Hepatocellular Carcinoma

HEPATOLOGY, October 1998, p. 1161-1165, Vol. 28, No.4
Meeting Report

Adrian M. Di Bisceglie, Robert L. Carithers Jr., and Gregory J. Gores

From the Department of Internal Medicine, Saint Louis University School of Medicine, St. Louis, MO

INTRODUCTION

The 3-day AASLD Clinical Research Single Topic Conference on Hepatocellular Carcinoma was aimed at exploring the pathogenesis of hepatocellular carcinoma (HCC) and developing strategies for prevention and early treatment of this form of cancer. An important goal of the meeting was to establish collaborative clinical research studies related to this important liver tumor.

EPIDEMIOLOGY, PATHOLOGY, AND PATHOGENESIS

Although the epidemiology of HCC has been studied extensively in some regions, few data are available from the United States. Dr. Craig Shapiro from the Centers for Disease Control and Prevention (Atlanta, GA) indicated that between 1979 and 1995 the reported incidence of HCC increased approximately 75%, from 1.2 to 2.1 per 100,000 population. In 1995, the cases comprised 78% whites, 14% blacks, and 8% Asians and Pacific Islanders. There was a distinct male preponderance among all ethnic groups, although this trend was most marked among Chinese Americans in whom the annualized rate of HCC among men was 26.7 per 100,000 and among women 6.2 per 100,000 population.

Although chronic viral hepatitis accounts for the large majority of HCC around the world, in the United States chronic Hepatitis B or C together account for no more than 30% to 40% of reported cases. Thus, there are many individuals in whom no obvious cause can be identified. Inherited metabolic diseases are well known to be risk factors for the development of HCC including hemochromatosis, -1-antitrypsin deficiency, glycogen storage disease, porphyria cutanea tarda, and tyrosinemia. Other conditions in which the link is weaker include Wilson’s disease and other forms of porphyria. Nonetheless, these inherited diseases are generally uncommon causative factors for HCC. Another factor that has been implicated is alcohol. Alcoholic cirrhosis may clearly result in HCC, but it is not certain whether alcohol is intrinsically carcinogenic or whether associated hepatocellular injury and regeneration, iron accumulation, or coexistent infection with the Hepatitis C virus (HCV) are required. Dr. Adrian Di Bisceglie (Saint Louis University School of Medicine, St. Louis, MO) urged further study of cases of HCC with no obvious cause to determine whether factors such as chemical carcinogenesis or underlying adenomas could be implicated.

Dr. Myron Tong (Huntington Memorial Hospital, Pasadena, CA) reviewed the progression from chronic viral hepatitis to HCC in the United States. Among those chronically infected with the Hepatitis B virus (HBV), he calculated the annual incidence of HCC to be 818 per 100,000; a very similar number to that noted in high incidence countries. Patients in the United States with HCV-related HCC are more likely to be older, white, and have a prior history of blood transfusion than individuals with HBV-related HCC.

Dr. Giovanna Fattovich (University of Verona, Verona, Italy) reviewed the association between HCV infection and HCC in Europe. She described 135 patients with posttransfusion Hepatitis C followed-up for a mean of 7.5 years at her center. Among these subjects, 104 developed chronic hepatitis with persistently increased serum aminotransferases, and 65 had a liver biopsy. Over time, 21 patients developed histological cirrhosis, but only 1 developed HCC 12 years after blood transfusion. She reviewed additional data on 384 patients in a Eurohep database with HCV-related cirrhosis. Over an average follow-up interval of 4 years (range, 1 to 11 years), 18% developed hepatic decompensation and 8% HCC. The 5-year risk of HCC was 7%, and the risk at 10 years was 14%. In a multivariate analysis, several independent factors were associated with an increased risk of HCC. These included older age, physical signs of cirrhosis, and elevated serum bilirubin more than 1.5 times the upper limit of normal. Interestingly, two factors in this group not associated with an increased risk of HCC were male gender and HCV genotype.

It is readily apparent that HCC has different manifestations in different parts of the world. Professor Michael Kew (University of the Witwatersrand Medical School, Johannesburg, South Africa) noted that patients with HCC from high incidence regions such as Southern Africa and the Far East often were younger and had larger tumors at initial presentation, which often grew very rapidly. In the United States, HCC is generally a late complication of clinically evident cirrhosis, whereas in high-incidence areas, pain and weight loss were caused directly by the tumor. HCCs occurring in Japan, China, and Europe often have a well-demarcated capsule, an important prognostic factor, whereas in the United States and Southern Africa, a capsule may be absent. Patients from high-incidence regions appear more likely to have elevated serum -fetoprotein (AFP) levels (93% vs. 70% above 20 ng/mL) and a higher median level (12,425 vs. 89 ng/mL).

The mechanism by which chronic HCV infection results in HCC is not known, although most cases occur in association with cirrhosis. Dr. Edward Tabor (Food and Drug Administration, Rockville, MD) speculated on ways by which HCV might be directly carcinogenic. Free radical-mediated injury that occurs as part of chronic liver damage might cause DNA damage and lead to HCC. Both HCV infection and HCC are associated with mutations in tumor suppressor genes such as p53 and RB. The core protein of HCV is thought to repress the promoters of p53 and RB, respectively. There is increasing evidence for a role of the X gene and its product in HBV-associated hepatocarcinogenesis. The X gene transactivates myc, fos, EGF-R, and transforming growth factor-B, whereas the X protein binds to and sequesters p53. HBV X gene transgenic mice frequently develop HCC. Sequencing of HBV DNA from HCC tissue and adjacent nontumorous liver tissue has shown a high rate of mutations in codon 130 and 131 of the X gene. This region of the genome also affects the core promoter in the overlapping genome of HBV, is essential for its transactivating function at codons 132 to 139, and is also adjacent to the region of X that binds to p53. Although the significance of this finding is not yet known, it suggests that mutations in HBV might be directly linked with the development of HCC.

Dr. Snorri Thorgeirsson (Laboratory of Experimental Carcinogenesis, National Cancer Insitute, Bethesda, MD) described the identification of a new gene related to control of hepatocyte growth, named BOG. This factor was identified by a subtractive cloning strategy using rat liver epithelial cells, one transformed and one nontransformed. The 19-kd protein product of this gene appears to suppress the growth-inhibiting effect of transforming growth factor-B1. The BOG gene is present in all mammalian species and expressed in various tissues.

SCREENING AND DIAGNOSIS

The rationale for screening for HCC is based on the concept that tumors usually begin as a single lesion, often have a long doubling time, are often encapsulated, and typically have a long asymptomatic stage and the concept that groups at high risk for developing HCC can been identified. The modalities potentially available for screening include AFP, ultrasound, serological markers, and a variety of other radiologic techniques. Serum AFP values are greater than 20 ng/mL in up to 90% of cases of HCC. Unfortunately, AFP has a low specificity, and levels may be elevated in pregnancy, germ cell tumors, and acute or chronic Hepatitis. A variety of other serological tests have been tested as potential screening tools for HCC although none is as sensitive and specific as AFP. Hepatic ultrasound has a greater sensitivity and specificity than AFP when used for HCC screening, although only one study has examined ultrasound as the sole modality of screening.

The two major strategies for HCC screening have been population based and clinic based. In population-based approaches, everyone in the population is screened, whereas in clinic-based approaches, strategies are dependent on the referral of patients. Although population-based programs are ideal, few such studies have been done. In one survey, 850 tumors were found among 3.5 million people in Shanghai. In the only randomized trial of screening, 14,794 HBV carriers were randomized to AFP determinations every 6 months versus no monitoring. Ultrasound was performed in those with AFP greater than 20 ng/mL. Although 18% of the tumors detected were less than 3 cm, no survival benefit was found. However, this study was complicated by the fact that many patients found to have small tumors did not have adequate access to definitive treatment.

Among the native Alaskan population, over 53,000 individuals (75% of the population) were screened for Hepatitis B markers and serum AFP estimated every 6 months in Hepatitis B surface antigen (HBsAg) carriers; those with levels greater that 15 ng/mL underwent ultrasound examination. Among the 1,487 HBsAg carriers, 349 (23%) had AFP elevations, 254 of whom were pregnant women. A total of 29 HCCs were diagnosed, 18 of which were resected. Among the resected patients, 7 were tumor free at 5 years and 2 additional patients were tumor free less than 5 years after surgery. In this study, AFP had a sensitivity of almost 97% with a specificity (excluding pregnancy) of 28%.

Dr. Morris Sherman (The Toronto Hospital, Toronto, Ontario, Canada), the principal investigator in a 9-year screening program in Canada, discussed the practical aspects of establishing a screening program for HCC. He noted that, to be truly effective, screening should result in a decrease in disease mortality. This has not yet been shown in HCC. Unfortunately, the opportunity to scientifically confirm this important aspect of HCC screening with randomized controlled trials may have been lost because of the reluctance of both patients and physicians to include an unscreened arm for comparison. If a program is to be effective, patients must accept screening and be willing to return for follow-up visits. This is very dependent on convenience, the level of anxiety induced, and the organization of the program. In the Canadian study, the dropout rate was 17% in its early phases; however, after a number of staff changes occurred, this rose to 50%.1 Finally, there must be an effective recall program and good confirmatory tests, and effective treatment must be available for those discovered to have HCC. The efficacy of hepatic resection, ethanol injection, and other modalities of therapy for HCC have distinct limitations. Furthermore, in patients found to have tumors during screening, only a small proportion actually undergo definitive therapy of the tumor because of age, liver function, general medical condition, and patient refusal.

These potential limitations of screening for HCC raise legitimate questions about whether it is ethical to offer screening when the benefits are unknown. This must be balanced against the general belief that progress in treatment of patients with HCC will only come from studies of small tumors. The costs of a screening program must also be considered. In most studies of HCC screening, the cost per life year gained ranges from $16,000 to $55,000. The cost of finding each tumor is $11,000 to $25,000, the cost for each year of life saved is $17,000, and the cost per life saved is greater than $350,000.

Three imaging procedures are in common use for diagnosing HCC: ultrasonography, helical computed tomography (CT) scans, and magnetic resonance imaging. Unfortunately the literature on most radiologic means of diagnosing HCC is extremely confusing. For example, in many studies it is assumed that all lesions within the liver are malignant, although typically only one is biopsied. Furthermore, there are few studies in which autopsy or explant correlation has been performed. As a consequence, the sensitivity and specificity of commonly used radiologic studies for detection of HCC is unknown and has probably been overestimated.

Ultrasound is inexpensive and widely available. However, results may be affected by the quality of the machine used, the diligence of the operator, and technical factors such as obesity and overlying bowel gas. In the limited studies available, ultrasound detects fewer than 50% of the lesions present on whole organ examination.

The helical CT is a rapid technique that allows scanning the liver twice, including both the hepatic arterial and portal venous phases. The addition of the hepatic arterial phase increases the number of lesions detected by 20% to 30% or more. The largest series of patients studied with histopathologic correlation of radiographic findings consisted of 424 patients undergoing evaluation for liver transplantation at the University of Pittsburgh, Pittsburgh, PA. Of these, HCC was found in 56 of the explanted livers removed at transplantation.2 Double helical CT scanning detected only 17% of tumors less than 1 cm, 29% of lesions 1 to 2 cm, and 63% of lesions greater than 2 cm. Dr. Jordi Bruix (Hospital Clinic, Barcelona, Spain) reported that magnetic resonance imaging and CT appear to have comparable sensitivities. Thus, magnetic resonance imaging appears to be gaining favor among many radiologists for evaluation of patients with possible HCC.

PREVENTION AND TREATMENT

A variety of approaches have been used for the prevention and treatment of HCC. Worldwide, HBV remains the most common cause of HCC because of widespread vertical transmission (maternal-infant spread at the time of birth) in Asia and Africa. With the advent of an effective Hepatitis B vaccine, it is now theoretically possible to prevent perinatal transmission of HBV and eliminate this cause of HCC with population-based vaccination programs. Dr. Anna Lok (University of Michigan Medical Center, Ann Arbor, MI) summarized current knowledge regarding the efficacy of this approach. Beginning in 1984, all newborns from mothers seropositive for HBsAg in Taiwan were vaccinated for HBV. Ten years after the initiation of this program, the number of children positive for HBsAg has fallen substantially, from 9.8% to 1.3%.3 Along with this decrease in the HBsAg carrier state, the incidence of HCC in children ages 6 to 14 has decreased by greater than 50%.4 Thus, HBV vaccination is the first example of a vaccine preventing a human cancer. Unfortunately, large-scale infant vaccination programs have not been widely implemented in other endemic areas for HBV, such as in large parts of Africa.

Treatment eliminating the virus from the liver could also potentially reduce viral-related HCC. Dr. Jay Hoofnagle (National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD) critically analyzed the current information regarding the ability of alfa-interferon treatment to reduce the development of HCC in cirrhotic patients. Nishiguchi et al.5 first reported in 1995 the results of a randomized controlled trial suggesting that alfa-interferon decreases the incidence of HCC in HCV-related cirrhosis. In a retrospective analysis, Mazella et al.6 reported that only 2.5% of 285 treated cirrhotics developed HCC over a follow-up period of 2.7 years compared with 9.8% of 92 untreated patients. Again, this effect was observed despite a sustained response rate of only 6% in the treated individuals. However, a large Eurohep study did not find a statistically significant difference in the diagnosis of HCC observed over 5 years in alfa-interferon treated versus nontreated cirrhotics. Dr. Hoofnagle concluded that presently existing data do not support the concept that alfa-interferon treatment of patients with cirrhosis eliminates the risk of HCC, but might decrease it somewhat.

Chemoprevention is defined as the use of specific natural or synthetic chemicals to reverse or suppress the complex processes involved in the initiation, promotion, or proliferation phases of oncogenesis. The goal of chemoprevention is, therefore, to prevent initiation of cellular transformation, eliminate premalignant lesions, prevent tumor progression, or block metastases. Chemoprevention can be either primary or secondary. Dr. Reuben Lotan (M.D. Anderson Cancer Center, Anderson, TX) reviewed the primary and secondary chemoprevention trials in progress for HCC in humans. The objective of primary chemoprevention is to prevent the formation of cancer in a patient at risk, whereas the aim of secondary chemoprevention is to block the development of metachronous lesions in a patient who has already had one lesion removed surgically.

Two agents, oltipraz and polyprenoic acid, are being used in human trials of chemoprevention for HCC. Oltipraz induces expression of glutathione-S-transferase and is being used in an endemic region of China in a primary chemopreventive trial.7 Increased intracellular glutathione-S-transferase helps detoxify aflatoxin B1, a carcinogen implicated in the genesis of p53 mutations in patients with HBV. Polyprenoic acid given for a defined time postoperatively, reduced the formation of metachronous lesions after surgery for resectable HCC.8 Polyprenoic acid is a modified retinoid; retinoids may induce apoptosis in initiated cells, promote differentiation of dedifferentiated cells, or block progression of cancers by functioning as anti-angiogenesis agents.

Dr. Brian Carr (University of Pittsburgh, Pittsburgh, PA) reviewed the biological characteristics of this tumor from the perspective of an oncologist. The multicentricity and propensity to vascular invasion of this malignancy were emphasized. In addition, HCC expresses many drug-resistance genes including p-glycoprotein, glutathione-S-transferase, and heat shock proteins, which account, in part, for the relative drug resistance of these tumors. The response to intravenous chemotherapy has been consistently discouraging and usually less than 20%. Although intrahepatic arterial chemotherapy consistently achieves response rates of 40% to 60%, complete responses are unusual, and survival is not enhanced. The toxicity of this approach combined with that of acceleration of the underlying liver disease appears to be a particular problem. Because neovascularity is a key feature of HCC, anti-angiogenesis agents are promising therapeutic agents.

Dr. Bruix reviewed the results of a study of percutaneous ethanol injection therapy conducted in Spain. Among 100 patients who were enrolled, 71 initially had successful therapy, which was maintained in 46 patients. The 5-year survival of all patients was approximately 20%; however, in Child’s class A patients who responded, the 5-year survival was 50%. Although the procedure was reasonably well tolerated, 4 patients developed portal vein thrombosis and there were two procedure-related deaths. Thus, this therapy may not be as innocuous as previously reported. Dr. Bruix also highlighted future alternative local ablative therapies that may be more effective than percutaneous ethanol injection therapy, including acetic acid injection, microwave therapy, and radiofrequency interstitial tumor ablation therapy.

Chemoembolization is widely used throughout the world for HCC. Dr. Alan Venook (University of California San Francisco, San Francisco, CA) described the techniques and outcomes of this approach. In his experience, 80% of patients experience a transient, subjective improvement in symptoms. Several participants commented on the dearth of randomized clinical trials and the varied techniques used. No studies have shown a long-term survival advantage in treated versus untreated patients. Objective treatment responses are difficult to assess because liquefaction necrosis precludes an accurate measurement of tumor volume. Furthermore, Dr. Bruix commented on a randomized clinical trial, which showed no survival benefit despite a partial tumor response to embolization in the treated patients. Better therapeutic options are needed for patients with advanced disease.

Dr. Bernard Langer (Toronto General Hospital, Toronto, Ontario, Canada) focused on resection of HCC and noted that improvements in patient selection, tumor selection, and operative technique have led to the current low surgical mortality rates and 5-year survival rates of 30% to 50% occurring after resection of HCC. Patients with two or fewer lesions, each less than 5 cm, which are encapsulated and have no evidence of macroscopic vascular invasion, are the best candidates for hepatic resection. Operative mortality in cirrhotics is approximately 10%, and there is a growing appreciation that patients with portal hypertension or even mild elevations of the bilirubin should not be subjected to hepatic resection.

The theoretical goals of liver transplantation for HCC include expanding the therapeutic options for patients with advanced liver disease and otherwise resectable HCC and increasing options for patients who are not currently resectable because the extent of resection would leave insufficient liver volume. Liver transplantation for HCC has fulfilled the first goal but not the second. Dr. Gregory Gores (Mayo Clinic, Rochester, MN) reviewed the evaluation of patients for liver transplantation with small HCC. Results after liver resection for small tumors (three lesions, each <5 cm) are excellent and equal to those obtained for cirrhosis alone. However, problems remain regarding preoperative staging. Current radiographic modalities (ultrasound, helical CT, and magnetic resonance imaging) frequently miss small and sometimes extensive infiltrating tumors. Limited experience with positron emission tomography scanning suggests that this technique could prove useful in the preoperative evaluation of patients with HCC. The question of whether patients with Child’s class A cirrhosis and a small HCC should undergo resection or liver transplantation remains controversial.

Although many centers use preoperative chemoembolization to prevent tumor growth while waiting for liver transplantation, neither the need for nor the efficacy of this approach has been rigorously evaluated. Dr. Myron Schwarz (Mt. Sinai Medical Center, New York, NY) described a protocol for patients with unresectable HCC because the extent of resection would leave insufficient liver volume. The protocol included patients with HCC of 5 cm or greater, comprising up to 75% of the liver volume, but without evidence for macroscopic vascular invasion. Patients received preoperative chemoembolization followed by intravenous adriamycin. Of the 78 patients entered into the study, 31 were never transplanted because of tumor progression or liver-related deaths. The 5-year disease-free survival was a discouraging 22%. The principal factor predictive of tumor recurrence was a tumor size greater than 8 cm. Thus, better adjuvant regimens are needed to expand the patient population with HCC that will benefit from transplantation. This experience also highlights the notion that results of liver transplantation for HCC should be assessed on an intent-to-treat basis to take into account those patients who died or experienced tumor progression while on the waiting list for transplantation.

RESEARCH PANELS

As a part of this meeting, four research panels were commissioned to develop clinical research plans in HCC with input from attendees, as summarized below:

1. A panel led by Dr. Robert Carithers (University of Washington Medical Center, Seattle, WA) developed plans for establishing a system for centralized collection of data on HCC similar to a registry. The focus of this panel was to discuss the essential ingredients of a data collection system for patients with HCC. It was agreed that for any registry to be effective, it had to be simple and easy to use but at the same time must incorporate the essential data needed to make the effort meaningful. Furthermore, the specific objectives of any registry should be well defined (for example, epidemiology, natural history, and response to therapy).

The data elements recommended for inclusion were fairly easy to define and included basic elements of history, physical examination, laboratory studies, the radiologic and pathological features of each lesion, classification systems for the extent of tumor spread and the severity of the underlying liver disease, the types of treatment used, and the outcome of each patient.

Much more difficult issues were how to implement such a database, who would enter the data, and of even greater importance, who would fund such an undertaking. Despite these obstacles, it was generally agreed that such a database would be a great stimulus to collaborative research on HCC, not only in the United States but throughout the world.

It was visualized that the ideal HCC registry would be based on the World Wide Web using an easy-to-use data entry format so that investigators throughout the world could collaborate to further our understanding of HCC.

2. The focus of the pathology research panel under the leadership of Dr. Swan Thung (Mt. Sinai Medical Center) was to develop means of assisting clinicians in making earlier diagnoses of HCC, when treatment might be more effective. Particular emphasis was placed on dysplastic nodules because they appear to represent the earliest stage of malignant transformation in the progression to well developed HCC.

The specific objective discussed by this panel was the development of collaborative research programs to establish well accepted pathological criteria for each of the different types of hepatic nodules, particularly premalignant lesions; to compare changes in preneoplastic lesions, HCC, and regenerative nodules; to characterize the cells of origin and evaluate aberrations in expression of proto-oncogenes, tumor suppressor genes, and growth factors in each lesion; and to perform molecular studies to determine the clonality of the various lesions.

In addition to the basic studies, the need for careful correlation between pathological findings, radiologic features, and clinical findings were stressed. These types of careful clinical studies are necessary to confirm the preneoplastic nature of the various lesions found within hepatic nodules.

3. The National Institutes of Health is preparing to conduct a study evaluating the effect of interferon therapy to prevent HCC in chronic Hepatitis C. This study is expected to yield much data on long-term follow-up. This panel, therefore, focused on two related areas, including screening and chemoprevention. It was recognized that a randomized, controlled trial of screening was not practical in the United States. Instead, the panel chose to focus on issues such as determining the best method of screening, developing new diagnostic and screening tools for HCC, and assessing which patients are likely to benefit most from treatment. A very exciting prospect is the possibility of chemoprevention of HCC in patients with underlying liver disease, encouraged by recent intriguing results using polyprenoic acid. Suitable compounds for chemoprevention should be sought and tested.

4. There was a great deal of interest in developing common trials of therapy for HCC and several such trials were proposed by a group led by Dr. Robert Gish (Pacific Medical Center, San Francisco, CA). A potential trial to treat HCC combined with (or before) liver transplantation was discussed by the panel. A major issue that surfaced was how to maintain patients who have HCC diagnosed before transplantation on the waiting list long enough to receive a liver transplant. The basic format of such a study would be to randomize patients who are discovered to have a tumor while on the waiting list to two alternate therapies. Possible therapies included chemoembolization or radiofrequency ablation versus no therapy. Most discussants were not comfortable with not offering any specific treatment, even recognizing that no therapeutic choice had a defined benefit. The end point for the study would be whether or not the patient remained suitable for transplantation based on the usual clinical criteria (single tumor <5 cm or <3 cm if up to 2 or 3 lesions, with no evidence of vascular invasion or extrahepatic disease). A crossover arm could be built into the trial if it appeared that patients had rapidly growing tumors not responding to therapy. The panel recognized that pretransplant therapies for HCC remain unproven but are moving toward broader use and concluded that it would be timely for a multicenter trial to be performed.

In general, attendees felt that HCC was an ideal disease around which to organize multi-institutional collaborative studies and that hopefully such studies would arise from this meeting.

Footnotes

Abbreviations: HCC, hepatocellular carcinoma; HCV, Hepatitis C virus; HBV, Hepatitis B virus; AFP, -fetoprotein; HBsAg, Hepatitis B surface antigen; CT, computed tomography.

The AASLD Clinical Research Single Topic Conference on Hepatocellular Carcinoma was held in St. Louis, MO, March 27 to 29, 1998

Received June 2, 1998; accepted June 5, 1998

Address reprint requests to: Adrian M. Di Bisceglie, Department of Internal Medicine, St. Louis University School of Medicine, 1402 S. Grand Blvd., St. Louis, MO 63124. Fax: (314) 268-5108

REFERENCES

1. Sherman M, Peltekian KM, Lee C. Screening for hepatocellular carcinoma in chronic carriers of Hepatitis B virus: incidence and prevalence of hepatocellular carcinoma in a North American urban population. HEPATOLOGY 1995; 22: 432-438[Medline].

2. Miller WJ, Baron RL, Dodd GD, Federle MP. Malignancies in patients with cirrhosis: CT sensitivity and specificity in 200 consecutive transplant patients. Radiology 1994; 193: 645-650[Medline].

3. Chen H-L, Chang M-H, Ni Y-H, Hsu H-Y, Lee P-I, Lee C-Y, Chen D-S. Seroepidemiology of Hepatitis B virus infection in children: ten years of mass vaccination in Taiwan. JAMA 1996; 276: 906-908[Medline].

4. Chang M-H, Chen C-J, Lai M-S, Hsu H-M, Wu T-C, Kong M-S, Liang D-C, et al. , and the childhood hepatoma study group. Universal Hepatitis B vaccination in Taiwan and incidence of hepatocellular carcinoma in children. N Engl J Med 1997; 336: 1855-1859[Medline].

5. Nishiguchi S, Kuroki T, Nakatani S, Morimoto H, Takeda T, Nakajima S, Shiomi S, et al.

Randomised trial of effects of interferon- on incidence of hepatocellular carcinoma in chronic active Hepatitis C with cirrhosis. Lancet 1995; 346: 1051-1055[Medline].

6. Mazzella G, Accogli E, Sottili S, Kesti D, Orsini M, Salzetta A, Novelli V, et al. Alpha

Interferon treatment may prevent hepatocellular carcinoma in HCV-related liver cirrhosis. J Hepatol 1996; 24: 141-147[Medline].

7. Jacobson LP, Zhang B-C, Zhu Y-R, Wang J-B, Wu Y, Zhang Q-N, Yu L-Y, et al. Oltipraz chemoprevention trial in Qidong, People’s Republic of China: study design and clinical outcomes. Cancer Epidemiol Biomarkers Prev 1997; 6: 257-265[Medline].

8. Muto Y, Moriwaki H, Ninomiya M, Adachi S, Saito A, Takasaki KT, Tanaka T, et al. Prevention of second primary tumors by an acyclic retinoid, polyprenoic acid, in patients with hepatocellular carcinoma. N Engl J Med 1996; 334: 1561-1567.

Copyright © 1998 by the American Association for the Study of Liver Diseases.