Taichi Shuto

Effects of Long-Term Postoperative Interferon-α Therapy on Intrahepatic Recurrence after Resection of Hepatitis C Virus–Related Hepatocellular Carcinoma: A Randomized, Controlled Trial

Hepatitis C virus (HCV) is an important cause of hepatocellular carcinoma in many areas of the world (1). The recurrence rate at 5 years after resection of HCV-related hepatocellular carcinoma has been reported as 70% to 80% (2-4). Recurrences after resection of hepatocellular carcinoma may be either metastases from the primary carcinoma or new foci of carcinomas (multicentric occurrence) (4-8). Chronic active hepatitis is a risk factor for recurrence, including multicentric carcinogenesis, after resection (4, 7, 8). In addition, the recurrence rate after the resection of HCV-related hepatocellular carcinoma is higher in patients with HCV viremia than in those without viremia (4). Interferon treatment decreases the incidence of hepatocellular carcinoma in patients with chronic hepatitis C (9-16). We conducted a randomized, controlled trial to evaluate whether postoperative therapy with interferon- would decrease the incidence of recurrence after resection of HCV-related hepatocellular carcinoma. Methods Patients and Trial Profile Between November 1993 and March 1997, 112 HCV-positive patients underwent resection for hepatocellular carcinoma at Osaka City University Hospital or Osaka City General Hospital, Osaka, Japan. Seventy-seven patients met the eligibility criteria: 1) single tumors less than 5 cm in maximum diameter on preoperative imaging; 2) detectable HCV RNA without hepatitis B surface antigen or HIV antibodies; 3) chronic hepatitis C or a ChildPugh score of A or B for compensated cirrhosis [17]; and 4) no severe thrombocytopenia (platelet count < 50 109 cells/L). Before resection, 31 of the 77 eligible patients gave written informed consent to participate in the trial; however, one patient was later excluded because resection was not curative. Forty-six patients declined to participate because they were unwilling or unable to agree to twice-weekly hospital visits for 2 years or were concerned about widely publicized side effects of interferon use. A curative operation was defined as complete resection of all macroscopically evident tumor tissue. Specifically, no tumors could be detected in the remnant liver on computed tomography performed 3 to 4 weeks after resection. In the one patient excluded from the trial after resection, a residual tumor was found on computed tomography 3 weeks after resection. In all, 30 patients were enrolled and randomly allocated to the interferon- group (n = 15) or a control group (n = 15). Randomization was done by using a random-numbers table, and patient assignments were withheld from the investigators. During the trial, no patient in the control group received any type of treatment; no patient in the interferon group received chemotherapy or any treatment other than interferon-. This study was done in accordance with the Helsinki Declaration and was approved by the ethics committees of our institutions. Interferon- Treatment Patients received 6 MIU of interferon- (human lymphoblastoid interferon, Sumiferon, Sumitomo Pharmaceuticals, Osaka, Japan) intramuscularly every day for 2 weeks, then 3 times weekly for 14 weeks, and finally twice weekly for 88 weeks (total dose, 1572 MIU). Interferon- therapy was started 5 to 15 weeks (mean, 9 weeks) after resection. In 1 of the 15 patients in the interferon- group, treatment was delayed until 7 months after resection at the patients request. Laboratory tests were done at least once monthly during interferon- therapy and at least once every 3 months thereafter for evaluation of the response to therapy and identification of side effects of interferon-. Serum HCV RNA was detected by using a method reported previously (9), and viral load was measured by using a branched-DNA probe assay (Quantiplex HCV-RNA, Chiron Corp., Emeryville, California). Response to Interferon- Therapy A complete response was defined as the absence of serum HCV RNA (virologic remission) and serum alanine aminotransferase (ALT) activity within the reference range ( 750 nkat/L [ 45 U/L]) for at least 6 months after interferon- therapy (biochemical remission) (9). A biochemical response was defined as a decrease in serum ALT activity to within the reference range but with persistently detectable serum HCV RNA. Nonresponse was defined as persistence of HCV RNA and no decrease in ALT activity. Detection of Recurrence Serum concentrations of -fetoprotein and protein induced by vitamin K absence and antagonist II were measured within 2 months of resection and every 3 months thereafter. Ultrasonography, computed tomography, magnetic resonance imaging, or some combination of these tests was done within 2 months after the operation and every 3 months thereafter. These examinations were done until the detection of recurrence or until the trial end point (final examination). When tumor recurrence was suspected on the basis of tumor markers, imaging, or both, angiography, ultrasonography-guided biopsy, or both were done to establish a definitive diagnosis. The radiologists who were responsible for diagnosis of tumor recurrences had no contact with the patients and had no information about the trial. Statistical Analysis Time to recurrence was measured from the time of resection to the detection of a recurrent tumor. The recurrence rates (including data on patients who could not complete interferon- therapy) were calculated by using the KaplanMeier method, and significance of between-group differences was assessed by using the log-rank test. All analyses were done by using SAS statistical software, version 6.12 (SAS Institute, Inc., Cary, North Carolina). Role of the Funding Source The funding sources had no role in the collection, analysis, and interpretation of the data or in the decision to submit the paper for publication. Results Most baseline variables were similar in the two groups, as were the operative procedures (Table). The patients who declined participation and the patients who were enrolled were similar with regard to most variables. Table. Patient Demographic and Clinical Characteristics Interferon- could not be administered to one patient because of premature ventricular contractions. Data on this patient were included in the interferon- group for all analyses. Interferon- administration could not be completed in 3 patients because of depression (n = 1, at 17 days), severe general fatigue (n = 1, at 11 months), and renal abscess (n = 1, at 10 months; this patient also had a recurrence detected in the same month). In 4 patients with recurrence, interferon- administration was stopped prematurely to allow treatment of the tumor. In the interferon- group, 2 patients were complete responders, 6 patients were biochemical responders only, and 7 patients were nonresponders. Among the control patients, postoperative serum ALT activity (except immediately after resection) was greater than the reference range, and serum HCV RNA was detected at the follow-up appointments. The median follow-up period (from the operation to the detection of recurrence or to the study end point) was 1087 days (25th and 75th percentiles, 514 and 1376 days) for patients receiving interferon- and 897 days (25th and 75th percentiles, 439 and 1105 days) for the controls. In all patients, examinations for detection of recurrence were done according to schedule. All patients without recurrence were still alive at the end of this trial. Recurrent tumors were detected in 5 patients (including the patient who could not tolerate interferon- therapy) in the interferon- group and in 12 control patients. In the interferon- group, recurrence was definitively diagnosed in 3 patients by using angiography and in 2 patients by using angiography and biopsy. In the control group, definitive diagnoses were made in 8 patients by using angiography, in 3 patients by using biopsy, and in 1 patient by using both methods. No recurrence was detected in either of the 2 complete responders in the interferon- group. One of the 6 biochemical responders and 4 of the 7 nonresponders had recurrent tumors. The recurrence rates increased along similar curves in the two groups in the 2 years after resection; thereafter, 6 controls but no patient receiving interferon- had recurrence (Figure). The recurrence rate was significantly lower in the interferon- group than in the control group (P=0.037). Figure. Recurrence rates in the interferon- ( solid line ) and control ( dotted line ) groups. P Discussion Persistent active hepatitis is a risk factor for development of hepatocellular carcinoma, indicating that treatment of active hepatitis may be important in the management of patients with HCV-related hepatocellular carcinoma (4). In our study, absence of interferon- therapy was a risk factor for recurrence. Although for the first 2 years of follow-up the recurrence rate increased for the two groups in similar curves, the recurrence rate in the interferon- group did not increase after that time. This finding suggests that patients in the interferon- group had been truly cured. Recurrent tumors detected within 2 years of the operation were likely to be either metastases from the primary carcinoma or carcinoma that appeared before or during interferon- therapy but had gone undetected at the operation. Our findings indicate that interferon- therapy does not suppress carcinoma itself. On the other hand, recurrences that appear more than 2 years after resection are likely to have arisen because of new carcinogenesis. Recurrence was detected in 1 of 8 patients with a biochemical remission (including the 2 complete responders) and in 4 of the 7 nonresponders. Recently, several investigators have shown that hepatocellular carcinoma is less likely to develop in patients in whom interferon- decreases ALT activity to within the reference range, even if HCV is not eradicated, probably because of decreased hepatocyte necrosis and liver fibrosis (9, 11, 12, 14-16). Thus, eradication of HCV may not be necessary for reduced risk for recurrenc

Cytokeratin 19 expression in hepatocellular carcinoma predicts early postoperative recurrence.

Clinicopathologic features and postoperative outcomes were investigated for patients who underwent curative surgery for biliary marker (CK7 and CK19)‐positive hepatocellular carcinoma (HCC). Of 157 HCCs, 93 were CK7(‐)CK19(+), 49 were CK7(+)‐CK19(‐), 1 was CK7(‐)CK19(+), and 14 were CK7(+)‐CK19(+). Semiquantitative analysis of expression levels demonstrated a significant correlation between CK7 and CK19 expression. Of various clinicopathologic parameters, tumor differentiation exhibited a significant correlation with CK7 and CK19 expression. All 15 patients with CK19‐positive HCC also had anti‐HBc. Log‐rank test revealed that CK7 expression, CK19 expression, high aspartate aminotransferase (AST) activity, low albumin concentration, portal invasion, intrahepatic metastasis, and severe fibrosis (cirrhosis) reduced the tumor‐free survival rate. Multivariate analysis demonstrated that CK19 expression, intrahepatic metastasis, and severe fibrosis were independent predictors of postoperative recurrence, while CK7 expression was not. Twelve of 15 patients with CK19‐positive HCC had tumor recurrence within 2 years after surgery, a significantly higher incidence of early recurrence than for CK19‐negative HCC. The incidence of extrahepatic disease, especially lymph node metastasis, was significantly higher for patients with CK19‐positive HCC. These findings indicate that CK19 expression is a predictor of early postoperative recurrence due to increased invasiveness.

Hepatitis C virus infection as a likely etiology of intrahepatic cholangiocarcinoma

Although hepatitis C virus (HCV)‐related cirrhosis has been suggested as a risk factor for intrahepatic cholangiocarcinoma (ICC), few sizeable studies have tested this hypothesis. We investigated ICC risk factors, with special reference to HCV infection. We conducted a hospital‐based case‐control study including 50 ICC patients and 205 other surgical patients without primary liver cancer. HCV seropositivity was detected in 36% of ICC patients and 3% of controls. By univariate analysis, the odds ratio (OR) for association of anti‐HCV antibodies with development was 16.87 (95% confidence interval (CI), 5.69 to 50.00). History of blood transfusion or diabetes mellitus, elevated serum total bilirubin, elevated aspartate aminotransferase and alanine aminotrans‐ferase, decreased serum albumin and decreased platelet count were identified as other possible ICC risk factors. By multivariate analysis, anti‐HCV antibodies (adjusted OR, 6.02; 95% Cl, 1.51 to 24.1), elevated alanine aminotransferase, decreased serum albumin, and decreased platelet count were found to be independent risk factors for ICC development. As liver status worsened, the adjusted OR for ICC tended to increase. HCV infection is a likely etiology of ICC in Japan.

Incidence and management of Bile leakage after hepatic resection for malignant hepatic tumors

BACKGROUND Bile leakage is one of the frequent and disturbing complications of hepatic resection. STUDY DESIGN Clinical records of the 363 patients who underwent hepatic resections without biliary reconstruction for hepatic cancers between January 1994 and June 2001 were reviewed. Postoperative bile leakage was defined as continuous drainage with a bilirubin concentration of 20 mg/dL or 1,500 mg/d lasting 2 days. Leakage that continued longer than 2 weeks or that required surgical intervention was defined as uncontrollable. Differences in incidence and frequency of uncontrollable leakage for the different types of hepatic resection, tumors, and underlying liver disease were investigated. Outcomes after treatment for uncontrollable bile leakage were also reviewed. RESULTS Postoperative bile leakage occurred in 26 of 363 patients (7.2%). Although the incidence in patients with cholangiocellular carcinoma (3/9 [33%]) was higher (p = 0.03) than in patients with hepatocellular carcinoma, rates of occurrence were similar among the different types of hepatic resection and underlying liver disease. Eight of the 26 patients (31%) had uncontrollable leakage. Two patients required reoperation to control leakage; one of these developed hepatic failure and died 2 months after surgery. Four patients underwent endoscopic nasobiliary drainage 21 to 34 days after hepatectomy, and the leakage resolved within 3 to 21 days. Fibrin glue sealing was effective in two patients whose leaking bile ducts were not connected to the common bile duct. CONCLUSIONS Although meticulous surgical technique can minimize the risk of postoperative bile leakage, some instances of leakage are unavoidable. Nonsurgical treatments, such as nasobiliary drainage or fibrin glue sealing, are preferable to reoperation.

Risk factors for postoperative delirium after liver resection for hepatocellular carcinoma.

We investigated risk factors for delirium in 100 patients who underwent liver resection for hepatocellular carcinoma. Postoperative delirium developed in 17 (17%). Univariate analysis revealed that advanced age (especially = 70 years old), a history of smoking, a decreased serum albumin concentration (especially < 3.8 g/dl), advanced cancer stage (II–IV), major hepatectomy, prolonged operating time, and large intraoperative blood loss were possible risk factors for postoperative delirium. When patients’ preoperative condition and laboratory test results were subjected to multivariate analysis, only advanced age [odds ratio (OR) 1.201; confidence interval (CI) 1.063–1.357] and a decreased serum albumin concentration (OR 0.151; CI 0.025–0.900) were independent risk factors for the delirium. The percentages of patients with high aspartate and alanine aminotransferase activities, a high indocyanine green retention rate at 15 minutes, a low platelet count, and advanced cancer stage (II–IV) were higher in patients with a low (< 3.8 g/dl), rather than high (= 3.8 g/dl) serum albumin concentration. These findings indicate that multiple factors, including advanced age, impaired liver function, and advanced cancer stage, affect the development of postoperative delirium after liver resection for hepatocellular carcinoma.

Risk Factors for Recurrence after Resection of Hepatitis C Virus-related Hepatocellular Carcinoma

Although there have been many studies of the risk factors for recurrence after resection of hepatocellular carcinoma (HCC), the subjects were patients with various viral status in the previous studies, and hepatitis C viremia has not been evaluated. We investigated risk factors, including hepatic C viremia and histologic findings of noncancerous hepatic tissue, for recurrence after resection of hepatitis C virus (HCV)-related HCC. A total of 223 patients who underwent liver resection for HCV-related HCC were studied. HCV viremia, laboratory data, degree of HCC malignancy, histologic findings in noncancerous hepatic tissue, preoperative interferon therapy, and operative methods were evaluated for recurrence risk by univariate and multivariate analyses. Serum levels of aspartate aminotransferase (AST), alanine aminotransferase (ALT), and total bilirubin, and the proportion of patients with a high histologic activity score (mild to severe active hepatitis) were significantly higher in patients with HCV viremia than in those without viremia. Serum albumin was significantly lower in patients with HCV viremia. By univariate analysis, older age (> 65 years old), HCV viremia, elevated AST (> 40 IU/L) and ALT (> 45 IU/L), large tumors (> 40 mm), multiple HCCs, moderately or poorly differentiated HCC, portal invasion, mild to severe active hepatitis, and lack of preoperative interferon therapy were risk factors for recurrence. Multivariate analysis showed that older age, HCV viremia, high AST, multiple HCCs, and portal invasion were independent risk factors. For HCV-related HCCs, not only the degree of maliganacy of the HCC but also HCV viremia and active hepatitis are risk factors for recurrence.

Surgical management of cystic hepatic neoplasms.

Background. The clinicopathologic features of cystic neoplasms of the liver, such as cystadenoma and cystadenocarcinoma, are poorly defined because these lesions are extremely rare. Methods. Clinicopathologic findings in six patients who underwent surgery for a cystic hepatic neoplasm were reviewed retrospectively to determine the best surgical strategy to treat this condition. Results. Five patients had a multilocular cyst or elevated lesions detected on preoperative imaging studies. The remaining patient had elevated serum concentrations of carcinoembryonic antigen and carbohydrate antigen 19-9, even though preoperative imaging studies of the hepatic cystic lesion revealed no imaging features of cystadenocarcinoma. Cytology failed to detect malignant cells in the cyst contents of any patient. One patient underwent palliative resection because pleural dissemination was present, and five patients underwent curative resection. Three patients had cystadenoma diagnosed in the surgical specimen. One patient with cystadenocarcinoma has survived for more than 8 years after a curative resection that was limited to the cyst wall, and another patient died of recurrence 13 months after surgery. Conclusions. When a cystic neoplasm is suspected of being cystadenocarcinoma, hepatic resection should be considered, because reliable criteria for distinguishing cystadenoma from cystadenocarcinoma clinically do not exist.

The clinical significance of lymph node metastases in patients undergoing surgery for hepatocellular carcinoma.

The frequency of lymph node (LN) metastasis in patients undergoing surgery for hepatocellular carcinoma (HCC) has rarely been studied. We evaluated the clinicopathologic characteristics and outcomes of six patients with nodal metastases from HCC among a total of 504 patients who underwent hepatic resection for HCC in our department over a 16-year period. The nodal metastases were diagnosed preoperatively in two patients. The average diameter of the resected tumors was 7.8 cm and all were confirmed as poorly differentiated HCC. All of the six patients had intrahepatic metastatic nodules and five also had portal vein invasion. One patient underwent limited resection, and the other five underwent bisegmentectomy. All of the regional LNs were removed in one patient, while only enlarged LNs were removed in the other five. One patient died of postoperative liver failure and the others all died later of intrahepatic or nodal recurrence. Our findings suggest that the prognosis of patients with nodal metastasis from HCC is generally poor, even if hepatic resection with regional LN dissection is performed.

Cytokeratin-19 fragments in serum (CYFRA 21-1) as a marker in primary liver cancer

Using an electrochemiluminescence immunoassay, CYFRA 21-1 concentrations were measured in sera from 187 patients with primary liver cancer (164 with hepatocellular carcinoma (HCC) and 23 with intrahepatic cholangiocarcinoma (ICC)) and 87 patients with benign liver diseases. Concentrations of CYFRA 21-1 were significantly higher in patients with ICC (5.0; interquartile range 3.1–10.7 ng ml−1) than in those with benign liver disease (1.4; 1.0–1.9; Mann–Whitney U-test, P<0.0001) or HCC (1.7; 1.1–2.7; Mann–Whitney U-test, P<0.0001). Using cutoff values selected for 95% specificity in the benign group (3.0 ng ml−1), CYFRA 21-1 showed higher sensitivity for ICC (87.0%) than three commonly used markers including α-fetoprotein (17.4%), carcinoembryonic antigen (34.8%), and carbohydrate antigen 19-9 (60.9%). Serum CYFRA 21-1 increased in ICC from stages I/II to IV (Kruskal–Wallis test, P=0.0102). CYFRA 21-1 concentration increased with extent of local invasion, but not nodal status. Serum CYFRA 21-1 represents a useful diagnostic test for ICC that offers high sensitivity. CYFRA 21-1 reflected differences in tumour burden, suggesting applicability to staging and follow-up.

Clinicopathological Criteria for Multicentricity of Hepatocellular Carcinoma and Risk Factors for Such Carcinogenesis

Multicentric occurrence is an important characteristic of hepatocellular carcinoma. We evaluated clinicopathological criteria for multicentric hepatocellular carcinoma and identified risk factors for such carcinogenesis. Subjects were 251 consecutive patients undergoing liver resection for hepatocellular carcinoma. One kind of multicentric hepatocellular carcinoma had at least one tumor consisting of well‐differentiated hepatocellular carcinoma, together with moderately or poorly differentiated hepatocellular carcinoma located in a separate region. The other kind had an area of well‐differentiated component around hepatocellular carcinoma with less differentiation in all occurrences. The outcome of patients with tumors classified in this way was studied. Univariate and multivariate analyses were done to identify risk factors for multicentric hepatocellular carcinoma. The cumulative survival rate was significantly higher in patients with multicentric hepatocellular carcinoma than in patients with hepatocellular carcinoma associated with intrahepatic metastasis. Analysis by Coxs proportional hazard model showed that multicentricity was not a factor in the outcome. The risk of multicentric occurrence increases with progression of chronic liver disease. Univariate analysis showed hepatitis C virus marker and hepatitis B core antibody to be risk factors. By multivariate analysis, the odds ratio for multicentric occurrence in patients infected with hepatitis C virus and with serum hepatitis B virus core antibody compared with patients without either hepatitis C virus or hepatitis B virus was 10.86. This ratio in patients with hepatitis C virus alone was 4.30. These criteria for multicentric hepatocellular carcinoma seem to be clinically useful. Hepatitis C virus infection with or without former infection by hepatitis B virus is a strong risk factor for multicentric hepatocarcinogenesis.