Prognostic Significance of Simultaneous Measurement of Three Tumor Markers in Patients With Hepatocellular Carcinoma

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2006;4:111 117 Prognostic Significance of Simultaneous Measurement of Three Tumor Markers in Patients With Hepatocellular Carcinoma HIDENORI TOYODA,* TAKASHI KUMADA,* SEIKI KIRIYAMA,* YASUHIRO SONE,* MAKOTO TANIKAWA,* YASUHIRO HISANAGA,* AKIHIRO YAMAGUCHI, MASATOSHI ISOGAI, YUJI KANEOKA, and JUNJI WASHIZU *Department of Gastroenterology and Department of Surgery, Ogaki Municipal Hospital, Ogaki, Japan Background & Aims: We conducted a prospective study to evaluate the significance of simultaneous measurement of 3 currently used tumor markers in the evaluation of tumor progression and prognosis of patients with hepatocellular carcinoma (HCC). Methods: Three tumor markers for HCC, alpha-fetoprotein (AFP), Lens culinaris agglutinin A reactive fraction of AFP (AFP-L3), and desgamma-carboxy prothrombin (DCP), were measured in the same serum samples obtained from 685 patients at the time of initial diagnosis of HCC. Positivity for AFP >20 ng/dl, AFP-L3 >10% of total AFP, and/or DCP >40 mau/ml was determined. In addition, tumor markers were measured after treatment of HCC. Results: Of the 685 patients, 337 (55.8%) were positive for AFP, 206 (34.1%) were positive for AFP-L3, and 371 (54.2%) were positive for DCP. In a comparison of patients positive for only 1 tumor marker, patients positive for AFP-L3 alone had a greater number of tumors, whereas patients positive for DCP alone had larger tumors and a higher prevalence of portal vein invasion. When patients were compared according to the number of tumor markers present, the number of markers present clearly reflected the extent of HCC and patient outcomes. The number of markers present significantly decreased after treatment. Conclusions: Tumor markers AFP-L3 and DCP appear to represent different features of tumor progression in patients with HCC. The number of tumor markers present could be useful for the evaluation of tumor progression, prediction of patient outcome, and treatment efficacy. Hepatocellular carcinoma (HCC) is one of the most common malignancies, especially in southern and eastern Asia. Currently in Japan, HCC is the third leading cause of death from cancer. The development of various scanning techniques and the identification of sensitive and specific tumor markers for HCC have contributed not only to detection of HCC but also to evaluation of progression of HCC and determination of patient prognosis. Three tumor markers specific for HCC are currently used clinically in Japan: alpha-fetoprotein (AFP), Lens culinaris agglutinin A reactive fraction of AFP (AFP-L3), and des-gamma carboxy prothrombin (DCP), which is also called protein induced by vitamin K absence-ii (PIVKA-II). Usefulness of the measurement of each of these tumor markers for detection and diagnosis of HCC, for evaluation of tumor progression, and for determination of patient prognosis has been reported. 1 4 However, these 3 tumor markers have not been evaluated together for evaluation of the progression of HCC and prediction of patient outcome. In the present study, we measured levels of these 3 tumor markers simultaneously at the time of HCC diagnosis, and we analyzed them with respect to tumor progression and patient survival. Patients and Methods Patients A total of 689 patients were diagnosed as having initial HCC (not recurrence) and treated at Ogaki Municipal Hospital between 1995 and 2004. Of these patients, 685 were enrolled in this prospective study; the remaining 4 patients were excluded because they were taking warfarin or vitamin K, which would influence the serum DCP level. The study group comprised 497 men and 188 women, with a mean age of 66.7 8.7 years (median, 67 years; range, 31 93 years). Characteristics of the patients are shown in Table 1. HBV infection was detected in 104 patients, HCV in 508 patients, and both HBV and HCV in 13 patients; no hepatitis virus was detected in the remaining 60 patients. Of the 685 patients, 193 underwent Abbreviations used in this paper: AFP, alpha-fetoprotein; AFP-L3, Lens culinaris agglutinin A reactive fraction of AFP; DCP, des-gammacarboxy prothrombin; HCC, hepatocellular carcinoma; LAT, locoregional ablative therapy; PIVKA-II, protein induced by vitamin K absence-ii; TACE, transcatheter arterial chemoembolization. 2006 by the American Gastroenterological Association 1542-3565/06/$32.00 PII: 10.1053/S1542-3565(05)00855-4

112 TOYODA ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 4, No. 1 Table 1. Clinical Characteristics of the Study Patients (n 685) Age (y) 66.7 8.7 Sex (female/male) 188 (27.4)/497 (72.6) a Etiology of underlying liver disease (HBV/HCV/HBV, HCV/non-HBV, non-hcv) 104 (15.2)/508 (74.2)/13 (1.9)/60 (8.7) a Child-Pugh class (A/B/C) 444 (64.8)/194 (28.3)/47 (6.9) a Albumin (g/dl) 3.44 0.60 Total bilirubin (mg/dl) 1.13 1.44 Prothrombin time (%) 82.9 17.6 15-minute retention of ICG (%) 22.9 17.0 Size of largest tumor (cm) 3.58 3.02 Number of tumors 1.86 1.77 Portal vein thrombosis ( / ) 593 (86.6)/92 (13.4) a AFP (ng/dl) 27.0 (0.8 2,402,000) b AFP-L3 (%) 0.5 (0 92.1) b DCP (mau/ml) 52.0 (10.0 75,000) b Treatment No treatment 90 (13.1) a Hepatectomy 193 (28.2) a LAT 166 (24.2) a TACE 196 (28.6) a Others 40 (5.9) a ICG, indocyanine green test. a Percentages are shown in parentheses. b Ranges are shown in parentheses. hepatectomy, 166 were treated by locoregional ablative therapy (LAT) including ethanol injection, microwave thermocoagulation, or radiofrequency ablation, and 196 were treated by transcatheter arterial chemoembolization (TACE). The study protocol was approved by the hospital ethics committee and was in compliance with the Helsinki Declaration. Written informed consent was obtained from all patients before the study for use of the pathology and laboratory data. Measurement of the Three Tumor Markers AFP, AFP-L3, and DCP were measured in the same serum sample at the time of HCC diagnosis. In patients who underwent hepatectomy, LAT, or TACE as a treatment for HCC, they were also measured approximately 1 month after the end of therapy. The serum AFP level was determined by enzyme-linked immunosorbent assay with a commercially available kit (ELISA-AFP; International Reagents, Kobe, Japan). A cutoff value of 20 ng/ml AFP was used to establish positivity for AFP, as proposed by Oka et al 5 and Koda et al. 6 Serum AFP-L3 was measured by lectin-affinity electrophoresis coupled with antibody-affinity blotting (AFP Differentiation Kit L; Wako Pure Chemical Industries, Ltd, Osaka, Japan) and was finally expressed as the percentage of AFP-L3 (AFP-L3 level/total AFP level 100). 7,8 The cutoff value used to establish positivity for AFP-L3 was AFP-L3 10%, as proposed by Shimizu et al. 9 The serum DCP level was determined by sensitive enzyme immunoassay (Eitest PIVKA-II kit; Eisai Laboratory, Tokyo, Japan) according to the manufacturer s instructions. 10 12 The cutoff value used to establish positivity for DCP was 40 mili arbitrary unit (mau)/ml, as proposed by Okuda et al. 13 Characteristics of HCC such as size of tumor, number of tumors, presence of portal vein thrombosis, and tumor stage according to American Joint Committee on Cancer (AJCC) and patient survival rates were compared according to tumor markers. In patients who underwent hepatectomy, LAT, or TACE, the changes in the status of tumor markers before and after treatment were analyzed. Statistical Analyses Data are expressed as mean standard deviation values or the median and range unless otherwise specified. Differences in proportions of number of patients between groups were analyzed by 2 test. Differences in quantitative values were analyzed by Student t test if the data were normally distributed; otherwise, differences were analyzed by Mann Whitney U test. The date of HCC diagnosis was defined as time zero for calculations of patient survival. Surviving patients and patients who died of causes other than liver disease were censored. Patients who died of HCC-related causes or liver failure were not censored. The Kaplan Meier method 14 was used to calculate survival rates, and the log-rank test 15 was used to analyze differences in survival. The Cox proportional hazards model 16 was used for multivariate analysis for factors that influenced patient survival. The variables analyzed were age, sex, Child-Pugh class (A vs B, C), initial treatment (no treatment vs hepatectomy, LAT, TACE, or other treatment), and the number of tumor markers present (none, 1, 2, or 3). The JMP statistical software package, version 4.0 (SAS Institute, Cary, NC), was used for all statistical analyses. All P values were derived from two-tailed tests, and P.05 was accepted as statistically significant. Results Serum AFP was above the cutoff level in 387 of the 685 patients (56.5%), serum AFP-L3 was above the cutoff level in 227 patients (33.1%), and serum DCP was above the cutoff level in 371 patients (54.2%). No tumor markers were above the cutoff level in 159 patients (23.2%). Only 1 of 3 tumor markers was above the cutoff level in 220 patients (32.1%; only AFP in 96 patients, only AFP-L3 in 14 patients, and only DCP in 110 patients). Two of 3 tumor markers were above the cutoff level in 153 patients (22.3%; AFP and AFP-L3 in 44 patients, AFP and DCP in 72 patients, and AFP-L3 and DCP in 15 patients). All 3 tumor markers were above the cutoff level in the remaining 153 patients (22.3%) (Figure 1).

January 2006 THREE TUMOR MARKERS FOR HCC 113 Figure 1. Distribution of patients with various patterns of positivity for the 3 tumor markers of HCC. Comparison of Tumor Characteristics and Patient Survival Between Patients Not Positive for a Tumor Marker and Those Positive for Only One Tumor Marker Characteristics of patients and HCCs were compared between patients not positive for a tumor marker and patients positive for only 1 tumor marker (AFP, AFP-L3, or DCP; Table 2). No significant difference was observed between patients not positive for a tumor marker and those positive for AFP only, except that there was a greater percentage of female patients in the AFP alone positive group (P.0138). In contrast, the tumor (largest tumor in cases of multiple tumors) was significantly larger in patients positive for DCP alone than in patients not positive for a tumor marker (P.0001). Tumors tended to be larger in patients positive for AFP-L3 alone than in patients not positive for a tumor marker (P.0614). In addition, the number of tumors was greater in patients positive for AFP-L3 alone than in patients not positive for a tumor marker (P.0075). In patients positive for DCP alone, the prevalence of HCC Figure 2. Survival of patients not positive for a tumor marker and patients positive for only 1 tumor marker (AFP, AFP-L3, and DCP) (patients not positive for a tumor marker vs patients positive for only DCP, P.0014). with portal vein thrombosis was significantly higher than that in patients not positive for a tumor marker (P.0018). There was no difference in the survival rate of patients not positive for a tumor marker and those positive for AFP alone or AFP-L3 alone. The survival rate of patients positive for DCP alone was lower than that of patients not positive for a tumor marker (P.0014; Figure 2). Tumor Characteristics and Patient Survival in Relation to the Number of Tumor Markers Characteristics of patients and HCCs in relation to the number of positive tumor markers present are shown in Table 3. The size and number of HCCs and the prevalence of portal vein thrombosis increased gradually along with an increase in the number of tumor markers present. The numbers of tumor markers present are shown in relation to tumor stages in Table 4. Significant association was found between these 2 factors. Table 2. Tumor Characteristics Among Patients Not Positive for a Tumor Marker and Those Positive for 1 Tumor Marker None (n 159) AFP alone (n 96) AFP-L3 alone (n 14) DCP alone (n 110) Age (y SD) 66.3 8.2 66.3 8.3 63.6 8.2 67.9 8.2 Sex (female/male) 37/122 37/59 3/11 25/85 Size of largest tumor a (cm SD) 2.24 1.41 2.17 1.20 3.99 3.90 3.94 3.00 Number of tumors b (n SD) 1.42 0.97 1.52 0.83 2.21 1.48 1.54 1.05 Portal vein thrombosis c 3 (1.9) 0 1 (7.1) 13 (11.8) NOTE. Percentages are shown in parentheses. SD, standard deviation. a None vs AFP-L3 alone, P.0614; none vs DCP alone, P.0001. b None vs AFP-L3 alone, P.0075. c None vs DCP alone, P.0018.

114 TOYODA ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 4, No. 1 Table 3. Tumor Characteristics According to the Number of Tumor Markers Present None (n 159) 1 marker (n 220) 2 markers (n 153) 3 markers (n 153) Age (y SD) 66.3 8.2 66.9 8.3 66.8 9.3 66.8 9.4 Sex (female/male) 37/122 65/155 44/109 42/111 Size of largest tumor a (cm SD) 2.24 1.41 3.18 2.61 3.72 3.18 5.57 3.69 Number of tumors b (n SD) 1.42 0.97 1.57 1.00 2.09 2.19 2.67 2.59 Portal vein thrombosis c 3 (1.9) 14 (6.4) 24 (15.7) 50 (32.7) NOTE. Percentages are shown in parentheses. SD, standard deviation. a None vs 1 marker, none vs 2 markers, and 2 markers vs 3 markers, P.0001. b None vs 2 markers, P.0105; 2 markers vs 3 markers, P.0189. c None vs 1 marker, P.0677; 1 marker vs 2 markers, P.0055; 2 markers vs 3 markers, P.0009. Survival rate differed significantly between groups. Survival rates decreased with an increase in the number of tumor markers present (no marker vs 1 marker, P.0181; 1 marker vs 2 markers, P.0141; 2 markers vs 3 markers, P.0001; Figure 3). Multivariate analysis showed the number of tumor markers present, as well as Child-Pugh class and treatment for HCC, to be independent factors affecting patient survival, although we found no increase in the risk ratio between patients not positive for a tumor marker and those positive for only 1 tumor marker (Table 5). Changes in the Number of Tumor Markers Present in Association With Treatment of Hepatocellular Carcinoma The number of tumor markers present was compared before and after treatment in patients who underwent hepatectomy, LAT, or TACE for HCC. As is shown in Figure 4, the number of tumor markers present decreased in association with treatment for HCC (hepatectomy, from 2.21 1.07 to 1.49 0.86, P.0001; LAT, from 1.99 0.95 to 1.76 0.87, P.0004; TACE, from 2.57 1.07 to 2.20 1.17, P.0001). The decrease was most marked in patients who underwent hepatectomy. Discussion In the present study, we measured 3 tumor markers, AFP, AFP-L3, and DCP, in serum obtained at the time of HCC diagnosis. These are all tumor markers for HCC that are currently used clinically, although another tumor marker, glypican-3, has been reported. 17,18 In the comparison of patients positive for only 1 tumor marker (AFP, AFP-L3, or DCP) and those not positive for a tumor marker, we did not find a difference in characteristics of HCCs or survival between patients not positive for a tumor marker and those positive for AFP alone. AFP is the tumor marker that is routinely and most widely used for monitoring the development of HCC. 1,5,19 It is also reported to be a marker for the recurrence of HCC after LAT or TAE. 6,20 However, AFP increases in association with hepatocyte regeneration also and is associated with serum ALT activity; AFP elevation is observed in up to 20% of patients with chronic hepatitis and in 20% 60% of patients with cirrhosis, even in the absence of HCC. 4 In addition, a recent study showed the significance of AFP as a marker of liver fibrosis. 21 AFP, therefore, does not always directly reflect the development of HCC. The lack of differences between patients not positive for a tumor marker and those positive for AFP alone indicates that the clinical significance of AFP as a marker of tumor progression of HCC is limited. AFP elevation, however, could be a marker for a high risk of multicentric recurrence of HCC. Elevated serum ALT activity is reported to be a risk factor for multicentric occurrence of HCC. 22,23 Because AFP elevation indicates enhanced liver regeneration associated with increased ALT activity and also indicates advanced liver Table 4. Tumor Stage and the Number of Tumor Markers Present Stage I (n 182) Stage II (n 261) Stage III (n 147) Stage IV (n 95) 0 markers (n 159) 69 (43.4) 68 (42.8) 19 (11.9) 3 (1.9) 1 marker (n 220) 61 (27.7) 101 (45.9) 45 (20.5) 13 (5.9) 2 markers (n 153) 44 (28.8) 51 (33.3) 34 (22.2) 24 (15.7) 3 markers (n 153) 8 (5.2) 41 (26.8) 49 (32.0) 55 (36.0) NOTE. Percentages are shown in parentheses.

January 2006 THREE TUMOR MARKERS FOR HCC 115 Figure 3. Patient survival according to the number of tumor markers present (patients not positive for a tumor marker vs patients positive for only 1 tumor marker, P.0181; patients positive for only 1 tumor marker vs patients positive for 2 tumor markers, P.0141; patients positive for 2 tumor markers vs patients positive for 3 tumor markers, P.0001). fibrosis, AFP elevation could be a predictor of multicentric recurrence of HCC after treatment. AFP, therefore, has significance in predicting a long-term outcome of patients with HCC. We found advanced HCC in patients positive for AFP-L3 alone or positive for DCP alone. In addition, we found that the features of HCC differed between patients positive for AFP-L3 alone and those positive for DCP alone. HCC in both groups was larger size than HCC in patients not positive for a tumor marker, although the increase in size did not reach statistical significance in Figure 4. Changes in the number of tumor markers present before and after treatment of HCC (hepatectomy and TACE, before treatment vs after treatment, P.0001; LAT, before treatment vs after treatment, P.0005). patients positive for AFP-L3 alone, possibly because of the small number of patients in this group. In contrast, the number of tumors was significantly greater in patients positive for AFP-L3 alone, whereas the prevalence of portal vein thrombosis was significantly higher in patients positive for DCP alone in comparison to that in patients not positive for a tumor marker. Elevation of AFP-L3 or DCP is reported to be a marker of advanced HCC. 13,24 30 DCP is reported to be an indicator of portal vein invasion of HCC. 28 We also found a higher prevalence of HCC with portal vein thrombosis in patients positive for DCP alone. The survival rate of patients Table 5. Multivariate Analysis for Factors Associated With Patient Survival Factor Parameter estimate Standard error X Risk ratio (95% confidence interval) P value Age 0.0067 0.0070 0.90 1.0067 (0.9930 1.0208).3418 Sex Male 1 Female 0.0011 0.0662 0.00 0.9989 (0.8791 1.1397).9870 Child-Pugh class A 1 B 0.1676 0.0685 5.86 1.1825 (1.0327 1.3511).0155 C 0.2712 0.1111 5.53 1.3116 (1.0477 1.6210).0187 Treatment No treatment 1 Hepatectomy 0.3880 0.1037 13.63 0.6784 (0.5536 0.8314).0002 LAT 0.3764 0.1069 11.89 0.6863 (0.5576 0.8482).0006 TACE 0.3573 0.0947 13.42 0.6995 (0.5822 0.8445).0002 Others a 0.3866 0.1430 7.87 0.6794 (0.5078 0.8923).0050 Number of tumor markers 0 1 1 0.1099 0.0933 1.41 1.1162 (0.9319 1.3445).2348 2 0.2770 0.0988 8.03 1.3192 (1.0887 1.6049).0046 3 0.7098 0.0943 60.65 2.0336 (1.6948 2.4553).0001 a Includes repeated hepatic arterial infusion chemotherapy and systemic chemotherapy.

116 TOYODA ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 4, No. 1 positive for AFP alone was similar to that of patients not positive for a tumor marker. Patients positive for DCP alone had a significantly lower survival than patients not positive for a tumor marker. Patients positive for AFP-L3 alone had a survival rate similar to that of patients positive for DCP alone, although the difference in survival between patients not positive for a tumor marker and those positive for AFP-L3 alone was not statistically significant, possibly because of the small number of patients positive for AFP-L3 alone. AFP-L3 and DCP, therefore, were indicators of different features of advanced HCC, and both were indicators of a poor patient prognosis. When results of the measurements of the 3 tumor markers were combined, the number of tumor markers present, there was an even distribution of patients with 1, 2, and 3 markers. The size and number of tumors and the prevalence of portal vein thrombosis increased gradually with the increase in the number of tumor markers present, and the number of markers present was well associated with the stages of HCC. The number of tumor markers present, therefore, could be an indicator of tumor progression as well as tumor stage. In univariate analysis of patient survival, we observed significant differences between patients positive for 1, 2, and 3 tumor markers and patients not positive for a tumor marker. According to multivariate analysis, the number of tumor markers present was independently associated with patient survival, as were Child-Pugh class and treatment of HCC. The risk ratio increased gradually with the increase in the number of tumor markers present, although there was no difference in the risk ratio between patients positive for 1 tumor marker and patients not positive for a tumor marker. AFP-L3 elevation and DCP elevation might reflect different features of advanced HCC, and AFP elevation might indicate a high risk of multicenter occurrence of HCC. Evaluation of these 3 tumor markers in combination, therefore, might serve as a good predictor of outcome in patients with HCC. With respect to treatment of HCC, the number of tumor markers present decreased after hepatectomy, after LAT, and after TACE. The decrease was more marked in patients who underwent hepatectomy than in those who underwent LAT; hepatectomy was superior to LAT in terms of reduction of the number of tumor markers. Unfortunately, we do not have data on the number of tumor markers present in patients treated by liver transplantation. Such analysis should be performed in the future. In summary, AFP-L3 and DCP elevation both indicated progression of HCC, but they reflected different features of the progression. The number of tumor markers present reflected tumor progression and can serve as a predictor of patient outcome. All 3 tumor markers can be assayed simultaneously in serum samples and do not require radiologic or pathologic evaluation. The evaluation of HCC by tumor markers does not, therefore, depend on the quality of radiologic apparatus or scanning skills. In addition, it is not influenced by a discrepancy between radiologic and pathologic evaluations of tumor extension such as portal vein thrombosis. Status of the 3 tumor markers can be useful as a laboratory indicator of tumor progression. References 1. Tsukuma H, Hiyama T, Tanaka S, et al. Risk factors for hepatocellular carcinoma among patients with chronic liver disease. N Engl J Med 1993;328:1797 1801. 2. Taketa K, Sekiya C, Namiki N, et al. Lectin-reactive profiles of alpha-fetoprotein characterizing hepatocellular carcinoma and related conditions. Gastroenterology 1990;99:508 518. 3. Liebman HA, Furie BC, Tong MJ, et al. Des-gamma-carboxy (abnormal) prothrombin as a serum marker of primary hepatocellular carcinoma. N Engl J Med 1984;310:1427 1431. 4. Fujiyama S, Tanaka M, Maeda S, et al. Tumor markers in early diagnosis, follow-up and management of patients with hepatocellular carcinoma. Oncology 2002;62:S57 S63. 5. Oka H, Tamori A, Kuroki T, et al. Prospective study of alphafetoprotein in cirrhotic patients monitored for development of hepatocellular carcinoma. Hepatology 1994;19:61 66. 6. Koda M, Murawaki Y, Mitsuda A, et al. Predictive factors for intrahepatic recurrence after percutaneous ethanol injection therapy for small hepatocellular carcinoma. Cancer 2000;88:529 537. 7. Taketa K, Endo Y, Sekiya C, et al. A collaborative study for the evaluation of lectin-reactive alpha-fetoprotein in early detection of hepatocellular carcinoma. Cancer Res 1993;53:5419 5423. 8. Yamashita F, Tanaka M, Satomura S, et al. Prognostic significance of Lens culinaris agglutinin A-reactive alpha-fetoprotein in small hepatocellular carcinoma. Gastroenterology 1996;111: 996 1001. 9. Shimizu K, Taniichi T, Satomura S, et al. Establishment of assay kits for determination of microheterogeneities of alpha-fetoprotein using lectin-affinity electrophoresis. Clin Chim Acta 1993; 214:3 12. 10. Mita Y, Aoyagi Y, Yanagi M, et al. The usefulness of determining des-gamma-carboxy prothrombin by sensitive enzyme immunoassay in the early diagnosis of patients with hepatocellular carcinoma. Cancer 1998;82:1643 1648. 11. Okuda H, Nakanishi T, Takatsu K, et al. Measurement of serum levels of des-gamma-carboxy prothrombin in patients with hepatocellular carcinoma by a revised enzyme immunoassay kit with increased sensitivity. Cancer 1999;85:812 818. 12. Nomura F, Ishijima M, Kuwa K, et al. Serum des-gamma-carboxy prothrombin levels determined by a new generation of sensitive immunoassays in patients with small-sized hepatocellular carcinoma. Am J Gastroenterol 1999;94:650 654. 13. Okuda H, Nakanishi T, Takatsu K, et al. Serum levels of desgamma-carboxy prothrombin measured using the revised enzyme immunoassay kit with increased sensitivity in relation to clinicopathological features of solitary hepatocellular carcinoma. Cancer 2000;88:544 549. 14. Kaplan EL, Meier P. Non parametric estimation for incomplete observation. J Am Stat Assoc 1958;53:457 481. 15. Petro R, Pike MC. Conservation of the approximation (0-E2)/E in the log rank test for srvival data on tumor incidence data. Biometrics 1973;29:579 584.

January 2006 THREE TUMOR MARKERS FOR HCC 117 16. Cox D. Regression models and life tables. J R Stat Soc 1972; 34:187 220. 17. Hippo Y, Watanabe K, Watanabe A, et al. Identification of soluble NH2-terminal fragment of glypican-3 as a serological marker for early-stage hepatocellular carcinoma. Cancer Res 2004;64: 2418 2423. 18. Capurro M, Wanless IR, Sherman M, et al. Glypican-3: a novel serum and histochemical marker for hepatocellular carcinoma. Gastroenterology 2003;125:89 97. 19. Ikeda K, Saitoh S, Koida I, et al. A multivariate analysis of risk factors for hepatocellular carcinogenesis: a prospective observation of 795 patients with viral and alcoholic cirrhosis. Hepatology 1993;18:47 53. 20. Okusaka T, Okada S, Ueno H, et al. Evaluation of the therapeutic effect of transcatheter arterial embolization for hepatocellular carcinoma. Oncology 2000;58:293 299. 21. Hu K-Q, Kyulo NL, Lim N, et al. Clinical significance of elevated alpha-fetoprotein (AFP) in patients with chronic hepatitis C, but not hepatocellular carcinoma. Am J Gastroenterol 2004;99: 860 865. 22. Tarao K, Rino Y, Ohkawa S, et al. Close association between high serum alanine aminotransferase levels and multicentric hepatocarcinogenesis in patients with hepatitis C virus-associated cirrhosis. Cancer 2002;94:1787 1795. 23. Tarao K, Takemiya S, Tamai S, et al. Relationship between the recurrence of hepatocellular carcinoma (HCC) and serum alanine aminotransferase levels in hepatectomized patients with hepatitis C virus-associated cirrhosis and HCC. Cancer 1997;79:688 694. 24. Aoyagi Y, Isokawa O, Suda T, et al. The fucosylation index of alpha-fetoprotein as a possible prognostic indicator for patients with hepatocellular carcinoma. Cancer 1998;83:2076 2082. 25. Kumada T, Nakano S, Takeda I, et al. Clinical utility of Lens culinaris agglutinin-reactive alpha-fetoprotein in small hepatocellular carcinoma: special reference to imaging diagnosis. J Hepatol 1999;30:125 130. 26. Hayashi K, Kumada T, Nakano S, et al. Usefulness of measurement of Lens culinaris agglutinin-reactive fraction of alpha-fetoprotein as a marker of prognosis and recurrence of small hepatocellular carcinoma. Am J Gastroenterol 1999;94:3028 3033. 27. Tada T, Kumada T, Toyoda H, et al. Relationship between Lens culinaris agglutinin-reactive alpha-fetoprotein and pathologic features of hepatocellular carcinoma. Liver Int 2005;25:848 853. 28. Koike Y, Shiratori Y, Sato S, et al. Des-gamma-carboxy prothrombin as a useful predisposing factor for the development of portal venous invasion in patients with hepatocellular carcinoma: a prospective analysis of 227 patients. Cancer 2001;91:561 569. 29. Sassa T, Kumada T, Nakano S, et al. Clinical utility of simultaneous measurement of serum high-sensitivity des-gamma-carboxy prothrombin and Lens culinaris agglutinin A-reactive alphafetoprotein in patients with small hepatocellular carcinoma. Eur J Gastroenterol Hepatol 1999;11:1387 1392. 30. Shimauchi Y, Tanaka M, Kuromatsu R, et al. A simultaneous monitoring of Lens culinaris agglutinin A-reactive alpha-fetoprotein and des-gamma-carboxy prothrombin as an early diagnosis of hepatocellular carcinoma in the follow-up of cirrhotic patients. Oncol Rep 2000;7:249 256. Address requests for reprints to: Hidenori Toyoda, MD, Department of Gastroenterology, Ogaki Municipal Hospital, 4-86 Minaminokawa, Ogaki, Gifu, 503-8502, Japan. e-mail: tkumada@he.mirai.ne.jp; fax: 81-584-75-5715. Presented in part at 40th Annual Meeting of the European Association for the Study of the Liver, April 2005.