–66) 38 (18–58) 39 (34–55) 33 (15–52) 32 (30–45) 28 (11–48) 28 (23–28) 22 (9–30) 24 (18–26) 19 (7–28) 19 (17–23) 13 (5–26) 17 (12–19) 2.2 (0.6–6.9) 1.9 (0.6–8.7)
p value <0.05 >0.05 >0.05 >0.05 >0.05 >0.05 >0.05 >0.05 >0.05 >0.05
* Uninvolved normal liver volume = total liver volume – planning target volume. y MDTNL (mean dose to normal liver) was derived using the formula, normal liver volume = total liver volume – planning target volume. z V5 = percentage of normal liver volume which received $5 Gy in total liver volume, and in V10-V40, the suffix indicates the dose ($) received by normal liver. x The RILD (radiation-induced liver disease) was estimated by an in-house Lyman model (n = 1.1, m = 0.28, TD50 [1] = 40.5 Gy). TD50 [1], the whole liver dose, which will produce 50% probability of toxicity; ‘‘m,’’characterizing the steepness of the dose response at TD50; and ‘‘n,’’ a volume effect parameter that indicates a larger volume effect as it increases (range, 0-1). { Range, 792 to 1,639 cm3. ** Range, 805 to 1,921 cm3.
a dosimetric study (unpublished data). That study demonstrated that for the tumor located close to the GI tract, the highest dose we could deliver was 50 Gy while keeping doses to the GI tract at tolerable levels. However, if the tumor location was distant from the GI tract, the highest doses that could be given were 52 Gy and 62 Gy for tumors of <10 cm and $10 cm, respectively, while both the GI tract dose and the liver dose constraints could be met. IMRT has been widely used for prostate and head and neck cancers but not for HCC. From our (unpublished) data and from that of other dosimetric studies (14–16), when comparing IMRT and 3DCRT for HCC, we found that tumor dose conformality by IMRT was superior to that by 3DCRT and that IMRT also could reduce MDTNL. In the current study, for patients with a tumor of $10 cm when the total dose was more than 48 Gy, or for those patients with a tumor of
Volume 79, Number 2, 2011
<10 cm but a tumor dose of more than 58 Gy, IMRT had to be used in 75% and 88% of patients, respectively, to satisfy the dose constraints of OAR. Otherwise, such high doses were impossible to deliver. However, to implement IMRT, we realized that IMRT was more sensitive to liver motion due to respiration (17), and hence, it was not appropriate to a moving target because of dose uncertainty. To eliminate liver motion, we used an ABC device to immobilize liver lesions and to make delivery of IMRT possible. In our recent report, the systematic and random set-up errors and intrafraction reproducibility of the diaphragm were in a range of 1.2 mm to 1.8 mm (10). Although, there was still some uncertainty of irradiation delivery, we thought that it was acceptable. Moreover, we used a simplified IMRT technique, which limited the number of subfields to #5 and the smallest subfield size to $5 cm 5 cm. Using these measurements, we could probably reduce the negative impact of IMRT, which was imposed by set-up error or liver reproducibility by a ABC. In spite of the fact that IMRT provided a better plan, we did not use IMRT for all patients because IMRT planning was time consuming. We needed more manpower and time to design, verify, and deliver IMRT plans, which in our very busy department, we could not afford. Therefore, for HCC, we generally use 3DCRT except when the desired dose constraints could not be met. In the current study, the acute toxicity was acceptable, with 11 cases of grade 1 hepatic toxicity and no case of a grade $3 GI toxicity in both groups. Unfortunately, one patient developed RILD and died as a result. Although in this case the radiation dose to the normal liver was limited to a tolerable level, we thought that the RILD was due to the patient’s hepatitis B infection-induced cirrhosis, and in addition, the patient’s associated diabetes mellitus probably made the hepatic tolerance worse. Considering that RILD was almost a fatal complication with 84% mortality in our previous series (7), prevention is critical. The best way to avoid RILD is to limit the hepatic dose to a tolerable level. From our previous analysis (11, 18), we found that tolerance to radiation therapy was much poorer in Chinese HCC patients than in metastatic liver cancer patients. This is because in over 90% of our patients, HCC developed on the basis of hepatitis B infection-induced cirrhosis, which impaired the liver’s capability to repair irradiation damage and to proliferate. According to our previous analysis of HCC treated by hypofractionation, we proposed
Table 3. Mean dosimetric parameters of stomach and duodenumy Stomach
Duodenum
Dose
Group I (<10 cm)
Group II ($10 cm)
Group I (<10 cm)
Group II ($10 cm)
Mean SD (Gy)* Max SD (Gy)*
5.4 6.4 27.0 20.4
9.9 10.5 31.3 17.7
10.6 12.1 24.0 20.0
14.4 13.7 28.2 19.6
* Mean values standard deviations (SD). p > 0.05 for all comparisons between the values of Group I and Group II.
y
Radiation dose escalation study for HCC d Z.-G. REN et al.
501
Table 4. Patient data and treatment-related toxicity in 40 hepatocellular carcinomas Dose level Group I (<10 cm) 46 Gy 50 Gy 54 Gy 58 Gy 62 Gy Group II ($10 cm) 40 Gy 44 Gy 48 Gy 52 Gy
No. of patients
No. of patients with DLT
No. of patients with non-DLT toxicity
4 4 4 4 4
None None None None None
None GI Gr 1 (1), H Gr 1 (1) GI Gr 1 (1), H Gr 1 and GI Gr 1 (1) GI Gr 1 (1), H Gr 1 and GI 1 (1), GI Gr 2 (1) GI Gr 1 (1), H Gr 1 and GI Gr 1 (2), GI Gr 2 (1)
4 4 4 8
None None None H Gr 5 (1)
None GI Gr 1 (1), H Gr 1 and GI Gr 1 (1) GI Gr 1 (1), H Gr 1 and GI Gr 1 (2) H Gr 1 and GI Gr 1 (3), GI Gr 2 (2)
Abbreviations: H = hepatic toxicity; Gr = grade.
that the liver’s tolerance for HCC associated with a liver function of Child-Pugh A was 23 Gy in terms of MDTNL and a tolerable DVH (11). In this study, we used conventional fractionation but chose hepatic dose constraints, which resulted from our hypofractionation study (7). The reason was based on the following considerations. From the hypofractionation study, we learned that a liver with hepatic cirrhosis tolerated irradiation more poorly than liver without cirrhosis. However, we really did not know exactly what the hepatic tolerance dose was for patients with cirrhosis, if treated by conventional fractionation, although we expected it would be higher than that by hypofractionation. In order to totally avoid RILD, we decided to use a stricter dose constraint. In the literature, Robertson et al. (19) used a V50% of normal liver to determine the total tumor dose. Because we were not convinced that this guideline was suitable for our patients with cirrhosis backgrounds and also because the purpose of this study was to escalate dose and obtain MTD, we did not apply the dose prescription of Robertson et al. In the current trial, we estimated that patients would be able to tolerate the dose with no severe hepatic injury. The RILD incidence of 2.5% (1/40 patients) in this study demonstrated the validity of our proposal for hepatic irradiation tolerance in HCC patients. Another measurement we exercised in irradiation planning was to limit the portal number to #5 to avoid a large volume of normal liver receiving low dose and to always have a part of the normal liver excluded from any beam. We expected that the protected normal liver would proliferate to compensate for the loss of hepatic function
during irradiation, which was demonstrated by our animal experiment (20). From the current study, we concluded that MTDs were 52 Gy and 62 Gy for patients with tumors of $10 cm and <10 cm in diameter, respectively. One question raised from our study was why patients with big tumors tolerated irradiation more poorly than those with small tumors while dose constraints were satisfied in both patients. As shown in Table 3, both mean doses and maximum doses to the stomach and duodenum were higher in patients with tumors of $10 cm than those with tumors of <10 cm, although the differences were not statistically significant. Also, there were no differences in MDTNL and DVH of normal liver in both groups, Therefore, for both groups, doses to liver, stomach, and duodenum were similar; however, the total doses already delivered to tumors were 40 Gy to 52 Gy in $10-cm tumors and 46 Gy to 62 Gy in <10-cm tumors; and OAR would not be able to tolerate doses increased to more than 52 Gy to $10-cm tumors. Moreover, there was 1,042 cm3 of uninvolved normal liver in patients with tumors of $10 cm, whereas there was 1,248 cm3 in those with tumors of <10 cm (p < 0.05) (Table 2). Thus, we thought that a smaller, uninvolved normal liver volume and probably higher doses to
Table 5. Failure patterns in 40 cases of HCC* Failure
No. of patients
% of patients
In field only Out-of-field only In field and out-of-field Out-of-field and distant Distant only
2 16 2 3 2
8 64 8 12 8
* Data indicate patterns of 25 failures in 40 patients with HCC after receiving a combination of transcatheter arterial chemoembolization and irradiation.
Fig. 1. Overall survival rates in 40 patients with hepatocellular carcinomas after receiving a combination of transarterial chemoembolization and irradiation.
502
I. J. Radiation Oncology d Biology d Physics
OARs would be the obstacle to further escalating the dose to patients with big tumors. As to the outcome of this study, IFPFS rates were 100% and 93%, and LPFS rates were 69% and 45%, respectively, at 1 year and 2 years, which implied that tumor control inside the field was excellent but the intrahepatic failure out-of-field was predominantly the failure pattern in 21 patients, which accounted for 84% of failures. We believe that it is attributed to the HCC biological behavior, i.e., the tendency to spread along the portal vein or multiple primary foci. Further efforts are needed to deal with the intrahepatic failures. DM was not a major cause of death in this trial with low DM rates (6% and 15% at 1 year and 2 years, respectively). The OS rates of 72% at 1 year and 62% at 2 years were encouraging and were greatly improved compared to OS rates of 60% at 1 year and 38% at 2 years in our latest publication (8). Moreover, more advanced tumors (median diameter of 10.5 cm) were encountered in the current study, whereas in
Volume 79, Number 2, 2011
the previous study, tumor diameter was 6.5 cm. It was obvious that the high tumor dose contributed to the improved survival (median tumor dose of 51 Gy in the current study vs. 43 Gy in the previous study [8]). Thus, our proposal is to deliver a dose as high as possible to HCC while keeping doses to OAR at tolerable levels. However, we thought that the patient selection could have contributed to the good outcome. Patients in this trial might have less potential to metastasize since only patients without distant metastases had been enrolled. CONCLUSIONS In summary, for HCC patients with Child-Pugh A liver function, MTD was 52 Gy and 62 Gy in patients with tumors of $10 cm and <10 cm in diameter, respectively, after TACE, which yielded an improved outcome and acceptable toxicity.
REFERENCES 1. Hawkins MA, Dawson LA. Radiation therapy for hepatocellular carcinoma: from palliation to cure. Cancer 2006;106:1653– 1663. 2. Cheng SH, Lin YM, Chuang VP, et al. A pilot study of three-dimensional conformal radiotherapy in unresectable hepatocellular carcinoma. J Gastroenterol Hepatol 1999;14:1025–1033. 3. Park W, Lim DH, Paik SW, et al. Local radiotherapy for patients with unresectable hepatocellular carcinoma. Int J Radiat Oncol Biol Phys 2005;61:1143–1150. 4. Park HC, Seong J, Han KH, et al. Dose-response relationship in local radiotherapy for hepatocellular carcinoma. Int J Radiat Oncol Biol Phys 2002;54:150–155. 5. Seong J, Park HC, Han KH, et al. Clinical results and prognostic factors in radiotherapy for unresectable hepatocellular carcinoma: A retrospective study of 158 patients. Int J Radiat Oncol Biol Phys 2003;55:329–336. 6. Dawson LA, McGinn CJ, Normolle D, et al. Escalated focal liver radiation and concurrent hepatic artery fluorodeoxyuridine for unresectable intrahepatic malignancies. J Clin Oncol 2000; 18:2210–2218. 7. Liang SX, Zhu XD, Lu HJ, et al. Hypofractionated three-dimensional conformal radiation therapy for primary liver carcinoma. Cancer 2005;103:2181–2188. 8. Zhou ZH, Liu LM, Chen WW, et al. Combined therapy of transcatheter arterial chemoembolization and three-dimensional conformal radiotherapy for hepatocellular carcinoma. Br J Radiol 2007;80:194–201. 9. Yang BH. Primary liver cancer. In: ChineseAnti-Cancer Association, editor. Guidelines for common cancers in China. Beijing: Associated Press of Peking Medical University and Peking Union Medical College; 1999. p. 389–479. 10. Zhao JD, Xu ZY, Zhu J, et al. Application of active breathing control in 3-dimensional conformal radiation therapy for hepatocellular carcinoma: The feasibility and benefit. Radiother Oncol 2008;87:439–444. 11. Liang SX, Zhu XD, Xu ZY, et al. Radiation-induced liver disease in three-dimensional conformal radiation therapy for
12. 13. 14.
15.
16.
17.
18.
19. 20.
primary liver carcinoma: The risk factors and hepatic radiation tolerance. Int J Radiat Oncol Biol Phys 2006;65:426–434. Fisher DR, Hendry JH. Dose fractionation and hepatocyte clonogens: Alpha/beta congruent to 1-2 Gy, and beta decreases with increasing delay before assay. Radiat Res 1988;113:51–57. Dawson LA, Lawrence TS, Ten Haken RK. In response to Dr. Tome and Dr. Fenwick. Int J Radiat Oncol Biol Phys 2004;58:1319–1320. Cheng JC, Wu JK, Huang CM, et al. Dosimetric analysis and comparison of three-dimensional conformal radiotherapy and intensity-modulated radiation therapy for patients with hepatocellular carcinoma and radiation-induced liver disease. Int J Radiat Oncol Biol Phys 2003;56:229–234. Thomas E, Chapet O, Kessler ML, et al. Benefit of using biologic parameters (EUD and NTCP) in IMRT optimization for treatment of intrahepatic tumors. Int J Radiat Oncol Biol Phys 2005;62:571–578. Eccles CL, Bissonnette JP, Craig T, et al. Treatment planning study to determine potential benefit of intensity-modulated radiotherapy versus conformal radiotherapy for unresectable hepatic malignancies. Int J Radiat Oncol Biol Phys 2008;72: 582–588. Fuss M, Salter BJ, Herman TS, et al. External beam radiation therapy for hepatocellular carcinoma: Potential of intensitymodulated and image-guided radiation therapy. Gastroenterology 2004;127:S206–S217. Xu ZY, Liang SX, Zhu J, et al. Prediction of radiation-induced liver disease by Lyman normal-tissue complication probability model in three-dimensional conformal radiation therapy for primary liver carcinoma. Int J Radiat Oncol Biol Phys 2006;65: 189–195. Robertson JM, Lawrence TS, Dworzanin LM, et al. Treatment of primary hepatobiliary cancers with conformal radiation therapy and regional chemotherapy. J Clin Oncol 1993;11:1286–1293. Zhao JD, Jiang GL, Hu WG, et al. Hepatocyte regeneration after partial liver irradiation in rats. Exp Toxicol Pathol 2009;61:511– 518.
doi:10.1016/j.ijrobp.2009.10.070
CLINICAL INVESTIGATION
Liver
THREE-DIMENSIONAL CONFORMAL RADIATION THERAPY AND INTENSITY-MODULATED RADIATION THERAPY COMBINED WITH TRANSCATHETER ARTERIAL CHEMOEMBOLIZATION FOR LOCALLY ADVANCED HEPATOCELLULAR CARCINOMA: AN IRRADIATION DOSE ESCALATION STUDY ZHI-GANG REN, M.D.,* JIAN-DONG ZHAO, M.D.,* KE GU, M.D.,* ZHEN CHEN, M.D.,y JUN-HUA LIN, M.D.,y ZHI-YONG XU, PH.D.,* WEI-GANG HU, PH.D.,* ZHEN-HUA ZHOU, M.D.,y LU-MING LIU, M.D.,y AND GUO-LIANG JIANG, M.D., F.A.C.R. (HON)* Departments of *Radiation Oncology and yIntegrative Oncology, Fudan University Shanghai Cancer Center, 270 Dong An Road, Shanghai, China, 200032 Purpose: To determine the maximum tolerated dose (MTD) of three-dimensional conformal radiation therapy (3DCRT)/intensity-modulated radiation therapy (IMRT) combined with transcatheter arterial chemoembolization for locally advanced hepatocellular carcinoma. Methods and Materials: Patients were assigned to two subgroups based on tumor diameter: Group 1 had tumors <10 cm; Group II had tumors $10 cm. Escalation was achieved by increments of 4.0 Gy for each cohort in both groups. Dose-limiting toxicity (DLT) was defined as a grade of $3 acute liver or gastrointestinal toxicity or any grade 5 acute toxicity in other organs at risk or radiation-induced liver disease. The dose escalation would be terminated when $2 of 8 patients in a cohort experienced DLT. Results: From April 2005 to May 2008, 40 patients were enrolled. In Group I, 11 patients had grade #2 acute treatment-related toxicities, and no patient experienced DLT; and in Group II, 10 patients had grade #2 acute toxicity, and 1 patient in the group receiving 52 Gy developed radiation-induced liver disease. MTD was 62 Gy for Group I and 52 Gy for Group II. In-field progression-free and local progression-free rates were 100% and 69% at 1 year, and 93% and 44% at 2 years, respectively. Distant metastasis rates were 6% at 1 year and 15% at 2 years. Overall survival rates for 1-year and 2-years were 72% and 62%, respectively. Conclusions: The irradiation dose was safely escalated in hepatocellular carcinoma patients by using 3DCRT/ IMRT with an active breathing coordinator. MTD was 62 Gy and 52 Gy for patients with tumor diameters of <10 cm and $10 cm, respectively. Ó 2011 Elsevier Inc. Hepatocellular carcinoma, Three-dimensional conformal radiation therapy, Intensity-modulated radiation therapy, Maximum tolerance dose.
Hepatocellular carcinoma (HCC) is one of the most common cancers in Asia. The treatment of choice for HCC is surgery, with 5-year survival rates of 50% to 90% in carefully selected patients. Unfortunately, only 20% to 25% of HCC patients are amenable to curative surgery at the time of diagnosis because there is no mass screening program for early detection of HCC in China. Nonsurgical treatment modalities, including transcatheter arterial chemoembolization (TACE) and tumor ablations (radiofrequency, high-intensity focused
ultrasound, etc.) have not been demonstrated to be very effective for locally advanced HCC. There was no place for radiotherapy in the management of HCC until the innovation of modern radiation technology, which included three-dimensional conformal radiation therapy (3DCRT) and intensity-modulated radiation therapy (IMRT). Since then, 3DCRT has been tried in the treatment of medically inoperable and technically unresectable HCC cases and has produced encouraging outcomes. Thus, radiation therapy has come to be recognized as having a role in the
Reprint requests to: Guo-Liang Jiang, Department of Radiation Oncology, Fudan University Shanghai Cancer Center, 270 Dong An Road, Shanghai, China, 200032. Tel.: 8621 64175590; Fax: 8621 64439052; E-mail: [email protected] Zhi-Gang Ren and Jian-Dong Zhao share first authorship. This study was supported by a grant from Ministry of Health, China (2004-826) and in part by National Science Foundation of China (grant 30800279).
A part of this study was presented at the 50th Annual Meeting of the American Society of Therapeutic Radiology and Oncology, Boston, MA, Sep 21-Sep 25- Sep 25, 2008. Conflict of interest: none. Acknowledgment—We thank Liu Taifu, M.D., F.A.C.R. (Hon.) for assistance with editing of this article. Received July 25, 2009, and in revised form Oct 28, 2009. Accepted for publication Oct 29, 2009.
INTRODUCTION
496
Radiation dose escalation study for HCC d Z.-G. REN et al.
palliative care of locally advanced HCC or a potentially curative option for early cases (1). Based on recent experience, we realized that HCC was not as radioresistant as we used to think, and its radiosensitivity was similar to that of epidermal carcinomas (2). Also, it was found there was a dose-response relationship. Doses of more than 50 Gy resulted in enhanced tumor response rates (3, 4), and an escalated irradiation dose was independently associated with improved outcome (5, 6). In our previous clinical studies, we also noticed that results for patients treated with higher 3DCRT doses yielded better local control and survival of HCC (7, 8). In order to further investigate increases of dose, we conducted a dose escalation study as a clinical phase I trial to obtain the maximum tolerated dose (MTD) in HCC patients. Moreover, from our dosimetric study, which compared 3DCRT with IMRT, we found that IMRT had more potential to increase tumor dose and simultaneously to spare normal liver (unpublished data). Therefore, in the current trial, we used both 3DCRT and IMRT. METHODS AND MATERIALS Patient eligibility This clinical trial was approved by the Research Ethics Board of Fudan University Shanghai Cancer Center and was registered with ClinicalTrials.gov (NCT 00848094). Inclusion criteria were patient age equal to or older than 18 years; histologically or cytologically confirmed HCC or clinical diagnosis based on clinical criteria proposed by the Chinese Society of Liver Cancer (9); presence of a solitary lesion without intrahepatic and distant metastases; a technically unresectable or medically inoperable lesion or a case in which surgery was declined; liver function level of Child–Pugh A; a Karnofsky performance status of $70; normal renal function and adequate bone marrow reservation; tolerance of an active breathing coordinator (ABC); and dose constraints of organs at risk (OAR) rated satisfactory as seen in 3DCRT/IMRT planning. The exclusion criteria included liver function of Child-Pugh B or C; multiple lesions in the liver or intrahepatic or extrahepatic metastases; indistinct tumor boundary on computed tomography/magnetic resonance imaging (CT/MRI); intolerance of an ABC; and dose constraints of OAR not possible to realize even with IMRT.
Treatment Patients received no more than four courses of TACE, which was followed by irradiation after an interval of 4 weeks. TACE was carried out using the Seldinger technique. The emulsion, including 5-fluorouracil (500-600 mg/m2), cisplatinum (30-40 mg/m2), epirubicin (Epiadriamycin) (40-60 mg/m2), and iodized oil, was injected first, followed by gelatin foam (Gelfoam) embolization. The interval between the TACE courses was 4 to 6 weeks. The methods for patient immobilization and application of ABC have been described previously (10). An ABC was used for all patients. Target volume definitions were as follows: on the simulation CT scan, the gross tumor volume (GTV) was delineated on the axial CT images with the aid of deposited iodinated oil, MRI, or contrastenhanced CT images. Clinical target volume was determined by adding a uniform margin of 5 to 8 mm around the GTV within the liver. The extra margin needed for internal target volume was individually determined depending on the reproducibility of the lesion/ liver position with the ABC. Planning target volume (PTV) was
497
formed by adding a margin of 6 mm around the internal target volume. The tumor dose was escalated from 40 Gy to 62 Gy as prescribed at the isocenter with inhomogeneity correction. The requirements for tumor dose coverage were (1) the PTV should be covered by a 95% isodose volume and (2) inhomogeneity was less than 10%. Dose constraints for OAR were as follows. (1) Liver: mean dose to normal liver (MDTNL) (where normal liver volume = total liver volume minus PTV) was limited to #23 Gy, and the dose-volume histogram (DVH) of the normal liver was within the tolerance area, i.e., V5 of <86%, V10 of <68%, V15 of <59%, V20 of <49%, V25 of <35%, V30 of <28%, V35 of <25%, and V40 of <20% (11); (2) stomach and duodenum: the volume receiving >50 Gy was limited to <1 cm3; (3) spinal cord: the point dose was <50 Gy; and (4) kidney: if one kidney received >20 Gy, >90% of the other kidney would be excluded from the primary beam. Planning with the 3DCRT technique was tried first. If it met the dose requirements and constraints, 3DCRT would be used. Otherwise, various inverse IMRT plans with a step-and-shoot technique were made until all dosimetric parameters were satisfied. When dose constraints of OAR could not be met even after applying IMRT, the patient would be withdrawn from the study. Coplanar and noncoplanar portals were designed so that the normal liver volume would receive as little radiation as possible. For the IMRT plan, we used a simplified IMRT technique, which met the following requirements: the number of portals was #5; the number of subfields in each portal was #5; and the smallest size of subfield was larger than 5 cm 5 cm.
Dose escalation strategy Patients were assigned to two subgroups based on their biggest tumor diameter, i.e., patients with tumors of <10 cm were assigned to Group I, and patients with tumors of $10 cm were assigned to Group II. Total tumor doses of 42 Gy (Group I) and 40 Gy (Group II) in 2 Gy per fraction, five fractions per week, were used as the initial doses. Escalation was carried out in each cohort at increments of 4 Gy tumor dose. The dose escalation started from the lowest dose level, and only one cohort was open for enrollment in a group. Subsequent cohorts received higher doses, up to a predetermined maximum of 62 Gy in Group I and 52 Gy in Group II. In each dose level, 4 patients were enrolled. The dose-limiting toxicity (DLT) was defined as a grade of $3 acute hepatic or gastrointestinal toxicity or any grade 5 treatment-related adverse event during irradiation or radiation-induced liver disease (RILD) within 4 months after irradiation. Toxicity was graded according to the Common Terminology Criteria for Adverse Events version 3.0 (National Cancer Institute). Diagnoses of hepatic injuries would be made only when there was no evidence of intrahepatic progression of HCC. The rule for termination of dose escalation was as follows. After irradiation was completed, patients had to be followed for 4 months to observe late complications, especially RILD. The higher dose level would not be started until 4 patients in the previous dose level demonstrated no DLT. If one patient experienced DLT, an additional 4 patients would be added to that dose-level cohort. When DLT occurred in 1 of 8 (1/8) patients, the dose escalation would be continued, and once DLT occurred in $2 ($2/4 or $2/8) patients, the dose escalation was terminated.
Evaluation and follow-up Physical examinations, complete blood cell counts, and blood chemistry analyses were performed weekly during radiotherapy.
498
I. J. Radiation Oncology d Biology d Physics
After patients completed irradiation, they were followed every 2 months for 4 months and then every 4 months thereafter. CT or MRI scans were taken on the follow-up visit. RECIST criteria were used to assess tumor response and progression. In-field failure was defined as disease progression inside the irradiated volume and out-of-field (intrahepatic) failure as outside the irradiated volume but limited to the liver. In-field progression-free survival (IFPFS) was defined as the time in which the HCC lesion within the irradiated volume remained stable after a decrease in size. The term local progression-free survival (LPFS) was also used, which included in-field and/or out-of-field failures. The other endpoints were median survival time (MST), overall survival (OS) rate, and distant metastasis (DM) rate. All endpoints were estimated using the Kaplan-Meier model, and the observation was started from the date of treatment.
RESULTS Dose escalation status Just before the start of the present study, we found from our previous clinical trial (8), in which TACE and 3DCRT were used and the patients receiving 42 Gy had been followed for over 4 months, no DLT was observed. Thus, it was evident that the total dose of 42 Gy could be tolerated by HCC patients. Therefore, we decided to skip the cohort that received 42 Gy and started with 46 Gy administered to Group I. The dose escalation was successfully carried out to the preplanned highest dose level (62 Gy), and no DLT was recorded. In Group II, no patients experienced DLT until 52 Gy was given, at which point the fourth patient developed DLT (RILD). Thus, an additional 4 patients were enrolled into the 52-Gy cohort, and no further DLT was found. The dose escalation was completed according to the original plan.
Patients From April 2005 to May 2008, a total of 58 HCC patients were enrolled, and 18 patients were excluded due to either intolerance to the ABC or because dose constraints of OAR could not be met. The remaining 40 HCC patients were enrolled in this study, with 20 patients in each group. Thirtytwo patients were diagnosed by histology or cytology and 8 patients by clinical diagnosis. Thirty-six patients were male and 4 patients were female, with a median age of 52.5 years (range, 33-68 years). The median Karnofsky performance status was 80 (range, 70-90). All patient conditions were associated with a liver function of Child-Pugh grade A and a single lesion without intrahepatic, regional node or extrahepatic metastases. According to AJCC staging (2006), there were 20 cases of T2 (stage II), 17 cases of T3 (stage IIIA), and 3 cases of T4 (stage IIIB). Hepatitis B virus infection was detected in 37 patients (93%). An alfa-fetoprotein concentration of >400 IU/ml was found in 17 (43%) patients, 20 IU/ml-400 IU/ml in 10 (25%) patients, and <20 IU/ ml in 13 (33%) patients. The median volume of GTV was 514 cm3 (range, 57-1,336 cm3), and the median maximum lesion diameter was 10.5 cm (range, 5.1-16.4 cm). Patient clinical characteristics are shown in Table 1.
Volume 79, Number 2, 2011
A median of two courses (range, 1-4) of TACE was administered. Irradiation was delivered by 3DCRT in 24 patients and by IMRT in 16 patients. When the total dose was $48 Gy or $58 Gy, IMRT had to be used in 75% and 88% of patients, respectively, for lesions of $10 cm and <10 cm, whereas, none of the patients needed IMRT when lower doses were delivered. For all 40 patients, the median tumor dose was 51 Gy (range, 40-62 Gy), and the MDTNL was 17 Gy (range, 922 Gy). Dosimetric parameters for normal liver, stomach, and duodenum, including the uninvolved normal liver volume, and mean doses to the organs, the maximum dose, and the V5-V40 range of the uninvolved liver are listed in Tables 2 and 3. Adverse events and toxicity The last follow-up was performed in January 2009, with a median follow-up time of 13 months (range, 2-40 months) for the entire group. Acute treatment-related toxicities of grades 1 and 2 were recorded in 11 (55%) patients in Group I and 10 (50%) patients in Group II, and no grade of $3 toxicity was observed (Table 4). The acute hepatic toxicities were mainly demonstrated by abnormalities in serum levels of alkaline phosphatase, glutamic-oxaloacetic transaminase, and glutamic-pyruvic transaminase. Gastrointestinal (GI) toxicities included nausea and diarrhea. All patients recovered from acute toxicities after appropriate treatments, and none of the patients needed to discontinue irradiation. Regarding late complications, one nonclassic case of RILD was observed in the 52-Gy cohort of Group II. The onset of RILD occurred 2 months after completion of irradiation, and the patient died of hepatic failure in 2 weeks. This patient had suffered from hepatitis B infection for over 10 years and was positive for hepatitis surface antigen (HBsAg), chronic HBV infection (HBeAg), and HB core antigen (HBcAb) at the diagnosis of HCC; also, the patient had a history of diabetes mellitus. The patient’s uninvolved normal liver volume was 827 cm3. The dose to normal liver was limited to a tolerable level (MDTNL of 13.09 Gy and DVH V5 of 43%, V10 of 35%,V15 of 30%, V20 of 28%, V25 of 24%, V30 of 19%, V35 of 16%, and a V40 of 12%). Local control, distant metastasis, and survival At the last follow-up visit, 24 patients were still alive (13 patients in Group I and 11 patients in Group II). Among the survivors there were 14 patients with local control and without evidence of intrahepatic and distant metastases (7 patients in Group I and 7 patients in Group II) and 10 patients whose outcomes were associated with failures, including in-field failures alone (2 patients in Group II), out-of-field failures alone (4 patients in Group I and 1 patients in Group II), both in-field and out-of-field failures (2 patients in Group I), and distant metastasis (1 patients in Group II). Sixteen patients died of out-of-field failure (4 patients in Group I and 7 patients in Group II), distant metastases (1 patients in Group II), both out-of-field and distant metastasis (3 patients in
Radiation dose escalation study for HCC d Z.-G. REN et al.
499
Table 1. Clinical characteristics in 40 hepatocellular carcinomas Patient characteristics Sex Female Male Age (yr) Median KPS 70 80 90 Diagnosis by Cytology Histology Clinical Serum AFP (ng/ml) #20 20-400 $400 HBV infection Positive Negative TACE Median course Median maximum tumor diameter (cm) GTV (cc) Median AJCC stage T1N0M0 (I) T2N0M0 (II) T3N0M0 (IIIA) T4N0M0 (IIIB)
Group I (<10-cm tumors) (range)
Group II ($10-cm tumors) (range)
Total
0 20
4 16
4 36
51 (36–68)
55.5 (33–68)
3 16 1
2 15 3
5 31 4
1 18 1
3 10 7
4 28 8
7 5 8
6 5 9
13 10 17
19 1
18 2
37 3
2 (1–3) 7.5 (5.1–9.5)
2 (1–4) 13.6 (10.5–16.4)
2 (1–4) 10.5 (5.1–16.4)
209 (57–474)
819 (506–1336)
514 (57–1336)
0 15 5 0
0 5 12 3
0 20 17 3
52.5 (33–68)
Abbreviations: AFP = alfa-fetoprotein; KPS = Karnofsky performance status; HBV = hepatitis B virus.
Group I), or RILD in 1 patient. Overall, there were 4 patients with in-field failures, 21 out-of-field failures, and 5 distant metastases (Table 5). For the entire 40 patients, IFPFS was 100% at 1 year and 93% at 2 years, and 1-year and 2-year LPFS rates were 69% and 44%, respectively. DM rates were 6% and 15% at 1 year and 2 years, respectively. MST was not reached because more than half of the patients were still alive in Group I and at 22 months (95% confidence interval, 11-33 months) in Group II. The 1-year and 2-year OS rates were 72% and 62%, respectively (Fig 1). DISCUSSION The fractionation and total doses of radiation therapy for HCC varied considerably in a series of publications. Dawson et al. (6) treated 43 patients with intrahepatic hepatobiliary cancer and liver metastases by 3DCRT with a median total dose of 58.5 Gy at 1.5 Gy per fraction twice daily and found that the escalated RT dose was independently associated with improved outcome. Also, we learned from a previous study of HCC that a higher dose yielded a better outcome, with an MST of 38 months and 12 months and a 3-year OS rate of 45% and 17%, respectively, for patients receiving $45 Gy and <45 Gy (8). Thus, we believed that higher irradiation
doses would be needed to produce better local controls. However, a high dose of hepatic irradiation would injure the liver, leading to RILD, which was almost a fatal complication (11). Therefore, we conducted the current clinical phase I trial to escalate irradiation dose and to determine what would be the MTD when irradiation was administered after TACE for HCC patients. When designing the protocol, we had to answer the following questions: What was the dose fractionation, hypofractionation or conventional? What was the preplanned highest dose level? Which irradiation technique should be used, 3DCRT or IMRT? With regard to fractionation, based on our previous study (7), hypofractionation by 5 Gy/fraction given every other day really produced a 3-year OS rate of 33%, but in 19 (19/128 [15%]) patients, RILD occurred, and 16 of 19 patients died as a result. Based on radiobiology studies, the liver is considered a late-responding normal organ with a low a/ b value (12, 13). Theoretically, conventional fractionation would lessen hepatic injury compared to the large fraction size. Therefore, we decided to use conventional fractionation in the current dose escalation study. The preplanned highest doses for escalation were 52 Gy and 62 Gy, respectively, for tumor sizes of <10 cm and $10 cm, which was determined by our experience with
I. J. Radiation Oncology d Biology d Physics
500
Table 2. Median dosimetric parameters of normal liver Parameter Uninvolved normal liver volume (cm3)* MDTNL (Gy)y V5 (%)z V10 (%) V15 (%) V20 (%) V25 (%) V30 (%) V35 (%) V40 (%) Probability of RILD (%)x
Group I (<10 cm)
Group II ($10 cm)
3
3
1,248 cm 1,042 cm (range){ (range)** 16 (9–22) 18 (15–20) 55 (27–80) 58 (41–75) 45 (21–66) 50 (35–66) 38 (18–58) 39 (34–55) 33 (15–52) 32 (30–45) 28 (11–48) 28 (23–28) 22 (9–30) 24 (18–26) 19 (7–28) 19 (17–23) 13 (5–26) 17 (12–19) 2.2 (0.6–6.9) 1.9 (0.6–8.7)
p value <0.05 >0.05 >0.05 >0.05 >0.05 >0.05 >0.05 >0.05 >0.05 >0.05
* Uninvolved normal liver volume = total liver volume – planning target volume. y MDTNL (mean dose to normal liver) was derived using the formula, normal liver volume = total liver volume – planning target volume. z V5 = percentage of normal liver volume which received $5 Gy in total liver volume, and in V10-V40, the suffix indicates the dose ($) received by normal liver. x The RILD (radiation-induced liver disease) was estimated by an in-house Lyman model (n = 1.1, m = 0.28, TD50 [1] = 40.5 Gy). TD50 [1], the whole liver dose, which will produce 50% probability of toxicity; ‘‘m,’’characterizing the steepness of the dose response at TD50; and ‘‘n,’’ a volume effect parameter that indicates a larger volume effect as it increases (range, 0-1). { Range, 792 to 1,639 cm3. ** Range, 805 to 1,921 cm3.
a dosimetric study (unpublished data). That study demonstrated that for the tumor located close to the GI tract, the highest dose we could deliver was 50 Gy while keeping doses to the GI tract at tolerable levels. However, if the tumor location was distant from the GI tract, the highest doses that could be given were 52 Gy and 62 Gy for tumors of <10 cm and $10 cm, respectively, while both the GI tract dose and the liver dose constraints could be met. IMRT has been widely used for prostate and head and neck cancers but not for HCC. From our (unpublished) data and from that of other dosimetric studies (14–16), when comparing IMRT and 3DCRT for HCC, we found that tumor dose conformality by IMRT was superior to that by 3DCRT and that IMRT also could reduce MDTNL. In the current study, for patients with a tumor of $10 cm when the total dose was more than 48 Gy, or for those patients with a tumor of
Volume 79, Number 2, 2011
<10 cm but a tumor dose of more than 58 Gy, IMRT had to be used in 75% and 88% of patients, respectively, to satisfy the dose constraints of OAR. Otherwise, such high doses were impossible to deliver. However, to implement IMRT, we realized that IMRT was more sensitive to liver motion due to respiration (17), and hence, it was not appropriate to a moving target because of dose uncertainty. To eliminate liver motion, we used an ABC device to immobilize liver lesions and to make delivery of IMRT possible. In our recent report, the systematic and random set-up errors and intrafraction reproducibility of the diaphragm were in a range of 1.2 mm to 1.8 mm (10). Although, there was still some uncertainty of irradiation delivery, we thought that it was acceptable. Moreover, we used a simplified IMRT technique, which limited the number of subfields to #5 and the smallest subfield size to $5 cm 5 cm. Using these measurements, we could probably reduce the negative impact of IMRT, which was imposed by set-up error or liver reproducibility by a ABC. In spite of the fact that IMRT provided a better plan, we did not use IMRT for all patients because IMRT planning was time consuming. We needed more manpower and time to design, verify, and deliver IMRT plans, which in our very busy department, we could not afford. Therefore, for HCC, we generally use 3DCRT except when the desired dose constraints could not be met. In the current study, the acute toxicity was acceptable, with 11 cases of grade 1 hepatic toxicity and no case of a grade $3 GI toxicity in both groups. Unfortunately, one patient developed RILD and died as a result. Although in this case the radiation dose to the normal liver was limited to a tolerable level, we thought that the RILD was due to the patient’s hepatitis B infection-induced cirrhosis, and in addition, the patient’s associated diabetes mellitus probably made the hepatic tolerance worse. Considering that RILD was almost a fatal complication with 84% mortality in our previous series (7), prevention is critical. The best way to avoid RILD is to limit the hepatic dose to a tolerable level. From our previous analysis (11, 18), we found that tolerance to radiation therapy was much poorer in Chinese HCC patients than in metastatic liver cancer patients. This is because in over 90% of our patients, HCC developed on the basis of hepatitis B infection-induced cirrhosis, which impaired the liver’s capability to repair irradiation damage and to proliferate. According to our previous analysis of HCC treated by hypofractionation, we proposed
Table 3. Mean dosimetric parameters of stomach and duodenumy Stomach
Duodenum
Dose
Group I (<10 cm)
Group II ($10 cm)
Group I (<10 cm)
Group II ($10 cm)
Mean SD (Gy)* Max SD (Gy)*
5.4 6.4 27.0 20.4
9.9 10.5 31.3 17.7
10.6 12.1 24.0 20.0
14.4 13.7 28.2 19.6
* Mean values standard deviations (SD). p > 0.05 for all comparisons between the values of Group I and Group II.
y
Radiation dose escalation study for HCC d Z.-G. REN et al.
501
Table 4. Patient data and treatment-related toxicity in 40 hepatocellular carcinomas Dose level Group I (<10 cm) 46 Gy 50 Gy 54 Gy 58 Gy 62 Gy Group II ($10 cm) 40 Gy 44 Gy 48 Gy 52 Gy
No. of patients
No. of patients with DLT
No. of patients with non-DLT toxicity
4 4 4 4 4
None None None None None
None GI Gr 1 (1), H Gr 1 (1) GI Gr 1 (1), H Gr 1 and GI Gr 1 (1) GI Gr 1 (1), H Gr 1 and GI 1 (1), GI Gr 2 (1) GI Gr 1 (1), H Gr 1 and GI Gr 1 (2), GI Gr 2 (1)
4 4 4 8
None None None H Gr 5 (1)
None GI Gr 1 (1), H Gr 1 and GI Gr 1 (1) GI Gr 1 (1), H Gr 1 and GI Gr 1 (2) H Gr 1 and GI Gr 1 (3), GI Gr 2 (2)
Abbreviations: H = hepatic toxicity; Gr = grade.
that the liver’s tolerance for HCC associated with a liver function of Child-Pugh A was 23 Gy in terms of MDTNL and a tolerable DVH (11). In this study, we used conventional fractionation but chose hepatic dose constraints, which resulted from our hypofractionation study (7). The reason was based on the following considerations. From the hypofractionation study, we learned that a liver with hepatic cirrhosis tolerated irradiation more poorly than liver without cirrhosis. However, we really did not know exactly what the hepatic tolerance dose was for patients with cirrhosis, if treated by conventional fractionation, although we expected it would be higher than that by hypofractionation. In order to totally avoid RILD, we decided to use a stricter dose constraint. In the literature, Robertson et al. (19) used a V50% of normal liver to determine the total tumor dose. Because we were not convinced that this guideline was suitable for our patients with cirrhosis backgrounds and also because the purpose of this study was to escalate dose and obtain MTD, we did not apply the dose prescription of Robertson et al. In the current trial, we estimated that patients would be able to tolerate the dose with no severe hepatic injury. The RILD incidence of 2.5% (1/40 patients) in this study demonstrated the validity of our proposal for hepatic irradiation tolerance in HCC patients. Another measurement we exercised in irradiation planning was to limit the portal number to #5 to avoid a large volume of normal liver receiving low dose and to always have a part of the normal liver excluded from any beam. We expected that the protected normal liver would proliferate to compensate for the loss of hepatic function
during irradiation, which was demonstrated by our animal experiment (20). From the current study, we concluded that MTDs were 52 Gy and 62 Gy for patients with tumors of $10 cm and <10 cm in diameter, respectively. One question raised from our study was why patients with big tumors tolerated irradiation more poorly than those with small tumors while dose constraints were satisfied in both patients. As shown in Table 3, both mean doses and maximum doses to the stomach and duodenum were higher in patients with tumors of $10 cm than those with tumors of <10 cm, although the differences were not statistically significant. Also, there were no differences in MDTNL and DVH of normal liver in both groups, Therefore, for both groups, doses to liver, stomach, and duodenum were similar; however, the total doses already delivered to tumors were 40 Gy to 52 Gy in $10-cm tumors and 46 Gy to 62 Gy in <10-cm tumors; and OAR would not be able to tolerate doses increased to more than 52 Gy to $10-cm tumors. Moreover, there was 1,042 cm3 of uninvolved normal liver in patients with tumors of $10 cm, whereas there was 1,248 cm3 in those with tumors of <10 cm (p < 0.05) (Table 2). Thus, we thought that a smaller, uninvolved normal liver volume and probably higher doses to
Table 5. Failure patterns in 40 cases of HCC* Failure
No. of patients
% of patients
In field only Out-of-field only In field and out-of-field Out-of-field and distant Distant only
2 16 2 3 2
8 64 8 12 8
* Data indicate patterns of 25 failures in 40 patients with HCC after receiving a combination of transcatheter arterial chemoembolization and irradiation.
Fig. 1. Overall survival rates in 40 patients with hepatocellular carcinomas after receiving a combination of transarterial chemoembolization and irradiation.
502
I. J. Radiation Oncology d Biology d Physics
OARs would be the obstacle to further escalating the dose to patients with big tumors. As to the outcome of this study, IFPFS rates were 100% and 93%, and LPFS rates were 69% and 45%, respectively, at 1 year and 2 years, which implied that tumor control inside the field was excellent but the intrahepatic failure out-of-field was predominantly the failure pattern in 21 patients, which accounted for 84% of failures. We believe that it is attributed to the HCC biological behavior, i.e., the tendency to spread along the portal vein or multiple primary foci. Further efforts are needed to deal with the intrahepatic failures. DM was not a major cause of death in this trial with low DM rates (6% and 15% at 1 year and 2 years, respectively). The OS rates of 72% at 1 year and 62% at 2 years were encouraging and were greatly improved compared to OS rates of 60% at 1 year and 38% at 2 years in our latest publication (8). Moreover, more advanced tumors (median diameter of 10.5 cm) were encountered in the current study, whereas in
Volume 79, Number 2, 2011
the previous study, tumor diameter was 6.5 cm. It was obvious that the high tumor dose contributed to the improved survival (median tumor dose of 51 Gy in the current study vs. 43 Gy in the previous study [8]). Thus, our proposal is to deliver a dose as high as possible to HCC while keeping doses to OAR at tolerable levels. However, we thought that the patient selection could have contributed to the good outcome. Patients in this trial might have less potential to metastasize since only patients without distant metastases had been enrolled. CONCLUSIONS In summary, for HCC patients with Child-Pugh A liver function, MTD was 52 Gy and 62 Gy in patients with tumors of $10 cm and <10 cm in diameter, respectively, after TACE, which yielded an improved outcome and acceptable toxicity.
REFERENCES 1. Hawkins MA, Dawson LA. Radiation therapy for hepatocellular carcinoma: from palliation to cure. Cancer 2006;106:1653– 1663. 2. Cheng SH, Lin YM, Chuang VP, et al. A pilot study of three-dimensional conformal radiotherapy in unresectable hepatocellular carcinoma. J Gastroenterol Hepatol 1999;14:1025–1033. 3. Park W, Lim DH, Paik SW, et al. Local radiotherapy for patients with unresectable hepatocellular carcinoma. Int J Radiat Oncol Biol Phys 2005;61:1143–1150. 4. Park HC, Seong J, Han KH, et al. Dose-response relationship in local radiotherapy for hepatocellular carcinoma. Int J Radiat Oncol Biol Phys 2002;54:150–155. 5. Seong J, Park HC, Han KH, et al. Clinical results and prognostic factors in radiotherapy for unresectable hepatocellular carcinoma: A retrospective study of 158 patients. Int J Radiat Oncol Biol Phys 2003;55:329–336. 6. Dawson LA, McGinn CJ, Normolle D, et al. Escalated focal liver radiation and concurrent hepatic artery fluorodeoxyuridine for unresectable intrahepatic malignancies. J Clin Oncol 2000; 18:2210–2218. 7. Liang SX, Zhu XD, Lu HJ, et al. Hypofractionated three-dimensional conformal radiation therapy for primary liver carcinoma. Cancer 2005;103:2181–2188. 8. Zhou ZH, Liu LM, Chen WW, et al. Combined therapy of transcatheter arterial chemoembolization and three-dimensional conformal radiotherapy for hepatocellular carcinoma. Br J Radiol 2007;80:194–201. 9. Yang BH. Primary liver cancer. In: ChineseAnti-Cancer Association, editor. Guidelines for common cancers in China. Beijing: Associated Press of Peking Medical University and Peking Union Medical College; 1999. p. 389–479. 10. Zhao JD, Xu ZY, Zhu J, et al. Application of active breathing control in 3-dimensional conformal radiation therapy for hepatocellular carcinoma: The feasibility and benefit. Radiother Oncol 2008;87:439–444. 11. Liang SX, Zhu XD, Xu ZY, et al. Radiation-induced liver disease in three-dimensional conformal radiation therapy for
12. 13. 14.
15.
16.
17.
18.
19. 20.
primary liver carcinoma: The risk factors and hepatic radiation tolerance. Int J Radiat Oncol Biol Phys 2006;65:426–434. Fisher DR, Hendry JH. Dose fractionation and hepatocyte clonogens: Alpha/beta congruent to 1-2 Gy, and beta decreases with increasing delay before assay. Radiat Res 1988;113:51–57. Dawson LA, Lawrence TS, Ten Haken RK. In response to Dr. Tome and Dr. Fenwick. Int J Radiat Oncol Biol Phys 2004;58:1319–1320. Cheng JC, Wu JK, Huang CM, et al. Dosimetric analysis and comparison of three-dimensional conformal radiotherapy and intensity-modulated radiation therapy for patients with hepatocellular carcinoma and radiation-induced liver disease. Int J Radiat Oncol Biol Phys 2003;56:229–234. Thomas E, Chapet O, Kessler ML, et al. Benefit of using biologic parameters (EUD and NTCP) in IMRT optimization for treatment of intrahepatic tumors. Int J Radiat Oncol Biol Phys 2005;62:571–578. Eccles CL, Bissonnette JP, Craig T, et al. Treatment planning study to determine potential benefit of intensity-modulated radiotherapy versus conformal radiotherapy for unresectable hepatic malignancies. Int J Radiat Oncol Biol Phys 2008;72: 582–588. Fuss M, Salter BJ, Herman TS, et al. External beam radiation therapy for hepatocellular carcinoma: Potential of intensitymodulated and image-guided radiation therapy. Gastroenterology 2004;127:S206–S217. Xu ZY, Liang SX, Zhu J, et al. Prediction of radiation-induced liver disease by Lyman normal-tissue complication probability model in three-dimensional conformal radiation therapy for primary liver carcinoma. Int J Radiat Oncol Biol Phys 2006;65: 189–195. Robertson JM, Lawrence TS, Dworzanin LM, et al. Treatment of primary hepatobiliary cancers with conformal radiation therapy and regional chemotherapy. J Clin Oncol 1993;11:1286–1293. Zhao JD, Jiang GL, Hu WG, et al. Hepatocyte regeneration after partial liver irradiation in rats. Exp Toxicol Pathol 2009;61:511– 518.