Primary liver cancer: Worldwide incidence and trends

(Continuation)

In the United States population, the cumulative life-time risk (CR) of liver cancer (hepatocellular carcinoma and intrahepatic bile duct carcinomas) is 0.88 percent in men and 0.42 percent in women, which is similar to that of esophageal cancer (CR=0.77 percent in men and 0.26 percent in women) or cancers of the larynx (CR=0.69 percent in men and 0.l7 percent in women). Despite the limitations of mortality data, liver cancer age-adjusted mortality rates follow a similar pattern and are consistent with the incidence data. Liver cancer is a tumor with high lethality, and a yearly fatality ratio approximately 1, indicating that most cases do not survive a year. In Europe, for the period 1990 to 1999 and, in the United States, for the period 1995 to 2000, liver cancer survival rates at the population level were analyzed from the unselected data of population-based cancer registries; the five-year relative survival rates (mortality from liver cancer adjusted for mortality from competing causes) were 6.5 percent and 8.3 percent, respectively. There are few differences in survival rates according to gender, suggesting a similar distribution of stage at diagnosis. In developing countries, liver cancer is inevitably fatal.

Liver Cancer Incidence In Different Ethnic Groups And Immigrant Populations

The lowest incidence rates are consistently found among whites (3.8 in men and 1.4 in women). Gradually increasing rates are found in the Japanese (5.5 in men and 4.3 in women), African American (7.1 in men and 2.1 in women), Hispanic white 9.8 in men and women), Filipino (10.9 in men and 2.4 in women), Chinese (16.2 in men and 5.0 in women), and Korean American (20.7 in men and 10.4 in women) populations. Among women, the high-risk ethnic groups have a 2- to 5-fold higher age-adjusted incidence rate of liver cancer than that in non-Hispanic white women. The cross-sectional rates are consistent with difference in risk factors and genetic susceptibility to liver cancer across ethnic groups.

Age And Gender Distribution – Age Specific Incidence Rates

In most high-risk areas, such as Southeast Asia (Qidong in China) or the West Coast of Africa (Bamako) in Mali, rates of liver cancer increase after 20 years of age and peak or stabilize at the age of 50 years and above.

In these countries, liver cancer is not a rare event at ages 20 to 35 years. Still, the incidence in Qidong is substantially higher at each age group than the corresponding incidence in Mali, a high-risk country in Africa, and the age of occurrence is significantly shifted toward the young age groups.

A second pattern is observed in high-risk countries with recent substantial degrees of development. For example, in the cancer registry in Shanghai, rates are lower than the corresponding rates in Qidong, and the age-specific incidence increases steadily with age, a pattern similar to that observed in low-risk areas. In the Osaka Cancer Registry in Japan, the incidence of liver cancer among men increases after the age range 45 to 50 years and reaches a plateau around 65 years of age.

During the last 2 decades, increases of primary liver cancer incidence rates have been reported from Australia, Central Europe, the United Kingdom, Japan, and North America. Decreasing trends were reported from Chinese population in Singapore and Shanghai, India, Sweden, and Spain.

Several reports from the United States population have investigated the consistency in trends using 3 different sources of information, namely liver cancer hospitalization rates, incidence rates from Cancer Registries, and mortality rates from the US population. The studies covered a time period from 1976 to 2003, and, for the analysis of cancer incidence, it was possible to restrict the comparison with the cases that had histopathogic confirmation of HCC (about 75 percent of reported cases). The study reported significant increasing trends for the Africa, white, and Hispanic populations. The analyses of the components of the trends suggested that, in the United States, there is a predominant cohort effect.

International trends in mortality have also been evaluated in 22 populations for the period 1979-1998. Among men, increases in mortality from liver cancer have been reported in the United States, Japan, Australia, Scotland, France and Italy, whereas decreasing trends have been reported in the United Kingdom. Trends among women were similar. Increases in the rates of cholangiocarcinoma have also been reported in the United States, Japan, Australia, Scotland, England, and Wales and among women in the United States, Australia, England and Wales.

Risk Factors For Primary Liver Cancer Hepatitis Viruses

Overall, 75 percent to 80 percent of cases of primary liver cancer are attributable to persistent viral infection with either HBV (50-55 percent) or HCV (25-30 percent). Worldwide, strong geographic correlations have been found between the incidence of liver cancer and the prevalence of hepatitis B surface antigen (HBsAg) or antibody to hepatitis C virus (anti-HCV).

Country-specific studies show consistent correlations between the prevalence of viral markers and the observed risk of liver cancer. In New Zealand, the Maori population has a higher incidence of primary liver cancer (12.8 in men and 3.6 in women) as compared with non-Maori populations (2.9 men and 1.3 in women). A study in the New Zealand population showed that the prevalence of HBsAg in Maori persons with primary liver cancer was 77 percent, in contrast to 6 percent among white persons with primary liver cancer. Similar results have been reported from the United States, where the prevalence of HBsAg among cases of primary liver cancer is higher among Asian patients born in the United States (80 percent) than among white patients (9 percent).

More specifically, one possible explanation for the differences in incidence rates in China and West Africa can be related to age of onset of infection. In many Asian countries, there is a high prevalence of HBsAg carrier mothers who also express hepatitis B "e" antigen (HBeAg) and remain HBV-DNA positive throughout the reproducing age intervals. In high-risk areas in Africa, the mother-to-child transmission rate is somewhat lower, and the child-to-child mode of transmission predominates. The impact of cofactors, perhaps with the exception of aflatoxin, in the high countries is marginal.

The reduction of risk observed in immigrants to lower risk countries probably reflects a combination of the reduction in the mother-to-child transmission of HBV linked to mixing populations, the reduction of exposure to contaminated blood products during medical interventions or other parenteral exposures, the lower prevalence of HBV (and possibly HCV) among siblings, and, more recently, the impact of the introduction of HBV vaccines.

In Egypt, the reported prevalence of HCV infection in the population has been estimated to be of the order of 5 percent to 9 percent in urban areas and 27 percent to 30 percent in the rural parts of the Nile delta. Epidemiologic studies have traced the source of HCV infection to the large scale treatment campaigns against schistosomal infestation conducted between 1920 and 1970. The treatment typically included several courses of intramascular or intravenous drugs against schistosoma that were delivered under nonsterile conditions. From current information, it is predictable that a cohort-related outbreak of HCV-related cirrhosis and HCC will occur in this population in the decades to come.

In the United States, the predominant cohort component of the increases in primary liver cancer incidence and mortality have been interpreted as a long-term consequence of increased HCV exposures in the period 1960-1970 through contact with contaminant blood and injection drug use in the relevant generations.

Alcohol

Chronic alcohol abuse and alcoholic cirrhosis have been recognized as a cause of primary liver cancer. However, it is not certain whether alcohol is a true carcinogen or if it acts as a cofactor in the presence of coexistent infection with HBV and/or HCV. Several epidemiologic studies among alcoholics have described a high prevalence of BV markers (16 percent-70 percent) and HCV markers (10 percent-20 percent) as compared with a background prevalence of close to 5 percent and less than 1 percent, respectively. These prevalence are even higher in primary liver cancer patients who are also alcoholics (27 percent to 81 percent of HBV markers and 50 percent to 77 percent of HCV markers), suggesting a complex interaction between alcohol and viral infections in the etiology of primary liver cancer. Studies conducted in northern Italy and Greece estimated that the attributable fraction of high levels of alcohol consumption, once adjusted for HBV and HCV status, were 45 percent in Italy and 15 percent in Greece

Aflatoxin

Primary liver cancer has been related to dietary aflatoxin exposure in countries where fungal infestation of crops and animal feed are common. Studies that have used detection of aflatoxin-albumin adducts, aflatoxin M1 in urine, aflatoxin-N7-guanine adducts in urine, or p53 specific mutations (G to T transversions at codon 249) have suggested that individuals who are HBsAg carriers and who are exposed to aflatoxin in their diets have an increased risk of progression to HC as compared with nonaflatoxin-exposed HBV carriers.

Oral Contraceptives

The occurrence of benign liver adenomas and occasional primary liver cancer among women who were long-term users of oral contraceptives has been documented in several studies. Case-control studies conducted in developed countries where a substantial number of women have used contraceptives for extended period of times have found relative risks between 1.6 and 5.5 among those who have ever used oral contraceptives, and some studies have shown a relationship with duration of use and risk of primary liver cancer. In one of the few studies conducted in the black population in South Africa, a country with high prevalence of HBV infection, no association between oral contraceptives and primary liver cancer was found. Analyses of the mortality trends of liver cancer in young women in the United Kingdom, United States, Japan, and Sweden have provided no support for a measurable effect of oral contraceptives on mortality because of liver cancer. A multicenter study conducted with 317 cases of liver cancer among women under 65 years and 1060 controls found an association between this tumor and duration of oral contraceptive use n the small subgroup of liver cancer cases without cirrhosis and with negative markers for HBV and HCV infection. Further studies are needed to clarify and quantify the role of oral contraceptives in liver cancer, and issue of considerable public health importance.

In addition to the use of oral contraceptive, other hormonal factors have been explored to explain the increased male susceptibility to primary liver cancer worldwide. A recent study in the United States found evidence for protective effects of parity, late menopause, use of hormonal replacement therapy, and early age at menarche. Except for the latter, all factors were independent of HBV status. Likewise, androgens have been postulated as a risk factor for primary liver cancer, but data to support this effect are lacking.

Tobacco

The association of cigarette smoking and primary liver cancer has been claimed from the results of some epidemiologic studies, notably in Japan. The evidence is, however, not consistent, and residual effects of HBV or HCV cannot be ruled out from the majority of the epidemiologic studies. Nevertheless, recent data from China and Taiwan have also reported an association of cigarette smoking with liver cancer independent of HBV status.

The gender-specific impact of alcohol, tobacco, and oral contraceptive use might explain the gender ration pattern as well as the steady increase in incidence by age observed in the population in which these exposures are common.

Other Hypotheses And Emerging Factors

Other factors that may modulate the long-term impact of BV or HCV infections on development of HCC include dietary factors, some chemicals (such as arsenic), some hereditary conditions (Haemochromatosis, Wilson’s disease, and others), and nutritional factors. Growing interest is currently focusing on the association of primary liver cancer with diabetes mellitus, obesity, and syndromes related to insulin resistance. Nonalcoholic fatty liver disease is being proposed as a risk factor for primary liver cancer. Because of the considerable prevalence of some of these conditions in Western populations, it is of importance to conduct research to describe properly the nature of the associations observed.

Opportunities For Prevention

By the year 2000, HBV vaccination program had been implemented in 135 countries, including the lesser developed areas in Africa and Asia. A major input from donor agencies has made it possible to provide HBV vaccine to all newborn in the last decade, and such efforts should be encouraged and supported. HBV vaccination trials initiated in the 1980s have already shown the ability of HBV vaccines to prevent the chronic carriage of HBsAg and the development of primary liver cancer when vaccination takes place at birth in Taiwan or as adults among HBsAg-negative men in Korea.

Screening of blood products for HCV markers in the countries that have introduced such programs has substantially reduced the rate of post transfusion hepatitis C. Awareness of the negative impact of unprotected sex, significant alcohol consumption, and smoking has probably had some impact in risk behavior modification. However, its potential impact on the incidence of primary liver cancer remains to be documented. Finally, the claim has been made that aflatoxin reduction has reduced the incidence of primary liver cancer in Singapore and Shanghai, although final proof has not been provided.

Conclusions

Primary liver cancer remains a major health problem with great geographic variability. Men are consistently more affected than women, and survival is poor worldwide. Increasing trends in incidence in some developed countries, including the United States, suggest an underlying cohort effect linked to HCV and HBV exposures. Reduction of primary liver cancer burden in most developing countries should give priority to HBV vaccination campaigns and to prevention of nosocomial spread of HBV and HCV. This implies reinforcing control of blood and blood products, as well as the use of sterile medical equipment. Patients with chronic hepatitis B infection may further benefit from reduction in the aflatoxin exposure in their diets. If achieved, aflatoxin reduction may also offer some protection to patients with hepatitis C. In low-risk populations, alcohol consumption may account for the majority of the primary liver cancer cases that are non reactive for viral markers.