Criteria	Colorectal Cancer
Disease is common	5%-6% lifetime risk
Early detection can prevent mortality	5-yr survival
	Stage I: near 100%
	Stage II: 80%
	Stage III: 30%-70%
	Stage IV: 10%
Screening methods are shown to be effective	See text
	FOBT RCTs
	Sigmoidoscopy and colonoscopy case-control studies
Resources are available to provide screening in U.S.	FOBT: Yes; primary care settingSigmoidoscopy: Yes; primary care or specialty clinicCT colonography: No; limited centers and fully trained radiologistsColonoscopy: Uncertain capacity (see text)



Resources are available to provide diagnostic tests for patients with positive screening	Colonoscopy resources are generally available if initial screening test positive
Screening is cost-effective	Models show cost-effectiveness
Screening methods are accepted by patients and providers	Yes: 50% adherence in U.S.

CT, computed tomography; FOBT, fecal occult blood test; RCT, randomized controlled trial.

Since 1985, there has been a slow but steady decline in CRC incidence in the United States. During this time period, screening rates have gradually increased to more than 50% of the U.S. population older than 50 years. From 2000–2005, the incidence of CRC declined by 6% in men and 9.4% in women. The annual percent reduction in mortality from 2002–2004 was 4.7%.¹

Several different screening programs exist, each with advantages, limitations, and degrees of uncertainty. All CRC screening programs share a common final pathway for the evaluation of positive tests: colonoscopy. The role of endoscopy in colon cancer detection and prevention is paramount. Effective CRC screening depends on the availability of high-quality colonoscopy. This chapter discusses the epidemiology and pathogenesis of CRC and reviews the rationale for screening and surveillance in both high-risk groups and average-risk individuals.

Epidemiology of Colorectal Cancer

Age, Gender, and Race and Ethnicity

CRC represents the third most common form of cancer (excluding skin) among men and women and is the second leading cause of cancer death overall in the United States. There is a progressive increase in risk associated with age (Table 35.2). Age-adjusted incidence and mortality of CRC are higher in men. There is some evidence that natural premenopausal hormones or postmenopausal hormone replacement therapy may exert some protective effect in women.^6,⁷

Table 35.2 Probability of Invasive Colorectal Cancer by Age and Gender

	Male	Female
<40 yr	0.08 (1/1329)	0.07 (1/1394)
40–59 yr	0.92 (1/109)	0.72 (1/138)
60–69 yr	1.60 (1/63)	1.12 (1/89)
≥70 yr	4.78 (1/21)	4.30 (1/23)
Lifetime	5.65 (1/18)	5.23 (1/19)
Risk of death	2.45 (1/41)	2.45 (1/41)

Data from Jemal A, Siegel R, Ward E, et al: Cancer statistics 2009. CA Cancer J Clin 59:225–249, 2009.

There is strong evidence that African American men and women are more likely to die from CRC than whites (Table 35.3).¹ African Americans are more likely to have advanced lesions at the time of diagnosis⁸ and are more likely to have proximal CRC compared with other races.^9,¹⁰ The reasons for increased mortality in African Americans are unknown. Delays in diagnosis could be due to socioeconomic status or poor access to health care, although some compelling data suggest that there are biologic differences.^11,¹² In a study of white and African American patients undergoing screening colonoscopy, African American men and women had higher age-adjusted prevalence of polyps larger than 9 mm.⁸ These data support the hypothesis of biologic differences.

Table 35.3 Incidence and Mortality of Colorectal Cancer by Race and Ethnicity

Variation by Country

Incidence rates vary around the world. The highest risk is in North America, Western Europe, Australia, New Zealand, Israel, and Japan (mortality 30 to 60 per 100,000). The lowest risk is found in India and most of Africa (mortality <10 per 100,000). Migrant studies have found rapid changes in incidence rates, suggesting that the disease is very sensitive to changes in environment. Incidence rates reach the levels of the host country within one or two generations.

Lifestyle and Medications

There is strong epidemiologic evidence that lifestyle factors have an impact on CRC risk. Diets that are high in fat and low in fiber and calcium may be associated with increased risk. Obesity, metabolic syndrome, low levels of physical activity, tobacco smoking, and heavy alcohol consumption have been associated with an increased risk of CRC. There is no evidence that modification of these factors reduces CRC risk; however, changes in most of these behaviors are generally good for overall health. There is strong evidence that individuals taking nonsteroidal antiinflammatory drugs and aspirin have a lower risk of developing colon adenomas and CRC.¹³ However, these drugs have significant adverse effects, and are not recommended for routine CRC prevention.

Polyps and Colorectal Cancer Risk

Most CRC develops from adenoma precursors. Adenoma prevalence increases steadily with age and exceeds 35% in adults older than 75 years. There is a 5% to 6% lifetime risk of CRC, so most patients with adenomas apparently do not develop clinically detectable cancer during their lifetime. Polyp size and histology are directly associated with malignant risk. The risk of high-grade dysplasia in polyps smaller than 5 mm is 1.1%; for polyps 5 to 9 mm, the risk is 4.6%; and in polyps larger than 9 mm, the risk is 20%.¹⁴ The risk of invasive cancer is less than 1% in polyps smaller than 1 cm and greater than 10% in larger polyps.¹⁵ These data suggest that patients most likely to develop cancer are patients with advanced adenomas. Evidence from the National Polyp Study¹⁶ supports the hypothesis that detection and removal of adenomas may prevent cancer incidence. All of these data have important implications for screening. Screening efforts directed at advanced adenomas would target patients with the greatest risk and potentially lead to significant reductions in cancer incidence.

Pathogenesis

Several genetic pathways may result in the development of adenomas and CRC. In 1990, a stepwise model of chromosomal instability was proposed by Fearon and Vogelstein.¹⁷ More than 80% of sporadic CRC is associated with mutation of the adenomatous polyposis coli (APC) gene on chromosome 5. The mutation of this tumor suppressor gene most likely promotes the development of adenomas. In familial adenomatous polyposis (FAP), there is a germline mutation of this gene. Normally, this gene acts to phosphorylate beta-catenin, which leads to its destruction. Accumulation of beta-catenin leads to unregulated cell proliferation and suppression of apoptosis.^18,¹⁹ The progression of adenoma to cancer seems to require additional mutations (K-ras, chromosome 18, TP53). The observation that not all tumors acquire the same mutations strongly suggests that there are several genetic pathways to malignant invasion.

A second genetic pathway (15% to 20%) of sporadic cancer is due to mutation of mismatch repair genes.²⁰ These genes normally repair errors in DNA replication. Germline mutations are found in patients with hereditary nonpolyposis colorectal cancer syndrome (HNPCC). The key feature of this mutation in HNPCC is rapid development of polyps that progress to malignancy. In sporadic cancers, acquired mutation of these genes leads to microsatellite instability and malignant progression. Patients with large serrated adenomas in the proximal colon may have BRAF mutations and hypermethylation, which may result in microsatellite instability and progression to malignancy.²¹

Another germline mutation in a base-excision repair gene, MYH, has been associated with recessive inheritance of multiple colorectal adenomas. The typical phenotype is a patient with 15 or more adenomas and no germline APC mutation.²²

High-Risk Groups

Recognition of inherited syndromes and other high-risk diseases associated with CRC is key to appropriate screening and management (Table 35.4). It is important for primary care providers to recognize these syndromes and refer patients for appropriate screening and surveillance.

Table 35.4 Risk Stratification

Perhaps the most important questions in the medical history are the following: (1) “Do you have a first-degree relative with CRC?” (2) “If yes, did any relative have cancer before age 50 years?” Patients with an index relative younger than 50 years old should be considered at risk for an inherited syndrome and warrant intensive screening at a young age. Recommendations for specific syndromes are summarized in Table 35.4.

Inherited Polyposis Syndromes

Patients with FAP have a germline mutation of APC on chromosome 5, predisposing them to adenoma formation. Most affected patients have more than 100 adenomas. All affected individuals (100%) develop CRC, and when the phenotype is recognized, colectomy should be considered. The average age of adenoma appearance is 16 years, and the age of CRC appearance is 39 years. In a variant of this syndrome, called attenuated FAP, mutations occur at either the 5′ or 3′ end of the gene. Phenotypically, patients have fewer polyps, with delayed onset of adenoma and cancer formation. Familial colon cancer in Ashkenazi Jews may be the result of a specific germline mutation of the APC gene (I1307K). This mutation seems to predispose to sporadic mutations at distant sites resulting in a high malignant risk.¹⁹

Family members should be advised to have genetic counseling and testing. Current tests detect 80% to 90% of FAP families. If the index patient has a retained rectum after subtotal colectomy, screening with sigmoidoscopy every 6 to 12 months is recommended. Index patients are at risk for other malignancies, the most common of which is duodenal or periampullary cancer (5% to 12% lifetime risk). Upper endoscopy with side-viewing instruments to visualize the ampulla is recommended beginning at age 20 to 25 years and should be repeated every 1 to 3 years. There is evidence that colon adenomas regress with cyclooxygenase-2 selective inhibition. Many experts recommend that patients use cyclooxygenase-2 inhibitors after colectomy, hoping to reduce the risk of upper gastrointestinal malignancy.

MYH-associated polyposis leads to development of multiple polyps that may be adenomatous or hyperplastic. Mutations of the MutY human homolog (MYH) gene, which encodes a member of the base excision repair system, occur. This system normally protects cells against the mutagenic effects of aerobic metabolism. Inheritance is autosomal recessive and should be considered in patients with 10 to 100 adenomas or hyperplastic polyps who have negative genetic testing for FAP.²²

Hereditary Nonpolyposis Colorectal Cancer Syndrome

HNPCC accounts for about 2% of all colon cancers. Affected individuals inherit one of several mismatch repair gene mutations. There is evidence that regular screening of kindreds can substantially reduce the risk of CRC.²³ The clinical definitions have been modified over the years from the original Amsterdam criteria²⁴ to the Bethesda guidelines,²⁵ which provide a much less rigid clinical definition. Clinical suspicion should be high if there is a family member with CRC before age 50, multiple generations with CRC, and relatives with HNPCC-associated cancers (endometrium, small bowel, ureter, or renal pelvis). The findings of the Finnish Cancer Registry²⁶ reinforce the need to be aware of other cancers that may develop in these kindreds, including endometrial (60%); stomach (13%); ovary (12%); bladder, urethra, and ureter (4.0%); brain (3.7%); kidney (3.3%); and biliary tract and gallbladder (2.0%).

Patients who meet the Bethesda criteria should have genetic counseling. If the index family member has specific mismatch repair gene mutation, testing can be performed in other family members to determine who is at risk. Family members should have colonoscopy surveillance every 2 years beginning at age 20 to 25 years and then annually beginning at age 40 years. In addition, screening for upper gastrointestinal cancers with endoscopy and pelvic examinations with transvaginal ultrasound are recommended every 1 to 2 years.

Familial Risk

Epidemiologic data ^27,²⁸ show that individuals with a first-degree relative with CRC have an increased risk of CRC. Twin studies estimated that 35% of CRC arose from inherited factors and 65% arose from environmental factors.²⁹ A meta-analysis considered risk associated with familial risk.³⁰ The relative risk of CRC with an affected first-degree relative was 2.4. With more than one relative, the risk was 4.2. If CRC was diagnosed before age 45 years in the index family member, the relative risk was 3.8; diagnosis at age 45 to 59 years was associated with a relative risk of 2.2. If diagnosis was made after age 59 years, the relative risk was 1.8. The risk associated with second-degree relatives is less certain because this is difficult to study. An analysis from the Utah registry found a risk of 1.5 for patients with second-degree relatives with CRC.

Based on these data, family members should be screened with colonoscopy. If the index family member had cancer after age 60 years, screening should be initiated at age 50 years with colonoscopy and performed every 10 years if colonoscopy is negative. If the family member had cancer before age 60 years, screening should be initiated with colonoscopy at age 40 and then performed every 5 years. There is also evidence that family members of patients discovered to have adenomas before age 60 years may be at increased risk for CRC,³¹ although this risk may be increased only if the index family member had advanced adenomas.³²

Ulcerative Colitis and Crohn’s Colitis

Although patients with inflammatory bowel disease represent less than 1% of all patients who develop CRC, patients with colitis represent a high-risk group. Risk is strongly associated with extent and duration of disease. The risk is very low in the first 8 years of disease. However, in patients with pancolitis, the risk increases by up to 0.5% to 1% per year so that after 35 years of disease, the cumulative risk of CRC may be 35%.³³ Patients with severe long-standing disease and primary sclerosing cholangitis may have a higher risk of developing CRC.³⁴

Cancers that develop in patients with colitis are often flat and infiltrating, making endoscopic detection difficult. Current recommendations are to perform complete colonoscopy, beginning by year 8 of disease, and obtain four-quadrant biopsy specimens of otherwise flat mucosa at 10-cm intervals throughout the colon to detect dysplasia. Raised lesions require more intensive biopsy. There are data that patients with low-grade or high-grade dysplasia in flat mucosa have a high likelihood of having CRC, and colectomy is generally recommended. There is controversy regarding the approach to low-grade dysplasia. As patients age, they are likely to develop sporadic adenomas, which by definition are adenomas with low-grade dysplasia. Distinguishing a sporadic adenoma from a dysplasia-associated lesion or mass is often difficult. There is consensus that if a raised lesion is seen, a biopsy specimen of the lesion and surrounding flat mucosa should be obtained. If the surrounding mucosa is nondysplastic, it is assumed that the raised lesion is most likely a sporadic adenoma.

There are no randomized clinical trials to evaluate surveillance. There is some evidence that survival is better in patients enrolled in surveillance.³⁵ Recommendations to perform colonoscopy every 1 to 2 years for life are based on the steadily increasing cumulative risk of CRC with time. Patients with colitis limited to the rectum and left colon have a lower risk. Surveillance beginning at age 15 years is recommended.

Screening Strategies for Average-Risk Individuals

Screening recommendations for average-risk asymptomatic individuals vary around the world. In Asia, there are no specific guidelines. In Europe and Canada, most countries offer fecal occult blood test (FOBT), and some offer sigmoidoscopy (United Kingdom, Italy, Norway) or colonoscopy (Germany, Austria, Poland, Italy). In the United States, three guidelines ²^–⁵ were published in 2008 and 2009 (Table 35.5).

Table 35.5 U.S. Colorectal Cancer Screening Recommendations for Average-Risk Individuals

Cancer screening is traditionally focused on early cancer detection (i.e., breast, prostate). However, because most CRC develops from adenoma precursors, there is an opportunity to prevent many cancers if precursor lesions are detected and removed.¹⁶ More recent guidelines in the United States have preferred the use of tests that are likely to detect polyps and prevent cancer.^2,⁵

Early Cancer Detection Tests

Fecal Occult Blood Test

Three randomized controlled trials showed that among patients who undergo FOBT compared with unscreened controls, cancers are discovered at an earlier stage, and mortality is reduced by 15% to 33%.³⁶^–³⁸ Standard guaiac fecal occult blood test (gFOBT) detects peroxidase activity of heme and is not specific for human blood. One-time testing with a standard guaiac test detects only 33% to 50% of patients with cancer (Table 35.6). A more sensitive guaiac test (SENSA) can detect more than 60% of patients with cancer (see Table 35.6).^2,^4,³⁹ Three separate stool samples per test have superior sensitivity compared with one or two samples. Patients with positive tests have a threefold to fourfold increased risk of cancer and should be referred for colonoscopy.

Table 35.6 Sensitivity (One-Time Test) of Colorectal Cancer Screening Tests

	Cancer	Advanced Adenomas*
STOOL-BASED TESTS
Standard gFOBT^39,^42,⁴³	33%–50%	11%
Sensitive gFOBT^2,^4,³⁹	50%–75%	20%–25%
Fecal immunochemical test^3,⁵	60%–85%	20%–50%
Stool DNA—old⁴²	51%	18%
Stool DNA—new^43,⁴⁴	≥80%	40%
STRUCTURAL EXAMINATIONS OF COLON
CT colonography⁴⁸	Uncertain; likely >90%	90% if >10 mm
Sigmoidoscopy^55,^57,⁵⁸	>95% distal colon	70%^†
Colonoscopy⁴⁵^–⁴⁷	>95%	88%–98%

CT, computed tomography; gFOBT, guaiac fecal occult blood test.

* Advanced adenoma defined as tubular adenoma ≥10 mm; adenoma with villous histology or high-grade dysplasia.

^† Assumes that if an adenoma is detected in the distal colon, the patient would have complete colonoscopy, which would result in detection of some proximal advanced adenomas.

From Lieberman D: Screening for colorectal cancer. N Engl J Med 361:1179–1187, 2009.

Fecal immunochemical test uses antibodies specific to human hemoglobin, albumin, or other blood components and is more specific for human blood. There are many commercially available tests, most of which detect more than 50% of cancers with one-time testing. There are several uncertainties, including the number of stool samples, proper preservation of samples while awaiting testing, and the performance characteristics of the test in clinical practice.^39,⁴⁰ It is unknown if fecal immunochemical test is superior to a less costly sensitive gFOBT.

Fecal blood testing has many important limitations. The detection rate for advanced adenomas is only 25% to 30%. One-time testing fails to identify most patients with important cancer precursor lesions. An effective program requires annual repeat testing for negative tests and referral for colonoscopy if tests are positive. Adherence in clinical practice is uncertain.⁴¹ Although the initial cost of testing is low, efforts to ensure adherence to repeat testing, referral for colonoscopy, and the management of detected cancers contribute to the programmatic cost. Patients should understand the limitations of the test and the fact that many cancers and high-risk adenomas go undetected.

Stool DNA Test

Gene mutations associated with colonic neoplasia can be detected in stool samples. Proof of principle studies in a large screening cohort showed that more than 50% of cancers could be detected.⁴² New versions of the test are promising but have not yet been evaluated in screening cohorts.^43,⁴⁴ The stool DNA test fails to identify most patients with advanced neoplasia and should be viewed as an early cancer detection test.

Imaging Studies of the Colon

Barium enema accurately identifies late stage cancer but is a poor test for important cancer precursor lesions and is rarely used for colon cancer screening today. Imaging with computed tomography (CT) colonography has largely replaced barium enema. Helical CT scans render two-dimensional and three-dimensional images of the colon.⁴⁵^–⁴⁸ CT colonography can identify correctly 90% of patients with polyps 10 mm or larger with a false-positive rate of 14% when performed by properly trained radiologists.⁴⁸ Quality in other clinical practice settings is unknown.

CT colonography is a new technology with many uncertainties, as follows:

Who should be referred for colonoscopy? There is consensus that all patients with polyps 6 mm or larger should be offered colonoscopy.² The management of patients with polyps 1 to 5 mm is controversial. CT colonography cannot accurately identify polyps of this size. Less than 2% of these patients have adenomas with advanced features, and cancer is rare.^49,⁵⁰ No studies have shown the safety of following such patients with repeat CT colonography.

Detection of flat polyps with CT colonography is unknown, some of which may harbor malignancy.⁵¹

Appropriate screening intervals after negative examinations or with possible polyps smaller than 6 mm are uncertain.

Radiation exposure could increase the risk of developing cancers.⁵² Although low-dose regimens are used, there is concern about cumulative x-ray exposure, and some countries do not allow imaging for screening purposes.

The evaluation of extracolonic findings could be an important driver of cost. Although 27% to 69% of persons screened with CT colonography have some other finding outside of the colon, most studies estimate that 5% to 16% of findings may be significant and require further evaluation.^2,⁴

Cost-effectiveness is uncertain. Several decision models find that screening with CT colonography would be costly. Key determinants of cost include the frequency of referral for colonoscopy and rate of evaluation for extracolonic findings.

Flexible Sigmoidoscopy

Case-control studies have found significant reduction in CRC mortality in patients who have undergone sigmoidoscopy—a benefit limited to the portion of the colon examined.⁵³ One small randomized trial showed a reduction in CRC incidence in patients exposed to sigmoidoscopy.⁵⁴ Screening colonoscopy studies show that sigmoidoscopy would fail to identify 30% or more of individuals with advanced neoplasia. There is evidence that sigmoidoscopy may be less effective with advancing age (>60 years old) because of higher prevalence of proximal neoplasia.⁵⁵ Large studies of sigmoidoscopy in the United States, United Kingdom, and Italy are not yet completed. Despite evidence of effectiveness, use of flexible sigmoidoscopy in the United States has declined over the past 10 years, coincident with the growth of screening colonoscopy.

Combined Flexible Sigmoidoscopy and Fecal Occult Blood Test

Although several cost-effectiveness analyses endorse the use of combined FOBT (annually) and sigmoidoscopy (every 5 years), there is little clinical experience with combined screening tests. The VA Cooperative Study ⁵⁶ found that the combination of one-time sigmoidoscopy and FOBT would identify 76% of patients with advanced neoplasia, which represented only a small incremental improvement over sigmoidoscopy alone. There are no data regarding adherence to this complex screening program and no evidence that it is being used in clinical practice.

Colonoscopy

Colonoscopy is the most commonly used screening test in the United States. Worldwide experience has shown feasibility in different populations in North America, Asia, and Europe.^55,⁵⁷^–⁶⁰ Among patients age 50 to 75 years being screened for the first time, 5% to 10% have advanced neoplasia, and 0.5% to 1.0% have cancer. Evidence from the National Polyp Study showed that detection and removal of adenomas could result in substantial reduction in incidence of CRC.¹⁶ Case-control studies have shown that in patients exposed to colonoscopy, both CRC incidence and mortality are reduced.⁶¹^–⁶³

All of the current screening guidelines endorse the use of colonoscopy screening. Several important uncertainties remain, as follows:

Interval cancers may occur after colonoscopy. In studies of patients with adenomas who had colonoscopy and polypectomy, 0.3% to 0.9% of patients developed cancer within 2 to 5 years of colonoscopy. There are several possible reasons. New fast-growing lesions with microsatellite instability may account for some interval lesions. Incomplete removal of neoplasia at baseline could lead to development of cancer at a prior polypectomy site. However, lesions missed at baseline colonoscopy represent the most common explanation in most interval cancers. Studies using CT colonography to evaluate optical colonoscopy showed that colonoscopy may miss 2% to 12% of polyps 10 mm or larger.⁴⁵^–⁴⁷ Flat lesions (see Chapter 36) may be difficult to detect at colonoscopy or incompletely visualized without chromoendoscopy.

The 10-year interval after a normal colonoscopy is based on expert opinion derived from indirect data. Although case-control studies have suggested a 10-year benefit from endoscopic examinations of the colon, concerns about interval cancers noted previously have cast some doubt on whether a 10-year interval is appropriate for all patients. Studies of follow-up colonoscopy 5 years after a negative baseline screening examination reveal a low rate of advanced neoplasia.^64,⁶⁵

Quality in clinical practice is uncertain and may contribute to miss rates at colonoscopy. Elements of quality ⁶⁶ are summarized in Box 35.1 and include both reporting indicators (does the element appear in the report, and is it reported in a reproducible manner) and performance indicators (examination completeness, polyp detection and removal, recommendations for follow-up, and adverse events). A study by the Clinical Outcomes Research Initiative⁶⁷ showed variability in reporting (despite use of a computerized report generator) and key outcomes such as completeness of examinations and polyp detection rates. Studies have shown that spending more time examining the colon during colonoscope withdrawal results in higher rates of polyp detection.⁶⁸ Further study is needed to validate key quality indicators.

Box 35.1

Quality Indicators for Colonoscopy

Reporting Indicators (Documentation in the Report)

Procedure indication: Appropriate indication and interval

Precautions requiring interventions (anticoagulation, ventricular defibrillator)

Sedation medications

Documentation of bowel preparation quality

Completeness of examination with documentation of landmarks (photographic documentation preferred)

Description of key findings, polypectomy method, and polyp retrieval

Communication of follow-up plan

Performance Indicators

Bowel preparation quality

Cecal intubation rate

Time of procedure

Polyp detection rate

Polyp retrieval rate

Unplanned events and interventions

Appropriate follow-up recommendations based on findings and pathology

Patient satisfaction

From Lieberman D, Nadel M, Smith R, et al: Standardized colonoscopy reporting and data system (CO-RADS): Report of the Quality Assurance Task Group of the National Colorectal Cancer Roundtable. Gastrointest Endosc 65:757–766, 2007.

New innovations in colonoscopy may make colonoscopy easier to perform (self-advancing instruments), able to view behind folds (3rd Eye, expanded wide vision angle), or detect polyps more accurately than white light alone (chromoendoscopy, narrow band imaging, autofluorescence). Further advances may enable “microscopic” endoscopy, which can characterize histologic features at the time of endoscopy. These innovations require further research.

Surveillance after Detection of Neoplasia

Surveillance after detection and removal of colon neoplasia (Table 35.7) should be viewed as part of a screening program. Recommendations for surveillance have an enormous impact on screening program cost and resource use.

Table 35.7 Surveillance Recommendations

Findings at Baseline Colonoscopy*	Recommended Interval for Colonoscopy
No polyps	10 yr
Hyperplastic polyps: rectum-sigmoid	10 yr
1–2 tubular adenomas <10 mm	5–10 yr
≥3 tubular adenomas	3 yr
Tubular adenoma ≥10 mm	3 yr
Adenoma with villous histology	3 yr
Adenoma with high-grade dysplasia*	3 yr
Invasive cancer^†	1 yr
Incomplete removal of neoplastic lesion	3 mo

* Assumes complete examination to cecum, with preparation adequate to detect polyps >5 mm and complete removal of visible polyps.

^† Assumes complete examination of the colon before resection of cancer. If complete examination was impossible, colonoscopy should be performed at 3 months to rule out synchronous lesions.

Data from Rex DK, Kahi CJ, Levin B, et al: Guidelines for colonoscopy surveillance after cancer resection: A consensus update by the American Cancer Society and the US Multi-Society Task Force on Colorectal Cancer. Gastroenterology 130:1865–1871, 2006; and Winawer SJ, Zauber AG, Fletcher RH, et al: Guidelines for colonoscopy surveillance after polypectomy: A consensus update by the US Multi-Society Task Force on Colorectal Cancer and the American Cancer Society. Gastroenterology 130:1872–1885, 2006.

Surveillance in Patients with Personal History of Colorectal Cancer

After the diagnosis and treatment of cancer, guidelines recommend surveillance colonoscopy within 1 year of curative resection, and every 3 to 5 years thereafter if the first colonoscopy is negative. This recommendation was based on a literature review of 24 studies, which found that 57 of 137 metachronous CRCs were found within 24 months of resection.⁶⁹ If a complete examination of the colon was impossible before treatment because of obstruction, a colonoscopy should be completed within 3 to 6 months of surgical treatment to rule out synchronous lesions.

Adenoma Surveillance

The rationale for surveillance after removal of adenomas is based on several lines of evidence, but connecting the dots of evidence is problematic. We know that patients with adenomas have an increased risk of developing CRC. Most CRC has an adenoma precursor lesion. Numerous studies have shown that patients with adenomas who undergo colonoscopy and polypectomy have a high likelihood of having adenomas detected during follow-up examinations.⁷⁰ Some of these lesions may have been present at the baseline examination but may not have been detected or were incompletely removed, whereas others may represent new lesions. The rationale for adenoma surveillance is based on the premise that adenomas can develop into CRC and that after adenomas are removed at baseline colonoscopy, more adenomas are commonly found during follow-up examinations.

There are several flaws in this reasoning, however. In their lifetime, 30% to 50% of individuals develop adenomas and only 5% to 6% develop cancer. Most patients with adenomas do not develop cancer and would be unlikely to benefit from surveillance. How should patients who are likely to progress to malignancy (and benefit from surveillance) be identified? Risk stratification based on patient characteristics and the index polyp would be ideal. There is some preliminary evidence that patients who have only one or two small tubular adenomas are at low risk for development of new advanced adenomas during 5 to 6 years of follow-up.⁵⁵ Patients with multiple adenomas, large adenomas (≥10 mm), or adenomas with villous histology have a higher risk of having advanced adenomas during surveillance.^55,^70,⁷¹ These data informed recommendations for adenoma surveillance, which are summarized in Table 35.7.⁷²

Complications of Screening and Surveillance

Colon screening with fecal testing is not associated with direct adverse events. It may be argued, however, that failure to detect cancer if present results in delays in cancer diagnosis, which may adversely affect outcomes. Imaging studies with CT colonography or barium are rarely associated with adverse events in screening examinations. Perforations are uncommonly reported but have generally occurred in symptomatic patients. Sigmoidoscopy is also quite safe and rarely associated with perforation or bleeding. CT colonography and sigmoidoscopy are associated with some discomfort.

If the initial screening test is positive, all screening programs ultimately lead to colonoscopy. The safety of colonoscopy in diverse practice settings is uncertain. The overall rate of adverse events is 3 to 5 per 1000 colonoscopies.⁷³^–⁸³ The clinical studies summarized in Table 35.8 have reported a limited set of adverse events, such as perforation (0.1 to 1 per 1000) and bleeding (0.5 to 6 per 1000). Most of these studies have included patients undergoing screening and diagnostic colonoscopy. Hospitalizations for other episodes such as cardiopulmonary events may occur in 1 in 1000 procedures and have been reported in a few studies. Few studies have fully evaluated events occurring within 30 days of colonoscopy.

Table 35.8 Serious Adverse Events of Colonoscopy: Studies from 2002–2009

Controversies in Screening and Surveillance

Age to Stop Screening

The 2008 U.S. Preventive Services Task Force guideline suggested that routine screening should stop at age 75 years and that no screening should be performed after age 85 years. This recommendation is based on a decision analysis that suggested benefit (prevention of colon cancer death) is reduced in elderly individuals, and potential for harm is increased. Nevertheless, CRC incidence and mortality continue to increase with advancing age. Many experts recommend that screening decisions be individualized based on patient comorbidity, likelihood of long-term survival, and a careful assessment of the risk of colonoscopy.

Customization of Screening

Women develop CRC and advanced neoplasia at a later age than men. The delay in onset of colon neoplasia may be related to natural hormonal protection before menopause or hormone replacement therapy after menopause. African Americans have a higher CRC incidence and mortality compared with whites ¹ and have higher age-adjusted cancer precursor lesions compared with whites.⁸⁴ Current guidelines call for all men and women to be screened beginning at age 50 years, although one expert group advocated initiation of screening at age 45 for African Americans. There is a rational basis for considering customization of screening for women (delay initiation) and for African Americans (early initiation). However, there is also concern that current screening guidelines are complex and that adding another layer of complexity with customization could paradoxically result in lower rates of screening.² The current guidelines place emphasis on ensuring that all individuals have access to screening.