D. Y. Kim and W. Park contributed equally to the current study.
Three-dimensional conformal radiotherapy for portal vein thrombosis of hepatocellular carcinoma
Version of Record online: 8 APR 2005
Copyright © 2005 American Cancer Society
Volume 103, Issue 11, pages 2419–2426, 1 June 2005
How to Cite
Kim, D. Y., Park, W., Lim, D. H., Lee, J. H., Yoo, B. C., Paik, S. W., Kho, K. C., Kim, T. H., Ahn, Y. C. and Huh, S. J. (2005), Three-dimensional conformal radiotherapy for portal vein thrombosis of hepatocellular carcinoma. Cancer, 103: 2419–2426. doi: 10.1002/cncr.21043
- Issue online: 18 MAY 2005
- Version of Record online: 8 APR 2005
- Manuscript Accepted: 25 JAN 2005
- Manuscript Revised: 5 JAN 2005
- Manuscript Received: 8 DEC 2004
- hepatocellular carcinoma;
- portal vein thrombosis;
- dose-response relation
Portal vein thrombosis (PVT) is a common complication in patients with advanced-stage hepatocellular carcinoma (HCC). The authors evaluated the impact of radiotherapy (RT) for PVT of HCC and analyzed the dose-response relation between RT and PVT.
Between March 1995 and December 2003, 59 patients diagnosed as HCC with PVT were included. The inclusion criteria were unresectable tumor with thrombosis in the main or first branch of the portal vein, liver function of Child–Pugh Class A or B, and an Eastern Cooperative Oncology Group performance status score of 0–2. The median age of the patients was 57 years (range, 36–78 years). A daily dose ranging from 2 to 3 gray (Gy) was administered using 6 or 10-megavolt (MV) X-rays, at 5 fractions a week, to deliver a total dose range of 30–54 Gy, which was a biologic effective dose of 39–70.2 Gy10 with an α/β ratio of 10.
Follow-up computed tomography scans showed a complete response (CR) in 4 of 59 patients (6.8%), a partial response (PR) in 23 patients (39.0%), no response (NR) in 28 patients (47.5%), and progressive disease (PD) in 4 patients (6.8%). The mean RT doses in the responders (CR and PR) and nonresponders (NR and PD) were 59.6 ± 5.6 Gy10 and 54.9 ± 8.5 Gy10, respectively (P = 0.036). The response rates in patients receiving < 58 Gy10 and ≥ 58 Gy10 were 20% and 54.6%, respectively (P = 0.034). The median survival duration and the 1-year and 2-year survival rates in the responders were 10.7 months, 40.7%, and 20.7%, respectively, and were 5.3 months, 25.0%, and 4.7%, respectively, in the nonresponders (P = 0.050).
RT induced a 45.8% objective response rate for PVT in patients with HCC. A dose-response relation was found to exist between the RT dose and PVT response. These results suggested that RT may be a treatment option for PVT in patients with HCC and that an RT dose ≥ 58 Gy10 should be recommended. Cancer 2005. © 2005 American Cancer Society.
Although portal vein thrombosis (PVT) is considered rare, occurring in only 1–5.7% of patients with cirrhosis,1 it is a common complication in patients with advanced-stage hepatocellular carcinoma (HCC), occurring in 20–70% of these patients.2, 3 PVT is a consequence of portal vein invasion by the tumor and causes serious results, such as portal vein hypertension, hematemesis from rupture of collateral vessels, ascites, and ischemic liver damage. Consequently, this is an extremely poor prognostic condition. Standard treatment regimens have not been established for patients with HCC with PVT.
Some investigators4–7 have noted that transcatheter arterial chemoembolization (TACE) was contraindicated for patients with HCC with PVT, because it had a potential risk of ischemic liver damage. On the contrary, other authors8 suggested that TACE might be safely performed in patients with PVT if they have good hepatic reserve or collateral circulation around the main portal vein. However, this result indicates that TACE has a limited role for patients with HCC with PVT and new strategies are needed.
Radiotherapy (RT) for HCC has been infrequently used in the treatment of HCC because the liver has a low tolerance for whole-organ irradiation.9, 10 However, some authors reported that the tolerance dose for the liver depended significantly on the volume of liver irradiated.11 Although the tolerance dose for whole-organ irradiation is low, a small volume of liver tissue can tolerate a higher dose of RT without serious hepatic problem. In addition, several recent reports12–16 have demonstrated that partial hepatic RT is feasible and showed promising responses for unresectable patients with HCC with or without PVT. With advances in three-dimensional (3-D) planning tools, 3-D conformal radiotherapy (3D-CRT) allows clinicians to escalate RT doses to the tumor and minimize RT doses to normal tissues, such as the normal parenchyma of liver, small bowel, and spinal cord. However, there is no consensus regarding RT for patients with HCC with PVT.
We have previously reported the results of 3D-CRT for unresectable HCC.17 Our preliminary results, as well as those of others,18, 19 suggested that a higher dose of local RT was tolerable and resulted in a higher response rate for locally advanced-stage HCC. However, the dose-response relation between the RT dose and PVT has not been established. The aims of our study were, therefore, to evaluate the impact of RT for PVT in patients with HCC and to analyze the dose-response relation between RT and PVT.
MATERIALS AND METHODS
The current study included 59 patients diagnosed with HCC with PVT at the Samsung Medical Center (Seoul, Korea) between March 1995 and December 2003. A diagnosis of HCC was based on 1 of the following 3 criteria: 1) histologic confirmation (n = 8); 2) a serum α-fetoprotein (AFP) level > 400 IU/mL, a radiologically compatible feature with HCC in ≥ 1 computed tomography scans (CT)/magnetic resonance imaging scans (MRI)/angiograms, and the presence of risk factors including hepatitis B virus (HBV) or hepatitis C virus (HCV) infection and cirrhosis (n = 33); and 3) a serum AFP level < 400 IU/mL, a radiologically compatible feature with HCC in ≥ 2 CT/MRI/angiograms, and the presence of risk factors including HBV or HCV infection and cirrhosis (n = 18).
The inclusion criteria were unresectable tumor with thrombus in the main or first branch of the portal vein, liver function of Child–Pugh Class A or B, and an Eastern Cooperative Oncology Group (ECOG) performance status score of 0–2. Our study was performed in accordance with the guidelines of our institutional review board. Informed consent was obtained from all patients. PVT was demonstrated using helical CT scans with contrast enhancement, MRI scans, or angiography. On contrast-enhanced CT scans, PVT was identified by the presence of a low-attenuation intraluminal filling defect adjacent to the primary tumor.
Patient characteristics are shown in Table 1. The median age of the patients was 57 years (range, 36–78 years). The median size of the tumor plus the PVT was 11 cm (range, 6–22 cm). In terms of previous treatments before RT, 35 patients had received TACE, 1 patient had received percutaneous ethanol injection therapy, 5 patients had received radiofrequency ablation, and 23 patients had not received any previous treatment.
|Characteristics||No. of patients(%)|
|≤ 60||39 (66.1)|
|> 60||20 (33.9)|
|No treatment||23 (39.0)|
|ECOG performance status|
|< 400||22 (37.3)|
|≥ 400||37 (62.7)|
|Tumor size (cm)a|
|≤ 8||10 (16.9)|
|≥ 12||23 (39.0)|
Patients underwent CT scans in a supine position for RT planning, with both arms raised above the head to facilitate use of lateral RT ports. CT scan data were transferred to a 3D RT treatment planning system (PROWESS, Alliant Medical Technology, Chico, CA). The primary tumor in the liver parenchyma, normal liver tissue, kidneys, spinal cord, and portal vein of each patient were contoured and reconstructed three dimensionally. The RT volumes and treatment angles were designed using a 3D view technique to minimize critical organ injury. A daily dose of 2–3 gray (Gy) was administered using 6 or 10 MV X-rays, at 5 fractions per week, to deliver a total dose of 30–54 Gy, which translates to a biologic effective dose (BED) of 39–70.2 Gy10 as the α/β ratio = 10. BED was calculated using a linear quadratic model to be equivalent to 2–3-Gy fraction treatments in respect of acute effects on normal tissues and tumors.20
In CT scan planning, the clinical target volume (CTV) was determined with a 1–1.5-cm margin of the primary mass and the thrombus, and the planning target volume had expanded by 0.5 cm of the CTV. The extra margin (1–1.5 cm) in the craniocaudal direction was added to cover the respiratory liver motion. 3D-CRT planning was designed under tentative guidelines so that the normal liver volume irradiated with more than one-half of the prescription dose should not be > 50% of the total liver volume. All patients were asked to perform shallow respiration to minimize respiratory motion.
PVT response was determined using serial CT scans 4–8 weeks after completion of treatment, and then every 2–3 months. Complete disappearance of PVT was defined as a complete response (CR), a > 50% reduction of thrombus in the greatest cross-sectional area was defined as a partial response (PR), a < 50% reduction of thrombus (including no change) was defined as no response (NR), and tumor growth was defined as progressive disease (PD). The objective response rates were calculated for the rate of CR or PR. Responders were patients with a CR or PR, whereas nonresponders were patients with an NR or PD.
Acute morbidity was evaluated weekly during treatment and 1 month after the treatment. Late morbidity was defined as occurring after 3 months. RT-induced liver disease (RILD) was defined as anicteric ascites and elevation of alkaline phosphatase levels > 2 times the pretreatment values in the absence of PD.21 If patients complained of melena or upper abdominal pain persisting for > 2 weeks during the follow-up period, suspected gastrointestinal (GI) toxicity was investigated by fiberoptic gastroduodenoscopy. GI toxicity was assessed using Common Terminology Criteria for Adverse Events, version 3.0.
Univariate logistic regression analysis was used to evaluate the association between PVT response and various parameters. Gender (male vs. female), ECOG performance status (0–1 vs. 2), presence of cirrhosis (no vs. yes), liver function of Child–Pugh class (Class A vs. Class B), serum level of AFP (< 400 IU/mL vs. ≥ 400 IU/mL), and the response of PVT (CR and PR vs. NR and PD) were considered as binary variables. Age, tumor size, and RT dose were analyzed as continuous variables. For multivariate analysis to evaluate the relation between the PVT response and various parameters, the stepwise procedure was performed using a logistic regression model containing all variables that attained or had a trend toward univariate statistical significance. The Fisher exact test was performed to evaluate the relation between primary tumor response and PVT response, as well as to evaluate the relation between RT dose and PVT response. Overall survival was calculated from the first date of RT. The probability of survival was calculated using the Kaplan–Meier method. The log-rank test was used to evaluate the prognostic influence of the PVT response. The statistical test was a two-sided test and was performed with SAS programs (version 8.01, SAS Institute Inc., Cary, NC). P ≤ 0.05 was statistically significant.
Response of Portal Vein Thrombosis
Of the 59 patients, a CR was observed in 4 (6.8%) patients, a PR in 23 patients (39.0%), NR in 28 patients (47.5%), and PD in 4 patients (6.8%). The objective response (CR and PR) rate was 45.8%. Figure 1 shows the CR of 1 of the 4 patients with a CR.
The results of univariate and multivariate analyses to evaluate the relation between PVT response and various parameters are summarized in Table 2. RT dose (BED) was the only significant parameter that predicted the response of PVT in univariate and multivariate analysis (P = 0.015 and P = 0.042, respectively). The mean (± standard deviation) RT doses in responders and nonresponders were 59.9 (± 5.6) Gy10 and 55.2 (± 8.6) Gy10, respectively. The other parameters, such as age, gender, performance status, cirrhosis, liver function, AFP level, and tumor size, had no significant association with the response of PVT. The objective response rates of PVT as a function of RT dose are summarized in Table 3. The objective response of PVT was observed in 3 of 15 patients (20%) with BED < 58 Gy10 and in 24 of 44 patients (54.6%) with BED ≥ 58 Gy10 (P < 0.034). Figure 2 shows the relation between primary tumor response and PVT response. The objective response rates of PVT were significantly higher in patients with a CR or PR compared with patients with a NR or PD of primary tumor (P < 0.001).
|Characteristics||No. of responders (n = 27)||No. of nonresponders (n = 32)||P valuea|
|Age (yrs, mean ± SD)||56.5 ± 8.7||56.3 ± 9.3||0.941||0.749|
|ECOG performance status|
|Tumor size (cm, mean ± SD)||11.5 ± 3.4||11.7 ± 4.3||0.789||0.472|
|BED (Gy10, mean ± SD)||59.9 ± 5.6||55.2 ± 8.6||0.015||0.042|
|PVT response||No. of RT doses (BED) (%)||P valuea|
|< 58 Gy10||≥ 58 Gy10|
|Responders (CR+PR)||3 (20)||24 (54.6)||0.034|
|Non-responders (NR+PD)||12 (80)||20 (45.4)|
At the time of analysis, 54 patients had died of disease and 5 patients were alive. The median follow-up period for all patients was 6 months (range, 2–45 months), and for living patients, 33 months (range, 17–45 months). In responders, the median survival duration was 10.7 months and the 1-year and 2-year overall survival rates were 40.7% and 20.7%, respectively. In nonresponders, the median survival duration was 5.3 months and the 1-year and 2-year survival rates were 25.0% and 4.7%, respectively (Fig. 3). The responders had a significantly higher overall survival rate than the nonresponders (P = 0.050).
No patient displayed Grade > 3 acute toxicity. For GI complications, 6 patients were scored as Grade 2 resulting from gastric or duodenal ulcers inside the RT field. The gastroduodenal toxicity was observed in 1 of 15 (6.7%) patients with BED < 58 Gy10 and in 5 of 44 patients (11.4%) with BED ≥ 58 Gy10 (P = 1.000). Of the 59 patients, RILD was observed in 0 of 15 patients (0%) with BED < 58 Gy10 and in 1 of 44 patients (2.3%) with BED ≥ 58 Gy10 (P = 0.100). Anicteric ascites without elevation of alkaline phosphatase levels developed in 12 patients and successfully resolved with or without diuretics.
PVT in HCC can lead not only to widespread dissemination of tumors throughout the liver, but also to marked deterioration of hepatic function. In previous reports,22, 23 the presence of PVT is the clinicopathologic parameter that most influences survival in multivariate analysis, and the median survival time of patients with HCC with PVT is approximately 2 months without any treatment. Furthermore, it is often the obstacle to treatment options.
Surgical resection has been used to treat patients with HCC with PVT.23–25 Kumada et al.25 reported 1-year and 2-year survival rates of 58% and 39% for for 13 patients with HCC with PVT, and suggested the prolongation of life span and quality of life (QOL). However, surgical treatment can only be performed for highly selected patients, because there is a potential risk of postoperative liver failure and early disease recurrence. Therefore, alternative treatments for patients with HCC with PVT have been investigated. TACE has been clinically considered to be contraindicated in patients with HCC with main portal trunk obstruction, because it could theoretically result in ischemic damage to normal liver unaffected by tumor.26 However, Lee et al.8 performed TACE in 31 patients with HCC with main portal vein obstruction, and reported that the rate of progressive hepatic insufficiency was not increased compared with an untreated group of patients. They demonstrated that TACE prolonged survival in patients with nodular HCC but did not increase survival in patients with diffuse HCC. Another treatment option, i.e., arterial infusional chemotherapy, has been investigated.27, 28 Sakon et al.27 reported the results of combined intraarterial 5-fluorouracil (5-FU) and subcutaneous interferon therapy in 11 patients with PVT. In 8 of these 11 patients, an objective response was observed with marked regression of the tumor and a decrease in tumor markers. Ando et al.28 reported good responses in 48 patients with PVT who were treated with hepatic arterial infusion chemotherapy consisting of low-dose cisplatin and 5-FU. The response rate was 48%, and the median survival times for responders and nonresponders were 31.6 and 5.4 months, respectively. However, these methods had technical problems associated with indwelling catheters and catheter-related sepsis. To date, standard treatment regimens have not been established for patients with HCC with PVT.
RT for the treatment of patients with HCC has been limited to palliation because of the low tolerance of the liver for RT. Some authors reported that the tolerance dose to the whole liver was 30–35 Gy.9, 10 However, this dose seems to be insufficient to control HCC.29, 30 Lawrence et al.11 demonstrated that the tolerance dose for 70% of the liver is approximately 42 Gy, 52 Gy for 50% of the liver, and approximately 70 Gy for 20% of the liver. Recently, therefore, local rather than whole-liver RT has been investigated for the treatment of HCC.12–14 Cheng et al.13 reported a median survival duration of 19 months, and a 2-year survival rate of 41% after local radiotherapy with TACE treatment of 25 patients with unresectable HCC. Seong et al.14 treated 30 patients with local RT starting 7–10 days after TACE and reported a median survival rate of 17 months, and a 2-year survival rate of 33.3%. Liu et al.31 reported that 3D-CRT in 44 unresectable patients with HCC, who had either failed with or were unsuitable for TACE, resulted in an objective response rate of 61.4%. They also noted that stage, PVT, pretreatment AFP level, and RT dose were significant prognostic factors for these patients. Furthermore, RT was reported to have some effect not only on the main tumor but also on the PVT complication. Cheng et al.15 reported that RT in 7 patients with HCC with PVT resulted in a median survival of 5 months (range, 2–15 months), suggesting that 3D-CRT should be considered as an alternative treatment option for patients with HCC with PVT. Tazawa et al.16 reported a retrospective study of combined therapy in 24 patients with HCC with PVT. The combined therapy with RT of 50 Gy in 2-Gy fractions and TACE showed a 50% response rate. They also reported that the survival rate was significantly better in the responders compared with the nonresponders. Ishikura et al.12 reported that 20 patients receiving combined therapy for PVT had a 1-year survival rate of 25% and a median survival time of 5.3 months. All of these studies suggested that RT might become the treatment of choice for PVT in patients with HCC. Contrary to these findings, Yamada et al.32 reported a prospective trial comprising 19 patients with HCC with PVT treated with a combination of 3D-CRT and TACE. They noted that there was no significant difference in cumulative overall survival between responders and nonresponders. However, our study demonstrated that the objective response rate was 45.8% and that the responders had a significantly higher overall survival rate than the nonresponders (P = 0.050).
In the advent of 3D planning systems, 3D-CRT allowed us to minimize irradiation of normal liver tissue, and therefore facilitates escalation of the RT dose without a significant increment in toxicity. In addition, detailed dose-volumetric information is available, which enables clinicians to study the dose-response relations for HCC. Robertson et al.33, 34 suggested that a dose-response relation might exist for HCC because hepatic control within the target volume was particularly encouraging in the high-dose local RT group. In contrast, their experience with whole-liver RT indicated there was tumor progression within the RT field in a substantial proportion of patients. Park et al.18 found that the RT dose was the most significant factor associated with tumor response in an analysis of 158 patients with HCC. The response rates in patients treated with doses < 40 Gy, 40–50 Gy, and > 50 Gy were 29.2%, 68.6%, and 77.1%, respectively. Similarly, Guo and Yu19 showed that higher irradiation doses resulted in higher survival rates in 107 patients with HCC. These data indicate that increasing the RT dose has important consequences for HCC. Although there have been studies reporting dose-response relations between RT and HCC, there have been no reports on such a relation between RT and PVT. In our study, the objective response of PVT was observed in 3 of 15 patients (20%) with BED < 58 Gy10 and in 24 of 44 patients (54.6%) with BED ≥ 58 Gy10 (P < 0.034). We also showed that overall survival was significantly higher in the responders compared with the nonresponders (P = 0.050). Although further studies are needed due to our relatively small study populations, our results supported the dose-response relation between RT and PVT in patients with HCC and demonstrated that the response of PVT was a significant prognostic predictor of patients with HCC with PVT.
Escalation of RT dose is frequently accompanied by increased toxicity. Park et al.18 reported that the rates of liver toxicity linked to doses of < 40 Gy, 40–50 Gy, and > 50 Gy were 4.2%, 5.9%, and 8.4%, respectively, and that the rates of GI complications were 4.2%, 9.9%, and 13.2%, respectively. In our study, gastroduodenal toxicity and RILD were observed in 1 of 15 patients (6.7%) patients and in 0 of 15 patients (0%) with BED < 58 Gy10, and in 5 of 44 patients (11.4%) and in 1 of 44 patients (2.3%) with BED ≥ 58 Gy10. Although the rate of GI toxicity and RILD increased as the RT dose increased, there was no statistical significance, which is believed to be primarily due to small sample size, the relatively low incidence of toxicity, and the disparity in numbers between patients who received RT dose < 58 Gy10 (n = 44) and ≥ 58 Gy10 (n = 15). In addition, a cirrhotic liver is frequently associated with HCC and with more severe RILD compared with a noncirrhotic liver, but there is little information in the literature that recommends clear treatment guidelines for clinical decisions regarding treatment of patients with HCC with cirrhosis. Thus, it is important to integrate an assessment of liver function into the RT treatment planning process for patients with cirrhosis. To minimize RILD in our study, a tentative guideline was followed, i.e., that the normal liver volume irradiated at more than one-half the prescription dose should not exceed one-half the total volume of the normal liver. RILD was observed in 1 of 59 patients (1.7%) in our study. However, larger studies and careful analysis of dose-volume histograms are needed to minimize RT-induced complications and to establish the treatment guidelines of RT for patients with HCC with PVT.
The current study has some limitations. First, because our study population was treated with heterogeneous dose-fractionated 3D-CRT and the sample size was relatively small, various clinical parameters and dose fractionation schedules, which could influence survival of patients with HCC with PVT, could not be thoroughly explored. The fractionated RT dose might have influenced the PVT response to RT due to the wide range of biologic properties of HCC, but a standard dose fractionation schedule has not been established for RT for patients with HCC with PVT until now. Therefore, further studies are needed to evaluate the standard dose fractionation schedule. Second, the current study evaluated the impact of RT for PVT in patients with HCC and analyzed the dose-response relation between RT and PVT. We did not analyze the outcomes, such as survival and QOL, between the two groups of patients who were treated with or without RT. Although our results suggest that the 3D-CRT for patients with HCC with PVT is promising in terms of response, further randomized studies are needed to evaluate the impact of 3D-CRT on the survival and QOL of patients with HCC with PVT.
The current study evaluated the effect of RT and analyzed the dose-response relation between RT and PVT for patients with HCC with PVT receiving 3D-CRT. RT induced a 45.8% objective response rate for PVT in patients with HCC. We also identified a dose-response relation between the RT dose and PVT response. The objective response of PVT was observed in 3 of 15 patients (20%) with BED < 58 Gy10 and in 24 of 44 patients (54.6%) with BED ≥ 58 Gy10. However, although further studies are warranted to establish the standard treatment regimens for patients with HCC with PVT, our data suggest that 3D-CRT may be a treatment option for patients with HCC with PVT and an RT dose ≥ 58 Gy10 should be recommended.
- 7Risk factors of early recurrence after curative hepatic resection for hepatocellular carcinoma. J Korean Cancer Assoc. 1999; 31: 165–172., , , et al.
- 8The safety and efficacy of transcatheter arterial chemoembolization in the treatment of patients with hepatocellular carcinoma and main portal vein obstruction. A prospective controlled study. Cancer. 1997; 79: 2087–2094., , , , , .
- 17Dose-response relationship of radiotherapy for locally advanced hepatocellular carcinoma. J Korean Cancer Assoc. 2000; 32: 918–924., , , et al.