Pediatric Colorectal Tumors

Updated: Jun 21, 2016
  • Author: Jaime Shalkow, MD, FACS; Chief Editor: Robert J Arceci, MD, PhD  more...
  • Print


At least 50% of the Western population will develop a colorectal tumor by age 70 years. In 10% of these individuals, the tumor progresses to malignancy. In adults, colorectal cancer is the second leading cancer that causes death worldwide. [1] This article discusses the classification, etiology, genetics, clinical presentation, and management of colonic tumors in children. These include polyps, sporadic colorectal carcinoma (CRC) and familial colon cancer (familial adenomatous polyposis [FAP]), and hereditary nonpolyposis colorectal cancer (HNPCC).

See the images below.

This picture depicts an abdominal CT scan of a 7 y This picture depicts an abdominal CT scan of a 7 year-old boy with a mucinous adenocarcinoma of the ascending colon. Note the thickness and increased vascularity of the colonic wall, as well as irregularities on the serosal surface. This cut also shows severe tumor infiltration of the colonic mesentery surrounding the mesenteric and retroperitoneal vessels.
Coronal CT scan demonstrating the profuse tumoral Coronal CT scan demonstrating the profuse tumoral infiltration of the ascending colonic mesentery surrounding mesenteric and portal vessels. Also note the thickness of the colonic hepatic flexure.
Surgical specimen after right hemicolectomy, inclu Surgical specimen after right hemicolectomy, including the terminal ileum up to the transverse colon. Mesenteric fat, vessels and lymph nodes were resected en block with the ascending colon. The large intestine has been opened longitudinally. Note the tumor on the right lower quadrant of the image, with severe thickness of the wall, areas of necrosis and hemorrhage, and some stippled calcifications.

See Benign or Malignant: Can You Identify These Colonic Lesions?, a Critical Images slideshow, to help identify the features of benign lesions as well as those with malignant potential.


Polypoid Disease of the Gastrointestinal Tract

Not all polyposis syndromes are familial. Familial polyposis syndromes are divided into 2 major groups based on the presence of adenomas or hamartomas. The inherited adenomatous polyposis syndromes include familial adenomatous polyposis (FAP) and Turcot syndrome; the familial hamartomatous polyposis syndromes include Peutz-Jeghers syndrome and juvenile polyposis.

Although juvenile polyps are common in children, adenomas are quite unusual. The latter are considered dysplastic precancerous lesions that are commonly seen in late adulthood. When discovered in children, they suggest one of several types of inherited colorectal cancer. [2]

Although the nomenclature is confusing, diffuse juvenile polyposis differs from juvenile polyposis coli. Diffuse juvenile polyposis is a syndrome with multiple polyps spread throughout the GI tract and presents in younger children (aged 6 months to 5 years); in juvenile polyposis coli, the polyps are confined to the rectosigmoid area and are typically found in older patients (aged 5-15 y). Hamartomatous polyps may also be found in patients with Cowden disease, Cronkhite, Bannayan-Riley-Ruvalcaba syndrome, and basal cell nevus syndrome. [3]

Colonic polyposis syndromes

See the list below:

  • Nonfamilial polyposis - Isolated juvenile polyps (inflammatory polyps)
  • Familial polyposis - Adenomas (FAP, Gardner syndrome, Turcot syndrome), hamartomas (juvenile polyposis, Peutz-Jeghers syndrome, Cowden disease, Cronkhite-Canada syndrome)

The lesions can be isolated to the intestine (eg, juvenile, lymphoid, familial adenomatous) or can involve other areas of the body (eg, Peutz-Jeghers syndrome, Gardner syndrome, Turcot syndrome). Most polyps of the GI tract are benign and result from hamartomas of the mucosa or lymphoid hyperplasia of the submucosal layer. However, adenomatous polyps represent a genetic alteration in the mucosa and have substantial malignant potential.

For study purposes, only the hamartomatous lesions and other nonfamilial lesions are discussed in this section. FAP is presented in detail below, with other cancer-predisposing entities.

Polyps occur in 1% of preschool-aged and school-aged children [4] and are the most frequent cause of rectal bleeding in toddlers and infants aged 2-5 years. Juvenile polyps are the most common (80%), followed by lymphoid polyps (15%).

Juvenile polyposis syndromes are classified as follows: [5]

Isolated juvenile polyps (nonmalignant) involves no family history of juvenile polyposis and fewer than 5 polyps confined to the colon.

Juvenile polyposis syndromes (malignant potential) are as follows:

  • Diffuse juvenile polyposis of infancy - Widespread polyposis of the entire GI tract in patients younger than 6 months.
  • Diffuse juvenile polyposis - Multiple polyps throughout the GI tract but concentrated in the stomach, distal colon, and rectum; usually occurs in patients aged 6 months to 5 years
  • Juvenile polyposis coli - Multiple polyps confined to the distal colon and rectum in patients aged 5-15 years

Lymphoid polyps (lymphoid nodular hyperplasia)

Lymphoid polyps (present in 15% of patients) are hyperplastic submucosal lymphoid aggregates, most likely due to a nonspecific infection (exposure to bacteria and viruses). Submucosal lymphoid tissue is prominent in children, particularly in the distal ileum (Peyer patches). These non-neoplastic polyps may occur in the rectum, colon, and terminal ileum. Macroscopically, they appear as firm, round, submucosal nodules that are smooth or lobulated. They are never pedunculated. They often have a volcano-like appearance with mucosal ulceration, which leads to occult blood loss. Histologically, they are hyperplastic lymphoid follicles with a large germinal center covered by colonic mucosa. They develop in young children, with a peak incidence at age 4 years.

Patients present with anemia or, less frequently, with severe rectal bleeding. Barium enema and colonoscopy findings are helpful (in 50% of patients), and biopsy findings confirm the diagnosis.

Surgery is indicated only for uncontrolled bleeding and intussusception that does not respond to enema treatment. Otherwise, expectant measures are adequate because these polyps are benign and spontaneously regress.

Isolated juvenile polyps

Also known as retention, inflammatory, or cystic polyps, isolated juvenile polyps are the most common types of polyps found in children (80%) and represent one of the most common sources of lower GI bleeding in this population. They are considered hamartomas and lack malignant potential. [6, 7] Juvenile polyps occur in approximately 1% of preschool-aged children. The peak incidence is in children aged 3-5 years; boys seek medical attention twice as often as girls do. [7] The polyps are solitary in 50% of patients; the remaining patients have 2-5 polyps. Approximately 40-60% of polyps are found in the rectosigmoid area; the remaining polyps are distributed throughout the proximal colon. [4] Isolated juvenile polyps are rarely seen after adolescence.

These polyps are smooth, reddish, and range from 2 mm to several centimeters in diameter. They often have an ulcerated surface, which accounts for the rectal bleeding. On cross-section analysis, cystic spaces filled with mucus are revealed. Some data suggest that that these polyps may result from structural rearrangement of the mucosa secondary to an inflammatory process. [8]

Patients most often present with hematochezia due to superficial ulceration of the polyp (93%), pain (10%), or rectal and/or polyp prolapse and encopresis. [4] Ten percent of juvenile polyps autoamputate with spontaneous cessation of rectal bleeding. [5, 7, 9] Colonoscopy is performed to eliminate juvenile polyposis (ie, >5 polyps). [10] The polyps can be endoscopically removed. When managing a prolapsed polyp, controlling the polyp stalk prior to resection is mandatory. Failure to control the polyp stalk can result in retraction, which makes hemostatic control very difficult. [7]

Juvenile polyposis syndromes

Diffuse juvenile polyposis of infancy

This entity occurs within the first months of life and is not familial. [11] Patients may present with diarrhea, rectal bleeding, intussusception, prolapse, bowel, protein-losing enteropathy, macrocephaly, clubbing of fingers and toes, and hypotonia. [6]

The entire GI tract is involved. One third of these patients have other congenital abnormalities such as Meckel diverticulum, malrotation, and heart lesions. [5]

Patients require total parenteral nutrition (TPN) and bowel rest, followed by selective resection.

Despite appropriate treatment, this disease is almost universally fatal; only 2 patients have been reported to survive after age 2 years. [6, 12]

Diffuse juvenile polyposis

Diffuse or familial juvenile polyposis was originally identified as isolated or multiple hamartomatous polyps that occur in the colon and rectum of children aged 6 months to 5 years. [13]

Patients present with bright red blood per rectum, anemia, abdominal pain, and rectal prolapse. Diffuse juvenile polyposis is inherited as an autosomal dominant trait; [5] thus, if a parent has the condition, the chance of having an affected child is 50%.

Hamartomas are malformed colonic mucosa arranged in a bizarre fashion. Typically, these are not considered premalignant unless they are part of a polyposis syndrome.

Patients with diffuse juvenile polyposis have a 50% lifetime risk of colorectal carcinoma (CRC). [14] This may be due to chronic inflammation that produces reactive hyperplasia, which then progresses to dysplasia or adenomatous changes. These polyps often have an ulcerated surface and demonstrate more epithelium with a villous or papillary configuration.

In addition to the aforementioned epithelial dysplasia occurring in juvenile polyps, adenomas are also often present. Thus, the approach to these patients is similar to that taken in patients with FAP. Some authors recommend monitoring these patients with an annual CBC count (to detect anemia due to GI bleeding), semiannual colonoscopy, and subsequent colectomy if severe dysplasia, bleeding, or rapid polyp formation occurs. Others advocate for prophylactic colectomy.

Associated congenital defects include cleft palate, malrotation, polydactyly, and cranial abnormalities.

Juvenile polyposis coli

A child with 3-10 colonic polyps, any number of polyps in GI tract outside of the colon, or one polyp and a family history of juvenile polyposis is considered to have the syndrome. [1]

Most patients have 50-100 colorectal polyps; they may also have gastric and small intestinal polyps.

Identifying patients with this syndrome is fundamental because of the high risk for carcinoma (17%) at an early age; the mean age at diagnosis of carcinoma is 35.5 years. [11]

Close long-term surveillance is important. The amount of polyps increases the risk of chronic bleeding, which subsequently leads to iron deficiency anemia, hypoproteinemia, and failure to thrive. [7, 12]

Macroscopically, these polyps resemble the isolated juvenile polyps; however, histologically, they have more epithelium with a villous or papillary configuration. Epithelial dysplasia can occur. Adenomas can also be found in conjunction with juvenile polyps. [6] Lobular polyps have a higher propensity for a more severe dysplasia (47%) than non-lobular polyps (10%). [15]

According to the St. Mark's Polyposis Registry in London, the cumulative risk for cancer in patients with a juvenile polyposis syndrome is 68% by age 60 years. [16] Because the entity is transmitted in an autosomal dominant fashion, patients with a juvenile polyposis syndrome and their families must receive long-term follow-up. [17]

Some authors advocate prophylactic total colectomy and rectal mucosectomy with an endorectal pull-through (ERPT), [15] whereas others recommend regular screening with colonoscopy and subsequent colectomy if severe dysplasia, rapid polyp formation, or bleeding occurs. [18]

When intussusception occurs in children older than 2 years, the discovery of a specific lead point is not uncommon (22%); however, lead points are only found in 2-8% of children within the usual age range (6-18 months). When a polyp is demonstrated as a lead point in a patient with intussusception, an evaluation may be indicated to identify polyposis syndromes. [1]

Some hamartomas do not appear to have any malignant potential. However, germline mutations and somatic inactivation of STK11, SMAD4, BMPR1A, and PTEN genes in hamartomatous polyposis syndromes create an epithelial environment favorable for neoplastic transformation. [2, 1]

Peutz-Jeghers syndrome

In 1921, Peutz reported on the association of intestinal polyps with mucocutaneous pigmented spots of the mouth, hands, and feet. [19] From 1944-1949, in a study of 20 patients, Jeghers defined the 2 main features of the syndrome as melanotic spots on the buccal mucosa and lips (with variable melanin pigmentation on the face and digits) and polyposis of the intestinal tract. [6] The melanotic spots range from brown to black and occur in the rectum, around the mouth, and on the lips, buccal mucosa, feet, nasal mucosa, and conjunctivae. These spots are typically present at puberty. [6]

The polyps most commonly appear in the small intestine (55%), followed by stomach and duodenum (30%) and the colorectal area (15%). Dramatic advances have occurred in the understanding of the genetic and molecular basis of the disease that apply to these polyps.

Although adenomas can occur concurrently in the syndrome, these polyps are mostly hamartomas of the muscularis mucosa. They appear as pedunculated lobulated lesions, measuring from a few millimeters to several centimeters. Peutz-Jeghers syndrome is inherited as an autosomal dominant trait, [19] but de novo cases can also develop. It affects all ethnic groups with equal sex distribution; [6] however, symptoms appear earlier in males (5-10 y) than in females (10-15 y). [19]

GI disturbances become apparent later. Patients usually present during early adolescence. Some patients present with an increased frequency of defecation, rectal bleeding, anemia, abdominal pain, vomiting, or recurrent episodes of intussusception. [19] Prolapse of rectal polyps in the first year of life, even in the absence of pigmentation, may indicate Peutz-Jeghers syndrome, at least in the familial cases.

Compared with the general population, patients with Peutz-Jeghers syndrome have a 13-fold increased risk of death due to GI cancer and a 9-fold increased risk for all other cancers. [6] The risk of death due to cancer by age 60 years is 50%, and it reaches 85% by age 70. Adenomatous and carcinomatous changes in the hamartomas have been reported. [20]

Peutz-Jeghers syndrome is closely related to early onset non-GI malignancies including breast, ovary, cervix, fallopian tubes, thyroid, lung, gallbladder, bile duct, pancreatic and testicular tumors. [3]

Screening tests to detect all these forms of cancer are recommended in children who present with abdominal pain or occult anemia and melanotic-pigmented spots. An aggressive screening and biopsy program should be undertaken, including an annual examination with CBC count, breast and pelvic examinations (with cervical smears and pelvic ultrasonography) in females, mammography at age 25 years, testicular examination in males, pancreatic ultrasonography, and biennial upper and lower endoscopy.

Extensive intestinal resections are contraindicated because of the recurrent nature of the polyps and the ensuing short-bowel syndrome that may result. Rapid growth, induration, severe dysplasia, villous changes, or polyps larger than 15 mm (which presents a much higher chance of having malignant transformation) suggest the need for a more aggressive intervention. [6]

Gardner syndrome

In 1962, Gardner and colleagues noticed extra-colonic manifestations in some kindred with polyposis. In this syndrome, the polyps are adenomatous rather than hamartomas. The associated extraintestinal tumors include desmoid cysts, cysts of the mandible, fibromas, osteomas, and hypertrophy of the retinal pigmented epithelium. [21] Bone tumors are most common (80%), followed by inclusion cysts (35%) and desmoid tumors (18%). [7] The syndrome is inherited in an autosomal dominant pattern. Osteomas are most frequently found in the skull and facial bones. Abnormal dentition is common. [6] Periampullary malignancies may develop during the third or fourth decades of life at rates much higher than in the general population. [7]

Gardner syndrome is considered a phenotypic variant of FAP, and different mutations on the adenomatous polyposis coli (APC) gene have been shown to be associated with this syndrome (APC polymorphism in exons 13 and 15). [22] Intestinal polyps have a 100% likelihood of undergoing malignant transformation. [23]

The natural history and treatment of patients with colonic polyps is the same as in those with FAP. Desmoid tumors of the abdominal wall and mesentery occur in 20% of patients with Gardner syndrome, usually appear 6-30 months after surgery for intestinal manifestations, and are the leading cause of death in patients who have undergone colectomy. Desmoid tumors are dense fibroblastic proliferations but can present with dysplasia and even fibrosarcoma. Treatment is challenging. When these tumors are small and well defined, excision is feasible with a recurrence rate of 10%; however, some are not identified until they become unresectable. Desmoids that involve the small bowel mesentery should be treated according to their symptoms and growth rate. Sulindac, tamoxifen, or vinblastine and methotrexate are adequate for slow-growing, mildly symptomatic tumors. Aggressive tumors require high-dose tamoxifen, or antisarcoma chemotherapy (doxorubicin and dacarbazine), and possibly radiation therapy. [24, 25]

Turcot syndrome

This syndrome, also considered a variant of FAP, includes multiple pediatric brain tumors (eg, gliomas, ependymomas) in families that also have an increased risk for polyposis and colon cancer. All patients with this syndrome develop carcinoma of the colon as young adults. [7] Colonic adenocarcinomas occur in the colonic polyps and in the mucosa between the polyps. Patients may present with chronic bloody diarrhea, hypoproteinemia, weight loss, anemia, malnutrition, bowel obstruction, and intussusception. Hamilton found that families with Turcot syndrome have mutations in APC or HNPCC genes. [26] The type of brain tumor correlates with the mutations, medulloblastomas in APC -related mutations, and microsatellite instability in families with glioblastoma multiforme. [6] In patients with a strong family history, begin diagnostic investigation during the second decade of life and continue annually.

Cronkhite-Canada syndrome

This is a variant of juvenile polyposis in which the GI polyps are associated with skin hyperpigmentation, alopecia, and nail changes (Cronkhite, 1955). Hair loss, and skin and nail changes may be evident long before GI symptoms appear. The hamartomatous polyps appear in the stomach and colon. Chronic diarrhea results in malabsorption, hypovitaminosis, hypoproteinemia, and fluid and electrolyte imbalance. Because patients with Cronkhite-Canada syndrome may develop colonic malignancy, close follow-up is recommended (see Syndromes associated with CRC). [7]

Typically, in all syndromes with increased risk of cancer, it is recommended to follow a screening schedule like the one used for patients with Peutz-Jeghers syndrome, in order to identify malignancies at an earlier stage.

Bloom syndrome

This is a rare, recessively inherited disease, in which growth retardation, accelerated aging, immunodeficiency, susceptibility to chromosome breaks, and a high frequency of malignant tumors are observed. [27] Patients with Bloom syndrome appear hypersensitive to various different DNA-damaging agents, such as UV light and irradiation. A generalized DNA repair defect is present, likely a defect in DNA ligation; thus, this process has been encompassed in diseases of DNA repair defects such as xeroderma pigmentosum, ataxia-telangiectasia, and Fanconi anemia.

The Bloom syndrome gene has been found in chromosome 15. In this regard, genes involved in DNA repair may be considered tumor suppressor genes. [28] No specific incidence of colorectal cancer in patients with Bloom syndrome has been described in the literature.

Cowden syndrome

This is an autosomal dominant transmitted disease with hamartomas of all 3 embryonal layers. Facial tricholemmomas, oral papillomas, multinodular goiter, and GI polyps with occasional GI cancer may also be found in patients with this syndrome. Fibrocystic breast disease and esophageal glycogenic acanthosis have been described. [25] These patients have a higher breast and thyroid cancer risk. Germline mutations have been identified in the PTEN gene. Lhermitte-Duclos disease is a variant of Cowden syndrome associated with cerebellar hamartomatous overgrowth. [3]

Treatment is directed toward alleviating symptoms of pain, bleeding, or obstruction. Polyps should be removed when symptomatic, and screening to detect subsequent development of more polyps is warranted.

Ruvalcaba-Myhre-Smith syndrome

This syndrome includes developmental abnormalities, microcephaly, and juvenile polyposis. It is a rare disease that occurs in males. No cancer has been reported in these patients. The polyps are removed when symptomatic, and family screening is advised.

Osler-Weber-Rendu syndrome

Also termed HHT, Osler-Weber-Rendu syndrome is an autosomal dominant familial disorder characterized by telangiectases and vascular malformations of the skin and mucous membranes and recurrent GI bleeding. It may also affect the brain, lungs, and liver. [29] The lesions are typically noticed in the first few years of life, and 50% of patients aged 10 years have had a GI bleed. A family history of the disease is reported in 80% of cases.

Telangiectases are usually present on the lips, oral and nasopharyngeal membranes, tongue, and perlingual areas. They also occur in the colon but are more common in the stomach and small bowel, where they tend to cause significant bleeding. [30]

In one study, 6 of 24 patients (25%) evaluated with HHT developed a colonic neoplasia, 3 had adenocarcinoma of the colon, and 3 more had multiple colonic polyps. [31] Elinav et al recommend lower GI tract evaluation for all patients with new-onset anemia or GI bleeding, even if blood loss may be a manifestation of GI HHT.

Oldfield syndrome

This syndrome refers to the association between sebaceous cysts and FAP. Patients present during adolescence with subcutaneous lesions typically located on the extremities, scalp, and face; they develop during adolescence. [32] These patients share the same chromosomal derangements as those with FAP (ie, germline mutations of the APC gene on band 5q21).

Serrated Polyposis Syndrome

This newly describe syndrome, also known as "hyperplastic polyposis syndrome", has an unknown molecular etiology and inheritance pattern, arising mostly as a sporadic syndrome, it confers a 35% risk for colorectal cancer.

The colonoscopic criteria for this syndrome includes one of the following:

  • At least 5 histologically confirmed serrated polyps proximal to the sigmoid colon, with at least 2 being greater than 1cm
  • Any number of serrated polyps proximal to the sigmoid colon, in a patient with a first-degree relative diagnosed with serrated polyposis syndrome
  • More than 20 serrated polyps scattered throughout the entire colon.

In order to prevent malignancy in these patients, complete colonoscopic detection and resection of polyps should be performed, in cases in which this is not feasible, surgical resection is recommended. [3]

MutYH-Associated Polyposis Syndrome

The most recently described adenomatous polyposis syndrome, MutYH-associated polyposis (MAP), has an autosomal recessive herency and requires an inherited mutation from each parent for the development of the disease, it is caused by a bi-allelic mutation in the MutYH gene, located in the chromosome 1p33-34, which encodes for a DNA glycosylase responsible for base excision repair, this mutation has been known to occur in 1-2% of the general population. [2]

The polyps found in MAP are typically small tubular or tubulo-villous adenomas, in untreated patients the risk for colorectal cancer is 80% at age 80; contrasting to FAP, the clinical presentation is usually an adult patient with tens to thousands of polyps, resembling a less severe FAP syndrome that presents during adulthood. [3, 1]

Colorectal carcinoma

Syndromes associated with CRC include the following:

  • Gardner syndrome - Polyposis, osteomas, and multiple sebaceous cysts
  • Turcot syndrome - Polyposis and brain tumors (gliomas, ependymomas)
  • Peutz-Jeghers syndrome - Colonic polyposis, ovarian tumors, and mucocutaneous pigmentation of lips, oral mucosa, and perioral region
  • Cronkhite-Canada syndrome - GI polyposis, skin hyperpigmentation, alopecia, and nail changes
  • Osler-Weber-Rendu syndrome - Juvenile polyps and hepatic telangiectasia
  • Oldfield syndrome - Polyposis and multiple sebaceous cysts
  • Bloom syndrome - Growth retardation, accelerated aging, immune deficiency, and malignant tumors
  • Cowden syndrome - Hamartomas, GI polyps, breast, thyroid, and GI cancer
  • Ruvalcaba-Myhre-Smith syndrome - Microcephaly and juvenile polyposis in males; no cancer
  • Serrated polyposis syndrome
  • MutYH associated polyposis syndrome

The screening schedule for patients with polyposis syndromes and increased risk of malignancy is as follows:

  • Symptoms related to polyps - Annually
  • Blood count to detect anemia - Annually
  • Breast and pelvic examinations with cervical smears and pelvic ultrasonography in girls - Annually
  • Testicular examination with ultrasonography in boys - Annually
  • Pancreatic ultrasonography - Annually
  • Esophagogastroduodenoscopy and colonoscopy - Biennially
  • Mammography - Recommended at ages 25, 30, 35, and 38 years; biennially until age 50 years; annually thereafter

Familial colon cancer

Familial colon cancer syndromes are divided into those associated with polyposis (familial polyposis coli) and Hereditary non-polyposis colorectal cancer (HNPCC). To provide a better description of the genesis of these 2 entities and their differences, their genetic principles are briefly discussed below.


Colorectal Carcinoma Genetics (Defects in Mismatch Recognition and Repair)

The unwinding and copying enzymes that replicate DNA, form a highly efficient and accurate replicative complex. However, this process is not perfect. Mistakes in base pairing occasionally occur. Some stretches of DNA are more likely to accumulate errors than others, particularly stretches of DNA that consist of tandem-repeat units. These areas are termed microsatellite regions. Certain patients have marked instability in the microsatellite repeats throughout their genomes; this instability leads to a failure to recognize and repair these nucleotide mismatches. Mismatch repair defects are an early step in the process leading to malignant transformation.

The progression from normal to dysplastic colonic epithelium begins with hyperplasia, followed by the development of adenomas and, finally, invasive carcinomas. Most mutations that occur in colon cancer develop after birth in single cells as a result of exposures to environmental influences or perhaps as a result of mistakes that cells make when they copy their DNA during cell division. Approximately 80% of annual cases of colorectal carcinoma (CRC) are not associated with hereditary factors.

The progression of adenoma to carcinoma depends on reproducible genetic alterations such as APC gene inactivation, K-ras oncogene activation, and p53 mutation.

Mutations in the APC gene, a tumor suppressor gene that controls tumor initiation, are present in 80-90% of patients with familial adenomatous polyposis (FAP). [24, 33] When the APC gene is mutated, the function of both APC alleles is lost. One allele is defective at birth in all cells, having been inherited from one parent; the other APC gene allele is mutated in individual colon cells during early childhood, supporting the 2-hit hypothesis by Knudson. [34]

Malignant progression from the development of hyperplasia takes 20-30 years. This is because the tumors have to accumulate other mutations in oncogenes and other tumor suppressor genes that convert the benign adenoma into a malignant tumor.

In contrast, defects in DNA repair, particularly a DNA repair system termed DNA mismatch repair, cause hereditary nonpolyposis colon cancer (HNPCC). The enzymes that copy DNA are not perfect and often make mistakes. This mismatch must be repaired in order to avoid mutations. The DNA mismatch repair system recognizes the DNA mismatch and repairs it. [35]

Patients with HNPCC do not have defects in the APC gene inherited from their parents. Benign tumors (ie, adenomas) develop at the same rate in these patients than in the general population; however, once a patient with HNPCC has an adenoma, it rapidly progresses because of the inherited DNA repair defect. Mutations involving tumor suppressor genes and oncogenes rapidly accumulate, and, as a result, only 3-5 years are needed for a benign tumor to progress to cancer. FAP may be considered a disease of tumor initiation, whereas HNPCC may be considered a disease of tumor progression. [36]


Familial Adenomatous Polyposis

As mentioned, adenomatous polyps represent a disturbing alteration in the mucosa and have substantial malignant potential. Adenomatous polyps occur in less than 3% of children with polyps. [6] However, familial adenomatous polyposis (FAP) accounts for less than 1% of all colorectal cancer. [25] The criteria for a diagnosis of FAP and, thus, for an increased risk of cancer in children with polyps include the following:

  • More than 5 polyps in the colon
  • Polyps throughout the GI tract
  • Any number of polyps associated with a family history of juvenile polyposis

FAP occurs in approximately 1 per 7,000 individuals. [34] The major feature of the syndrome is extensive polyposis, which is defined by at least 100 visible adenomatous polyps in the large intestine. [6] Some patients have thousands of polyps. The rate of development of colorectal carcinoma (CRC) after the third decade of life is nearly 90%. [35]

These patients also have a greatly increased risk of upper GI malignancies (eg, duodenal and periampullary adenocarcinomas), thyroid cancer (occurring in 1% of patients with FAP), [37] and hepatoblastoma (occurring in 1 case per 250 persons with FAP, compared with 1 case per 100,000 persons in the general population).{Ref41, 42}

FAP has been divided into 2 types: the sparse type and the profuse type. The former one is characterized by hundreds of polyps; whereas the latter is characterized by thousands of polyps. Patients with the profuse type tend to have adenocarcinoma at an earlier age. [6] In addition, depending on the codons that are mutated within the gene, patients may develop desmoid tumors or congenital hypertrophy of retinal pigment epithelium (CHRPE). [38] Environmental factors may affect manifestations of FAP; identical mutations may result in different phenotypes in different patients. [38]

Adenomatous polyps progress through dysplasia to complete transformation. Extension of the neoplastic cells into the basement membrane of the colonic epithelium represents carcinoma in situ. Because the colonic mucosa does not contain lymphatics, metastases do not occur until the tumor invades the submucosa through the muscularis mucosa. The duodenal mucosa may also be involved with adenomatous polyps; periampullary adenocarcinoma develops in the duodenal mucosa in 2.9% of patients with FAP. [39] Although gastric polyps may occur in patients with FAP, the polyps are usually benign hamartomas, and no evidence of neoplastic transformation has been reported. [6] The use of capsule endoscopy may play a role in the screening for small intestinal polyps.

FAP is inherited as an autosomal dominant trait; a 10% incidence of new mutations is reported. FAP is caused by a deletion in the APC gene (tumor suppressor) on the long arm of chromosome 5. [36] Specific genetic alterations have been identified in most of the 30% of patients who do not test positive for mutations in the APC gene with routine testing. [24] Recent studies indicate the presence of mosaicism in approximately 15% of such cases. [40]

All untreated patients with FAP develop colon cancer. The average age at which these patients develop cancer is 39 years; malignant transformation occurs by age 20 years in 7%, and by age 25 years in 15% of patients. [6]

Patients usually present during early adolescence. Approximately 90% are asymptomatic but are identified during routine surveillance because of a family history of FAP. Some patients present with an increased frequency of defecation, rectal bleeding, anemia, and abdominal pain. The diagnosis is confirmed by endoscopic biopsy findings. Most who present with symptoms already have a malignant condition. [6] Because many FAP carriers have few polyps but still develop early colorectal cancer, surgery is indicated even if polyposis does not develop. After prophylactic surgery, carriers require screening of their upper GI tracts and rectums (if rectal mucosa is left in place) to evaluate for malignancy. [35] Surgical removal of the entire colonic mucosa prevents the development of CRC. [9]

Total proctocolectomy with permanent ileostomy is not advocated because of the physiologic and psychologic impact of a permanent stoma in a young patient, and the risk of bladder atony, impotence, and retrograde ejaculation due to destruction of nervi erigentes during the pelvic dissection. [35] Total abdominal colectomy with ileorectal anastomosis is also not ideal because 44% of patients require subsequent treatment for rectal polyps that develop in the remaining mucosa, and because the cumulative risk of developing rectal cancer is 10% at age 50 years and 29% by age 60 years. [6]

In a 30-year review of FAP, Nikitin et al found that coloproctectomy with preservation of the anal sphincter and coloproctectomy with ileoanal pull-through resulted in the development of anal canal cancer in 4.1% of patients, whereas 10.7% of patients developed cancer after colectomy with preservation of various colonic segments. [41] They found that occurrence of cancer is not related to sex, age, length of preserved colonic segment, presence of cancer in the removed colonic segment, or postoperative follow-up period. Presence of polyps in the colonic segments preserved during surgery significantly increased the risk of cancer.

Total colectomy with a rectal mucosectomy and endorectal pull-through (ERPT) is the procedure of choice. [7] Some authors suggest that an ileal reservoir is not absolutely necessary because it increases the risk of pouchitis (23%) and that patients who undergo straight (nonreservoir) pull-through develop a neoreservoir within 24 months, which decreases the frequency of stools. [6]

Laparoscopic techniques for total abdominal colectomy with ileorectal anastomosis have been described as safe and effective. [42] Gastroduodenoscopy and flexible endoscopic surveillance of the pelvic pouch must be performed annually in these patients. [25] Rapid growth, induration, severe dysplasia, villous changes, or polyps larger than 1 cm (which presents a much higher chance of having malignant transformation) suggest the need for more aggressive intervention.

Sulindac (Clinoril), a nonsteroidal anti-inflammatory drug (NSAID), was found to reduce the number of polyps in patients with FAP. [43] However, later studies demonstrated that the effect was only partial and that CRC may still develop. Apparently, the mechanism of action is induction of apoptosis in the abnormally proliferating epithelial cells. Studies with specific cyclooxygenase-2 (COX-2) inhibitors (eg, celecoxib [Celebrex]) have also shown some efficacy. [44]

The American Society of Colon and Rectal Surgeons have published management guidelines and practice parameters for patients with FAP. Patients with FAP or people with personal or family risk factors for FAP, should be referred to center registries and genetic counselors with experience in the multidisciplinary management of these individuals. [45]


Hereditary Nonpolyposis Colon Cancer (Lynch Syndrome)

Hereditary nonpolyposis colon cancer (HNPCC) is a syndrome of early onset colorectal cancer in multiple family members. In contrast to familial adenomatous polyposis (FAP), HNPCC does not have a specific phenotype, and malignancy develops in the absence of adenomatosis of the colon and rectum. Lynch syndrome is the most common hereditary colon cancer syndrome and accounts for about 2-3% of all colorectal cancer cases. [40] The lifetime risk of developing colorectal cancer for patients with Lynch syndrome is 60-80%. [40]

The expression of the disease may be limited to the colon (Lynch syndrome I) or may coexist with extracolonic tumors (Lynch syndrome II). [46] These other tumors include endometrial (second most common), uterine, ovarian, stomach, pancreatic, and genitourinary cancers. [47] They usually manifest in the second decade of life. [35] Patients with Lynch syndrome have a 50% lifetime risk of developing cancer and a 3-fold increased incidence of CRC compared with the general population. Female patients with Lynch II syndrome should also undergo vaginal ultrasonography, endometrial aspiration, and serum CA-125 assessment annually beginning at age 30 years. [25]

Most CRCs in patients with HNPCC demonstrate microsatellite instability. Lynch used the term replication error positive to describe such tumors. [38, 25] The first genetic cause of the syndrome was identified in chromosome 2; since then, additional loci have been described. More than 90% of these mutations are in 2 genes, MSH2 and MLH1, which are located on chromosome arms 2p and 3p, respectively. These genes are inherited in a dominant fashion, with 90% penetrance. [46] Patients suspected of carrying the mutation may be tested for mismatch repair gene mutations in commercial laboratories.

HNPCC patients have a better prognosis than those with sporadic CRC. [38, 25, 48] The average age of colorectal cancer onset among HNPCC gene carriers appears to be at 69 years. [49] Synchronous and metachronous tumors are frequent.

Patients present with lower stages of disease at diagnosis than patients with sporadic CRC, and distant metastases at diagnosis are also less frequent. One possible explanation is that the large number of mutations that accumulate in these cells results in production of abnormal products recognized as foreign by the host. Alternatively, a high mutation rate may actually hinder tumor dissemination through derangement of functions critical to that process. [48]

The Amsterdam criteria for defining HNPCC include the following: [50, 4]

  • Onset of colorectal, endometrial, small bowel, ureter or renal pelvis cancer in at least 3 individuals spanning 2 generations
  • At least one of these individuals is a first-degree relative of the other two
  • At least one of these individuals must have a diagnosis prior to age 50 years
  • Familial Adenomatous Polyposis has been excluded

The stringency of the Amsterdam criteria likely excludes a number of individuals with an inherited predisposition to colorectal cancer. FAP must be excluded. Some patients with this syndrome satisfy the Amsterdam criteria for HNPCC, but their cancer cells do not display MSI-2. These patients seem to have a distinct form of the syndrome recently referred to as "familial colorectal cancer type X." They have a lesser risk for colorectal cancer and a later onset. They do not carry the risk for extracolonic malignancies. [51, 25]

HNPCC malignancies occur in the cecum and ascending colon more often than in other sites (70%); they have a higher incidence of poorly differentiated and mucin-producing tumors (signet cell). Colonoscopy screening is recommended biannually, beginning at age 25 years. [52, 25, 53, 54] Lynch proposed treatment with subtotal colectomy rather than hemicolectomy or a segmental resection because the risk of a second or third primary cancer of the colon is 45% over 10 years. Patients who are poorly compliant with colonoscopic surveillance may be candidates for prophylactic colectomy. Patients who have undergone subtotal colectomy must be informed that they require lifelong endoscopic evaluation of their remaining rectal segment. [38] They also face the risk of extracolonic cancers

Lynch reported that colonic polyps can be identified in as many as 17% of first-degree relatives during colonoscopic screening. Adenomas are more likely to grow and progress to invasive cancer in this patient population.

The optimal management strategy for HNPCC gene carriers has yet to be established. Evidence suggests that surveillance colonoscopy and polypectomy are effective in reducing the risk of invasive cancer for these patients. [40] It seems wise to advise biannual colonoscopy starting at age 25-30 years and annually after age 40 years. Currently, experts have not reached consensus regarding the role of prophylactic subtotal colectomy for patients with HNPCC. Given the high penetrance of the disorder and the high rate of synchronous and metachronous disease among mutation carriers, a strong case may be made for prophylactic colectomy.

HNPCC is usually underdiagnosed due to the prompt surgical treatment required to provide symptomatic relief. Patients who present with colorectal cancer before 50 years of age have more symptoms associated with this malignancy, such as bleeding and abdominal pain. This is a critical problem for the surgeon, given that an incomplete assessment for Lynch Syndrome derives in a deficient care and counseling to the patient, prevents a more extensive surgical approach (hemicolectomy or total colectomy), and may overlook synchronous extracolonic tumors, failing to provide complete treatment. [5]

Subtotal colectomy with a rectal mucosectomy and endorectal pull-through (ERPT) has not been studied in this population. Patients who elect to undergo a subtotal colectomy require colonoscopic surveillance of the remaining rectum. Patients who elect not to consider prophylactic surgery must commit to lifelong surveillance. At present, both of these are reasonable management strategies.


Sporadic Colorectal Carcinoma

Adenocarcinoma of the colon and rectum is the most common cancer of the GI tract [55] and carries the second highest mortality rate. [1] Approximately 150,000 new cases are diagnosed annually in the United States; less than 1% of which (approximately 80 cases per year) occur in patients younger than 20 years of age. [35, 9, 56]

It is known that the incidence of adenocarcinoma of the colon increases significantly after the 5th decade of life, and continues to increase as age increases, although it is important to mention that 12% of colon cancer cases present in patients younger than 50 years. [6]

In children, colorectal cancer is the second most common cancer of the alimentary tract after liver tumors; the incidence of colorectal cancer is 1.3-2 cases per million population. [55, 57] Most patients present during the second decade of life. [55, 58, 59] The youngest patient ever reported was 9 months of age. [60] Children are more likely than adults to have advanced-stage at presentation, as well as unfavorable tumor histology (mucinous), which confers them a poor outcome. [61, 60]

The development of carcinoma of the colon appears to be associated with several predisposing factors, as follows:

  • Familial polyposis syndromes
  • Hereditary nonpolyposis syndromes
  • Inflammatory bowel disease (ulcerative colitis)
  • Previous ureterosigmoidostomy
  • Previous radiation therapy
  • Possible exposure to environmental chemicals and radiation
  • Obesity and Adipokines

Another predisposing factor is a high fat and low fiber diet. Fecal material in the bowel of patients consuming low-fiber diets has a long transit time; this association is indicated by a higher colorectal cancer incidence, associated with a parallel increase in economic development and adoption of Western diets and lifestyles, which is responsible for the higher incidence in developed countries. However, this association has not been supported in children. [4]

A new cross-sectional study by Comstock, et al. showed a positive association between increased BMI, increased waist circumference, leptin, IP-10, and TNF-alpha with the development of multiple (more than 3) colonic polyps and tubular adenomas, showing a significant association between obesity and colorectal tumors. [7]

Approximately 75% of patients represent sporadic cases, 10-20% have familial colon cancer without a defined genetic pattern, and 1% of patients have a genetic polyposis syndrome. [36] The oncogene ß-catenin has been identified in 15% of colorectal adenocarcinomas. ß-catenin creates a dedifferentiated stem cell–like phenotype that remains in a state of constant proliferation, which is an important contributor of cancer invasion and metastasis. The serum protease, urokinase plasminogen activator (uPA) initiates a cascade of proteolytic steps that degrade the extracellular matrix, playing a role in adhesion, migration, and invasion. It coordinates cell growth and, when found in tumor cells, indicates aggressive tumor growth and poor patient survival.

In patients with sporadic colorectal carcinoma (CRC), 1% are secondary to inflammatory bowel disease, and 5-6% present with hereditary nonpolyposis syndromes. There is a clear-cut association between colon cancer and patients with ulcerative colitis (UC); the risk is 20-fold higher than in the general population, being related to the length of time with UC diagnosis, [62, 63] increasing 1-2% per year after 10 years. Patient's age and extent of disease at diagnosis are strong independent risk factors for colorectal cancer in patients with UC. Cancer risk is higher for patients with childhood-onset inflammatory bowel disease when compared with adult onset counterparts. There is a cumulative incidence of 5-10% at age 20, and 12-20% at age 30 years. [64]

Ulcerative proctitis likely does not increase the risk of colonic or rectal cancer. [64] The current practice of surveillance and early surgery may have and impact on prognosis. [65]

In patients with ulcerative colitis who develop colonic malignancies, cancer usually occurs at a young age, with synchronous tumors being prevalent. Furthermore, patients with colitis involving the entire colon are more likely to develop cancer than those with left-sided colitis only. In these patients, the risk of cancer begins to increase 15-20 years after the onset of symptoms. [66] Patients with ulcerative colitis who develop colonic strictures should be considered to have carcinomas until proven otherwise, and a stricture in these patients is an indication for surgical intervention. [7] In a report by Lashner, 11 of 15 patients with strictures were found to have carcinomas on biopsy findings. [63]

Patients who have had the disease for a minimum of 7 years should undergo surveillance endoscopy at least every other year. [7] Colectomy is indicated for the following reasons: [64]

  • A macroscopic lesion with overlying low-grade dysplasia
  • A low-grade dysplasia in multiple foci
  • Persistent unifocal low-grade dysplasia found on repeated examinations
  • High-grade dysplasia

Patients with Crohn disease, carry a risk for colon cancer 20 times greater than the general population; this risk is also related to the amount of bowel involved. These patients also carry an increased risk for lymphoma and small intestinal tumors. [65] Their colorectal cancer risk probably results from chronic inflammation or dysplastic rectal mucosa (left after surgery). Thus, close surveillance is important in such cases. [6] Contrast enemas and colonoscopy are the mainstays of surveillance. Biopsies should be performed on suspicious areas and should also be performed randomly during the colonoscopy. The risk for cancer in patients with Crohn disease is less frequent than in those with ulcerative colitis. [67] Some malignancies related to inflammatory bowel disease might be secondary to immunomodulation therapy used to treat the latter.

Patients who undergo urinary diversion with ureterosigmoidostomy are at risk for the development of colon cancer (5% of patients), likely because of chronic inflammation caused by the mixture of feces and urine at the implant site. [68] Although many types of urinary diversion have been described, almost all of them include an intestinal segment as a reservoir. An exception is ureterostomy, in which the ureters open directly on the skin. However, these stomas are not continent and difficult to care for. They also tend to stenose frequently. The Bricker diversion includes an ileal conduit as a reservoir and has been used worldwide after cystectomy for cancer.

Another option for diversion is the Mainz 2 pouch, which uses a sigmoid pouch as a reservoir, in which the ureters are implanted with an antireflux mechanism. This seems to decrease the incidence of cancer in the colonic segment, which always presents at the site of the ureteral implant and is believed to be secondary to the carcinogenic effect of nitrosamines present in the infected urine secondary to reflux.

Children with other abdominal malignancies who have received radiation therapy also pose an increased risk for early development of colorectal cancer in the radiation field. It is likely that radiation induces mutations in the P53 gene.{Ref63, 74} The Children’s Oncology Group screening guidelines recommend that children and adolescents who receive 30 Gy or more of pelvic radiotherapy be screened for colorectal cancer after 10 years or at age 35 years, whichever is later. [61]

Colorectal cancer in pediatric patients has a significantly worse survival rate than adults. This is likely due to delayed diagnosis, advanced stage at presentation, and increased incidence of high-grade tumors. Approximately 60-86% of pediatric and adolescent patients have Dukes stage C or D. [61] The increased frequency of mucinous variants and preponderance of right-sided lesions contribute to the advanced stage at diagnosis. [55, 69]

In the largest cohort of pediatric colorectal cancer patients, a study by Poles et al compared management and outcomes of 918 pediatric colorectal cancer patients (Ped) with the management and outcomes of 157,779 early onset adult (EA) and 1,304,085 older adults with colorectal cancer tumors (OA). The study found that patients ≤50 presented more frequently with stage 3 and 4 disease (Ped: 62.0%, EA: 49.7%, OA: 37.3%) and rectal cancer (Ped: 23.6%, EA: 27.5%, OA: 19.2%). Pediatric histology was more likely signet ring, mucinous, and poorly differentiated. Initial treatment was usually surgery, but patients ≤50 were more likely to have radiation (Ped: 15.1%, EA: 18.6%, and OA: 9.2%) and chemotherapy (Ped: 42.0%, EA: 38.2%, and OA: 22.7%). Children and older adults showed poorer overall survival at 5 years when compared to early onset adults. The study also added that despite standard oncologic treatment, age ≤21 was a significant predictor of mortality. [70]


In adults, colorectal cancer is the second leading cause of cancer death in the Western world. [71] It is rare in populations with limited meat intake; therefore, it is less common in Africa and Asia. [35, 72] In children and adolescents, these tumors may occur in any site in the large bowel and are not usually associated with a family history of colon cancer. [55, 73, 74, 75, 72] No sex predilection is reported.


Colon cancer is triggered by a series of point mutations and genetic alterations that progressively cause normal cells to transform into adenomas that progressively become dysplastic, resulting in carcinoma foci. [76] These mutations occur in a certain sequence that determines the clinical characteristics of the tumor.

Colorectal cancer arises from the mucosal surface of the bowel, generally at a site of an adenomatous overgrowth or polyp. The tumor may penetrate the bowel wall and even perforate the serosa into the omental fat, lymph nodes, liver, ovaries, and other loops of bowel. Some lesions cause bowel obstruction. Synchronous lesions may be present, with same or different histology and stages of development. Carcinoma in situ may occur in one or more polyps. Patients with synchronous primary tumors have the same prognosis as patients with single colon cancers. [77]

The epidermal growth factor receptor (EGFR) is abnormally expressed in colorectal cancer cells (72-82%). It promotes cell division, migration, angiogenesis, and inhibits apoptosis. [71] Thus, EGFR plays an important role in the pathogenesis of colorectal cancer. Its expression is associated with poor survival and increased risk of metastasis. Italiano et al demonstrated EGFR expression in colorectal cancer metastatic cells. [71] Monoclonal antibodies and low molecular weight tyrosine kinase inhibitors may show to be useful in the therapeutic armamentarium for patients with colorectal cancer.

A rare "flat-type" colorectal tumor has been reported. This form tends to be more aggressive despite the small size of the tumor, since it shows high-grade dysplasia and progresses rapidly to invasive cancer. Inactivation of p53 and 17p-LOH have been described in this tumor. [78] They are poorly differentiated and contain pools of mucin. Mucin-producing or signet ring adenocarcinoma is the predominant cell type; it occurs in 50% of pediatric cases compared with a 5% occurrence reported in adults. [79] These tumors may become huge. The differential diagnosis includes malignant carcinoid, leiomyosarcoma, malignant fibrous histiocytoma, and metastatic tumor from other sites.

In a series reported by LaQuaglia, the interval from symptom onset to diagnosis was not a significant predictor of mortality; however, tumor grade was. [58] Signet ring tumors, which are more common in children and adolescents than adults, behave more aggressively and are associated with earlier penetration of the bowel wall and extension along peritoneal surfaces, which suggests more aggressive tumor biology. The mucin absorbs water, swells, and invades tissues, thus promoting spread of malignant cells. Mucin also interferes with the mucopolysaccharide-coating immune recognition of carcinoma cells. [80] Children have a worse prognosis when compared with adult patients who are at the same disease stage. [1]

Clinical presentation

Signs and symptoms may be absent or innocuous. However, almost all children present with abdominal pain, and 40-70% have nausea and vomiting. Most (76%) [61] also have altered bowel habits and weight loss. Due to its rarity, colorectal cancer is frequently overlooked in the differential diagnosis of these symptoms in children. [81] Unlike in adults, colorectal cancer in children is evenly distributed throughout the colon, with up to 60% of tumors arising in the right colon. [55, 58, 1]

The signs and symptoms of colorectal cancer are related to its primary site within the large bowel. Tumors involving the cecum and ascending colon, which may be associated with familial colon carcinoma, develop large masses before symptoms appear.

Constipation or diarrhea and change in the caliber of stools are more common with left-sided lesions, and may present before the development of tarry stools or rectal bleeding, which are observed in a third of patients. Approximately 20-30% of patients have anorexia and weight loss. [6]

Clinical and laboratory investigation

In the absence of rectorrhagia or hematochezia, patients may test positive for occult blood in the stool; however, screening for fecal occult blood has not proven to be of significant value for the treatment of pediatric patients. [35] Hepatic function abnormalities may be related to metastatic involvement of the liver. Anemia is due to blood loss or malnutrition. Although fewer than 75% of colon carcinomas in children produce carcinoembryonic antigen (CEA), levels of this protein should be determined. CEA may be a useful tool in identifying recurrent disease after resection, and an increase in CEA during follow-up, is also related to a higher mortality rate. [7, 8]

A new urine metabolites screening test for colonic adenomas is under development, showing a favorable alternative to conventional fecal-based screening tests. Preliminary results show that this urine test has a sensitivity of 82.7% and a specificity of 51.2%, with a negative predictive value of 88.5%, making it a future alternative for excluding colonic polyps. [9]

In the near future, blood tests including miRNA analysis of 5-genes implicated in colorectal cancer could constitute a new screening test. Preliminary reports show a predicted specificity of 70-95% and a predictive sensitivity of 83-91%, although more multicenter studies are required. [10]

Imaging studies

Conventional radiographic studies include barium enema with air contrast to define the tumor. Abdominal and chest CT scans define spread to the liver, lungs, or enlarged lymph nodes, as well as pelvic metastases, especially to the ovaries. Colorectal cancer has the ability to metastasize through various routes, including transmural invasion and spread by continuity, intraluminal extension, and hematogenous, lymphatic, and transperitoneal routes. Kaste et al analyzed 32 patients with peritoneal metastatic implants from different primary tumors and found that 22% of these patients had colorectal cancer. [82]

CT scanning may be unable to detect intra-abdominal metastases because of lesion size, paucity of intra-abdominal fat, contiguity with the primary tumor, ascites, implant location, and adequacy of bowel opacification. Current CT scanners are able to detect implants as small as 5 mm in diameter. MRI may further improve detection.

Colonoscopy is useful in locating the site of lesions within the large bowel. The entire length of the colon should be evaluated. Transrectal ultrasonography may help determine the extent of invasion and resectability of rectosigmoid cancer. Intraoperative ultrasonography of the liver may reveal metastases not observed in other imaging studies. [7] Radioisotope studies should include a bone scan; if results are positive, bone marrow aspiration and biopsy are indicated to determine spread to the marrow.

Advanced imaging techniques are indicated for specific patient groups, in 2014, the European Society of Gastrointestinal Endoscopy (ESGE) strongly recommended that conventional screening with white light colonoscopy in high-risk patients should be performed, as well as pancolonic conventional or virtual chromoendoscopy for patients suspected or known to have Lynch syndrome, and serrated polyposis syndrome.

The ESGE also recommends to have all patients with longstanding colitis undergo periodic pancolonic chromoendoscopies, with either 0.1% methylene blue or 0.1%-0.5% indigo carmine and targeted biopsies, replacing the common practice of non-targeted four quadrant biopsies. [11]


In 1932, Cuthbert Esquire Dukes, the Director of the Research Laboratory at St. Mark's Hospital in London, indicated that growth of colorectal cancers followed an orderly and predictable fashion. [83] He created a staging classification system that was later modified in 1954 by Astler and Coller. [84] The extent of the disease is determined using the modified Dukes staging scheme. According to this classification, in stage A, only the mucosa and submucosa are affected; in stage B, the disease is limited to the bowel wall; in stage C, the disease is limited to the lymph nodes; and in stage D, the patient has distant metastases, peritoneal implants, direct invasion of other viscera, or surgically unresectable tumors. Then the American Joint Committee on Cancer (AJCC), published the most commonly used system to evaluate prognosis in colorectal cancer. [1]

AJCC TNM staging of colorectal cancer is as follows: [1]

Table. (Open Table in a new window)

Primary Tumor (T) Nodal Involvement (N) Distant Metastasis (M)
TX: Primary tumor cannot be assessed. Nx: Regional lymph nodes cannot be assessed. MX: Presence of distant metastasis cannot be assessed.
T0: No evidence of primary tumor is present. N0: No evidence of regional lymph node metastases is present. M0: No evidence of distant metastasis is observed.
Tis: Carcinoma in situ is present. N1A: Metastasis in 1 pericolic or perirectal lymph nodes is present. M1A: Distant metastasis is present in a single organ.
T1: Tumor cells invade the submucosa N1B: Metastasis in 2-3 pericolic or perirectal lymph nodes is present. M1B: Distant metastasis is present in multiple organs.
T2: Tumor cells invade the muscularis propria. N1C: Metastasis in subserosa, mesentery, or non peritonealized pericolic or perirectal tissue without lymph node metastasis.  
T3: Tumor cells invade the muscularis propria into nonperitonealized pericolic or perirectal tissues. N2A: Metastasis in 4-6 pericolic or perirectal lymph nodes.  
T4A: Tumor cells perforate the visceral peritoneum . N2B: Metastasis in 7 or more pericolic or perirectal lymph nodes is observed.  
T4B: Tumor cells directly invade and adhere other organs and structures.    

The AJCC prognostic stages are defined as follows: [1]

Table. (Open Table in a new window)

Stage Primary Tumor (T) Nodal Involvement (N) Distant Metastasis (M)
0 Tis N0 M0
I T1 N0 M0
  T2 N0 M0
IIA T3 N0 M0
  T1 N2A M0
  T2/T3 N2A M0
  T1/T2 N2B M0
  T3/T4A N2B M0
  T4B N1A/N1B/N1C/N2A/N2B M0
IVA Any T Any N M1A
IVB Any T Any N M1B


Surgery is the only effective modality to cure these patients, although adjuvant chemotherapy extends life. Few patients with extensive metastatic disease are cured.

Because it has been shown to be of prognostic significance, complete surgical resection with aggressive lymph node dissection is essential for cure; [60] thus, it is the goal of surgery. Adequate retroperitoneal lymph node biopsies cannot be overemphasized. [61] However, only 40-69% of pediatric patients are candidates for curative resection, a much lower number than in adults.{Ref6, 38, 60, 63} Debulking is of little benefit for patients with extensive metastatic disease. Occasionally, resections of bulky tumors or metastases offer palliation.

General surgical principles apply. Biopsy findings are required to confirm the diagnosis of colorectal cancer. A biopsy sample can be obtained by colonoscopy, laparoscopy, or laparotomy. [85] Staging procedures include performing a biopsy of known enlarged lymph nodes, a biopsy of the ovaries in female patients, resection of the omentum, and liver biopsies. Ligating the mesenteric vein that drains the tumor area early during surgery, before manipulating the tumor, is most important to avoid dissemination.

When the tumor is confined to the cecal area or right colon, a right hemicolectomy is performed; if the lesions are located from the splenic flexure to the sigmoid colon, a left hemicolectomy is indicated. Low anterior resections should be performed in patients with tumors located 5 cm above the anal verge. If present, liver metastases may be resected at the primary procedure or at a later time. Lobectomy is not necessary if margins can be preserved, with a lesser resection. [7]

New techniques for early stage colorectal cancer and large rectal adenomas are available. Recent studies indicate that a Transanal Endoscopic Microsurgery is feasible for this kind of tumors, providing a less than 10% rate of complications requiring conversion to open surgery, a negative margin in 85% of patients, and a recurrence rate of 9%. Extensive training is required in order to provide adequate treatment using this novel technique. [12]

Chemotherapy and radiation therapy

Patients with unresectable tumors diagnosis should undergo only biopsy and neoadjuvant chemotherapy, with or without radiation therapy.

The general consensus for surgery after neoadjuvant radiotherapy for locally invasive colorectal tumors, is an interval period of 6-weeks following radiation therapy, although recent studies show that a longer interval period could be more beneficial to patients. [13]

Intraoperative radiation therapy (IORT) is advocated for tumors that have metastasized to the mesentery or mesenteric lymph nodes. IORT is performed while the bowel is displaced from the peritoneal cavity. [35] Several chemotherapeutic regimens have been used, including 5-fluorouracil, [75] irinotecan, and folinic acid (leucovorin) rescue. [86] Other active agents in adults with colorectal cancer include oxaliplatin, irinotecan, bevacizumab, cetuximab, and capecitabine. [14]

Bevacizumab, an anti-VEGF monoclonal antibody, has been shown to have an improved response in metastatic colorectal cancer, when added to chemotherapeutic regimes such as IFL (Irinotecan + 5-Fluorouracil/Leucovorin bolus), FOLFOX (Oxiplatin + 5-Flourouracil/Leucovorin infusion) or FOLFIRI (R). These treatment options benefit only patients who underwent primary tumor resection, previous to the chemotherapeutic regime. [15]

Survivors of colorectal cancer have an increased risk for leukemia and other secondary malignancies. [87, 88, 89, 90] Boice reported a 2-6% risk of developing leukemia after treatment with semustine (nitrosoureas). [88] Alkylating agents also increase the risk for secondary leukemia. [88] In addition, pediatric patients with cancer treated with radiotherapy for solid tumors have a 3% incidence of secondary malignant tumors. [53] Survival for children with colorectal cancer is dismal; the 5-year overall survival rate in different series is 5-28%. [58, 91]

Genetic testing and chemopreventive agents

Genetic testing for cancer susceptibility has positive attributes. Family members who have tested negative for the particular APC or MMR gene are spared the repeated medical and endoscopic examinations that otherwise would have to be performed in every member of each family. They are also spared the anxiety associated with not knowing whether they are affected. Surveillance can then be concentrated on those who have inherited a mutant gene.

Even though genetic testing studies for screening are lacking, expert opinions recommend that genetic testing should be performed to every patient with newly diagnosed colorectal cancer, in order to prevent morbidity and mortality in direct relatives. Implementation of universal screening for Lynch syndrome has been declared as part of the 2020 objectives of the Office of Public Health Genomics in the United States. [16]

Future prospective studies should include a systematic family history and testing for potentially relevant genetic conditions if indicated. They should also address the risk of colorectal cancer in family members and promote sensible family screening guidelines. [61]

Most cancers have an increased glycolytic pathway (Warburg effect). Proteomic studies have demonstrated this pathway enhancement in colorectal cancer culture cells. Xhuezhi et al also reported on decreased gluconeogenesis, suppressed glucuronic acid pathway, and an impaired tricarboxylic acid cycle. [1] This provides an insight to the tumorigenesis of colorectal cancer and possible specific target therapies to be developed.

Chemopreventive agents may one day inhibit the development of adenomas. NSAIDs have well documented effect on shrinking existing adenomas in patients with FAP and potentially inhibit their formation. [34] Although NSAIDs may have non-COX-mediated pathways, aspirin inhibits COX enzymes in the conversion of arachidonic acid to prostaglandins. Although COX-1 is thought to produce cytoprotective prostaglandins in the GI tract, COX-2 is expressed in response to growth factors, mitogens, and cytokines and is found in 50% of colorectal adenomas and in 85% of cancers. [25] COX-2-specific NSAIDs inhibition is believed to be protective against epithelial transformation. NF-kB transcription factor translocation producing apoptosis is another mechanism of aspirin.

Specific COX-2 inhibitors are no longer used because of their cardiovascular side effects. Bevacizumab is a monoclonal antibody that targets vascular endothelial growth factor-A (VEGF-A), which is believed to be critical in cancer angiogenesis. Cetuximab is another monoclonal antibody that targets the EGFR, which is involved in cancer cell proliferation, degradation of the extracellular matrix (invasiveness), tumor migration, and endothelial proliferation. It may well become a good genetic target treatment for colorectal cancer. Collaborative multi-institutional pediatric clinical trials are needed to evaluate the prognosis, optimal treatment response, and the basic biology of childhood-onset colorectal cancer.