Japanese Journal of Clinical Oncology 32:455-460 (2002)
© 2002 Foundation for Promotion of Cancer Research
Prognostic Factors in Patients with Hepatocellular Carcinoma Treated by Transcatheter Arterial Embolization
Divisions of 1 Hepatobiliary and Pancreatic Oncology and 4 Diagnostic Radiology, National Cancer Center Hospital, Tokyo, 2 Cancer Information and Epidemiology Division, National Cancer Center Research Institute, Tokyo and 3 Department of Epidemiology and Biostatistics, Kyoto University School of Public Health, Kyoto, Japan
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Background: Transcatheter arterial embolization induces marked antitumor response in patients with hepatocellular carcinoma, but the survival benefit of transcatheter arterial embolization remains to be determined. This study investigated prognostic factors in patients with advanced hepatocellular carcinoma treated by transcatheter arterial embolization.
Methods: A total of 128 consecutive patients with non-resectable hepatocellular carcinoma, who had undergone transcatheter arterial embolization between May 1990 and August 1998, were analyzed to investigate prognostic factors.
Results: Median survival time and survival proportions at 1, 3 and 5 years were 3.3 years, 92.0, 54.6 and 23.4%, respectively. By multivariate analysis using the accelerated failure time model, age <60 years, hepatitis C virus antibody positivity, serum albumin >3.5 g/dl, absence of portal vein invasion and serum 
-fetoprotein level <400 ng/ml were significantly associated with favorable survival. For clinical application, we also propose a prognostic equation with combination of specific prognostic factors, by which survival curves of each patient could be predicted directly.
Conclusion: The findings of the current study may be helpful in predicting the life expectancy of hepatocellular carcinoma patients treated by transcatheter arterial embolization and in designing future clinical trials of transcatheter arterial embolization for hepatocellular carcinoma.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Transcatheter arterial embolization (TAE) is a widely used treatment for advanced hepatocellular carcinoma (HCC), because TAE induces a marked antitumor effect in HCC. Furthermore, TAE can be administered for any type of HCC, regardless of size, location or number of tumors, if hepatic reserves are sufficient (1). However, the survival benefit of TAE in patients with advanced HCC remains to be determined. Four of five randomized controlled trials failed to show a survival benefit for advanced HCC, although one trial showed a benefit (2–6). Moreover, optimum TAE procedures and proper patient selection for TAE have not yet been fully clarified. If more suitable TAE methods and more appropriate candidates for TAE can be elucidated, TAE may be able to demonstrate better results in advanced HCC. The present study was conducted to evaluate a number of variables that may affect survival in patients with advanced HCC treated by TAE. The analysis of prognostic factors may be helpful in predicting the life expectancy of HCC patients treated by TAE and in designing future clinical trials of TAE for HCC.
|
PATIENTS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Patients
The study subjects were 128 consecutive advanced HCC patients with no indications for surgical resection or local ablation therapy such as percutaneous ethanol injection, who had undergone TAE between May 1990 and August 1998 at the National Cancer Center Hospital, Tokyo, Japan. They had not received any prior treatments for HCC except hepatic resection. Diagnosis of HCC was made by either histological examination (74 patients, 58%) or typical computed tomographic (CT) and/or angiographic findings (54 patients, 42%). Patients with extrahepatic metastasis or tumor thrombus in the main portal trunk were excluded in this study. Before TAE, a chest X-ray, ultrasonography and CT of the abdomen were performed to exclude the presence of extrahepatic metastasis and bone metastasis was ruled out by bone scintigraphy if the patients had symptoms such as pain. None of the patients received any other treatments for HCC during the follow-up period.
Methods
TAE was performed by selectively introducing a catheter into the proper, right, left hepatic artery or a segmental branch of the hepatic artery and injecting a mixture of an iodized oil (Lipiodol; Guerbet, Paris, France) and anti-cancer agents followed by 0.5–1 mm fragments of a gelatin sponge block (Gelfoam; Upjohn, Kalamazoo, MI). Doxorubicin 30 mg (Adriamycin; Adria Laboratories, Dublin, OH), mitomycin 8 mg (Mitomycin-C; Kyowa Hakko Kogyo, Tokyo, Japan) plus doxorubicin 30 mg or zinostatin stimalamer 4 mg (SMANCS; Yamanouchi, Tokyo, Japan) were used as anti-cancer agents.
Follow-up
The antitumor effects of TAE were assessed by CT 1 month after treatment and subsequent follow-up examinations including CT, -fetoprotein (AFP) and biochemistry were performed at least every 3 months. TAE was performed when relapse of the treated lesions and/or new hepatic lesions were seen.
Factors Analyzed
Pretreatment clinical variables were investigated for their relationship to survival by the univariate and multivariate analysis. The pretreatment variables were chosen by considering possible effects on the prognosis as indicated by previous investigations (7–24) or suggested from our own clinical experience. Each variable, which was classified as host-related, tumor-related and treatment-related, was divided into two subgroups. The host-related variables were as follows: age (<60 or 
60 years); gender (male or female); alcohol abuse, ethanol intake 80 g/day for 5 years (positive or negative); hepatitis B surface antigen (positive or negative); hepatitis C virus antibody (HCVAb) (positive or negative); prior hepatectomy (positive or negative); serum total bilirubin (<2.0 or 2.0 mg/dl); serum albumin (
3.5 or >3.5 g/dl); glutamic oxaloacetic transaminase [82 IU/l (2.5 times upper normal limit) or >82 IU/l]; glutamic pyruvic transaminase [70 IU/l (2.5 times upper normal limit) or >70 IU/l]; serum lactic dehydrogenase (<500 or 500 IU/l); and cholinesterase (normal limit: 155–391 IU/l) (<100 or 100 IU/l). The tumor-related variables were as follows: number of tumors (single or multiple); tumor distribution (unilateral or bilateral); tumor size (25 or >25%); portal vein (PV) invasion (positive or negative); and serum AFP level (<400 or 400 ng/ml). The treatment-related variable was the anti-cancer agents used (doxorubicin, mitomycin + doxorubicin or zinostatin stimalamer). The number of tumors, tumor distribution and portal vein invasion were examined by ultrasonography, CT and/or angiography. In this study, PV invasion indicated tumor thrombus in the PV except for the main portal trunk. Tumor size was estimated on CT from the total tumor area relative to the entire liver.
Statistical Analysis
Survival time was defined as the time from the date of the first TAE to date of death or the last date of follow-up. Survival curves were estimated by the Kaplan–Meier method (25). Then, 95% confidence intervals (95% CI) for the survival proportion were calculated using standard errors based on Greenwood’s equation (26).
To evaluate the influence of the potential prognostic factors on survival, median survival times were estimated by the Kaplan–Meier method for each factor as a univariate analysis. In addition, accelerated failure time model (AFTM) regression analysis with Weibull distribution (26) was used for univariate and multivariate analysis. In AFTM, survival times are modeled as a function of prognostic factors so that survival times can be predicted for patients with combination of specific prognostic factors. The effect of potential prognostic factors can be evaluated from the ratio of survival time. This is more directly interpretable than the hazard ratio that is estimated in the Cox model (27). All variables considered in univariate analysis were entered in the multivariate model and variable selection was not conducted.
Furthermore, predicted survival curves for all patients were estimated by AFTM. The final model for prediction included variables with P < 0.05 by the downward variable selection procedure. All analyses were conducted by SAS (SAS Institute, Cary, NC). All P-values quoted are two-sided.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Patients’ Characteristics
Patients’ characteristics are shown in Table 1. Of the 128 patients with HCC, 102 were men and 26 women (3.9:1). The median age was 63 years (range:36–82 years). Hepatitis B surface antigen and HCVAb were positive in 20 (16%) and 102 (80%), respectively, and 62 patients (48%) had a history of hepatic resection. Ninety patients (70%) had multiple tumors and tumor size >25% was found in 18 patients (14%). Median serum AFP level was 24 (range: 2–201 000) ng/ml. Eighty-seven patients (68%) had received TAE using doxorubicin, 17 patients (13%) using mitomycin + doxorubicin and 24 patients (19%) using zinostatin stimalamer. There was a median of two TAE procedures (range: 1–11) during the follow-up period.
|
Survival
The median survival time, 1 year, 3 year and 5 year survival proportions in all 128 patients were 3.3 years (95% CI: 2.9–3.9 years), 92.0% (95% CI: 87.2–96.7%), 54.6% (95% CI: 44.6–64.6%) and 23.4% (95% CI: 13.9–32.9%), respectively (Fig. 1). At the time of analysis, 74 patients had died due to tumor progression and/or hepatic decompensation and one patient had died due to acute hepatic failure related to TAE 1 month after treatment.
|
Univariate and Multivariate Analysis
Median survival times and ratios of survival time for univariate analysis are shown in Table 2. As for host-related factors, serum albumin
3.5 g/dl and cholinesterase <100 IU/l were associated with more than 30% shorter survival times. As for tumor-related factors, multiple tumors, bilateral tumor distribution, tumor size >25% and serum AFP level 400 ng/dl were associated with more than 30% shorter survival times. There was no substantial difference in survival times among patients with the three chemotherapeutic regimens (doxorubicin, mitomycin + doxorubicin and zinostatin stimalamer). Results of multivariate analysis are also shown in Table 2. Age 60 years, HCV Ab negativity, serum albumin level 3.5 g/dl, presence of PV invasion and serum AFP level 400 ng/ml were associated with more than 30% shorter survival times.
|
Prediction
The prediction equation of survival probability at time t for patients with a set of specific prognostic factors is as follows:
S(t) = exp(–exp{2[log(t) – (7.65 – 0.29 x Age + 0.38 x HCV Ab – 0.45 x Albumin – 0.65 x AFP – 0.43 x tumor distribution)]})
where Age = 1 when age 60 years, HCVAb = 1 when HCVAb shows positivity, Albumin = 1 when serum albumin 3.5 g/dl, AFP = 1 when AFP 400 ng/ml and tumor distribution = 1 when tumors distribute bilaterally, otherwise each variable is counted as 0. The Weibull shape parameter was 0.50 (95% CI: 0.41–0.59), which indicates that the death rate (= hazard) for patients treated by TAE increased with time. Using this equation, predicted survival curves can be calculated. As an example, Fig. 2 shows the predicted survival curves for patients with combination of specific prognostic factors. The median survival time was also calculated for patients by solving the equation with S(t) = 0.5.
|
We showed that age, HCVAb, serum albumin, AFP and PV invasion were associated with more than 30% shorter survival times in multivariate models. In contrast, the prediction equation for patient survival times included a different set of variables, that is, age, HCVAb, serum albumin, AFP and tumor distribution. This often happens in statistical analysis. In the former model, all potential prognostic factors were included without variable selection, since the objective of the model is to estimate each factor’s effect while adjusting for the effects of all the other variables. In the latter model, the model was constructed to predict each patient’s survival curve by the smallest number of variables with the smallest error. The latter model did not contradict the former and each model indicated potentially large effects of tumor distribution and PV invasion on patient survival.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
TAE has been widely used for non-resectable HCC and is currently the mainstay of non-surgical treatments for HCC. However, the efficacy of TAE on survival in patients with advanced HCC is still debated, because four of five randomized controlled trials (2–6) comparing TAE versus no treatment, hormonal therapy or chemotherapy failed to show survival improvement in patients with advanced HCC. Various reasons for the negative results in survival were mentioned as follows: TAE may impair the baseline prognosis because of deleterious effects on liver function or early relapse seen in most patients may counteract the potential benefits of initial response to TAE (1). However, the most important problem is that the appropriate candidates for TAE and optimal TAE methods have not yet been fully clarified and this affected our evaluation of the survival benefits of TAE.
The general indications for TAE are HCC patients with adequate hepatic reserve, no extrahepatic metastasis and no tumor thrombus in the main portal trunk, who are contraindicated for surgical resection or local ablation therapy. This study was carried out to evaluate factors affecting survival in 128 patients who met these indications. This single-institution study was undertaken with the same unified methods for tumor staging and identical procedures for treatment, follow-up and supportive care throughout, thus enabling us to confirm important prognostic factors, predict life expectancy and design future clinical trials of TAE for advanced HCC.
In this study, prognostic factors were analyzed by multivariate analysis using AFTM, but not the Cox proportional hazard model. Cox regression analysis showed similar results in terms of P-value and corresponding hazard ratio (data not shown). Since AFTM could provide more information such as predicted survival curves and ratio of survival times, it is more appropriate for the study to investigate the influence of potential prognostic factors on survival times if the model shows a good fit to the data and the assumption holds, as in this study.
Concerning host-related factors, age 60 years, HCV Ab negativity and serum albumin level 3.5 g/dl were unfavorable prognostic factors in the multivariate analysis (7–16). With regard to age, this result was consistent with some previous reports demonstrating that advanced age is an adverse prognostic factor in HCC patients regardless of tumor stage and therapeutic methods (7–10). Serum albumin is one of the most important factors in evaluating hepatic reserve (11–16) and it was therefore included in the Okuda staging system (15) and Child–Pugh classification (16) The finding that HCVAb positivity was one of the factors that contributed to a better outcome in this study was unexpected. This was partly due to residual confounding, because HCVAb positivity was strongly associated with favorable tumor-related factors, such as smaller tumor size and tumor number (data not shown).
Of tumor-related factors, serum AFP level and PV invasion were identified as significantly associated with shorter survival times in the multivariate analysis and these findings were compatible with previous reports (11–13,17–24). Serum AFP level is one of the most useful tumor markers as a prognostic factor in HCC patients (11–13,17–23). The level of AFP has been reported to be correlated with tumor burden (21). It has also been reported that poorly differentiated HCC produce more AFP (22,23). Therefore, higher AFP levels may reflect a more advanced stage and greater malignant potential of tumors. PV invasion, which is also reflection of the biologic aggressiveness of the tumor (11,12,24), is presumed to result in intrahepatic metastasis because it often occurs via the PV.
For clinical application of these findings, we proposed a prediction equation that can be used to predict the survival curve of each patient directly. These predicted survival curves can be easily calculated, because they are based on variables obtained during routine clinical examinations. This equation, therefore, can be used to stratify patients with advanced HCC before TAE according to predicted survival. Several reports have been published regarding staging systems in HCC that appear to have prognostic significance (7,8,10–12,15,17). However, as far as we know, this is the first report concerning a prediction equation designed to assess the prognosis of individual HCC patients treated by TAE. These findings may be helpful in predicting life expectancy in HCC patients treated by TAE and furnish more information for stratifying patients, when future TAE trials are designed.
|
FOOTNOTES |
|---|
+ For reprints and all correspondence: Masafumi Ikeda, Hepatobiliary and Pancreatic Oncology Division, National Cancer Center Hospital, 5–1–1 Tsukiji, Chuo-ku, Tokyo 104-0045, Japan. E-mail: sokada@ncc.go.jp
Abbreviations: TAE, transcatheter arterial embolization; HCC, hepatocellular carcinoma; CT, computed tomography; AFP, -fetoprotein; HCV Ab, hepatitis C virus antibody; PV, portal vein; AFTM, accelerated failure time model
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
1 Okada S. Transcatheter arterial embolization for advanced hepatocellular carcinoma: the controversy continues. Hepatology 1998;27:1743–4.[Web of Science][Medline]
2 Pelletier G, Ducreux M, Gay F, Luboinski M, Hagege H, Dao T, et al. A randomized trial of hepatic arterial chemoembolization in patients with unresectable hepatocellular carcinoma. J Hepatol 1990;11:181–4.[Web of Science][Medline]
3 Groupe d’Etude et de Traitement du Carcinome Hepatocellulaire. A comparison of lipiodol chemoembolization and conservative treatment for unresectable hepatocellular carcinoma. N Engl J Med 1995;332:1294–6.
4 Pelletier G, Ducreux M, Gay F, Luboinski M, Hagege H, Dao T, et al. Treatment of unresectable hepatocellular carcinoma with lipiodol chemoembolization: a multicenter randomized trial. J Hepatol 1998;29:129–34.[Web of Science][Medline]
5 Bruix J, Llovet JM, Castells A, Montana X, Bru C, Ayuso MC, et al. Transarterial embolization versus symptomatic treatment in patients with advanced hepatocellular carcinoma: results of a randomized, controlled trial in a single institution. Hepatology 1998;27:1578–83.[Web of Science][Medline]
6 Lin DY, Liaw YF, Lee TY, Lai CM. Hepatic arterial embolization in patients with unresectable hepatocellular carcinoma – a randomized controlled trial. Gastroenterology 1988;94:453–6.[Web of Science][Medline]
7 Mondazzi L, Bottelli R, Brambilla G, Rampoldi A, Rezakovic I, Zavaglia C, et al. Transarterial oily chemoembolization for the treatment of hepatocellular carcinoma: a multivariate analysis of prognostic factors. Hepatology 1994;19:1115–23.[Web of Science][Medline]
8 Calvet X, Bruix J, Gines P, Bru C, Sole M, Vilana R, et al. Prognostic factors of hepatocellular carcinoma in the West: a multivariate analysis in 206 patients. Hepatology 1990;12:753–60.[Web of Science][Medline]
9 Falkson G, Cnaan A, Schutt AJ, Ryan LM, Falkson HC. Prognostic factors for survival in hepatocellular carcinoma. Cancer Res 1988;48:7314–8.
10 Okada S, Okazaki N, Nose H, Yoshimori M, Aoki K. Prognostic factors in patients with hepatocellular carcinoma receiving systemic chemotherapy. Hepatology 1992;16:112–7.[Web of Science][Medline]
11 The Cancer of the Liver Italian Program (CLIP) Investigators. A new prognostic system for hepatocellular carcinoma: a retrospective study of 435 patients. Hepatology 1998;28:751–5.[Web of Science][Medline]
12 Llado L, Virgili J, Figueras J, Valls C, Dominguez J, Rafecas A, et al. A prognostic index of the survival of patients with unresectable hepatocellular carcinoma after transcatheter arterial chemoembolization. Cancer 2000;88:50–7.[Web of Science][Medline]
13 Farinati F, De Maria N, Marafin C, Herszenyi L, Del Prato S, Rinaldi M, et al. Unresectable hepatocellular carcinoma in cirrhosis: survival, prognostic factors and unexpected side effects after transcatheter arterial chemoembolization. Dig Dis Sci 1996;41:2332–9.[Web of Science][Medline]
14 Shijo H, Okazaki M, Higashihara H, Koganemaru F, Okumura M. Hepatocellular carcinoma: a multivariate analysis of prognostic features in patients treated with hepatic arterial embolization. Am J Gastroenterol 1992;87:1154–9.[Web of Science][Medline]
15 Okuda K, Ohtsuki T, Obata H, Tomimatsu M, Okazaki N, Hasegawa H, et al. Natural history of hepatocellular carcinoma and prognosis in relation to treatment. Study of 850 patients. Cancer 1985;56:918–28.[Web of Science][Medline]
16 Pugh RN, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 1973;60:646–9.[Web of Science][Medline]
17 Akashi Y, Koreeda C, Enomoto S, Uchiyama S, Mizuno T, Shiozaki Y, et al. Prognosis of unresectable hepatocellular carcinoma: an evaluation based on multivariate analysis of 90 cases. Hepatology 1991;14:262–8.[Web of Science][Medline]
18 Hatanaka Y, Yamashita Y, Takahashi M, Koga Y, Saito R, Nakashima K, et al. Unresectable hepatocellular carcinoma: analysis of prognostic factors in transcatheter management. Radiology 1995;195:747–52.
19 Taniguchi K, Nakata K, Kato Y, Sato Y, Hamasaki K, Tsuruta S, et al. Treatment of hepatocellular carcinoma with transcatheter arterial embolization. Analysis of prognostic factors. Cancer 1994;73:1341–5.[Web of Science][Medline]
20 Savastano S, Miotto D, Casarrubea G, Teso S, Chiesura-Corona M, Feltrin GP. Transcatheter arterial chemoembolization for hepatocellular carcinoma in patients with Child’s grade A or B cirrhosis: a multivariate analysis of prognostic factors. J Clin Gastroenterol 1999;28:334–40.[Web of Science][Medline]
21 Nomura F, Ohnishi K, Tanabe Y. Clinical features and prognosis of hepatocellular carcinoma with reference to serum alpha-fetoprotein levels: analysis of 606 patients. Cancer 1989;64:1700–7.[Web of Science][Medline]
22 Kondo F, Wada K, Nagato Y, Nakajima T, Kondo Y, Hirooka N, et al. Biopsy diagnosis of well-differentiated hepatocellular carcinoma based on new morphologic criteria. Hepatology 1989;9:751–5.[Web of Science][Medline]
23 Brumm C, Schulze C, Charels K, Morohoshi T, Kloppel G. The significance of alpha-fetoprotein and other tumour markers in differential immunocytochemistry of primary liver tumours. Histopathology 1989;14:503–13.[Web of Science][Medline]
24 Ikeda K, Kumada H, Saitoh S, Arase Y, Chayama K. Effect of repeated transcatheter arterial embolization on the survival time in patients with hepatocellular carcinoma. An analysis by the Cox proportional hazard model. Cancer 1991;68:2150–4.[Web of Science][Medline]
25 Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457–81.[Web of Science]
26 Kalbfleish JD, Prentice RL. The Statistical Analysis of Failure Time Data. New York: Wiley 1980.
27 Cox DR. Regression models and life tables [with discussion]. J R Stat Soc Ser B 1972;34:187–220.
Received June 18, 2002; accepted July 25, 2002
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C. M. Pacella, G. Francica, F. M. L. Di Lascio, V. Arienti, E. Antico, B. Caspani, F. Magnolfi, A. S. Megna, S. Pretolani, R. Regine, et al. Long-Term Outcome of Cirrhotic Patients With Early Hepatocellular Carcinoma Treated With Ultrasound-Guided Percutaneous Laser Ablation: A Retrospective Analysis J. Clin. Oncol., June 1, 2009; 27(16): 2615 - 2621. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Buijs, J. A. Vossen, C. Frangakis, K. Hong, C. S. Georgiades, Y. Chen, E. Liapi, and J.-F. H. Geschwind Nonresectable Hepatocellular Carcinoma: Long-term Toxicity in Patients Treated with Transarterial Chemoembolization--Single-Center Experience Radiology, October 1, 2008; 249(1): 346 - 354. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||