Ionizing radiation can induce DNA damage, thereby increasing risk for cancer development (Bolus, 2008; Wall et al., 2006; Williams, 2008). Higher risks are associated with younger age at exposure, and females have somewhat higher risks of cancer from radiation exposure than males do. Some occupations and geographic sites are associated with increased levels of exposure, but the role of ionizing radiation in cancer disparities has not been well studied. Several recent studies report greater occupational exposure among African American than Caucasian workers in the Savannah River Site nuclear power plant, with African Americans more likely to have detectable radiation exposure on their monitors (Angelon-Gaetz, Richardson, & Wing, 2010). Ionizing radiation is used in a variety of screening and diagnostic tests at doses designed to minimize patient risk, but the cumulative effects are not negligible, especially in individuals who have frequent and/or numerous tests at a young age (Wall et al., 2006). The radiation exposure that occurs with CAT scans, angiograms, and nuclear medicine studies is much higher than with mammograms or simple X-rays, and their long-term risk is not fully known, in part because cumulative exposure is not generally tracked in the United States, but exposures from imaging tests can be substantial (Goodman et al., 2008). Ionizing radiation is used at high doses to treat some forms of cancer. It increases the risk of secondary cancers, and thus its use is generally restricted to life-threatening diseases in which the benefit of radiation far outweighs the risks (Doi, Mieno, Shimada, & Yoshinaga, 2009; Li et al., 2010; Shuryak, Hahnfeldt, Hlatky, Sachs, & Brenner, 2009). Improvements in the methods used to deliver therapeutic radiation have reduced its carcinogenicity by more effectively focusing its delivery to the cancers being targeted, with less damage to normal cells. Ultraviolet (UV) radiation in sunlight can also cause DNA damage, thereby increasing the risk of several forms of skin cancer (McPhail, 1997). Skin pigment reduces the ability of UV radiation to penetrate skin and cause damage; thus, lightskinned individuals are more at risk for damage through sunburn and excessive tanning, and this can lead to premalignant conditions such as actinic keratosis, as well as skin cancers.
Therapies that damage DNA remain a primary modality for treating many types of cancer, with the hope that cancer cells, due to their higher growth rates, will be more affected than normal cells. These therapies increase the risk of secondary cancers, with some combinations being more carcinogenic than others (Gururangan, 2009). People with inherited defects in DNA repair are at increased risk for secondary cancers induced by chemotherapy and/or radiation. The treatment for many forms of cancer includes the use of radiation and/or cytotoxic chemotherapy, recognizing that the greatest chances of cure are dependent on eradication of all cancer cells. The numbers of longterm cancer survivors have increased steadily, but this trend comes at a price. The cumulative incidence of subsequent cancers approaches 15% at 20 years after diagnosis of primary cancer, representing a three- to tenfold increased risk compared with the general population. Some treatment programs have a much higher risk of subsequent cancers, and when possible, less carcinogenic treatment options have replaced some that have especially high rates of secondary malignancies.
BIOLOGY OF BREAST CANCER DISPARITIES
Disparate Incidence and Mortality Rates
Breast cancer is the most commonly reported cancer of women in the United States, with an estimated 230,480 new cases of invasive cancer expected to affect women in 2011 (American Cancer Society, 2011). Overall mortality rates for breast cancer began a steady decline in the 1990s, due to a combination of increased screening and the utilization of therapies that were developed in response to improved understanding of breast tumor biology and the molecular mechanisms driving disease progression (Newman & Martin, 2007). Caucasian women have the highest age-adjusted incidence rates of breast cancer in comparison with all racial/ethnic groups over the age of 45 years. However, African American women have the highest incidence rates among women under the age of 45 years (Amend, Hicks, & Ambrosone, 2006; Newman & Martin, 2007; Polite, Dignam, & Olopade, 2005). Furthermore, African American women have the highest mortality rates from breast cancer and a lower 5-year survival than Caucasian women (American Cancer Society, 2010; Field et al., 2005). By contrast, Hispanic, Asian/Pacific Islander, and American Indian/Alaska Native women have lower incidence and mortality rates of breast cancer than both Caucasian and African American women (American Cancer Society, 2010).
Disparities in Breast Cancer Stage and Tumor Characteristics
Early studies undertaken to examine the impact of race and ethnicity on breast cancer were carried out when characterization of breast cancer primarily involved staging and microscopic appearance of the tumors. These studies demonstrated that African American and Hispanic patients were more likely to present with advanced tumors that had more aggressive histologic features than Caucasian patients, but they were unable to determine whether this was due to delays in seeking medical intervention or whether there were biologic differences in the forms of breast cancer that developed. For example, the National Cancer Institute’s Black/White Cancer Survival Study, which was a cohort of 1,130 women (518 Caucasian and 612 African American) aged 20-79 diagnosed with primary breast cancer between January 1, 1985 and December 31, 1986 (Eley et al., 1994), reported that African American women were almost twice as likely as Caucasian women to be diagnosed at advanced stages: 30% of African Americans presented with stage III or IV cancers compared to 18% of Caucasians. The microscopic appearance of breast cancer reports the degree to which breast cancers resemble normal breast structures (differentiation), how rapidly the cells are replicating (mitotic index), and how much the normal architecture of cells is altered (atypia). The Black/White Cancer Survival Study reported that African Americans more commonly had poorly differentiated tumors (20% vs. 14%) and tumors with high-grade nuclear atypia (15% vs. 10%) in comparison to Caucasians, and 18% of African Americans had high-grade mitotic activity in comparison to 10% of Caucasians, following adjustment for age and stage (Chen et al., 1994; Eley et al., 1994).
Recent studies continue to show that African American women are significantly more likely to present with later-stage tumors than Caucasian women, and are also more likely to display tumor characteristics associated with a poor prognosis, irrespective of age and stage of disease (Curtis, Quale, Haggstrom, & Smith-Bindman, 2008; Porter et al., 2004). African American women are also more likely to be diagnosed with breast cancer at a younger age than Caucasian women, with 35% of African Americans diagnosed under the age of 50 years compared to 21% of Caucasians (Fiel et al., 2005). This raises the possibility that hereditary syndromes and/or distinct environmental exposures might be contributing to earlier development of breast cancer and the worse outcome in African Americans compared to Caucasians. In addition, African Americans are 40% more likely to be diagnosed with inflammatory breast cancer when compared to Caucasians. Inflammatory breast cancer is a very aggressive form of breast cancer in which tumor cells grow and spread to remote sites very rapidly. Prolonged inflammation is well known to increase the risk of developing certain forms of cancer, but the inflammation in this condition is not known to precede the development of cancer. Inflammatory breast cancer makes up approximately 2% of total breast cancer cases and thus plays only a minor role in the disparities in outcome that occur (Hance, Anderson, Devesa, Young, & Levine, 2005).
Older African Americans often have a worse outcome when diagnosed with breast cancer compared to their Caucasian counterparts (Curtis et al., 2008). African Americans on Medicare had a 30% increased risk of death from breast cancer when compared to Caucasian patients, after adjustment for hormone receptor status, tumor size, nodal status, and menopausal status.
There is considerable variability in how patients with breast cancer are managed throughout the United States, but this variability is eliminated for patients on clinical trials, who are closely monitored while following strict protocols. When African American patients on clinical trials were compared to Caucasian patients on the same trials, they experienced worse disease-free survival [HR = 1.56; 95% CI (1.15-2.11)] and overall survival [HR = 1.95; 95% CI (1.36-2.78)] after adjusting for hormone receptor status, tumor size, menopausal status, nodal status, and baseline absolute neutrophil count (Hershman et al., 2009). Collectively, these and other studies suggest that there may be differences in breast cancer biology and progression that contribute to the disparities in disease outcome.
While Hispanic women have a lower incidence of and mortality rate for breast cancer than Caucasians, those who develop breast cancer tend to have a more aggressive breast cancer phenotype. When the tumors of 4,885 Caucasian women, 1,016 African American women, and 777 Hispanic women were collected from 31 hospitals nationwide, both African American (49%) and Hispanic (48%) women were more likely than Caucasian women (39%) to have high levels of cell replication and less cellular differentiation seen within their tumors (Elledge, Clark, Chamness, & Osborne, 1994).
The Role of Inherited Genes in Disparities
BRCA1 and BRCA2 are proteins that play an important role in DNA repair (Kwei et al., 2010; Olopade et al., 2008; Powell & Kachnic, 2008; Zhang & Powell, 2005). People who inherit mutant forms of either BRCA1 or BRCA2 have a 40%-80% lifetime risk of developing breast cancer. BRCA1 and BRCA2 mutation frequencies vary by geographic region and ethnicity (Fackenthal & Olopade, 2007). Within the United States, the Ashkenazi Jewish population has the highest reported frequencies of BRCA1 and BRCA2 mutations, mainly due to three founder mutations (BRCA1 187delAG and 5385insC, BRCA2 6174delT) that have a combined carrier frequency of 1 in 40 for this population. By comparison, the overall prevalence of BRCA1 and BRCA2 mutations is estimated to be between 1 in 400 and 1 in 800 (Petrucelli et al., 2010).
Patients taken from the Northern California Breast Cancer Family Registry were used to assess the BRCA1 mutation carrier frequency within various racial and ethnic groups who developed breast cancer within the United States. These patients were under the age of 65 and were diagnosed with invasive breast cancer between January 1, 1995 and December 31, 2003. Five racial/ ethnic groups, including 549 Caucasians (encompassing both Ashkenazi Jewish Caucasians and non-Ashkenazi Jewish Caucasians), 444 Asians, 393 Hispanics, and 341 African Americans, were tested for BRCA1 mutations. After the Ashkenazi Jewish patient population (8.3%), Hispanics (3.5%) had the highest frequency of inherited BRCA1 mutations, followed by non-Ashkenazi Caucasians (2.2%), African Americans (1.3%), and then Asians (0.5%). Interestingly, this study also reported that African American patients (17%) diagnosed under the age of 35 had a significantly higher BRCA1 mutation frequency when compared to the other non-Ashkenazi Jewish populations (Hispanics [8.9%], Asians [2.4%], Caucasians [7.2%]) in the same age range. Additionally, this study found that the types of BRCA1 mutations varied among the different racial/ethnic groups. The most frequent BRCA1 alterations in Hispanics and Caucasians (including both the Ashkenazi Jewish and non-Ashkenazi Jewish populations) were frame-shift mutations, whereas in African Americans, the most prevalent BRCA1 alterations were missense mutations (John et al., 2007). Other studies have also reported unique and distinct BRCA1 and BRCA2 mutations within the African American population (Ademuyiwa & Olopade, 2003; Olopade et al., 2003). Awareness of different types of mutations helps in screening patients for inherited defects that predispose to breast and other cancers. African Americans also appear to have a higher incidence of unclassified variants of BRCA1 and BRCA2 in comparison to other racial/ethnic populations. These variants are sequences within the gene that differ from sequences found in most people, yet they are not known to adversely affect the structure of the protein and hence are not currently considered mutations. Due to the undetermined significance of these variants, it is unknown whether they modify breast cancer risk and survival.
With the advent of high-throughput sequencing and whole-genome technologies, studies show that the major genetic component contributing to breast cancer predisposition is likely caused by multiple and common low-penetrance genes that act in conjunction to modify breast cancer risk (Olopade et al., 2008). Genome-wide association studies have led to the identification of single nucleotide polymorphisms (SNPs) within a small number of genes, primarily within Caucasian populations with estrogen receptor positive breast cancers. A primary example of this phenomenon was the identification of four SNPs in the FGFR2 gene that were highly associated with breast cancer risk in a study of 1,145 postmenopausal women of European ancestry with invasive breast cancer and 1,142 controls (Hunter et al., 2007). Additional SNPs that are significantly associated with breast cancer risk have been identified in four genes (CASP8, TNRC9, MAP3K1, and LSP1) and three genomic regions (2q35, 8q24, and 5p12). These findings were validated in additional cohorts of Caucasian women; however, in studies of other racial/ethnic populations, the SNPs had a modest effect in an Asian cohort and discordant results in cohorts of African ancestry (Olopade et al., 2008). These studies indicate that associations between genomic variants and cancer risk vary among racial/ethnic populations and, further, they demonstrate the need to assess the relationships between SNPs and disease risk within diverse patient cohorts. The identification of these modifiers of cancer risk may further elucidate why differences exist in genetic susceptibilities to cancer among individuals and populations.
Differences in Proteins That Regulate Cellular Proliferation, Apoptosis, and DNA Repair
The changes in DNA that occur in cells that become cancerous lead to changes in the expression of many proteins, some of which play critical roles in the malignant behavior of the cells. These include cyclins, proteins that serve as co-factors to enzymes that regulate progression through the cell cycle (Malumbres & Barbacid, 2009). Alterations that occur in cancer can lead to high levels of several cyclin proteins independent of the usual signals for their production. In general, women with breast cancer who exhibit low levels of cyclin E expression and high levels of cyclin D1 expression are significantly less likely to die from their breast cancer than women with high levels of cyc-lin E and low levels of cyclin D1 (Porter et al., 2004). African American women had higher levels of cyclin E (OR [odds ratio] = 4.3) and lower levels of cyclin D1 expression (OR = 0.5) than Caucasian women in their breast cancer cells, potentially contributing to the worse outcome for African American women.
One of the most common events in the development of cancer involves the tumor suppressor protein p53 that is mutated or eliminated in over 50% of human cancers (Green & Kroemer, 2009). Protein p53 is a transcription factor that is induced when abnormal DNA is present or under cellular stress, and it leads to expression of multiple proteins, including proteins that inhibit DNA replication, help repair damaged DNA, and induce programmed cell death—all functions that are designed to protect the integrity of the DNA. When the function of p53 is disrupted, cells with damaged DNA are more likely to be propagated and acquire more mistakes. The mutations in the p53 gene (TP53) that most commonly occur in tumors lead to a stabilization of the protein and a loss of its ability to function as a transcription factor, so that high levels of p53 protein reflect loss of p53 function. Patients with stage I and II breast cancers displayed no difference in the frequency of TP53 gene alterations between African Americans (20%) and Caucasians (19%), but there were differences in the types of TP53 alterations that were encountered (Blaszyk et al., 1994; Shiao, Chen, Scheer, Wu, & Correa, 1995). When more advanced stages were included in the analysis, tumors of African Americans (OR = 1.7) more frequently displayed high levels of mutant p53 protein expression when compared to the tumors of Caucasians (Porter et al., 2004). In general, tumors that display high levels of mutant p53 have a more aggressive phenotype, are less likely to respond to adjuvant therapy, and thus might contribute to the worse outcome of African American patients.
Differences in Hormonal Factors That Contribute to Breast Cancer Disparities
Estrogen is known to play an important role in the development and progression of many cases of breast cancer, especially those that express high levels of estrogen receptor (ER) and progesterone receptor (PR). Both genetic and environmental factors affect the levels and types of estrogen that are present. Estrogens are synthesized from cholesterol, with estradiol being the predominant form (Taioli et al., 1999, 2010). The ovaries make most of the estradiol until menopause, but estradiol is also made by fat cells and may play a role in the increased rate of ER positive breast cancer in obese postmenopausal women (Brown & Simpson, 2010).
Two mutually exclusive pathways are involved in metabolizing estradiol. One pathway utilizes the enzyme CYP1A1 to generate an inactive metabolite (2-hydroxyestrone), and the other pathway uses CYP3A4 to generate a metabolite that continues to have estrogenic activity (16a-hydroxyestrone) (Masi & Olopade, 2005). There is an inherited variant of CYP3A4 that is more active and leads to higher levels of the 16a-hydrox-yestrone metabolite, and it was more common in African American girls (62%), when compared to Hispanic (52%) and Caucasian girls (17%) in a study of 137 healthy 9-year-old girls (Kadlubar et al., 2003). This high-activity variant was associated with earlier onset of puberty. These data suggest that higher physiological levels of active estrogen, along with earlier onset of puberty, may contribute to the higher prevalence of both early onset breast cancer and breast cancer-specific mortality in African American women (Masi & Olopade, 2005).
Differences in Molecular Characteristics of Breast Cancers
The alterations in DNA that lead to breast cancer lead to changes in gene expression, which can be assessed using complementary DNA (cDNA) microarrays, a method that allows evaluation of hundreds of genes simultaneously.
Comparison of breast cancer specimens to normal breast cells has revealed that some breast cancers are derived from cells that line the breast ducts (luminal cells), whereas others arise from cells that are beneath the luminal cells (basal or myoepithelial cells). Six distinct subtypes of breast cancer have been identified that have distinct patterns of gene expression: luminal A, luminal B, HER2-enriched, basal-like, normal breast-like, and claudin-low (Perou et al., 2000; Prat et al., 2010). Luminal A tumors have relatively high expression of genes, such as the ER gene, that are normally expressed by cells that line the breast ducts (luminal cells) (Carey et al., 2006; Oh et al., 2006; Prat et al., 2010; Sorlie et al., 2001). Luminal B tumors show low to moderate expression of ER, whereas they express high levels of genes that induce cell proliferation and block programmed cell death (apoptosis), reflecting a more aggressive form of breast cancer compared to luminal A cancers. The HER2-enriched subtype demonstrates overexpression of ERBB2/HER-2/neu and low levels of ER and ER-associated genes. The basal-like subtype generally does not express ER, PR, or ERBB2/HER-2/neu, whereas it has high levels of expression of other proteins, such as keratin 5, keratin 17, and laminin. The normal breast-like subtype is distinguished by high expression of genes characteristic of basal epithelial and adipose cells, along with low expression of genes characteristic of luminal epithelial cells.
Of all the subtypes, luminal A tumors have the best prognosis, demonstrating the highest overall survival and relapse-free survival (Prat et al., 2010; Sorlie et al., 2001). The tumors are slower to metastasize than some of the other forms, and treatments that attack the ability of hormones to drive these tumors have been available for many years. Basal-like, claudin-low, and HER2 classified tumors are more aggressive tumors that metastasize more readily, leading to a worse outcome both in terms of overall survival and relapse-free survival (Di Cosimo & Baselga, 2010). The use of herceptin has dramatically improved the prognosis of patients with HER2-classified tumors, indicating that knowledge of factors that drive tumor growth and survival can be exploited for developing new treatments (Mukai, 2010). Basal-type tumors metastasize early and have been especially difficult to treat, but new insights into their biology are leading to more effective treatments, such as the poly ADP-ribose polymerase (PARP) inhibitors in conjunction with new combinations of cytotoxic chemotherapy (Anders et al., 2010; Di Cosimo & Baselga, 2010).
Molecular characterization using cDNA microarrays reveals that African American women are more likely to have basal-like tumors and less likely to have luminal A or B tumors than Caucasians, providing strong evidence that some of the disparities are not just due to differences in access to care and in tumor management. In the Carolina Breast Cancer Study of 496 cases (196 African American and 300 non-African American) of invasive breast cancer, the prevalence of basal-like tumors was significantly higher in African Americans (26%) than in non-African Americans (16%) (Carey et al., 2006). Additionally, this high prevalence of basal-like tumors in African Americans was mainly seen in premenopausal women, irrespective of stage at diagnosis. This study also found that basal-like tumors were more likely to have high nuclear and histological grade, as well as a high mitotic index, than other tumor subtypes, after adjustment for age, race, and stage. Furthermore, the basal (44%) and HER2 (43%) subtypes had a higher percentage of p53 mutations in comparison to luminal subtypes (luminal A, 15%; luminal B, 23%), thereby offering some insights into why early studies showed that African Americans were more likely to have tumors that displayed unfavorable microscopic features. It is currently not known why African Americans are more prone to develop basal-like tumors than Caucasians. It is also not known why luminal A breast cancers predominate in Asian and Caucasian populations, and are more common in postmenopausal than in premenopausal women. However, further study of the mechanisms driving the individual breast cancer subtypes and the environmental exposures that likely modify these processes will lead to improved understanding of this phenomenon.
BIOLOGY OF COLORECTAL CANCER DISPARITIES
Incidence and Mortality Rates
Colorectal cancer is the third leading cause of new cancer cases and cancer deaths in both men and women within the United States, with an estimated 141,210 cases expected to occur in 2011. Overall incidence and mortality have decreased in the past two decades as a result of improved screening and improvements in treatment. Screening can lead to the removal of premalignant polyps, thereby decreasing the likelihood that cancer will develop. However, incidence rates are increasing about 2% per year in adults under the age of 50, a population that is not recommended for screening unless in high-risk circumstances (American Cancer Society, 2010). African Americans have the highest age-adjusted incidence and mortality rates of any other racial/ethnic group, whereas Asians/Pacific Islanders and American Indians/Alaska Natives have the lowest rates. Both incidence and mortality rates within African Americans are declining, but the decline has been slower than in Caucasians, leading to an increasing divergence, especially in mortality rates (American Cancer Society, 2009, 2010). For those who get colorectal cancer, African Americans (OR = 1.2), American Indians (OR = 1.2), Hispanics (especially Mexicans, OR = 1.2), and Hawaiians (OR = 1.3) are more likely to die than Caucasians, with the greatest disparity in risk of death in early-stage cancers (stages I and II), even while adjusting for age, stage, and treatment (African Americans, OR = 1.4; Hispanics, OR=1.4; Hawaiians, OR = 1.4) (Alexander et al., 2004; Chien, Morimoto, Tom, & Li, 2005). In a study of 574 patients (224 African American and 350 Caucasian) from the University of Alabama at Birmingham Hospital and the Birmingham Veterans Affairs Hospital tumor registries, African Americans with high-grade tumors were three times more likely to die of colon cancer within 5 years of surgical resection when compared to Caucasians with high-grade tumors [HR = 3.05, 95% CI (1.32-7.05)], following adjustment for race, gender, age, hospital, stage, and anatomic site. The African Americans and Caucasians within this study had similar proportions of high-grade tumors at diagnosis, a similar prevalence of comorbid conditions, and a similar frequency of deaths due to causes other than colorectal cancer, suggesting that differences in the aggressiveness of the cancers play an important role in the disparate survival outcomes among these populations (Alexander et al., 2005). Furthermore, in a large population study (33,464 Caucasians, 6,024 African Americans, 1,618 Asian/Pacific Islanders, and 911 American Indian/Alaska Natives or other unidentified racial/ethnic groups), patients diagnosed under the age of 50 were more likely to present with distant disease and poorly differentiated tumors and were more likely to be African American (Fairley et al., 2006). SEER data has also shown that the differential in incidence and survival outcome between African American and Caucasian patients is the largest in younger cohorts (under 50 years of age) (Polite, Dignam, & Olopade, 2006). Collectively, these studies suggest that African Americans are developing cancers with more aggressive phenotypes, thereby leading to their escalated rates of colorectal cancer mortality.
The Role of Genetics in Colorectal Cancer Disparities
Approximately 15% of all colorectal cancers occur in people who have an inherited risk for colorectal cancer. The majority of these are due to inherited mutations in one of the DNA mismatch repair genes (MLH1, MSH2, PMS2, and MSH6) that lead to the development of Lynch syndrome (Kinzler & Vogelstein, 1996). The changes in DNA mismatch repair that occur in this syndrome can affect particular areas of the DNA called "microsatellites," leading to a specific type of genomic instability called microsatellite instability (MSI). MSI can also occur in response to acquired changes in DNA repair, but the presence of MSI within colon cancer cells suggests the presence of either inherited defects in DNA mismatch repair enzymes or acquired mistakes in the function of this pathway. A few studies have examined MSI in colorectal cancers from African Americans and Caucasians to determine if mismatch repair pathway dysfunction is playing a role in the higher rates of young-onset cases and proximally located colon tumors in young African Americans compared to Caucasians. These small studies reported that MSI incidence was more than twofold greater in African Americans when compared to Caucasians (Ashktorab et al., 2005; Ionov, Peinado, Malkhosyan, Shibata, & Perucho, 1993), but larger studies found that the rate in African Americans was approximately 20%, a frequency similar to what has been reported in the U.S. population (Cunningham et al., 2001; Hampel et al., 2005). It remains to be determined whether there are racial differences in the frequency of people harboring defective DNA mismatch repair genes, since not all cases of MSI are due to an inherited defect.
Another genetic syndrome that leads to colorectal cancer is familial adenomatosis polyposis (FAP). The syndrome results from an inherited defect in the APC gene, and people who inherit defective APC develop hundreds of polyps, each with the potential to progress to cancer. Nearly 100% of people who inherit mutant APC develop cancer by the time they are in their forties. The large number of polyps that occur in this syndrome make carriers easy to identify, but FAP is sufficiently rare that it does not play a large role in the racial disparities that exist.
Methylenetetrahydrofolate reductase (MTHFR) is an enzyme involved in folate metabolism, which has been inversely linked to colorectal cancer risk (Le Marchand, Wilkens, Kolonel, & Henderson, 2005). Folate is important in making the building blocks for DNA, and folic acid deficiency leads to double-strand chromosome breaks by extensive incorporation of uracil into DNA. This occurs because there is a deficiency of thymidine when folate is low, so its precursor, uracil, is substituted. Folate also plays a functional role in DNA methylation, so patterns of methylation of the genome change during folate deficiency, and this can alter gene expression (Ames, 1999; Goelz, Vogelstein, Hamilton, & Feinberg, 1985). A common polymorphism, C677T, in the MTHFR gene was previously identified and shown to result in a temperature-sensitive enzyme that functions with decreased activity (Weisberg et al., 1998). In a study of 2,843 cases and controls from the Multiethnic Cohort Study, the MTHFR 677TT genotype was associated with a 23% decreased risk of colorectal cancer (Le Marchand et al., 2005). There was an even stronger association at high levels of folate intake. The study also reported differences in the allele frequency among racial/ethnic groups, with the T allele being the lowest in African Americans and Hawaiians when compared to Hispanics, Japanese, and Caucasians. The differences in the frequency of this allele might contribute to the higher incidences of colorectal cancer in Hawaiian and African American populations.
Environmental Factors Contributing to Colorectal Cancer Disparities
The development of colorectal cancer is dependent on the accumulation of multiple changes in the DNA. Irrespective of whether the colorectal cancer is sporadic or in people with genetic risk, the majority of cancers arise in polyps, so that removal of polyps is likely responsible for the decreasing incidence of colorectal cancer that has occurred in recent years. Insights into the steps involved in the development of cancer come from comparing the DNA in colorectal cancer to DNA in polyps and in unaffected colon or rectum. Such studies show that over 90% of polyps contain an acquired change in the APC gene (either genetic or epigenetic), and additional changes in the DNA are found when polyps grow very large. Not all polyps progress to cancer, but those that do contain many more genetic changes that are responsible for the ability of cells to invade and metastasize.
Epidemiologic studies have shown that alcohol consumption, smoking, diabetes, obesity, and high meat intake are associated with increased risks of colorectal cancer (Akhter et al., 2007; Gapstur, Potter, & Folsom, 1994; Huxley et al., 2009). These exposures can lead to inappropriate DNA methylation and alkylation that promote carcinogenesis (Cross & Sinha, 2004; Slattery, Schaffer, & Edwards, 1997). Therefore, the dietary and lifestyle patterns that are common to specific racial/ethnic populations may induce altered frequencies and/or spectra in genetic alterations among different populations, leading to variations in cancer risk and clinicopathologic features.
Whole-genome analysis of tumors offers insights into some of the changes that occur in colorectal cancer in different patient populations. When 15 colorectal cancer samples taken from African American patients at Howard University Hospital were compared to a previously published analysis of 22 Caucasian colorectal cancer cases from Germany, many of the genetic changes were similar, but there were a few differences (Ashktorab et al., 2010; Lassmann et al., 2007). The ATM gene, whose encoded protein functions in the DNA damage response, was frequently amplified in Caucasian tumors but not in any of the African American tumors. In addition, the DCC gene, whose encoded protein functions as a receptor that can induce programmed cell death (Rodrigues, De Wever, Bruyneel, Rooney, & Gespach, 2007), was primarily amplified in Caucasians but deleted in African Americans (Takayama, Miyanishi, Hayashi, Sato, & Niitsu, 2006). They also reported that the STS gene, involved in promoting the growth of human breast cancer cells, was deleted in Caucasians and amplified in African Americans (Ashktorab et al., 2010). The functional consequences of these differences are not known, but they indicate that differences exist in the types of changes that occur in tumors from various patients, underscoring the need for characterizing both the causes and the consequences of these differences.
Researchers have made tremendous progress in understanding many of the biologic changes that occur in cancer and some of the factors that initiate and perpetuate various forms of cancer. Vaccines, antibiotics, and other interventions are helping reduce some of the cancers that are initiated by infectious agents, and government regulations are reducing exposures to known carcinogens that once were more commonly encountered by the public. Multiple genes that are inherited and lead to increased risk for cancer have been identified, offering the opportunity to screen for people who are at increased risk. Screening for some forms of cancer has improved outcome through identifying cancers before they have spread, and clinical trials have helped define combinations of treatment that improve outcome. Knowledge of how specific genes and biologic processes contribute to tumorigenesis offers the opportunity for developing interventions that reduce risk and/or allow early detection of premalignant and malignant changes. Various forms of cancer are being characterized by some of the specific mutations and changes in gene expression that occur, offering more detailed understanding of biologic similarities and differences in cancers that were previously only characterized by their microscopic appearance. For those who get cancer, specific pathways that are critical to tumor cell survival and growth are now being targeted in many forms of cancer, leading to more effective treatments, with fewer long-term side effects. These and other advances in detection and treatment are responsible for the more than 10 million cancer survivors in the United States today.
Population studies have alerted the nation to the disparities in cancer incidence and mortality rates among the diverse racial/ethnic populations within the United States. Health care access plays a vital role in cancer health disparities, as unequal access plagues many, especially minority and impoverished populations. Biologic factors also contribute to the differences in incidence and outcome of cancer in different racial and ethnic groups. Racial and ethnic classifications do not necessarily align with population ancestry and hence are limited in their ability to identify people who might have some shared genetic background. African Americans are a heterogeneous group with admixture from African, European, and American Indian populations. Asian/Pacific Islander and Hispanic designations include a variety of ancestries within each racial/ethnic construct. To begin to refine our understanding of how genes contribute to tumor risk, researchers have begun to incorporate ancestry informative markers (AIMs) into their analysis of diverse populations, as a method to reduce bias associated with population stratification (Nassir et al., 2009).
Elimination of the disparities that exist in cancer incidence and outcome is an important goal, but it is not sufficient. There are approximately 1.5 million people who get cancer in the United States each year, and more than 550,000 people die of it. Elimination of cancer disparities would have a valuable impact on reducing these numbers. Through knowledge, it should be possible to develop more effective interventions that decrease incidence and mortality from cancer, to benefit people from all racial and ethnic groups.