Durban Surgeons

Specialist surgeons in Durban KZN South Africa

General surgeon Durban

General and laparoscopic surgeons

Vascular surgeons

Bariatric surgery Durban

Weight loss surgery Durban

Surgical anatomy

Alimentary Tract
Abdomen and its Contents
Breast, Skin and Soft Tissue
Endocrine System
Organ Transplantation
Pediatric Surgery
Surgical Critical Care
Surgical Oncology (including Head and Neck Surgery)
Trauma/Burns and Acute Care Surgery
Vascular Surgery

Provides evaluation and treatment of the full range of complex general surgery problems. Working in conjunction with colleagues in other specialties, and collaboration with referring physicians as essential to optimal patient care.

APPLIED SURGICAL ANATOMY

Radical surgical resection remains the primary method of treatment of most colorectal tumours, and offers the greatest chance for cure, especially in early stage disease. A thorough knowledge of this anatomy is essential for optimal early and late surgical results and to achieve both low morbidity and mortality rates with complex procedures, and prolonged overall and disease-free survival, improving quality-of-life. The course of lymphatic drainage in any part of the colon or rectum is determined by its blood supply.

The colon comprises the caecum, appendix, and ascending, transverse, descending and sigmoid colon. The caecum, and transverse and sigmoid colons are considerably mobile, while the ascending and descending colons are retroperitoneal and fixed due to fusion of their mesentries with the peritoneum of the posterior abdominal wall. Characteristics of the colon include peritoneum-enclosed collections of fat called appendices epiploicae and the arrangement of its longitudinal muscle into 3 bands: the tenia omental (posterior) on the posterolateral border of the ascending and descending colons and the anterior border of the transverse colon; the tenia libera (anterior) on the antimesenteric border of the caecum and ascending, descending and sigmoid colons, but the inferior border of the transverse colon where it is covered by the attachment of the greater omentum; and the tenia mesocolica (lateral) on the posteromedial side of the caecum and the ascending, descending and sigmoid colons, and the posterior border of the transverse colon, at the attachment of the transverse mesocolon. Between the tenia coli are sacculated outpouchings called haustra coli.

Vasculature:

The mesenteric arteries originate from the primitive ventral segmental arteries, all but 3 of which regress: The 10^th segmental artery gives rise to the celiac artery that supplies the foregut, the 13^th gives rise to the superior mesenteric artery (SMA) to supply the midgut, and the 21st or 22nd artery gives rise to the inferior mesenteric artery (IMA) to supply the hindgut.

The SMA emerges from beneath the pancreas and crosses the third part of the duodenum, running between the two layers of small bowel mesentery, and dividing into right and left branches. The right branches include the inferior pancreaticoduodenal artery (which loops upward to communicate with the pancreatoduodenal arcade), the middle colic artery, the right colic artery, and the ileocolic artery. The ileocolic is the most constant, as the right colic and the middle colic could be absent. The 15–20 left branches are jejunal (4–6) and the ileal (11–14).

The superior rectal artery continues the downward course of IMA in the base of the sigmoid mesocolon to the back of the upper end of the rectum, where it bifurcates into two vessels, adjacent to the inferior portion of the pouch of Douglas and opposite the level of S3. The larger right branch divides into two main branches, which run down the right anterior and right posterior aspects of the rectum and supply the posterior and lateral surfaces, while the left continues undivided down the left lateral aspect and supplies the anterior surface of the rectum. These branches descend on the rectal wall and subdivide until they penetrate the muscle coat to reach the submucosa in which they proceed downward as straight vessels, which run in the columns of Morgagni and terminate above the anal valves as a capillary plexus.

The middle rectal artery usually arise from the anterior divisions of the internal iliac arteries and proceed medially and forward below the pelvic peritoneum, in the tissue of the lateral ligaments to reach the rectal wall, where they anastomose with the branches of the superior and inferior rectal vessels.

The inferior rectal arteries arise from the internal pudendal branches of the internal iliac arteries as they are lying in Alcock’s canal in the fascia or the outer walls of the ischiorectal fossa and run medially and slightly forward dividing into branches which penetrate the external and internal anal sphincters and reach the submucosa and subcutaneous tissues of the anal canal. They communicate with branches from the opposite side and possibly from the middle rectal arteries of both sides, leaving a potentially vessel-deficient area in the dorso-caudal region of the rectal ampulla, where most anastomotic leaks after low anterior resection of the rectum are observed.

Venous drainage of the caecum, appendix, ascending colon and most of the transverse colon is via the superior mesenteric vein, whereas the distal ⅓ to ½ of the transverse, the descending and the sigmoid colons are drained by the inferior mesenteric vein. Veins from the upper two-thirds of the rectum are drained by the superior rectal vein, which empties into the portal system via the inferior mesenteric vein. Veins from the lower third of the rectum are drained by the middle and inferior rectal veins, which empty into the systemic venous circulation via the internal iliac veins. The venous drainage of the rectum may explain why tumours of the lower rectum and/or anal canal can give directly systemic (i.e., pulmonary) metastases without hepatic metastases and why rectal varices may develop in patients with portal hypertension, complicating surgery for rectal cancer.

Lymphatic drainage of the colon begins with intramural lymphatics and progresses to extramural lymphatics. Intramural drainage begins in continuous lymphatic plexuses in the submucous and subserous layers of the colonic wall, which begin as blindended capillaries in the mucosa, from where progressively larger vessels connect with lymphatic networks in the submucosa, muscularis externa, and subserosa, eventually draining into the extramural lymphatics. The four levels of extramural colonic lymph nodes are the epicolic, paracolic, intermediate and principal nodes. The epicolic nodes lie on the colon itself, the paracolic nodes lie along the marginal artery between it and the colon, the intermediate nodes line the main colic vessels and their branches, and the principal nodes lie on the superior and inferior mesenteric vessels. Broadly, lymph from the appendix and caecum drains to ileocolic nodes via appendicular and cecal lymphatic vessels, and then to the superior mesenteric nodes. Drainage of the ascending and transverse colons is to the superior mesenteric lymph-nodes via the iliocolic, left colic and middle colic nodes, which communicate with the adjacent celiac and upper lumbar nodes. Although lymphatics draining the transverse colon communicate with those of the greater omentum and can also drain into nodes at the splenic hilum, this is a very rare route for colonic cancer spread. There is no communication between the lymph vessels of the transverse colon and those of the stomach. The left ⅓ and splenic flexure of the transverse colon are dually drained and the descending and sigmoid colons are drained by the inferior mesenteric nodes. All these nodes drain to the intestinal lymph trunk and then into the cisterna chyli.

As with the colon, continuous lymphatic plexuses in the submucous and subserous layers of the rectal wall provide for intramural lymphatic drainage. Lymph from the upper third of the rectum drains to superior rectal nodes after transversing pararectal nodes, and then to the inferior mesenteric nodes. The lymphatic drainage of the remainder of the rectum and anal canal is dependent on its relation to the mucocutaneous junction. The part proximal to the mucocutaneous junction either drains superiorly, parallel to the middle rectal artery and its branches on the corresponding side wall of the pelvis, or traverses the levator ani muscle to follow the inferior rectal artery; both pathways lead to internal iliac nodes, common iliac nodes, and the lumbar trunks. Lymphatic drainage from the anal canal inferior to the mucocutaneous junction is exceptional in that it does not parallel blood vessels. The collecting ducts pass anteriorly and superiorly in the perineum; together with lymphatic channels from perianal skin, they pass to superficial inguinal nodes, and then to the lumbar trunks via external iliac nodes,

The colon is responsible for converting 1000-2000ml of isotonic chyme that it receives from the ileum daily into 200ml of semisolid faeces by removal of 90% of the fluid content. The chyme enters the caecum via the ileocaecal valve, which is compressed shut during colonic contraction to prevent reflux back into the ileum. It relaxes with gastric emptying (gastroileal reflex). Movement of contents is facilitated by 3 types of contractions:

Up to 400mEq of sodium is absorbed daily via NaK-ATPase, along with chloride in exchange for bicarbonate. Potassium is actively secreted into the colonic lumen, and then passively diffuses back into the mucosa.

The most striking aspect of the epidemiology of colon cancer is its geographic variation: age- adjusted incidence varies by up to 12-fold among different countries, with industrialized nations having the highest incidence, and South American countries and China having low incidences. This wide variation is largely attributed to differences in diet and other environmental factors. The majority of epidemiological data about colon cancer originates from the USA, which conveniently is the prototype of an industrialised nation. The incidence in Japan, previously much lower than that in the US, has recently increased with industrialization and adaptation of a Western diet. Interestingly, descendants of Japanese immigrants to America have a high incidence of colon cancer similar to that of other Americans, which is attributed to dietary and other environmental adaptations. The lifetime risk for colon cancer in the US is approximately 1 in 17, where it is responsible for approximately 10% of all cancer mortality. This incidence has decreased by approximately 20% in the last 25 years, while the mortality has decreased by approximately 30%. American blacks have a small increased risk for colon cancer compared with whites, while American Indians have a significantly lower risk. The incidence is slightly higher in American men than women, and rises sharply with age, beginning at age 50, which has been attributed to accumulation of random somatic mutations with age. 90% of cases occur after age 50, and only 4% of cases occur before age 40.

Colon cancer is the best understood complex (multistep) cancer in terms of molecular genetics. The process of carcinogenesis begins with the development of specific types of neoplastic polyps (commonly hyperplastic and adenomatous) in the colonic mucosa.

Colon cancer arises from mucosal colonic polyps, and polyp histology determines its malignant potential. The two common histologic types of polyp are hyperplastic and adenomatous. Hyperplastic polyps contain an increased number of glandular cells with decreased cytoplasmic mucus, but lack nuclear hyperchromatism, stratification, or atypia, while adenomatous nuclei are hyperchromatic, enlarged, cigar-shaped, and crowded together in a palisade pattern. Adenomas in turn may be classified as tubular or villous. Tubular adenomas are composed of branched tubules, whereas villous adenomas contain digitiform villi arranged in a frond. Tubulovillous adenomas contain both elements. Most colon cancers arise from conventional adenomas (adenoma-to-carcinoma sequence) as demonstrated by epidemiologic, clinical, pathologic, and molecular genetic findings.

The adenoma-Carcinoma model was developed by Vogelstein and Fearon in 1988 to describe the process of transformation of an adenoma to a carcinoma and to relate the various stages of this sequence to the influences of specific genetic abnormalities. The result is a multistep model that identifies APC gene abnormality as a precipitant in the transformation of normal epithelium to dysplastic/hyperproliferative epithelium. Mutation of k-ras leads to early/intermediate adenoma formation, and loss of DCC allows progression to an advanced adenoma. Mutations of p53 are a late event that allows for transformation of the dysplastic adenoma to a carcinoma.

1. Operative specimens containing colon cancer frequently contain one or more synchronous adenomas.

2. The risk for colon cancer increases markedly with increasing number of adenomatous polyps within the colon.

4. Patients with FAP who have hundreds or thousands of adenomatous colonic polyps inevitably develop colon cancer if colectomy is not performed.

5. Patients who have adenomatous polyps larger than 1 cm diagnosed by barium enema who do not undergo colonoscopic polypectomy develop colon cancer at a rate of 1% to 1.5% per annum.

The adenoma-carcinoma model may explain a large portion of carcinoma development, however there are many carcinomas which do not exhibit the above mutations, and the model is not universally applicable to all carcinomas.

Although many hyperplastic polyps have little or no association with colon cancer, a minority appear to be linked to colon cancer by means of the recently reclassified (sessile) serrated adenoma, which previously was characterized as a hyperplastic polyp (serrated adenoma-to-carcinoma theory). Serrated adenomas arise within hyperplastic polyps but differ from ordinary hyperplastic polyps by abnormal proliferation of crypt epithelium and by nuclear atypia. They appear to transform into colon cancer via a different pathway from that of conventional adenomas and to result in a recognizably different form of colon cancer. With differentiation of serrated adenomas (high-risk) from hyperplastic polyps in the reclassified nomenclature, the remaining conventional hyperplastic polyps are believed to harbor a negligible risk for developing colon cancer.

Unlike conventional adenomas, serrated adenomas frequently have BRAF genetic mutations and exhibit extensive DNA methylation but lack adenomatous polyposis coli (APC) gene mutations. The serrated adenoma is a precursor lesion of colorectal carcinoma with high microsatellite instability (MSI-H), which constitutes approximately 15% of sporadic colon cancer. The specific genetic defects responsible for the serrated adenoma are, however, unknown.

Evidence supporting the theory that serrated adenomas can transform to cancer:

1. Serrated adenomas share the same genetic mutations characteristic of sporadic MSI-H cancers, suggesting a common molecular pathway.

2. Serrated adenomas are sometimes found contiguous to areas of severe dysplasia, suggesting that the dysplasia arose from this precursor lesion.

3. Patients with hyperplastic polyposis who have 30 or more hyperplastic polyps distributed throughout the colon (or at least five hyperplastic polyps proximal to the sigmoid colon, at least two of which are greater than 1 cm in diameter) frequently have serrated adenomas and frequently develop colon cancer.

4. Pathologic studies have demonstrated MSI-H colorectal cancers develop in the same area of the proximal colon where serrated adenomas were previously identified by colonoscopy.

FAP is inherited as a classic Mendelian single autosomal dominant gene and results in germline mutation of the APC gene located on chromosome 5q, which is carried one allele in all somatic cells, including colonocytes. Colonic adenomas form when the second APC allele is lost or undergoes mutation in an individual colonocyte. Studies of this uncommon syndrome have led to breakthroughs in understanding the molecular basis of the transformation of sporadic adenomas to colon cancer. Patients develop hundreds or thousands of adenomatous polyps throughout the colon beginning after puberty and inevitably develop colon cancer. In attenuated FAP, APC mutations occur at certain sites, particularly the extreme proximal or distal ends of the APC gene.

In contrast, patients with hereditary nonpolyposis colon cancer (HNPCC) typically have only a few colonic polyps, often in the right colon, which transform to cancer in middle age. It results from mutations of one of the mismatch repair genes, such as hMLH1, hMSH2, and hMSH6. It is clinically diagnosed with the Amsterdam II criteria:

Our current understanding of colon cancer pathogenesis reveals a series of genetic mutations leading to progressively disordered local DNA replication and accelerated colonocyte mitosis, accumulation of which results in transition from normal mucosa frank carcinoma. Approximately 15% of sporadic colon cancers result from dysfunctional mismatch repair genes. In the HNPCC syndrome, the mismatch repair genes malfunction because of genetic mutation, while in sporadic serrated adenomas, the cause is often DNA hypermethylation.

Spontaneous somatic APC mutation in colonocytes underlies the development of sporadic adenomatous polyps and accounts for 80-85% of sporadic colon cancers. These mutations occur early in adenoma development and often are found in aberrant crypt foci, the earliest identifiable dysplastic crypts. Overall, these adenomas generally remain benign, and malignant transformation requires further genetic alterations. Specific mutations of k-ras activate a signal pathway from the cell membrane to the nucleus to promote colonocyte replication, and are associated with exophytic growth of adenomas in the transition to carcinoma. Approximately 50% of sporadic adenomas contain APC mutations, while approximately 50% of colon cancers demonstrate k-ras mutations.

The p53 gene is responsible for arresting the cell cycle after DNA injury to facilitate either repair or apoptosis. Absence of this function results in unrestrained replication of genetic errors, leading to genomic instability and resulting in loss of heterozygosity (LOH). This has been found to promote the progression of advanced adenoma to frank carcinoma. Approximately 50% of colonic lesions with high-grade dysplasia and approximately 75% of frank cancers exhibit p53 mutations.

The DCC gene encodes for a pathway that promotes apoptosis and suppresses tumours. Loss of the normal DCC gene is important in the transition from an intermediate to an advanced adenoma. DNA hypermethylation can inactivate suppressor genes, promoting cancer. 25% of colon cancers are associated with methylation and inactivation of p14 (an inducer of the p53 pathway), which produces the same cancer phenotype as p53 mutation. Methylation of the tumour suppressor gene p16, occurs in approximately 35% of colon cancers.

Colon cancers are classified as well differentiated, moderately well differentiated, or poorly differentiated based on the degree of preservation of normal glandular architecture and cytologic features. 20% of colon cancers are poorly differentiated and have a poor prognosis. 15% are classified as mucinous, or colloid, because of prominent intracellular accumulation of mucin, which may become so abundant that the cells’ nuclei are displaced peripherally, resulting in the “signet ring” variant of mucinous carcinoma, which is markedly aggressive and has a poor prognosis.

Colon cancer associated with HNPCC has unusual histopathologic features, such as mucinous differentiation, prominent lymphocytic reaction, and a medullary growth pattern. The medullary form of colon cancer, previously classified as an undifferentiated carcinoma, is characterized by sheets of eosinophilic and polygonal cells heavily infiltrated with small lymphocytes and devoid of glandular elements.

Other cancers of the colon are rare. Kaposi’s sarcoma can involve the colon as part of disseminated disease with AIDS. Non-Hodgkin’s lymphoma of the colon is rare and may also be AIDS-associated. Carcinoid of the lower GIT usually occurs in the rectum or appendix but rarely may present in the rest of the colon.

Colon cancer can occur in a pedunculated polyp, sessile polyp, mass, or stricture. Small polyps rarely contain cancer. Only approximately 1% of diminutive polyps contain cancer. Cancer in a sessile polyp may metastasize earlier than cancer in a pedunculated polyp because of closer proximity to the lymphatic drainage and flat lesions may be more biologically aggressive than pedunculated polyps because of cellular growth into the colonic wall rather than the colonic lumen.

Flexible sigmoidoscopy results in substantial reduction of the incidence of left-sided colon cancer but reduces the incidence of right-sided colon cancer negligibly, indeed the relative frequency of right-sided colon cancer has gradually increased during the past several decades, which has been attributed to decreased frequency of left-sided cancer resulting from polypectomy of premalignant left-sided polyps at flexible sigmoidoscopy.

TX – primary tumor cannot be assessed.
T0 – no evidence of primary tumor.
Tis – carcinoma in situ: intraepithelial or invasion of lamina propria.
T1 – tumor invades submucosa.
T2 – tumor invades muscularis propria.
T3 – tumor invades through muscularis propria into subserosa or into nonperitonealized pericolic or perirectal tissues.
T4 – tumor directly invades other organs or structures and/or perforates visceral peritoneum.

NX – regional lymph nodes cannot be assessed.
N0 – no regional lymph node metastasis.
N1 – metastasis in 1–3 regional lymph nodes.
N2 – metastasis in 4 or more regional lymph nodes.

MX – distant metastasis cannot be assessed.
M0 – no distant metastasis.
M1 – distant metastasis.

Colon cancer spreads by local invasion to contiguous organs or distantly by lymphatic or hematogenous invasion. Carcinoma in situ, or high-grade dysplasia, implies cancer that is confined to the mucosa without penetration of the muscularis mucosa, and is unlikely to metastasize because the lymphatic and vascular channels are below the muscularis mucosa. Traditionally, invasive colon cancer has been staged according to the Dukes’ classification, later modified by Astler and Coller, with stage A penetrating beyond the muscularis mucosa into the submucosa. Stage B1 extends beyond the submucosa into the muscularis propria; stage B2 extends through the muscularis propria into the serosa; stage C has regional lymph node metastases; and stage D has distant metastases.

Currently, the most oncologically relevant system is the 4-stage TNM classification, although staging by either scheme correlates highly with cancer prognosis. This classification is helpful particularly in endosonographic staging of colon cancer.

Stage I in the TNM classification corresponds to Dukes’ A or B1 lesions, stage II corresponds to Dukes’ B2 lesion, stage III corresponds to Dukes’ C lesion, and stage IV corresponds to Dukes’ D lesion. 20% of patients initially present with Dukes’ D colon cancer, with identified metastases. Another 30% of patients have no metastases detected preoperatively or intraoperatively but eventually develop gross recurrence, presumably from initially undetected micrometastases after apparently curative surgery. The most common sites of gross metastases are the regional lymph nodes and liver. Colon cancer metastasizes early to the liver because of venous drainage of the colon via the portal system. Other sites, including the lungs, peritoneum, pelvis, and adrenals, typically become involved only after hepatic or lymphatic metastases occur. In contrast, rectal cancers which originate below the peritoneal reflection penetrate early into adjacent pelvic structures due to the lack of a serosa.

Symptoms and signs of colorectal cancer are vague and nonspecific, with overt clinical features usually appearing only with advanced disease. Conversely, when typical signs and symptoms do occur, they are usually due to diseases other than malignancy. In a large review of patients who had colorectal polyps, about two-thirds were asymptomatic, and many of the others had symptoms that were likely coincidental and not caused by the polyps. The most common symptoms attributable to polyps are rectal bleeding, abdominal pain, and change in bowel habits, however, rectal bleeding in a patient who has a small colonic polyp is more likely caused by other conditions, especially haemorrhoids. Large polyps rarely form the leading edge of a colonic intussusception. Polyps >1 cm in diameter are more likely to produce symptoms, while polyps < 0.5 cm rarely produce symptoms. Rectal polyps may be palpable by digital rectal examination. Much less than half of adenomas cause faecal occult blood, and large adenomas are much more likely than small adenomas to do so. A benign colonic polyp rarely causes iron deficiency anemia; iron deficiency anemia is more common with a malignant polyp because of greater chronic blood loss. Small adenomas are believed to grow very slowly, unless they undergo malignant transformation, and it is estimated that only 5% of adenomas progress to colon cancer. Risk factors for such progression include advanced adenomas with significant dysplasia, diameter >1 cm, and villous features.

Serrated adenomas are relatively uncommon, accounting for 0.6% of 18000 colonic polyps in a 1990 pathologic study. The revised histologic criteria, however, may result in 18% of polyps originally classified as hyperplastic being reclassified as serrated adenomas. Serrated adenomas frequently are right-sided and sessile. They may be flat and therefore difficult to identify at conventional colonoscopy using white light. The clinical presentation of serrated adenomas is largely unstudied and unknown. It is likely that most serrated adenomas are asymptomatic and do not produce fecal occult blood. Studies to date have not stratified and differentiated serrated adenomas from ordinary hyperplastic polyps, but in studies of hyperplastic polyps, < 5% produced guaiac positive stool.

Symptoms depend on cancer location, cancer size, and presence of metastases. Left colon cancers are more likely than right colon cancers to cause partial or complete intestinal obstruction because the left colonic lumen is narrower and tends to contain better formed stool due to reabsorption of water in the proximal colon [6]. Large exophytic cancers also are more likely to obstruct the colonic lumen. Partial obstruction produces constipation, nausea, abdominal distension, and abdominal pain. Partial obstruction occasionally and paradoxically produces intermittent diarrhoea as stool moves beyond the obstruction. Distal cancers sometimes cause gross rectal bleeding, but proximal cancers rarely produce this symptom because the blood becomes mixed with stool and chemically degraded during colonic transit. Bleeding from proximal cancers tends to be occult, and patients may present with iron deficiency anemia without gross rectal bleeding. The anemia may produce weakness, fatigue, or palpitations. Advanced cancer, particularly when metastatic, can cause cancer cachexia [6], characterized by a symptomatic tetrad of involuntary weight loss, anorexia, muscle weakness, and a feeling of poor health.

Less common symptoms include nausea and vomiting, malaise, anorexia, and abdominal distention. Although colon cancer can present with diarrhea or constipation, a recent change in bowel habits more likely is from colon cancer than chronically abnormal bowel habits. Physical signs are often non-specific, occur later in the disease and encompass features of chronic anaemia, malnutrition, hypoalbuminaemia, Gl obstruction, or rarely an abdominal mass. Other uncommon features include a palpable Virchow’s lymph node in the left supraclavicular space, and hepatomegaly from hepatic metastases.

50% of patients with colon cancer are anaemic, however anaemia is common, so only a small minority of patients who have anemia have colon cancer. Iron deficiency anemia of unknown origin, particularly in the elderly, warrants evaluation for colon cancer. Advanced colon cancer may result in hypoalbuminemia from malnutrition. Liver enzymes are usually are within normal limits in patients who have colon cancer, however ALP and LDH levels may increase with hepatic metastases.

Carcinoembryonic antigen (CEA) is a glycoprotein oncofoetal tumor-associated antigen first described in 1965, expressed by more than 90% of colorectal adenocarcinomas, but not increased in the serum of more than 90% of patients. Serum levels can be easily measured, however the biological function of CEA remains unclear. The CEA level is not sensitive enough for screening, however preoperative testing is useful for prognosticating and as a baseline for comparison with postoperative levels. An elevated preoperative CEA is an indicator of poor prognosis, correlating with extent of disease and risk of recurrence; however this almost always normalises with apparently complete colon cancer resection; thus failure to normalize postoperatively suggests incomplete resection, and a sustained and progressive rise after postoperative normalization strongly suggests cancer recurrence. Patients who have this finding require prompt surveillance colonoscopy to exclude colonic recurrence and abdominal imaging to exclude metastases. Smokers may have an elevated CEA without malignant disease and smoking may affect the accuracy of CEA results. The normal CEA range is < 2.5 ng/ml (depending on the assay used); Levels > 10 ng/ml suggest extensive disease and levels > 20 ng/ml suggest metastatic disease.

Less frequently used markers include CA 19-9 which is used to monitor gastrointestinal post-resection for recurrence and is non-specific for colorectal cancer; and CA 195, which is rarely used to detect colon cancer - changing levels may indicate progression or regression of tumour load.

Patients with colon cancer can present with acute large-bowel obstruction, most commonly from exophytic intraluminal growth and uncommonly from intussusception or volvulus. Features include abdominal pain, nausea and vomiting, obstipation, abdominal tenderness and/or distention, and decreased bowel sounds. Colon cancer rarely causes ischemic colitis due to dilatation proximal to malignant obstruction or malignant infiltration of blood vessels. It may present with acute perforation and generalized peritonitis and rarely can perforate slowly to form a walled-off inflammatory mass or abscess with localized peritoneal signs. It may penetrate into adjacent organs, such as the bladder or small bowel, leading to fistula formation. Factors promoting colonic perforation include disruption of mucosal integrity resulting from transmural malignant extension or colonic ischemia and increased intraluminal pressure resulting from colonic obstruction. Colonic obstruction or perforation is a poor prognostic indicator. Colon cancer occasionally causes gross rectal bleeding because of cancerous mucosal ulceration. Approximately 6% of metastatic adenocarcinomas with an unknown primary eventually are shown to arise from the colon.

Adenomas are more common in men and their incidence increases with patient age. They are rare in patients younger than 40 years old, except in patients who have genetic syndromes. Classically, about 15% to 25% of asymptomatic patients age 50 or older have adenomas detected at screening colonoscopy, although autopsy studies reveal prevalences up to 33%, rising to 50% in patients more than 70 years old. Adenomatous polyps are relatively asymptomatic, early colon cancer has few signs and symptoms but may be curable, while advanced cancer has more easily detectable signs and symptoms but is often incurable. This provides the rationale for mass screening of the general population before symptoms or signs occur, in an attempt to detect and remove premalignant premalignant lesions before malignant transformation can occur, and to detect established cancer at an early stage. Colorectal cancer screening is a rapidly evolving field because of breakthroughs in both the understanding of disease pathophysiology and in the technology for colon cancer screening and therapy. Established screening and surveillance colonoscopy regimens can prevent mortality from colorectal cancer, yet large proportions of eligible patients tend to go unscreened worldwide. Patients refuse screening because of embarrassment, fear of potential complications, reluctance to undergo invasive tests when asymptomatic, denial, and potential economic costs, and the majority of primary physicians in multiple countries studied tend to not refer patients appropriately for screening. Patients thereafter tend to become progressively harder to recruit for colon cancer screening tests as the more compliant patients have already undergone screening. The residual unscreened patients tend to be poor or ethnic minorities who have limited access to health care.

Faecal occult blood testing (FOBT) has previously been the mainstay of screening for colon cancer and colonic polyps. It is a colorimetric assay of a guaiac-based reaction catalyzed by the pseudoperoxidase present in blood, and is based on increased microscopic rectal bleeding in patients who have colon cancer compared with patients without colonic disease. Patients who have colon cancer, however, have a range of microscopic bleeding with considerable overlap with normal controls, which gives the test only moderate specificity. Specificity may be increased by avoiding ingestion of broccoli, cauliflower, or red meats and by discontinuing aspirin therapy for 3 days before the test. Sensitivity is improved by performing stool tests on three different occasions, because colon cancer typically bleeds intermittently; avoiding ascorbic acid for several days before the test, because ascorbic acid inhibits the guaiac reaction; and performing the test on fresh stool.

It is currently unclear whether supplemental iron causes a falsely positive FOBT, however most centres offering the test advise withholding iron therapy for several days prior. Even in ideal research studies, only 5% to 10% of patients who have fecal occult blood have colon cancer, and another 20% to 30% have colonic adenomatous polyps. The sensitivity is particularly low for adenomas that are small or are located in the proximal colon. Although true-positive tests can lead to early colon cancer detection and cure, false-positive tests result in a large number of unnecessary colonoscopies. Despite these flaws, however, FOBT has remained a common traditional screening test because of test safety, simplicity, noninvasiveness, and low cost.

Despite the large number of false positives, it has been demonstrated that annual screening by FOBT, with referral for colonoscopy if the test was positive, reduces mortality from colon cancer. In one study of 46,551 patients, patients who had annual screening by FOBT had a 5.88 per 1000 cumulative mortality from colon cancer compared with an 8.83 per 1000 cumulative mortality in control subjects not undergoing such screening. Unexplained fecal occult blood mandates further evaluation of the colon to exclude colon cancer or polyps in any patient more than 40 years old.

The fecal immunochemical test (FIT) has been developed and extensively tested by the Japanese, and involves the use of antibodies to detect the globin protein within human haemoglobin in stool. These tests are more specific than guaiac tests, because they:

Patient participation in FIT is simple because of the absence of dietary restrictions and the ease of sampling by merely wiping a brush within the toilet bowl after defection. It does, however, require processing by a centralized laboratory and costs substantially more than guaiac tests.

FIT is significantly more sensitive than guaiac testing, and in one colonoscopic study of 2512 patients, detected 87.5% of cancers compared to 54.2% for guaiac testing, and 42.6% of advanced adenomas, compared to 23.0% for guaiac testing. Screening with FIT followed by colonoscopy for positive screening substantially reduces colon cancer mortality, and in one study of 42,150 Japanese patients followed for 13 years, mortality decreased by 70% in those patients screened by FIT compared with unscreened controls. Although the sensitivity of FIT for advanced adenomas is superior to that for guaiac testing, this rate is much lower than the 90% or higher sensitivity of screening colonoscopy, which is therefore superceding FIT.

Barium enema was a historically important alternative to colonoscopy, but its role is being superceded by virtual colonoscopy (CT colonography). It is only about 80% sensitive at detecting colon cancer and is much less sensitive at detecting colonic polyps. In a study of 580 patients undergoing both barium enema and colonoscopy, barium enema detected only 32% of colonic polyps less than 6 mm in diameter, and only about 50% of larger (at least 6 mm) colonic polyps. Barium enema also does not permit histologic characterization of an identified lesion because of an inability to perform biopsies, and it does not permit therapeutic removal of polyps. Thus detection of a moderate or large polyp at barium enema necessitates follow-up colonoscopy for polypectomy.

Sigmoidoscopy decreases mortality from rectosigmoid colon cancer, with a 60% reduction of rectosigmoid cancer in patients undergoing one or more rigid sigmoidoscopies in the prior decade compared with unscreened controls matched for age and sex. Patients who die from colon cancer have a much lower frequency of having undergone screening flexible sigmoidoscopy (10%) than age- and sex-matched controls without colon cancer (30%) (P<.05). Therefore flexible sigmoidoscopy every 3 to 5 years with annual FOBT has been recommended for screening.

Currently, however, its role is diminishing, because the proximal half of the colon, where up to 50% of cancer and polyps are located, is not visualized endoscopically, and proximal lesions usually do not have synchronous distal lesions. Reduction in colon cancer mortality is largely limited to left-sided cancers, and it is still inadequate for complete assessment of distal colon cancer, because 3% to 5% of these patients have a synchronous proximal cancer.

Colonoscopy is a highly specific test for colon cancer. At colonoscopy, polyps are removed and masses are biopsied for a pathologic diagnosis. Early colon cancer may occur in an adenomatous polyp and, therefore, may be difficult to distinguish by colonoscopy from a nonmalignant adenomatous polyp. For example, a 2cm wide villous adenoma has an approximately 40% chance of harbouring cancer. Polyp risk factors for malignancy include villous rather than tubular histology, large size, sessile morphology, and increasing number of colonic polyps. Advanced colon cancer typically appears as a large, exophytic mass because of intraluminal growth or as a colonic stricture because of circumferential growth. A colonic stricture, however, may be benign. Malignancy is suggested when a colonic stricture is ulcerated, indurated, asymmetric, and friable and has irregular or overhanging margins. The colonoscopic appearance of a stricture is suggestive but not definitive. Pathologic examination of multiple colonic biopsies and cytologic analysis of stricture brushings usually are diagnostic.

Colonoscopy is recommended for screening of patients more than 50 years old at average risk for colon cancer or colonic polyps. Colonoscopy is highly sensitive at detecting large (greater than 1 cm) colonic polyps, with a miss rate of only 6%, and it is moderately sensitive at detecting diminutive (less than 0.6 cm) polyps, with a miss rate of about 27%. Colon cancers tend to be larger than adenomatous polyps and rarely are missed. Colonoscopy is also highly specific, and has the advantage of polyp removal and biopsy of masses.

The adenoma-to-carcinoma sequence strongly suggests that screening colonoscopy with polypectomy of adenomas should substantially prevent colon cancer, which has been strongly supported by clinical trials. In the National Polyp Study, 699 patients underwent surveillance colonoscopy at 1, 3, and every 2 subsequent years after detecting at least one adenomatous polyp at an index colonoscopy, and had a 76% to 90% decline in the incidence of colon cancer compared with historical reference groups, with all cancers being detected early. Disadvantages of colonoscopy as a screening test are that it is resource intensive, expensive, somewhat invasive, and uncomfortable, it entails a small, but significant, risk of serious complications, and requires a team, including a technician, nurse, and highly trained colonoscopist. Colonic preparation and dietary restrictions are necessary for 24 hours before the test. Diagnostic colonoscopy has about a 0.4% complication rate the most common of which are GI bleeding and colonic perforation.

Identification of premalignant or potentially malignant polyps has become more complicated recently with the realization that serrated adenomas, which are often flat and inconspicuous, have a risk of malignant transformation equivalent to, or potentially greater than, that of conventional adenomas, and the realization that conventional colonoscopy frequently misses colonic polyps. The detection rate of adenomas varies greatly among experienced colonoscopists (3-fold at the Mayo clinic and up to 10-fold in other studies). Small polyps, which are frequently missed, occasionally exhibit advanced histology or even harbour frank cancer, and occasionally even large polyps are missed.

1. A properly prepared colon is essential for colonoscopy. 4 L of polyethylene glycol solution should be drunk over 4 hours. Fleet’s phospho-soda can precipitate renal failure or electrolyte abnormalities, and should be used with extreme caution in patients who have even mild renal insufficiency, always with background fluid support.

2. Adequate sedation is a prerequisite for colonoscopy. Propofol provides deeper sedation and a faster recovery time than conventional sedation with benzodiazepines and opiates, and may be useful in selected patients. A large colonoscopic study has shown gastroenterologists and nurses can administer propofol safely.

3. Confirmation of cecal intubation is essential. Missed polyps that are not removed at an index incomplete screening colonoscopy may continue to grow without colonoscopic intervention for 10 years until the next recommended screening colonoscopy, which may be a reason for early interval cancers after reportedly normal index colonoscopy. Depth assessment by ruler markings are unreliable as the tubing may be looped, and caecal landmarkds should be used for confirmation: the appendix appears as a slit-like, semilunar, or ovoid opening near the caput of the cecum on the (medial) colonic wall closest to the middle of the ileocecal valve; the ileocecal valve is a characteristic round, smooth, soft, homogeneous mound lying on a prominent fold on the medial colonic wall just distal to the caput of the cecum; air or stool may be seen entering the colon via the ileocecal valve; the ileocecal valve may contain a central dimple above its opening, and may appear yellowish due to fatty infiltration - because it is perpendicular to the colon, it is not seen directly, but may be intubated blindly; the merger of the three teniae coli at the cecal base produces a characteristic appearance; transillumination with the colonoscopic light source during cecal intubation may demonstrate light visible in the deep right lower quadrant of the abdomen; external manual compression in the same abdominal region may assist with localising the colonoscope; a purplish-blue are on one side of the colonic wall at an acute turn may be found at the hepatic and splenic flexures.

5. A withdrawal time of at least 6 min is recommended to ensure adequate colonic examination and to maximize the detection of adenomas. The detection rate of polyps correlates directly with the mean time of colonoscopy, and in a prospective study of 7882 colonoscopies among 12 colonoscopists, colonoscopists who had a mean withdrawal time greater than 6 min had, compared with colonoscopists who had a shorter withdrawal time, a significantly higher detection rate of neoplastic polyps (28.3% versus 11.8%, P<.001), and of advanced neoplasia (6.4% versus 2.6%, P<.005;).

Colonoscopic maneuvers can improve polyp detection around sharp turns, areas of colonic spasm, especially in the sigmoid colon or with severe diverticulosis; and behind large folds. Such areas should be reintubated beyond the poorly visualized area as required. Colonic spasm is avoided by not irritating the colon by meticulous colonoscopic technique that avoids excessive colonic air insufflation, colonic loops, and traumatic intubation. It occurs in areas of severe colonic diverticulosis, particularly in the sigmoid colon where the lumen is relatively narrow even without diverticulosis. It may be addressed by gentle air insufflation to distend the spastic area or instillation of warm water. Areas behind a fold not visualized in the orthograde (unretroflexed) position can be viewed by an experienced colonoscopist retroflexing the colonoscope.

New technologies that enhance the detection rate of colonic polyps include chromoendoscopy and narrow band imaging. In chromoendoscopy a vital dye, either methylene blue or indigo carmine, is sprayed on the colon via a catheter in the instrument channel to highlight areas of abnormal mucosa before the area of the colon is examined. It may be useful in detecting diminutive adenomas. High magnification chromoendoscopy combines chromoendoscopy with an advanced colonoscope that provides zoom optics with up to 100 times magnification. Chromoendoscopic spraying highlights the colonic pits because of their retention of the dye. The high magnification then permits detailed analysis of the anatomy of the superficial colonic pits. Kudo has classified six colonic pit patterns, labeled I, II, III s, III l, IV, and V:

Type II pits are stellate or papillary. Type I and II are characteristic of ordinary hyperplastic polyps. Type III s pits are tiny and round, while type III l pits are large and tubular.

Type IV pits resemble the sulci or gyri on the surface of the cerebral cortex. Adenomas typically exhibit type III l or type IV pits.

Type V pits are typical of advanced neoplasia. Several studies have shown that pit pattern analysis is reliable at histologic classification of colonic polyps, with sensitivities of 90-95% and specifitities of 72-87%, which is much higher than conventional colonoscopy.

Narrow-band imaging involves the use of specific fractions of the light spectrum; however recent studies have suggested that this may not offer improved visualisation of lesions compared to white light, and may be superceded in clarity by newer high-definition colonoscopes. Other techniques for histologic characterization of polyps include autofluorescence, and confocal laser microscopy.

It is impossible to accurately determine polyp histology solely by colonoscopic criteria, therefore most units have adopted a general policy of removing all polyps detected at colonoscopy. With improved technology, polyp detection is increasing, and research into guidelines for determining polyp histology based solely on colonoscopic criteria is currently underway. Polyps are currently characterized at colonoscopy according to size, color, number, segmental location, intramural location (mucosal versus submucosal), presence or absence of a stalk (pedunculated versus sessile), and superficial appearance, which provides important clues concerning polyp histology and malignant potential.

Colonoscopic appearance, however, only moderately correlates with polyp histology.

Pathologic examination of a colonoscopic biopsy provides an indication of polyp histology, but is subject to sampling error. A polyp is definitively classified by pathologic examination of the entire polyp after polypectomy. Colonoscopic polypectomy is diagnostic and therapeutic for noncancerous adenomatous polyps.

The differential diagnosis of numerous polypoid colonic masses detected at colonoscopy includes familial adenomatous polyposis (FAP), attenuated FAP, hyperplastic polyposis, juvenile polyposis, Peutz-Jeghers syndrome, pseudopolyposis, diffuse colonic hemangiomatosis, and pneumatosis coli. These conditions are differentiated by clinical, radiologic, colonoscopic, and histologic findings.

In patients with FAP, the colonic mucosa is carpeted by hundreds or thousands of adenomatous polyps. In patients with attenuated FAP, only about thirty adenomatous polyps are present, usually in the proximal colon, which tend to be flat growths because of intramural, rather than intraluminal, growth.

Hyperplastic polyposis is characterized by 30 or more polyps in the colon, a predominantly right-sided polyp distribution, and a positive family history.

Juvenile polyposis is characterized by a family history of juvenile polyposis, more than five juvenile polyps in the colon, multiple juvenile polyps throughout the rest of the GI tract, and polyp development at a young age. This syndrome is associated with a greatly increased risk of colon cancer.

In Peutz-Jeghers syndrome, multiple hamartomatous polyps, which characteristically contain abundant branching smooth muscle, occur throughout the GI tract. Patients characteristically have perioral and oral hyperpigmentation caused by melanin deposition. This syndrome is associated with a greatly increased risk of colon cancer. Pseudopolyps represent islands of variably inflamed residual mucosa surrounded by a background of previously sloughed off mucosa. It is associated most commonly with ulcerative colitis. Other colonoscopic findings of ulcerative colitis, including mucosal erythema, granularity, blunting of the normal vascular pattern, friability, mucopus, mucosal hemorrhage, and superficial ulcerations may be present.

At colonoscopy, hemangiomas often appear as multiple, violet-blue, sessile, polypoid lesions. They are associated with characteristic dermatologic lesions in the blue rubber bleb nevus syndrome. In pneumatosis coli, multiple air-filled cysts are present in the colonic submucosa. Colonoscopy reveals multiple, pale, cystic, round polypoid masses with overlying intact mucosa.

Early colon cancer may occur in an adenomatous polyp and be difficult to distinguish by colonoscopy from a nonmalignant adenomatous polyp. For example, a 2 cm wide villous adenoma has an approximately 40% chance of harboring cancer. Polyp risk factors for malignancy include villous rather than tubular histology, large size, sessile morphology, superficial ulceration, multinodularity, depressed lesions, and increasing number of colonic polyps. The colonoscopic nonlift sign is associated with infiltrative malignancy. Advanced colon cancer typically appears as a large, exophytic mass because of intraluminal growth, or as a colonic stricture because of circumferential growth. Colonic strictures may be benign, and malignancy is suggested when a colonic stricture is ulcerated, indurated, asymmetric, and friable, and has irregular or overhanging margins. The colonoscopic appearance is not definitive, and pathologic examination of multiple colonic biopsies and cytologic analysis of stricture brushings are usually diagnostic. Atypical polyps or flat lesions should be treated aggressively, however, to exclude a dysplasia associated lesion or mass (DALM). Polypectomy should not be performed on a lesion exhibiting the characteristic findings of an inverted diverticulum because of the risk of colonic perforation.

Large polyps that are likely or obviously malignant should be sampled extensively by multiple biopsies to increase the diagnostic yield, but not removed in toto by polypectomy at an initial colonoscopy to avoid the extra risks of polypectomy when cancer surgery likely will be required subsequently

Endomucosectomy or endoscopic mucosal resection (EMR) combines the conventional snare polypectomy with submucosal injection to cut through the middle or deep submucosa and remove lesions lying in the deep mucosa or in the submucosa. It is being used increasingly to remove colorectal lesions, particularly large adenomas, as it provides an alternative to surgery for deeper superficial lesions without evident penetration of the deep muscle layer, regional lymph nodes, or distant metastases. Sessile villous adenomas, adenomas with carcinoma in situ (T0 lesions), and early (T1N0M0) cancers invading the submucosa are appropriate. The tumour is characterized by endoscopy, sampled by endoscopic biopsy, and locally staged by endosonography before endomucosectomy is considered. Patients are evaluated for the suitability of endomucosectomy based on tumor size, endoscopic appearance, pathology of the initial endoscopic biopsy, and the estimated depth of tumor penetration (T stage). Endomucosectomy usually is applied to polypoid (protruding) lesions, but sometimes can be applied to flat or even minimally depressed lesions provided the above criteria are satisfied. Endomucosectomy is advantageous over endoscopic ablative therapy (using argon plasma coagulation or photodynamic therapy) because the entire treated specimen is removed and available for histologic analysis and pathologic staging. It has lower morbidity and mortality than surgical resection.

The technique involves deep submucosal injection of normal saline to permit deep resection of the submucosa. Lesions that lifts during submucosal injection is amenable to endomucosectomy; those that do not (nonlift sign) have a high risk of invasive carcinoma

Colon cancer incidence and survival has improved only moderately during the past two decades despite the manifest efficacy of colonoscopic polypectomy at cancer prevention. This failure is caused by insufficient implementation of colonoscopy screening partly because of the expense, invasiveness, discomfort, and risks of colonoscopy. New simpler, less invasive, and safer tests are being designed to overcome these barriers to universal screening for colon cancer.

DNA from colon cancer is shed into the fecal stream in greater quantities than DNA from normal colonic mucosa, and is degraded with normal cell breakdown over time or by contiguity with stool during colonic passage. PCR amplification can detect minute quantities of cancerous DNA in stool specimens with a sensitivity from 52% to 91%. Detected cases tend to have histologically advanced and clinically symptomatic colon cancer. The sensitivity of detecting adenomas is much lower than that for colon cancer (13% in a large prospective colonoscopic study). Genetic stool screening has the potential test advantages of noninvasiveness and user friendliness, but requires improvement and further testing in large clinical trials. The key requirement for colon cancer screening tests is a high sensitivity in detecting adenomas. In the future, molecular genetics may play other important roles in managing sporadic colon cancer, including:

In virtual colonoscopy, CT images are obtained in the prone and supine positions during a prolonged breath hold, and then reformatted into two-dimensional images in the three orthogonal (axial, sagittal, and coronal) planes, or reconstructed into three-dimensional endoluminal (virtual colonoscopy) images that simulate the conventional colonoscopic view. This avoids many of the undesirable features of colonoscopy of test invasiveness, patient discomfort, and need for sedation and analgesia, and entails fewer risks. Application of digital subtraction technology to stool and administration of oral contrast to tag stool, however, may obviate the need for oral laxatives.

CT colonography is relatively safe, The risk of colonic perforation from colonic insufflation with air is small but significant (1 per 1700 studies).

A single CT colonography, using the most modern machines and a low-dose radiation protocol, exposes the patient to about half the radiation dose of a conventional barium enema. The accuracy of virtual colonoscopy is controversial, with wide discrepancy between various studies Because of the inability to acquire biopsies or institute management with virtual colonoscopy, detection of a polyp or mass at virtual colonoscopy currently requires follow-up colonoscopy for polypectomy or biopsy.

This test involoves swallowing a videocapsule, which traverses the gut by peristalsis to provide a wireless endoscopy by radiofrequency transmission. The videocapsule contains a miniaturized image capturing system, battery, light source, and transmitter, all of which are contained within an 11x30mm capsule. It is attractive as a screening test for colon cancer because of examination simplicity, noninvasiveness, minimal patient discomfort, and apparent relative safety. Its limitation is the short battery-life of 8 hours, insufficient to view the entire colon. When this is corrected, videocapsule endoscopy could provide an initial screening examination of the colon, with positive tests being referred to endoscopy. Videocapsule endoscopy had a sensitivity of 77% and specificity of 70%, compared with colonoscopy.

Because of the importance of high sensitivity and relative specificity of a screening test to prevent cancer, colonoscopy has emerged as the screening test of choice. Professional organizations generally have not endorsed virtual colonoscopy (CT colonography) for colon cancer screening because of highly variable and conflicting data about test sensitivity. Patients at average risk for colonic adenomatous polyps or cancer undergo screening colonoscopy every 10 years, or alternative screening tests at periodic intervals.

Patients who are members of high-risk groups undergo periodic surveillance more frequently, with colonoscopy as the recommended test. The age of beginning surveillance and the frequency of surveillance depend upon the age of onset of the increased cancer risk and the quantitative risk of cancer. Aside from periodic screening or surveillance, patients require colonoscopy to exclude colon cancer, adenomatous polyps, or other colonic diseases for specific indications.

The most prominent environmental and demographic risk factors for colon cancer are obesity, physical inactivity, alcoholism, smoking, and a diet that is high in fats or low in fruits and vegetables. Environmental factors presumably modulate the risk for genetic mutations responsible for colon cancer, although the precise molecular mechanisms currently are unknown. Dietary fiber may reduce the risk for colon cancer, but this effect is somewhat controversial. Proposed mechanisms include decreased mucosal exposure to intraluminal carcinogens resulting from stimulated intestinal transit, decreased concentration of carcinogens in stool due to increased stool bulk, increased concentrations of anticarcinogenic short-chain fatty acids, and stabilization of insulin levels due to delayed starch absorption that otherwise might promote colonic carcinogenesis. The identification of environmental risk factors is potentially important clinically because modification or elimination of an identified risk factor could lower the cancer risk.

Environmental risk factors, however, are believed to exert their effects after long-term (chronic) exposure for decades. This may explain the weak protective effects of most reported experimental environmental interventions, such as dietary modifications, that last only several years.

Nonsteroidal anti-inflammatory drugs (NSAIDs) reduce cellular proliferation, slow cell cycle progression, and stimulate apoptosis, and are believed to reduce adenoma formation and inhibit colon cancer development by inhibiting the cyclooxygenase enzymes required for the synthesis of prostaglandin E2, which in turn promotes tissue inflammation, cellular proliferation, and tumor growth. NSAIDs also may retard carcinogenesis by effects on cell adhesion and apoptosis. Case-control and cohort epidemiologic studies also provide evidence of decreased adenoma incidence or decreased colon cancer mortality with chronic NSAID use, particularly use of aspirin. In a prospective study of more than 600,000 adults over a 6-year period, the relative mortality from colon cancer was approximately 0.6 in men and 0.58 in women who used aspirin 16 or more times per month compared with nonusers of the same gender. Women who take aspirin for at least 20 years have a relative risk of 0.56 for developing colon cancer compared with nonusers. Chronic aspirin use is associated with a one-third reduction of the risk for adenomas compared with controls detected at follow-up colonoscopy, and a smaller, but still significant, reduction in the incidence of recurrent adenomas at colonoscopy.

Other NSAIDs, such as sulindac, seem to cause similar reductions in colon cancer or colon polyp incidence, although the effects are less well studied compared with aspirin. COX-2 is believed to mediate cell proliferation and tumor growth. Hence, selective COX-2 inhibitors may block adenoma formation and cancer development. Celecoxib, a selective COX-2 inhibitor, shows some promise in causing regression of colonic adenomas in patients who have FAP, leading to a 28% reduction in the mean number of rectal polyps. The effects of NSAIDs on sporadic adenomas generally are less dramatic. Although data support that NSAIDs inhibit colonic carcinogenesis, the optimal specific NSAID, NSAID dosage, and duration of treatment are unknown. The role of COX-2 selective inhibitors versus nonselective COX inhibitors needs to be analyzed and defined better.

CT is the standard modality to image the abdomen in patients who have CRC. It is relatively accurate at detecting liver metastases, with an accuracy of approximately 85%. CT is much more sensitive at detecting large than small hepatic lesions. It is only moderately accurate at T staging and the accuracy for T staging was only 74% in a large multicenter study. CT errors typically occur from underestimating the T stage. CT is approximately only 50% to 70% accurate in N staging of rectal cancer.

MRI is more accurate than CT in detecting focal liver metastases, particularly small metastases, from colon cancer because of the typically sharp contrast between metastatic lesions and the normal liver on MRI. Administration of contrast agents, such as superparamagnetic iron oxide, improves test sensitivity further. MRI also is more specific for hepatic metastases than CT. Hepatic metastases have a much shorter T2 sequence than hepatic hemangiomas or cysts. Hepatic metastases typically demonstrate rapid and strong enhancement with intravascular contrast due to enhanced vascularity but may enhance inhomogeneously due to hypovascular areas within metastases. Despite these advantages, CT is the standard test because of lower cost, greater machine availability, and more widely available expertise in image interpretation. MRI traditionally is reserved for characterizing ambiguous hepatic lesions detected by abdominal ultrasound or CT.

Endosonography has been used for T and N staging of rectal cancer because of the relative inaccuracy of CT for this staging. Preoperative evaluation of the T stage and the N stage has a great impact on the therapy for rectal cancer. Patients who have superficial cancer (T1N0) can be treated by local endoscopic or transanal resection without wide excision. Patients who have T2N0 lesions are treated surgically without preoperative adjuvant therapy. Patients who have deep intramural involvement (T3 or T4) or who have nodal involvement (N1 or N2) receive radiation and possibly chemotherapy before surgery. Patients who do not have rectal sphincter involvement may avoid a colostomy. Endoscopic ultrasound (EUS) is more accurate than CT or MRI for T staging, with an accuracy of 91% compared with 71% for CT (P = 0.02). Rectal endosonography has an approximately 85% accuracy for T staging which also compares favorably to the 70-80% of MRI. Tumors generally appear as homogeneous hypoechoic masses that disrupt the normal five-layer ultrasonographic structure of the rectal wall. Errors in endosonographic T staging may result from inflammation in tissue adjacent to cancer. Endosonography is more accurate for staging T1, T3, and T4 lesions than T2 lesions because of difficulty in assessing cancer invasion through the muscularis propria. It has 75% accuracy for N staging, which is greater than the 55% to 65% for CT and 60% to 65% for MRI. At endosonography, malignant lymph nodes tend to be large (>1 cm), to be hypoechoic, to have sharply demarcated borders, and to be round rather than ovoid or flat. Inflamed lymph nodes occasionally mimic these sonographic features.

Rectal ultrasound has become the standard preoperative imaging modality for local T and N staging of rectal cancer because of relatively high accuracy but has not yet been shown to prolong survival. The rectum easily is accessible via an ultrasound probe using a rigid probe inserted blindly or an echoendoscope inserted under endoscopic guidance and the procedure is safe. Endosonographic findings have been shown to result in modification of treatment plans. The accuracy of endoscopic ultrasound is operator dependent. Other factors affecting accuracy include the ultrasound frequency, with higher frequency improving the resolution but decreasing the depth of penetration; the tumor location, with reduced accuracy for tumors low in the rectum; and prior radiotherapy because of increased wall echogenicity after radiation. Locally recurrent rectal cancer is important to detect early so that patients can undergo salvage surgery for possible cure. EUS currently is the most reliable imaging study for detecting recurrence. The clinical benefit of early detection of rectal cancer recurrence, however, is limited by the low cure rate of salvage surgery. Colonic endosonography technically is more demanding and time consuming than rectal endosonography. Most patients who have colon cancer without distant metastases undergo colonic resection, regardless of T or N stage. Conventional endosonography produces a 2-D image. The recently developed 3-D endosonography may provide better spatial information that improves diagnostic accuracy.

Approximately 20–25% of patients with CRC will have liver metastasis at diagnosis, and as many as 60% of patients who develop metastases will have liver metastases. On average, 5-year survival for these patients is 2–8% without surgical resection, although with improved detection and treatment options, rates of death from CRC are decreasing. Surgery remains the treatment of choice for patients with resectable hepatic colorectal metastases; in patients who qualify for complete surgical resection, approximately 25–35% will be alive at 5 years. However, the majority (80–90%) of patients with hepatic metastases are poor candidates for surgery due to unfavorable location, size, or number of metastases; insufficient liver reserve; or extrahepatic disease.

Large phase 3 studies are lacking in neoadjuvant therapy of CRC liver metastases, however the impact of perioperative chemotherapy on survival is illustrated by a recent trial in patients with resectable gastric cancer, in which perioperative chemotherapy (epirubicin, cisplatin, and continuously infused 5-FU) conferred a 25% higher likelihood of overall survival (P = 0.009); a higher 5-year survival rate (36% versus 23%), and a 34% higher likelihood of progression-free survival (P < .001) than surgery alone, illustrating the potentially profound impact of perioperative

chemotherapy on a generally intractable tumour type. The EPOC study, published in 2007, showed improved progression-free survival with the use of neoadjuvant therapy over surgery alone. At a median follow-up of 3.9 years, resected patients receiving FOLFOX4 (6 cycles before

and 6 cycles after surgery) had an absolute survival advantage of 7.2% over patients receiving surgery alone (P = 0.058).

Postoperative, or adjuvant, systemic therapy has become routine and standard for stage III colon cancer, and should also be strongly considered in stage II patients. It is generally recommended for any medically fit patient who has stage II cancer with unfavourable factors, including colonic perforation, unfavorable histology, colonic obstruction, or lymphovascular invasion. The optimal choice of adjuvant chemotherapy has recently changed from a 6-month course of 5-fluorouracil (5FU)-based chemotherapy alone to a 6-month course of infusional 5FU plus leucovorin (LV) and oxaliplatin (FOLFOX) based on a large trial of adjuvant systemic therapy for resected stage II or III colon cancer, which demonstrated an increase in disease-free survival at 3 years from 72.9% to 78.2% (P = 0.002) with addition of oxaliplatin to FU/LV. Toxicities were comparable between the two groups, with the exception that oxaliplatin is associated with a much higher rate of paresthesia: 12.4% versus 0.2% grade 3 (serious) toxicity, which persisted at a grade 3 level in 1.1% of treated patients at 1 year of follow-up.

Approximately 20% or more of patients who have CRC initially present with metastatic disease (mCRC). Approximately 35% of patients who present with stage III, 20% of patients with stage II, and 5% to 10% of patients with stage I cancer eventually relapse and subsequently die from mCRC. The most common sites of metastases are the liver, lung, peritoneum, and retroperitoneum. Many advances have occurred recently in the treatment of mCRC. Active agents for mCRC in addition to the original 5FU include irinotecan, capecitabine, oxaliplatin, bevacizumab, cetuximab, and panitumumab. The goals of systemic therapy of mCRC include palliation of symptoms, prolongation of life, and, in selected cases of liver-only metastases, tumor regression to facilitate surgical resection of these metastases. The median survival of a patient who has mCRC has improved during the last decade from less than 1 year (with only 5FU-based therapy) to approximately 2 years (with multiagent systemic therapy).

The choice of appropriate first-line therapy for mCRC depends on their overall functional status, as well as their liver, kidney, and bone marrow function. Efficacy of chemotherapy for metastatic cancer is measured in terms of “response to therapy”, which is defined as a reduction, usually by some specified percentage, in cancer size. Reduction in the serum level of a tumor marker, such as CEA, can be confirmatory, but is an inadequate criterion. A partial response is defined as a 30% decrease in the longest dimension of each measurable tumor deposit, using unidimensional, or response evaluation criteria in solid tumours (RECIST), criteria. A complete response is complete disappearance of all clinically detectable disease. The response rate (RR) is the percentage of patients who meet either criterion. Measures used to determine the duration of benefit include: (1) progression-free survival, which is the time from the start of treatment to the date the disease worsens, and (2) overall survival, which is the length of time patients are alive after diagnosis. 5FU, often modified by LV, has been used for half a century as a standard agent for mCRC. Irinotecan was approved in 1996, and subsequently, oxaliplatin, capecitabine, bevacizumab, cetuximab, and panitumumab have also been approved. 5FU blocks the enzyme thymidylate synthase, which is essential for DNA synthesis. LV, also known as folinic acid, enhances the antineoplastic effects of 5FU. Both LV (FOL, folinic acid) and 5FU (F, fluorouracil) can be combined with irinotecan (IRI) or oxaliplatin (OX) with the treatment acronyms FOLFIRI or FOLFOX, respectively. These alternative treatments consist of administration of a bolus of 5FU, LV, and either oxaliplatin or irinotecan. The patient is then sent home with a 2-day infusion of low-dose 5FU, administered by a small, lightweight, portable pump, usually worn on a belt or shoulder strap, infused through a centrally placed catheter. The patient or health care provider can simply disconnect the catheter after the 2-day infusion.

Capecitabine is an oral fluoropyrimidine with a similar mechanism of action and efficacy as 5FU.

Irinotecan is a derivative of camptothecin, found in Camptotheca acuminata, a plant native to China. It potently inhibits topoisomerase I, an enzyme that facilitates the uncoiling and recoiling of DNA during replication by cleaving one strand and subsequently reattaching that strand. Oxaliplatin is a platinum chemotherapy that inhibits DNA replication and transcription by forming inter- and intra-strand DNA adducts/cross-links. In patients who have mCRC, optimal chemotherapy consists of initial administration of a fluoropyrimidine and oxaliplatin or irinotecan (eg, FOLFOX or FOLFIRI). Tournigand and colleagues and Colucci and colleagues performed randomized trials in which patients received either FOLFIRI followed by FOLFOX, or vice versa. In the Tournigand and colleagues study, FOLIRI was found to have an RR of 56% and an 8.5-month median progression- free survival (mPFS), whereas FOLFOX had an RR of 54% and an mPFS of 8 months. Colucci and colleagues found that FOLFIRI had an RR of 31% and FOLFOX had an RR of 34%. Both regimens had an mPFS of 7 months. Both investigators concluded that the regimens had similar efficacy when used as first-line therapy. Either FOLFOX or FOLFIRI can therefore be considered standard options for first-line treatment of mCRC. These regimens are typically given with bevacizumab.

Bevacizumab is a monoclonal antibody that binds to vascular endothelial growth factor ligand to inhibit angiogenesis. Its antineoplastic effect is ascribed to regression of microvascular density, inhibition of neovascularization, and normalization of grossly abnormal tumor vasculature that permits more effective chemotherapy delivery to the tumor. Addition of bevacizumab to irinotecan, FU, and LV for mCRC improves progression-free survival from 6.2 months to 10.6 months, improves the response rate from 35% to 45%, and improves overall survival from 15.6 to 20.3 months. Most recently, Saltz and colleagues reported on a randomized trial that found that the addition of bevacizumab to oxaliplatin-based chemotherapy significantly improved progression-free survival from 8.0 to 9.4 months. The addition of bevacizumab did not improve the response rate.

In 2004, the FDA approved cetuximab, the chimeric (human/mouse) monoclonal antibody targeting epidermal growth factor receptor (EGFR), for treatment of mCRC with irinotecan and as a single agent for patients intolerant of irinotecan-based therapy. In a single-arm, nonrandomized trial, cetuximab as a single agent had a 9% RR and a 35% rate of minor response or disease stability for at least 12 weeks. When cetuximab was combined with irinotecan, the response rate was 22.9% versus 10.8% for irinotecan alone. Cetuximab causes an acneform rash on the face and upper body in more than 80% of patients.

In 2006, the FDA approved panitumumab, a monoclonal antibody to EGFR, that unlike cetuximab is fully humanized (not chimeric). It is indicated for patients who have mCRC that has progressed on or following 5FU, oxaliplatin, and irinotecan-containing regimens. In a large randomized trial of panitumumab versus best supportive care for mCRC, a response rate of 8% was found. Like cetuximab, panitumumab causes an acneform skin rash. As a fully human monoclonal antibody, panitumumab entails a lower risk for serious infusion reactions than the 3% rate observed with cetuximab. The relative activity of cetuximab versus panitumumab and the relative activity of panitumumab when given with chemotherapy are currently unknown.

Many solid tumours, including CRC, are associated with obesity and the metabolic syndrome of hyperglycemia and insulin resistance. Paradoxically, malignancy and its various treatment modalities induce a catabolic state and impair the function of the GI tract, resulting in weight loss. These metabolic abnormalities increase surgical complication rates, including impaired wound healing and thromboembolism. An increased body mass index also presents challenges to the medical and radiation oncologist when calculating the ideal dose of therapy. Patients who have CRC should be provided with a customized dietary plan to maintain an ideal BMI. Such patients who consume a diet high in fruits, vegetables, and non-red meats have an improved outcome when compared with those who consume a traditional Western diet high in red meat and refined grains.

There is conflicting data in the literature suggesting that transfusion may increase the risk of recurrence on an immunological basis. Another explanation is that the need for transfusion may be related to the stage of disease, with patients with more advanced disease who are already at higher risk for recurrence requiring more transfusions. There is evidence that improving haemoglobin levels with pre-operative iron supplementation reduces need for perioperative transfusion and improves perioperative outcome. Management of anaemia at this stage should be on an individualized basis.

The use of mechanical bowel preparation for colonic surgery has been extensively studied, with current data suggesting no difference between preparing or not preparing the bowel. Evidence suggests increased rated of anastomotic breakdown/sepsis with poorly prepared bowel. There is no clear evidence to support orthograde washout over retrograde washout or vice versa. Since the colon is a significant reservoir of organisms which are potentially lifethreatening following translocation, there is probably specific value in bowel preparation of a subgroup of patients who are at risk for septicaemia and who are expected to have a prolonged ileus. The majority of polyethylene glycol preparations are isotonic and cause minimal fluid shifts, however sodium phosphate preparations may cause fluid and electrolyte abnormalities, especially in patients with renal impairment, and should be used with caution.

Standard antibiotic prophylaxis measures apply to surgery for colorectal malignancy. Agents should cover the common colonic aerobes (Escherichia coli) and anaerobes (Bacteroides fragalis).

Oral antibiotics include neomycin (which is not absorbed from the GIT) and metronidazole, which has an oral bioavailability almost equivalent to that of the intravenous preparation. Intravenous antibiotics include second or third generation cephalosporins combined with metronidazole. Dosing is at induction, with a second dose for prolonged surgery based on the half-life of the drug.

Patients undergoing surgery for colorectal malignancy often have multiple risk factors for DVT, including malignancy, abdominopelvic mass, dehydration with obstructing lesions, prolonged immobilization and prolonged surgery. Standard multimodal DVT prophylaxis measures should be implemented in all such patients.

The extent of resection is determined by the location of the lesion, its blood supply, and draining lymphatic system, and the presence or absence of direct spread to adjacent organs. Lymphatics should be resected as completely as possible. Restoration of GI continuity may be with sutures or staples, and anastomoses should be well perfused and tension-free.

It has been shown histologically that micrometastasis does not extend beyond 4cm in the bowel wall, therefore resection margins of 5cm are generally accepted to be safe.

The issue of ‘‘high’’ vs. ‘‘low’’ ligation of IMA has been a highly controversial topic since the turn of the 20th century. ‘High’’ ligation of the IMA (defined as ligation of this vessel at the point where the artery arises from the abdominal aorta, under cover of the 3rd part of the duodenum) is preferred by many surgeons, no matter where the tumor is sited in the left colon. At least theoretically, ‘‘high’’ ligation of the IMA could improve the lymphatic clearance, and seems to be a reasonable extension of the operation, albeit being associated with a higher risk of hypogastric nerve injury. The inferior mesenteric vein is ligated separately at the same level or higher. Subsequently, perfusion of the left part of the transverse colon (which is used to construct the anastomosis) is based on the middle colic artery; at the same time, the main blood supply to the rectum through the superior rectal artery is interrupted, and any part of the distal colon retained below the tumor have to be supplied entirely by the middle and inferior rectal vessels. Many surgeons feel that this latter supply is sufficient to supply the upper rectum and rectosigmoid up to 10 cm above its peritoneal reflection, but it would probably be unwise to rely on it to supply any greater extent of bowel. Therefore, when the IMA is ligated at its origin, the resection must ensure that the distal stump of the bowel is not too long. Most surgeons anastomose at the level of upper rectum in the pelvis, resulting in a dissection extending from the upper abdomen to the pelvic cavity, which may impose greater strain on the patient compared to ‘‘low’’ ligation of IMA. Most studies reveal that ‘‘high’’ ligation is not associated with significant improvement of survival compared to ‘‘low’’ ligation of IMA (ligation immediately distal to the left colic artery) when the most proximal lymph nodes were involved. However, improved survival advantage of 13.8% and 5.7% was found in 2 studies by Rosi and Grinell respectively. The decision is usually individualised, acknowledging that ‘‘high’’ ligation facilitates a tension-free anastomosis and is compulsory when a common trunk between left colic artery and first sigmoid artery exists (27% of cases). “High’’ ligation requires extended mobilization and resection without clear long-term advantages, and is contraindicated in older or poor-risk patients.

In patients with AAAs, the IMA is often occluded or severely stenosed due to severe atheromatous disease, and high ligation of the IMA carries little risk of colon ischemia. In contrast, if ligature of the IMA is too far distal, it may interrupt critical collateral flow from the left colic to sigmoid branches of the IMA. In an attempt to preserve colon arterial supply, revascularization of the IMA during AAA repair has been proposed when there is insufficient collateral flow. This is important when performing a colon resection in patients with previous AAA repair, where the left colon usually relies on the marginal artery and collaterals with the SMA and middle/inferior rectal arteries.

Surgical resection is the cornerstone of curative therapy for rectal cancer. The primary goal of treatment is complete eradication of the primary tumour along with the adjacent mesorectal tissue, the superior rectal arterial pedicle, and its associated lymph nodes. The extent of resection margins has steadily decreased over the past two decades from the traditional 5 cm margin down to 1cm or less. Although intramural micrometastases have occasionally been observed beyond 2cm, recent studies have not shown increased recurrence rates from 1cm margins with well-differentiated lesions. The optimal resection margins are likely to be based on a combination of factors and therefore should be individualized for each patient. Good mobilization of the rectum may provide extra length and facilitate sphincter and anastomotic preservation with low lesions.

The intraoperative distinction between benign adhesion and malignant invasion is difficult. 50% of adherence between a malignant rectal mass and an adjacent structure is due to malignant invasion. En-bloc resection of adherent masses provides a decrease in local recurrence rates of 69% compared with 18% for separation, and the 5-year survival rate for patients who had en-bloc resection of adherent lesions is 50%, compared to 20% for patients whose lesions were separated. Any adjacent organ or tissue that has been invaded by tumour should therefore be resected en-bloc with the tumour, without any attempt at separation.

Local excision of rectal tumours is an evolving therapy which still requires further study. Available information suggests that it may be equivalent to radical resection in selected patients. It is indicated for T1 and T2 lesions, and is sufficient for T1 lesions with negative margins, but must be supplemented with chemoradiotherapy for T2 lesions with negative margins. T2 lesions with positive margins following local resection should be managed with radical resection. T1 lesions with poor prognostic features should also be supplemented with chemoradiotherapy. Local excision also has a role as minimally invasive surgery in the palliation of advanced rectal tumours in high-risk patients who may not tolerate more extensive procedures.

Abdomino-perineal resection: This is the complete excision of the rectum together with the anus by concomitant dissection through the abdomen and the perineum, with closure of the perineal wound and creation of a permanent end-stoma on the abdominal wall. It is indicated for patients:

Low anterior proctosigmoidectomy with colorectal anastomosis (“low anterior resection”):

This is the resection of the rectum via the abdominal approach, almost always including the sigmoid colon because ligation of the IMA results in variable blood supply to it. Anastomotic results equivalent to those obtained by handsewing may be achieved by using a circular stapler for the colorectal anastomosis, however, high-risk anastomoses may be covered with a proximal stoma.

Proctosigmoidectomy has been revolutionized by the development of the total mesorectal excision described by Heald in 1998, which involves complete rectal excision, with complete removal of the mesorectum and its lymphatic structures. The use of TMRE has resulted in:

The principle of TMRE is the use of sharp dissection under direct vision in the avascular, areolar plane between the fascia propria of the rectum, and the parietal fascia overlying the pelvic wall structures, thus allowing autonomic nerve preservation and good haemostasis without violating the meorectal envelope. The dissection is carried down through Waldeyer’s fascia posteriorly and through or anterior to Denonvillier’s fascia anteriorly.

A sphincter-sparing coloanal anastomosis may be performed when sparing of the sphincters in a low lesion may be possible without violating oncological principles, for example for a low lesion made resectable by neoadjuvant therapy. Loss of the rectal reservoir with simple coloanal anastomosis often results in varying degrees of incontinence and frequent small-volume bowel movements (“clustering”), which may improve to a degree with time and dietary modification. Two techniques have been developed to address this problem:

Patients may need an ileal pouch or permanent ileostomy if they require proctocolectomy for inflammatory bowel disease or for familial colorectal cancer. An ileal pouch is constructed from the terminal ileum and anastomosed to the top of the anus, which allows evacuation per anus and is generally preferred to an end ileostomy. In contrast to colitis, patients do not have urgency and so are not inconvenienced by the frequency. Some patients may experience pouch dysfunction, most commonly in the form of pouchitis; however this usually settles with antibiotics.

The need for ileostomy or colostomy may be permanent or temporary. Colostomies are generally easier to manage than ileostomies, however with either stoma, quality of life may matche that of the general population. The expertise of a specialist stomal therapy nurse is key to the management of these patients.

These are constructed for patients following panproctocolecomy for inflammatory bowel disease, or in the context of colorectal cancer, the presence of cancer with extensive inflammatory bowel disease, or inherited cancer syndromes such as FAP. They are commonly chosen by, and work well in, younger patients with good anal sphincter function.

• The original S-pouch often requires insertion of a catheter per anus to empty it, and is generally no longer created.

• The quadruplicated W-pouch was popular; its greater volume was thought to reduce the evacuation frequency.

• The J-pouch is now the most widely used. It has the advantage that it can be formed using a stapler to reduce operation time. There is good evidence that a duplicated J-pouch with 20 cm limbs achieves results almost equivalent to the W-pouch.

Pouch–anus anastomosis: the upper half of the anal canal is lined by rectal-type mucosa that is often diseased; this can be removed by mucosectomy followed by hand-sewn anastomosis of the pouch to the mid-anal canal. Alternatively, the pouch is anastomosed to the top of the anal canal, preserving all the anal canal mucosa, using a circular end-to-end stapler. Stapling reduces trauma, with less minor anal leakage particularly at night, and is now the most popular technique.

Staged surgery may be used. The operation potentially has three stages. Removal of the colon and rectum, formation of the pouch, and, if used, closure of a temporary covering loop ileostomy 8–12 weeks after pouch formation.

Normal pouch function: most patients with a pouch empty it four to eight times in 24 hours (500–700 ml of stool of a porridge-like consistency). There should be no urgency, and patients should be able to defer evacuation for 20–30 minutes. About 50% of patients have to empty their pouch once during the hours of sleep and a few twice. Evacuation should be spontaneous and without pain. Most patients have perfect continence. Minor nocturnal leakage occurs in 10% of patients with a stapled pouch–anus anastomosis and in 20–40% of those with a handsewn anastomosis. A few patients suffer minor leakage by day. Pouch function varies between individuals, and fluid intake and the timing and composition of meals affect the pattern of pouch emptying. However, patients are generally not concerned about the precise number of motions because they no longer

Suffer urgency and can defer evacuation. More importantly, their quality of life (including energy level) is significantly improved – their disease has been ‘removed’ and their physical well-being restored. Patients with a pouch are encouraged to eat a normal diet.

Fertility: there is increasing evidence that female fertility may be adversely affected by pouch surgery, and hence, where possible, decisions may be deferred until after completion of family.

Pouch dysfunction: function outside the limits defined above is considered abnormal. Most patients with a pouch will experience an episode of poor function. Some problems settle spontaneously and most can be resolved with a thorough and systematic approach. Dysfunction can arise from three sites:

Pouchitis is the most common cause of significant dysfunction with an incidence of 20–40%. Of patients who develop pouchitis, 75% have intermittent attacks that are usually straightforward to manage. The remaining 25% suffer persistent symptoms or frequent attacks, and the symptoms usually develop in the first 6 months after pouch formation. The diagnosis of pouchitis is made on the triad of:

• clinical symptoms, including loose, frequent motions (often with blood), fever, malaise, anorexia and, occasionally, extraintestinal manifestations

If secondary causes of pouchitis are excluded, a 7–10-day course of metronidazole or ciprofloxacin is usually effective. Antibiotics probably work by altering the bacterial flora rather than reducing the total number of bacteria. Similarly, some cases of chronic pouchitis have been shown to respond well to treatment with probiotics although the longer term efficacy may be limited.

Beans, beer, caffeinated beverages, chocolate, leafy green vegetables, raw fruits and vegetables, spicy food, wholemeal food, cereal, alcohol, citrus fruits and juice

Apple sauce, bananas, boiled rice, cheese, smooth peanut butter, tapioca, white bread, potatoes, suet pudding, pasta

Beer, carbonated beverages, dried beans and peas, milk and milk products, onions, cabbage, broccoli, sprouts

Mushrooms, sweetcorn, potato skins, nuts, tomato skins, raw fruit skins, celery strings

In contrast to studies concerning breast carcinomas, in which patients with occult nodal tumor cells (OTC) or micrometastases had a significantly reduced survival time, at present, the prognostic relevance of OTC in lymph nodes of colorectal carcinomas is not really elucidated. The incidence of micrometastases in CRC seems to parallel the histological grade, with none in well-differentiated lesions, but up to 50% in poorly differentiated lesions. Pathological nodal understaging has been shown to exist, and the potential value of node mapping and SLN biopsy for colon cancer is to upstage node-negative patients to node positive to ensure accurate therapy.

Several recent studies have demonstrated the surgical feasibility of laparoscopical resections of colorectal cancer, reproducing endoscopically all those resections usually performed with an open abdomen, and reporting low mortality and morbidity rates. However, the low number of enrolled patients and the low period of follow-up do not allow sure conclusions about the real effectiveness of this procedure. It has been shown that the rates of recurrent cancer were similar after laparoscopically assisted colectomy and open surgery colectomy, suggesting that the laparoscopic approach is an acceptable alternative to open surgery for colon cancer. A review of the recent literature suggested that the overall incidence of port-site recurrence after laparoscopic colorectal cancer surgery was approximately 1%, in contrast to the previously suggested 4%. In large series the conversion rate ranges from 7% to 25%, while in smaller ones from 2% to 41%.

Without treatment, the median survival for colorectal liver metastases (CRLM) is 6-8 months, varying with the extent of disease . The prognosis is best with isolated, single-lobe metastasis or when limited in number. Even for the best prognostic groups, very few survive five years with no treatment. Surgery is the only treatment that offers cure for CRLM. Less than 20% of patients were considered suitable for attempted curative resection. Recently, a variety of strategies have been developed to optimise curative treatment, including improved preoperative staging techniques, new standards for surgical resection, novel surgical strategies, modern systemic neoadjuvant chemotherapy and an emerging role for ablative therapies.

Many studies consistently show five year survival rates following liver resection of 30-50%, Recurrence after five years is unusual and hence the vast majority of patients who survive five years can be considered cured of the disease Resectability is now being determined by what remains, rather than what is removed. Specifically, CRLM should be determined as resectable if: (i) The disease can be completely resected. (ii) Two adjacent liver segments can be spared with adequate vascular inflow and outflow and biliary drainage. (iii) The volume of the liver remaining after resection i.e. the ‘‘future remnant liver’’ (FRL) will be adequate. In those with an otherwise normal liver, the safe FLR volume is 20%. New surgical strategies have been employed to improve resectability. Portal vein embolization induces atrophy of the liver to be resected and hypertrophy of the liver that will remain (i.e. increases the future liver remnant). Similarly, two-stage hepatectomies involves delayed rehepatectomy after hypertrophy of the residual liver and may be used for large bilateral lesions in which a one stage resection of all the involved segments would lead to liver failure. Re-resection and even serial resections of CRLM is feasible and 5 year survival rates mirrors that observed for patients following a single resection. Reported morbidity and mortality rates are also similar to those reported after initial hepatectomy. Extrahepatic colorectal metastases may be resected with curative intent, such as direct diaphragmatic invasion, adrenal metastases and lung metastases that are few in number and readily resectable. Over the last decade, there has been considerable development of ablative techniques including cryo, radiofrequency, microwave or laser ablation and high-intensity focused ultrasound (HIFU), however, their roles in management of liver metastases are still being evaluated.

Pulmonary metastases from CRC often represent systemic and uncontrolled tumour growth, however in a number of patients lung disease is limited and the patient remains well. 5- and 10-year survival rate of 44% and 25% are possible when the metastases are resectable. Chemotherapy is the only alternative treatment but very rarely leads to survival beyond 24 months. Pulmonary metastases which are suitable for resection are usually detected on chest radiography- they are rarely a cause of symptoms and the majority of patients have otherwise healthy lungs. CT scans supplemented by PET scans usually confirm the diagnosis, but percutaneous biopsy is sometimes necessary to exclude a primary lung cancer. The criteria for resecting CRC pulmonary metastases are (1) the primary tumour is controlled or is controllable; (2) complete resection is possible; and (3) the patient has adequate pulmonary reserve to tolerate the planned resection. Surgical approaches include posterolateral thoracotomy, staged bilateral thoracotomies, median sternotomy, clamshell incision, and video-assisted thoracic surgery. The majority of patients post resection well with few complications. Following resection favourable prognostic factors include a long disease-free interval, small number and small size of metastases, a normal carcinoembryonic antigen level, and an absence of concomitant liver metastases and mediastinal lymph node spread. Surgery for pulmonary metastases of CRC remains the best means of local control and the best way to render the patient disease-free. Patients with complete resection of pulmonary metastases have an improved long-term survival when compared to patients with unresected metastases.

Despite the fact that elderly people represent a major proportion of colorectal cancer patients, they are frequently undertreated both for early and advanced-stage disease. Patient preference is often the main reason for not receiving adequate treatment, as many elderly patients regard themselves as being unfit for rigorous or demanding therapies. It has been demonstrated by numerous reports that there is significant benefit from chemotherapy and surgery in the adjuvant and palliative setting, irrespective of patient age. Oncologic studies consistently support the notion that elderly patients have equal potential gains with younger patients from adjuvant and palliative chemotherapy. The toxicity profile of most regimens is mostly unchanged at older-age, with certain exceptions for specific drug combinations.

Evidence concerning the use of surgery is more difficult to interpret, as it is more common for elderly people to have comorbid conditions not allowing a surgical procedure to be performed, or not allowing rapid and uneventful recovery after surgery. Elderly people also frequently postpone medical consultation, leading to a greater proportion of advanced-stage disease and therefore, more extensive or frequently emergency operation. In the systematic review of the Colorectal Cancer Collaborative Group, 11% of patients younger than 65 years underwent emergency surgery, as opposed to 15% in the 65–74 age group, 18% in the 75–84 age group, and an alarming 29% in the older than 85 years group. The data reviewed suggest that surgery may be of benefit to selected patients, after thorough medical evaluation.

Palliative care is defined by the World Health Organization (WHO) as ‘‘the active total care of patients whose disease is not responsive to curative treatment. Control of pain, other symptoms, and psychological, social, and spiritual problems is paramount. The goal of palliative care is achievement of the best quality of life for patients and their families’’. CRC is the second leading cause of cancer-related mortality in the United States, with 147,000 new cases and 57,000 deaths occurring each year. Approximately 20% of CRC patients will present with irresectable locally advanced or metastatic disease, and 50% of patients who undergo apparently curative resection for localized disease eventually die of metastatic disease. The median survival for patients with stage IV CRC is between 6 and 9 months from the time of diagnosis, and in this time they may present with a variety problems, from psychological issues to physical symptoms such as obstruction, bleeding, perforation, pain, weight loss, and fatigue, all affecting and diminishing quality of life. There is a broad range of multidisciplinary treatment options available for palliative management. Reports of recent advances in systemic chemotherapy for CRC have demonstrated an almost 2 year median survival for patients with Stage IV. Noncurative surgery may delay systemic chemotherapy that could otherwise result in regression of both the primary tumour and metastatic foci. The main factors that influence treatment plan decisions are patient age and comorbidities, aggressiveness of tumor biology including response to chemotherapy, local extent of disease, and severity of patient symptoms.

Once the diagnosis of disseminated stage IV disease has been made, additional studies will likely

add little to patient care, add to treatment costs and may worsen discomfort. 20% of patients with CRC will present with stage IV CRC, with a reported 8% 5-year survival, and 75–90% of these patients will present with unresectable disease. While there is consensus regarding noncurative resection for palliation of symptoms such as bleeding, perforation, and obstruction, the approach to asymptomatic patients or patients who present with symptoms that may not respond to surgical resection is less well defined. Surgical resection in asymptomatic stage IV CRC may lead to unnecessary delay in the initiation of systemic treatment. Moreover, such patients are at higher risk for a perioperative complications including anastomotic leaks, surgical site infections, venous thromboembolic phenomena, and pulmonary morbidities, which may significantly delay or preclude the patient from receiving systemic chemotherapy. In the majority of patients with incurable stage IV CRC, there is no advantage to resection of the primary tumor in asymptomatic patients and chemotherapy is preferable to surgery. Advocates of a more proactive initial surgical treatment argue for the benefit of prophylactic surgery to avoid future complications such as bleeding, perforation, and obstruction with a mortality rate of 27% that is associated with emergency operation in patients with stage IV CRC. Multiple studies have demonstrated benefit from this course of management in patients who are fit for surgery. Patients with advanced rectal tumors may present with advanced symptoms from cancers which may obstruct, perforate, bleed, and/or fistulize to adjacent organs or skin. Bowel habits are often compromised with tenesmus, fecal incontinence, constipation or intractable diarrhoea, and local invasion of nerves may cause intractable pain. Extension to the ureters or bladder may result in obstructive uropathy. Since any of these symptoms may be so profound, palliative abdominoperineal resection (APR), low anterior resection (LAR), or Hartmann’s procedure (particularly in cases with poor sphincter function but

Pelvic exenteration is a particularly radical form of surgical palliation. Extended pelvic resection in the form of partial or complete pelvic exenteration in patients with locally advanced rectal disease that invades surrounding organs has been reported with various survival rates and incidences of morbidity and mortality. Notoriously high rates of morbidity (40–75%) and expected readmissions have led many to question its use for both palliative and curative indications. However, pelvic exenteration for patients with otherwise curable locally advanced rectal cancer may result in 25–62% 5-year survival. Furthermore, successful urinary and/or intestinal reconstruction is possible in 33% of patients after en-block rectal and bladder resection for such T4 rectal cancer.

One-third of the patients with CRC presents with near or complete bowel obstruction. Patients with

bowel obstruction tend to have more advanced disease and bowel obstruction is a poor prognostic sign. At laparotomy 30% of patients who present with malignant bowel obstruction will be deemed irresectable due to extensive local or systemic disease, and/or associated comorbidities. When patients present with acute malignant large bowel obstruction, an emergent diverting colostomy may be required, with a possible curative or palliative resection performed at a later stage. Mortality and morbidity rates in the setting of emergency surgical decompression approach 20% and 50%, respectively, and up to 50% of stomas are ultimately not reversed.

The advantages of laparoscopic surgery over open surgery include less incisional pain, faster recovery, and shorter hospital stay. It is also thought to be less immunologically stressful due to attenuation of the perioperative operative stress response. Laparoscopy can be technically demanding, especially in emergent cases were obstructed large and small bowel as well as frail tissues may be involved. Relative contraindications for laparoscopic colectomy may include previous abdominal surgery, extreme central obesity, very large fixed tumors, and tensely distended obstructed bowel. A 22% conversion rate in patients who underwent laparoscopic colectomy for palliation vs. 11% in patients who underwent laparoscopic resection for cure has been reported. Laparoscopic palliation for CRC can be safe and feasible with satisfactory outcomes and should be part of the armamentarium of surgeons who treat patients with advanced CRC.

Patients with stage IV incurable CRC who are not candidates for surgery may benefit from endoscopic palliation including laser therapy and self-expanding metallic stents. Endoscopic laser therapy using the neodymium-doped yttrium aluminium garnet (Nd:YAG) laser has been used to palliate symptomatic colorectal lesions since the early 1980s. Laser treatment can successfully provide symptom relief for patients with bleeding, mucus discharge, or tenesmus and can effectively maintain luminal patency in obstructing tumors. The ability to endoscopically palliate colorectal malignancies has gained tremendous progress with the utilization of self-expending metallic stents (SEMS), which have become widely utilized in the setting of colorectal malignancies for several indications: SEMS are indicated

Patients with CRC who deem inoperable due to advanced disease or patients with potentially curable disease in whom surgery is precluded because of concomitant high-risk medical problems, may benefit from palliative endoscopic laser therapy, which may also be used to decompress obstructed leftsided colonic tumors as a bridge to surgical resection. Patients without complete bowel obstruction receive bowel preparation. The procedure is mostly used for left-sided colon and rectal lesions and is contraindicated for obstruction secondary to extraluminal tumors. Multiple

treatment sessions are often required to achieve significant and ongoing symptom relief. Laser therapy was shown to be effective for palliation of bleeding and obstructing symptoms.

The value of routine history and physical examination is uncertain and only 20% of recurrences are identified on routine history and physical examination. More than 45% of the recurrences discovered within the first 3 years after resection are exposed by symptoms that occur between physician visits and tests and 24% of patients with symptoms suggestive of recurrent disease delay reporting these signs to their physician until the next scheduled appointment. Digital rectal examination has a sensitivity of 74% to 86% and a specificity of 98% to 100% in detecting local recurrence. More than 90% of colorectal surgeons follow-up patients every 3 to 6 months for the first 3 years, and more than 80% continue to follow-up their patients annually after that.

CEA has value as a tumor marker in patients with known cancer and is used most widely to monitor patients for recurrent disease after curative resection of colorectal tumors. After curative resection, an increased CEA level should decrease to normal levels in 4 to 8 weeks, failing which incomplete resection of the primary tumor should be suspected. No study has shown prolonged survival in patients monitored with CEA, and the serum level that should trigger further work-up to detect recurrence or lead to surgery is unknown. A single CEA value of greater than 5 ng/mL will trigger an evaluation by most clinicians, but it remains controversial whether detection of rising CEA levels leading to early detection (lead time, 3–8 months) of tumor recurrence improves long-term survival. 30% of all colorectal cancer recurrences do not produce CEA, and studies have shown only a 0-3% improvement in long-term survival with early recognition of recurrent disease.

In patients for whom resection of liver metastasis is indicated, it currently is recommended that CEA levels be monitored every 2 to 3 months for 2 years or more after resection.

The current American Society of Clinical Oncology recommendations are as follows:

There are no studies specifically addressing the usefulness of any portion of the FBC or LFT, changes in which lag behind other serologic and radiographic features of liver metastasis. 50% of colorectal metastases is to the liver; the sensitivity of ALP for liver metastasis in patients with colorectal cancer is 77%, but the false-positive rate is 34%. The false-negative rate in most studies is greater than 4%. Because of low specificity, routine use of LFTs in surveillance of colorectal cancer is not recommended.

5-10% of patients who undergo resection of colorectal cancer will develop pulmonary metastasis. Results of resection of pulmonary metastasis from colorectal cancer are encouraging, therefore it is imperative to detect these lesions as early as possible. A chest radiograph has been the primary screening modality for pulmonary metastasis, however it has a low frequency of first detecting pulmonary metastasis (3–20%). Despite the low cost of chest radiographs, the advantage in outcome is small. There are no randomized studies showing a beneficial effect of yearly, 6-month, or 3-month chest radiographs. The current available data do not support routine chest radiographs in the follow-up evaluation of colorectal cancer, however, increasing CEA levels or specific pulmonary symptoms in patients should warrant a chest radiograph. Furthermore, patients with rectal carcinoma, who have a higher propensity for lung metastasis compared with colon cancer, may warrant a lower threshold for obtaining a chest radiograph.

Ultrasound is operator dependent and can be influenced by certain patient factors such as bowel gas overlying the liver and obesity. It has a low sensitivity of 57% in detecting liver metastasis, which decreases to 20% for liver lesions less than 1 cm in diameter. The reported incidence of local recurrence in rectal carcinoma after surgical resection is 3% to 32%, therefore transrectal ultrasound may be effective in diagnosing recurrent rectal cancer. However, ultrasound is unable to distinguish between surgical/radiation scarring, benign and malignant masses. In men, transrectal ultrasound is limited to patients with sphincter-preserving surgery. However, in women, transvaginal ultrasound can be used even if proctectomy has been performed. The accuracy of transrectal ultrasound in detecting local rectal cancer recurrence is 80% to 85%, with reported sensitivities and specificities of 90-100%.

The use of ultrasound-guided fine-needle aspiration (EUS FNA) of lymph nodes is not helpful in preoperative staging, however it may have a role in diagnosing early recurrence, with sensitivity of 80%, a specificity of 100%, and an accuracy of 83%.

Currently, a computed tomography (CT) scan is the preferred method for defining local recurrence of colorectal cancer. It has a sensitivity of 95%. CT scans can evaluate the anastomosis, tumor bed, regional lymph nodes, and the liver. The sensitivity of CT scans in detecting liver metastasis from colorectal carcinoma is 80% to 90%. The false-negative rate for detecting liver metastasis by CT scan is 40%. Routine surveillance CT scan provides little additional benefit when combined with CEA monitoring. Based on current data, the use of periodic CT scan in the postoperative patient with colorectal carcinoma is unproven. However, the CT scan is an extremely useful test to

evaluate further any suspicious signs and symptoms in a patient (ie, increasing CEA level, abdominal pain, abnormal LFTs). Further studies are needed to evaluate the role of routine CT scanning in patients with non–CEA-producing tumors.

Positron emission tomography (PET) is a functional imaging modality that is based on physiologic and metabolic changes within tissue to allow for the detection of disease, in contrast to conventional imaging techniques such as CT scan which show morphologic changes. (18F)-fluoro-2-deoxy-D-glucose (FDG) is a radioactive glucose analog, and its accumulation is proportional to the rate of glucose transport and metabolism within a given cell. Its use in oncology is based in part on the fact that tumor cells have increased glucose use. Hence, PET has emerged as an important tool in the work-up of patients with cancer and it is useful in assessing patients with suspicion for recurrence based on increased CEA levels, or physical or radiologic findings that are indeterminate.

Multiple studies have shown PET CT to have a positive predictive of 90-95% and the negative predictive value of 100%, with a sensitivity of 80-97% and specificity of 76%. It may also predict irresectable disease in 96% of patients. Disadvantages include the normal physiologic accumulation of FDG in the stomach, bowel, kidneys, ureter, and bladder contributing to low specificity, variability in interpretation of images, and cost.

Immunoscintigraphy uses radiolabeled monoclonal antibodies to specific tumor antigens. In the case of colorectal carcinoma, the most commonly used monoclonal antibodies are anti-CEA antibodies, B72.3, and anti 17-1A antibodies. Sensitivities (25%) and specificities (20-90%) are very low with a false positive rate of 80%. It therefore has no role in diagnosis, but may later find a role in differentiating suspicious lesions as being benign or malignant, or localising recurrent disease suggested by high CEA levels but negative imaging with other means.

Surveillance of the colon after resection of colorectal cancer is essential to detect metachronous lesions or intralumenal anastomotic recurrent cancer at an early stage. All patients should undergo a full endoscopic evaluation of the colon before undergoing resection for colorectal cancer to document a cancer and polyp-free colon before resection. A second or synchronous malignancy at a different location within the colon or rectum is found in 3% to 7% of patients about to undergo colorectal cancer resection, and another 25% of these patients will have adenomas that require removal, therefore if a preoperative colonoscopy of the entire colon is not performed, a postoperative colonoscopy after resection is mandatory. Colonoscopy is rarely the first indicator of recurrence (0–19% of patients), however, 80% of anastomotic recurrence is found at endoscopy. Although colonoscopy is not the most common test to first indicate recurrence, it is useful to detect new polyps or new cancers especially in patients with a history of rectal cancer. It is unclear as to how often a postoperative surveillance colonoscopy should be performed, however most colorectal surgeons perform a colonoscopy every 6 to 12 months for the first 5 years after primary resection of colorectal cancer

Despite advances in technology to detect recurrent disease, such as CEA assays, immunoscintigraphy, PET scans, and CT scan, no postoperative surveillance study unequivocally has proven a survival benefit to these patients. The survival advantage from surveillance is expected to be small, and there is a need for large, well designed clinical trials to better define the optimal postoperative surveillance program for patients with colorectal cancer. Follow-up guidelines are controversial. A typical approach used at Memorial Sloan Kettering, and endorsed by ASCO, is to examine the patient every 3 to 4 months for the first several years postoperatively, with CEA monitoring at each visit, and annual CT scans of the chest, abdomen, and pelvis for the first 3 years. Full colonoscopy is performed at 1, then 3, and then every 5 years after resection, unless colonoscopic findings or other factors necessitate more frequent colonoscopy.

Colon cancer incidence and survival has improved only moderately during the past 2 decades despite the manifest efficacy of colonoscopic polypectomy at cancer prevention. New simpler, less invasive, and safer tests are being designed to overcome these barriers to universal screening for colon cancer. Potentially exciting screening or diagnostic tests still in the experimental stage include stool genetic markers and videocapsule endoscopy of the colon. The role of CT colonography (virtual colonoscopy) in the screening of colon cancer currently is unclear and may be clarified by further studies as the technology matures.

1 Sakorafas GH, Zouros E, Peros G: Applied vascular anatomy of the colon and rectum: Clinical implications for the surgical oncologist. Surgical Oncology (2006) 15, 243–255

2 Drake RL: Anatomy of the Colon, in Baker RJ, Fischer JE (eds) Mastery of Surgery, 4e Lippencott Williams & Wilkins (2001):1461-1465

3 Cappell MS: Pathophysiology, Clinical Presentation, and Management of Colon Cancer Gastroenterol Clin N Am 37 (2008) 1–24

4 Cappell MS: Reducing the Incidence and Mortality of Colon Cancer: Mass Screening and Colonoscopic Polypectomy. Gastroenterol Clin N Am 37 (2008) 129–160

5 Lawes :Chemotherapy for colorectal cancer—an overview of current management for surgeons. EJSO (2005) 31, 932–941

6 Asmis TR: Systemic Therapy for Colon Cancer. Gastroenterol Clin N Am 37 (2008) 287–295

8 Dionigi G: The application of sentinel lymph node mapping in colon cancer. Surgical Oncology (2007) 16, S129–S132

9 Rovera F: Colorectal cancer: The role of laparoscopy. Surgical Oncology (2007) 16, S65–S67

10 Nesbitt C: Factors that influence treatment strategies in advanced colorectal cancer. EJSO 33 (2007) S88eS94

11 Warwick R: Resection of pulmonary metastases from colorectal carcinoma. EJSO 33 (2007) S59eS63

12 Leen E: Radiofrequency ablation of colorectal liver metastases. Surgical Oncology (2007) 16, 47–51

13 Golfinopoulos V: Treatment of colorectal cancer in the elderly: A review of the literature. CANCER TREATMENT REVIEWS (2006) 32, 1–8

14 Wasserberg N: Palliation of colorectal cancer. Surgical Oncology (2007) 16, 299–310

15 Keighley MR, Lee JR, Ambrose NS, Indications for bowel preparation in colorectal cancer, Int Adv Surg Oncol, 2003;6:257-70

16 Heald RJ, Husband EM, Ryall RD. The mesorectum in rectal cancer surgery: the clue to pelvic recurrence, Br J Surg 69:613, 1982

17 Vogelstein B, Fearon ER, Hamilton SR, et al. Genetic alterations during colorectal-tumor development. N Engl J Med 1988;319:525–32.

18 Rex DK, Lieberman D. ACG colorectal cancer prevention action plan: update on CT-colonography. Am J Gastroenterol 2006;101:1410–3

19 Mahmoud N: Colon and rectum. In: Townsend (ed) Sabiston Textbook of Surgery, 17^th ed, Elsevier-Saunders (2004) 1401-1481

20 Kaiser AM: Tumors of the colon. In Maingot’s Abdominal Operations, 11^th ed: Zinner MJ, Ashley SW (eds) McGraw-Hill (2007) 625-659

21 Dadal KM: Cancer of the rectum. In Maingot’s Abdominal Operations, 11^th ed: Zinner MJ, Ashley SW (eds) McGraw-Hill (2007) 693-726

22 Bullard KM: Colon, rectum and anus. InScwardtz’s Textbook of surgery. 8^th ec: Brunicardi FC (ed) McGraw-Hill 2005 1055-1117

23 Ferndale L: Colorectal Cancer: Surgical seminar (2004). UKZN academic press. 2004

24 Singh A: Colrectal Cancer: Surgical seminar (2006), UKZN academic press. 2006

25 Abir F: The postoperative surveillance of patients with colon cancer and rectal cancer. Am J S 192 (2006) 100–108

RISK FACTOR	PROPOSED MECHANISM
Epidemiology
Advancing age	Acquired colonocyte mutations accumulate with age. Sharp increase in cancer incidence from 5^th decade onwards
Living in industrialized nations	Dietary and environmental carcinogens
Diet
Low fruit and vegetable consumption	Anticarcinogenic substances in fruits and vegetables (eg, folic acid)
Obesity	Carcinogens in an unhealthy diet or role of abnormal insulin levels in carcinogenesis
Social habits
Smoking cigarettes	Carcinogens present in tobacco
Alcohol	May promote cell proliferation and inhibit DNA repair
Genetics/family history
FAP	Develops hundreds of adenomatous colonic polyps. Inevitably develops colon cancer resulting from small but significant risk for malignant transformation in each adenoma. (100% risk of cancer)
Gardner’s syndrome	Variant of FAP
HNPCC (Lynch syndrome)	Mutant mismatch repair gene leads to accumulation of genetic mutations, including mutations of tumor suppressor genes (80%)
Peutz-Jeghers syndrome	Syndromic hamartomatous polyps occasionally may transform to adenomas
Juvenile polyposis	Syndromic juvenile polyps can transform to adenomas and then cancers over time
Family history of nonsyndromic colon cancer	Postulated shared genetic factors leading to mild susceptibility to colon cancer and possibly shared environmental factors
Hyperplastic polyposis	Genetic mutation in hyperplastic polyposis seems to predispose to colon cancer
Inflammatory bowel disease
Chronic Crohn’s colitis	Dysplasia and genetic mutations associated with mucosal injury and repair
Chronic ulcerative colitis	Dysplasia and genetic mutations associated with cell injury and repair. 10-20% risk after 20 years
History of prior neoplasia
Colonic adenomatous polyps	Precursor lesions of colon cancer
Prior colon cancer	Genetic predisposition or environmental factors
Other
Pelvic radiation	Carcinogenic effects resulting from radiation induced mutations
Streptococcus bovis bacteremia	May promote colonocyte proliferation
Ureterosigmoidostomy	Carcinogens excreted in urine or colonic mucosal proliferation during repair after urine-induced mucosal injury (100-500 x increased risk)
Acromegaly	Growth hormone promotes proliferation of preexisting colonic adenomas and cancers

Questionable or controversial risk factors for colon cancer:
Physical inactivity	Physical activity may stimulate immunosurveillance and stimulate intestinal peristalsis to decrease mucosal contact with fecal carcinogens
Low calcium	Calcium binds to bile acids that otherwise are potentially colonotoxic
High fat	Various theories (eg, increased bile secretion)
High red meat	Animal fat in red meat or carcinogens (eg, nitrosamines) in cooked meat
Low selenium	Selenium can help neutralize toxic free radicals due to antioxidant effects
Low folate	Folate needed for DNA synthesis and repair
Low carotenoid diet	Carotenoids can help neutralize free radicals (antioxidant effects)
Low fiber diet	Dilution of carcinogens in stool due to increased stool bulk and stool water with a high fiber diet
Breast cancer	Shared reproductive hormonal or environmental factors
Diabetes mellitus	Insulin may modulate colonocyte proliferation
Prior cholecystectomy	Continuous colonic exposure to potentially carcinogenic bile acids after cholecystectomy

Gene	Chromosome location	Normal function of encoded protein	Clinical manifestations of mutation
APC gene	5q	Regulates cell growth and apoptosis	Homozygous somatic mutation associated with colonic adenomas
K-ras gene family	Various chromosomes	Encodes a small guanosine triphosphate binding protein on cell membrane involved in transduction of mitogenic signals across cell membrane	Mutated in approximately 50% of colon cancers. May act in an intermediate stage of carcinogenesis. Mutation common in hyperplastic polyps.
p53 gene	17p	Regulates G1 cell cycle and apoptosis	Critical in transition from late adenoma to early cancer.
DCC gene	18q	Encodes a neural cell adhesion molecule; facilitates apoptosis, tumor suppressor	Believed to promote progression to frank carcinoma.
Mismatch repair genes	Located on several chromosomes	Recognize errors in nucleotide matching on complementary chromosome strand and initiate excision of erroneous strand	Progressive accumulation of mutations throughout the genome in affected cells leading to hypermutability and genetic chaos. Mutations of oncogenes or tumor suppressor genes can lead to colon cancer.

Dukes’ Classification (Astler-Coller modification)
Carcinoma in situ / high grade dysplasia	Intramucosal carcinoma that does not penetrate the muscularis mucosae
Stage A	tumors invade through the muscularis mucosae into the submucosa but do not reach the muscularis propria.
Stage B1	tumors invade into the muscularis propria.
Stage B2	tumors completely penetrate the smooth muscle layer into the serosa.
Stage C	tumors encompass any degree of invasion but are defined by regional lymph node involvement.
Stage C1	tumors invade the muscularis propria with less than 4 positive nodes.
Stage C2	tumors completely penetrate the smooth muscle layer into the serosa with 4 or more involved nodes.
Stage D	lesions with distant metastases

Stage I	93%
Stage IIA	85%
Stage IIB	72%
Stage IIIA	83%*
Stage IIIB	64%
Stage IIIC	44%
Stage IV	8%

Durban Surgeons

STAGING

CLINICAL FEATURES

SCREENING

MANAGEMENT

PROGNOSIS

REFERENCES