|Home | About | Journals | Submit | Contact Us | Français|
Colorectal cancer (CRC) is the most frequently diagnosed cancer among Alaska Native (AN) people, and the second leading cause of cancer death. The incidence rate for the combined years 1999 through 2003 was 30% higher than the rate among U.S. whites (USWs) for the same period. Current incidence rates may serve to monitor the impact of screening programs in reducing CRC in the AN population.
Incidence data are from the Alaska Native Tumor Registry and the National Cancer Institute Surveillance, Epidemiology, and End Results (SEER) Program. We compared AN CRC incidence, survival rates, and stage at diagnosis with rates in USWs for cases diagnosed from 2005 through 2009. Relative survival calculations were produced in SEER*Stat by the actuarial method.
The CRC age-adjusted incidence rate among AN men and women combined was higher than those in USW men and women (84 vs. 43/100,000; P < .05; AN:USW rate ratio [RR] = 2.0). The greatest differences between rates in AN people and USWs were for tumors in the hepatic flexure (RR = 3.1) and in the transverse (RR = 2.9) and sigmoid (RR = 2.5) regions of the colon. Rectal cancer rates among AN people were significantly higher than rates in USWs (21 vs.12/100,000). Five-year relative survival proportions by stage at diagnosis indicate that the CRC 5-year relative survival was similar in AN people and USWs for the period 2004 through 2009.
The high rate of CRC in AN people emphasizes the need for screening programs and interventions to reduce known modifiable risks. Research in methods to promote healthy behaviors among AN people is greatly needed.
Since the early 1970s, colorectal cancer (CRC) has been the most frequently diagnosed cancer among Alaska Native people (AN) and the second leading cause of cancer death in that population. Among other U.S. populations, high rates of CRC are reported among blacks (USBs), followed by whites (USWs), with comparatively low rates among the Asian/Pacific Islander population.1–3 Reports of American Indian and Alaska Native populations (AI/AN), typically combined as a single group, have CRC incidence rates lower than that of USBs and USWs.4,5 The Alaska Native Tumor Registry (ANTR) has reported higher rates of CRC among AN people compared with that in major U.S. populations for more than 30 years.6
From 1969 to 2003, CRC incidence increased 25% among AN people, whereas among USWs, incidence has declined since 1986.3 Current incidence rates of CRC may serve to monitor the impact of expanded CRC screening and cancer education programs in effect in recent years. A description of CRC among AN people and current incidence rates and trends are presented in this report.
The AN population of 130,000 comprises many tribes and ethnic groups across Alaska that may be broadly grouped as Inupiat and Yupik Eskimo (northern and western coasts); Athabascan (interior region), Tlingit, Haida, and Tsimshian Indian (southeast); and Aleut (Aleutian Island chain and Alaska peninsula) populations. Approximately 23% of AN people live in the state's largest urban center, Anchorage.
Cancer information is from the National Cancer Institute Surveillance, Epidemiology and End Results (NCI SEER) Program SEER*Stat and the ANTR.7,8 The ANTR has participated in the National Cancer Institute (SEER) Program since 1999.9 Tumor data and diagnostic, treatment, and demographic information are from electronic medical records, medical charts, pathology reports, and provider dictations. Abstracting of colon and rectal cancer cases followed strict coding and staging guidelines established by the SEER program, and data are processed through a standard set of computerized edits.
The morphology and topography classifications for colon and rectal cancer are based on the International Classification of Diseases for Oncology (ICD-O), second- and third-edition coding.10 Rectosigmoid junction site cancers are grouped with cancers of the rectum. Proximal colon sites are defined as cancers of the cecum, ascending colon, hepatic flexure, and transverse and splenic flexure (ICD-O topography codes C180 and C182–C185) and exclude the appendix, while distal colon sites include cancers of the descending and sigmoid colon regions (ICD-O C186, C187). Lymphoma, carcinoids, and Kaposi's sarcoma of the colon or rectum were excluded from the analyses.
Cancer stage at diagnosis is described using the SEER summary stage 2000 coding system of in situ, local, regional, and distant, derived from the Collaborative Staging System begun in 2004.11,12 Population data are from National Center for Health Statistics for AN people (bridged series) and USWs, available on the NCI SEER*Stat website.13 All incidence rates include invasive cancers and are age-adjusted to the standard 2000 U.S. population using the direct method and are expressed as per 100,000 population.14 Statistical significance was noted for overlapping 95% confidence intervals of AN and USW age-adjusted rates.
Rates are based on 5-year averages, except for survival calculations where a 6-year period was used to include adequate case numbers for analysis. Five-year relative survival rates were calculated in SEER*Stat by the actuarial method using AN and USW population–based life tables for each population. Rate ratios are AN rates divided by USW rates.
During the 5-year period 2005 through 2009, 301 cases of invasive and 36 cases of in situ CRC were diagnosed among AN people (Table 1). Ninety-eight percent of CRC cases had histologically confirmed pathology; less than 1% of cases were based on death certificate data only. The CRC incidence rate in AN people was two times the rate in USWs (84 vs. 43/100,000; P < .05; AN:USW rate ratio [RR] = 2.0). Age-specific CRC rates for 10-year age groups were significantly higher among AN people than among USWs for every age group except the oldest (80+ years; Figure 1).
Rates of colon cancer were similar between AN men and women, but exceeded the rates in USW men and women. The AN:USW disparity in incidence rates is statistically significant (P < .05) across the major colon topographic subsites, except for tumors arising in the ascending and descending colon. The greatest differences between rates in AN people and USWs were for tumors in the hepatic flexure (5 vs. 2/100,000, AN:USW) and the transverse (9 vs. 3/100,000) and sigmoid (22 vs. 9/100,000) regions of the colon.
The proportions of cancers arising in the proximal (41% of all CRC), distal (29%), and rectal (28%) regions of the large bowel in the AN population were similar to those in the USW population. However, AN men and women showed higher incidence rates for both the proximal and distal regions of the colon than did USW men and women.
Rates of rectal cancer were significantly higher in AN men (31 vs. 13; P < .05) than in AN women. The rectal cancer incidence rate in AN men, but not in AN women, was significantly higher than the rate in USWs (31 vs.15/100,000 in men, respectively; RR = 2.0).
During the diagnosis years 2005 through 2009, 90% of CRCs among AN people were identified as adenocarcinomas (ACA), and 6% were cystic, mucinous, or mucin-producing ACA. USWs had a similar proportion of ACA (85%) and of cystic, mucinous, and mucin-producing ACA (9%) for the same period. Approximately 9% of the ACA identified in AN patients and 8% in USW patients were found in polyps. Seventy-three percent of CRCs diagnosed in AN people and 59% of CRCs in USWs were described by pathologists as of moderately differentiated grade (grade 2).
From 2005 through 2009, the incidence rates for patients in whom CRC was at the in situ stage at diagnosis were found to be more than 4 times higher among AN people (8.7/100,000) than among USWs (1.9/100,000). Among the invasive cancers, the incidence rate was highest for regional stage disease among AN people (36/100,000), more than two times the USW rate (15/100,000). Similarly, AN rates for local (28/100,000) and distant (18/100,000) stage disease were twice USW rates (17 and 8/100,000 for local and distant disease, respectively). Figure 2 shows a comparison of the percentage of CRCs by stage at diagnosis between AN people and USWs. A greater proportion of CRCs was diagnosed as in situ in AN people than in USWs (11% vs. 4%) and fewer cancers in AN people in the localized stage (30% vs. 38%). The percentages of CRCs diagnosed at regional and distant stages were more similar between AN people and USWs.
Five-year relative survival proportions by stage of disease at diagnosis from 2004 through 2009 are shown in Figure 3. At each stage of diagnosis, the 5-year relative survival of CRC was similar between the AN and USW populations.
A comparison of AN and USW CRC rates for a 40-year period is shown in Figure 4. A statistically significant downward trend in USW rates begins in the 1980s, a time when AN rates were rising. AN rates appear to have decreased since the 1990s, but a test for a downward trend did not achieve statistical significance.
During the past 40 years, CRC incidence has increased dramatically in AN people and continues to account for the greatest cancer burden among AN people. The two-fold disparity between AN people and USWs in CRC incidence and mortality is increasing, primarily due to declining incidence rates among USWs. We were unable to detect statistical significance in rate changes over the recent 10-year period, although rates appear to have declined since the late 1990s. Survival rates for CRC have improved in the United States since 1960, more so among USWs than USBs.15 The lower survival rate among USBs has been attributed in part to later stage at diagnosis.15–17 Elevated CRC incidence in AN people does not seem to be due to a later stage at diagnosis; AN people have similar proportions of CRC diagnosed at the distant stage, as do USWs.
Chronic disease risk factor data for AN people collected in a prospective study, (Education and Research Towards Health [EARTH]) have provided some baseline information on potential levels of risk for cancer among this population. The study recruited over 3800 participants from three regions in Alaska and collected medical measurements and information on diet, lifestyle, and behavioral health, and chronic diseases. Dietary shifts away from traditional foods to a more general U.S. diet were noted. The EARTH study reported that fewer than 25% of AN participants acquired all or almost all of their food from subsistence means.18 Traditional AN diet includes wild game, marine mammals, fish, berries, and wild greens; is high in protein and fats, including omega-3 fatty acids; and lacks preservatives.18,19 Some dietary studies in other populations have indicated that a diet high in fruit and vegetable consumption may lower the risk of CRC, but results have been mixed.20–22 If such a diet is indeed beneficial, the seasonal availability of these foods in Alaska may reduce the overall protective effect. Farm-raised red meats and processed foods, associated with an increased risk of CRC, have become more available to AN people living in rural areas.23,24 Information on fruit and vegetable intake of AN people is collected in the Alaska Behavioral Risk Factor Surveillance Survey (BRFSS), a telephone survey conducted annually since 1991. In the 2009 BRFSS, 14% of AN people reported eating five or more servings of fruits and vegetables per day, compared to 25% reported by Alaska's non-native population.25
Other potential risk factors include tobacco use, diabetes, and obesity. CRC incidence has been shown to be higher among smokers.26 The BRFSS has consistently reported a relatively high proportions of smokers among AN people (2009 survey: 39% among AN people, compared to the Alaska non-native population current smoking rate of 19%).25
Type 2 diabetes has been associated with increased risk of CRC. Among AN people, the disease, once well below the rate in USWs, is increasing, particularly in the 15- to 34-year age group.27 Incidence rates of diabetes doubled from 16 per 10,000 population from 1986 through 1990 to 33 per 10,000 in the more recent period 2002 through 2006. The 2009 BRFSS data for AN people indicated that 6% of AN people surveyed reported diabetes (Alaska non-native: 6%).25
Being overweight has also been associated with an increased risk of CRC.28 The 2009 BRFSS data show that among AN people, 41% reported being overweight, with a BMI between 25 and 29.9 (Alaska non-native, 38%), and 32% met the definition of obese, with a BMI of 30 or more based on reported height and weight (Alaska non-native, 24%).25 Forty-six percent did not meet the recommendations for moderate physical activity. Diabetes coupled with the increases in body weight, shifts towards a higher fat diet and reduced levels of physical activity, as reported through the BRFSSs, are likely contributing risk factors to CRC and other chronic diseases in AN people.
Hereditary risk factors, such as familial adenomatous polyposis (FAP) and hereditary nonpolyposis colorectal cancer (HNPCC), also known as Lynch syndrome, may account for 5% to 10% of CRCs in the general U.S. population.4 HNPCC, a result of germ-line mutations in DNA-mismatched repair genes, does not appear to be unusually high in AN people. A review of archived CRC cancer tissue specimens indicated levels of expression of DNA-mismatched repair genes in AN people similar to those found in a study of other U.S. populations.29 Further studies of risk factors, including possible genetic factors are needed to explore possible causes of high CRC incidence rates in AN people.
Smoking cessation and CRC screening programs to address modifiable risk factors for CRC are current approaches to reducing CRC incidence and mortality. CRC control is a national priority for the National Comprehensive Cancer Control Program (NCCCP), and the resulting funding provided by the Centers for Disease Control (CDC) to states, tribes, and tribal organizations has allowed for enhanced interventions since 1998.30 Screening for CRC has been identified as a priority among tribal healthcare providers throughout Alaska. Since 1998, enhanced funding has allowed for CRC screening. The greater proportion of in situ cancers seen in AN CRC diagnoses compared with USW proportions may be due to an increase in screening in several areas of the state as well as efforts to target first-degree relatives of individuals diagnosed with CRC.
Alaska Tribal Health System efforts have focused on addressing barriers to early detection and increasing screening opportunities for AN people, particularly in rural areas. One approach has been to provide additional itinerant colonoscopy screening clinics in conjunction with patient navigators at regional healthcare facilities not providing screening services and training for midlevel providers to perform flexible sigmoidoscopies at regional health facilities.31,32 In addition, more than 300 Community Health Aides and Practitioners working in rural Alaska communities received extensive cancer education training. A variety of culturally sensitive CRC educational materials have been made available to AN people living in communities throughout Alaska.33 BRFSS data from 2002 to 2010 indicate CRC screening (persons responding that they had a sigmoidoscopy or colonoscopy) was reported by approximately half of the individuals aged 50 years and older who responded to the surveys. The 2011 survey indicated that this proportion had increased to 65% among AN people.34 A study of CRC screening rates in American Indian/AN people who used the Indian Health Service for healthcare from 1996 through 2004 found the highest screening rates of 5 U.S. geographic areas to be among AN people.35 Recently, three Alaska regional tribal groups were awarded a CDC colorectal cancer control program (CRCCP) grant to enhance efforts to increase CRC screening through existing Breast and Cervical Cancer Program offices.36
Elevated CRC incidence rates among persons under the age of 50 are of particular interest to providers, since screening is not recommended until age 50 for those of average risk. Age-specific incidence rates for AN people are double those of USW for ages 30 to 39 (11.3 vs. 5.6/100,000) and 40 to 49 (45.1 vs. 19.8/100,000). In an effort to identify persons at increased risk, an outreach program has been developed to identify family members of CRC patients who could benefit from early or more frequent screening.
A high prevalence of Helicobacter pylori in the AN population, often resulting in chronic gastritis, limits the use of guaiac tests to detect blood in the stool.37 In an ongoing study to determine the feasibility of using a human hemoglobin-specific stool blood assay, the fecal immunochemical test (FIT), the specificity and sensitivity of the test are compared that of the guaiac fecal occult blood test (FOBT) among study participants who undergo colonoscopy.38 The FIT test, if shown to have high test sensitivity and specificity for CRC, could greatly benefit AN people living in rural areas through early detection.
Although the rate of colon cancer in AN men may have declined since the mid-1990s, the high rate of CRC in AN people emphasizes the need for screening programs and interventions to reduce known modifiable risks. Certain behaviors that contribute to increased risk of CRC are being addressed through smoking cessation and diabetes programs; however, changing behaviors to reduce risk from excess weight, excess alcohol, physical inactivity, and diet remain the major challenges for public health intervention. Research in methods to promote healthy behaviors among AN people is greatly needed.
The authors thank the healthcare and medical records staff of the Alaska Native Medical Center, regional facilities throughout the state, and staff from central cancer registries of Alaska and Washington who contribute valuable data to the Alaska Native Tumor Registry. Funding for the Alaska Native Tumor Registry and analysis was provided in part by the National Institutes of Health, National Cancer Institute, Surveillance, Epidemiology and End Results (SEER) Program under an Inter-agency Agreement with the Indian Health Service (NCI Y1-PC-0064-01.
This research was funded in part through the National Cancer Institute, Surveillance, Epidemiology and End Results (SEER) Program, Interagency Agreement No. Y1-PC-0064-01.
Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.