Hepatocellular Carcinoma in Keio Affiliated Hospitals – Diagnosis , Treatment , and Prognosis of this Disease –

161 Abstract We have retrospectively investigated patients with hepatocellular carcinoma (HCC) (912 cases) treated in the Affiliated Hospitals of Keio University (the Keio Association for the Study of Liver Diseases: KASLD) and here we review the recent diagnosis and treatment of HCC. HCC is a major cause of death in Japan and a major etiology of this disease is chronic viral infection such as hepatitis C virus (HCV) and hepatitis B virus (HBV). Screening of HCC by imaging studies and measurement of serum tumor markers successfully prolonged survival of the patients in Japan. The prognosis of this disease has been determined by both tumor factors and degree of liver function, and its staging is usually established with a recent system such as Japan Integrated Staging Score (JIS). The 5-year survival rate of JIS stage 0, 1, 2 and 3 were 68.3%, 51.9%, 25.8% and 16.6%, respectively in our cohort. Multivariate analysis using Cox proportional hazard models showed that age (>65), HCV infection, tumor number, TB (>1.0), AFP (>20) and PIVKA-II (>40) were significant factors affecting survival among the entire patients. Major treatment strategies are hepatic resection, radio frequency ablation and transarterial chemo-embolization, but alternative treatments such as radiation, chemotherapy, and their combination have been used to reduce tumor sizes resulting prolongation of the survival or maintenance of patients’ quality of life, while liver transplantation has not been popular in Japan. However, the overall survival continues to decrease from year to year, and does not show a plateau phase in Kaplan-Mayer curve. These results suggest that the best way to improve survival can be achieved by prevention of the disease. Antiviral therapies have decreased the incidence of HCC, indicating that treatment for chronic hepatitis is the best way to prevent HCC development at present. (Keio J Med 58 (3) : 161-175, September 2009)


Introduction
Hepatocellular carcinoma (HCC) is the third and fourth leading cause of cancer-related death in Japanese men and women, respectively, and is the 5 t h most common cause of cancer worldwide.The incidence of HCC has increased in Japan in the past several decades and has recently increased dramatically in the United States and other Western countries, except for a small group of patients whose mortality showed favorable trends. 1 The increase is primarily related to the spread of hepatitis C virus (HCV) infection, which peaked in the 1960s and 1980s in Japan and the United States, respectively.Considerable attention has been paid to the emerging HCC epidemic, which is often associated with hepatitis C and hepatitis B, and end-stage liver cirrhosis (LC) which occurs as a result of repetitive cycles of inflammation and repair and is associated with the development of HCC.End-stage liver cirrhosis and HCC can both seriously impact the quality of life of patients.
In Western countries, liver transplantation has been used successfully to treat patients with HCC; 2 however, the shortage of donor livers and other technical reasons limit the option for transplantation in some patients. 3,4 n Japan, liver transplantation for HCC is less common due to the shortage of cadaveric donor organs. 5Additionally, strategies to reduce tumor growth are important for patients who are awaiting liver transplantation.For that purpose, ablation techniques or chemoembolization to control local tumor growth are very useful.Unlike with other cancers, the treatment indication strategy for HCC depends not only on the surgical staging of cancer growth and degree of spread but also on residual reserved liver function.These considerations determine whether treatment modalities other than surgery, such as local ablation techniques, should be employed.Thus, the treatment of HCC has progressed dramatically during the past few decades since our last review, 6 and the prognosis of HCC has improved in parallel with our understating of the nature of the disease.[10][11] In the present report, we summarized data of patients with HCC who had been previously admitted to the Keio University Hospital and the Keio Affiliated Hospitals, which was integrated as the Keio Association for the Study of Liver Diseases (KASLD), and review the treatments and prognoses of this disease.

Patients Studied
In 2007, the core members of the fourth KASLD meeting resolved to collect clinical data from HCC patients through the affiliated hospitals of Keio University to better understand recent trends in HCC treatment.The data were analyzed retrospectively.Survival rates were analyzed using the Kaplan-Meier product limit survival method and the log-rank test was used to determine statistical significance.Data from patients in different groups were compared using the Mann-Whitney U test for continuous variables.Categorical variables were analyzed by x 2 test or Fisher's exact test.We used Cox proportional hazard models to examine the potential determinants of survival.The start date used for survival analysis was the date of the first diagnosis.All statistical analyses were performed using the Statistical Program for Social Sciences (SPSS 13.0 for Windows; SPSS, Inc., Chicago, IL).Statistical significance was defined by a P value less than 0.05.
The list of the hospitals and the number of patients included are summarized in Table 1.Baseline characteristics of the patients are summarized in Table 2.The age distribution of patients with HCC (at the time of diagnosis) is shown in Figure 1; a majority of the patients were elderly, probably because the main cause of HCC is HCV infection, and progression of liver fibrosis depends on age and duration of infection. 1 2, 1 3However, elderly Table 1 Patients and hospitals enrolled in this statistics for the KASLD patients develop HCC more often than younger patients despite their lower grade of fibrosis, 1 4 suggesting that the impact of aging is stronger for fibrosis for development of HCC.It has been said that male gender is the most significant risk factor of HCC.The peak age of female patients was older than male patients and aging factor may be a main causing factor in female patients.

Etiology of HCC
The major clinical risk factor for development of HCC is cirrhosis of the liver.The major risk factors for the development of LC are HBV-and HCV-infection and chronic heavy alcohol drinking. 15HCV has been previously shown to account for the majority of cases of HCC in Japan, 16 and our findings were similar; the frequency of HCV was 69% (Fig. 2). Figure 3 shows the proportion of cases with liver diseases due to different causes among patients admitted to Keio University Hospital during the last 4 years.It is clear that HCC was the major cause of admissions due to liver diseases.HBV was the second leading cause of HCC (seen in 10% of cases) and the infections were acquired early in life through transmission from mother to offspring in most HBV cases (Fig. 2).Chronic alcohol use of greater than 80 g/day for more than 10 years increased the risk of HCC 5-fold, 1 7 and chronic alcohol use in HBV and HCV infection (seen in 1.78% of cases in our study) doubles the risk as compared with either infection alone.Figure 4 shows the causes of LC in patients consulting to Keio University Hospital.Chronic HCV infection was a major cause of cirrhosis; this finding is comparable to nation-wide in Japan, but differs from the trend in Western countries, where alcoholic cirrhosis remains a major cause of LC.Insulin-resistance and the metabolic syndrome are thought to be predisposing factors non-alcoholic steatohepatitis (NASH) and cirrhosis in 10% to 20% of cases, 1 8 and diabetes and obesity have been shown to double the risk of HCC. 19These patients might be included in the unknown category in Figure 2. The cause of LC in 1998 did not include NASH; however, in 2008, NASH was included as one of the causes in our study (Fig. 4).

Screening of HCC
Although potentially curative treatment for early stage of HCC was achieved with a 3-year survival rate in the range of 60-80%, the rate is as low as 10% for late-stage disease worldwide.These observations strongly suggest that screening practices allowing for improved diagnosis at an early stage of this disease could have a large impact on survival.Recently, prospective cohort studies and a randomized controlled trial (RCT) suggested that HCC screening of patients with chronic liver diseases can improve survival rates; 2 0-2 2 however, this notion remains controversial. 2 3 Many clinicians support screening for HCC and believe that it would be unethical to perform a randomized controlled trial to demonstrate its efficacy. 24he Japan Society of Hepatology (JSH) recommends HCC screening by a tumor marker measurement (presence of 2 of 3 markers: alpha-fetoprotein [  ery 3 months, in addition to screening once every 6 months examination by abdominal ultrasonography (US) and annually by computed tomography (CT) scanning for patients with chronic hepatitis, and imaging studies every 3-4 months in patients with LC. 2 5 Although the American Association for the Study of Liver Diseases (AASLD) and European Association for the Study of Liver Diseases (EASL) both recommend HCC screening, the National Cancer Institute (NCI) has not renewed its statement that "based on fair evidence, screening would not result in a decrease in mortality from hepatocellular cancer".Following the rapid increase in the number of patients with HCC in Western countries, hepatologists in these countries recognized the importance of management.The number of HCC cases in Japan seems to have peaked several decades earlier than in Western countries.
The effectiveness of US depends on the experience of the examiner, the equipment available, the presence of LC, and the size of the tumor.It has generally been reported that the sensitivity of US for detecting tumor nodules is greater than 60% and the specificity is greater than 90%, and that the sensitivity decreases in cirrhotic livers.A serum AFP level greater than 20 ng/mL (the upper limit of normal) has low sensitivity for detecting HCC (25-65%) and is not suitable as the sole screening test.Elevated serum AFP levels are sometimes observed in patients with chronic liver disease with a high degree of hepatocyte regeneration.In such cases, AFP-L3 usually distinguishes true HCC from such a condition described above.Recent studies have reported that HCC surveillance strategies using both AFP and US improve the cost-effectiveness (in quality adjusted life years) of therapy. 26,27 t is thought that high-risk patients for HCC include all patients with cirrhosis and some HBV-infected patients irrespective of cirrhosis (age >40 years in men and >50 years in women).

Diagnosis of HCC
Because screening for HCC has recently become common in Japan, small or early-stage HCC is often diagnosed prior to the presentation of specific clinical symptoms.In contrast, right upper quadrant abdominal pain, early satiety and weight loss are recognized as the textbook manifestations of HCC in Western countries.These symptoms are thought to occur in more advanced cases.Histological examination is not usually required for the diagnosis of HCC, unlike with cancers in other organs.A consensus statement -that a diagnosis of HCC can be made based on imaging findings of nodules greater than 2 cm in diameter-had been previously issued by the EASL and United Network for Organ Sharing criteria. 28e can diagnose nodules greater than 2 cm in diameter as HCC if any 2 imaging studies, including US, CT, magnetic resonance imaging (MRI) or hepatic angiography, show increased vascularity.Alternatively, 1 positive imaging result and an AFP level greater than 400 ng/mL are diagnostic.0][31] This finding originated from previous investigations on the correlation between blood supply and the grade of malignancy of hepatic nodules. 32][35] Imaging for lesions less than 2 cm diameter and cases of indeterminate radiologic findings do not have sufficient accuracy to distinguish HCC from other conditions.Serial US or other radiologic examinations are recommended every 3 months or alternatively, fine needle biopsy can be performed.There is a small but definite risk (0.003%-5%) of tumor seeding from an invasive biopsy. 8urthermore, the false-negative rate from biopsy of lesions less than 2 cm diameter is approximately 30% to 40%. 8][38] As was already described above, AFP, AFP-L3, and PIVKA-II are the most commonly used serum tumor markers.Each marker alone has poor diagnostic ability for detecting HCC at any level of pretest risk, but can be useful in combination with imaging criteria.

Natural History of HCC
Prospective studies have shown that HCC most often develops through a progressive pathway from premalignant nodular lesions in the cirrhotic liver to cancerous lesions. 6The annual risk for HBV, HCV, and alcohol-induced cirrhosis are estimated at 2-3%, 1-8%, and 1%, respectively.HCC can also develop in the absence of cirrhosis in HBV carriers at a rate of 0.26% to 0.6% per year.
Several risk factors for the development of HCV-related HCC have been identified such as age, degree of liver fibrosis, male sex, alcohol consumption, positivity for HBsAg, and high AFP level.1][42][43] Recent large prospective studies have confirmed that coffee consumption decreases the risk of liver cancer, 4 4-4 6 however, it has not been determined that coffee drinking really decrease the incidence of HCC.Other risk factors for development of HCC, including obesity and type 2 diabetes, have been described elsewhere. 47,48 n our cohort, BMI significantly affected the survival of patients with HCC; notably, severe obesity (BMI 35-40) was associated with a 5-fold higher 16-year death rate. 47lthough treatment intervention was included in the course of patients with HCC, we analyzed which factors affected the survival of patients with HCC (Table 3).Among the entire cohort of patients, survival was significantly affected by age (<55, <65), HCV+/-, causes of HCC, tumor size, tumor number, Child-Pugh stage, TNM stage, CLIP score, JIS stage, BMI (<25), treatment modality, Milan criteria, total bilirubin levels (TB), albumin level (ALB), prothrombin time, AFP level, and PIV-KA-II level.Among patients with HCV infection, age (<55), tumor size (3 cm≤), and BMI did not significantly impact survival.These findings suggest that HCC patients with HCV infection are older and their tumor sizes are smaller as compared with patients with HCC associated with other etiologies.The tumors found in HCV-infected patients were smaller because these patients were intensely screened and the tumors were detected earlier.Multivariate analysis using Cox proportional hazard models showed that age (65≤), HCV infection, tumor number, TB (1.0≤), AFP (20≤) and PIVKA-II (40≤) were significant factors affecting survival among the entire patient cohort (Table 4a).Among patients with HCV infection, tumor number, age (65≤), tumor size, tumor number, TB (1.0≤), and AFP (20≤) were significant factors (Table 4b).Table 4a shows that survival rates were significantly higher among HCC patients with HCV infection versus other etiologies.This finding seems curious, but it suggests that HCC screening was routinely  performed for the HCV-infected patients and that this screening system worked well to improve survival.There were 56 patients with non-B, non-C (NBNC) HCC.Although number of these patients was insufficient to conclude the result, survival was significantly affected by age (65≤), gender and the disease staging (Table 5).
Tumor stage seemed to be more advanced among these patients as compared with cases of HCC associated with other etiologies, presumably because they had not been previously screened for HCC.But as shown in this table, there is no significant characteristic of the patients in this category (no significant high-risk factor); therefore, screening of these patients is very difficult.

Treatment and Prognosis of HCC
In Japan, the management of HCC involves multiple disciplines, including hepatology, surgery, diagnostic and interventional radiology and pathology.Liver transplantation has not been frequently performed so far in Japan, and surgery and loco-regional treatment, such as ablation, remain the 2 major curative modalities for earlystage HCC.Before choosing a modality of treatment, tumor staging should be evaluated precisely.HCC staging is also important for the prognosis of HCC.Two disease processes, namely, the tumor and the underlying cirrhosis, complicate the prognosis of HCC simultaneously, and many studies have suggested that cirrhosis rather than the size of the tumor mass is the main determinant of the prognosis.The median survival time of untreated patients with newly diagnosed HCC is weeks to months in Western countries, where screening of HCC only recently became popular.Survival time could be improved by screening for HCC among high-risk groups.A number of factors are associated with a poor outcome: male sex, advanced age, etiologic agent, presence of more than 1 risk factor, tumor size, tumor number, doubling time of nodules, vascular invasion, and distant metastasis.The complicated and heterogeneous nature of HCC has led to the development of a variety of staging systems.
Overall survival curve in our cohort is shown in Figure 6a.One-, 3-, and 5-year survival rates were 86%, 64%, and 44%, respectively, which were higher than those obtained from the Japanese nationwide surveillance (year 2002-3) made by the Liver Cancer Study Group of Japan (LCSGJ) including 96,404 HCC cases.Figure 6b shows the survival curve in patients with chronic HCV infection.The survival rates at 1-, 3-, and 5-year were 91%, 70%, and 49%, respectively, which were higher than those of whole patients.This result also indicates that screening of patients with chronic liver diseases makes better outcome.The measurement of serum AFP and PIVKA-II levels is an easy way of screening.In our case series, elevation of serum levels of AFP (>20; normal limit) (Fig. 7a) and PIVKA-II (>40; normal limit) (Fig. 7b) significantly affected the survival of patients with HCC.It was suggested that AFP levels more than 400 ng/mL and/or PIVKA-II level more than 400 mAU/mL were significant bad prognostic markers.The 5-year survival rates of these patients were around 20%.
The Tumor Node Metastasis (TNM) classification, mostly used for solid tumors, has extreme limitations because it does not include the severity of underlying cirrhosis or residual liver function.Many prognostic staging systems have been established; the first attempt by Professor Okuda in Japan, 49   pan Integrated Staging Score (JIS).It remains controversial as to which of these systems is best for staging HCC.
The TNM system is popular for surgical patients but this system is not suitable for staging HCC because it does not include variables related to liver function.This system was revised in 2002 and the new version has been demonstrated to be more accurate. 50,51 he prognosis of patients with HCC in our cohort was determined by Kaplan-Meyer analysis and it differed significantly according to TNM staging (Fig. 8a) and Child-Pugh classification (Fig. 8b).These findings suggest that the prognosis of patients with HCC depends not only on tumor progression but also on reserved liver function.The Okuda staging system 49 was the first staging system to include both tumor stage and factors related to liver function, such as tumor size, ascites, bilirubin level and albumin level.However, while commonly accepted in Western countries until a few years ago, this system has proven to be inadequate for patients with early HCC.The system was first proposed in 1985 based on a retrospective cohort 850 patients in Japan, where cases were typically advanced at the time of initial diagnosis.It failed to include recently identified factors of importance such as vascular invasion, portal vein thrombosis, and AFP levels.The CLIP system 5 2 includes many fac-tors and was found to be useful even in Japan 53 , but the inclusion criteria are too broad, which would reduce its value in cases of early HCC.The prognosis of our cohort was well correlated with this staging system (Fig. 9).However, the prognosis of patients in advanced disease, such as CLIP scores 4-6, did not well differentiated each other.The BCLC system, 2 8 derived from the results of surgical treatment of early HCC and the natural history of untreated HCC, has also been validated by several groups worldwide, 5 4-5 6 and in patients who underwent ablation. 57The CUI system was established based mainly on HBV-associated patients but only 10% of the patients received curative treatment, and it is considered to be complicated and non-validated. 58The GRETCH system was based on a study in France which included mainly patients with alcoholic liver disease; 59 its validity appears to be poor. 5 5, 5 6, 6 0, 6 1Lastly the JIS system was proposed in 2003 by Kudo et al. using a combination of the Child-Pugh score and the modified TNM classification according to LCSGJ. 62This system is more suitable for patients with early HCC, and a recently modified JIS system proposed by Nanashima et al. 63 was shown to be useful for predicting the prognosis of patients receiving resection. 64In Japan, patients with HCC, who always receive screening at the high-risk stage and the tumor is al-  ways diagnosed in its early phase, may be best evaluated using the JIS system.Patients with HCC in our cohort were analyzed according to the JIS scoring system and it showed that the prognosis was well distinguished, especially in the early stages (Fig. 10).It was suggested that the prognosis of HCC is better predicted by the JIS scoring system than the CLIP scoring system in our cohort or Japanese patients with HCC (Fig. 9 and 10).A treatment modality is selected according to a proposed algorithm for the treatment of HCC. 10,65 ery early HCC, presenting with a single lesion <2 cm, is treated with resection and radiofrequency ablation (RFA), which both offer similar 5-year survival rates.RFA is less invasive and is associated with fewer complications and lower cost. 66The choice of therapy depends on the tumor location, degree of liver function, and the presence of comorbidities.Liver function is preserved in early HCC when there is a solitary nodule or up to 3 nodules, each ≤ 3 cm diameter.Such patients are candidates for liver transplantation according to the Milan criteria 67 but they can also be treated by resection or RFA with the possibility of a long-term cure.Patients with compensated cirrhosis and without HCC-related symptoms or vascular invasion who fall outside of the criteria for very early or early HCC (intermediate stage) are always treated with transarterial chemoembolization (TACE), which leads to a 23% improvement in the 2-year survival.Down-staging by single or combined therapy with TACE and RFA greatly improves posttrasplantation outcome if the patients are carefully selected. 68Patients with vascular invasion or extrahepatic spread are considered to be at an advanced stage and TACE may increase survival among some of these patients.However, insufficient treatment by TACE seems to increase in the expression of vascular endothelial growth factor (VEGF) and/or basic fibroblast growth factor (b-FGF) in the liver, inducing the possibility of an active angiogenesis and an extended invasiveness of HCC. 6 9 We compared the survival rates of patients in JIS score 0-1 (n=128) according to the first treatment modality (Table 6).Five-year survival rates of Fig. 6 Overall survival rates for whole patients (a) and patients with chronic hepatitis C virus (HCV) infection (b).One year, 3-year, and 5-year survival rates of whole patients were 86%, 64%, and 44%, respectively, while those of HCV-infected patients were 91%, 70%, and 49%, respectively.patients received hepatic resection, RFA, TACE, and percutaneous ethanol injection (PEI) were 89.3%, 70.8%, 44.6%, and 43.2%, respectively.
][72][73][74][75] A recent RCT showed that sorafenib, a multikinase inhibitor of raf kinase, VEGF2, VEGF-receptor (R)3, and platelet-derived growth factor receptor (PDGFR)-β kinases, improved overall survival by about 3 months among patients at BCLC stage C as compared with placebo. 76Another option may be focused radiation therapy (hypofractionated stereotactic radiotherapy), which has also been considered recently for the treatment of HCC, especially in cases with portal thrombosis or lesions just under the diaphragm. 7 7Carbon ion radiotherapy has been utilized in some institutions in Japan; a favorable result was obtained but the procedure is still restricted to a small proportion of HCC. 78Interferon (IFN) is an antitumor drug candidate and is used as systemic treatment for HCC due to chronic HCV-or HBV-infection.0][81][82][83][84] Recent studies reported favorable clinical outcomes with IFN usage in cases of advanced-stage HCC. 47,85,86 Terapies combining anti-cancer agents, delivered systemically or transarterially, in combination with TACE or that without embolization (TAI) have been utilized in Japan under institutional ethical approval, but no clear evidence has been reported.

The Future of HCC Treatment
The mechanisms of HCC pathogenesis are gradually clarifying.Several tumor suppressors or protooncogenes are known to be mutated in HCC, including p53, Rb, and insulin-like growth factor (IGF), 4 8, 8 7 and intracellular signaling pathways, such as VEGF, endothelial growth factor (EGF), and their down-stream signaling cascades, have been shown to contribute to the growth of HCC. 88enome-wide analysis of gene transcription in the liver has recently revealed that some gene expression profiles consistently change from chronic hepatitis to HCC, 8 9 suggesting that some key molecules to accelerate or inhibit tumor development might be discovered in future.Imatinib, an inhibitor of several tyrosine kinases, is another agent that targets hepatic cancer progenitor cells. 90hese new agents are very encouraging candidates for the treatment for HCC in the future.Immunologic stimulation may be another future therapeutic candidate for HCC. 91It has also been demonstrated that it is possible to predict HCC based on expression-profiling of formalin-fixed, paraffin-embedded liver tissue. 9 2Several possible preventive agents for HCC have been identified, such as retinoids, 9 3-9 5 branched-chain amino acid supplementation, 9 6 vitamin K, 9 7 and anti-viral therapies, 9 8-1 0 2 and these agents may help to overcome the difficulties in treating advanced cases of HCC.A recent investigation has indicated that therapy to resolve hepatic fibrosis may also be useful for preventing the development of HCC. 1 0 3 Among these various therapeutic modalities, resolution of liver cirrhosis using heat shock protein 47 is the most promising in my opinion, 1 0 4 because this therapy may not only prevent HCC development but also improve liver function by resolution of liver cirrhosis.
The overall survival of patients with HCC continues to decrease from year to year, and does not show a plateau phase in Kaplan-Mayer curve, as was shown in our cohort, which is trend that is quite different from that observed with other types of cancer.These results suggest that the best way to improve survival of patients with HCC through prevention of the disease.It has been confirmed that IFN therapy for chronic HCV infection decreases the incidence of HCC.Nucleoside or nucleotide analogues for chronic HBV infection also decrease the incidence of HCC.This outcome has been encouraging to hepatologists and we continue to treat patients with chronic liver diseases as we have done for chronic hepatitis C during the last few decades [105][106][107][108][109][110][111] until HCC is finally eradicated.Table 6 The survival rates of patients with HCC in the JIS score 0-1 RFA: radio frequency ablation therapy, TACE: trasarterial chemo-embolization, PEI: percutaneous ethanol injection therapy.

Fig. 2 Fig. 3
Fig. 2 Causes of hepatocellular carcinoma in our cohort.Fig. 3 Proportion of cases with liver diseases in patients admitted to Keio University Hospital during the last 4 years.HCC: hepatocellular carcinoma, LC: liver cirrhosis, CH: chronic hepatitis, AH: acute hepatitis, ALC: alcoholic liver diseases, Meta: tumor metastasis from other organs, LD: liver dysfunction, NAFLD: non-alcoholic fatty liver diseases, DILI: drug-induced liver injury, AIH: autoimmune hepatitis.

Fig. 1
Fig. 1 Age distribution of patients with hepatocellular acrcinoma admitted in the affiliated hospitals of Keio University.

Fig. 5
Fig. 5 Typical small hepatocellular carcinoma detected by computed tomography scanning.The tumor is stained at arterial phase (a) and washed out early at late phase (b).Allows show the tumor.

Fig. 7
Fig. 7 Overall survival rates according to alpha-fetoprotein (AFP) levels (a) and vitamin K antagonist (PIVKA)-II levels (b).Statistically significant difference was observed among 4 categories in AFP (p<0.0001 by log-rank test) and among 3 categories in PIVKA-II (p<0.0001 by log-rank-test).

Fig. 9
Fig. 9 Overall survival rates according to Cancer of Liver Italian Program (CLIP) scores in whole patients.Statistically significant difference was observed among 7 categories (p<0.0001 by log-rank-test).

Fig. 10
Fig. 10 Overall survival rates according to Japan Integrated Staging (JIS) scores in whole patients.Statistically significant difference was observed among 6 categories (p<0.0001 by log-rank-test).

Fig. 8
Fig. 8 Overall survival rates according to Child-Pugh grades (a) and TMN stages (b) in whole patients.Statistically significant difference was observed among categories in both systems (p<0.0001 by log-rank test).

Table 2
Baseline characteristics of the patients

Table 3
Investigated factors and their significance in correlation to the survival of patients with HCC (by log-rank test) HCV: hepatitis C virus antibody, HBsAg: hepatitis B virus antigen, BMI: body mass index, DM: diabetes mellitus, TB: serum total bilirubin level, Alb: serum albumin level, PT: prothrombin time, ICG: indocyanin green test.: different between whole patients and HCV infected patients.

Table 4
Significant factors affected survival of patients with HCC found by multivariate analyses HCV: hepatitis C virus antibody, TB: serum total bilirubin level.

Table 5
Investigated factors and their significance in correlation to the survival of patients with non-B, non-C HCC HCV: hepatitis C virus antibody, HBsAg: hepatitis B virus antigen, BMI: body mass index, DM: diabetes mellitus, TB: serum total bilirubin level, Alb: serum albumin level, PT: prothrombin time, ICG: indocyanin green test.: Significantly different factors.