Abstract
Colorectal cancer is the third leading cause of cancer-related deaths in women. Colorectal cancer is a preventable disease with accepted screening modalities that have been proven to save lives. As women are more likely than men to develop right-sided colon cancers, colonoscopy is the preferred screening test in women. Currently, women are less likely to undergo colorectal cancer screening than men. Frank discussions addressing the fear or embarrassment of endoscopic screening are important in helping women overcome these barriers. Enhanced education of both practitioners and patients targeted to improve colorectal cancer screening adherence will improve early diagnosis and patient survival.
Colorectal cancer (CRC) is the third leading cause of cancer-related deaths in women in the USA (Figure 1). In 2006, there will be an estimated 75,810 new cases of CRC diagnosed in women and the disease will account for 27,300 deaths [1]. For both sexes, there is a lifetime incidence of 6%. If diagnosed early, the 5-year survival rate is greater than 90%, as opposed to metastatic disease, which has a 5-year survival rate of less than 5% [2].
2002 US mortality statistics in women.
Pathogenesis
Significant genetic advances in the past decade have improved our understanding of CRC pathogenesis. Much of this knowledge comes from the genetic study of hereditary colon cancer, particularly familial adenomatous polyposis (FAP) and hereditary nonpolyposis colorectal cancer (HNPCC) [3]. It is now known that colon cancer arises from two pathways: an accumulation of mutations in specific genes that control cell growth and DNA repair, which accounts for 80% of all CRC; or from microsatellite instability [4,5]. The malignant transformation process occurs over a number of years. First, there is transformation of normal colonic epithelial cells to small adenomatous polyps; these grow to larger and more dysplastic polyps, which finally develop into invasive adenocarcinoma. The cumulative risk of developing invasive carcinoma from adenomatous polyps increases with time; the risk of carcinoma is 5% at 5 years, 14% at 10 years and 35% at 20 years of follow-up [6]. Polyps at highest risk for malignant transformation include those with a diameter greater than or equal to 1 cm, those with more villous features or those with high-grade dysplasia.
Risk factors
Age
Epidemiological studies have shown an increase in CRC beginning at age 40 years, with a sharp rise between the ages of 50 and 55 years. The risk doubles with each succeeding decade and continues to rise exponentially [7]. The lifetime risk of diagnosis of and death from CRC is similar in both sexes, although the age-adjusted mortality and incidence is higher in men [8]. Data from the Surveillance, Epidemiology and End Results program demonstrate that the CRC risk of a 50-year-old white man is similar to that of a 55-year-old white woman.
Family history
Family history of CRC and polyps increases the likelihood of CRC, with the cancer tending to occur at an earlier age [9]. On average, people with an affected first-degree relative have the same risk of CRC as patients without a family history; however, these patients develop CRC approximately 10 years earlier [10], thus providing the rationale for screening in this population to begin at age 40 years.
Inflammatory bowel disease
Patients with a history of ulcerative colitis (UC) or Crohn's disease have an increased risk of developing CRC [11,12]. CRC is observed in 3.5–5.7% of all patients with UC and in 1.9% (95% confidence interval [CI]: 1.4–2.5) of those with Crohn's disease [12–14]. Established risk factors include duration of the disease, extent of diseased colon, age of onset, stenotic disease, inadequate pharmacological therapy and folate deficiency [15–19].
Genetic syndromes
Genetic syndromes account for approximately 5% of CRCs [20]. FAP, with an autosomal dominant inheritance, occurs due to a mutation in the tumor suppressor gene adenomatous polyposis coli. Patients with this syndrome have a 100% colon cancer risk if left untreated, with most cancers occurring before the age of 40 years. Germline mutations in mismatch repair genes lead to another genetic syndrome, HNPCC. Cumulative risk for CRC in HNPCC, if left untreated, is approximately 60%.
Diabetes mellitus
Observational studies have shown an approximately twofold higher incidence of CRC in patients with diabetes mellitus [22]. This is secondary to hyperinsulinemia, which has been recognized as a growth factor that promotes proliferation of colon cancer cells both indirectly and directly through insulin growth factor-1 [23,24]. Recently, Elwing and colleagues found that women with Type 2 diabetes mellitus had higher rates of colorectal adenomas compared with lean and nondiabetic women, adding further evidence that Type 2 diabetes may be an important contributor in the evolution of the adenoma–carcinoma sequence [25].
Pelvic irradiation
Pelvic radiation for benign or malignant causes is associated with a 2.2–3.6-fold increase in CRCs [26]. Generally the cancers occur on average 15–20 years after radiation treatment.
Diet
Data regarding fat intake are inconsistent, with some large cohort studies suggesting a positive association between fat intake and colon carcinogenesis, while others do not [27–29]. However, an association between higher intakes of processed meats or red meat and CRC has been found [30–32]. Data from prospective studies have shown a 12–17% increased risk of CRC with consumption of 100 g/day of red meat [30,31].
Female-specific cancers in relation to colorectal cancer
Endometrial cancer
In 2006, approximately 41,200 women were diagnosed with endometrial cancer, accounting for 6% of all cancers. The average age of diagnosis is between the ages of 55 and 65 years. An association between endometrial cancer and CRC occurs specifically in patients with HNPCC. In these patients, endometrial cancer can be the sentinel cancer [33]. Patients with HNPCC have a 40–60% chance of developing endometrial cancer [34,35]. Owing to this, experts recommend screening colonoscopy in patients diagnosed with endometrial cancer before the age of 40 years. On the other hand, women with HNPCC, or those with a substantial likelihood of being a mutation carrier (i.e., a mutation is known to be present in the family) with a suspected autosomal dominant predisposition to colon cancer, should consider beginning annual testing for early endometrial cancer detection at age 35 years [36]. A recent retrospective report by Schmeler and colleagues confirmed the efficacy of prophylactic hysterectomy for the prevention of endometrial cancer in patients with HNPCC [37].
Ovarian cancer
The lifetime risk of ovarian cancer increases from a baseline of 1.5% to approximately 5–10% in HNPCC carriers. The average age of onset of ovarian cancer in patients with HNPCC differs markedly from the general population, occurring approximately 16 years earlier. A review by Watson and colleagues found that the type of cancer is more likely to be epithelial, detected at an earlier stage and moderately or well differentiated at the time of diagnosis [38]. At the time of hysterectomy, HNPCC patients are recommended to undergo a bilateral salpingo–oophorectomy [39].
Breast cancer
Patients with a rare syndrome known as Cowden's disease, or the multiple hamartoma syndrome, have a 25–50% risk of breast cancer. These patients may be at an increased risk of colon cancer, although no specific CRC screening strategies have been recommended. Data regarding CRC risk in patients with a prior diagnosis of breast cancer are conflicting. Two meta-analyses have suggested a 20% increased risk of CRC in patients with a prior history of breast cancer, in contrast to a large retrospective cohort study of 17,415 breast cancer patients and 69,660 matched controls, which demonstrated a 20% reduced risk of CRC (odds ratio [OR]: 0.80; 95% CI: 0.56–1.15) [40,41]. At present, women with a prior diagnosis of breast cancer can follow average-risk screening guidelines for CRC.
Colorectal cancer during pregnancy
CRC during pregnancy is rare, with a reported incidence of approximately 0.002% [7,42]. In pregnancy, the diagnosis of CRC is challenging for numerous reasons. CRC is generally a disease of middle age, with 92–96% of colon cancers occurring in patients aged 50 years or older. In addition, the presenting signs/symptoms of CRC, such as rectal bleeding, iron deficiency anemia and constipation, are often attributed to the normal complications/manifestations of pregnancy. This may lead to a delay in diagnosis and allow the cancer to progress to an advanced stage. Unfortunately, CRC generally does not present with symptoms until advanced disease.
Carcinogenesis of colon cancer in pregnancy is not clear; however, studies suggest that the increased levels of estrogen, progesterone and prolactin related to pregnancy stimulate the development of CRC [43–45]. Primary-care providers and obstetricians should refer the pregnant patient with significant gastrointestinal symptoms to a gastroenterologist for evaluation. Expert opinion recommends postponing endoscopy until the second trimester [46]. Case series of lower endoscopy have demonstrated the safety of flexible sigmoidoscopy; however, data on colonoscopy are more limited [47,48]. The gastroenterologist should be prepared to perform sigmoidoscopy (preferably without endoscopic medications) for significant lower gastrointestinal symptoms such as persistent rectal bleeding. Once diagnosis is made, the aim of treatment is to start therapy for the mother as soon as possible and to deliver the baby at the earliest time allowable. The management mandates a multidisciplinary approach involving experts in obstetrics, neonatology, gastrointestinal surgery and medical oncology.
Colorectal cancer screening methods in average-risk patients
Recommended screening modalities
Five modalities are currently recommended by the US multisociety task force for CRC screening for average-risk screening (
Fecal occult blood testing
Testing involves two samples from three consecutive stools using either a guaiac-based test with dietary restriction (avoiding red meat, turnips or horseradish) or an immunochemical test without dietary restriction. One-time FOBT has a sensitivity for detecting advanced neoplasia (CRC and any polyps >1 cm) of 4.9% versus three consecutive stool samples, which improves sensitivity to 23.9% [50]. In a large national survey, a third of primary-care providers utilized only a single sample digital rectal exam rather than the recommended collection of two samples from three consecutive stools as their primary CRC screening modality [51]. It is important to reinforce the proper usage of FOBT, since, when properly performed, repeated annual testing can detect up to 92% of cancers [52]. In addition, three large randomized, controlled trials have demonstrated that the use of FOBT leads to an 18–33% reduction in CRC deaths [52–54].
Screening guidelines for average-risk men and women.
Checks for hidden blood in three consecutive stool samples and should be administered every year.
Physicians use a flexible, lighted tube (sigmoidoscope) to visually inspect the interior walls of the distal colon and rectum. This should be administered every 5 years.
Test that uses a series of X-rays of the colon and rectum (taken after the patient is given an enema containing barium dye, followed by an injection of air in the lower bowel). This should be administered every 5 years.
Combination of two of the above modalities.
Physicians use a flexible, lighted tube (colonoscope) to visually inspect the interior walls of the rectum and the entire colon. This test should be administered every 10 years. During this procedure, samples of tissue may be collected for closer examination, or polyps may be removed. Colonoscopies can be used as screening tests or as follow-up diagnostic tools when the results of another screening test are positive.
Adapted from [89].
Flexible sigmoidoscopy
Flexible sigmoidoscopy is a cost-effective method of CRC screening that has been shown, in case–control studies, to reduce CRC mortality by 59–70% [55–57]. However, there are reasons to not recommend flexible sigmoidoscopy as the optimal test for CRC in women. Women are more likely than men to have a limited flexible sigmoidoscopy examination secondary to differences in anatomy [58–60]. In women, perhaps due to the uterus and ovaries, the sigmoid colon appears to have an increased angulation as the sigmoid descends toward the left inguinal area [61]. This can lead to increased difficulty in the procedure and increased pain, since flexible sigmoidoscopy is performed without conscious sedation. However, the strongest reason to not recommend flexible sigmoidoscopy as the CRC test of choice is that proximal cancer lesions will be missed, since the procedure only examines approximately 60 cm of the distal colon. This issue is particularly pertinent to women with the recent findings of the Colorectal Neoplasia Screening with Colonoscopy in Average-Risk Women at Regional Naval Medical Centers (CONCeRN) trial. This demonstrated that 65% of advanced neoplasia would be missed in women if flexible sigmoidoscopy was the only utilized screening modality [61]. The CONCERN trial was a parallel study to the Veterans Affairs Cooperative study group 380 trial, which was performed in men and demonstrated that only 35% of advanced neoplasia would be missed if flexible sigmoidoscopy was utilized for CRC screening [62]. Women are less likely to have distal lesions than men (35 vs 65%); therefore, flexible sigmoidoscopy alone is insufficient for CRC screening.
Barium enema
Barium enema is recommended as a CRC screening test every 5 years. It is advantageous due to its low cost and its ability to examine the whole colon. However, in the National Polyp Study, barium enemas missed 50% of adenomas greater than 1 cm in size [63]. Furthermore, specificity is poor and the best screening interval is unknown.
Flexible sigmoidoscopy & fecal occult blood testing
Combining flexible sigmoidoscopy with FOBT enhances screening efficacy. Lieberman and colleagues found that combining both modalities increased the sensitivity as a CRC screening tool by up to 76% versus 70 and 24% sensitivity for flexible sigmoidoscopy and FOBT, respectively [50].
Colonoscopy
In 2000, federal law entitled Medicare beneficiaries to a screening colonoscopy every 10 years. Since that time, several health plans have followed suit and now cover colonoscopy. Colonoscopy is the preferred modality for CRC screening as it allows full visualization of the colon, detection and removal of polyps, and diagnosis of colon cancer. Winawer and colleagues demonstrated a lower cumulative incidence of CRC following polypectomy relative to the general population registries. After a mean follow-up of nearly 6 years, only five cancers were detected, compared with the expected rate of 20.7–48.3 cases, demonstrating a 76–90% reduction in CRC [64]. Additional European cohort studies have shown a 66–80% reduction in CRC incidence after polypectomy relative to those without CRC screening [65,66]. These findings support the principle that colonoscopic polypectomy prevents CRC.
For reasons similar to those for flexible sigmoidoscopy, colonoscopy can be more difficult in women than in men. Women have increased colon length compared with men, despite their smaller stature and smaller abdominal cavity. In addition, a history of hysterectomy can increase the likelihood of an incomplete colonoscopy by 53%.
Emerging screening tests
Two modalities, fecal DNA and computed tomography (CT) colonography, deserve mention as possible future screening modalities. Fecal DNA is a simple, noninvasive test that involves the collection of 30 g of stool to detect tumor-specific products. Initial studies comparing patients with advanced, symptomatic CRC were promising, with sensitivity and specificity for CRC of 91 and 93%, respectively [67]. However, Imperiale and colleagues performed a randomized, controlled trial of fecal DNA in asymptomatic, average-risk patients, which demonstrated a sensitivity and specificity of 51.6 and 94%, respectively, for CRC, and a sensitivity and specificity for patients with advanced neoplasia (defined as CRC, polyp >1 cm, villous component or high-grade dysplasia) of 18.2 and 94%, respectively [68]. The other future modality, CT colonography, uses thin-section helical CT with 3D reconstruction. Patients still require bowel preparation and the procedure requires air insufflation, which can lead to discomfort. Numerous clinical trials have yielded mixed results in the average-risk screening population. The four largest trials have shown sensitivities for polyps with a diameter over 1 cm ranging from as low as 48% to as high as 94%. The trial by Pickhardt and colleagues yielded the highest results and was technically different from other trials in that it relied most heavily on the 3D reconstruction, and stool was labeled using dilute barium and gastrografin [69]. However, due the wide disparities between the trials, CT colonography is still not endorsed as a primary modality for CRC screening. Further confirmatory studies are required.
Recommended colorectal cancer screening in high-risk patients
Patients considered to be high-risk include those with a family history of colon cancer, adenomatous polyps, FAP or HNPCC, and a history of inflammatory bowel disease. Patients with a family history of colon cancer and adenomatous polyps diagnosed at age less than 60 years, or two first-degree relatives of any age, should undergo colonoscopy at age 40 years, with subsequent colonoscopies every 5 years. Patients with a first-degree relative aged over 60 years, or two second-degree relatives, with colon cancer or adenomatous polyps should be screened as average-risk persons, starting at age 40 years. It is important to highlight that patients with one second-degree or any third-degree relative with CRC have the same risk of CRC as average-risk patients. Patients with pancolitis UC and Crohn's disease should undergo colonoscopy 8–10 years after diagnosis for CRC screening, whereas patients with left-sided UC should undergo surveillance colonoscopy after 15–20 years of disease. Patients who are gene carriers or at risk for FAP should undergo sigmoidoscopy annually, beginning at age 10–12 years. Gene carriers or those at risk for HNPCC should undergo colonoscopy every 1–2 years, beginning at age 20–25 years or 10 years younger than the family's earliest case [49].
Predictors of colorectal cancer screening adherence
Numerous studies have identified predictors of CRC screening adherence. In assessing demographics, only age and higher level of education have correlated with increased CRC screening adherence. Patients between the ages of 65 and 85 years are most likely to be compliant [70,71]. Findings from the 1987 and 1992 National Health Interview Surveys (NHIS) demonstrated that compliance with CRC screening modalities increased with higher educational levels. Previous compliance to non-CRC-related cancer screening tests also correlates with an increased likelihood of CRC screening adherence. Specifically, adherence to mammography and Papanicolaou smear within the past year is associated with improved CRC screening [72–75]. These associations may be explained by patient motivation and increased interactions with a healthcare provider. Patients' personal attitudes toward cancer and the benefits of screening also have a large effect on CRC screening adherence. In the NHIS, patients' acknowledgment of the benefits of colon cancer screening, for example that colon cancer is preventable and curable, positively affects compliance. Patients with a family history of CRC are consistently more likely to undergo CRC screening; individuals are generally twice as likely to be adherent than those without a family history [72–74].
Women are different from men.
Women are more likely to develop right-sided colon cancers
Women are less likely to be compliant with colorectal cancer (CRC) screening
Women with female providers are more likely to be compliant with CRC screening
Women are more likely to be embarrassed by endoscopic CRC screening
Women are more likely to have fear of discomfort for endoscopic CRC screening
Women are more likely to prefer a female physician for their endoscopic CRC screening
Physician recommendation has been cited as a major factor influencing a patient's decision to undergo CRC screening [78–83]. Under specific physician recommendations, patients are 12–17-fold more likely to comply with CRC screening [76,84]. Weitzman and colleagues demonstrated that persistent recommendations and a strong personal directive from a doctor facilitated FOBT and flexible sigmoidoscopy use, overcoming patient reluctance to screening [78]. The physician's gender may also affect CRC screening adherence, as it has been seen to do so in other female cancer screening. In breast and cervical cancer, data from the gynecological and primary-care literature have demonstrated that patients seen by female physicians are more likely to undergo screening than patients seen by male physicians [85,86]. This association has also been seen in CRC screening adherence. In a recent retrospective study, Menees and colleagues found that having a female physician was associated with an increased prevalence of CRC screening in women at the time of upper endoscopy and increased CRC screening completion 6 months later [87].
Barriers to colorectal cancer screening
Despite overwhelming evidence of reduced CRC-related mortality through current screening practices, CRC screening adherence continues to be insufficient. Two national surveys, the Behavioral Risk Factor Surveillance System and the NHIS, found 57 and 42.5% adherence rates, respectively, for all accepted screening methods [88,89]. Consistently, women lag behind men in CRC screening adherence [90]. Women's low adherence to CRC screening is thought to be secondary to inadequate understanding of the health benefits that screening provides, perceived low risk of CRC, failure of a physician to recommend the procedure or poor insurance coverage [72,78,91,92]. Reports of primary-care physicians' CRC screening practices vary widely. In a rural population, discussions for CRC screening occurred in only 14% of the eligible population, whereas a recent national survey of physicians demonstrated that 98% of the polled primary-care physicians recommended CRC screening in their patients [93,94]. Embarrassment and psychological fear of discomfort associated with the procedure may reduce CRC screening, especially in the female population [72,95–98]. Weinberg and colleagues noted that the women who were more likely to be concerned regarding screening-related pain were those who were older, less likely to believe in screening guidelines or that screening would reduce CRC risk, and more likely to be anxious and embarrassed by CRC screening [98]. The gender of the endoscopist performing the procedure may also be a reason for reduced compliance. Recent studies have shown that 43–48% of women have a preference for gender-concordant endoscopists [96,99,100]. Reasons listed for a female gender preference included less embarrassment and the perception that a female endoscopist was more empathetic, a better listener and technically better [96]. In 5% of women, female endoscopist preference translates into an absolute barrier to colonoscopy [96].
Conclusion
CRC is a preventable disease, with accepted screening modalities that save lives. Despite this, women are less likely to be adherent to CRC screening than men. Women are also more likely to develop right-sided colon cancers, thus reinforcing the need for colonoscopy as the preferred screening test. Frank discussions addressing the fear or embarrassment of endoscopic screening are important in helping to overcome these barriers. Guaranteeing a female endoscopist at the time of colonoscopy may help alleviate some of these concerns. Further education of both practitioners and patients will continue to improve CRC screening adherence.
Future perspective
Programs targeted at universal screening with the removal of barriers, including fear of discomfort, embarrassment and cost, will allow for early detection and will save lives. Technological advancements in new and refined procedures, including smaller endoscopy equipment, CT colonography, genetic markers and cancer-specific antigens, will hopefully lead to more accurate and less invasive screening for colon and rectal cancers.
Executive summary
Colorectal cancer (CRC) is the third leading cancer killer in US women.
CRC can often be prevented. Regular screening tests can find precancerous colorectal polyps so they can be removed before they turn into cancer.
Screening can find CRC early, when treatment can be very effective: screening saves lives.
Polyps and CRC may not cause symptoms, especially at first.
Both men and women are at risk. Women are at higher risk for right-sided colon cancers not detected by screening sigmoidoscopy. Therefore, colonoscopy is the screening method of choice for women.
Many insurance plans, including Medicare, help pay for CRC screening.
Physicians must counsel their patients regarding the need for CRC screening and relieve their fears and anxieties. Some women may prefer referral to a female endoscopist.