Each year, the American Cancer Society estimates the number of new cancer cases and deaths expected in the United States in the current year and compiles the most recent data on cancer incidence, mortality, and survival rates based on incidence data from the National Cancer Institute and mortality data from the National Center for Health Statistics. Incidence and mortality rates are age standardized to the 2000 US standard million population. A total of 1,368,030 new cancer cases and 563,700 deaths are expected in the United States in 2004. Incidence rates stabilized among men from 1995 through 2000 but continued to increase among females by 0.4% per year from 1987 through 2000. Mortality rates have decreased by 1.5% per year since 1992 among men, but have stabilized from 1998 through 2000 among women. Cancer death rates continued to decrease from the three major cancer sites in men (lung and bronchus, colon and rectum, and prostate) and from female breast and colorectal cancers in women. In analyses by race and ethnicity, African-American men and women have 40% and 20% higher death rates from all cancers combined compared with White men and women, respectively. Cancer incidence and mortality rates are lower in other racial and ethnic groups than in Whites and African Americans for all sites combined and for the four major cancer sites. However, these groups generally have higher rates for stomach, liver, and cervical cancers than do Whites. Furthermore, minority populations are more likely to be diagnosed with advanced stage disease than are Whites. Progress in reducing the burden from cancer can be accelerated by applying existing cancer control knowledge into practice among all segments of the population.
Cancer is a major public health burden in the United States and in other developed countries. Currently, one in four deaths in the United States is due to cancer. In this article, we provide an overview of the most frequently used cancer statistics, including frequency, incidence, mortality, and survival rates for the year 2004.
MATERIALS AND METHODS
We obtained mortality data from 1930 to 2001 from the National Center for Health Statistics.1 We obtained incidence data (1975 to 2000), five-year relative survival rates, and data on lifetime probability of developing cancer from the Surveillance Epidemiology and End Results (SEER) program of the National Cancer Institute, covering approximately 10% to 14% of the US population.2–5 We obtained population data from the US Census Bureau.6 Causes of death were coded and classified according to the International Classification of Diseases (ICD-8, ICD-9, and ICD-10).7–9 Cancer cases were classified according to the second edition of the International Classification of Diseases for Oncology.10
Estimated New Cancer Cases
The precise number of cancer cases diagnosed each year in the nation and in every state is unknown because complete cancer registration has not yet been achieved in many states. Consequently, for the national estimate, we first estimate the number of new cancer cases occurring annually in the United States from 1979 through 2000 by using age-specific cancer incidence rates collected by the SEER program2 coupled with population data reported by the US Census Bureau.6 We then forecast the number of cancer cases expected to be diagnosed in the United States in the year 2004 using an autoregressive quadratic time trend model fitted to the annual cancer case estimates.11 For estimates of new cancer cases in individual states, we relied on state cancer death statistical data and assumed that the ratio of estimated cancer deaths to cases in each state was equal to that in the United States.
Estimated Cancer Deaths
For this report, we used a new method for estimating the number of cancer deaths expected to occur in the United States and in each state in the year 2004. The method is described by Tiwari et al. in an accompanying article in this issue of CA.12 We based our projections on underlying causes of death recorded on death certificates as reported to the National Center for Health Statistics.1 The numbers of cancer deaths recorded annually from 1969 to 2001 in the United States and in each state were fitted using a state-space model to forecast the number of cancer deaths expected to occur in 2004.
We provide mortality statistics for the leading causes of death and deaths due to cancer in the year 2001. Causes of death for 2001 were coded and classified according to ICD-10.7 This report also provides updated statistics on trends in cancer incidence and mortality rates, probability of developing cancer, and five-year relative survival rates for selected cancer sites based on data from 1973 through 2000.3 All age-adjusted incidence and mortality rates are standardized to the 2000 US standard population and expressed per 100,000 population.
The long-term incidence trends (1975 to 2000) presented in the table are adjusted for delays in reporting. Delayed reporting affects the most recent one to three years of incidence data (in this case, 1998 to 2000) as certain types of cancer continue to be reported. Cancers that are frequently reported late are those diagnosed in outpatient settings, such as melanoma and prostate cancers. The National Cancer Institute has developed a method to account for expected reporting delays for all cancer sites combined and for several specific cancer sites when long-term incidence trends are analyzed.13 Delay-adjusted trends provide a more accurate assessment of trends in the most recent years for which data are available.
Expected Numbers of New Cancer Cases
Table 1 presents estimated numbers of new cancer cases expected in the United States in 2004 by sex. The estimate of approximately 1.4 million new cases of invasive cancer does not include carcinoma in situ of any site except the urinary bladder, nor does it include basal and squamous cell cancers of the skin. More than one million cases of basal and squamous cell skin cancer, approximately 59,390 cases of breast carcinoma in situ, and 40,780 cases of in situ melanoma are expected to be newly diagnosed in 2004. Table 2 shows the estimated numbers of new cancer cases for each state and selected cancer sites.
Figure 1 lists the most common cancers expected to occur in men and women in 2004. Among men, cancers of the prostate, lung and bronchus, and colon and rectum account for more than 55% of all newly diagnosed cancers. Prostate cancer alone accounts for approximately 33% (230,110) of incident cases in men. Based on cases diagnosed between 1992 and 1999, approximately 86% of these estimated new cases of prostate cancer are expected to be diagnosed at local or regional stages, for which the five-year relative survival rate equals 100%.
The three most commonly diagnosed cancers among women in 2004 will be those of the breast, lung and bronchus, and colon and rectum, accounting for approximately 55% of estimated cancer cases in women. Breast cancer alone is expected to account for 32% (215,990) of all new cancer cases among women.
Expected Numbers of New Cancer Deaths
Table 1 also shows the expected numbers of cancer deaths in 2004 for men, women, and both sexes combined. It is estimated that approximately 563,700 Americans will die of cancer, corresponding to over 1,500 deaths per day. Cancers of the lung and bronchus, prostate, and colon and rectum in men, and cancers of the lung and bronchus, breast, and colon and rectum in women continue to be the most common causes of cancer death. These four cancers account for one half of the total number of cancer deaths among men and women (Figure 1). Lung cancer surpassed breast cancer as the leading cause of cancer death in women in 1987. Lung cancer is expected to account for 25% of all female cancer deaths in 2004. Table 3 provides the estimated number of cancer deaths in 2004 by state for selected cancer sites.
Trends in Cancer Incidence and Mortality Rates
Figures 2 to 5, FIGURE 3, FIGURE 4, FIGURE 5 depict long-term trends in cancer incidence and mortality rates for all cancers combined and for selected cancer sites by sex. Table 4 shows incidence and mortality patterns for all cancer sites and for the four common cancer sites based on joinpoint analysis. Trends in incidence were adjusted for delayed reporting in the table, but not in the figures (Figures 2 and 3, FIGURE 3). Cancer death rates for all cancer sites combined decreased by 1.5% per year from 1992 to 2000 in males, but they stabilized from 1998 to 2000 in females after declining by 1.3% from 1995 to 1998 (Table 4). The leveling of rates among females is in part due to the implementation of ICD-10 rules for coding and selecting the underlying cause of death. Approximately 0.7% more deaths were classified as cancer using ICD-10 rules than using ICD-9 rules,14 creating discontinuity in the most recent cancer mortality rate trend.15,16 Delay-adjusted cancer incidence rates stabilized in men from 1995 to 2000 and increased by 0.4% per year from 1987 to 2000 in women (Table 4).
Mortality rates have continued to decrease across all four major cancer sites in men and in women (Table 4). The incidence trends are mixed, however. Lung cancer incidence rates are declining in men but still increasing among women, although the rate of increase has slowed. Colorectal cancer incidence rates have stabilized since the mid 1990s in males and females. Female breast and prostate cancer incidence rates have continued to increase, although at a slower rate than in previous years, in part due to increased screening using mammography and prostate-specific antigen testing, respectively. The recent increase in female breast cancer incidence rates involves both small and large tumors, 17 which may reflect increased use of hormone replacement therapy, an increased prevalence of obesity, or both.
Changes in the Recorded Number of Deaths Due to Cancer From 2000 to 2001
A total of 553,768 cancer deaths were recorded in the United States in 2001, the most recent year for which actual data are available. Nearly 700 more cancer deaths were recorded in 2001 than in 2000, reflecting growth and aging of the population. Cancer accounted for approximately 23% of all deaths, ranking second only to heart disease (Table 5). When cause of death is ranked within each age group, categorized in 20-year age intervals, cancer ranks as one of the five leading causes of death in each age group among both males and females. Cancer is the leading cause of death among women aged 40 to 79 years and among men aged 60 to 79 years (Table 6). Overall, among men and women combined, cancer is the leading cause of death before age 65, followed by heart disease (Figure 6). A total of 163,546 persons younger than 65 years died of cancer in the United States in 2001, compared with 117,346 persons who died of heart disease.
Table 7 presents the leading site-specific causes of cancer death by age for males and females in 2001. Among men younger than 40 years, leukemia is the most common fatal cancer, whereas cancer of the lung and bronchus predominate in men aged 40 years and older. Colorectal cancer is the second most common fatal cancer among men aged 40 to 79 years, and prostate cancer is the second most common among men aged 80 and older. Among women, the leading causes of cancer death are leukemia among those younger than 20 years, breast cancer among those aged 20 to 59 years, and lung cancer in those 60 years and older.
From 2000 to 2001, the number of recorded cancer deaths increased by 993 in men, whereas it decreased by 316 in women. The total number of cancer deaths decreased slightly across all of the four major cancer sites, except for lung cancer in females. Five hundred ninety more women in the United States died of lung cancer in 2001 than in 2000 (Table 8).
Cancer Occurrence by Race and Ethnicity
Cancer incidence and mortality rates vary considerably among racial and ethnic groups (Table 9). For all cancer sites combined, African American men have a 25% and 43% higher cancer incidence and mortality rate, respectively, than do White men. African-American women have lower incidence rates than do White females for all cancer sites combined, and yet they have 20% higher mortality rate. For the specific cancer sites listed in Table 9, incidence and mortality rates are consistently higher in African Americans than in Whites, except for breast cancer (incidence) and lung cancer (mortality) among women. Death rates from stomach and cervical cancers among African Americans are more than twice the rates in Whites. Factors that contribute to mortality differences in African Americans and Whites include differences in exposure (eg, hepatitis C for liver cancer), access to regular screening (breast, cervical, and colorectal cancers), and timely, high-quality treatments (many cancers). The higher breast cancer incidence rates among White females are thought to reflect a combination of more frequent mammography, delayed childbearing, and perhaps greater use of hormone replacement therapy.
Among other racial and ethnic groups, cancer incidence and mortality rates are lower for all cancer sites combined and for the four most common cancer sites than those for Whites and African Americans. However, incidence and mortality rates for cancers of the uterine cervix, stomach, and liver are generally higher in minority than in White populations. Stomach and liver cancer incidence and mortality rates are more than twice as high as in Asian/Pacific Islanders than in Whites, reflecting increased exposure to infectious agents such as H pylori and hepatitis C.
Historical information to adjust for delays in reporting is not available for all racial and ethnic groups. From 1992 to 2000, delay-unadjusted incidence rates for all cancer sites combined decreased by 3.1% per year among American Indians/Alaskan Natives, by 1.0% per year in African Americans, by 0.6% among Whites, and by 0.5% among Asian/Pacific Islanders, whereas incidence rates stabilized among Hispanic-Latinos (data not shown). Similarly, the death rate from all cancers combined decreased annually from 1992 through 2000 by 1.5% per year in Asian/Pacific Islanders, 1.3% among African Americans, 0.9% among Whites, and 0.3% among Hispanic-Latinos, but it stabilized among American Indians/Alaskan Natives.3
Lifetime Probability of Developing Cancer
The lifetime probability of developing cancer is higher for men (45%) than for women (38%) (Table 10). However, because of breast cancer, women have a slightly higher probability of developing cancer before the age of 60. It is noteworthy that these estimates are based on the average experience of the general population. Thus they may overestimate or underestimate individual risk because of differences in exposure, family history, or both.
Cancer Survival Rates by Race
A poorer probability of survival once cancer is diagnosed contributes to the higher mortality rates among African-American men and women. African Americans are less likely than Whites to be diagnosed with cancer at a localized stage, when the disease may be more easily and successfully treated, and they are more likely to be diagnosed with cancer at a regional or distant stage. This is true for most of the common cancer sites (Figure 7). Furthermore, for nearly every cancer site, African Americans have lower five-year relative survival rates than do Whites at each stage of diagnosis (Figure 8), suggesting the possible influences of disparities in access and receipt of quality health care and of differences in comorbid conditions. The extent to which these factors, individually or collectively, contribute to the overall differential survival rate is not entirely clear.18 However, recent findings suggest that African Americans and Whites have similar disease outcomes when they receive similar cancer treatment and medical care.19
The relative five-year survival rates for most of the selected cancer sites and all cancer sites combined have improved notably over time (Table 11). This is true for both Whites and African Americans. Cancer sites without significant improvement in survival rates in the past 25 years include the uterine corpus, cervix, larynx, liver, lung, pancreas, stomach, and esophagus. For nearly all cancer sites, survival rates for African Americans remain substantially lower than survival rates for Whites.
The relative survival rate cannot be calculated for other racial and ethnic groups because accurate life expectancy data for these population subgroups are not available. However, based on cause-specific survival rates of persons with cancer from 1988 to 1997 in the SEER areas, most minority populations have an elevated probability of dying of cancer within five years of diagnosis, compared with non-Hispanic Whites after accounting for differences in age and stage at diagnosis.20 For most cancers, minority populations are also less likely to be diagnosed at a localized stage and more likely to be diagnosed at distant stage, compared with White populations. Based on all solid tumors diagnosed with stage information from 1996 to 2000 in the SEER areas, 30% of the cases among American Indians/Alaskan Natives, 26% in African Americans and Hispanic-Latinos, and 25% among Asian/Pacific Islanders were diagnosed at distant stage, compared with 22% in Whites.4,5
Cancer in Children
Cancer is the second leading cause of death among children between the ages of 1 and 14 years in the United States; accidents are the most frequent cause of death in this age group (Table 12). The most commonly occurring cancers found in children (0 to 14 years) are leukemia (particularly acute lymphocytic leukemia), tumors of the central and sympathetic nervous systems, lymphoma, soft tissue sarcomas, and renal tumors.3 In the past 25 years, the five-year relative survival rate has improved significantly for many childhood cancers, including non-Hodgkin lymphoma, acute lymphocytic leukemia, acute myeloid leukemia, and Wilms' tumor (Table 13). The five-year relative survival rate among children for all cancer sites combined improved from 56% for patients diagnosed in 1974 to 1976 to 78% for those diagnosed in 1992 to 1999.3
STUDY LIMITATIONS AND FUTURE CHALLENGES
Estimates of the expected numbers of new cancer cases and cancer deaths should be interpreted with caution. These estimates may vary considerably from year to year, particularly for less common cancers and in states with smaller populations. Unanticipated changes may occur that are not captured by our modeling effort. The estimates of new cancer cases are based on incidence rates for the geographic locations that participate in the SEER program and, therefore, may not be representative of the entire United States. For these reasons, we discourage the use of estimates to track year-to-year changes in cancer occurrence and death. The recorded number of cancer deaths and cancer mortality rates from the National Center for Health Statistics and cancer incidence rates from SEER are generally the preferred data sources for tracking cancer trends, even though these data are three and four years old, respectively, when the estimates are calculated.
Despite these limitations, the American Cancer Society estimates do provide evidence of current patterns of cancer incidence and death in the United States. Such estimates will assist in continuing efforts to reduce the public health burden of cancer.