Hepatocellular carcinoma (HCC) is a major cause of cancer death in many Asian countries^1,2). Since almost all patients have extensive intrahepatic spread of the tumor and/or metastatic disease at the time of diagnosis, the disease is usually unresctable. Although some of these patients were treatable surgically, the disease is actually too extensive to perform surgical attempt or which will recur after resection. Therefore, they might be considered candidates for systemic or regional therapy. Until now, several reports for systemic chemotherapy of HCC had shown an unsatisfactory response of less than 10%^3,4). Therefore, the development of new active anticancer agents is essential, Cis-platinum has a broad spectrum of antineoplastic action and there has been many reports demonstrating favorable effects for the treatment of various malignant diseases^5,6). Interferon has been shown to have a powerful antiproliferative effect on the human hepatoma cell line^{7, 8)}. In this study, we report our experience of IFN-α plus cis-platinum treatment in 30 patients with HCC. The current study was also undertaken to evaluate the survival time of patients with HCC and find prognostic factors which allow the selection of patients with a life expectancy long enough to undergo therapy.

Of the 30 patients treated with IFN-α plus cis-platinum, 4 patients (13.3%) showed partial response, the median survival time and cumulative survival rate at 3 months and 6 months were 33.0 weeks and 93.5%, 75.0%, respectively. When compared with IFN-α plus cis-platinum group, no antitumor therapy group showed that the median survival time and cumulative survival rates were 14.0 weeks and 84.6%, 57.7% at 3, 6 months respectively. This study demonstrated that IFN-α plus cis-platinum was of value in prolongation survival and inducing tumor regression in patients with inoperable HCC. In an earlier study of the natural history of patients with inoperable HCC, the median survival in the group receiving no antitumor therapy was only 3.5 weeks¹⁴⁾. The longer survival of the group receiving no antitumor therapy in this study is most likely attributed to good Karnovsky score status above 70%. But the median lengths of survival of Okuda¹⁵⁾ and Attali¹⁶⁾ in untreated patients with HCC were 16.0 weeks and 11.0 weeks, respectively, which were not different from that of this study. Although this study was made retrospectively, these results made it possible to evaluate the clinical effect of IFN-α plus cis-platinum and the selection of patients was considered to be unbiased.

Until recently, though the treatment of HCC remained still unsatisfactory. Adriamycin has been the most frequently administered drug for treatment of HCC among various chemotherapeutic modalities^{17, 18)}. Several reports demonstrated the response rates of Adiramycin to HCC have varied over a wide range from 15–34%^19,20). It seemed that the comparison of variable results was difficult because dosage and definition of response differed widely. Recently. Johnson²⁰⁾ reported that Adriamycin induced tumor regression was 3.3%–10% in patients with HCC and similar results were reported in many studies^{21, 22)}. But, there have been many reports concerning the high incidence of fatal side effects, such as cardiotoxicity^{23, 24)}. Therefore. alternative therapy for treatment of HCC is essentially needed. Interferon has been shown to have a powerful antitumor effect, correlated with its antiproliferative against a human hepatoma cell line in vivo and vitro⁷⁾, and immunomodulatory properties⁸⁾. In clinical study, IFN-α induced tumor regression of 7–41% was observed²⁵⁾. Cis-platinum has a broad spectrum of antineoplastic action correlated with inhibition of DNA synthesis²⁶⁾ and there has been several reports demonstrating favorable effects⁵⁾. The response rate of cis-platinum to HCC was 8.8–15.4%⁶⁾. Compared with previous trials, showing Adriamycin plus IFN-α therapy induced tumor regression of 50% in only 3.3% of patients with HCC²⁷⁾, a 13.3% objective regression rate in this study, anticipating synergistic effect for IFN-α plus cis-platinum combination. was shown to be superior to Adiramycin plus IFN-α therapy. In fact, the response rate of Adriamycin plus IFN-α was actually less than what would be anticipated from reported experiences using either drugs alone. In context, Adriamycin plus IFN-α therapy seems that there was no synergistic interaction between two drugs for treatment of HCC. But IFN-α and cis-platinum therapy might seems that there is synergistic effect. The reason for the difference in the response rate, median survival time and survival rate between IFN-α plus cis-platinum and previous several trials was unknown. Several possible explanations exist. First, the number of patients evaluated in the previous study, involving Adriamycin plus IFN-α therapy, was relatively small compared with those in this study. Second, there may be an ethnic and/or regional difference in the biological characteristics of HCC itself in the tumor responsiveness to therapy. Previous studies in North American²⁸⁾ or Chinese patients^{24, 25)} showed no benefit or minimal beneficial effect of IFN-α or cis-platinum. Third, the longer survival rates and median survival time in this study, we would like to suggest three points. First, it was mainly in the treated group that significant prolongation of survival was observed. Second, despite the terminal progression of tumors, four patients of these treated group showed significant tumor shrinkage compared with the radiologic dimensions of the tumors before therapy. Third, although we observed that IFN-α plus cis-platinum increased the survival of patients with HCC, we found no objective evidence of correlation between tumor regression and survival. However, despite the absence of a benefit between tumor regression and survival, it should be noted that the IFN-α plus cis-platinum achieves partial response in a few patients and this prompts slower tumor progression, which may delay the appearance of tumor-related symptoms. Our observation also suggests that survival is probably more related to the advanced status of the disease of these patients on entry into the study than to progression of the tumor.

This study was also carried out to evaluate the factors affecting survival, thus enabling us to confirm important prognostic factors, predict the long them outcome and plan future clinical trials of the therapy of HCC. In our univariate analysis, 4 of 12 variables tested had a significant prognostic value, such as the absence of portal vein thrombus, alkaline phosphatase less than 280 U/L, total bilirubin less than 2.0 mg/dl and serum triglyceride less than 155 mg/dl showing significantly longer survival using log-rank tests.

Age, sex, tumor size, metastasis, presence or absence of cirrhosis, serum albumin more than or equal to 3.5 gm/dl, no ascites and serum AFP less than 400 ng/ml had no significant effect on survival. Recently, Sutton and co-workers²⁹⁾ reported that age, sex, ascites and cirrhosis were found to have a prognostic value in their analysis. In our study, neither ascites nor cirrhosis affect the survival of HCC. This is unbelievable because ascites and cirrhosis in patients with HCC were usually considered poor prognostic variables³⁰⁾. The reason why ascites or cirrhosis did not predict survival remains obscure. The absence of the prognostic value of ascites or cirrhosis might be due to the small number of cirrhosis or the late stage of HCC in this study. AFP did not appear in this study to be a parameter significantly related to survival. The prognostic value of AFP levels has already been debated in the past with different conclusion^31–33). A close relation between AFP levels and the tumor burden has been denied by Cotton³²⁾ and Ebara³¹⁾, whereas Sheu³¹⁾ found a significant correlation between the progressive rise of AFP and the increase of the neoplastic mass. In this study, we confirmed that there was no statistical correlation between AFP and tumor progression.

Our multivariate analysis showed that only three factors, the treated group with IFN-α plus cis-platinum, portal vein thrombus and alkaline phosphatase had an independent prognostic variable to survival. Chlebowski²⁸⁾ reported that serum bilirubin affects prognosis of HCC. Calvet³⁴⁾ also reported that serum bilirubin was a predictor of survival. It is generally agreed that elevated serum bilirubin level has been shown to adversely influence th survival of patients with HCC. regardless of treatment. In this study, serum bilirubin was not found to be significant in multivariate analysis. However, bilirubin was found to be significant in the univariate analysis. Although not all the variates identified as significantly related to survival, using univariate analysis, this discrepancy may have occurred because the influence of the individual factor on each other was not in account for the univariate analysis but was in the multivariate analysis.

Hypercholesterolemia is an important paraneoplastic manifestation in patients with HCC³⁵⁾. The incidence of Hypercholesterolaemia in HCC patients has been reported as varying from 11 to 38%³⁶⁾.

According to Hwang³⁷⁾, who reported that hypercholesterolemia or hypertriglyceridemia HCC patients tended to have large tumor volume, none of the small HCC patients had high cholesterol or triglyceride levels, and serum triglyceride levels in HCC patients fell to the normal range after surgical resection or chemotherapy of the tumor, and rose to abnormal level again after tumor recurrence.

These findings may indicate that re-elevated serum cholesterol or triglyceride levels in HCC patients were associated with tumor recurrence and the presence of a viable tumor mass.

Although the exact mechanism of hypercholesterolemia or hypertriglyceridemia in HCC patients has not been fully understood, by the view of Hirayama³⁸⁾, he postulated that there may be autonomous cholesterol or triglyceride biosynthesis inn the tumor cell, followed by cholesterol or triglyceride release into the circulation, causing hypertriglyceridemia or hypercholesterolemia.

Therefore, serum cholesterol or triglyceride levels may be a clinically important prognostic significancy in therapy with IFN-α and cis-platinum.

We, thus, conclude that IFN-α and cis-platinum combination therapy may lead to higher response rates and the improvement of survival for inoperable HCC patients and above mentioned favorable prognostic factors may be also be useful in the design and analysis of future clinical trials for systemic chemotherapy of HCC.