Fax: (770) 488-4639.
Article first published online: 3 NOV 2008
Published 2008 by the American Cancer Society
Supplement: Assessing the Burden of HPV-Associated Cancers in the United States
Volume 113, Issue Supplement 10, pages 2855–2864, 15 November 2008
How to Cite
Watson, M., Saraiya, M., Benard, V., Coughlin, S. S., Flowers, L., Cokkinides, V., Schwenn, M., Huang, Y. and Giuliano, A. (2008), Burden of cervical cancer in the United States, 1998–2003. Cancer, 113: 2855–2864. doi: 10.1002/cncr.23756
The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention.
This article is a US Government work and, as such, is in the public domain in the United States of America.
- Issue published online: 3 NOV 2008
- Article first published online: 3 NOV 2008
- Manuscript Accepted: 7 MAY 2008
- Manuscript Revised: 6 MAY 2008
- Manuscript Received: 14 APR 2008
- Cooperative Agreement. Grant Number: U50 DP424071-04
- Centers for Disease Control and Prevention
- cervical cancer;
- human papillomavirus;
- human papillomavirus vaccine;
Recent interest in human papillomavirus (HPV)-associated cancers and the availability of several years of data covering 83% of the US population prompted this descriptive assessment of cervical cancer incidence and mortality in the US during the years 1998 through 2003. This article provides a baseline for monitoring the impact of the HPV vaccine on the burden of cervical cancer over time.
Data from 2 federal cancer surveillance programs, the Centers for Disease Control and Prevention (CDC)'s National Program of Cancer Registries and the National Cancer Institiute's Surveillance, Epidemiology, and End Results Program, were used to examine cervical cancer incidence by race, Hispanic ethnicity, histology, stage, and US census region. Data from the CDC's National Center for Health Statistics were used to examine cervical cancer mortality by race, Hispanic ethnicity, and US census region.
The incidence rate of invasive cervical cancer was 8.9 per 100,000 women during 1998 through 2003. Greater than 70% of all cervical carcinomas were squamous cell type, and nearly 20% were adenocarcinomas. Cervical carcinoma incidence rates were increased for black women compared with white women and for Hispanic women compared with non-Hispanic women. Hispanic women had increased rates of adenocarcinomas compared with non-Hispanic women. The South had increased incidence and mortality rates compared with the Northeast.
Disparities by race/ethnicity and region persist in the burden of cervical cancer in the US. Comprehensive screening and vaccination programs, as well as improved surveillance, will be essential if this burden is to be reduced in the future. Cancer 2008;113(10 suppl):2855–64. Published 2008 by the American Cancer Society.
Cervical cancer is the second most common cancer among women in developing countries and the seventh most common cancer among women in developed countries.1 In the US in 2004, there were approximately 12,000 incident cases of invasive cervical cancer and 3850 deaths because of the disease.2 Since 1973, when federally funded, population-based cancer registration began in the US, cervical cancer incidence and mortality rates have declined.3, 4 It is believed that this decline is attributable in large part to the wide use and acceptance of cervical cytology screening, and the decline is most notable in squamous cell carcinomas (SCCs), which are detected more readily with the Papanicolaou (Pap) test than adenocarcinomas (ACs).4
Despite decreased incidence and mortality rates of cervical cancer, racial/ethnic and geographic disparities persist. Black and Hispanic women have higher rates of cervical cancer incidence and mortality than other groups; in the US, it is the sixth most common cancer among black and Hispanic women and 13th most common among white women.2, 5‒7 Incidence and mortality rates of cervical cancer also vary geographically, with higher rates reported in the South, Appalachia, and areas bordering Mexico.5, 8
It has been established that human papillomavirus (HPV) is a necessary cause of cervical cancer.9 The International Agency for Research on Cancer has identified HPV types 16 (HPV-16), HPV-18, HPV-31, HPV-33, HPV-35, HPV-39, HPV-45, HPV-51, HPV-52, HPV-56, HPV-58, HPV-59, and HPV-66 as oncogenic types in the cervix.10 Of these types, HPV-16 and HPV-18 account for approximately 70% of cervical cancers.11 High-risk HPV DNA testing has been recommended as an adjunct to Pap test screening and as a management tool for certain types of Pap test results by some medical organizations.12, 13 A quadrivalent HPV vaccine protecting against HPV-16, HPV-18, HPV-6, and HPV-11 currently is licensed and available in the US, and a bivalent vaccine protecting against HPV-16 and HPV-18 is currently under US Food and Drug Administration review. Among women without prior evidence of infection with HPV-16 and HPV-18, these vaccines have high efficacy in preventing high-grade cervical intraepithelial neoplasia (CIN-2/CIN-3) resulting from infection with these HPV types.14, 15
This report describes the incidence, stage, histology, and mortality of cervical carcinoma in the US during the years 1998 through 2003, confirming and adding to findings from our previous study. The current article incorporates an additional year of data, uses comprehensively described histology groupings agreed upon by a scientific working group, adds information on mortality, and places results in the context of this supplement to Cancer describing the burden of HPV-associated cancers in the US.5
MATERIALS AND METHODS
Data from 2 federal cancer surveillance programs, the Centers for Disease Control and Prevention (CDC)'s National Program of Cancer Registries (NPCR) and the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) Program, were used to examine cervical carcinoma incidence. NPCR and SEER data from 39 registries, which cover 83% of the US population for the years 1998 through 2003, met publication criteria for the United States Cancer Statistics report for all years from 1998 through 2003.16 Mortality data are from the CDC's National Center for Health Statistics; all US deaths during 1998 through 2003 were included.16 More information on the data sources in this article can be found in the report by Watson et al in this supplement 2008.17
We examined invasive cervical cancer cases (International Classification of Diseases, 3rd edition site codes C53.0, C53.1, C53.8, and C53.9) diagnosed from 1998 through 2003. We did not examine in situ cervical tumors, which have not been reportable to NPCR and SEER since 1996. All analyses were limited to cervical carcinomas by using specific histologies defined in Table 1, which made up 98.4% of all cervical cancers in the dataset.17 Cancer stage was presented as SEER Summary Stage 1977 for cases diagnosed before 2001 or as SEER Summary Stage 2000 for cases diagnosed in 2001 or later. Changes that affected the coding of stage for cervical cancer were minor.18
|Cervical Carcinoma Histology Definitions||ICD-O-3 Codes|
|Squamous cell and transitional cell carcinomas||8050-8084, 8120-8131|
|Adenocarcinomas||8140-8149, 8160-8162, 8190-8221, 8260-8337, 8350-8551, 8570-8576, 8940-8941|
|Adenosquamous and glassy cell||8560, 8015|
|Small cell/neuroendocrine||8013, 8041-8045, 8240-8246|
|Other specified carcinomas||8014, 8030-8040, 8046, 8090-8110, 8150-8157, 8170-8180, 8230-8239, 8247-8255, 8340-8347, 8561-8562, 8580-8671|
|Unspecified carcinomas||8010-8012, 8020-8022|
Rates were calculated in SEER*Stat, standardized to the 2000 US Standard Population, and expressed per 100,000 women; limits of confidence intervals were 95% and were based on the Gamma method using the modification described by Tiwari et al.19 Age-adjusted rates of invasive cervical carcinoma incidence are reported for race, Hispanic ethnicity, histology, stage, and US census region. Racial groups examined include white, black, and Asian-Pacific Islander (API); ‘all races’ also includes other and unknown categories. Hispanic ethnicity includes women from all race categories identified as Hispanic in the medical record or by use of an algorithm.20 Whites were used as the referent group for race comparisons; non-Hispanics were used as the referent group for Hispanic ethnicity. Reported differences in rates were based on statistically significant differences in rate ratios with a P value <.05.
There were 65,074 cases of invasive cervical carcinoma in our dataset during the years 1998 through 2003, with an incidence rate of 8.9 per 100,000 women (Table 2). Cervical carcinoma was rare in women aged <30 years, with an average of approximately 680 cases per year. Women aged <50 years had >50% of the burden of cervical carcinoma each year, with an average of 5933 cases annually. Equally important, a significant proportion of the burden was among women aged >50 years (an average of 4912 cases per year). Black women had increased rates of cervical carcinoma compared with white women and were diagnosed at an older median age (49 years vs 47 years for white women). Hispanic women had increased rates of cervical carcinoma compared with non-Hispanic women and were diagnosed at a younger median age (45 years vs 48 years for non-Hispanic women). SCCs made up 72.1% of cervical carcinomas; 19% were ACs. Adenosquamous carcinomas and small cell/neuroendocrine carcinomas (SCNEs) made up only 3.8% and 1.1% of cervical carcinomas, respectively. Just over half of all cervical carcinomas (52.3%) were diagnosed at the local stage, with an additional 31.5% diagnosed at the regional stage. Only 9.1% of cervical carcinomas were diagnosed at the distant stage, and the remaining 7.1% were unstaged. Age at diagnosis increased with later stage at diagnosis. Age-specific rates of cervical carcinoma incidence by race/ethnicity are shown in Figure 1.21 API women had the lowest rates of cervical carcinoma at age <45 years, after which white women had the lowest rates. Rates for white women peaked at ages 40 to 44 years and then leveled off, whereas rates for black, Hispanic, and API women peaked after age 65 years.
|Variable||Average Annual Incidence Count||Incidence Rate (95% CI)||Percent||Median Age, y|
|All ages||10,846||8.9 (8.8-9)||100||47|
|Squamous cell carcinoma||7824||6.4 (6.4-6.5)||72.1||47|
|Adenosquamous carcinoma||413||0.3 (0.3-0.4)§||3.8||45|
|Small cell-neuroendocrine||123||0.1 (0.1-0.1)§||1.1||47|
|Other carcinomas||421||0.3 (0.3-0.4)§||3.9||50|
Black women had increased rates of SCC and decreased rates of AC compared with white women (Table 3). The rates and percent distribution of cervical carcinoma histology among API women were similar to those among white women. Hispanic women had increased rates of SCC, AC, and adenosquamous carcinoma compared with non-Hispanic women. SCNE rates were increased in black and API women compared with white women. Cervical carcinoma histology by age at diagnosis is shown in Figure 2. SCC incidence rates rose sharply, peaked at ages 40 to 44 years, and showed a steady decline after approximately age 65 years. AC incidence rates rose more gradually, also peaked at ages 40 to 44 years, and remained relatively stable thereafter.
|SCC||AC||ASC||SCNE||Other Carcinomas||All Carcinomas|
Greater than half of all cervical carcinomas (52.9%) were diagnosed at the localized stage (Table 4). However, white women had a greater proportion of disease diagnosed at the localized stage (54.7%) than black and API women (44% and 48%, respectively). Black women had a greater percentage of unstaged cervical carcinomas (8.7%) than white women (6.4%). Although Hispanic women had higher overall rates of cervical carcinoma, the percentage distribution of carcinoma by stage was similar to that of non-Hispanic women. For each race/ethnic group, most cervical carcinomas in women aged <50 years were diagnosed at the localized stage (Fig. 3). After age 50 years, cervical carcinoma of all stages declined for white women. However, for black women aged >50 years, localized cervical carcinomas remained high, and carcinomas that were diagnosed at regional and distant stages also increased. Among API and Hispanic women aged >50 years, the rates of invasive cervical carcinoma diagnosed at the localized stage declined, whereas carcinomas diagnosed at regional and distant stages increased.
Overall, the South had a higher rate of cervical carcinoma than other census regions; this increase was statistically significant (using the Northeast as the referent group) (Table 5). Hispanic women had higher rates than non-Hispanic women in all regions. Black women had increased rates of invasive cervical carcinoma compared with white women in the Northeast, Midwest, and South. For white and black women, the South had the highest rates of cervical carcinoma, whereas the Midwest had the highest rates for API and Hispanic women.
There were 24,707 deaths from cervical cancer in our dataset during the years 1998 through 2003, with a death rate of 2.7 per 100,000 women (Table 6). Death rates from invasive cervical cancer rose with age for all races/ethnicities (Fig. 1). Disparities persisted in cervical cancer mortality as well as incidence. Black women had the highest mortality rates from cervical cancer of all races. Hispanic women also had increased cervical cancer mortality rates compared with non-Hispanic women. The South census region had increased rates of mortality because of cervical cancer using the Northeast as the referent group. The median age at death from cervical cancer did not vary greatly by race. Hispanic women had a median age of 51 years for cervical cancer mortality, which was lower than that for non-Hispanic women.
|Variable||Average Annual Mortality||Mortality Rate||Percent||Median Age, y|
The results of this study confirm recent reports indicating higher cervical cancer incidence rates among black and Hispanic women and women living in the South census region.5, 6, 22 Although the findings of previous analyses produced similar results, for the current report, we used standardized criteria and methodology that were established for this supplement of Cancer expressly for surveillance purposes—to characterize the prevaccine burden of all HPV-associated cancers, including cervical cancer, with the hopes of using similar criteria in subsequent years.
Racial/ethnic and geographic disparities in cervical cancer incidence and age at diagnosis may be caused by various patterns of screening and follow-up/treatment of preinvasive lesions. According to some national surveys, Pap testing rates among black women are similar to those of white women, whereas other surveys indicate that older black women report Pap testing less frequently than white women.23‒25 Thus, higher incidence rates that increase with age in black women may be attributable to less screening or lack of follow-up of abnormal test results.26 To our knowledge, there are no national surveys that ask about follow-up of abnormal Pap tests to assess potential disparities, although special studies indicate mixed results.26, 27 For Hispanic women, differences may be attributable to decreased likelihood of screening and lack of follow-up.23, 28 Recent analysis of screening data controlling for age, race, and socioeconomic status (SES) indicate that approximately 10% of foreign-born women from Mexico, Central America, and South America report having never been screened.29 Racial/ethnic and geographic differences in cervical cancer incidence rates also may reflect differences in access to healthcare, such as the distance a patient has to travel to a healthcare facility, language barriers, the availability of culturally appropriate and sensitive healthcare, and health insurance coverage.30 Increased rates of cervical carcinoma incidence and mortality in the South, even among white women, may be attributable to differences in demographics. Rates of poverty are higher in the South, and women without a usual source of healthcare, women without health insurance, and women of lower SES are less likely to have received Pap testing at the recommended intervals.23, 31‒33 We were not able to examine rates of cervical cancer among American Indian and Alaska Native (AI/AN) women as a separate race group in this analysis.17 However, recent research indicates that AI/AN women have higher rates of cervical cancer than non-Hispanic white women.34
HPV infection is extremely common among sexually active women and usually resolves with no treatment; persistence of the virus and progression to invasive cervical cancer is rare and typically occurs over years or decades.35, 36 However, cofactors that vary by race/ethnicity, such as tobacco use and high parity, have been implicated that may facilitate the progression from HPV infection to cervical cancer.37 In addition, low SES varies by race/ethnicity and is associated with increased rates of cervical cancer, most likely because of reduced access to care.38, 39 Thus, variations in cofactors by race/ethnicity may be responsible in part for variations in incidence rates of cervical carcinoma,40 although this information is not available in the cancer registry system.
We confirmed previous findings that cervical carcinoma histology varied by race/ethnicity. Higher rates of invasive SCC among black and Hispanic women may be caused by lower rates of Pap testing or inadequate follow-up.24, 26, 41 Higher rates of AC among Hispanic women may be caused by younger age at first intercourse, higher parity, higher body mass index, or other, unexplored cofactors.42, 43 Although adenosquamous carcinoma and SCNE cervical carcinomas are rare, disparities by race/ethnicity are a cause for concern because of their poor prognoses and may indicate clues regarding etiology.44, 45
We confirmed that black women had higher proportions of invasive cervical carcinoma diagnosed at later stages than other women, as in previous studies, suggesting that cervical cancer is not being detected and treated as early in this group as in other racial/ethnic groups.5, 6 Also, the finding that black and Hispanic women had higher rates of unstaged cervical cancers may confirm previous reports, which also indicated that black and Hispanic women were more likely to receive less aggressive treatment or no treatment compared with white women.46, 47 Cancers may be clinically unstaged for many reasons: death of the patient before staging, refusal of diagnostic or treatment procedures, or less aggressive workup because of advanced age or comorbid conditions.48 Higher mortality rates among black women, especially in the South, may be attributable to higher incidence rates, later stage at diagnosis, and/or disparities in treatment.26, 49
Cancer registry data have some limitations, which had an impact on our study. Although population-based cancer registries provide an excellent system with which to measure invasive cancers, they do not capture information on HPV types related to cancer or on individual risk factors, such as smoking. Reporting of race and Hispanic ethnicity uses data from medical records, which may or may not be accurate. However, application of the North American Association of Central Cancer Registries' Hispanic Identification Algorithm to incidence data increased the identification of individuals likely to be Hispanic. In addition, only 63% of the population in the South census region was included in our dataset. However, an analysis of cervical cancer rates that included more states from this region but was limited to more recent years demonstrated that the rates in this region remained higher than the rates in other regions.50 Thus, case counts for the South very likely were underestimates, but rates generally are representative of the region as a whole. Finally, bias may exist in the collection of information on cancer stage and may affect rates, particularly for unstaged cancers; accuracy of cancer registry data on stage at diagnosis is relatively low compared with other demographic variables.51
Another limitation may be that population denominators include all women, not just those with an intact cervix and therefore at risk for cervical carcinoma. In 2006, Merrill estimated cervical cancer rates in SEER by using data from Behavioral Risk Factor Surveillance System and the Hispanic/Latino Identification Algorithm and excluding women who underwent hysterectomies from the denominator; in that report, the rates of cervical cancer and racial disparities increased substantially.52 More research is needed to determine the applicability of that method or other methods to determine the population at risk for developing cervical cancer.
This study, which assesses the overall burden of cervical cancer in a majority of the US population, sets the stage for monitoring the impact of the HPV vaccine and other emerging technologies to detect potential changes in the burden of cervical cancer over time. Although this report covers a large proportion of the US population, improved collection of cancer registry data will be essential to continued monitoring of the burden of invasive cervical cancer in all areas of the US. Histology categories in this article are described comprehensively and are based on current knowledge; thus, these findings could be of interest in future etiological studies and could provide important baseline data for assessing the eventual impact of HPV vaccine. In addition, by focusing on cervical cancer rates and burden in younger women, we may be able to see the impact of the vaccine in this age group. Our addition of mortality data reminds us of the importance of screening and prevention programs, because an average of >4000 women per year died from a largely preventable disease during the years of our study.
The HPV vaccine and HPV DNA testing have the potential to reduce disparities in the incidence and mortality of cervical cancer in the US. However, rates of cervical cancer currently are highest among those who can least afford prevention. Current disparities in access to care do not bode well for future uptake of the vaccine among lower income girls and women. Thus, vaccination programs as well as increased screening and follow-up of low SES, underinsured and uninsured women, and those in rural and medically underserved areas will be important in the reduction of the burden of cervical cancer in the US. Surveillance of cervical cancer and precancerous lesions is essential to ensuring that vaccination and screening programs are effectively reducing the burden of cancer in the US.
- 2US Cancer Statistics Working Group. United States Cancer Statistics: 2004 Incidence and Mortality. Atlanta, Ga: Department of Health and Human Services, Centers for Disease Control and Prevention and National Cancer Institute; 2007.
- 10IARC Monographs on the Evaluation of Carcinogenic Risks to Humans: Human Papillomaviruses . Lyon, France: World Health Organization, International Agency for Research on Cancer; 2007..
- 14Efficacy of a prophylactic adjuvanted bivalent L1 virus-like-particle vaccine against infection with human papillomavirus types 16 and 18 in young women: an interim analysis of a phase 3 double-blind, randomised controlled trial. Lancet. 2007; 369: 2161–2170., , , et al.
- 16US Cancer Statistics Working Group. United States Cancer Statistics: 2003 Incidence and Mortality. Atlanta, Ga: Centers for Disease Control and Prevention; 2006. Available at: http://www.cdc.gov/cancer/npcr/uscs/. Accessed October 23, 2007.
- 18members of the NAACCR Collaborative Research Work Group. Site-specific Comparison of Summary Stage 1977 and Summary Stage 2000 Coding. Springfield, Ill: North American Association of Central Cancer Registries; 2005. Available at: http://www.naaccr.org/index.asp?Col_SectionKey=11&Col_ContentID=397. Accessed on February 11, 2007., , , , , and
- 20North American Association of Central Cancer Registries (NAACCR) Latino Research Work Group. NAACCR Guideline for Enhancing Hispanic-Latino Identification: Revised NAACCR Hispanic/Latino Identification Algorithm [NHIA v2]. Springfield, Ill: North American Association of Central Cancer Registries; 2005.
- 21Population Projections of the United States by Age, Sex, Race, and Hispanic Origin: 1995 to 2050, U.S. Bureau of the Census, Current Population Reports, P25–1130. Washington, DC: US Government Printing Office; 1996..
- 31The topography of poverty in the United States: a spatial analysis using county-level data from the Community Health Status Indicators project. Prev Chronic Dis [serial online]. 2007; 4: A111..
- 48YoungJL,RoffersSD,RiesLAG,FritzAG,HurlbutAA, eds. SEER Summary Staging Manual-2000: Codes and Coding Instructions. NIH Publication No. 01–4969. Bethesda, Md: National Cancer Institute; 2001.
- 50US Cancer Statistics Working Group. United States Cancer Statistics: 1999‒2004 incidence and mortality web-based report. Bethesda, Md: Department of Health and Human Services, Centers for Disease Control and Prevention and National Cancer Institute; 2007. Available at: www.cdc.gov/uscs. Accessed on March 25, 2008.
- 53Summary Stage: Data Effects of the Changes in 2000. Springfield, Ill: North American Association of Central Cancer Registries; 2003..