Epithelial Neoplasms of the Large Intestine

Published on 30/06/2015 by admin

Filed under Pathology

Last modified 30/06/2015

Print this page

rate 1 star rate 2 star rate 3 star rate 4 star rate 5 star
Your rating: none, Average: 0 (0 votes)

This article have been viewed 1432 times

Chapter 27

Epithelial Neoplasms of the Large Intestine

Mark Redston

David K. Driman


The most common neoplasms of the large intestine are adenomas (see Chapter 22). Adenomas are the precursor of most primary malignant epithelial neoplasms of the large intestine. Although abundant clinical, morphologic, and genetic evidence suggests that primary epithelial malignant neoplasms are a heterogeneous group of tumors, most clinicians consider these neoplasms together to be “colorectal carcinoma.”1,2 Much of the discussion in this chapter refers to colorectal carcinoma (CRC) as a generic, single disease, with the recognition that this is an oversimplification. In practice, approximately 85% of CRCs are typical adenocar­cinomas; relatively distinct histologic subtypes form the remainder (Box 27.1). Recent advances in genetic testing for hereditary colorectal cancer syndromes and for sensitivity to targeted therapies also highlight the clinical relevance of emerging molecular classification systems.

Box 27.1

World Health Organization Classification of Epithelial Neoplasms of the Large Intestine

Dysplasia (Intraepithelial Neoplasia)

Modified from Bosman FT, Carneiro F, Hruban RH, Theise ND, eds. WHO Classification of Tumours of the Digestive System. Vol 3. In: Bosman FT, Jaffe ES, Lakhani SR, Ohgaki H, eds. World Health Organization Classification of Tumours. 4th ed. Lyon, France: IARC; 2010.

Clinical Features

Early-stage CRC is typically diagnosed only at the time of screening or other, unrelated colonoscopy and does not usually manifest with symptoms, signs, or other laboratory findings. Advanced cancers are more likely to result in clinical symptoms, including a change in bowel habits, constipation, abdominal distention, hematochezia, or tenesmus (rectosigmoid lesions). It is important to appreciate that these “colorectal-type” symptoms and signs apply predominantly to left-sided cancers; right-sided cancers often manifest insidiously with nonspecific systemic symptoms and signs such as fatigue, weight loss, and anemia. Only approximately 40% of patients have localized disease at presentation. Approximately 40% have regional metastases, and approximately 20% have distant metastases.3 Endoscopy with biopsy is the standard diagnostic approach. Computed tomography and magnetic resonance imaging (MRI) are used to assess depth of invasion, regional spread, and distant metastases; rectal MRI is the gold standard to assess the extent of local spread in rectal cancers, with transrectal ultrasound an alternative for early lesions.

A variety of screening recommendations have been proposed and endorsed by the American Gastroenterological Association, American Medical Association, and American Cancer Society (Table 27.1). Standard guidelines are modified in individuals with a personal or family history of colorectal adenoma or carcinoma.4 Screening is clinically effective and cost-effective, yet its use remains relatively low.5 Screening colonoscopy provides the added benefit of polyp removal, and it is well established that polypectomy can prevent CRC.6 An estimated 90% of CRC deaths could potentially be prevented by implementing a combination of (1) strategies aimed at improving screening, polyp management, and early diagnosis of CRC; (2) lifestyle modifications, including dietary change and increased exercise; and (3) chemoprevention.3,79

Table 27.1

Guidelines for Early Detection of Adenomas and Colorectal Cancer in Average-Risk Individuals

Method Interval*
Tests That Detect Adenomatous Polyps and Cancer
Flexible sigmoidoscopy Every 5 yr
Colonoscopy Every 10 yr
Double-contrast barium enema Every 5 yr
Computed tomography colonography Every 5 yr
Tests That Primarily Detect Cancer
Guaiac-based fecal occult blood test Annual
Fecal immunochemical test Annual
Stool DNA test Interval uncertain

* Beginning at age 50 yr.

 These tests should be encouraged if resources are available and the patient is willing to undergo an invasive procedure.

From Levin B, et al. Screening and surveillance for the early detection of colorectal cancer and adenomatous polyps, 2008: a joint guideline from the American Cancer Society, the US Multi-Society Task Force on Colorectal Cancer, and the American College of Radiology. Gastroenterology. 2008;134:1570-1596.


Malignant epithelial tumors of the colon and rectum are the fourth most common cancer in men worldwide (after lung, prostate, and stomach), and the third most common in women (after breast and uterine cervix), accounting for 9.4% of all cancers in 2008, with more than 1 million new cases diagnosed each year.3,10,11 There is marked variation in the age-standardized incidence, with a 25-fold difference between high-risk regions (affluent countries including Australia, New Zealand, Europe, the Americas, and Japan) and low-risk regions (developing countries including Africa, India, and other parts of southeast Asia) (Fig. 27.1).11 The likely role of environmental and lifestyle influences, particularly diet, alcohol intake, and physical activity, in the genesis of these differences is supported by abundant data. There are also significant global differences in the age at onset of CRC, with a mean age of only 50 years in developing countries.

In the United States, there were an estimated 103,170 new cases of colon cancer and 40,290 new cases of rectal cancer in 2012.10 CRC is the third most common cancer in both men and women and the fourth most common cancer overall, representing about 9% of all cancers.3 Overall, it is the second leading cause of cancer death, behind only lung cancer, and it is the leading cause of cancer death among nonsmokers.3 The lifetime risk for development of CRC is estimated at about 5%.3,12 U.S. Surveillance, Epidemiology, and End Results (SEER) Program statistics reveal an incidence of 30.45 per 100,000 for colon carcinoma and 12.16 per 100,000 for rectal carcinoma in 2009.4,13 CRC is significantly more common in men (combined age-adjusted incidence, 54.0 per 100,000, versus 40.2 per 100,000 in women); this difference is more striking for rectal cancer than for colonic cancer, 3,4 and the increased incidence in men is apparent only after the age of 50 years. Despite the higher incidence in men, women live longer, so there are a similar number of total cases and cancer deaths in men and women.3 Since 1987, the incidence of colon cancer in the United States has been slowly falling.14 The incidence increases with age, with approximately 9% of cases occurring before age 50 years and only approximately 1% before age 35 years.3,4

In addition to the global variation in CRC incidence, there are significant regional and ethnic differences in incidence within the United States. The incidence varies by approximately 1.5-fold between high-risk regions (predominantly the northeast Atlantic coast) and low-risk regions (predominantly the South and Midwest).3 Table 27.2 presents the differences related to racial and ethnic backgrounds. The incidence is highest among African Americans and lowest among individuals of Native American origin.3

Table 27.2

Incidence and Mortality Rates of Colorectal Cancer by Race and Ethnicity, United States, 2004-2008*

Race/Ethnic Group Incidence Mortality
Male Female Male Female
White 54.6 40.3 20.1 14.0
African American 66.9 49.7 30.5 20.4
Asian American/Pacific Islander 42.4 32.7 13.3 9.9
American Indian/Alaska Native 51.5 41.5 19.8 14.0
Hispanic/Latino 48.6 34.2 15.5 10.3


* Rates per 100,000, age-adjusted to the 2004 U.S. standard population.

From American Cancer Society. Cancer Facts and Figures 2012. Atlanta: American Cancer Society; 2012:44.


The risk of CRC is influenced by both endogenous (constitutional) and exogenous (environmental) factors (Table 27.3).15 For the practicing surgical pathologist, genetic predisposition and long-standing inflammatory bowel disease (IBD) have the most direct clinical impact, and these topics are discussed later. Age, as discussed previously, is the most powerful risk factor. CRC is predominantly a disease of late middle-aged and elderly individuals.3 The increased risk in males is thought to be related to differences in hormonal milieu.16

Table 27.3

Risk Factors for Colorectal Cancer*

Factor Relative Risk
Family history (first-degree relative) 1.8
Physical inactivity (<3 hr/wk) 1.7
Inflammatory bowel disease (physician-diagnosed Crohn’s disease, ulcerative colitis, or pancolitis) 1.5
Obesity 1.5
Red meat 1.5
Smoking 1.5
Alcohol (>1 drink/day) 1.4
High vegetable consumption (≥5 servings/day) 0.7
Oral contraceptive use (≥5 yr) 0.7
Estrogen replacement (≥5 yr) 0.8
Multivitamins containing folic acid 0.5

* Modifiable factors are in bold text.

From American Cancer Society. Cancer Facts and Figures 2002. Atlanta: American Cancer Society; 2002:20. [Data from Colditz et al., 2000.15]

Remaining risk factors are largely related to lifestyle and, importantly, are modifiable, suggesting the potential for interventions aimed at significantly reducing the incidence of CRC.17 The five most convincingly implicated lifestyle factors are obesity, physical activity, and ingestion of red meat, processed meat, and alcohol.18 Of these modifiable risk factors, diet has been the most extensively studied, and although there is little doubt that elevated risk is consistently associated with a “Western” type of diet, it has been difficult to determine which components are most important. Diets with a high calorie intake and those rich in meat, particularly animal fat, have been implicated in many studies.1921 Possible mechanisms for this effect include the production of heterocyclic amines, stimulation of higher levels of fecal bile acids, production of reactive oxygen species, and elevated insulin levels.22,23 In addition to high-risk factors, there are inverse associations with vegetable and fiber consumption, although fiber falls out of some multivariate analyses when adjusted for other dietary risk factors.2426 This effect could be related to anticarcinogens, antioxidants, folate, induction of detoxifying enzymes, binding of luminal carcinogens, fiber fermentation to produce volatile fatty acids, or reduced contact time with epithelium because of faster transit.24,25 Several studies, including a large pooled multivariate analysis, have found that high folate intake is associated with a decreased risk of CRC, providing some of the most direct evidence of dietary risk factor relationships.27,28 Finally, alcohol intake has been associated with an increased risk of CRC.29

There is an inverse association between use of nonsteroidal antiinflammatory drugs and CRC risk.7 Smoking exposure is associated with CRC, although the relative risk is less than for many other tobacco-related cancers,30 and some studies consider cigarette smoking only a suggestive risk factor.18 Sedentary lifestyle,31 long-standing IBD32 (see Colitis-Associated Neoplasia), pelvic irradiation,33 and ureterosigmoidostomy are also associated with an increased risk of CRC.

Finally, there is evidence that there are important differences in the epidemiologic risk factors associated with different subtypes of CRC. There has been a trend in recent years toward the development of more proximal cancers,13 which may relate to changes in epidemiologic risk factors. In fact, most of the lifestyle factors are associated specifically with increased risks of colon cancer, and only age and gender are predictive of an increased risk of rectal cancer.34 Furthermore, there are molecular biologic differences between right- and left-sided CRCs that would support different epidemiologic associations.35 More recently, research has focused on the complex interactions of hormones, energy balance, intestinal flora, and inflammation.21

Genetic polyposis syndromes account for fewer than 0.5% of all incident CRCs (Table 27.4). Nonpolyposis forms of hereditary colorectal carcinoma have a much higher overall contribution to the causation of CRC and are discussed in detail later in this chapter.


Progression from Adenoma to Carcinoma

Most, if not all CRCs arise from adenomas, either conventional adenomas, sessile serrated adenomas/polyps (SSA/Ps), or traditional serrated adenomas (in decreasing order of frequency).36,37 Residual adenoma is identified in approximately 10% to 30% of CRCs; in the remainder, the adenomas are presumably overgrown by cancer.38 There are distinct associations between the histologic type of precursor lesion and the subsequent pathogenesis of the derived CRC. These indicate that there are two broad pathways involved in neoplastic progression in the colorectum: the conventional adenoma pathway and the serrated adenoma/polyp pathway.

The conventional adenoma pathway accounts for approximately 70% to 80% of all CRCs and is much more prevalent in the left colon and rectum than in the right colon. Conventional adenomas typically precede cancer by approximately 15 years.39 The prevalence of conventional adenomas in the U.S. population is approximately 25% by age 50 years, and 50% by age 70 years, and these adenomas have a high lifetime risk of progression if not removed.40 Exact risks of progression are not known; however, one study estimated a 10% to 15% chance of progression during 10 years for a 1-cm conventional adenoma.41 Endoscopic removal of conventional adenomas decreases the incidence of subsequent CRC in treated patients.6,42

The serrated pathway has been increasingly recognized in the past 10 years, and it is estimated to account for approximately 20% to 30% of all CRCs.4347 Most CRCs arising in the serrated pathway develop from SSA/Ps, particularly those located in the right colon. CRC is typically preceded by the development of foci of true dysplasia within these polyps.45,48,49 Historically, conventional endoscopic screening programs have been less effective at reducing right-sided CRC, in large part because the risk of progression of serrated polyps, almost all of which were previously diagnosed as hyperplastic polyps, was not recognized.50 The exact risk of progression of these polyps is yet to be fully characterized, although the malignant potential appears sufficient to warrant complete endoscopic removal.37 Challenges remain for endoscopic surveillance programs, because serrated polyps are difficult to recognize and completely remove endoscopically,51 and because the underlying genetic instabilities that drive these lesions have the propensity to result in CRCs in relatively small polyps (see later discussion). Traditional serrated adenomas may also precede CRC; however, these lesions are much less common, and the risk of progression appears to be lower. Conventional hyperplastic polyps, particularly those of the left colon, rarely, if ever, progress to cancer.

Aberrant crypt foci represent the earliest stages of colorectal neoplasms and are present before the development of grossly apparent adenomatous polyps.39 Aberrant crypt foci are microscopic lesions most readily identified by examination of methylene blue–stained, stripped mucosal sheets under a dissecting microscope. They are characterized by a localized collection of crypts that show an increase in crypt diameter and an increased number of lining epithelial cells, which imparts a serrated or slitlike appearance. Histologic sections of aberrant crypt foci reveal a range of findings, such as normal or only mildly hyperplastic epithelium, features more typical of serrated polyps, or, rarely, true dysplasia (the latter being similar to microscopic adenomas incidentally identified in patients with familial adenomatous polyposis).52 Aberrant crypt foci may also be visualized endoscopically, although the technical challenges of these methods have prohibited routine clinical applications.53

Reports of very small (<1 cm) carcinomas that lack any evidence of residual adenoma have raised the possibility that some cancers may arise de novo.54 These cancers represent fewer than 5% of all CRCs. Some studies suggest that they are more likely to be of higher grade than ex-adenoma carcinomas, with a higher risk of lymphatic and blood vessel invasion.54 However, other studies do not support that hypothesis. Hornick and colleagues report that small carcinomas without a dysplastic component shared clinical and molecular characteristics with small carcinomas containing a minimal dysplastic component and with conventional larger carcinomas.55 It is also possible that some of these lesions represent rapid progression from a small adenoma to cancer because of the early acquisition of high-grade genetic alterations (e.g., aneuploidy, TP53 mutations).54

Genetic Model of Colorectal Cancer Progression

Human cancers are characterized by an accumulation of a variety of genetic alterations, including mutations that either activate oncogenes or inactivate tumor suppressor genes.5658 This accumulation of genetic alterations is a critical event in the progression from adenoma to carcinoma and likely begins in aberrant crypt foci and other precursor cells that may not manifest morphologic features of neoplasms.5861 To accumulate the array of genetic alterations typical of most CRCs, tumor cells must acquire mutations and epigenetic alterations at an increased rate compared with normal crypt epithelial cells.62 Increased acquisition and tolerance of mutations is a hallmark of CRC development, and is referred to as genome instability.63,64 Genes involved in the maintenance of the genome have been likened to “caretakers” of the genome.65 There are at least two general forms of genome instability important to the development of colorectal neoplasia: chromosomal instability (CIN) and microsatellite instability (MSI).66 The third major driving pathway for the development of CRC is the widespread accumulation of epigenetic gene silencing (CpG island methylator phenotype, or CIMP).43,66,67

Chromosomal Instability.

CIN is characterized by a persistently increased rate of gains and losses of chromosomal material; it is present in 70% to 80% of CRCs.68 The acquisition of abnormalities in the number of whole chromosomes results in aneuploidy. In addition to whole-chromosome abnormalities, CRCs have other forms of somatic copy number abnormalities, including abnormalities of whole chromosomal arms as well as focal gains and losses. The underlying genetic basis of aneuploidy in human cancers is poorly understood, although most studies have focused on genes involved in regulation of mitotic spindle assembly and segregation. CIN is a major underlying genetic aberration in the conventional adenoma–carcinoma progression pathway and is therefore the predominant form of instability in left-sided neoplasms.66,69

Microsatellite Instability.

MSI is characterized by widespread alterations in the sizes of repetitive DNA sequences. It is present in approximately 10% to 15% of CRCs.70 MSI is caused by defective DNA mismatch repair (MMR) (Fig. 27.2) (see Hereditary Nonpolyposis Colorectal Cancer). In addition to alterations in the sizes of repetitive DNA sequences, MSI results in markedly increased rates of mutation of coding sequences (somatic hypermutation). In general, CRCs with MSI do not harbor abnormalities in chromosomal number or the focal regions of subchromosomal gains or losses that typify cancers with CIN. In most CRCs with MSI, the underlying defect in MMR function is caused by epigenetic CpG island hypermethylation of the promoter region of the MLH1 gene. This is a characteristic feature of many CRCs arising in the serrated neoplastic pathway, and most of these cancers are high-frequency CIMP (discussed later).44,71 MSI is also the mechanism that underlies the progression of Lynch syndrome cancers, which are caused by inherited defects in DNA MMR (see later discussion). In Lynch syndrome, MSI develops in conventional adenomas and drives rapid progression to cancer.

CpG Island Methylator Phenotype.

CpG island methylator phenotype (CIMP) is the acquisition of widespread hypermethylation of CpG dinucleotides in the promoter regions of genes.45,72,73 Referred to as an epigenetic alteration (because it does not change the DNA sequence), this is a major mechanism of inactivation of tumor suppressor genes such as TP16, CDHI, and MLH1. Widespread CpG island methylation in a single cancer stands in stark contrast to the very limited methylation silencing that occurs in most CRCs and is known as high-frequency CIMP (CIMP-H).67,70,74 CIMP-H is a characteristic feature of CRCs that arise in the serrated pathway, and it is present in 20% to 30% of CRCs, including almost all cancers that also have MLH1 hypermethylation silencing. The underlying genetic basis of the CIMP-H phenotype is poorly understood, but there is evidence that genetic factors and environmental exposures (e.g., smoking, estrogen withdrawal) may be associated with the development of serrated pathway carcinomas.45 A working hypothesis is evolving wherein genetic and epidemiologic factors contribute to abnormal methylation events in serrated polyps of the right colon, which predisposes them to methylation silencing of MLH1, MGMT, and other important genes. Presumably, these events drive progression from dysplasia to adenocarcinoma43,7577 (Table 27.5). The subsequent carcinomas that develop are referred to as “serrated” adenocarcinomas by some authorities (see later discussion), and they are often found to be MSI-H or CIMP-H, or both, on molecular phenotyping.78,79

Implications of Progression Pathways for Screening and Prevention.

Screening colonoscopy is more effective for the prevention of left-sided than right-sided CRC, and this difference is thought to be caused by the predominance of the conventional adenoma–carcinoma progression pathway in the left colon compared with the serrated adenoma–carcinoma progression pathway in the right colon.37,50,80 The failure of screening to prevent right-sided colon cancer to the same degree as left-sided cancer appears to be related to the difficulties that exist in the endoscopic identification of SSA/Ps as well as their particular molecular pathogenetic characteristics. Serrated polyps are more difficult to completely excise, which could directly lead to colonoscopic screening failures.51,81,82 Serrated polyps are also at greater risk of undergoing rapid progression in a relatively small lesion, secondary to the acquisition of DNA MMR deficiency and MSI.37 Indeed, studies of interval cancers have found that they are much more likely to exhibit MSI and CIMP-H, both features of CRCs arising in the serrated pathway.83,84

Molecular Pathway Classification of Colorectal Cancer.

A molecular-pathologic classification of CRC has been proposed that is based, in large part, on the presence or absence of CIN (i.e., CIN− versus CIN+; diploid versus aneuploid), the MMR status (microsatellite stable [MSS] versus MSI), and the status of the methylation pathway (CIMP+ versus CIMP−)69,75,85 (see Table 27.5). Although the clinical relevance of the aneuploid and (MMR)-deficient pathways has been delineated (see Prognostic Factors), the possible clinicopathologic associations of the methylator pathway have not been as well defined.

WNT Signaling Abnormalities.

APC was first identified as the gene mutated in individuals with familial adenomatous polyposis.86 In this syndrome, affected individuals inherit one mutant copy of APC that is functionally inactive. In tumorigenesis, the second allelic copy of the APC gene is also inactivated, which fulfills Knudsen’s paradigm for tumor suppressor gene inactivation. APC mutations are also present in 70% to 80% of sporadic CRCs, where mutations occur early during neoplastic development, and even in dysplastic aberrant crypt foci.58,86 The apparent necessity of APC inactivation for the development of early adenomas has resulted in its designation as a “gatekeeper” gene of colorectal neoplasia.65

APC is an important component of the Wingless (WNT) pathway, an evolutionarily conserved signaling cascade that is critical to embryonic development and intestinal epithelial renewal.86,87 APC mutations abrogate its role in binding β-catenin, thereby releasing β-catenin from phosphorylation regulation by glycogen synthase kinase-3β GSK3β and allowing it to accumulate in the nucleus, where it is involved in activating transcription of a number of other downstream targets, such as cyclin D and Myc.86 Recent microarray studies suggest that the transcriptional profile of β-catenin activation resembles the program of intestinal crypt stem cells. APC is also involved in cytoskeletal interactions, and, although the most important potential role in tumorigenesis has not been precisely defined, APC has been directly implicated in cell–cell adhesion, migration, chromosomal segregation (genome stability), and apoptosis.58,86

Further evidence of the importance of WNT signaling abnormalities in colorectal tumorigenesis is evidenced by the presence of mutations in the β-catenin gene (CTNNB1) in some tumors that lack APC mutations.58 In contrast to the inactivating APC mutations, the CTNNB1 mutations target amino acid residues integral to phosphorylation, resulting in oncogenic activation of WNT signaling.58

Buy Membership for Pathology Category to continue reading. Learn more here