| |
Liver Cancer - Liver cancer mortality rates Double Since 1990
-
Disparities in liver cancer occurrence in the United States by race/ethnicity and state
|
| |
| |
Download the PDF here
The burden of liver cancer in the United States is significant and is expected to increase in the decades to come. Wide disparities in liver cancer death rates by sex, race/ethnicity, and state persist, reflecting differences in the prevalence of major risk factors and, to some extent, inequalities in access to high-quality care.. From 1990 to 2014, death rates rose by 57% in blacks, 69% in Hispanics, and 82% in NH whites; in AIs/ANs, the death rate more than doubled, from 5.4 (per 100,000) to 11.9.....Compared with NH whites (20%), survival was statistically significantly lower in blacks (16%) and higher in APIs (27%) (Table 1).....According to NHANES, blacks in the United States are more likely than other races/ethnicities combined to have chronic HCV infection.....The black-white racial disparity was the highest in the District of Columbia, in which the black rate for both sexes combined (13.3 per 100,000) was more than 4 times higher than that for NH whites (3.2 per 100,000)....the proportion of all HCCs during 2000 through to 2011 attributable to HCV was 17% in whites, 21% in Hispanics, 30% in Asians, and 36% in blacks," the investigators state
A major factor contributing to this increase is the comparatively high prevalence of hepatitis C virus (HCV) infection among those born during 1945 through 1965, also called "baby boomers."[2, 6] The sustained rise in obesity and type II diabetes over the past several decades has also likely contributed to the increasing liver cancer trend.[7]
"among US Medicare recipients during 2000 through 2011, metabolic disorders (including excess body weight, diabetes, impaired glucose tolerance, nonalcoholic fatty liver disease, and metabolic syndrome) accounted for the largest proportion of hepatocellular carcinomas (HCCs) (32%), followed by chronic HCV infection (21%), excessive alcohol drinking (13%), smoking (9%), and chronic HBV infection (4%).[28] HCC is the major histologic type of primary liver cancer, accounting for greater than 70% of all liver cancers in the United States......Currently, the CDC recommends one-time HCV testing for people born from 1945 to 1965.[39] However, only 14% of this cohort has reported HCV testing, with little variation between Asians, NH whites, blacks, and Hispanics......underdiagnoses of HCV cases, in addition to obstacles to care.......According to NHANES, blacks in the United States are more likely than other races/ethnicities combined to have chronic HCV infection.......Overall, the proportion of all HCCs during 2000 through 2011 attributable to HCV was 17% in whites, 21% in Hispanics, 30% in Asians, and 36% in blacks.......Smoking increases liver cancer risk by approximately 50%......A large-scale, prospective study in the United States has demonstrated that diabetes increases the risk of liver cancer by 2.3-fold in men and 1.4-fold in women......Among states with available data, liver cancer death rates were statistically significantly higher in black men than in NH white men in all states except Mississippi, New Mexico, Arkansas, and South Carolina. The black-white racial disparity was the highest in the District of Columbia, in which the black rate for both sexes combined (13.3 per 100,000) was more than 4 times higher than that for NH whites (3.2 per 100,000).....The liver cancer death rate in both sexes combined during 2010 through 2014 was highest (≥8 per 100,000) in the District of Columbia, Hawaii, Louisiana, and Texas and lowest (<5 per 100,000) in Iowa, Montana, South Dakota, Nebraska, Utah, Vermont, and North Dakota (Fig. 7). Consistent with a lower liver cancer death rate in NH whites, the death rate was generally higher in states with a lower proportion of NH whites. Among states with comparable proportions of NH whites, those with higher poverty levels tended to have higher liver cancer death rates. For example, the rate was higher in Kentucky (6.1 per 100,000; population, 87% NH whites) than in North Dakota (3.8 per 100,000; 89% NH whites)......HCV infection prevalence in the United States is higher among those with lower education or income and among ages 20-59 years is nearly twice as high among men as among women....HCV infection accounts for approximately 20% of liver cancers worldwide....According to NHANES, the prevalence of chronic HCV infection (defined as positive HCV RNA testing) in the United States was 1% during 2003 through 2010, representing approximately 2.7 million people.[6, 38] However, this number is likely an underestimate, because populations with very high infection prevalence (eg, heavy drug users and incarcerated or homeless people) are not represented in NHANES.[37, 38] A study incorporating these populations estimated that at least 3.5 million adults were infected with HCV as of 2010.[38] People born between 1945 and 1965 ("baby boomers") have an HCV prevalence of approximately 2.6% and account for about 80% of HCV-infected cases,[6] which explains the increased age-specific death rates for individuals ages 50 to 64 years in the current study. Major risk factors for chronic HCV infection in the United States are injection drug use and receipt of a transfusion before 1992.[6] Other factors that increase risk include being born to an HCV-infected mother, receiving high-risk tattoos or body piercing, needle-stick injuries, having sexual contact with infected people, and some medical conditions (eg, transplantation, hemodialysis, and human immunodeficiency virus infection).[39] HCV infection prevalence in the United States is higher among those with lower education or income and among ages 20-59 years is nearly twice as high among men as among women.[2, 6].......More than 20% of patients with nonalcoholic fatty liver disease, a condition highly associated with excess body weight and diabetes, will develop nonalcoholic steatohepatitis, a more serious type of nonalcoholic fatty liver disease in which fat accumulation in the liver causes inflammation and cell damage, thereby increasing the risk of liver cancer....Consistent with a substantial increase in the number of people with excess body weight and associated metabolic disorders in the United States, the estimated proportion of HCCs attributable to these factors increased from 26% during 2000 through 2003 to 36% during 2008 through 2011.
Liver cancer incidence and death rates were highest for AIs/ANs (15.2 and 11.9 per 100,000, respectively) and were more than double those in NH whites (6.3 and 5.5 per 100,000, respectively), among whom the rates were lowest. Death rates increased with age up to ages 80 to 84 years in all racial/ethnic groups except blacks, among whom mortality peaked at ages 60 to 64 years (Fig. 1)."
---------------------
Disparities in liver cancer occurrence in the United States by race/ethnicity and state
A Cancer Journal for Clinicians June 9 2017
Abstract
Liver cancer is highly fatal, and death rates in the United States are increasing faster than for any other cancer, having doubled since the mid-1980s. In 2017, it is estimated that the disease will account for about 41,000 new cancer cases and 29,000 cancer deaths in the United States. In this article, data from the Surveillance, Epidemiology, and End Results (SEER) Program and the National Center for Health Statistics are used to provide an overview of liver cancer incidence, mortality, and survival rates and trends, including data by race/ethnicity and state. The prevalence of major risk factors for liver cancer is also reported based on national survey data from the Centers for Disease Control and Prevention. Despite the improvement in liver cancer survival in recent decades, only 1 in 5 patients survives 5 years after diagnosis. There is substantial disparity in liver cancer death rates by race/ethnicity (from 5.5 per 100,000 in non-Hispanic whites to 11.9 per 100,000 in American Indians/Alaska Natives) and state (from 3.8 per 100,000 in North Dakota to 9.6 per 100,000 in the District of Columbia) and by race/ethnicity within states. Differences in risk factor prevalence account for much of the observed variation in liver cancer rates. Thus, in contrast to the growing burden, a substantial proportion of liver cancer deaths could be averted, and existing disparities could be dramatically reduced, through the targeted application of existing knowledge in prevention, early detection, and treatment, including improvements in vaccination against hepatitis B virus, screening and treatment for chronic hepatitis C virus infections, maintaining a healthy body weight, access to high-quality diabetes care, preventing excessive alcohol drinking, and tobacco control, at both the state and national levels.
Introduction
Liver and intrahepatic bile duct cancer (hereafter liver cancer) is the fifth and eighth leading cause of cancer death in men and women, respectively, accounting for approximately 41,000 cancer cases and 29,000 deaths in the United States in 2017.[1] Liver cancer death rates are increasing at a faster pace than any other cancer.[2, 3] An average 19 years of life are lost per death because of a relatively young median age at diagnosis (63 years) and a high fatality rate.[2] Liver cancer incidence has been rising in the United States since at least the mid-1970s,[2, 4] a trend that is projected to continue through at least 2030.[5] A major factor contributing to this increase is the comparatively high prevalence of hepatitis C virus (HCV) infection among those born during 1945 through 1965, also called "baby boomers."[2, 6] The sustained rise in obesity and type II diabetes over the past several decades has also likely contributed to the increasing liver cancer trend.[7]
The incidence of liver cancer varies by race/ethnicity and state, mainly because of differences in the prevalence of major risk factors and, to some degree, because of disparities in access to high-quality care.[1, 2, 8] However, there is limited information on contemporary liver cancer mortality and survival rates in the United States, particularly by race/ethnicity. Recently, the Annual Report to the Nation on the Status of Cancer series (hereafter the Annual Report) featured liver cancer incidence rates and trends but provided limited information on mortality.[2, 3] A small number of other studies that examined liver cancer mortality and/or survival in the United States did not report death rates by either state[9, 10] or race/ethnicity.[3, 11]
In this report, we examine trends in liver cancer incidence, survival, and mortality in the United States and provide contemporary state-level death rates for 5 broad racial and ethnic groups. State-level statistics are particularly important, as they can inform state cancer-control planning, early detection, and prevention efforts. By using data from a population-based subset of the United States, we also describe disparities in survival rates and trends by race/ethnicity and stage. Prevalence and trends in major risk factors for liver cancer, as well as interventions to prevent or mitigate their burden, are also reviewed in detail.
Cancer Occurrence
National Level
Incidence and mortality
Average annual liver cancer incidence (2009-2013) and death (2010-2014) rates for the most recent 5 years of available data in the United States were 7.7 and 6.3 per 100,000, respectively. Both incidence and death rates were notably 2- to 3-fold higher in men than in women for each of the 5 major racial/ethnic groups (Table 1). Liver cancer incidence and death rates were highest for AIs/ANs (15.2 and 11.9 per 100,000, respectively) and were more than double those in NH whites (6.3 and 5.5 per 100,000, respectively), among whom the rates were lowest. Death rates increased with age up to ages 80 to 84 years in all racial/ethnic groups except blacks, among whom mortality peaked at ages 60 to 64 years (Fig. 1).
Liver cancer incidence and mortality patterns are similar, with rates overall increasing since at least 1975 for incidence and since 1980 for mortality.[4] Data by ethnicity are only available since 1992 for incidence (Fig. 2) and since 1990 for mortality (Fig. 3). From 1990 to 2014, death rates rose by 57% in blacks, 69% in Hispanics, and 82% in NH whites; in AIs/ANs, the death rate more than doubled, from 5.4 (per 100,000) to 11.9. In stark contrast, death rates declined slightly in APIs, from 10.8 (per 100,000) in 1990 to 9.6 in 2014. As a result of this divergent trend, a crossover between APIs and AIs/ANs occurred, in that death rates in AIs/ANs were about 40% lower than those in APIs in 1990 but were 10% higher in 2014 (Fig. 3). In all racial/ethnic groups, changes in liver cancer death rates were in the same direction among men and women (Table 2).
Overall, our findings confirm and expand upon results on liver cancer occurrence from the Annual Report series and other published results.[2, 3, 11, 27] Despite rising death rates, however, liver cancer incidence rates in AIs/ANs did not show any increase after 2009 in our analysis. This is likely because AI/AN mortality data in our analysis were from CHSDA counties nationwide, whereas incidence trends were based only on cases from CHSDA counties in areas covered by SEER 13. Indeed, the Annual Report, which used nationwide CHSDA incidence data, reported a statistically significant, continuous increase in liver cancer incidence among AI/AN men from 2003 to 2012,[2] consistent with death rate trends in our analysis.
Stage distribution and survival
Figure 4 shows liver cancer stage distribution by race/ethnicity. The proportion of liver cancers diagnosed at the localized stage ranged from 40% in AIs/ANs to 45% in APIs (approximately 15% of cases were unknown stage). The 5-year liver cancer-specific survival rate was 21% overall for patients diagnosed during 2006 through 2012, ranging from 4% for distant stage disease to 37% for localized disease (Fig. 5). There was little difference in 5-year survival between men and women. Compared with NH whites (20%), survival was statistically significantly lower in blacks (16%) and higher in APIs (27%) (Table 1). Absolute differences in stage-specific survival by race/ethnicity were largest for localized disease, for which survival was highest in APIs (46%) and lowest in blacks (30%) and AIs/ANs (28%) (Fig. 5). Notably, although there was little difference in stage distribution between NH whites and blacks, the 5-year survival rate in blacks was approximately 5% lower for all stages combined (Table 1).
The 5-year survival rate for liver cancer has been increasing since at least 1992 in all races except for AIs/ANs, in whom the trend is inconsistent, in part because of sparse data (Fig. 6). Compared with NH whites, 5-year survival has increased more rapidly in APIs and has lagged in blacks, whereas trends in Hispanics have been similar. By stage, the most significant increase occurred for localized disease, for which the 5-year survival rate increased from 17% for patients diagnosed in the early 1990s to 39% for those diagnosed during 2006 through 2012 in the SEER 13 areas (Fig. 6). A more modest increase in survival was observed for regional-stage liver cancers, which rose from 8% to 14% over the same period. No improvement has been observed for distant-stage disease.
Consistent with our results, earlier studies found a lower 5-year survival rate in blacks than in whites[27] and an increase in 5-year relative survival in all races combined and in studied races (whites, blacks), largely limited to localized-stage and regional-stage liver cancer.[3]
State-Level Mortality
Similar to national patterns, liver cancer death rates in most states were 2- to 3-fold higher in men than in women (Table 3). Racial/ethnic disparities in liver cancer mortality were present in most states, including those with relatively less poverty and low liver cancer death rates, in which low rates among whites masked markedly higher rates among racial/ethnic minority groups (eg, Minnesota and Utah). Among states with available data, liver cancer death rates were statistically significantly higher in black men than in NH white men in all states except Mississippi, New Mexico, Arkansas, and South Carolina. The black-white racial disparity was the highest in the District of Columbia, in which the black rate for both sexes combined (13.3 per 100,000) was more than 4 times higher than that for NH whites (3.2 per 100,000). Liver cancer death rates were generally higher in APIs than in NH whites across states, although in some states with fewer liver cancer deaths among APIs, the difference was not statistically significant. Disparity patterns for Hispanics were less consistent than those for other minority groups, in that liver cancer death rates were higher than in NH whites in only about one-half of states with available data.
The liver cancer death rate in both sexes combined during 2010 through 2014 was highest (≥8 per 100,000) in the District of Columbia, Hawaii, Louisiana, and Texas and lowest (<5 per 100,000) in Iowa, Montana, South Dakota, Nebraska, Utah, Vermont, and North Dakota (Fig. 7). Consistent with a lower liver cancer death rate in NH whites, the death rate was generally higher in states with a lower proportion of NH whites. Among states with comparable proportions of NH whites, those with higher poverty levels tended to have higher liver cancer death rates. For example, the rate was higher in Kentucky (6.1 per 100,000; population, 87% NH whites) than in North Dakota (3.8 per 100,000; 89% NH whites).
The geographic patterns of incidence described in the Annual Report were generally comparable to the mortality patterns in our study, with some differences. For example, in the Annual Report, Alabama and Louisiana ranked among the second and third quartile of states, respectively, in terms of liver cancer incidence rates[2] but were among the 10 states with highest liver cancer death rates in our analysis. These differences may be related to variation in access to care across states.
Risk Factors
Current Status and Trends
The majority (approximately 60%) of liver cancers in the United States are attributed to potentially modifiable risk factors, although these proportions vary somewhat because of differences in data sources, study periods, and methodology.[28-30] Makarova-Rusher et al estimated that, among US Medicare recipients during 2000 through 2011, metabolic disorders (including excess body weight, diabetes, impaired glucose tolerance, nonalcoholic fatty liver disease, and metabolic syndrome) accounted for the largest proportion of hepatocellular carcinomas (HCCs) (32%), followed by chronic HCV infection (21%), excessive alcohol drinking (13%), smoking (9%), and chronic HBV infection (4%).[28] HCC is the major histologic type of primary liver cancer, accounting for greater than 70% of all liver cancers in the United States.[11] However, another recent study indicated that 24% of all liver cancer deaths in adults ages 35 years and older in 2011 were caused by smoking.[30] Although the magnitude of association between HBV/HCV infection and HCC is much greater (relative risk [RR], 22-60) than that of excessive alcohol drinking (RR, 7.3), metabolic disorders (RR, 3.8),[28] or smoking (RR, 1.5),[31] these factors account for a substantial number of HCCs in the United States because of their higher population prevalence. Rare genetic disorders account for only 2% to 3% of HCCs.[28, 29]
Differences in risk factor prevalence accounts for much of the variation in liver cancer rates by sex, age group, state, and race/ethnicity. For example, higher liver cancer death rates in Texan Hispanics (predominantly of Mexican origin) than in Floridian Hispanics (mostly of Cuban or Puerto Rican origin) may reflect higher prevalence of chronic HCV infection, excess body weight, and nonalcoholic fatty liver disease in Mexican Americans than in Hispanics of Cuban or Puerto Rican origin, as well as differences in immigration history.[32, 33] Among populations with a large proportion of recent immigrants (APIs and Hispanics), aggregated liver cancer rates are influenced by high background HCV/HBV prevalence in some countries of origin.[34, 35] Conversely, acculturation across generations in these populations also increases liver cancer risk, as several lifestyle risk factors are more common in the United States (eg, obesity, alcohol consumption, and smoking) than in many parts of Asia and Latin America.
Hepatitis C virus
HCV infection accounts for approximately 20% of liver cancers worldwide,[36] with prevalence ranging from 3.5% or more in many countries in Central, East, and Western Asia and North Africa to less than 1% in several higher income countries in Western Europe.[37] According to NHANES, the prevalence of chronic HCV infection (defined as positive HCV RNA testing) in the United States was 1% during 2003 through 2010, representing approximately 2.7 million people.[6, 38] However, this number is likely an underestimate, because populations with very high infection prevalence (eg, heavy drug users and incarcerated or homeless people) are not represented in NHANES.[37, 38] A study incorporating these populations estimated that at least 3.5 million adults were infected with HCV as of 2010.[38] People born between 1945 and 1965 ("baby boomers") have an HCV prevalence of approximately 2.6% and account for about 80% of HCV-infected cases,[6] which explains the increased age-specific death rates for individuals ages 50 to 64 years in the current study. Major risk factors for chronic HCV infection in the United States are injection drug use and receipt of a transfusion before 1992.[6] Other factors that increase risk include being born to an HCV-infected mother, receiving high-risk tattoos or body piercing, needle-stick injuries, having sexual contact with infected people, and some medical conditions (eg, transplantation, hemodialysis, and human immunodeficiency virus infection).[39] HCV infection prevalence in the United States is higher among those with lower education or income and among ages 20-59 years is nearly twice as high among men as among women.[2, 6]
According to NHANES, blacks in the United States are more likely than other races/ethnicities combined to have chronic HCV infection, although the prevalence for other individual minority groups is not available because of sparse data.[6] Some surveys and reports from state health departments indicate that HCV infection is generally more common in AIs/ANs, APIs, and Hispanics than in NH whites but may vary by geographic location.[34, 35, 40-43] Overall, the proportion of all HCCs during 2000 through 2011 attributable to HCV was 17% in whites, 21% in Hispanics, 30% in Asians, and 36% in blacks.[28]
Although representative information on HCV prevalence across states is not available, states with higher liver cancer incidence rates have generally reported higher HCV infection prevalence.[40, 42, 43] Higher prevalence of HCV infection in nonwhites and those with lower income is consistent with higher liver cancer death rates in states with both a higher proportion of nonwhites and high poverty.
Hepatitis B virus
Although HBV infection is the most common cause of liver cancer globally, accounting for over one-half of liver cancer cases worldwide, it accounts for less than 5% of liver cancers in the United States.[36] HBV infection prevalence is 5% or greater in many low- and middle-income countries in sub-Saharan Africa, Central and East Asia, and the Western Pacific, as well as a small number of countries in other regions, but it is generally low (<2%) in high-income countries.[44, 45] Transmission of HBV infection in endemic countries is most common from mother to infant (vertical transmission), whereas in other countries, it mainly occurs through direct contact with body fluids, such as sharing syringes for drug use, needle-stick injuries, and sexual contact (horizontal transmission).[44]
Since the 1990s, the overall prevalence of chronic HBV infection (defined as positive for hepatitis B surface antigen) in the United States has been stable at 0.3% to 0.4% in individuals ages 20 years and older but has declined substantially (from 0.2% to 0.03%) in those ages 6 to 19 years, largely reflecting the influence of HBV vaccination uptake among children.[22] During 2011 to 2012, an estimated 850,000 people in the United States had chronic HBV infection.[22] However, prevalence varies greatly by race/ethnicity, ranging from 0.1% in NH whites and Mexican Americans to 0.6% in blacks and 3.1% in Asians.[22] Prevalence is also higher among foreign-born individuals (1.1%) compared with those born in the United States (0.1%),[22] with those born in China and other highly endemic countries having substantially higher prevalence (≥9%).[46] Consequently, 93% of Asians with chronic HBV infection are foreign-born,[22] and the proportion of all HCCs during 2000 through 2011 attributable to HBV was 18% in Asians, compared with <3% in whites, blacks, and Hispanics.[28] A considerable proportion of liver cancers among Pacific Islander and AI/AN adults may also be attributable to HBV infection because of their high prevalence of HBV infection, although these data are unavailable.[47, 48]
Metabolic disorders and excess body weight
Prospective studies have demonstrated that liver cancer risk increases by 26% per 5 kg/m2 increase in body mass index (BMI) in the United States,[7] where 69% of people ages 20 years and older have excess body weight (BMI ≥25 kg/m2).[49] From the 1960s to 2013-2014, the prevalence of overweight (BMI 25.0-29.9 kg/m2) among adults remained high but stable at 32% to 34%,[50] while obesity (BMI ≥30 kg/m2) tripled from 13% to 38%, and class 3 obesity (BMI ≥40 kg/m2) increased 8-fold, from 1% to 8%.[21] These increases occurred in both men and women until the late 1990s, but since have been mainly confined to women.[21, 50] Currently, overweight prevalence is higher in men (38%) than in women (30%), but more women than men are obese (40% vs 35%) and class 3 obese (10% vs 6%).[21] The prevalence of obesity has substantially increased in all races/ethnicities with available historical information.[50] Among women, obesity prevalence ranges about 3-fold to 5-fold, from a low of 12% in Asians to 38% in NH whites, 47% in Hispanics, and 57% in blacks.[21] Among men, obesity prevalence in NH whites (35%), blacks (38%), and Hispanics (38%) is similar but is nearly 3-fold the prevalence in Asians (13%).[21]
Prevalence of obesity by state and race/ethnicity is shown in Figure 8. In most cases, patterns of obesity prevalence in each racial/ethnic group correspond to liver cancer death rates. For example, both obesity prevalence and liver cancer mortality in the District of Columbia are among the lowest in the country in NH whites but are very high in blacks. Obesity may also contribute to substantially higher liver cancer death rates in blacks and Hispanics than in NH whites in Texas. In some other instances, however, this pattern is more equivocal. For example, although NH whites in Hawaii have lower obesity prevalence than NH whites in the continental United States, their liver cancer death rate is higher. These inconsistencies likely reflect differences in the prevalence of other major contributors to liver cancer, notably viral hepatitis infection.
A large-scale, prospective study in the United States has demonstrated that diabetes increases the risk of liver cancer by 2.3-fold in men and 1.4-fold in women.[7] Consistent with the increase in excess body weight, the prevalence of diabetes has increased from 9.8% in the late 1980s to 12.4% in 2011-2012, and it is estimated that an additional 36.5% are prediabetic.[23] The increase in diabetes over 2 decades has been more modest in NH whites (from 8.6% to 9.5%) than in blacks and Mexican Americans (from approximately 17% to 21%). Although trend data are not available for other minority groups, the contemporary prevalence of diabetes was relatively high in all Hispanics combined (18.7%) and in Asians (16.5%),[23] despite lower BMIs in the latter.[21] Diabetes prevalence is also high and increasing rapidly in those with low education and income.[23, 51]
More than 20% of patients with nonalcoholic fatty liver disease, a condition highly associated with excess body weight and diabetes, will develop nonalcoholic steatohepatitis, a more serious type of nonalcoholic fatty liver disease in which fat accumulation in the liver causes inflammation and cell damage, thereby increasing the risk of liver cancer.[52] There is little information on trends in nonalcoholic steatohepatitis prevalence from population studies, partly because diagnosis may require invasive procedures. However, a few studies have reported a rising trend in the prevalence of nonalcoholic fatty liver disease in the United States, consistent with the increases in excess body weight and diabetes prevalence, although there may be variations in reported prevalence based on diagnostic criteria used in those studies.[24, 53] An analysis of NHANES data reported an increase in the prevalence of nonalcoholic fatty liver disease from 18% during 1988 through 1991, to 29% during 1999 and 2000, and to 31% during 2011 and 2012.[24] Among the 3 racial/ethnic groups in that analysis, Mexican Americans had the highest prevalence, followed by NH whites and blacks. An analysis of data from more than 1.3 million veterans using different diagnostic criteria also demonstrated an increase, although the reported prevalence was lower (an increase from 6% in 2003 to 18% in 2011); racial patterns were similar to those in NHANES.[52]
Alcohol drinking and tobacco smoking
Alcohol consumption increases liver cancer risk by approximately 10% per drink per day (eg, a 3-fold increased risk for someone consuming 8 drinks per day).[54] Excessive drinking includes binge drinking (≥4 drinks per occasion for a woman or ≥5 drinks per occasion for a man) and heavy drinking (≥8 drinks per week for a woman or ≥15 drinks per week for a man).[55] Binge drinking is the most common type of excessive alcohol drinking in the United States-17% of adults (or 38 million people) binge drink.[55] On the basis of the National Survey on Drug Use and Health, binge drinking prevalence among individuals aged 50 years and older has increased by approximately 20% in the last decade.[26] Men, those with a higher income, and tobacco and illicit drug users are most likely to binge drink. Compared with NH whites, the self-reported prevalence of binge drinking is similar in blacks, lower in Asians, and slightly higher in Hispanics.[26] As illustrated in Figure 9, states with the highest prevalence of binge drinking are in the North and Midwest, whereas the intensity of binge drinking is highest in the central South and parts of the Midwest.
Smoking increases liver cancer risk by approximately 50%.[31] Smoking prevalence among men is highest in AIs/ANs (38%), followed by blacks (21%), NH whites (17%), and APIs and Hispanics (12% to 13%). Prevalence in women is generally much lower than in men within the same racial/ethnic group (24% in AIs/ANs, 13% in blacks, 7% in Hispanics, and 3% in APIs), with the exception of NH white women (16%).[25] Despite overall reductions in smoking in the United States, prevalence is highest and declines are weakest among individuals with lower education levels and incomes,[25] who have a higher likelihood of other liver cancer risk factors. States with the highest smoking prevalence are in the Southern region among NH whites, in Southern states and the Midwest in black men, and in the Midwest in black women.[56]
Interventions
Many liver cancer cases could potentially be prevented by alleviating exposure to risk factors or mitigating the risk among those exposed. The following is a broad summary of current strategies along with their limitations and includes interventions intended for low-income individuals, minorities, and other high-risk groups. Tailored, culturally appropriate application of these interventions for different subpopulations at the state and local level is essential but remains a challenge, particularly considering the disparities in health care access and limitations in public health resources.
Hepatitis virus infection
The Institute of Medicine has established 4 general groups of recommendations for the prevention and control of HBV and HCV, including improvements in infection surveillance, awareness, infection care and services, and HBV vaccination.[57] More representative data are required on HBV and HCV prevalence among subpopulations in the United States, including data by state and from high-risk populations (eg, the homeless). Increasing awareness among policy makers, health care providers, and the general population is an essential part of any strategy to control HBV and HCV infection. Awareness is even more important for those at risk and the individuals/organizations that serve them, because many are unaware of viral hepatitis infection screening, HBV vaccination, or the availability of treatment to cure HCV infection. Currently, the CDC recommends one-time HCV testing for people born from 1945 to 1965.[39] However, only 14% of this cohort has reported HCV testing, with little variation between Asians, NH whites, blacks, and Hispanics.[58] In contrast, a recent assessment indicated that HCV testing among the AI/AN 1945 to 1965 birth cohort served by the Indian Health Service increased 4-fold during 2012 through 2015, from 8% to 33%, likely due in part to the implementation of integrated electronic clinical decision-support tools at Indian Health Service facilities as well as other provider-based initiatives.[59] Some professional organizations also recommend screening for high-risk individuals,[39] including the incarcerated population, which may be one of the cost-effective ways to reduce transmission of HCV infection.[60]
Although interferon and/or ribavirin have been used with some success to treat HCV infection, the introduction of oral direct-acting antivirals in 2014 represented major progress, resulting in very high cure rates with shorter term treatment and substantially fewer adverse side effects.[61, 62] Although future studies are needed to determine the long-term effects of these new drugs, they are likely to reduce HCV complications, including HCC. However, underdiagnoses of HCV cases, in addition to obstacles to care and the high cost of direct-acting antivirals, are major barriers. For example, only slightly more than one-third of those with known HCV infection receive follow-up care, and less than 10% are successfully treated.[39] The variation in reimbursement of direct-acting antivirals because of their high cost further contributes to suboptimal access in states with more restricted coverage.[39]
Unlike HCV, treatments for HBV infection, including pegylated α-interferon and nucleos(t)ide analogues, may halt the progression of active liver disease but do not cure the infection.[62, 63] Thus, although these treatments have been able to reduce the risk of HBV-related HCC, they require long-term administration.[64] HBV vaccination is the most effective means of preventing HBV infection and liver cancer and has been part of routine childhood vaccinations in the United States since 1982.[65-67] In 2014, HBV vaccination coverage in children ages 19 to 35 months in the United States was 91% to 93% in NH whites, blacks, Hispanics, and Asians, and it was slightly higher in Pacific Islanders (95%) and AIs/ANs (99%).[68] Despite historically high HBV infection prevalence among Pacific Islanders and Alaska natives, comprehensive universal infant vaccination has been very successful in lowering the childhood infection rate in these populations.[47, 48] For example, HBV prevalence in children ages 2 to 11 years in the US-affiliated Pacific Island countries decreased from 8.4% in the 1980s to 0.2% in the 2000s.[48] Although greater than 90% of adolescents in the United States received the HBV vaccine in 2015, coverage varied across states, ranging from 83% in Idaho to 98% in New Hampshire.[69] HBV vaccination of high-risk adults is also recommended, but coverage remains low at only 50% in the United States.[57] Both HBV and HCV infection can be prevented through the expansion of comprehensive programs to reduce transmission through high-risk behaviors (eg, using shared syringes).[57]
Excess body weight
Consistent with a substantial increase in the number of people with excess body weight and associated metabolic disorders in the United States, the estimated proportion of HCCs attributable to these factors increased from 26% during 2000 through 2003 to 36% during 2008 through 2011.[28] Excess body weight in childhood is a major contributor to the obesity epidemic, because, for many children, it extends into adulthood; it is also a risk factor for type II diabetes and other metabolic disorders associated with liver cancer in adults.[70, 71] Therefore, preventing excess body weight in children should be a major focus of any weight-control intervention at the population level. Among children and adolescents ages 2 to 19 years, approximately one-third have excess body weight and 17% are obese, although there is substantial variation by race/ethnicity and across states.[49, 72, 73]
The etiology of childhood obesity is complex and is influenced by multiple behavioral, cultural/environmental (eg, family, school), socioeconomic, and genetic factors, which can lead to unhealthy diet, physical inactivity, and adverse changes in metabolism.[70, 74-77] The relative effect of these factors may vary across subpopulations, including by race/ethnicity.[75-77] Because children generally spend a considerable amount of time in school (including preschool), school-based interventions can provide a great opportunity for weight control, such as education about healthy lifestyles and increasing physical activity and access to affordable, healthy food while reducing access to unhealthier foods at school.[74] However, the effects of school-based interventions may be difficult to sustain long term, and studies have shown that additional environmental strategies could improve outcomes. For example, family-based interventions may help to improve diet, increase physical activity, and reduce screen time at home.[74-76]
Among adults, intensive lifestyle interventions (including diet, physical activity, and behavior therapy) to promote healthier diet and physical activity have been shown to be successful in maintaining or reducing body weight and diabetes risk among participants in several studies.[78-80] However, these interventions have not been as successful at the population level, including when offered through primary care providers.[81-83] This inconsistency may be due in part to participation bias, but it may also reflect primary care providers' time constraints and insufficient training for delivering effective behavioral counseling, which could be remedied through appropriate instruction or referrals.[81-84] When lifestyle interventions are not successful for adequate weight loss, individuals with obesity may be eligible for pharmacotherapy, and those with class 3 obesity may benefit from bariatric surgery.[85]
The use of modern communication technologies, such as social and mobile tools, may have the potential to increase the efficiency and accessibility of behavioral interventions, but studies to date have produced mixed results.[86-88] Some regulations and policies may also help weight-control strategies in both children and adults, including mass media campaigns (increasing awareness and education to change unhealthy behaviors), food labeling and advertising regulation, taxation of nonessential high-calorie foods and sugary beverages, and improvements in urban structure (eg, increasing public transportation).[89-92] Overall, maintaining normal weight may be more challenging for minorities and individuals with low income, because they often live in neighborhoods with more limited access to healthier foods and appropriate environments for physical activity.[75, 76] For example, poor neighborhoods usually have limited sports facilities and programs (in schools and outside) and fewer or unsafe sidewalks.[75, 76]
Increased awareness, the availability of multiple recommendations and guidelines for maintaining a healthy body weight, and the implementation of some community-level interventions and regulations may have contributed to a modest decrease in obesity in children ages 2 to 5 years and stable trends in children ages 6 to 11 years as well as adult men in the United States during the last decade.[21, 49, 72] However, to adequately address the obesity epidemic and associated metabolic disorders, comprehensive implementation of successful interventions and additional research to identify more efficient strategies at the population level, especially in racial/ethnic minorities, are needed.[72]
Other risk factors
In addition to reducing weight and increasing physical activity, appropriate diabetes care can prevent or delay complications like nonalcoholic fatty liver disease. However, undiagnosed diabetes remains a substantial challenge in the United States; one-third of patients with diabetes in the most recent NHANES survey were diagnosed by laboratory tests used in the survey and had not been previously diagnosed.[23] Undiagnosed diabetes was more common in Asians and Hispanics (50%) than in blacks (37%) and NH whites (32%). This highlights the importance of access to preventive medicine and routine blood sugar tests. However, disparities in access to high-quality diabetes care persist even among those diagnosed with the disease, largely driven by lack of insurance coverage,[93] particularly among Hispanics overall and lower income groups.[94, 95]
For preventing excessive alcohol drinking, the Community Preventive Services Task Force recommends increasing alcohol excise taxes; regulating alcohol outlet density and the days and hours of alcohol sale; maintaining government controls over alcohol sales; using computers, mobile phones, and other electronic devices for interventions; and holding retailers accountable for harms when they illegally sell/serve alcohol.[55] Effective policies for reducing tobacco control have been discussed comprehensively in several publications.[31] Briefly, the policies recommended by the World Health Organization's Framework Convention on Tobacco Control include tobacco taxation, smoke-free laws, warning about the dangers of tobacco (warning labels and media campaigns), marketing bans, and assistance with smoking cessation.[96] Smoking prevalence across states is highly correlated with tobacco-control policies in states.[56, 97] Although smoking may play a smaller role than some other risk factors in the burden of liver cancer, reducing tobacco use will also prevent numerous deaths from many other cancers and chronic diseases.[30, 97]
Treatment and Survival
HCC has a very poor prognosis. The overall median survival for patients with untreated disease is 4 months, ranging from 2 months for late-stage cancers to 14 months for very early stage cancers.[98] Potentially curative treatments are available for early stage liver cancers, including surgical resection, ablation, and liver transplantation.[99-102] Currently, HCC is the leading diagnosis among liver transplant recipients (27% in 2015), and the number of liver transplants because of liver cancer continues to increase.[103] Transarterial embolization and transcatheter arterial chemoembolization are usually used to treat patients with intermediate-stage liver cancer who do not have tumor-related symptoms, vascular invasion, or metastasis and may increase survival.[100-102, 104] Improvements in and access to treatments are the likely reasons for a substantial increase in localized-stage liver cancer survival and, to a lesser degree, regional-stage survival, as demonstrated in our analysis. Systemic treatment with sorafenib (an oral multikinase inhibitor) is used for more advanced HCCs, except for end-stage cancers, which are generally considered for palliative care only. However, sorafenib can increase survival for only a few months.[100-102] This, along with the lack of any specific treatment for end-stage HCC, explains the little improvement in 5-year survival for advanced cancers over the entire study period in our analysis.
Variation in treatment may account in part for differential stage-specific survival by race and ethnicity. Disparities in treatment for patients diagnosed with potentially curable, early stage disease have been documented in several studies. A study of patients diagnosed from 1998 to 2011 with clinical stage I/II HCC indicated that slightly less than 40% underwent surgery. A lower probability of undergoing surgery was observed for black and uninsured or Medicaid-insured patients in multivariate models that included stage and comorbidity score. In that study, the median survival of patients who underwent surgery was 48.3 months compared with 8.4 months for those who did not undergo surgery.[105] Other studies have noted pronounced disparities in the receipt of liver transplantation among patients who were eligible for this treatment. One such study of patients with early stage HCC reported to SEER registries from 1998 to 2010 who were eligible for liver transplantation based on the Milan criteria indicated that black, Asian, and Hispanic patients were substantially less likely to receive this treatment than NH whites.[106] A study in California by race and ethnicity examined HCC treatment and survival patterns in greater detail for Asian subpopulations. Consistent with previous studies, that study indicated that the probability of receiving surgical treatment was lower for black and Hispanic patients than for NH white patients. In addition, the study found considerable variations in surgical treatment among Asian subpopulations, with the lowest rates reported for Laotian/Hmong and Cambodian patients and the highest rates reported for Chinese, Korean, Japanese, and Vietnamese patients.[107] Those studies suggested that considerable progress in reducing liver cancer mortality could be achieved by ensuring equitable access to high-quality treatment, especially for patients with early stage disease for whom surgery potentially can be curative.
Part of the improvement in survival for localized-stage cancers may also be related to earlier detection. The American Association for the Study of Liver Diseases has developed guidelines for the surveillance of high-risk individuals, such as certain patients with cirrhosis, HBV- or HCV-related chronic active hepatitis, or other chronic liver diseases that meet certain criteria.[100, 101] However, although studies have suggested that such surveillance is cost-effective and could lead to a greater percentage of liver cancers diagnosed at earlier stages, it remains controversial.[100, 101, 108-110] Because of a lack of high-quality data from clinical trials, it is unclear whether liver cancer surveillance could translate into lower mortality from the disease,[100, 101, 110] demonstrating the need for further research.
Limitations
Population-based cancer data in the United States are generally available only for the 5 broad groups presented herein. However, liver cancer rates can vary substantially by subgroup within these large, heterogeneous populations. For example, as pointed out by the Office of Management and Budget in 1997, the broad category of Asian and Pacific Islander is problematic because it includes subpopulations with vastly disparate cultural, socioeconomic, and health profiles, which lead to substantial variations in cancer occurrence. In addition, caution is warranted when interpreting death rates and cancer-specific survival for racial groups other than NH whites and blacks, because both racial/ethnic misclassification on death certificates and nonrandom censoring because of biases in patient follow-up are more likely to occur. One recent analysis indicated that APIs and Hispanics were substantially more likely to be censored (lost to follow-up) within the first 5 years after a cancer diagnosis than NH whites, particularly for cancers with an unfavorable prognosis.[111] Finally, liver cancer incidence and mortality rates in AIs/ANs were based on a relatively modest number of outcomes and should be interpreted with caution.
Conclusions
The burden of liver cancer in the United States is significant and is expected to increase in the decades to come. Moreover, despite some improvements in localized and regional disease survival rates in the 2 most recent decades of available data, the overall prognosis for liver cancer remains poor. Wide disparities in liver cancer death rates by sex, race/ethnicity, and state persist, reflecting differences in the prevalence of major risk factors and, to some extent, inequalities in access to high-quality care. However, most liver cancers are potentially preventable, and interventions to curb the rising burden of liver cancer and reduce racial/ethnic disparities should include the targeted application of existing knowledge in prevention, early detection, and treatment, including improvements in HBV vaccination, screening and treatment of HCV, maintaining a healthy body weight, access to high-quality diabetes care, prevention of excessive alcohol drinking, and tobacco control.
|
|
| |
| |
|
|
|
