INTRODUCTION. Jpn J Clin Oncol 1997;27(4)

Transcription

1 Jpn J Clin Oncol 1997;27(4) Therapeutic Results of Resection, Transcatheter Arterial Embolization and Percutaneous Transhepatic Ethanol Injection in 3225 Patients with Hepatocellular Carcinoma: a Retrospective Multicenter Study,,,,,,,, From the Departments of 1 Surgery, 2 Internal Medicine and 3 Radiology, National Cancer Center Hospital East, Kashiwa, Chiba, Japan, and 4 Department of Environmental and Occupational Health, Toho University School of Medicine, Tokyo, Japan The outcome in 3225 patients with hepatocellular carcinoma (HCC) was studied in groups with equivalent prognosis treated with resection, transcatheter arterial embolization (TAE), and percutaneous transhepatic ethanol injection (PEI). Significant factors for better clinical background included a tumor diameter of 30 mm, tumor number 3 and (clinical) Stage I. In patients with Stage I disease having tumors of 30 mm and 3 in number, survival after resection and PEI did not differ, while survival after TAE was significantly worse. In those patients with Stage II disease, survival after PEI was significantly better than after resection or TAE. In patients with Stage I or II disease having tumors 31 in size and 3 in number, survival after resection was significantly better than after TAE. In patients with Stage I disease having tumors of 31 mm and 4 in number, survival after resection was significantly better than after TAE. Our conclusions are as follows. Firstly, resection or PEI is recommended for patients with Stage I disease having 3 tumors all 30 mm in size. Secondly, PEI is recommended for patients with Stage II disease having 3 tumors all 30 mm in size. Thirdly, for patients with Stage I disease having tumors 31 mm or larger in size, whatever the number of tumors, resection should be selected rather than TAE. Key words: hepatocellular carcinoma hepatectomy embolization, therapeutic percutaneous transhepatic ethanol injection INTRODUCTION The most commonly used treatments for hepatocellular carcinoma (HCC) are resection, transcatheter arterial embolization (TAE) and percutaneous transhepatic ethanol injection (PEI). Since 80% of hepatocellular carcinomas are associated with liver cirrhosis, the cancer must be treated while liver function is maintained. Treatment methods should be selected on the basis of liver function and neoplastic involvement. The criteria and methods by which specific treatments are selected, should be clarified. Resection Received September 26, 1996; accepted February 13, 1997 For reprints and all correspondence: Shigeru Imoto, Department of Surgery, National Cancer Center Hospital East, 5 1, Kashiwanoha 6-chome, Kashiwa, Chiba 277, Japan Abbreviations: AFP, α-fetoprotein; HCC, hepatocellular carcinoma; ICGR15, indocyanine-green retention at 15 min; PEI, percutaneous transhepatic ethanol injection; TAE, transcatheter arterial embolization. is the most consistently effective local treatment. Intraoperative ultrasound examination enables resection to be performed safely and reliably (1 2). Reported survival rates after resection range from 39% to 51% at 3 years and from 25% to 49% at 5 years (3 11). Another effective treatment for HCC is TAE. The liver receives a double blood supply from the portal vein and hepatic artery. Since the cancer lesion receives only an arterial blood supply, embolization of the hepatic artery causes selective necrosis of the cancer lesion. Yamada et al. (12) performed TAE in 789 patients with unresectable HCC and obtained a 5-year survival rate of 6%. In 66 patients with tumors <50 mm in size, however, significantly better survival rates were achieved by TAE. In PEI, ethanol is injected into the tumor under ultrasonic or computed tomographic guidance to induce coagulation necrosis. Ebara et al. (13) treated 95 patients with HCC <30 mm in size; he reported a 5-year survival rate of 28%. Many authors have reported that PEI is an effective treatment that could replace surgical therapy (14 17).

2 252 Hepatocellular carcinoma therapy Table 1. Clinical stage * Factor Stage I Stage II Stage III Ascites Negative Well controlled Poorly controlled Total bilirubin (mg/dl) Albumin (g/dl) > <3.0 ICG R15 (%) < >40 Prothrombin aetivity (%) > <50 * According to the Liver Cancer Study Group in Japan (6). Although the effectiveness of these various treatments has been reported, the criteria by which specific treatments are selected have not been clearly established. For this reason, a retrospective multicenter study was undertaken to compare survival of patients who had undergone resection, TAE or PEI for the treatment of HCC. MATERIALS AND METHODS SUBJECTS AND CLINICAL MATERIAL As part of a multicenter study, all patients with HCC treated from 1985 to 1990 in 18 institutes were reviewed. In all patients, the diagnosis of HCC was made on the basis of histologic or angiographic findings combined with a rise in the serum levels of α-fetoprotein (AFP). The subjects were 3225 patients comprised of 2647 men and 585 women. The patients ranged in age from 10 to 87 years, with a mean ± SD of 59.9 ± 8.8 years (mean ± SD). Resection was performed in 1656, PEI in 302, TAE in 1067, chemotherapy in 186 and radiation in 14. The survey was accomplished in September PROGNOSTIC FACTORS Factors thought to influence survival were selected and then classified to obtain survival rates using the Kaplan Meier method. The significance of the differences was evaluated by the log-rank test with univariate analysis. The following prognostic factors were tested: age, sex, performance status (PS), tumor size, tumor number, serum levels of bilirubin and albumin, indocyanine-green retention at 15 min (ICGR15), prothrombin activity, clinical stage (Table 1) according to the Liver Cancer Study Group in Japan (6), AFP, presence of portal vein or hepatic vein tumor thrombus, and ascites. With each factor related to the survival rate used as variables, stepwise multiplevariate analysis was performed. Multiple regression analysis based on the Cox proportional-hazard model was applied to calculate the relative-risk ratio between each factor and the survival rate. METHODS Patients with HCC were classified on the basis of these results into groups with similar factors to evaluate the survival rate with each treatment. Since serum levels of bilirubin and albumin are included in the factors used for clinical stage, liver function was evaluated on the basis of the clinical stage, and the survival rates were statistically analyzed by the log-rank test and the generalized Wilcoxon test. Recurrence was diagnosed by two different diagnostic modalities. Figure 1. Survival by clinical stage in 3162 patients. ( Stage is clinical stage.) RESULTS SURVIVAL BY THERAPEUTIC METHOD The 3- and 5-year survival rates after resection (62.2% and 41.3%) and PEI (69.3% and 46.1%, respectively) did not differ significantly. However, the survival rate after TAE (25.6% and 9.4%, respectively) was significantly lower than those of either the resection or PEI (P < 0.01). The survival rate after chemotherapy (6.8% and 4.5%) and radiation (7.1% and 0%) was significantly worse than after the other treatments. The background of the patients who underwent resection, PEI, TAE and other treatments are shown in Table 2. While there are clearly significant differences in clinical stage among four different treatment-groups, survival should be evaluated with the same prognostic factors. UNIVARIATE ANALYSIS OF PROGNOSTIC FACTORS Age and sex were not significantly related with prognosis. Factors that were associated with increased survival included a performance status of 0 1, tumor size 30 mm, tumor number 3, absence of ascites, serum bilirubin 2 mg/dl, serum albumin >3.5 g/dl, ICGRl5 <15%, prothrombin activity >80%, clinical Stage I, AFP <399 ng/ml and absence of tumor embolism in the portal vein or hepatic vein. MULTIVARIATE ANALYSIS OF PROGNOSTIC FACTORS (TABLE 3) Factors that were particularly significant were clinical stage, tumor number and tumor size. The total number of patients for each item varied because of missing data. SURVIVAL BY CLINICAL STAGE (Fig. 1) Survival was poorer with more advanced clinical stage. In patients with Stage I disease, the 3- and 5-year survival rates were 59.7% and 39.9%, respectively. In patients with Stage II disease, the 3- and 5-year survival rates were 38.9% and 22.6% and with Stage III disease, these rates were 24.7% and 8.9%, (P < 0.01).

3 Jpn J Clin Oncol 1997;27(4) 253 Table 2. Background of the patients Resection TAE PEI Others P-value Patients (n) Age (mean ± SD) ± ± ± 7.53 NS Sex (male/female) 1356/ / /70 168/32 NS Stage (I/II/III) 1033/490/54 372/543/ /153/23 19/108/73 <0.05 TAE, transcatheter arterial embolization; PEI, percutaneous ethanol injection; NS, not significant. Figure 2. Survival by number of tumors in 3021 patients. Figure 3. Survival by tumor size in 2902 patients. Table 3. Multivariate analysis of prognostic factors for survival time in 3225 patients with HCC using Cox s proportional-hazard model Variables Hazard ratio (95% CI) P-value Clinical stage ( ) Tumor number ( ) Tumor size (mm) ( ) HCC, hepatocellular carcinoma; CI, confidence interval. SURVIVAL BY NUMBER OF TUMORS (Fig. 2) In patients with 3 tumors, the 3- and 5-year survival rates were 55.3% and 35.5%, respectively. These rates were 19.6% and 7.8% in patients with 4 10 tumors, and 16.9% and 6.3% with 11 tumors. Survival was significantly poorer in patients with 4 tumors than in those with 3 (P < 0.01). SURVIVAL BY TUMOR SIZE (Fig. 3) Survival was significantly poorer in patients with tumors 31 mm or larger in size than with tumors 30 mm or smaller (P < 0.01). SURVIVAL OF STAGE I PATIENTS HAVING 3 TUMORS, ALL 30 MM (Fig 4) The 3- and 5-year survival rates for patients who underwent resection (77.9% and 54.7%, respectively) and PEI (83.7% and 53.8%) were not significantly different. However, the survival rates in patients treated by TAE (46.8% and 16.9%) were Figure 4. Survival of patients with Stage I disease having tumors 30 mm in size and 3 in number. RESC, resection; PEI, percutaneous ethanol injection; TAE, transcatheter arterial embolization. significantly poorer than those in patients who underwent either resection or PEI (P < 0.01). The 3- and 5-year survival rates for patients in whom only one portion was resected (78% and 61.7%, respectively) and those in whom two or three portions were resected (82.1% and 50.5%) were not significantly different. DISEASE-FREE SURVIVAL OF STAGE I PATIENTS WITH 3 TUMORS, ALL 30 MM (Fig. 5) The 3- and 5-year disease-free survival rates in patients who underwent resection (42.7% and 25.9%, respectively) and in those who underwent PEI (39.5% and 12.5%) were not significantly different. The disease-free survival rates in patients treated by

4 254 Hepatocellular carcinoma therapy Figure 5. Disease-free survival in patients with Stage I disease having tumors 30 mm in size and 3 in number. For abbreviations, see legend to Fig. 4. Figure 7. Disease-free survival in patients having Stage II disease with tumors 30 mm in size and 3 in number. For abbreviations, see legend to Fig. 4. Figure 6. Survival in patients with Stage II disease having tumors 30 mm in size and 3 in number. For abbreviations, see legend to Fig. 4. Figure 8. Survival in patients with Stage I disease having tumors 31 mm in size and <3 in number. For abbreviations, see legend to Fig. 4. TAE (16.9% and 0%) were significantly poorer than those who underwent either resection or PEI (P < 0.01). SURVIVAL IN STAGE II PATIENTS WITH 3 TUMORS, ALL 30 MM (Fig. 6) The 3- and 5-year survival rates in patients who underwent PEI (65.3% and 45.2%) were significantly better than in those who underwent resection (59.1% and 36.1%) (P < 0.05) and those who underwent TAE (35.0% and 14.9%) (P < 0.01). DISEASE-FREE SURVIVAL IN STAGE II PATIENTS WITH 3 TUMORS, ALL 30 MM (Fig. 7) The 3- and 5-year disease-free survival rates in patients who underwent resection (30.6% and 12.3%), in those who underwent PEI (29.9% and 6.6% ) and in patients who underwent TAE (18.2% and 4.8%) were not significantly different. SURVIVAL IN STAGE I PATIENTS WITH 4 TUMORS, ALL 30 MM IN SIZE Because only 21 patients underwent resection and 16 underwent TAE, statistical evaluation was not possible. SURVIVAL IN STAGE I PATIENTS WITH 3 TUMORS, SOME 31 MM (Fig. 8) The 3-and 5-year survival rates in patients who underwent resection (62.5% and 45%) were significantly better than in those who had TAE (31.2% and 20.5%) (P < 0.01). The 5-year survival rates in 144 patients who underwent a one-portion resection (42%) and 26 patients who underwent a two-portion resection (36%) were not significantly different. SURVIVAL IN STAGE II PATIENTS WITH 3 TUMORS, SOME 31 MM (Fig. 9) The 3- and 5-year survival rates in patients with Stage II disease who underwent resection (58.4% and 40.1%) were significantly better than in those who underwent TAE (26.8 % and 10.2%) (P < 0.01). SURVIVAL IN STAGE I PATIENTS WITH 4 TUMORS, SOME 31 MM (Fig. 10) The 3- and 5-year survival rates in patients who underwent resection (31.2% and 17.3%) were significantly better than in those who had TAE (18.1% and 4%) (P < 0.05).

5 Jpn J Clin Oncol 1997;27(4) 255 Figure 9. Survival in patients with Stage II disease having tumors 31 mm in size and <3 in number. For abbreviations, see legend to Fig. 4. Figure 11. Survival rate in patients with Stage II disease having tumors 31 mm in size and >4 in number. For abbreviations, see legend to Fig. 4. DISCUSSION Figure 10. Survival in patients with Stage I disease having tumors 31 mm in size and >4 in number. For abbreviations, see legend to Fig. 4. SURVIVAL IN STAGE II PATIENTS WITH 4 TUMORS, SOME 31 MM (Fig. 11) The 1- and 3-year survival rates in patients who underwent resection (58.6% and 37.5%) and in patients who had TAE (38.4% and 0%) were not significantly different. OPERATIVE MORTALITY RATE BY STAGE IN PATIENTS WITH RESECTION The 30-days and 90-days mortality rates in patients with Stage II disease was significantly higher than those with Stage I disease (P < 0.05). Table 4. Operative mortality rate Clinical stage Patients No. who died No. who died (at operation) (n) within 30 days within 90 days Stage I (1 4%) 40 (3 9%) Stage II (3.3%) 43 (8.8%) Significance P < 0.05 P < 0.01 Totals (1.96%) 83 (5.45%) Although various treatments are used for HCC, their therapeutic results have only rarely been compared. Ohnishi et al. (18) compared TAE and surgery and reported a median survival of 35 months for resection and 28.8 months for TAE. Yoshimi et al. (19) compared the results of resection in 66 patients with those of TAE in 29 and noted no difference in survival rate. Castells et al. (20) compared 33 patients, with a solitary HCC <40 mm in size, who underwent resection and 130 similar patients who underwent PEI. The 5-year survival rates in patients who underwent resection (44%) and in those who were treated by PEI (34%) were not significantly different. Yamasaki et al. (21) compared the results of hepatectomy, TAE and PEI, for an HCC <30 mm and noted similar survival rates after hepatectomy and PEI. The survival rates after TAE were worse than those after hepatectomy or PEI. These previous evaluations were made retrospectively based on small numbers of patients, and the comparisons between resection, TAE, and PEI were inadequate. It is necessary to analyze prognostic factors in a large number of patients in sufficient detail and to evaluate the result of each method of treatment between groups with similar prognostic factors. Although various prognostic factors have been reported (22 24), no conclusion has been drawn as to which are significant. In the present study, the significant factors for better prognosis included clinical stage, tumor number and tumor size. The prognostic factors identified in this study suggested that, therapeutic results in patients with tumors 30mm in size and 3 in number, should be compared with those in patients with more tumors of larger size independently of clinical stages. In patients with Stage I disease having tumors 30 mm and 3 in number, no significant difference in overall survival and disease-free survival rates was found between resection and PEI, but survival rates after TAE were significantly poorer. These results suggested that resection or PEI should be performed in patients with Stage I disease having tumors 30 mm and 3 in number. In patients with clinical Stage II disease having tumors 30 mm and 3 in number, survival after PEI was significantly better than after

6 256 Hepatocellular carcinoma therapy resection. These results suggested that PEI should be performed in patients with clinical Stage II disease having tumors 30 mm and 3 in number. Ebara et al. (13) and Vilana et al. (16) proposed tumors 30 mm in size and 3 in number as indications for PEI, mainly because of technical limitation such as the inability to inject an effective volume of ethanol into the whole area of the tumor. From the viewpoint of prognosis, this indication was correct. Joseph et al. (17) performed CT for evaluation of the therapeutic effect of PEI and demonstrated necrosis in the absence of either peripheral modularity or vascularity. An evaluation of the survival and recurrence rates in patients with complete tumor necrosis will be necessary in the future. Castells et al. (20) compared survival after PEI and after resection. While no difference was found in survival rate, the recurrence rate at 2 years after resection was 45% and after PEI it was 66%. Ebara et al. (13) reported that the recurrence rate at 2 years after PEI was 60.8%. In our study, disease-free survival rates in patients who underwent resection and in those who underwent PEI were not significantly different. However, the disease-free survival rate after resection tended to be better than that after PEI after 3 years. These results suggested that if no operative death occurred, disease-free survival rate of resection would be better than that after PEI. When recurrence occurred, various types of treatment were applied depending on recurrence pattern and liver function. According to Takayasu et al. (25), multiple hepatic recurrence was most common (64%) and solitary recurrence was observed in only 16% of patients who underwent hepatic resection. Ebara et al. (13) reported that no recurrence of the originally treated tumors was found, but in 48% of the patients, new lesions developed. Takayasu et al. (25) reported that one of the significant prognostic factors was a recurrent pattern. Suenaga et al. (26) recommended re-resection in patients with good liver function having 3 tumors. We think that resection or PEI should be selected in patients having 3 recurrent tumors, but that TAE should be selected in patients with multiple recurrent tumors. We have no data concerning which treatment modalities are the best to prolong survival after recognition of recurrence. Taking into consideration the recurrence rate, and site of recurrence, selection of either PEI or resection, with reference to their advantages and disadvantages, requires future prospective studies. For treatment of patients having tumors 31 mm and 3 in number, resection should be performed. While survival in patients with clinical Stage I disease who underwent a one-portion resection and in those who had a two- or three-portion resection were not significantly different, resection is recommended in patients having tumors located in both lobes of the liver. In patients with clinical Stage II disease having tumors located in both lobes of the liver, TAE should be done rather than resection of the two portions. In the patients having tumors 31 mm and 4 in number, survival after resection was significantly better than after TAE in our study. While resection was shown to be the most consistently effective method of treatment, the significant problem is operative death. Reported operative mortality has ranged from 2.3% to 7% (5,7,11). Castells et al. (20) reported that no patients died after undergoing PEI, while three patients died after resection. In the present study, the mortality rate of patients with clinical Stage I disease was significantly lower than that of those with Stage II disease. Since the operative mortality rate of patients with clinical Stage II disease was significantly worse than those with Stage I disease in the present study, resection should be selected, taking the site of the tumor into consideration in patients with clinical Stage II disease. APPENDIX Participant members of this study: Kouichi Akamatsu, Third Department of Internal Medicine, School of Medicine, Ehime University; Masatoshi Okazaki, Department of Radiology, School of Medicine, Fukuoka University; Eizo Okamoto, First Department of Surgery, Hyogo College of Medicine; Yo Sasaki, Department of Surgery, Osaka Medical Center for Cancer and Cardiovascular Diseases; Kyuichi Tanikawa, Second Department of Internal Medicine, School of Medicine, Kurume University; Toshimichi Nakayama,Second Department of Surgery, School of Medicine, Kurume University; Takao Tsuji, First Department of Internal Medicine, Okayama University Medical School; Satoshi Nakano, Department of Gastroenterology, Ogaki Municipal Hospital; Yoshiyuki Shimamura, Division of Surgery, National Matsudo Hospital, Masatoshi Makuuchi, The Second Department of Surgery, School of Medicine, Shinshu University; Tadatoshi Takayama, Division of Surgery, National Cancer Center Hospital Central; Toshimasa Asahara, Second Department of Surgery, School of Medicine, Hiroshima University; Kazuki Ito, Department of Gastroenterology, Shizuoka General Hospital; Tetsuo Takayama, Department of Internal Medicine, Sakashita Hospital; Minoru Tanada, Department of Surgery, National Shikoku Cancer Center Hospital; Hiroki Taniguchi, First Department of Surgery, Kyoto Prefectural University of Medicine; Kenji Nakamura, Department of Radiology, Osaka City University Medical School; Kiyoshi Yoshikawa, Department of Surgery, Osaka Rosai Hospital; Makoto Watanabe, Second Department of Internal Medicine, Shimane Medical University. Acknowledgement This study was supported by the Comprehensive 10-year Strategy for Cancer Control by the Ministry of Health and Welfare. References 1. Makuuchi M, Hasegawa H, Yamazaki S. Intraoperative ultrasonic examination for hepatectomy. Jpn J Clin Oncol 1981;11: Ryu M, Watanabe Y, Ozaki M, Yamamoto H, Usui S, Hirasawa H, et al. Intraoperative ultrasonography for hepatic resection. Asian J Surg 1984; 7: Okuda K, Ryu M, Tobe T. Surgical management of hepatoma: the Japanese experience. In: Wanebo KJ, editor. Hepatic and Biliary Cancer. New York: Marcel Decker, 1987: Nagao T, Goto S, Kawano N, Inoue S, Mizuta T, Morioka Y, et al. Hepatic resection for hepatocellular carcinoma. Ann Surg 1986;205: Nagasue N, Kohno H, Chang YC, Tanimura H, Yamanoi A, Uchida M, et al. Liver resection for hepatocellular carcinoma: results of 229 consecutive patients during 11 years. Ann Surg 1993;217:

7 Jpn J Clin Oncol 1997;27(4) The Liver Cancer Study Group in Japan. Primary liver cancer in Japan. Ann Surg 1990;221: Chen MF, Hwang TL, Jeng LBB, Jan YY, Wang CS, Chou FF, et al. Hepatic resection in 1230 patients with hepatocellular carcinoma. Arch Surg 1988;124: Iwatsuki S, Startzl T. Personal experience with 411 resections. Ann Surg 1988;208: Franco D, Capussoti L, Smadja C, Bouzari H, Meakins J, Kemeny F, et al. Resection of hepatocellular carcinoma; results in 72 patients with cirrhosis. Gastroenterology 1990;98: Bismuth H, Chiche L, Adam R, Castaing D, Diamond T, Dennison A, Liver resection versus transplantation for hepatocellular carcinoma in cirrhotic patients. Ann Surg 1993;218: Tang Z, Yu Y, Zhou X, Ma Z, Yang R, Lu J, et al. Surgery of small hepatocellular carcinoma; analysis of 144 cases. Cancer 1989;64: Yamada R, Kishi K, Sonomura T, Tsuda M, Nomura S, Satoh M. Transcatheter arterial embolization in unresectable hepatocellular carcinoma. Cardiovasc Intervent Radiol 1990;13: Ebara M, Ohto M, Sugiura N, Yoshikawa M, Okuda K, Kondo F, et al. Percutaneous ethanol injection for the treatment of small hepatocellular carcinoma; study of 95 patients. Gastroenterol Hepatol 1990;5: Livrahghi T, Balondi L, Rapaccini S, Torzilli ASG. Percutaneous ethanol injection in the treatment of hepatocellular carcinoma in cirrhosis. Cancer 1992;69: Shiina S, Tagawa K, Unuma T, Fujino H, Uta Y, Niwa Y, et al. Percutaneous ethanol injection therapy of hepatocellular carcinoma; analysis of 77 patients. Am J Roentgenol 1990;155: Vilana R, Bruix J, Bru C, Asyuso C, Sole M, Rodes J. Tumor size determines the efficacy of percutaneous ethanol injection for the treatment of small hepatocellular carcinoma. Hepatology 1992;16: Joseph F B, Baumgarten D A, Bernardino M. Hepatocellular carcinoma: CT appearance after percutaneous ethanol ablation therapy. Work in progress. Radiology 1993;186: Ohnishi K, Tanabe Y, Ryu M, Isono K, Yamamoto Y, Usui S, et al. Prognosis of hepatocellular carcinoma smaller than 5 cm in relation to treatment: study of 100 patients. Hepatology 1987;7: Yoshimi F, Nagao T, Inoue S, Kawano N, Muto T, Gunji T, et al. Comparison of hepatectomy and transcatheter arterial chemoembolization for the treatment of hepatocellular carcinoma: Necessity for prospective randomized trial. Hepatology 1992;16: Castells A, Bruix J, Bru C, Fuster J, Vilana R, Navasa M, et al. Treatment of small hepatocellular carcinoma in cirrhotic patients: A cohort study comparing surgical resection and percutaneous ethanol injection. Hepatology 1933;18: Yamasaki S, Hasegawa H, Makuuchi M, Takayama T, Kosuge T, Shimada K. Choice of treatments for small hepatocellular carcinoma: Hepatectomy, embolization or ethanol injection. Gastroenterol Hepatol 1991;6: Okuda K, Otsuki 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: Calvet X, Bruix J, Gines P, Bru C, Sole M, Vilana R, Rodes J. Prognostic factors of hepatocellular carcinoma in the West: a multivariate analysis in 206 patients. Hepatology 1990;12: Rosellini SR, Arienti V, Nanni O, Ugenti F, Tassinari M, Camporesi C, et al. Hepatocellular carcinoma. Prognostic factors and survival analysis in 135 Italian patients. Hepatology 1992;16: Takayasu K, Wakao F, Moriyama N, Muramatsu Y, Yamasaki S, Kosuge T, et al. Postresection recurrence of hepatocellular carcinoma treated by arterial embolization: analysis of prognostic factors. Hepatology 1992;16: Suenaga M, Sugita H, Kokuba Y, Uehara S, Kurumiya T. Repeated hepatic resection for recurrent hepatocellular carcinoma in eighteen cases. Surgery 1994;115:452 7.