http://onlinelibrary.wiley.com/doi/10.3322/caac.21166/full

Introduction
[1] Cancer is a major public health problem in the United States and many other parts of the world. One in 4 deaths in the United States is due to cancer. In this article, we provide the expected numbers of new cancer cases and deaths in 2013 nationally and by state, as well as an overview of current cancer statistics using data through 2009, including incidence, mortality, and survival rates and trends. We also estimate the total number of deaths averted as a result of the decline in cancer death rates since the early 1990s, and provide the actual reported numbers of deaths in 2009 by age for the 10 leading causes of death and the 5 leading cancer types.

Expected Numbers of New Cancer Cases

Incidence and Mortality Data
Mortality data from 1930 to 2009 in the United States were obtained from the National Center for Health Statistics (NCHS).1, 2 There are several sources for cancer incidence data. The Surveillance, Epidemiology, and End Results (SEER) program of the National Cancer Institute reports long-term (beginning in 1973), high-quality, population-based incidence data covering up to 26% of the US population. Cancer incidence rates for long-term trends (1975-2009), 5-year relative and cause-specific survival rates (2002-2008), and estimations of the lifetime probability of developing cancer (2007-2009) were obtained from SEER registries.3-6 The North American Association of Central Cancer Registries (NAACCR) compiles and reports incidence data from 1995 onward for cancer registries that participate in the SEER program or the Centers for Disease Control and Prevention's National Program of Cancer Registries (NPCR). Incidence data for state-level rates (2005-2009), trends by race/ethnicity (2000-2009), and estimated new cancer cases in 2013 were obtained from NAACCR.7 Cancer cases were classified according to the International Classification of Diseases for Oncology.8 All incidence and death rates are age-standardized to the 2000 US standard population and expressed per 100,000 persons.
Cancer incidence rates in this report are delay-adjusted whenever possible in order to account for anticipated future corrections to registry data due to inherent delays and errors in case reporting. Delay-adjusted rates primarily affect the most recent years of data for cancers that are frequently diagnosed in outpatient settings (eg, melanoma, leukemia, and prostate) and provide a more accurate portrayal of the cancer burden in the most recent time period.9 For example, melanoma incidence rates adjusted for delays in reporting are 14% higher than unadjusted rates in the most recent reporting year. Delay-adjusted rates are available for SEER registry data and were obtained from the National Cancer Institute. Incidence trends presented for the most recent 5 years (2005-2009) are based on delay-adjusted rates from SEER 13 cancer registries.4

Projected Cancer Cases and Deaths in 2013
The precise number of cancer cases diagnosed each year in the nation and in every state is unknown because cancer registration is incomplete in some states. Furthermore, the most recent year for which incidence and mortality data are available lags 3 to 4 years behind the current year due to the time required for data collection, compilation, quality control, and dissemination. Therefore, we project the numbers of new cancer cases and deaths in the United States in 2013 in order to provide an estimate of the contemporary cancer burden. The methods for projecting both new cases and deaths in the current year were recently modified so estimates should not be compared from year to year.
We projected the number of new invasive cancer cases that will be diagnosed in 2013 (with the exception of urinary bladder, for which in situ cases are included) using a 2-step process that first estimates complete incidence counts by state during years for which observed data are available, and then projects these counts 4 years ahead for the United States overall and each state individually.10 To estimate counts for each state through 2009, we used a spatiotemporal model based on incidence data for 1995 through 2009 from 49 states and the District of Columbia that met NAACCR's high-quality data standard for incidence, covering about 98% of the US population.11 (Minnesota cancer registry data could not be included in the model because county-level data were unavailable.) This method accounts for expected delays in case reporting and considers geographic variations in sociodemographic and lifestyle factors, medical settings, and cancer screening behaviors as predictors of incidence. A temporal projection method (the vector autoregressive model) was then applied to the estimated counts to obtain projections for 2013. For the complete details of this methodology, please refer to Zhu et al.10
To estimate the numbers of new breast carcinoma in situ (female) and melanoma in situ cases in 2013, we first estimated the number of in situ cases occurring annually from 2000 through 2009 in the United States by applying the age-specific incidence rates in the 18 SEER areas to the corresponding US population estimates provided in SEER*Stat.12 We then projected the total number of cases in 2013 based on the annual percent change from 2000 through 2009 generated by the joinpoint regression model.13
We estimated the number of cancer deaths expected to occur in 2013 in the United States overall and in each state using the joinpoint regression model based on the actual numbers of cancer deaths from 1995 through 2009 at the state and national levels as reported to the NCHS.1 For the complete details of this methodology, please refer to Chen et al.14

Other Statistics
The estimated numbers of cancer deaths averted in men and women due to the reduction in overall cancer death rates were calculated by applying the 5-year age-specific cancer death rates in the peak year for age-standardized cancer death rates (1990 in men and 1991 in women) to the corresponding age-specific populations in the subsequent years through 2009 to obtain the number of expected deaths in each calendar year if the death rates had not decreased. We then summed the difference between the number of expected and observed deaths in each age group and calendar year for men and women separately.

Trends in Cancer Mortality
Cancer death rates decreased by 1.8% per year in males and by 1.5% per year in females during the most recent 5 years of data (2005-2009). These declines have been consistent since 2001 and 2002 in men and women, respectively, and are larger in magnitude than those occurring in the previous decade (Table 5). Death rates peaked in men in 1990 (279.8 per 100,000), in women in 1991 (175.3 per 100,000), and overall in 1991 (215.1 per 100,000). Between 1990/1991 and 2009, cancer death rates decreased 24% in men, 16% in women, and 20% overall. Figure 6 shows that as a result of almost two decades of consistent declines in cancer death rates, about 1,177,300 cancer deaths were averted, 152,900 of these in 2009 alone.
Death rates continue to decrease for the 4 major cancer sites (Figs. 4 and 5). Over the past two decades of data, death rates have decreased from their peak by more than 30% for cancers of the colorectum, female breast, and male lung, and by more than 40% for prostate cancer. The decrease in lung cancer death rates—among men since 1990 and among women since 2002—is due to the reduction in tobacco use,20 while the decrease in death rates for female breast, colorectal, and prostate cancers largely reflects improvements in early detection and/or treatment.17, 21, 22 Over the past 10 years of data (2000-2009), the largest annual declines in death rates were for chronic myeloid leukemia (8.4%), cancers of the stomach (3.1%) and colorectum (3.0%), and non-Hodgkin lymphoma (3.0%).

Cancer Survival
African Americans are less likely to survive cancer than whites. The 5-year relative survival is lower among African Americans for every stage of diagnosis for nearly every type of cancer (Fig. 7). These disparities may result from inequalities in access to and receipt of quality health care and/or from differences in comorbidities. As shown in Figure 8, African Americans are less likely than whites to be diagnosed with cancer at a localized stage, when treatment is usually less extensive and more successful. The extent to which factors other than stage at diagnosis contribute to the overall survival differential is unclear.28 A study of Medicare-insured patients showed that African Americans remain less likely than whites to receive standard cancer therapies for lung, breast, colorectal, and prostate cancers.29 Some studies suggest that African Americans who receive cancer treatment and medical care similar to that of whites experience similar outcomes.30

Conclusions
In 2009, Americans had a 20% lower risk of death from cancer than in 1991, when cancer death rates peaked. Despite this substantial progress, all demographic groups have not benefitted equally, particularly for cancers such as colorectal and breast, for which mortality declines have been attributed to earlier detection and improvements in treatment. Further progress can be accelerated by applying existing cancer control knowledge across all segments of the population, with an emphasis on those groups in the lowest socioeconomic bracket as well as other disadvantaged populations.