Introduction

Colorectal cancer (CRC) is one of the most common human malignancies, and despite progress in screening and treatment strategies, it remains a significant public health problem in the developing and underdeveloped world. Although significant progress has occurred in developed countries, thanks to screening, early removal of neoplastic polyps, and advances in clinical treatments, there is still a significant proportion of patients who experience tumor relapse and ultimately die from their disease. In parallel with this, there have been significant developments in our understanding of molecular pathobiology of CRC and its application to patient care over the last decade. This chapter reviews this progress in our understanding of the biology of CRC, its diagnosis and treatment and summarizes the advances that have been achieved in the clinical management of this disease.

CRC is the second most common cancer in women and the third most common cancer in men worldwide, resulting in more than 1 million cases each year. It is expected that approximately 143,460 new cases of CRC will be diagnosed within the United States and 450,000 in Europe in 2012. In addition, with approximately 51,690 estimated deaths in 2012 in the United States and 232,000 deaths in Europe, CRC remains the second most common cause of cancer death within the Western world.1,2 Countries and regions with a western lifestyle, for example, North America, Europe, Australia, and New Zealand, have traditionally had the highest rates of CRC,³ whereas lowest rates are found in Africa and Asia.³ Within economically developed nations, the lifetime risk of developing CRC is 1/20.² In economically transitioning countries such as in Eastern Europe, CRC incidence is increasing substantially (Fig. 77-1).⁴ This increase appears to reflect a change in physical activity and diet. By contrast, the United States is the only country where the incidence is declining for both males and females (Fig. 77-2). The overall incidence of CRC declined by 22% between 1975 and 2000.⁵ This decline has been attributed largely to increasing screening rates, which expanded from 27% to 46% of the eligible population between 1987 and 2003.⁶ Celebrity profile and media attention also helped increase the uptake of screening among the general population.⁷ In 2012, the American Cancer Society released new guidelines in relation to early detection and screening for CRC.⁸

The incidence of CRC is relatively rare in the under 40 years of age group and increases rapidly in older cohorts. The incidence is 15 times greater in adults older than age 50 years compared with those younger than age 50 years.⁹ CRC is 25% more likely to develop in men than in women, and the rate is 20% higher in African Americans compared with whites.¹⁰ The burden of CRC is likely to increase with increasing life expectancy in both males and females.

Inherited Colon Cancer Syndromes

As many as 20% of patients with CRC are likely to have a close relative with the disease and, as such, can be considered as familial in nature.13,14 Indeed, it is likely that 5% to 10% of all CRCs are syndromic. Molecular genetic testing leading to subsequent clinical management recommendations currently exists for patients with a variety of CRC syndromes, including hereditary nonpolyposis colorectal cancer (HNPCC), familial adenomatous polyposis (FAP), MYH-associated polyposis, and the hamartomatous polyposis syndromes (primarily Peutz-Jeghers syndrome, juvenile polyposis, and Cowden disease).^17–17 In addition to CRC, these syndromes are accompanied by other extracolonic manifestations, some of which represent extracolonic neoplasias. The risk of CRC in an individual with a first-degree relative with CRC is increased threefold compared with those with no family history. High-risk individuals should undergo genetic counseling, with the potential for genetic testing and a program of screening by colonoscopy.

Familial Adenomatous Polyposis

FAP is a rare autosomal dominant disorder that accounts for 1% to 2% of all CRC and is characterized by the presence of hundreds to thousands of adenomas in the colorectal mucosa by 20 to 30 years of age (Fig. 77-3).^21,22 An attenuated form of FAP (AFAP) is recognized, with fewer polyps (on the order of 100 or less), and both the presentation and the development of carcinoma are delayed.²³

In general, FAP is the result of an inherited defect in one allele of the APC tumor suppressor gene; however, up to 20% of cases could represent de novo mutations without an apparent family history. When the second allele is inactivated through mutation early in childhood, the colonic epithelium begins a process of uncontrolled hyperplasia and polyposis. The lifetime risk of CRC is nearly 100%, with 90% of cases occurring by 45 years of age²⁰; lifetime risk appears to be correlated with the number of adenomas.²⁴ In addition to the risk of CRC, other phenotypic manifestations of the disorder can include tumors of the central nervous system (also known as Turcot syndrome),²⁵ epidermal cysts, osteomas, dental abnormalities, mesenteric or abdominal wall desmoid tumors (also known as Gardner syndrome),²⁶ congenital hypertrophy of the retinal pigmented epithelium, and upper gastrointestinal tumors, especially periampullary duodenal carcinoma. Several of these phenotypic features have been correlated with specific mutation sites.²⁷

The most common genetic abnormality results in a stop codon, leading to the production of a truncated and nonfunctional APC protein. The in vitro production of this abnormal protein now serves as a genetic screening procedure for the disease, following appropriate genetic counseling. In the absence of preventive measures, invasive CRC occurs in FAP patients at an average age of 42 years. This provides a compelling rationale for a variety of preventive interventions, including early endoscopic screening, proctocolectomy, colectomy with ileorectal anastomosis, lifelong upper and lower intestinal surveillance, and participation in clinical trials evaluating the potential benefits and risks of chemopreventive options.

MYH-Associated Polyposis

This is a polyposis syndrome characterized by mutation in the MYH gene, which is involved in the DNA proofreading machinery of the genome; its failure allows for the accumulation of further mutations, particularly in the APC gene.30–30 The polyposis of this syndrome tends to be less pronounced than FAP.

Hamartomatous Polyposis Syndromes

Colorectal hamartomas are frequent in Peutz-Jeghers syndrome (PJS), juvenile polyposis syndrome (JPS), and Cowden disease.³¹ These syndromes are very rare, with incidences below 1 per 100,000 of the population.

Other Genetic Factors

In addition to high-penetrance genes such as those described previously, a variety of low-penetrance polymorphic genes have been investigated for their potential to influence CRC risk, particularly when considered in the context of concomitant dietary or environmental exposures.³² Specifically, polymorphisms associated with an increased risk of CRC included the NAT2 fast phenotype (70% increased risk); the GSTT1 null genotype (40% increased risk); a number of mutant alleles of ALDH2; any of four rare variants of the protooncogene HRAS1; and the a2, a5, and a13 alleles of TNFα. More recent studies using whole-genome scanning approaches have also identified single-nucleotide polymorphisms locus associated with an increased risk of CRC at the 8q24 locus.^33,34

Familial clustering of CRC is a common occurrence, even in those without a definable genetic predisposition. For example, twin studies suggest that heritable factors are important in CRC development, even in the absence of known genetic predispositions.³⁵ Two recent meta-analyses of observational studies demonstrated an approximate relative risk of 2.25 in individuals with a single first-degree relative affected by CRC and a 4 to 4.25 relative risk in those with more than one affected relative or with a family member who developed CRC before 45 years of age.^36,37 Population lifetime CRC risk estimates for a 50-year-old increased from 1.8% to 3.4% with one affected relative and to 6.9% with two or more affected relatives. Risks were also elevated approximately twofold for individuals with first-degree relatives who harbored colorectal adenomas, further supporting the thesis that colorectal adenomas and carcinomas are linked.

Family History of Colorectal Cancer or Adenomatous Polyps

Screening is recommended for individuals with a first-degree relative who developed CRC before the age of 50 years. Screening should commence at the age of 40 years or 10 years earlier than the earliest cancer.8,38,39 There is newer evidence that having a first-degree relative develop an advanced polyp (>1 cm or villous histology) before the age of 50 years also increases the risk of developing CRC.⁴⁰ This has led some organizations (including the American College of Gastroenterology) to include it as grounds for screening of targeted individuals.^8,41

Prior Polyps; Inflammatory Bowel Disease

The risk of CRC in individuals with a previously diagnosed adenomatous polyp is particularly increased if the polyps are large (>1 cm) or multiple (>2) or advanced (high-grade dysplasia or villous histology) and requires additional follow-up.42,43 With pancolonic inflammatory bowel disease (ulcerative colitis or Crohn disease), the risk of CRC increases with the duration of illness (2% at 10 years, 18% at 30 years).^44–47 The extent of the involved colon is an important determinant of risk and no increased risk is seen when it is confined to the rectum.⁴⁵ Screening is recommended to begin 10 years after the diagnosis is made and the frequency to be increased in subsequent decades or according to the degree of ongoing inflammation.

Diabetes Mellitus; Obesity

There is an association between risk of developing diabetes and CRC. This is partly explained by the shared risk factors of physical inactivity and obesity (see below), but there would appear to be an additional independent risk factor.⁴⁸

Obesity, particularly abdominal obesity, is a risk factor for CRC, independent of physical activity.⁴⁹ The association is stronger in men than in women. Men in the highest quintile (body mass index [BMI] >30) have double the risk compared with those in the lowest quintile (BMI <23).^50,51 The relationship is linear (Fig. 77-4).⁵² The mechanism appears to be related to insulin resistance and hyperinsulinism.

Alcohol

Most studies demonstrate a link between high alcohol intake and risk of CRC or adenoma. Consumption of an average 2 to 4 alcoholic drinks per day resulted in a 23% increased risk of CRC, compared with those who drank less than 1 alcoholic drink per day.⁵³

Smoking

An association with CRC has been demonstrated, with a relative risk of 1.26 (1.11 to 1.43) compared with nonsmokers, with the risk persisting up to 25 years after stopping smoking.⁵⁴ It is estimated that 15% to 20% of CRC can be attributed to smoking,⁵⁵ and the link is strongest for rectal cancer.⁵⁶

Genetic studies have revealed reductions in CRC risk for polymorphisms in a number of genes, including homozygous carriers of the C677T variant allele of methylenetetrahydrofolate reductase (20% to 30% reduced risk) and heterozygous or homozygous carriers of an intron 3 polymorphism of the p53 gene.³²

Colon cancer is usually quite insidious in its development. Asymptomatic patients are suspected of having colon cancer because of iron-deficiency anemia or a positive FOB test that is found during routine physical examination. With increasing awareness of the benefits of screening colonoscopy, evidence of cancer could be discovered at the time of this examination. Symptoms of CRC are a result of the growth of the tumor into the lumen or adjacent organs. Symptomatic patients may have episodes of bleeding per anum (hematochezia or melena) or increased frequency of bowel dysfunction-constipation or diarrhea and/or vague abdominal discomfort. Weight loss is less common unless the disease is advanced, but fatigue and weakness are frequent. On occasion, the degree of anemia can be severe enough to result in symptoms of cardiac ischemia in patients with occult coronary artery disease and syncope, which leads to the finding of anemia and evidence of bleeding from the colon tumor.

Fatigue and iron-deficiency anemia are more common with right-sided lesions. Gross bright red blood via the rectum, increasing constipation, and cramping abdominal pain are symptoms that suggest a left-sided lesion. Advanced tumors can present with symptoms that are indicative of extension to other organs, such as the bladder (pneumaturia). A perforated colon cancer must always be considered when the patient is seen emergently with abdominal pain, free air, and diffuse peritonitis. At times, the emergent status presents the surgeon with the difficult task of differentiating between an inflammatory mass secondary to a perforated colon cancer of the sigmoid colon or upper rectum and a perforated diverticulitis. Obstruction, however, is the most common acute problem, requiring emergency surgery for colon cancer (with an incidence of approximately 30%). Although obstruction almost always occurs secondary to the tumor that is obstructing the lumen, the tumor can cause intestinal intussusception.

As previously noted, colonoscopy is the most accurate diagnostic test for CRC, as it provides direct visualization, a fairly accurate determination of location, and the opportunity to obtain tumor tissue for histologic examination. The possibility of a second occult primary cancer (synchronous CRC) is approximately 5%, the probability for the presence of additional polyps is approximately 20% to 40%. Both can be diagnosed during colonoscopy and must be cleared or tattooed so that they can be dealt with at the time of surgery.

Staging

The anatomic extent of disease at presentation (stage) is the strongest predictor of survival for patients with CRC and forms the basis of appropriate patient management. The tumor, node, metastasis (TNM) staging system of the American Joint Committee on Cancer (AJCC) and the International Union Against Cancer is considered the international standard for CRC staging.¹³⁰ In contrast with other staging systems for CRC, the TNM system is continuously updated on the basis of existing data, multidisciplinary in design, allows for the incorporation of all technologic approaches to staging, and has a comprehensive set of rules of application that ensure uniform use. The predictive accuracy of TNM staging can be increased through incorporation of validated prognostic features, such as lymphatic or venous invasion, into the overall assessment process. Indeed, although the basic categories for tumor direct extension, distant metastases, and lymph node involvement have not changed since the last edition, a number of new subdivisions have occurred that may have clinical significance (Table 77-4, Fig. 77-5).

Table 77-4

Surgical Stage and Survival Rates in Colorectal Cancer

AJCC/UICC Cancer Staging*				Comparisons			5-Year Survival Rates¶
AJCC/UICC	Tumor	Regional Lymph Nodes	Distant Metastases	Comparison Dukes†	Astler, Collier‡	SEER§
Stage 0	Tis	N0	M0	—	Limited to mucosa	In situ
Stage I	T1	N0	M0	Dukes A	Extending into submucosa	Localized	93.2
	T2	N0	M0	Dukes A	Extending into muscularis propria	Localized
Stage IIA	T3	N0	M0	Dukes B	Extending through muscularis propria	Regional	84.7
Stage IIB	T4A	N0	M0	Dukes B	Extension through bowel wall	Regional	72.2
Stage IIC	T4B	N0	M0	Dukes B	Extension through bowel wall	Regional
Stage III	Any T	N1-2	M0	Dukes C	Limited to or extension through bowel wall with involved lymph nodes (LNs)	Regional	–
Stage IIIA	T1, T2	N1	M0	Dukes C	Limited to bowel wall with involved LNs	Regional	83.4
	T1	N2a	M0	Dukes C	Limited to bowel wall with involved LNs	Regional
Stage IIIB	T3, T4	N1	M0	Dukes C	Extension through bowel wall with involved LNs	Regional	64.1
	T2-T3	N2a	M0	Dukes C	Limited to or extension through bowel wall with involved LNs	Regional
	T1-T2	N2b	M0	Dukes C	Limited to bowel wall with involved LNs	Regional
Stage IIIC	T4a	N2a	M0	Dukes C	Extension through bowel wall with involved LNs	Regional	44.3
	T3-T4a	N2b	M0	Dukes C	Extension through bowel wall with involved LNs	Regional
	T4b	N1-2	M0	Dukes C	Extension through bowel wall with involved LNs	Regional
Stage IV	Any T	Any N	M1	—	Distant metastases	Distant	8.1
Stage IVA	Any T	Any N	M1a	—	Distant metastases	Distant
Stage IVB	Any T	Any M	M1b	—	Distant metastases	Distant

^*International Union Against Cancer. UICC TNM classification of malignant tumors. 7th ed. New York: Wiley-Blackwell; 2010.

^†Dukes CE. The classification of cancer of the rectum. J Pathol Bacteriol 1932;35:323.

^‡Astler VB, Coller FA. The prognostic significance of direct extension of carcinoma of the colon and rectum. Ann Surg 1954;139(6):846–52.

^§Steinhorn SC, Kopecky KJ, Myers MH, Ball C. Characteristics of colon cancer patients reported in population-based tumor registries and Comprehensive Cancer Centers. J Natl Cancer Inst 1983;70(4):629–34.

^¶O’Connell JB, Maggard MA, Ko CY. Colon cancer survival rates with the new American Joint Committee on Cancer sixth edition staging. J Natl Cancer Inst 2004;96(19):1420–5.

In the TNM system, the designation “T” refers to the local extent of the untreated primary tumor, “N” to the status of the regional lymph nodes, and “M” to distant metastatic disease at the time of diagnosis and initial workup. The prefix “p” is used to denote the pathological determination of a staging parameter (i.e., T, N, or M), as contrasted with the clinical determination, which is designated by “c.” By convention, pathologically assigned T and N parameters are considered more accurate than clinically assigned parameters, but require a resection of the primary tumor or biopsy that is adequate to evaluate the highest pT category and nodes that are adequate to validate lymph node metastasis, respectively. Assignment of M0, signifying a global determination of the absence of distant metastasis anywhere in the body, is impossible by pathological means. Therefore, pM0 is not part of the TNM staging system. A pM1 designation, however, is assigned when any distant metastasis is confirmed by tissue or cytologic examination. Clinical classification (cTNM) is based on evidence acquired through a variety of techniques that include, but are not limited to, physical examination, radiologic imaging, endoscopic examination, biopsy, and surgical exploration.

Clinical staging is performed during initial evaluation of the patient before any cancer-directed therapy is initiated and, once assigned, is not changed in the medical record on the basis of subsequent information, remaining as documentation of the basis for treatment planning. It may be revised only if more accurate clinical information becomes available (e.g., higher-resolution imaging studies) before initiation of treatment or before making the decision not to treat the patient. Stage evaluation continues through the first course of surgery or 4 months following presentation, whichever is longer. Staging ends at the time of diagnostic workup if a decision is made not to treat the patient or if pathological classification is not possible. When pathological staging information becomes available following surgical resection, a combined pathological and clinical stage (typically pT, pN, cM) can be constructed.

The definitions of the individual TNM categories and stage groupings and stage-related survival are shown in Figure 77-5, and comparisons of the TNM system to other historical staging systems are shown in Table 77-4. All tumors should be verified microscopically as carcinomas to confirm the appropriateness of TNM staging, which applies only to primary colorectal carcinomas, and to plan appropriate treatment.

Other tumor types that may resemble colorectal carcinoma clinically include colorectal lymphomas, carcinoid tumors, gastrointestinal stromal tumors, metastatic tumors that exhibit tropism for the gastrointestinal tract (e.g., melanoma), malignancies of adjacent organs that directly invade the colorectum (e.g., cancers of the ovary, endometrium, bladder, or prostate), or appendiceal tumors. Benign lesions that can mimic CRC include adenomas, hamartomas, solitary rectal ulcers, stercoral ulcers, endometriomas, and Crohn disease or diverticular disease producing mural strictures.

In detail, these are the individual TNM categories,

(a) Tis: Denotes carcinoma in situ (involving only the mucosa); (b) T1: through the muscularis mucosa and extends into the submucosa; (c) T2: into the muscularis propria; T3: through the muscularis propria and into the external aspects of the colon or rectum; T4a: through the serosa/visceral peritoneum; T4b: attached to or invading into nearby organs.

(a) N0: No cancer in lymph nodes; (b) N1: Involvement of 1 lymph node (N1a), of two to three lymph nodes (N1b) or submucosal satellites with no involved lymph nodes (N1c); (c) N2: involvement of four to six lymph nodes (N2a) or seven or more (N2b).

(a) M0: No distant metastases, M1a in only one organ, M1b in more than one organ.

The TNM categories are translated into the several staging groups, as indicated in Table 77-4.

When establishing the TNM status of a CRC, a number of considerations should apply. Because the colorectal mucosa is architecturally unique and lacks stromal lymphatics, tumor invasion of the lamina propria has no associated risk of regional nodal metastasis (unlike upper gastrointestinal malignancies) and is included in the definition of pTis. Extramural extension of the tumor within lymphatics or veins does not count as local spread of tumor as defined by T3, but is instead denoted as L1 or V1 disease, respectively, in the L or V classification schema of the TNM system. In contrast, discrete smooth-contoured extramural tumor nodules of any size are classified as replaced lymph nodes, and each is counted separately in the N category on the basis of evidence that the number of pericolonic tumor-involved nodes correlates inversely with disease-free survival.¹³¹ Among the features that define T4 tumors, serosal penetration is the most severe.¹³² It is also possible to classify a tumor as pT4 on the basis of positive cytologic specimens of touch preparations from the serosa overlying the primary tumor. Direct invasion of adjacent organs or structures or other segments of the colorectum by way of the serosa or mesocolon (e.g., invasion of the sigmoid colon by carcinoma of the cecum) all should be classified as pT4. In contrast, intramural (longitudinal) extension of tumor from one subsite (segment) of the large intestine into an adjacent subsite or into the ileum (e.g., for a cecal carcinoma) or anal canal (e.g., for a rectal carcinoma) does not affect the pT classification.

Stage-related outcome data are based on pN assignment by conventional histologic staining of lymph nodes that are identified on routine macroscopic examination. Because many nodal metastases in CRC are found in small lymph nodes (<5 mm in diameter), ¹³³ diligent searching for lymph nodes in resection specimens is essential. Indeed, the number of lymph nodes retrieved during the pathological examination of a CRC resection specimen is now considered, by many, to be one of the surrogates of the quality of pathological assessment. Over the last few years, there have been numerous opinions on the number of lymph nodes that represents adequate assessment, anywhere between 7 and 21.¹³⁴ The number of lymph nodes that are recovered from resection specimens varies widely and depends on several factors: while some are unavoidable (patient factors such as age and anatomic variation), others are part of the quality of care (surgical resection technique or diligence of the pathologist in harvesting all existing nodes).

The consensus statement of the College of American Pathologists¹³⁵ suggests that 12 lymph nodes be considered the minimum acceptable harvest from a careful specimen dissection. This is also the number suggested by the Royal College of Pathologists in the United Kingdom, which includes a mean number of 12 lymph nodes retrieved as one of the three measures of standard for a satisfactory colorectal pathology service (the other two being a frequency of serosal involvement of at least 20% in colon cancers and 10% for rectal cancers, and a frequency of extramural venous invasion of at least 25%).¹³⁶ If fewer than 12 nodes are found after careful gross examination, additional techniques (i.e., visual enhancement techniques such as fat clearing) may be considered. It has been further recommended that all grossly negative or equivocal lymph nodes be submitted entirely for microscopic examination and that involvement of grossly positive lymph nodes be confirmed by either complete or partial microscopic examination.

To be included in the N category, a lymph node must be within the regional lymphatic drainage area of the primary tumor. Metastasis in other lymph nodes is classified as M1. Occasionally, however, a CRC may involve more than one site or subsite by continuous longitudinal extension. For example, a cecal carcinoma might extend across the ileocecal valve into the ileum. In these cases, the regional lymph nodes are defined as those of all involved sites and subsites. In rare cases, the regional nodes of the primary tumor site are free of malignancy, but the nodes in the drainage area of an organ directly invaded by the primary tumor contain metastases. In this circumstance, the lymph nodes of the invaded site are considered as those of the primary site and are classified in the N category.

Increasingly, attention is being focused on alternative methods of detection of very small amounts of metastatic tumor in draining lymph nodes, namely isolated tumor cells (ITCs) and micrometastases, and the TNM classification is beginning to focus on these methodologies. To ensure uniform pathological assessment and data collection, small numbers of tumor cells that are detected only by special techniques (including molecular means) or that are seen histologically but measure 0.2 mm or less in diameter are defined as ITCs. According to International Union Against Cancer (TNM) recommendations,¹³⁰ ITCs are classified as N0, but allow suffixes to discern the presence or absence of these, such as: pN0(i−)—negative morphologic findings for ITC; pN0(i+)—positive morphologic findings for ITC; pN0(mol−)—negative nonmorphologic findings for ITC; and pN0(mol+)—positive for nonmorphologic findings for ITC. In contrast, small amounts of metastatic tumor that measure greater than 0.2 mm but less than 2.0 mm are defined as “micrometastases” and classified as N1 or M1. The number of lymph nodes that are involved by micrometastases or ITCs should be stated in the pathology report. To date, the data on the prognostic impact of micrometastases or ITCs in CRC are conflicting.^139–139 Although systematic reviews¹⁴⁰ and retrospective studies¹⁴¹ indicate their validity, only when the results of ongoing prospective studies are revealed¹⁴² will it be possible to determine their full utility.

Systematic reviews suggest that the use of sentinel lymph node analysis (the lymph node closest to the primary lesion) in colon cancer remains controversial.143,144 In any case, the TNM classification also allows for an analysis of ITCs in sentinel lymph nodes in a similar fashion to regular resections, namely pN0(i−)(sn), pN0(i+)(sn), pN0(mol−)(sn), and pN0(mol+)(sn).

Positive peritoneal fluid cytology or tumor that is present (only) in lymphatic vessels of a distant site is considered pM1 disease. Excluded from pM1 designation are ITCs that are found in the bone marrow and tumor foci in the mucosa or submucosa of adjacent bowel, also known as satellite lesions or skip metastases. However, the latter must be distinguished from synchronous primary tumors.

For tumors that are treated prior to staging, stage-related prognosis is altered in comparison to tumors that are staged prior to treatment. This is especially relevant to rectal cancers that are treated with neoadjuvant radiation. However, the rules of application of TNM allow for description of the extent of posttreatment disease using the same TNM parameters modified by the prefix “y.” The classification of residual disease also may be a strong predictor of posttreatment outcome, and the ypTNM classification provides a standardized framework for the collection of data that are needed to accurately evaluate new therapies. Other symbols used in the TNM classification are: the suffix “m” in parenthesis, indicating multiple primary tumors at a single site; the prefix “r,” denoting recurrent tumor; and the prefix “a,” indicating that the classification is established at autopsy.

Stage-related prognosis is also modified by the presence of residual tumor in the patient after primary surgical resection. Stage-related prognosis in disease with no (or limited hepatic) distant metastasis at presentation is predicated on complete eradication of all detectable tumor with cancer-directed surgery. The postoperative residual disease status of the patient, reflecting the efficacy of primary treatment, is categorized by a system known as the R classification, as follows¹³⁰:

• RX—Presence of residual tumor cannot be assessed

• R0—No residual tumor

• R1—Microscopic residual tumor

• R2—Macroscopic residual tumor

The R classification has stage-independent prognostic significance in that R1 and R2 are adverse prognostic factors compared to R0 status at any stage of disease. The relevance of the R classification to the pathologist is primarily related to resection margin evaluation. By convention, tumor that is present microscopically or macroscopically at a resection margin corresponds to R1 or R2, respectively. However, R0 status cannot necessarily be assigned to a case with negative resection margins because R0 refers to absence of residual tumor anywhere in the patient, including distant metastatic sites. Residual tumor is directly linked to the concept of tumor regression after treatment, a concept mentioned in some minimum dataset documents.¹⁴⁵ There is emerging evidence that this is predictive of outcome when resection margins are clear,¹⁴⁶ but there is uncertainty over the best way for it to be assessed. It is, therefore, recommended that at present, only complete regression, or the presence of minimal residual tumor is recorded.

The pertinent margins of a CRC resection specimen include the proximal and distal transverse margins, the mesenteric margin, and, when appropriate, the circumferential (radial) margin. The circumferential resection margin (CRM) represents the retroperitoneal or perineal adventitial soft-tissue margin of the excised bowel. For all segments of the large intestine that are either incompletely encased (ascending colon, descending colon, upper rectum) or not encased (lower rectum) by peritoneum, the CRM is created by surgical dissection of the retroperitoneal or subperitoneal aspect, respectively, at operation.

When the distance between the tumor and the nearest longitudinal margin is 5 cm or more, anastomotic recurrences are very rare.¹⁴⁵ In low anterior rectal resection specimens of rectal cancers, however, wide distal cuffs of normal mucosa can be hard to achieve, owing to anatomic constraints. In this circumstance, a margin of 2 cm is accepted as adequate, and in many cases, distal margins of 1 cm or less also prove sufficient, especially for T1 and T2 tumors.

In rectal carcinoma, the CRM has been demonstrated to be the margin of greatest importance in predicting risk of local recurrence, which is itself a strong predictor of survival.¹⁴⁷ In fact, multivariate analyses have suggested that tumor involvement of the CRM is the single most critical factor in predicting local recurrence in rectal cancer.¹⁴⁸ Emerging data on CRM involvement in the ascending or descending colon suggest a similar relationship to risk of local recurrence.¹⁴⁹ Therefore, routine assessment of the CRM is recommended in all applicable CRCs, and measurement of the distance from the tumor to the nearest CRM, representing the “surgical clearance” around the tumor, is suggested.¹⁵⁰ On the basis of published data from clinical trials, the risk of local recurrence is strongly increased if tumor is present 1 mm or less from the nonperitonealized surface of the specimen.¹⁵¹ By convention, therefore, a positive CRM is now defined as one that is 1 mm or less from the closest approach of tumor. Some data have suggested that the risk of local recurrence also is significantly increased with clearances of 2 mm or less.¹⁵² In contrast, the risk of recurrence is very low with clearance of more than 2 mm and can be considered “negative.”

For segments of the colon that are completely encased by a peritonealized (serosal) surface (e.g., cecum, transverse, and sigmoid colon), the mesenteric resection margin may be a relevant “radial” margin, as tumors may extend to this margin with (pT4) or without (pT3) penetrating the serosal surface. It should be examined when the point of deepest penetration of the tumor is on the mesenteric aspect of the colon, especially when the mesentery has been trimmed close to the colonic wall. For tumors that are limited to an antimesenteric peritonealized aspect of the bowel, the mesenteric margin usually is not relevant.

Other aspects in the pathological analysis that may be of use clinically include:

• Measurement of extramural spread beyond the muscularis propria, both because it is related to prognosis in rectal cancer,¹⁵³ and because of preoperative imaging auditing purposes.¹⁵⁴

• Recording of tumor involvement of the nonperitonealized resection margin to decide on postoperative adjuvant therapy.¹⁴⁹ Grading of the surgical plane of resection in rectal cancer specimens.^153,155

• Recording of degree of tumor regression in rectal cancer patients with preoperative chemoradiotherapy.¹⁵⁶ Recording whether presence and type of tumor perforation within the specimen (serosal vs. retro-/infraperitoneal).

Pathology Markers for Colorectal Cancer

There are no special histopathologic studies, such as histochemical or immunohistochemical assays, that are performed routinely for colon cancer. However, in many instances further studies are necessary.

Positive immunostaining for cytokeratin 20 and negative for cytokeratin 7, although nonpathognomonic, are associated with CRC. This association may be of use in the context of a broad differential diagnosis,¹⁶¹ or to detect small clusters or single infiltrating carcinoma cells (<5 cells) at the invasive edge (so-called tumor budding), as the survival rate for stage II patients with tumor budding might not be significantly different from the survival rate for all stage III patients.^164–164 Other tests that may be relevant in CRC fall under the framework of molecular diagnostics and/or personalized medicine and are indicated in Table 77-6.

Imaging Modalities for Staging of Colon Cancer

The use of imaging for the staging of colon cancer can be divided into two broad categories: (a) detection and staging of the primary tumor and (b) determination of the extent of metastatic disease. For the detection of the primary cancer and of polyps greater than 1 cm in diameter, colonoscopy is now routinely used. When for technical reasons (e.g., partially obstructing cancer, tortuous colon, poor preparation) it is not possible to complete colonoscopy to the ileocecal valve, air-contrast barium enema or virtual colonoscopy are options for completing the evaluation of the colon, although they lack the capacity for biopsy or removal of polyps (Figs. 77-6 and 77-7). The emerging technology of capsule endoscopy has been found to have low sensitivity for detection of polyps and cancer and although less invasive, has the disadvantage that no biopsy or polyp removal can occur during the examination.^129,165 For the staging of potential extracolonic metastatic disease, computed tomography (CT) scan of the thorax/abdomen/pelvis, magnetic resonance imaging (MRI), and positron emission tomography (PET) have been used with varying levels of success.^166,167

Computed Tomography

Contrast-enhanced helical CT scan of the thorax/abdomen/pelvis is considered the mainstay of preoperative imaging of colon cancer patients. CT scan can demonstrate regional tumor extension and tumor-related complications (e.g., obstruction, perforation, fistula formation). In addition, CT scan combines a high sensitivity for the detection of lung and liver metastases with availability, safety, and the ability to aid in the detection of peritoneal disease and metastatic lymph nodes. The sensitivity of CT scan is, however, low for abnormalities that adhere to peritoneal and visceral surfaces, such as peritoneal carcinomatosis or low-volume tumor. An added advantage of CT scan as a preoperative imaging modality for hepatic surgery is that CT angiography can be obtained at the same time as a conventional diagnostic CT scan (Figs. 77-8 through 77-11). This can be important in screening for hepatic arterial and venous anomalies prior to hepatic resection or if the surgeon is considering inserting a hepatic artery infusion pump. CT angiography is an excellent modality for determining the hepatic arterial, portal venous, and hepatic venous anatomy and for assessing variants of these vessels.¹⁶⁸

Magnetic Resonance Imaging of the Liver

Compared with CT, MRI is probably slightly more sensitive and specific for detection and characterization of colon cancer metastases in the liver.171–171 A recent meta-analysis concluded that MRI was the preferred first-line imaging study for evaluating CRC liver metastases in patients who have not previously undergone therapy.¹⁷² MRI is also indicated in evaluating patients with central nervous system symptoms to determine the presence of metastases and for evaluating difficult-to-image pelvic wall recurrences (Fig. 77-12).

Fluorodeoxyglucose Positron Emission Tomography

Fluorodeoxyglucose positron emission tomography (FDG-PET) imaging does not appear to add significant information to CT scans for routine preoperative staging.173,174 The main value of PET scanning in patients with CRC is as an adjunct to the other imaging modalities in the setting of:

1. Localizing site(s) of recurrence in patients who have rising CEA and nondiagnostic conventional imaging following primary treatment. In this context, PET scan can potentially detect occult metastatic disease, thereby selecting patients who may benefit from salvage surgery. In a retrospective analysis of 105 patients who underwent PET scanning and subsequent CT scan plus other conventional diagnostic studies, the investigators found a higher sensitivity (87% vs. 66%) and specificity (68% vs. 59%) for PET scanning compared with CT scan plus other conventional studies¹⁷⁵ for detection of recurrence/metastases.

2. Evaluation of patients who are thought to be candidates for resection of isolated (liver or other) metastases. Flamen and colleagues reported in their study that PET scanning led to a potentially curative resection in 14 of 50 patients with unexplained rising CEA and completely normal or equivocal conventional diagnostic workup.¹⁷⁶

Compared with CT imaging, PET has, however, a lower sensitivity for the detection of hepatic metastases, but a higher sensitivity for the detection of extrahepatic metastases.¹⁷⁷ The superiority of FDG-PET compared with CT scan for detection of extrahepatic disease was also suggested in a retrospective analysis that used a scoring system to weigh the individual studies according to the quality of data and the clinical impact of the radiographic findings.¹⁷⁸ The pooled sensitivity and specificity for the detection of extrahepatic disease for FDG-PET and CT scan were 91% and 98% and 61% and 91%, respectively. The percent change in clinical management from the performance of PET was 25% (range: 20% to 32%). Consequently, many centers now use a combination of CT and PET scanning to stage patients before potential hepatic resection (Figs. 77-13 and 77-14).¹⁷⁹

Recent treatment with chemotherapy may alter the sensitivity of PET for the detection of liver metastases, potentially as a result of decreased cellular metabolic activity of the tumor. Hence, PET results need always to be interpreted in light of recent obtained therapies.

Intraoperative Ultrasound (Open and Laparoscopic)

Intraoperative ultrasound is considered the most sensitive and specific imaging test for detection of hepatic metastases from colon cancer (Fig. 77-15). It is far more sensitive for the detection of metastatic lesions than are inspection and palpation of the liver by the surgeon, particularly for deep tumors.^180,181 Liver metastases as small as 3 mm can be detected routinely by intraoperative ultrasound.¹⁸² Additional advantages of intraoperative ultrasound are its ability to guide biopsies of suspicious hepatic lesions and its use as an aid in both liver resection planning and guiding tumor ablation (e.g., cryoablation or radiofrequency ablation) (Fig. 77-16). Intraoperative ultrasound has a very limited role in the evaluation of extrahepatic metastases from colon cancer because of the highly targeted nature of the examination. Occasionally, retroperitoneal lymph nodes and other metastatic deposits can be imaged if specific abnormalities are noted on preoperative imaging or if an abnormality is noted during inspection and palpation of the abdomen.

Laparoscopic ultrasound can be used for the same purposes as open intraoperative ultrasound of the liver. Because of lack of access to some areas of the liver, the sensitivity of laparoscopic ultrasound appears to be slightly less than that of open intraoperative ultrasound, but laparoscopic ultrasound has the advantage of being a less invasive approach.¹⁸³ This approach is particularly useful in patients who have a high likelihood of unresectable liver disease or peritoneal disease or in patients who are to undergo another minimally invasive treatment for known liver metastases, such as tumor ablation.

Molecular Pathogenesis (the Molecular Basis of Colorectal Cancer)

The knowledge of the molecular pathogenesis of CRC is extensive and, yet, paradoxical. Despite the significant amount of genetic information accumulated over the last number of years, we lack a comprehensive, unified molecular classification for CRC. In 1988, Vogelstein and colleagues proposed a multistep model outlining the key pathogenic changes in oncogenes and tumor suppressor genes for CRC development.¹⁸⁴ This stepwise model of understanding carcinogenesis, matching phenotypic and genotypic changes, provided a prototype framework for the study of oncogenesis in CRC and many other cancer types, and fuelled our understanding of cancer in general. However, while the “Vogelgram” highlights the involvement of specific oncogenes and tumor suppressor genes in the adenoma to carcinoma transition, the precise sequence of genes proposed (Fig. 77-17) required refinement, as it became clear that other molecular paradigms were necessary to explain CRC carcinogenesis. Currently, our understanding of the molecular etiology of CRC, underpinned by the latest technological advances, is informed by a series of diverse pathways and mechanisms which, it is hoped in the future, will be integrated in a single, coherent model for CRC carcinogenesis.

The Epidermal Growth Factor Receptor Pathway

A key pathway in CRC pathogenesis is the epidermal growth factor receptor (EGFR) pathway (Fig. 77-18). EGFR is a transmembrane receptor tyrosine kinase of the ErbB or HER (human epidermal receptor) family, made up of four related receptors: EGFR (ErbB1/EGFR/HER1), ErbB2 (HER2/NEU), ErbB3 (HER3), and ErbB4 (HER4).^185,186 EGFR can become activated by ligand-dependent or ligand-independent mechanisms, overexpression, or mutation. Ligands for EGFR include epidermal growth factor (EGF), transforming growth factor alpha (TGF-α), amphiregulin, betacellulin, and epiregulin. Ligand binding to EGFR induces a conformational change in EGFR that results in activation of its tyrosine kinase domain, phosphorylation of its tyrosine residues, and recruitment of several adaptor proteins (Fig. 77-18). Dependent on the genetic background of the tumor, activated EGFR can result in increased signaling through the MAPK, PI3K-Akt, and STAT pathways, mediating multiple cellular responses, including apoptosis, proliferation, invasion, DNA repair, and survival.¹⁸⁷ EGFR is overexpressed in 25% to 77% and expressed in 80% to 90% of screened patients with CRC.¹⁸⁸ In addition, overexpression of EGFR is associated with poor prognosis but is not predictive for response to EGFR-targeted therapies.

APC-β Catenin-Wnt Signaling Pathway

In CRC, the majority of mutations in the APC gene cause a truncation of the APC protein and therefore a loss of its normal function to degrade cytoplasmic β-catenin. This degradation occurs when β-catenin is bound to the APC–AXIN–GSK-3β complex. GSK-3β acts to phosphorylate β- catenin, leading to its conjugation to the ubiquitin protein and degradation. β-Catenin stability is regulated by diacylglycerol-independent protein kinase C-like kinase activity, which is required for β-catenin ubiquitination (Fig. 77-19).¹⁹⁷ The absence of a functioning APC protein leads to the accumulation of β-catenin in the cytoplasm and its translocation to the nucleus. β-Catenin plays an important role in cell–cell adhesion by linking cadherin receptors to the actin cytoskeleton.

β-Catenin is also related to the Wnt-signaling pathway that determines cell fate, specifically during development. Activation of the Wnt receptor results in phosphorylation and inactivation of GSK-3β, which then prevent GSK-3β from phosphorylating β-catenin.¹⁹⁸ The nuclear accumulation of β-catenin causes it to complex with Tcf-LEF (T-cell factor–lymphocyte enhancer factor). The β-catenin LEF protein complex can activate genes with Tcf-LEF promoter recognition sites, including c-Myc, cyclinD1, PPARS, matrilysn, Fra-1, UPAR, c-Jun, PML, and gastrin. These changes result in enhanced cell proliferation and inhibition of apoptosis.^199,200

Surgical Treatment

Radical resection of the tumor-bearing segment of colon, with wide margins and removal of the lymphatic drainage of the tumor, is the standard for curative therapy. The extent of resection is determined by the tumor size, location, histologic grade, and tumor extension into the colon wall and into adjacent tissue or organs. Historically, resections have been quite radical with respect to the removal of the mesocolon-bearing draining lymphatics at risk for tumor spread; however, there is no major survival advantage to extended lymph node dissection.²⁵³ Limited segmental resections are indicated primarily when the surgery is deemed palliative.

Mechanical Bowel Preparation

Historically, it was believed that mechanical bowel preparation (MBP) was necessary to perform a safe colonic resection. In 1972, Hughes suggested that patients undergoing elective colonic resection may not require MBP; however, this idea was not widely accepted.265,266 In 2011, a Cochrane review, which included 18 studies involving almost 6000 patients, concluded that the absence of MBP is not associated with a greater risk of perioperative morbidity or mortality.²⁶⁷ There was no statistically significant difference in the rate of anastomotic leakage between patients receiving MBP and those who did not. Overall, there was no increase in the risk of intraabdominal or extraabdominal complications; however, there was some evidence that its use may be associated with a greater risk of wound infections. This may be partly the result of a suboptimal bowel cleanse that may leave intraluminal content, which is liquid and, therefore, more likely to spill out intraoperatively. MBP is unpleasant for patients and can result in significant electrolyte imbalance and dehydration.²⁶⁸ Increasingly, MBP is no longer routinely used for elective colonic resection, unless there is a specific indication, such as the need to do an on-table colonoscopy.

Perioperative Nutrition and Fasting and Intraoperative Fluid Management

A period of fasting is a prerequisite for safe general anesthesia to avoid the risk of aspiration of gastric contents. Most anaesthetic associations now advocate the avoidance of solids for at least 6 hours preoperatively and a minimum of 2 hours for clear liquids. This guideline contrasts with the historical prolonged fasting time and “Nil by mouth from midnight” policy. The evidence now clearly shows that prolonged fasting before surgery is detrimental.²⁶⁹ The state of fasting is associated with insulin resistance and there is a correlation between postoperative insulin resistance and length of stay.²⁷⁰ The metabolic response to surgery is one of hypermetabolism and insulin resistance. So, one of the key components of most ERPs is the administration of a carbohydrate drink both the night before surgery and 2 hours before the operation. There is no evidence that this increases the risk of aspiration but it has been shown that in doing so, postoperative insulin resistance is significantly reduced. The implications of this are important, because hyperglycemia postoperatively is associated with increased morbidity and mortality in some patients.

Historically, patients undergoing major colorectal resections received large volumes of intraoperative fluid. There is evidence that excess administration of crystalloids in the perioperative period increases morbidity.²⁷¹ One of the central components of enhanced recovery is goal-directed fluid replacement. Intraoperatively, fluid requirements can be more accurately determined through the use of an intraesophageal Doppler probe. This approach monitors changes in intraoperative stroke volume, and thus the administration of excess, or inadequate fluid, can be avoided. It facilitates the optimization of cardiac function, thus reducing cardiac morbidity.^272,273

Early feeding is a key component of a fast-track recovery program. There is no gain from leaving a nasogastric tube in situ after routine elective surgery. This practice does not reduce the incidence of ileus and it is associated with an increased risk of pneumonia.²⁷⁴ The routine use of intraabdominal drains should be avoided.²⁷⁵

Thromboembolism Prophylaxis

The routine use of thromboembolism prophylaxis is another component of all ERPs. There is a greater risk of thromboembolic complications after colon and rectal surgery than other general surgical procedures. In 1988, a meta-analysis confirmed that the risk of deep vein thrombosis, pulmonary embolism, and resulting fatality is reduced by the routine administration of subcutaneous heparin in general surgical patients.²⁷⁶ Use of adjuncts, including compression stockings and intermittent pneumatic compression, also afford additional protection,²⁷⁷ and early mobilization is paramount.

Laparoscopic Surgery

Approximately 50% of those diagnosed with CRC have potentially curable disease.²⁷⁸ Twenty percent will have metastatic disease at the time of presentation and thus any surgery is likely to be palliative. About one-fifth of CRCs are seen as emergencies. Surgical resection involves removal of the segment of disease-containing colon with a margin of a least 5 cm on either side of the tumor, with the adjacent colonic mesentery containing the regional lymphatic drainage and lymph nodes. In 2000, the National Cancer Institute (NCI) issued guidelines on the surgical management of colon cancer. The resection should include wide excision of the mesentery, with ligation of the feeding artery close to its origin, with a minimum of 12 nodes removed.²⁷⁹ Moreover, the extent of resection must be determined by tumor size, location, histologic grade, and tumor extension into the colon wall and into adjacent tissue or organs.

Until the early 1990s, all colorectal surgery involved a full laparotomy, generally via a midline or paramedian incision. However, since 2000, increasing numbers of CRC resections are being carried out as laparoscopic or laparoscopically assisted procedures. Initially, the use of laparoscopic approach for colorectal resection was considered experimental with the first laparoscopic colonic resection performed in 1991.²⁸⁰ The initial uptake of laparoscopic CRC resection has been slow, a result of initial studies showing higher incidence of port site recurrence.^281,282 More recent studies have disproved the higher incidence of port site metastasis associated with laparoscopic CRC surgery. Data from the main randomized, multicenter trials, which included almost 3000 patients, have provided evidence that laparoscopic resection does not compromise outcomes.^283,284 Other obstacles to the uptake of this novel surgical approach include higher hospital costs and a long learning curve for surgeons.

The rate of laparoscopic colon cancer resections have increased over the last few years with more than 25% of cases completed laparoscopically. This statistic is compared with 8.3% and 17.2% of major resections of CRC completed laparoscopically in the 2006-2007 and 2007-2008 periods, respectively.²⁸⁵ It is now recommended that if both are available, a laparoscopic resection should be offered as an alternative to an open resection.²⁸⁶ It is anticipated that the laparoscopic approach to CRC resection will continue to increase as more surgical trainees are exposed to this as a routine part of their training. Since 2005, several large, randomized, controlled trials including, the Clinical Outcomes of Surgical Therapy study group (COST),²⁸⁷ the Colon Cancer Laparoscopic or Open Resection (COLOR)²⁸⁸ study, and the Conventional versus Laparoscopic-Assisted Surgery in Colorectal cancer (CLASSIC)²⁸⁴ have been carried out, which have provided the basis for this change in practice. Similar to National Institute for Clinical Excellence (NICE) in the United Kingdom, the American College of Colorectal Surgeons support its use.

Laparoscopic Resection Approaches for Colorectal Cancer

Figures 77-20 through 77-23 depict the four classic resections for colon cancer. A right hemicolectomy (Fig. 77-20) is the standard approach for tumors involving the caecum and right colon. The mesocolon and its vascular structures are transected at the base of the mesocolon along the superior mesenteric artery. The ileocolic and right colic arteries are the branches of the superior mesenteric artery that are ligated, along with branches of the middle colic artery. The middle colic artery is usually preserved unless the tumor is located in the hepatic flexure of the colon. In this situation, the resection is extended to include the middle colic artery.

For a right hemicolectomy, the ileum is divided approximately 10 cm from the ileocecal valve. Continuity of the bowel is restored by anastomosis of the terminal ileum to the transverse colon. Although the greater omentum is often removed in part or completely to facilitate resections of the right and transverse colon, there is no evidence that this measure improves survival.

Cancer involving the transverse colon requires mobilization of both the hepatic and splenic flexures. The middle colic artery is transected at the superior mesenteric artery (see Fig. 77-21. When the tumor is located in the splenic flexure region, the surgeon must be more concerned about adequate collateral blood supply through the marginal artery. This concern is increased in elderly patients who have a history of significant vascular disease. Tumors that are located in the left colon are managed by a left hemicolectomy (see Fig. 77-22). The inferior mesenteric artery is ligated near its origin on the aorta. The more distal the location of the cancer, the more mobilization of the upper rectum will be needed. It is not always necessary to mobilize the splenic flexure, but if there is any concern that the anastomosis will be under tension, then it must be performed to provide adequate length. Cancers that involve the sigmoid portion of the colon are resected by what is commonly termed a sigmoid colectomy or a high anterior resection. The more distal sigmoid tumors require resection of the upper rectum to achieve an adequate distal margin.

As has been noted, in colon surgery it is the arterial blood supply that usually dictates the extent of resection. The rule is to provide a minimum of a 5- to 6-cm margin of colon in the resection. Evidence suggests, however, that it is rare to find colon cancer spread beyond 1.2 cm from the gross margin of the tumor.

Figure 77-23 illustrates a subtotal colectomy and ileorectal anastomosis. A subtotal colectomy might be required in patients who harbor synchronous cancer in the right and left colon and patients seen with colon cancer at a young age. Other indications for subtotal colectomy include the presence of an obstructing left-sided cancer, which has led to cecal perforation in the presence of a competent ileocecal valve, extensive diverticular disease, and a right colon lesion with extension into the sigmoid colon or, similarly, a cancer of a redundant sigmoid colon that invades into the right colon.

There are disadvantages to this type of resection and anastomosis. Firstly, the rectum remains in situ and therefore ongoing surveillance of this segment will be required. Secondly, given the short length of large intestine remaining, increasing stool frequency and poor function can be major problems. An alternative to this is a segmental resection of each tumor-containing segment with the formation of two anastomoses. When a resection, of any type, is carried out laparoscopically, the vascular division is generally carried out intracorporeally and the colonic dissection and anastomosis, extracorporeally. An exception to this is an anterior resection in which the bowel is often divided and reanastomosed intracorporeally with the use of laparoscopic and perirectum stapling devices.

Restoring Bowel Continuity

The restoration of bowel continuity can be performed in a number of ways. The bowel can be rejoined in an end-to-end, side-to-side, or side-to-end manner. The anastomosis can be fashioned by hand sewing or the use of a stapling device. There is no evidence that one technique is superior to the other. Schwab and colleagues reported the results of an in vitro study comparing initial bursting strength for three different anastomotic techniques.²⁹⁹ These authors performed 10 hand-sutured anastomoses, 10 biofragmentable ring anastomoses, and 10 stapled anastomoses of fresh human colon that was harvested at time of colon cancer surgery. They measured the pressure that was required to burst the anastomosis and found no significant difference. They concluded that because there were no differences, hand-sutured anastomoses were certainly the least expensive and should remain the standard. However, a stapled anastomosis is generally less time-consuming, thus the decision is generally dictated by surgeon preference.

What is of utmost importance in performing an anastomosis is ensuring that there is adequate blood supply, that it is tension free, and that it has an adequate lumen. These conditions reduce the risk of anastomotic dehiscence and the attendant risk of morbidity and mortality. Generally, an anastomotic leak rate of less than 5% is acceptable.³⁰⁰ In addition to technique being critical to the integrity of the anastomosis, factors such as malnutrition, hypoalbuminemia and cardiovascular disease all increase the risk of anastomotic dehiscence.³⁰¹

Another issue in relation to technique is timing of vascular ligation. Turnbull was one of the early proponents of what is called a “no touch technique.” This technique involves ligation of the lymphovascular pedicle before tumor mobilization to avoid dissemination of tumor cells intraoperatively.³⁰² Although there is evidence that this approach does reduce the number of cells disseminated intravascularly, it has not been demonstrated that this translates into a survival benefit.

Another issue for debate is the use of a cytocidal agent, such as povidone-iodine, for washout of the rectal stump. Theoretically, this could reduce the risk of anastomotic or local recurrence. It is commonly performed by surgeons following anterior resection and before anastomosis of the left colon to the rectal stump. The Association of Coloproctology of Great Britain and Ireland included this recommendation in their 2007 guidelines as good clinical practice. However, this practice is less common in the United States. Although in vitro use of povidone-iodine does effectively kill tumor cells, the presence of blood makes this less effective in vivo.³⁰³ In a small study, Agaba and colleagues failed to demonstrate a reduction in local recurrence rates as a result of intraluminal povidone-iodine washout.³⁰⁴

Surgical Management of Lymph Nodes in Colorectal Cancer

Lymph node metastasis is the most important prognostic factor in locoregional CRC. As was described, colon resection has classically involved taking the mesocolon of the resected bowel at the origin of the vessels supplying the segment of colon to be removed. This ensures removal of the first level of lymph nodes (often called the epicolic and pericolic nodes) and the intermediate or second level of nodes in the resection when the vascular branches supplying the segment are taken at the superior mesenteric artery and inferior mesenteric artery. The rationale for these more or less standard resections is based on the assumption that lymphatic tumor dissemination follows an orderly, sequential route of first-order nodes, second-order nodes, and so forth.

It is known, however, that metastases can be found in distant node groups—paraaortic, celiac, and porta hepatis—even when regional nodes appear to be normal. Supporting this observation is a report of administration of 25-mCi ^99mtechnetium (Tc)-tagged fab–fragment–antibody to CEA before the patients underwent colon resection. Intraoperatively, a hand-held scintillation probe was used, and surgeons removed all nodes that were identified as suspicious. Seven of 20 patients were upstaged, and metastatic spread was found at distant sites of retroperitoneum and renal hilum.³⁰⁵

The accuracy of regional node analysis and its prognostic value are directly proportional to the thoroughness of the surgical technique in removing all regional nodes and the pathological examination of the resection specimen in identifying and harvesting all regional lymph nodes for microscopic assessment (see “Staging” and Table 77-4). The AJCC and the CAP have recommended examination of at least 12 lymph nodes to assign stage II.³⁰⁶ It has been shown that a minimum of 12 to 18 lymph nodes must be examined to accurately predict regional node negativity in CRC.^307–315 For this reason, it has been suggested that 12 lymph nodes be considered the minimal acceptable harvest from a careful specimen dissection, and the National Quality Forum has accepted this recommendation as a quality indicator.³¹⁶ Increasingly, however, the evidence indicates that the greater the number of nodes that are examined, the greater is the likelihood that metastasis will be found. In one study of T3 tumors, nodal metastases were found in 22% of cases if fewer than 15 lymph nodes were harvested from the specimen, whereas 85% of cases with 15 or greater recovered nodes showed metastasis.³⁰⁷

In a study of more than 2400 pT3 CRCs that were resected at a single institution over 45 years, Goldstein and colleagues showed by mathematical analysis that the predictive probability of identifying a single positive node increased in a continuous manner as the number of recovered lymph nodes increased, suggesting that there is no minimum number of nodes that accurately or reliably stages all patients.³⁰⁹ More importantly, it has been shown that clinical outcome is linked to the lymph node harvest. Numerous studies show that conventional pathological examination of increased numbers of lymph nodes is itself associated with an increased survival advantage in stage II disease,^308–314 indicating a positive effect of optimal mesenteric resection by the surgeon, optimal lymph node harvest from the resection specimen by the pathologist, or both.

The large study Intergroup Trial INT-0089, which included 3759 patients with stage III or high-risk stage II disease, showed that overall survival and cause-specific survival both increased as more lymph nodes were examined, suggesting that the number of lymph nodes that are analyzed is itself an independent prognostic variable in CRC.³¹⁰ In contrast, survival is significantly worse for patients with stage II colon cancer, roughly equivalent to stage III disease, in cases with fewer than 7 to 9 lymph nodes harvested from the resection specimen, suggesting clinically significant understaging of disease when too few nodes are sampled.^319–319

Despite the strength of the data demonstrating the critical nature of adequate lymph node assessment in CRC, significant variation in lymph node recovery from resection specimens exists within and among medical centers.322–322 In addition, the average number of lymph nodes examined per specimen is often found to be lower than the minimal recommended number, suggesting that a large number of patients with CRC are staged inadequately.³²³ The number of lymph nodes that are recovered from resection specimens is dependent on several factors. Surgical technique, surgery volume, and patient factors, such as age and anatomic variation, alter the actual number of nodes in a resection specimen, but the diligence and skill level of the pathology examination in identifying and harvesting lymph nodes in the resection specimen also are major factors.³²⁴ Because it has been shown that many nodal metastases in CRC are found in small lymph nodes (<5 mm in diameter), diligent search for lymph nodes is required on gross examination of resection specimens.³⁰⁸

Sentinel node biopsy is an approach that was initially used for melanoma patients and increasingly used for breast cancer patients as a means of improving staging accuracy and reducing the extent of surgical resection needed. However, as previously discussed, sentinel node approaches in CRC have unacceptably high false-negative rates.³²⁵ As a result, there is currently no role for sentinel node biopsy in the management of CRC.

Surgical Management of Obstructing Colorectal Cancer

Approximately 20% of patients with a colonic cancer will present as an emergency and bowel obstruction accounts for 80% of these cases.326,327 Most commonly, bowel obstruction is caused by a left-sided tumor but approximately 20% will arise in the right colon. Typically, patients will present with colicky abdominal pain and distention. They may also have a varying degree of alteration in bowel habit and vomiting. A variety of diagnostic investigations are available. These include plain x-ray of abdomen, CT scan of abdomen, and water-soluble contrast enema. It is important to note the diameter of the caecum in this situation; if the ileocecal valve is competent the small bowel will not be distended. In this setting a caecum of greater than 9 to 10 cm is at imminent risk of perforation and intervention should occur without delay. Endoscopy is not recommended as a diagnostic tool in this situation because of the risk of perforation.

Although the use of a water-soluble contrast medium as an enema may prove useful in the diagnosis of obstruction, barium should not be used. It can become inspissated in the colon distal to the obstruction and if a perforation occurs, entry of barium into the peritoneal cavity can lead to a life-threatening barium peritonitis. Obstructing colonic tumors can be managed in a variety of ways. If the tumor is right-sided, a standard right hemicolectomy is usually performed. However, if the lesion is left-sided, several options are available, including the following:

Although a Hartmann procedure is the treatment of choice in the systemically unwell patient with severe intraabdominal sepsis, it is no longer optimal management in all cases of obstruction. Not only does it necessitate a second operation to restore intestinal continuity, a reversal of the Hartmann procedure is associated with significant morbidity and mortality. A systematic review in 2010, which included more than 6000 patients, confirmed this with morbidity and mortality of up to 50% and 5%, respectively.³²⁹ In addition, a recent Spanish multicenter study found that almost 60% of patients having a Hartmann procedure for obstruction never had it reversed.³³⁰

Increasingly, there is a move toward performing a primary anastomosis with or without a defunctioning loop ileostomy. In this situation, the proximal colon usually needs to be lavaged either through a proximal enterotomy or the placement of a dedicated lavage device via the distal end. A number of studies confirm the safety of this approach without an increase in the risk of morbidity, and specifically no increase in the rate of anastomotic leak.333–333 Some advocate subtotal colectomy in the setting of obstruction, because it allows for the removal of any synchronous tumors. However, the incidence of such lesions is low at 2% to 4%.^334,335 In addition, a randomized study found bowel function after subtotal colectomy and ileorectal anastomosis was generally not as good as after left hemicolectomy.³³⁶ To avoid missing any synchronous lesions it is possible to perform on-table colonoscopy after colonic lavage.

Finally, endoscopic placement of a self-expanding metal stent is another option that has become increasingly used in recent years.337–341 By relieving the acute obstruction it can act as a bridge to elective resection for operable tumors, obviating the need for emergency surgery. In doing so, it may facilitate a primary anastomosis and avoidance of a stoma.^337,338 Emergency surgery for obstruction is associated with significant morbidity, and mortality in the region of 20%.^341–341 However, the use of stents is not without complication and a recent meta-analysis revealed a lower success rate than previously thought and a perforation rate of approximately 6%.³³⁷ Tumor perforation is a particular concern given that this automatically converts the tumor stage to a T4 and may influence long-term survival.

Surgical Management with Involvement of Adjacent Organs

When a colon cancer is adherent to an adjacent structure, surgery should involve resection of the attached viscera, partially or fully, in continuity with the tumor. Although it may appear that the tumor is simply adherent to the adjacent viscera via inflammatory adhesions, in more than 50% of cases there will be actual tumor invasion. Colonic tumors most commonly invade the small bowel, urinary bladder, and abdominal wall.³⁴² By performing an en bloc resection, cancer-specific survival is significantly improved. In a follow-up study of more than 150 patients who underwent multivisceral resection for locally advanced disease, no patient who had an R1 or R2 resection survived 5 years.³⁴³ In contrast, almost 80% of patients who had an R0 resection survived 5 years. Surgically disrupting these sites of adherence can lead to the spillage of tumor cells into the peritoneal cavity. This spillage accounts for the increase in the risk of recurrence and decrease in survival when this “tumor plane” is disrupted.

When the tumor involves a loop of small bowel, this should be resected in continuity with the tumor. Similarly, a tumor that is adherent to the abdominal wall should be resected along with a portion of the abdominal wall. When involving the bladder, again a partial or possibly even a complete cystectomy is required. Although malignancy accounts for approximately 20% of colovesical fistulae, less than 1% of colonic tumors fistulate to the bladder.³⁴⁴ Depending on the site in the bladder and the extent of bladder involvement, resection may necessitate full cystectomy and formation of an ileal conduit.³⁴⁵

Surgical Management of Perforated Colorectal Cancer

Colonic tumors infrequently present with perforation, with incidence ranging from 2% to 9%.346,347 Perforation secondary to a tumor has a poor prognosis and is associated with greater mortality than colonic perforations caused by other pathologies. Surgical mortality ranges from 30% to 40%.³⁴⁶ The higher mortality is caused by a combination of underlying malignancy and sepsis. Free perforations, in which there is widespread purulent or fecal contamination throughout the peritoneal cavity, occur in one-third of patients. In two-thirds of patients, the perforation is “contained” when there is little in the way of contamination beyond a localized abscess. Patients seen with a perforated colon are often systemically unwell secondary to sepsis, making resuscitation the priority in the first instance.

Historically, the optimal surgical management is resection of the perforated segment with formation of an end colostomy (Hartmann procedure). This procedure remains the treatment of choice in the patient who has widespread contamination and/or is hemodynamically compromised as a result of sepsis. However, similar to cases of obstruction, resection with primary anastomosis and proximal defunctioning is now deemed appropriate in certain cases. Chen and colleagues published data comparing the outcomes for in a group of patients with perforated diverticulitis who were treated with a Hartmann procedure versus primary anastomosis and defunctioning. They demonstrated comparable morbidity and mortality between the two groups, provided anastomosis was performed in a stable patient.³⁴⁸ Although several authors have demonstrated the safety of this approach in perforated diverticular disease,^349,350 there is little evidence available regarding its appropriateness in the case of perforated colonic cancer. However, it seems reasonable to apply the same principles to its management.

Surgical Management of the Malignant Colon Polyp

Malignant polyps are defined as adenomas that contain invasive carcinoma which penetrates through the muscularis mucosa into the submucosa. They can be classified as sessile or pedunculated, and are, in essence, pT1 lesions. They do not include polyps that contain intraepithelial or intramucosal carcinoma because these do not carry the potential for metastasis. In various colonoscopic polypectomy series, the incidence of these polyps ranges from 2% to 12%.351,352 With the introduction of screening colonoscopy, the prevalence of these early cancers is likely to increase.³⁵³ Two classification systems exist that can assist in determining the likelihood of an individual polyp having associated involved lymph nodes, thus the need for further formal colonic resection can be determined.

The Haggitt classification refers to the depth of invasion, in the stalk of a pedunculated polyp or the submucosa of a sessile polyp. It ranges from 1 to 4. It is estimated that the risk of nodal metastases is less than 1% for Haggitt levels 1, 2, and 3.351,352,354 Consequently, these lesions are potentially adequately treated by polypectomy alone. However, higher rates of lymph node involvement have been reported with Haggitt level 3 lesions.³⁵⁴ Malignant sessile polyps are all Haggitt level 4 and have a much greater incidence of lymph node metastasis, with reports ranging from 10% to 25%.^357–357

Kitajima and colleagues described an additional classification system for sessile polyps and those pedunculated polyps that invade the submucosa (Haggitt level 4).³⁵⁸ It is broken down into three groups—SM (submucosa) 1, 2, and 3—depending on the depth of submucosal invasion. The greater the depth, the greater the risk of lymph node metastases. In addition to depth, there are a number of other adverse histologic features which will influence ultimate treatment. These include malignant polyps with pedunculated morphology that have unfavorable histologic criteria (partial polyp resection, poorly differentiated carcinoma, vascular or lymphatic invasion, margin of resection <2 mm, or depth of submucosal invasion ≥3 mm from muscularis mucosae).^359,360 For malignant polyps with pedunculated morphology, less than Haggitt level 3 with favorable histologic criteria, polypectomy is considered to be curative; however, close endoscopic surveillance is required.

Critical to the management of this group of patients is a discussion at a local multidisciplinary team meeting, taking into account not only the pathological features, but also patient factors, including comorbidities. The frail elderly patient or those with significant comorbidities may be best served by endoscopic resection alone, if considered high risk for major resection.

Surgical Management of Synchronous Metastatic Disease

Approximately 20% of patients with CRC have metastatic disease at the time of diagnosis.³⁶¹ Metachronous lesions will develop in another 25%. Some authors have suggested that the presence of synchronous lesions predicts a worse outcome than for patients in whom metastases develop at a later date.^362,363 The liver is the most common site of distant metastases. Peritoneal disease is rare, but other sites of potential spread include the ovaries, adrenals, lung, bone, brain, and kidneys. Although at one time a diagnosis of liver metastases automatically led to a palliative approach, treatments have now evolved that offer small numbers of these patients a potential for cure.

Perioperative Chemotherapy

The question as to whether perioperative chemotherapy treatment has an impact on survival in patients undergoing CRC liver metastasectomy was investigated in the European Organization for Research and Treatment of Cancer (EORTC) phase III 40983 (EPOC trial) study. In this trial, 364 patients with up to 4 CRC liver metastases were randomly assigned to perioperative FOLFOX4 (oxaliplatin 85 mg/m² and LV5FU2), 6 cycles before and 6 cycles after surgery (Chemotherapy Arm), or surgery alone (Surgery Arm); 3-year progression-free survival (PFS) results were presented at the meeting of the American Society of Clinical Oncology (ASCO) in 2007.³⁷² Only 303 patients, evenly balanced between the two Arms, actually underwent resection, and 115 of 151 in the Chemotherapy Arm received postoperative chemotherapy. In the Surgery Arm, 152 patients received resection. The 3-year PFS rate was 28.1% in the Surgery-only Arm versus 35.4% for the patients who were randomly assigned to the Chemotherapy Arm (P = 0.058 for intention-to-treat [ITT] analysis). Excluding the ineligible patients (6%), the PFS rate was 28.1% versus 36.2% (P = 0.041), and for those who actually underwent resection, the 3-year PFS rate was 33.2% versus 42.4% (P = 0.025). The observed median overall survival (OS) was higher in the Chemotherapy Arm (61 months) than in the Surgery Arm (54 months) in all randomly assigned patients, and was 64 months in the Chemotherapy Arm versus 55 months in the Surgery Arm in eligible patients. The 5-year OS rate was 51.2% (Chemotherapy Arm) versus 47.8% (Surgery Arm) (all randomly assigned) and 52.4% (Chemotherapy Arm) versus 48.3% (Surgery Arm) (eligible patients) (hazard ratio [HR] = 0.88; 95% confidence interval [CI], 0.68 to 1.14; P = 0.34) (all randomly assigned).

After progression, second-line treatment with chemotherapy was given in 59% (Chemotherapy Arm) and 77% (Surgery Arm) of the patients with cancer progression and repeat surgery occurred in 46% (Chemotherapy Arm) and 40% (Surgery Arm), respectively. The investigators of the study concluded that perioperative chemotherapy with FOLFOX4 improves PFS but does not significantly improve OS over surgery alone. A potential benefit may have been diluted by second treatment lines and by more deaths unrelated to cancer in the Chemotherapy Arm.

So, it remains uncertain whether there is a benefit from neoadjuvant chemotherapy treatment compared with initial resection for patients seen with potential resectable CRC liver metastases and the decision is often made on a case-by-case basis. In addition, the optimal chemotherapy regimen to be used before resection of the liver metastases has not been established. In the CRYSTAL study (which is discussed in greater detail in “Monoclonal Antibodies Targeting EGFR–Cetuximab”), the rate of surgery for metastasis and the rate of R0 resection were both higher in patients with KRAS wild-type tumors who received cetuximab plus FOLFIRI compared with FOLFIRI alone (7.9% vs. 4.6%; P = 0.0633 and 5.1% vs. 2.0%; P = 0.0265, respectively).³⁷³ In the OPUS trial, when the analysis was limited to patients with KRAS wild-type tumors, resectability was increased from 4% to 10%, but these data are based on very small numbers (3 of 73 with FOLFOX versus 6 of 61 for FOLFOX plus cetuximab).³⁷⁴ Similarly, bevacizumab only improved moderately the resectability of liver metastases when added to XELOX or FOLFOX in a large randomized trial (8.4% vs. 6.1% with chemotherapy alone).³⁷⁵

The role of hepatic intraarterial chemotherapy in conjunction with systemic chemotherapy treatment before liver metastasectomy has only been evaluated in smaller studies.376,377 Although the results of these studies seemed promising, this approach is not commonly used by most clinicians; further large randomized trials are needed to evaluate a potential role of neoadjuvant hepatic intraarterial chemotherapy in the preoperative setting.³⁷⁸

Most patients undergoing liver resection will receive adjuvant chemotherapy treatment. A recent pooled analysis from two phase III studies (Fédération Francophone de Cancérologie Digestive [FFCD] Trial 9002 and the EORTC/National Cancer Institute of Canada Clinical Trials Group/Gruppo Italiano di Valutazione Interventi in Oncologia [ENG] trial), evaluated the role of postoperative chemotherapy (5-fluorouracil [5-FU]+ leucovorin [LV]) following resection of CRC liver or lung metastases.³⁷⁹ A total of 278 patients (Chemotherapy, n = 138; Surgery, n = 140) were included in this analysis. Median PFS was 27.9 months in the Chemotherapy Arm as compared with 18.8 months in the Surgery Arm (HR = 1.32; 95% CI, 1.00 to 1.76; P = 0.058). Median OS was 62.2 months in the Chemotherapy Arm compared with 47.3 months in the Surgery Arm (HR = 1.32; 95% CI, 0.95 to 1.82; P = 0.095). Although statistically not significant, the data provided a proof of principle of benefit for post resection chemotherapy treatment. The chemotherapy used in these studies was considered to be inferior and the more active combination FOLFOX (oxaliplatin plus 5-FU+LV) is now used as adjuvant treatment following resection of CRC liver metastases (see also “Adjuvant Oxaliplatin Combinations”).

Radiofrequency Ablation

Radiofrequency ablation (RFA) is sometimes applied following macroscopically incomplete resection of CRC liver metastases. A recent retrospective review of the RFA literature for CRC liver metastases reported a wide range of 5-year survival (14% to 55%), and local recurrence rates (3.6% to 60%).³⁸⁰ There are no randomized prospective trials in patients with potentially resectable liver metastases, comparing RFA with hepatic resection with or without postoperative chemotherapy treatment. The retrospective comparative series suggest lower local recurrence rates, and better PFS, and, in some cases, better OS, following resection.³⁸¹ However, these studies are biased, because patients treated with resection alone versus RFA differ across several important clinical and treatment-related variables. Definitive conclusions cannot be made in the absence of prospective randomized trials directly comparing both approaches.

Managing Complications of Colorectal Cancer Surgery

Postoperative ileus is one of the most commonly encountered complications following colonic resection. In fact, some degree of ileus is essentially a normal physiological response in most cases, but there is a variation in how clinically significant this is.³⁸⁵ The majority will run a benign self-limiting course. There is no evidence that routine use of a nasogastric tube (NGT) prevents a prolonged ileus. In fact, as discussed previously, early feeding and avoidance of a NGT are key components of most fast track and ERPs. Other contributing factors are avoidance of opiate analgesia and fluid overload.³⁸⁶ The use of such multimodal approaches are associated with a reduction in time to passage of flatus and a shorter length of hospital stay.³⁸⁷ However, in the setting of a prolonged ileus, where the patient has persistent vomiting and is still unable to tolerate oral diet, several days postoperatively, more aggressive measures are appropriate. Importantly, in the first instance, is to exclude an underlying cause, such as a postoperative collection or mechanical obstruction. This is generally best done with the use of CT with oral and intravenous contrast.³⁸⁶ Such patients will generally need a period of fasting and placement of an NGT. The dilemma in these cases is when to instigate total parenteral nutrition. Consensus expert opinion is, that in the absence of preoperative nutritional deficiency, it is reasonable to wait at least 5 days before starting total parenteral nutrition.^387,388 Total parenteral nutrition is not without complication, and in particular, there is risk of sepsis from the presence of a central venous catheter to facilitate feeding.

Anastomotic dehiscence is associated with significant morbidity and reported mortality rates range from 6% to 18%.389,390 Although subclinical leaks are probably more common, clinically significant leaks will occur in approximately 4% to 5% of cases.³⁹¹ The median time of diagnosis is 7 days postoperatively. Anastomotic leak is associated with a poor function in the longer term, as well as increasing the risk of a permanent stoma.³⁹² Patients with an anastomotic leak can present in a variety of ways ranging from severe sepsis with generalized peritonitis, to simply having a prolonged ileus or onset of an arrhythmia such as atrial fibrillation. It is important to have a low index of suspicion to diagnose a leak early, allowing for it to be managed expediently. Although some patients will require reexploration, take down of the anastomosis, and formation of a defunctioning stoma, many can be managed conservatively. Those in whom the leak is “contained,” both clinically and radiologically, can often be managed with intravenous antibiotics or radiologically guided percutaneous drainage if there is evidence of a collection.

Wound infections are a common postoperative problem after colonic resection. Risk factors for wound infection may be patient-dependent or related to operative technique and extent of surgery. Patient factors include obesity, smoking,393,394 diabetes mellitus, and immunosuppression. Poor surgical technique and fecal contamination also increase the risk.³⁹⁵ Currently, wound infection rates vary from 5% to 10%. In the elective setting, the infection rate should be less than 5%. This reduction from historically higher rates is associated with the perioperative administration of antibiotics. Many centers now routinely administer a single dose of a broad-spectrum antibiotic, at the time of induction of anesthesia. There is no benefit from the routine administration of additional postoperative doses.³⁹⁶

Wound infections typically present around the fifth postoperative day with swelling and erythema of the wound with or without systemic signs of sepsis. If there is evidence of cellulitis, intravenous antibiotics should be administered. In some cases, it may be necessary to remove several sutures or staples from the skin to allow drainage of a subcutaneous fluid collection. This type of wound can then generally be managed with daily packing and allowed to heal by secondary intention. If there is a more extensive superficial wound dehiscence with a large volume exudate, it may be appropriate to use a negative pressure vacuum dressing to allow more efficient management of the wound discharge and expedite healing.³⁹⁷ If there is concern that the underling muscle sheath is not fully intact, the use of a vacuum assisted device should be avoided because of the potential risk of forming an enterocutaneous fistula. From the available heterogenous studies, the risk seems to range between 6% and 25%.³⁹⁸ Wound dehiscence is uncommon, but the risk is significantly increased for those undergoing emergency surgery for colon cancer.³⁹⁹

Critical to improving a patient’s outcome in CRC is the surgeon’s training, experience, and operative volume. One such report involved a retrospective analysis of 15,427 admissions in northern Illinois undergoing segmental colon resection between 1994 and 1997 performed at 76 nonfederal hospitals with 514 operating surgeons. The authors determined inpatient mortality, complications of surgery, and hospital length of stay. They found that American Board of Surgery certification and increasing years of experience were associated with reduced mortality. Added colorectal certification and site of training did not factor significantly into the outcomes.⁴⁰⁰ In contrast, the COST, which was a randomized, controlled trial examining laparoscopic versus open surgery for colon cancer, did not demonstrate any difference in complications or long-term outcomes based on surgical volume.⁴⁰¹ But when looking at outcomes in relation to rectal cancer, Borowski and colleagues did demonstrate improved cancer survival for those having their surgery in high volume centers.⁴⁰²

Managing Uncommon Colonic Tumors

More than 98% of tumors of the large intestine are adenocarcinomas. Less common tumors are gastrointestinal stromal tumors (GISTs), carcinoids, and melanoma.

GISTs are a type of mesenchymal tumor, previously more commonly known as leiomyomas and leiomyosarcomas. They originate from the interstitial cell of Cajal, the “pacemaker cells” of the gastrointestinal tract, located in the muscularis propria. They all have malignant potential; between 10% and 30% of GISTs are overtly malignant. Tumors less than 2 cm are almost always benign.⁴⁰³ More than 90% have a mutation in the c-kit protooncogene, with the remaining 10% having a mutation in platelet-derived growth factor receptor α (PDGFRα). The mean age of diagnosis is 60 years and GISTs are more common in males. The main criteria that aid in the determination of risk of recurrence and prognosis are, site, size, and mitotic index.⁴⁰⁴ Tumors less than 2 cm with a mitotic index of less than 5 mitoses per 50 high-powered fields have virtually no risk of recurrence. Unless the tumor is not amenable to resection, preoperative biopsy should be avoided because of the potential risk of bleeding and potential for seeding of tumor cells.⁴⁰⁵

Historically, the mainstay of treatment was resection with a reported 5-year survival rate of 28% to 35%,⁴⁰⁶ largely because of the high recurrence rate even with an R0 resection. However, a Japanese study of 60 patients who had a laparoscopic or laparoscopic-assisted resection demonstrated a 5-year survival rate of 89% to 100% for those with tumors between 2 and 5 cm.⁴⁰⁷ The introduction of imatinib has contributed significantly to the improvement in outcome for metastatic tumors and those not amenable to resection.⁴⁰⁸

There is also evidence for the use of imatinib as an adjuvant treatment for those tumors with a high risk of recurrence. Colonic GISTs are relatively uncommon, accounting for less than 5% of cases.409,410 Low-risk colonic GISTs generally have a good prognosis after complete resection.⁴¹¹ Those deemed irresectable may undergo downstaging first with a imatinib before attempted resection or debulking.⁴¹⁰

Carcinoid tumors of the colon are rare and account for less than 8% of all gastrointestinal carcinoids.413–413 They are neuroendocrine tumors that originate in the Kulchitsky cells of the crypts of Lieberkühn and belong to the amine precursor uptake and decarboxylation system. The majority of colonic carcinoids arise in the caecum and ascending colon and most will require colonic resection.⁴¹⁴ Five-year survival rate is approximately 70%, but drops to less than 30% in the presence of distant disease. Less than 10% of these patients will present with carcinoid syndrome.⁴¹⁴ Surgery remains the only potentially curative treatment for colonic carcinoid tumors with no distant metastases. If the tumor is not amenable to resection, palliative debulking procedures or intestinal bypass of the tumor can be performed.⁴¹⁵

Surgical Management of Tumors of the Appendix

Approximately 1% of appendicectomy specimens contain a neoplasm. Eighty percent to 90% of these are carcinoids, with the remaining 10% to 20% including mucinous cyst adenoma, adenocarcinoma, lymphosarcoma, paraganglionoma, and GISTs.⁴¹⁶ Most present as acute appendicitis, the tumor leading to occlusion of the appendix lumen in the same way as a fecalith. The finding of a tumor is usually an incidental one, as a consequence of histologic examination of the surgical specimen.

Carcinoids, which account for 50% to 75% of appendiceal tumors, typically occur at the tip, but can also arise at the base of the appendix. Generally they have low-grade malignant potential and outcomes are good. For tumors with a diameter smaller than 1 to 2 cm, appendicectomy is considered an adequate treatment. However, tumors with diameters larger than 2 cm should be treated with subsequent right hemicolectomy because of the risk of lymph node metastases.⁴¹⁷ Adenocarcinoma of the appendix should similarly be treated with right hemicolectomy. Prognosis is typically worse when compared with that for a primary colonic adenocarcinoma. This is because the tumor has often spread to the peritoneal surface by the time of presentation.

Mucinous cyst adenocarcinoma is an uncommon tumor of the appendix that is potentially difficult to manage if it ruptures. When this occurs, it leads to the development of Pseudomyxoma peritonei, which is characterized by the deposition of mucin pools containing tumor cells.⁴¹⁸ Pseudomyxoma peritonei can also arise from mucinous ovarian tumors. Patients with this condition typically are seen with a distended abdomen caused by mucinous ascites. Aggressive accumulation of malignant mucin in the peritoneal cavity can lead to intestinal or ureteric obstruction, malnutrition and respiratory compromise.

Management of Pseudomyxoma peritonei is best performed in specialist centers. The now well-established Sugarbaker technique,⁴¹⁸ involves complete surgical tumor removal (also known as complete cytoreduction), combined with intraoperative heated chemotherapy, followed by postoperative intraperitoneal chemotherapy. The operation takes approximately 10 hours to complete and includes:

One group has reported complete cytoreduction in 66% of patients (n = 289) with Pseudomyxoma peritonei, secondary to a mucinous appendiceal tumor. The reported 5- and 10-year survival rate was 87% and 74%, respectively. This outcome was compared with a poor outcome of 34% (5-year survival) and 3% (10-year survival) for debulking only.⁴¹⁹ When complete cytoreduction is not possible, there may still be some benefit from debulking surgery. The majority of patients will be suitable for one of these two approaches.⁴²⁰