Carcinoma of the Colon

Neil H. Hyman

There is a tremendous literature on cancer, but what we know for sure about it can be printed on a calling card.

—AUGUST BIER

Excluding cancer of the skin, colorectal carcinoma is the second most common malignancy found in most Western countries. In women, lung and breast cancers are more common, whereas in men, lung and prostate cancers are more frequently observed. In 2010, the American Cancer Society estimated that 141,210 new cases would be diagnosed in the United States, and 49,380 deaths would occur.⁹⁴¹ The chance of colorectal carcinoma developing during the life of an infant born in the United States today is approximately 5%.

The incidence of colorectal cancer had been relatively stable for the 40 years before the last two decades, even as death rates were falling. However, the incidence now appears to be decreasing significantly, with an approximately 3% drop observed in the past 10 years alone. This largely reflects increased screening, with adenoma detection and removal.²⁰⁹^,⁷⁹² Further, the death rate from colorectal cancer has dropped by more than 30% since 1990, owing to earlier detection and better treatment.⁹⁴¹ The trends in cancer incidence, mortality, and patient survival in the United States are derived from the SEER Program (Survey of Epidemiology and End Results) of the National Cancer Institute.⁹⁴¹

▶ ETIOLOGY AND EPIDEMIOLOGY

The epidemiology of large bowel cancer has become a major area of investigative interest. Interpopulation and nationality differences were the inspiration for Burkitt’s hypothesis on the contributory role of a low-residue, high-carbohydrate diet in causing colorectal cancer.¹³⁴^,¹³⁷ His observations generated additional reports about the incidence and mortality of large bowel cancer, and other hypotheses have been suggested⁹²^,³⁹⁶^,³⁹⁸^,⁴⁰²^,⁴⁴²^,⁷⁴²

Distribution and Nationality

The mortality rates for colorectal cancer in Western European countries are generally high.²⁶⁵^,⁸⁴⁶ Scotland is the leader in the world, with rates much higher than even those of England.⁹¹² Spain and Portugal, conversely, have relatively low rates, more consistent with those of Eastern Europe.²⁶⁵ The two populations with risks similar to those of Western Europe are the people of Israel and the Chinese in Singapore. In Israel, a considerable difference has existed between Israelis born in Europe and those born in North Africa or Asia. The incidence in the former is 2.5 times that of the latter. With the exception of Singapore, Asia has a low incidence. African countries generally have a very low frequency of colorectal cancer; in Latin American countries, the incidence rates vary.²⁰⁷^,⁴⁴³^,¹⁰⁷¹

In addition to the differences in the incidence of colorectal cancer from one population to another, observations
have been made concerning the variations in the distribution within the colon and rectum. It has been postulated that low-risk populations have a relatively increased incidence of right-sided cancers, whereas relatively highrisk communities have an increased risk of left-sided malignancies.109,134,137,399

A difference in the incidence of cancer has been observed from country to country when urban populations are compared with rural populations.⁹³^,¹⁸⁶,²⁵⁴,⁵⁷³^,⁷²⁹^,⁹⁶⁴^,¹⁰¹² The nature of the urban-rural gradient with respect to risk has been extensively investigated in the United States. By comparing metropolitan and nonmetropolitan counties, Haenszel and Dawson showed that in the United States, increased risk for large bowel cancer is found in urban populations in each major region of the country.⁴⁰¹

With respect to immigrants, colorectal cancer is less common among Japanese Americans than among white Americans.⁹⁴⁵ However, the rates for Japanese Americans are higher than those for Japanese living in Japan. The children of these immigrants have an incidence approximating that of native white Americans.⁴⁰² A similar phenomenon has been seen in other populations.⁹⁶³

Race

African American men and women have a higher incidence of colorectal cancer than that of Whites, although African Americans and Whites within the same community and region of the country may have a similar rate.²²⁰^,¹¹³⁵ The risk for large bowel cancer in Asian American and Pacific Islander, Native American, Alaska native, and Hispanic/Latino appears to be lower than Whites.⁹⁴¹ There is a higher death rate among African American men and women because they are less likely to be diagnosed at a localized stage, often have more comorbidities and may have diminished access to quality health care.⁹⁴¹ Studies suggest that African Americans who receive medical and oncologic treatment comparable to that of Whites appear to achieve the same outcomes.

▶ SOCIOECONOMIC STATUS AND OCCUPATION

Some studies have shown a higher death rate for colorectal cancer in more affluent people.¹⁸⁷^,⁷⁵² Colombia, a country with a low incidence, reported a higher rate in this group.⁴⁰⁰ However, the vast majority of reports suggest worsened cancer survival in patients with a lower socioeconomic status, typically attributed to patient comorbidities, advanced stage at diagnosis, and diminished access to high-quality care.¹¹²⁶ Two large studies performed in the United States found that patients of lower socioeconomic status were more likely to have advanced cancer at diagnosis.¹⁴³^,¹⁸⁵ Interestingly, a publication from Ontario demonstrated that significant disparities in survival persist even without a difference in stage at presentation, suggesting other factors are playing a role, such as tumor biology and comorbidities.¹⁰⁶ Elevated body mass index (BMI), obesity, and low physical activity appear to increase the risk of colorectal cancer and may be important confounding variables.⁴⁹⁹ Occupation has been investigated as a possible causative factor.⁶⁸⁴^,⁷⁵² The relative affluence associated with some occupations appears to be the reason why certain professionals have a higher incidence of colorectal cancer.

▶ RELIGION

In studies of the religion of patients, Jews have a higher incidence than people of other religions.³⁹⁶^,³⁹⁷^,⁷³⁷^,⁹¹³ Members of the Church of Jesus Christ of Latter-Day Saints (Mormons) have a low incidence.¹³¹^,²⁸²^,⁶¹³ Their religion prohibits the use of tobacco, alcohol, tea, and coffee. Seventh-Day Adventists have a significantly lower rate of colorectal cancer; their church proscribes tobacco and alcohol.⁵⁶⁸^,⁵⁶⁹^,⁷⁹⁴

▶ ALCOHOL AND TOBACCO

The data regarding alcohol as an independent risk factor for colorectal cancer is conflicting.¹⁷⁵^,³²³^,⁵⁹²^,⁷⁸¹ For example, a prospective study of Japanese men in Hawaii revealed an association between the consumption of alcohol and rectal cancer, attributable to a monthly consumption of beer of 500 oz (15 L) or more.⁸⁰⁵ On the other hand, moderate alcohol intake (especially wine) has been found to be protective against the development of distal colorectal cancer.²¹⁰ Two meta-analyses on cigarette smoking and colorectal cancer have demonstrated an increased risk among cigarette smokers; the pooled risk estimate ranged from 1.07 to 1.25.¹⁰⁷^,⁵⁷⁵ Current smokers have been shown to have a relative risk of 1.7 for adenomas when compared with nonsmokers. This effect was even more pronounced for advanced or multiple adenomas in a “dose-dependent” relationship, suggesting a direct role in colorectal carcinogenesis.⁷²¹ A report from Quebec, Canada evaluated the effects of smoking on the risk of colorectal cancer according to anatomic subsite.⁹²³ A positive association with cigar smoking and rectal cancer was observed. There was no statistically significant association with cigarette smoking, but there was a “positive association” with proximal colon cancer.

Diet

Diet is the epidemiologic area that has received the most attention since the mid-1980s 20 years, especially since Burkitt’s observations (see Biography, Chapter 27).¹³³,¹³⁴,¹³⁵,¹³⁶ and ¹³⁷ He and Painter postulated that a high content of fiber was the primary factor responsible for the low incidence of colorectal cancer in African natives.⁷⁷⁷ In essence, their theory states that whatever carcinogen is ingested or produced should be present in a relatively diluted form, and when the transit time is decreased, it is excreted rapidly. Fleiszer and colleagues and Chen and coworkers found that parenteral administration of dimethylhydrazine, a known colon carcinogen, conferred great protection on rats, provided their dietary fiber was increased.¹⁷³^,³⁰³

Other studies have demonstrated correlations between colorectal cancer and additional dietary factors. For example, Nigro (see Biography, Chapter 25) and colleagues demonstrated that cancer in the animal model can be inhibited by an increased fiber intake only when the fat content is relatively low.⁷⁴²,⁷⁴³ and ⁷⁴⁴ They presented a program for possible prevention of colorectal cancer: a 10% reduction in fat consumption, the addition of 25 g of dietary fiber per day, and plant steroids.⁷⁴² These substances have been shown to inhibit the cancer that can be induced by carcinogens. One theory holds that inositol hexaphosphate (phytic acid), an abundant plant seed component present in many fiber-rich diets, is one of the specific agents responsible for suppression of colon carcinogenesis.³⁶²

The mechanism by which increased dietary fiber achieves protection from the development of large bowel cancer remains speculative. Fiber may provide protection by increasing stool bulk and dilution of putative carcinogens in the colonic lumen, by more rapid transit with diminished exposure to injurious agents, and by fermentation of fiber to short-chain fatty acids.⁵⁸² Because fiber is heterogeneous, its mechanism of action within the gut may vary. McIntyre and colleagues studied the effects of three types of dietary fiber in fermentative production of butyrate in the distal colon to ascertain the influence on tumor mass in a rat model of bowel cancer.⁶⁶³ They observed that the fiber associated with high butyrate concentrations in the distal large bowel is protective against large bowel cancer, whereas soluble fibers that do not raise distal butyrate concentrations are not protective.

Sengupta and coworkers conducted a literature search on the effect of dietary fiber on tumor incidence through the use of the MEDLINE database of all case-control, longitudinal, and randomized, controlled studies published in English between 1988 and 2000, as well as animal model studies in the period, 1986 to 2000.⁹¹⁷ Thirteen of 24 case-control studies demonstrated a protective effect of dietary fiber against colorectal neoplasms; conversely, only 3 of 13 longitudinal studies in various cohorts demonstrated a protective effect of fiber. The animal studies were more impressive; 15 of 19 demonstrated protection against tumor induction when compared with controls. A meta-analysis of 25 prospective studies concluded that a high intake of dietary fiber (in particular, cereal fiber and whole grains) was associated with a modest reduction in the risk of colorectal cancer.⁴¹

Fecal Bile Acids

Population-based studies have demonstrated that a Western diet is associated with high levels of fecal secondary bile acids, primarily deoxycholic acid and lithocholic acid, the same trend as has been observed in patients with colon carcinoma.⁴⁶^,⁶⁶² Elevated secondary bile acid concentrations exert detrimental effects both on colonic epithelial structure and function.²³⁵ Fecal bile acid concentration is increased by dietary fat and decreased by dietary cereal fiber. Others have shown an unambiguous connection between the fecal bile acid level and the incidence of dimethylhydrazine-induced colon cancer.⁷¹⁶ Hill and associates determined that the feces of people in Western countries exhibit a high concentration of bile acids when compared with the feces of residents of African and Eastern countries.⁴⁴⁴

Cholesterol

Some investigators have demonstrated a strong correlation between colorectal cancer and a high intake of animal fat and protein.²⁸^,⁴⁵⁵^,⁴⁵⁶ Others opine that the consumption of red meat, or total or saturated fat, has only a weak association with the development of colorectal cancer.¹⁰²⁰ Populations with a high consumption of beef generally have the highest incidence of bowel cancer.²⁵³^,²⁵⁴ The epidemiologic evidence is conflicting, but there seems to be an inconsistent relationship of colorectal cancer with respect to fat and sugar consumption, serum cholesterol, and serum β-lipoprotein.¹¹⁹^,⁶³³^,⁷³⁶^,¹⁰²⁸^,¹⁰⁹² Winawer and colleagues performed a time-trend, case-control study in which serum cholesterol was determined at several intervals before the diagnosis of colon cancer.¹¹⁰² They concluded that although individuals in whom colorectal cancer developed had the same levels of serum cholesterol as the general population initially, during the 10 years before the diagnosis of cancer was established, they demonstrated a decline in cholesterol values. This took place in a population of “control” individuals whose serum cholesterol levels tended to increase with age.

A nested case-control study of 520,000 Western Europeans was performed to examine the association between serum levels of total cholesterol and its constituent lipoproteins and colorectal cancer. This included 1,238 incident cases of colorectal cancer during the study period. High concentration of high-density lipoprotein (HDL) was inversely related to the risk of colon cancer (but not rectal cancer); there was no robust association between other blood lipid concentrations and the risk of colorectal cancer.

Bacteria

Bacteria are thought to play a role in the causation of colorectal cancer; their action on ingested fat or metabolites may be a critical factor. Hill and associates demonstrated that people in the United States and Great Britain have a higher colony count of anaerobic flora and a lower count of aerobic bacteria.⁴⁴⁴ Others have confirmed this observation.²⁶ The similarity between the chemical structure of bile salts and the carcinogen, methylcholanthrene, has been observed. It is not unreasonable to hypothesize that the action of bacteria on bile salts may produce a substance capable of inducing malignant degeneration. Burkitt, in fact, postulated that one of the reasons for the preponderance of carcinoma in the distal bowel is the higher concentration of bacteria in this location.¹³⁴ Nonpathogenic bacteria play an important role in preserving the function and integrity of the gut’s mucosal barrier. Other bacteria produce toxic metabolites that can cause cell mutations and affect intracellular signal transduction. As such, the intestinal microflora may be a modifiable target for diminishing the risk of colorectal cancer. For example, colorectal cancer patients have higher bacterial counts in the Bacteroides/Prevotella group.⁹²⁹ Probiotics continue to receive attention as potential chemoprotective agents, but there remains little convincing evidence of efficacy to date. Limited studies have suggested a possible increase in colorectal cancer in patients with Helicobacter pylori infection.¹¹³⁸

Cholecystectomy

Because of the clinical evidence for an increased quantity of secondary bile acids in the feces of patients with bowel cancer and experimental studies demonstrating that secondary bile acids promote chemical carcinogenesis, cholecystectomy has been implicated as a possible precipitating factor.¹⁰³⁷^,¹⁰⁵⁶ This operation increases secondary bile acids in the enterohepatic circulation.¹⁰³⁷ Other reports have failed to confirm a relationship between gallbladder removal and the subsequent development of colorectal cancer.²^,⁶^,⁹⁵^,⁵⁰⁴ There is, however, some opposing evidence to suggest that more than 10 years following cholecystectomy older women may have an increased risk, especially for right-sided lesions.⁶³²^,⁷⁰⁶ Johansen and coworkers evaluated 40,000 patients with gallstones identified in the Danish Hospital Discharge Register.⁴⁹⁵ A borderline significant association was seen between gallstones and cancer of the colon. Jorgensen and Rafaelsen believe that cholecystectomy per se is not responsible for the apparent association, but rather that gallstone disease itself accounts
for the relationship.⁵⁰¹ A meta-analysis of 33 case-control studies showed a pooled relative risk with cholecystectomy of 1.34 for the development of colorectal cancer; the risk for proximal colon cancer was the most significant.³³⁶

Jejunoileal Bypass

With respect to jejunoileal bypass, an operation that in animals promotes the development of chemically induced bowel cancers, there has been no evidence to date to suggest an increased risk in humans,⁶⁶¹ despite profound alterations in transit time, bile salt metabolism, and fecal flora. In a long-term follow-up study (up to 17 years), Sylvan and associates examined these patients postoperatively by means of colonoscopy and biopsy as well as by flow cytometric DNA analysis.⁹⁹¹ These investigators were not able to verify any colorectal malignant transformation.

Ulcer Surgery

An association has been reported between colorectal cancer and prior peptic ulcer surgery, specifically truncal vagotomy.⁷¹⁷ Mullan and colleagues observed increased proportions of chenodeoxycholic acid and lithocholic acid as well as decreased proportions of cholic acid in the duodenal bile of these individuals.⁷¹⁷ They proposed that abnormalities in bile acid metabolism as a consequence of vagotomy could explain the increased risk for the development of colorectal cancer. The cancer risk associated with peptic ulcer surgery was assessed in a cohort of 1,992 surgical patients who were seen in a peptic ulcer clinic in Glasgow, most of whom had undergone vagotomy and a drainage procedure. The risk of colorectal cancer in long-term follow-up was no greater than the general population, irrespective of procedure.⁴⁹⁰

▶ ASPIRIN

There is considerable evidence to suggest that regular use of aspirin and other nonsteroidal anti-inflammatory agents reduces the risk for the development of colorectal cancer.¹⁰⁰⁰^,¹⁰²⁰^,¹⁰²¹ Giovannucci and colleagues determined the rates of colorectal cancer among women in the Nurses’ Health Study who reported regular aspirin use, comparing the rates in this group with those of women who stated that they did not use aspirin.³³⁷ They concluded that the risk for colorectal cancer is reduced only after 10 or more years of aspirin use. Thun and coworkers found that dietary consumption of vegetables and grains and regular use of aspirin were the only factors having an independent and statistically significant association with prevention of colon cancer.¹⁰²⁰ An overall analysis of four studies assessing the impact of aspirin in the general population (n = 69,535) showed no protective effect for the first 10 years of follow-up.¹⁹⁷ However, analysis of the studies involving higher dose aspirin (300 to 1,500 mg/daily) demonstrated a 26% reduction in the incidence of colorectal cancer over a 23-year follow-up period.¹⁹⁷

Estrogen

Most epidemiologic studies have reported an inverse association between postmenopausal hormone (PMH) therapy and colorectal cancer risk.¹⁶² In a report by Paganini-Hill of 7,701 women who were initially free of cancer and used estrogen replacement therapy, there was a statistically significant reduction in the incidence of colorectal cancer and colorectal cancer deaths compared with those individuals who did not take this replacement medication.⁷⁷⁵ The impact of hormone replacement therapy may be more pronounced based on the molecular pathway of carcinogenesis. In the Iowa Women’s Health Study, PMH reduced colorectal cancer incidence by 18% overall.⁵⁸⁰ The relative risk for microsatellite unstable-low (MSI-L)/MSS tumors was 0.6 when PMH use exceeded 5 years, whereas there was no protection afforded for microsatellite unstable-high (MSI-H) tumors.⁵⁸⁰

Inflammatory Bowel Disease

Patients with inflammatory bowel disease, either ulcerative colitis (UC) or Crohn’s colitis, are at increased risk for the development of colorectal cancer; the estimates of this risk has varied considerably in different studies but appears to accrue beginning 8 to 10 years after the onset of symptoms (see Chapters 29 and 30).⁶¹⁴^,⁶⁸² It seems clear that the duration and extent of disease are key risk factors.⁴⁹⁴^,⁵⁵²^,¹⁰⁵³^,¹¹⁰⁹ A Swedish population-based study of 3,000 UC patients found a relative risk of colorectal cancer of 1.7 for proctitis, 2.8 for left-sided disease, and 14.8 for pancolitis.²⁷⁵ With respect to Crohn’s disease, some reports have been published implicating an association between regional enteritis and small bowel carcinoma.⁵⁷⁸^,⁷¹²^,⁸⁷⁵^,⁹²⁷ The relationship between granulomatous colitis and large bowel cancer is perhaps less emphasized but appears to be real (see Chapter 30).¹¹⁴¹

Radiation

There have been considerable differences of opinion about the risk for the development of colorectal cancer following pelvic irradiation. One report demonstrated an increased risk for women who were irradiated for gynecologic cancer.⁸⁸⁸ Additional studies are required for an accurate assessment of any relationship, especially in light of the increased application of neoadjuvant therapy.

Immunosuppression

Immunosuppressive therapy, especially following organ transplantation, is associated with an increased risk for the development of malignant tumors, possibly including that of the colon and rectum. Renal transplant patients may have up to twice the risk of colon cancer but do not seem to have an increased risk of rectal cancer. A surveillance colonoscopy program for these individuals is recommended.

Appendectomy

McVay reported an appreciably increased incidence of colorectal carcinoma in patients who had undergone appendectomy.⁶⁶⁸ He suggested that the relationship could be explained by immunologic factors. Others have failed to substantiate this relationship.³⁸³^,⁴⁵⁷^,⁴⁶⁷ However, a prior history of appendectomy has been found to be an independent risk factor for decreased survival, worsening the prognosis for those in whom carcinoma of the cecum subsequently developed.²⁹

Ureterosigmoidostomy

Numerous authors have recognized the relationship between ureterosigmoidostomy and carcinoma of the colon at the site of the ureteral implant into the colon.³⁹⁴^,⁴¹¹^,⁵⁴³^,⁶¹⁶^,⁶⁸¹^,⁹⁶²^,⁹⁸¹^,¹⁰⁴²^,¹⁰⁸² The incidence of carcinoma ranges from 2% to 15%, with a lag time in the 20-year range.⁴⁵ The cause may be related to
bathing of the colonic mucosa by urine, the presence of a carcinogen in the urine, the by-products of the interaction of colonic bacteria, and urine or the effects of the ureter, itself, implanted into the colon. Alterations in the mucus glycoproteins in the surrounding mucosa has been described.⁹³⁴

HENRY T. LYNCH (1928-PRESENT)

Henry Lynch was born in Lawrence, Massachusetts on January 4, 1928 and grew up in New York. Although underage, he enlisted in the U.S. Navy in March of 1944 and served in the European and South Pacific theaters. He received his bachelor’s degree from the University of Oklahoma, a master’s degree in clinical psychology from Denver University, and was nearing completion of a PhD in human genetics when he entered medical school at the University of Texas Medical Branch in Galveston, from which he received his MD degree in 1960. He completed several post-degree medical training programs and was on the faculty of the University of Texas: MD Anderson Cancer Center in Houston before coming to Creighton University School of Medicine in Omaha, Nebraska in 1967. Since that time, he has been Chairman of Preventive Medicine and Public Health at Creighton, advancing to full professor in 1972. During the 1960s, when cancer was considered an almost solely environmentally caused disease, Lynch demonstrated mendelian inheritance patterns for a previously unrecognized form of colon cancer (hereditary nonpolyposis colorectal cancer, now known as Lynch syndrome) and for the hereditary breast-ovarian cancer syndrome, which he subsequently helped link to the BRCA1 and BRCA2 genes. In addition, he provided some of the first findings of hereditary malignant melanoma and of familial aspects of prostate and pancreatic cancer. The purpose of this work has been to enable physicians to identify high-risk patients more quickly and accurately, leading to earlier and more effective surveillance, management, and treatment. Henry Lynch’s contributions to cancer research have been recognized by many awards from such groups as the American Cancer Society, Bristol-Myers Squibb, the American Association for Cancer Research, the Susan G. Komen Breast Cancer Foundation, the American Society of Clinical Oncology, and the Jacqueline Seroussi Memorial Foundation for Cancer Research.

It is suggested that this type of diversion of the urinary tract be abandoned, especially in young patients with benign disease.⁹²² Periodic endoscopic evaluation of the bowel is required. Consideration should be given to resecting that area of the colon, with conversion to another diversionaryprocedure.²⁸³^,⁵³³^,⁹⁶²

Congenital Urinary Tract Anomalies

Atwell and coworkers believed they had recognized an association between a family history of congenital anomalies of the urinary tract and the development of colorectal cancer at a young age.³⁸ Their observations should be taken with caution because the numbers reported were small.

Extracolonic Tumors

With respect to the incidence and risk for the development of a metachronous colorectal cancer following an extracolonic primary tumor, one study demonstrates that a patient with breast cancer has the same risk for a colorectal malignancy as for a second primary tumor in the opposite breast.¹¹

An association between sebaceous gland tumors and internal malignancies has been called the Muir-Torre syndrome. The incidence of colorectal malignancies is estimated to be almost 50%.¹⁰⁵⁹ Muir-Torre syndrome appears to be a variant of Lynch syndrome, necessitating genetic evaluation, counseling, and appropriate workup for associated malignancies as described next.⁶⁰⁸

Genetic Predisposition

Genetic influences have been known for some time to be an independent risk factor for the development of colorectal cancer. Of all colorectal cancer cases, 2% to 5% can be attributed to known genetic disorders, such as hereditary nonpolyposis colorectal cancer (Lynch syndrome) or familial adenomatous polyposis (FAP).⁴⁸⁴ However, another 25% of colorectal cancer patients have at least one first-degree relative with colorectal cancer, without a known, defined genetic predisposition.¹⁰⁹⁵ These individuals appear to have double the risk of colorectal cancer when compared with the general population.¹⁴²

This observation could conceivably be a consequence of common environmental exposures, primary genetic factors, the interaction of environment and heredity, or simply of chance.⁶⁰⁶ However, in a prospective study, Rozen and colleagues confirmed the relationship of the family history, even when one member harbors a large bowel neoplasm.⁸⁶⁶ As a consequence, they and others have advocated a screening program for such family members.⁷²⁷ Conversely, there is no evidence to suggest a greater frequency of bowel tumors in spouses of colorectal cancer victims. Fuchs and coworkers conducted a prospective study of almost 120,000 patients who had not been previously examined by colonoscopy or sigmoidoscopy and who provided data on first-degree relatives with colorectal cancer.³¹⁸ The relative risk of cancer in persons with affected first-degree relatives, compared with those without a family history, was 1.72 for one relation and 2.75 for two or more. This increased risk for disease was especially evident among younger individuals (5.37).

Rapid advances in the identification of genetic events that are important in colonic carcinogenesis have been made in the past few years.¹³ Specifically inherited abnormalities such as that for familial adenomatous polyposis have been discussed in Chapter 22. Both acquired and genetic anomalies (ras gene point mutations; c-myc gene amplification; allelic deletion at specific sites on chromosomes 5, 17, and 18) seem to be capable of mediating steps in the progression from normal to malignant colonic mucosa (see Figure 22-15
and Chapter 22).¹³ Chromosomal studies have succeeded in identifying a gene on chromosome 18q that is altered in colorectal cancers.²⁹¹ Allelic deletions have been found to occur in more than 70% of such tumors. These are thought to signal the existence of a tumor suppressor gene in the affected region.²⁹¹ Specifically, an abnormality of the p53 tumor suppressor gene, the most commonly mutated gene in human cancer, is thought to be critical to the development of the majority of human tumors.⁹⁶⁷ In essence, the presence of the gene through its product (the p53 protein) acts to induce cell cycle arrest or apoptosis in response to DNA damage.⁹⁶⁷

Lynch Syndromes

Excluding the polyposis syndromes, carcinoma of the colon has been reported in cancer families, the so-called cancer family syndrome or hereditary nonpolyposis colorectal cancer (HNPCC).⁵⁹⁴^,⁶⁰¹^,⁶⁰³^,⁶⁰⁴^,⁸⁰⁶ This hereditary predisposition has been reported to account for at least 3% of colorectal cancer cases worldwide.⁴⁰⁹ In a prospective multicenter study from Finland, Mecklin and coworkers investigated family history and other risk factors during a 1-year period for all new patients in whom colorectal cancer was diagnosed.⁶⁷³ Lynch and colleagues estimate that the risk for development of colorectal cancer is three times greater than that of the general population if one has a first-degree relative with this condition.⁶⁰⁶ Familial colorectal cancer, however, requires the presence of the disease in two or more first-degree relatives. Itoh and colleagues noted a sevenfold increased risk for colon cancer.⁴⁷⁵ Unfortunately, it has been shown that although family cancer history is commonly obtained during the initial surgical consultation of patients with colorectal cancer, there is a tendency to underestimate the extent and its implications.⁸⁷¹

Based on early observations by Warthin,¹⁰⁶⁸ Lynch and colleagues have defined two clinical variants: Lynch syndrome I or HNPCC and Lynch syndrome II or hereditary site-specific nonpolyposis colonic cancer (HSSCC).⁶⁰⁰

Lynch syndrome I is characterized by the following features:

Autosomal dominance
Early age at onset
Predominance of proximal bowel involvement
Multiple primary colon tumors

Lynch syndrome II is characterized by the same features but, additionally, shows an excess of other adenocarcinomas, particularly involving the endometrium and the ovary.⁶⁰⁴,⁶⁰⁶^,⁶¹² Others have added stomach, small bowel, and urinary tract cancers to the spectrum.⁶⁰²^,⁶¹¹^,¹⁰⁵¹ Itoh and coworkers noted that the risk for breast cancer was increased fivefold, and the lifetime risk was estimated at 1 in 3.7 for first-degree relatives of persons with Lynch syndrome II.⁴⁷⁵ Even carcinoma of the larynx has been suggested to be associated.⁶⁰⁵

In the era of molecular genetics, Lynch syndrome is defined in terms of a germ line mutation in a DNA mismatch repair (MMR) gene.¹⁰¹ The clinical distinctions originally described are largely arbitrary, and the list of associated extracolonic malignancies continues to grow.⁶⁰⁸ These include endometrium, ovary, stomach, hepatobiliary tract, pancreas, upper tract urothelial tumors, brain (Turcot’s syndrome), and sebaceous adenomas (Muir-Torre syndrome).

It is useful at this point to define three terms that have entered the literature in this field: Amsterdam criteria, Bethesda guidelines, and microsatellite instability. These criteria are useful in identifying patients who should undergo MSI testing and/or molecular genetic testing to identify germ line mutation in one of the four MMR genes (MLH1, MSH2, MSH6, and PMS2).

Amsterdam Criteria

In 1991, the following clinical criteria (Amsterdam criteria) were established to facilitate consistency in research and are often applied in diagnosing HNPCC¹⁰⁵¹:

Three or more cases of colorectal cancer in a minimum of two generations
One affected individual a first-degree relative of the others with colorectal cancer
One case of colorectal cancer diagnosed before age 50 years
Exclusion of a diagnosis of familial adenomatous polyposis The criteria have since been modified as follows:
Two cases of colorectal cancer when families are small (one younger than 55 years old)
Two cases of colorectal cancer and one of endometrial cancer or other early-onset cancer

Bethesda Guidelines

The Bethesda guidelines were developed in 1997 from the results of a National Cancer Institute workshop on HNPCC.⁸⁵⁴ These guidelines include all of the criteria described in Amsterdam I and Amsterdam II. Because it was believed that the Amsterdam criteria alone led to an underestimation of the true incidence of HNPCC, additional criteria were added. These include histopathologic (signet ring cell, poorly differentiated), morphologic (right-sided), and less selective clinical criteria. Additionally, the guidelines were proposed to assist in the selection of patients whose tumors should be analyzed for microsatellite instability.¹¹²⁹

Microsatellite Instability

Colorectal cancers demonstrate increased rates of intragenic mutation, characterized by generalized instability of short, tandemly repeated DNA sequences known as microsatellites.³⁸⁵ A high frequency of microsatellite instability (defined as 40% or more of the microsatellite loci) has been found in most patients with HNPCC. This is because of the inactivation of MMR function by the subsequent loss of the second allele that results in length variations of short sequences in HNPCC colorectal cancers.¹¹²⁹ This alteration of dinucleotide repeats in microsatellite sequences, MSI or replication error, is used as a diagnostic criterion of MMR deficiency.⁷⁵⁵ Early-age-at-onset colorectal cancer has been demonstrated to be correlated with high-frequency microsatellite instability tumor status.⁸¹³ It has also been shown to be a marker for predicting development of metachronous colorectal carcinoma after surgery.⁹³⁷

Specific findings of the patients in accordance with Lynch and coworkers were as follows: mean age at the initial colon cancer diagnosis was 44.6 years; of first colon cancers, 72.3% were located in the right side of the colon and only 25.0% were in the sigmoid and rectum; 18.1% of the patients had synchronous colon cancer, with a risk for metachronous lesions at 10 years of 40.0%.⁶⁰⁶ Studies have shown the existence of a genetic defect with a population frequency of 19.0% that is transmitted in a mendelian, dominant mode.²⁶⁶ This defect may predispose the bowel epithelium to the effects of fecal carcinogens.

What clinical clues should lead a physician to suspect the diagnosis of HNPCC? The following have been determined:

Early onset of carcinoma of the colon, especially in the proximal bowel (in the absence of multiple colonic polyps)
Presence of multiple primary cancers (e.g., of the colon, endometrium, ovary)
Having a first-degree relative with early-onset cancers integral to Lynch syndrome II⁶¹⁰

It must be remembered, however, that in the experience of Mecklin and Järvinen, only 40% of patients had a positive family history at the time the tumor was diagnosed.⁶⁷¹ Some have suggested that the flat adenoma, a slightly raised lesion with adenomatous tubules concentrated near the luminal surface, or even a small, flat carcinoma may represent markers for the syndrome (see Chapter 21).¹⁰⁴,⁴⁶⁵^,⁵⁴⁹ When cancers do develop, the incidence of the mucinous type is high.³^,⁶⁰⁶^,⁶⁷⁴ Svenden and colleagues suggest that young individuals with metachronous colorectal cancer developing after a previous diagnosis of colorectal carcinoma could in fact have HNPCC.⁹⁸⁷ The possibility of HNPCC should certainly be considered in adolescents in whom colorectal cancer is diagnosed.⁶²² Such consideration would inevitably lead the surgeon to make a recommendation concerning the nature of the operation. For example, some have suggested that because there is a high risk for the development of a metachronous colorectal cancer following a limited resection, a total colectomy may be indicated.¹⁰⁴⁷ Others concur that close relatives of early-onset cases warrant more intensive colonoscopic screening at an earlier age than do relatives of patients in whom disease is diagnosed at an older age.⁴⁰⁴^,⁴⁰⁵

Screening

The American Society of Colon and Rectal Surgeons (ASCRS) established a task force that led to the publication of Practice Parameters for the identification and testing of patients at risk for dominantly inherited colorectal cancer. The following are the conclusions of that collaborate group⁹⁵⁸^,⁹⁵⁹:

Take a family history.
Document a suspicious pedigree. Request medical records to confirm the diagnosis.
Identify criteria for genetic testing (Amsterdam, Bethesda, microsatellite instability in tumors).
Offer surveillance to families not meeting the aforementioned criteria for genetic testing.
Adhere to all protocols for genetic testing, including institutional review board, informed consent, and counseling.

Surveillance

Because the lifetime risk for the development of colorectal cancer approaches 80% in HNPCC, a surveillance program is recommended. In Lynch syndrome I, the surveillance approach is directed to the bowel exclusively. However, with Lynch syndrome II, one must also be aware of the increased risk for the development of extracolonic tumors. Annual colonoscopy is generally believed to be the preferred screening modality for bowel cancer in these individuals, although some believe that this is too frequent; evaluation of the stool for occult blood is unsatisfactory for this purpose.⁴⁵³^,⁵⁴⁸^,⁶⁰⁶ Because of the increased risk for harboring benign and malignant tumors, colonoscopy is recommended for screening asymptomatic individuals with first-degree relatives having colon cancer, even in the absence of one of the Lynch syndromes.³⁸⁹^,⁹⁷⁸ Indeed, colonoscopy has superceded prophylactic surgery in those with an inherited mutation.⁵⁵⁷ Lynch and associates point out the potential for adverse medicolegal consequences because of failure to diagnose colorectal cancer.⁶⁰⁹

Green and colleagues performed colonoscopic screening on 61 asymptomatic individuals with an affected first-degree relative who had HNPCC.³⁷² Neoplasms were found in 15% and malignancies in 3%. Because of the high incidence of multiple lesions, consideration should be given to the performance of subtotal or total colectomy if surgery becomes necessary for colon cancer.¹⁵⁴ A regular, annual endoscopic follow-up of the residual rectum is still necessary, of course.⁶⁷² The risk for development of rectal cancer has been estimated to be 3% every 3 years after abdominal colectomy for the first 12 years.⁸⁵⁵ There is no doubt that the familial cancer risk associated with early-onset disease outside of the recognized cancer predisposition syndromes is markedly increased.⁴⁹⁶

With respect to Lynch syndrome II patients, annual pelvic examinations are recommended beginning at age 25, including endometrial aspiration biopsy and ovarian ultrasonography.⁶⁰⁶ Prophylactic hysterectomy with bilateral salpingo-oophorectomy should be considered in postmenopausal women and in those who have completed childbearing.⁶⁰⁶

Lynch and Lynch have made a plea for the establishment of computerized registries, such as have been developed for familial adenomatous polyposis, to transmit information about the diagnosis, surveillance, and management of hereditary colon cancer syndromes.⁵⁹⁹^,⁶⁰⁷ Recognizing highrisk families and individuals who would benefit from surveillance should help reduce the incidence of this common malignancy.¹¹²

In the Netherlands, families with HNPCC are monitored in an intensive surveillance program. Of the 35 cancers detected while patients were on the program, all but 2 were reported as identified at a local stage.²⁴⁵

Genetic Testing and Counseling

It is known that HNPCC is caused by germ line mutations in one of four DNA MMR genes: hMSH2, hMLH1, hPMS1, or hPMS2 (see also Chapter 22). It is estimated that defects in two of the known MMR genes, hMSH2 and hMLH1, account for 90% of mutations found in HNPCC families.⁶⁹^,⁸¹⁵ Although many mutations in these genes have been found in HNPCC kindreds, thereby complying with the so-called Amsterdam criteria, little is known about the involvement of these genes in families not satisfying these criteria but showing clear-cut familial clustering of colorectal cancer and other cancers.¹⁰⁸⁷ Wijnen and colleagues found hMSH2 and hMLH1 mutations in 49% of the kindreds that fully complied with the Amsterdam criteria, whereas a disease-causing mutation could be identified in only 8% of the families in which the criteria were not satisfied fully.¹⁰⁸⁷ These results imply that there are significant consequences to genetic testing and counseling in the management of colorectal cancer families.

Once the diagnosis has been established, the importance of genetic counseling has been strongly emphasized by Lynch and colleagues and by others.⁶¹¹^,⁹⁷⁵ It is recommended that all families with suggestive pedigrees should be referred to a geneticist for genetic testing. If the test result is negative for carrying the gene, the family member’s cancer risk drops to that of the general population.⁶⁷⁷ Conversely, the lifetime cancer risk for a gene carrier is approximately 90%. However,
individuals need to know the implications and consequences of genetic test results before acquiescing to the testing.¹¹²²

Commercial testing is available in the United States through OncorMed.^* The company has produced a protocol for testing. People who meet the following inclusion criteria should be tested:

A person with colorectal cancer who has three relatives with colorectal cancer (at least one being a first-degree relative to the other two)
A person with colorectal cancer who has two or more firstor second-degree blood relatives with colorectal cancer
A person with colorectal cancer with onset before 30 years of age
A person with colorectal cancer with onset between 30 and 50 years who has at least one other first- or seconddegree relative with colorectal cancer
A person with colorectal cancer with multiple colon primary tumors
A person with colorectal cancer and another related primary cancer
A relative of an individual with a documented MSH2 or MLH1 mutation

The following people should not be tested:
A person younger than 18 years old
A person with a known diagnosis of ulcerative colitis for 7 or more years, familial adenomatous polyposis/Gardner’s syndrome, hereditary flat adenoma syndrome, Peutz- Jeghers syndrome, familial juvenile polyposis syndrome, or hereditary discrete polyp-carcinoma syndrome
A cognitively impaired person or one unable to provide informed consent
Someone who has a psychological condition precluding testing

Treatment

The Standards Task Force of the American Society of Colon and Rectal Surgeons established practice parameters for the treatment of patients with dominantly inherited colorectal cancer (HNPCC).⁹⁶⁰ The following guidelines have been published:

Treatment must be preceded by thorough counseling about the nature of the syndrome, its natural history, its extracolonic manifestations, and the need for compliance with all recommendations for management and surveillance.
When patients with HNPCC as defined by genotype or compliance with Amsterdam I criteria are diagnosed with more than one advanced adenoma or a colon cancer, they should be offered the options of prophylactic total colectomy and ileorectal anastomosis or hemicolectomy plus yearly colonoscopy.
Patients with HNPCC who have rectal cancer should be offered the options of total proctocolectomy and ileopouch anal anastomosis or anterior proctosigmoidectomy, assuming that the sphincters can be saved.
Female patients with HNPCC and uterine cancer in their families may be offered prophylactic hysterectomy once childbearing is complete or when they undergo surgery for other intra-abdominal conditions.

▶ AGE AND GENDER

The incidence of carcinoma of the colon and rectum increases with age, but the progression also varies by anatomic site, population, and sex. In our experience, the mean age at diagnosis for men was 63 years, and for women, 62 years.²⁰⁴ Cook and associates computed the slopes of the logarithm of the incidence against the logarithm of the age from a number of cancer registries and demonstrated that the slopes of the curves for colon and rectal cancer for men were consistently higher than the slopes of curves for women in almost every population.¹⁹⁶ They noted, furthermore, that this variation in male-female difference was greater for colon than for rectal carcinoma.

In women, colorectal cancer ranks third in the United States in number of cancer deaths, 9%. Lung (26%) and breast (15%) are first and second, respectively.⁹⁴¹ The 2011 estimated cancer incidence is third, after breast and lung (30%, 14%, and 9%, respectively).

In men, colorectal cancer (8%) ranks third, after lung (28%) and prostate (11%), for deaths. Prostate cancer is now the single most common cancer (29%); lung is second (14%), and colorectal, third (9%). Men have a preponderance of rectal cancer and a slight excess of cancer of the descending and transverse colon. The incidence of cancer of the ascending colon and cecum is essentially the same for both sexes according to one report,¹⁶⁰ but according to another, women were found to have more right-sided tumors.⁹⁸⁰

▶ SYMPTOMS AND SIGNS

Change in Bowel Habits

Change in bowel habits is the most frequent complaint of patients with colorectal cancer. The change may be as insignificant as that from a bowel movement every other day to one daily. All too often, people place little emphasis on this observation until a profound alteration occurs. Generally, a more distal lesion creates more obvious symptoms than a proximal one. The reasons for this are threefold: first, it is more “difficult” for formed stool in the distal colon to pass through an area of narrowing than for the relatively liquid stool present in the proximal bowel; second, the lumen of the bowel itself is larger proximally than distally; and third, because of the presence of other symptoms (bleeding, pain, discharge), the patient is more likely to pay attention when a distal tumor produces a change in bowel habits.

Bleeding

Bleeding is the second most common symptom of colorectal cancer. It may be overt or occult. The blood may be bright red, purple, mahogany, black, or inapparent. The more distal the location of the lesion, the less altered the blood will be, and the redder it will appear. Although bleeding can represent a relatively early sign of cancer of the bowel, it is often a neglected symptom. Helfand and colleagues performed a prospective cohort study of 201 individuals who mentioned rectal bleeding as part of their review of systems evaluation and then determined whether such a complaint merits investigation for significant pathology.⁴³⁴ They identified 24% with “serious disease,” including benign and malignant neoplasms and inflammatory bowel disease. The authors concluded that physicians should ask all adults about visible rectal bleeding
and should visualize the entire colon in those who manifest such symptoms. Individuals frequently attribute bleeding to hemorrhoids, particularly if they have had prior difficulty with hemorrhoids. For this reason, it is important to treat bleeding hemorrhoids, so that the presence of this symptom succeeds in alerting the patient to seek medical attention.

Conversely, the physician may mistakenly attribute the bleeding to hemorrhoids. Nothing is more tragic than the misdiagnosis of a potentially curable cancer because of inadequate examination or investigation. Too often, patients are managed with suppositories, creams, and laxatives, and only when symptoms become severe enough is proper investigation undertaken.

Mucus

The presence of mucus, either as a discharge (implying a distal lesion) or mixed with the stool, is another symptom; it often accompanies bleeding. The presence of mucus and bleeding should be considered a highly suggestive combination that necessitates bowel investigation.

Pain

Rectal pain is an unlikely presenting symptom of cancer. The most common reasons for anorectal pain are thrombosed hemorrhoids, anal fissure, abscess, and proctalgia fugax. When rectal cancer produces pain, the lesion usually is very distal or very large. Pain may result from infiltration of the sensitive anal canal or from sphincteric invasion. Such invasion may produce tenesmus, a painful urgency to defecate.

Abdominal pain resulting from tumor implies an obstructing or partially obstructing lesion. This pain is usually colicky in nature and may be associated with abdominal distension, nausea, or vomiting. Intestinal obstruction is a presenting complaint in 5% to 15% of individuals with colorectal cancer. Back pain from retroperitoneal extension of a tumor of the ascending or descending colon is an unusual and late sign.

Mass

A palpable or visible abdominal mass in the absence of other signs and symptoms implies a slow-growing, infiltrative process that may be much more amenable to surgical extirpation than might otherwise be anticipated. Such tumors often metastasize quite late in the course of disease.

Weight Loss

Weight loss, in the absence of other symptoms, is a poor prognostic sign. Inanition and loss of strength and appetite suggest metastatic disease, most commonly to the liver. Presentation with symptoms of metastatic disease occurs in approximately 5% of patients with colorectal cancer. Hepatomegaly is a frequent observation, but pulmonary, cerebral, and osseous metastases as isolated findings may reveal an occult colorectal primary on investigation.

Peritonitis

Perforation with peritonitis is an unusual presentation today (except in certain hospitals that serve an indigent population). Differentiating carcinoma from perforated diverticulitis, particularly with a sigmoid lesion, may be extremely difficult (see Surgical Treatment).

“Appendicitis”

Rarely, carcinoma of the cecum can obstruct the lumen of the appendix and cause signs and symptoms of acute appendicitis.⁷⁴⁵ An even more uncommon phenomenon is perforation of the appendix from obstructing carcinoma of the more distal bowel.⁹⁸⁶ The development of a fecal fistula after appendectomy should lead the physician to suspect an underlying malignancy. Although such presentations are uncommon, any individual older than 50 years old with a presentation of acute appendicitis should be evaluated carefully at the time of surgery for underlying carcinoma.

Inguinal Hernia

It has been thought that inguinal hernia in older men is associated with colorectal carcinoma, especially if the hernia is of relatively short duration.²²⁶^,⁶⁵⁰^,⁸¹²^,¹⁰¹³ Because of this observation, some have advised routine colonic examination for all patients before herniorrhaphy. However, Brendel and Kirsh reported no such association.¹¹⁴ The high incidence of colorectal neoplasms in the general population suggests that the theoretical relationship is more likely to be coincidental. That said, it is known that carcinoma of the colon can present, albeit rarely, as an incarcerated hernia.⁴⁵¹ It is therefore prudent to perform colonoscopic examination prior to undertaking hernia repair in patients experiencing a change in bowel habits or other symptoms suggestive of underlying bowel pathology.

Septicemia

Septicemia from Streptococcus bovis is associated with gastrointestinal neoplasms, especially colonic neoplasms.⁷⁵ Kline and colleagues prospectively studied individuals with sepsis caused by this organism.⁵²⁷ Eight of 15 who completed gastrointestinal evaluation, including colonoscopy, were found to harbor colon carcinoma. The differential diagnosis of a fever of unknown origin includes colorectal cancer and demands appropriate evaluation.⁶⁶⁵ Additionally, fevers of undetermined origin in individuals with known colorectal carcinoma should lead one to investigate the possibility of bacterial endocarditis.⁸⁶¹ Secondary infections of hepatic metastases in an individual with a known primary tumor of the colon are well recognized. However, it is much less appreciated that colonic cancer can be an underlying cause of pyogenic liver abscesses in the absence of metastases. After the usual causes have been excluded, an asymptomatic colon cancer should be considered in the differential diagnosis.¹⁰⁰⁵ The presence of an organism normally found in the colon should heighten the clinician’s suspicion.

Cutaneous Manifestations

Nonmetastatic cutaneous presentations of colorectal cancer have been reviewed by Rosato and associates.⁸⁵⁹ They noted a number of conditions associated with gastrointestinal malignancy, including acanthosis nigricans, dermatomyositis, pemphigoid, and others (see Chapter 9). Such manifestations are rare, but any disseminated skin condition that is unresponsive to conventional therapy should encourage the physician to consider gastrointestinal investigation.

Cutaneous metastases from colorectal carcinoma are extremely unusual except, of course, in the incision or port site
(see later). Even rarer is the individual who presents with skin lesions as the initial complaint.²²⁵ Certainly, the incidence must be less than 1%. Biopsy will usually clarify any confusion as to the diagnosis.

Intussusception

Intussusception in the adult is always a condition that requires surgical treatment. This is in contradistinction to children, for whom medical management (e.g., reduction by barium enema) may result in cure. Patients usually present with signs and symptoms of intestinal obstruction.

Nagorney and associates reviewed the Mayo Clinic experience with 144 cases of adult intussusception treated at that institution since 1910.⁷²⁴ Almost 9 in 10 were associated with a definitive pathologic process: malignant neoplasm, benign tumor, metastatic lesion, or Meckel’s diverticulum (Figures 23-1 and 23-2). Two-thirds of the colonic intussusceptions were associated with primary carcinoma of the colon, whereas only one-third of the intussusceptions of the small intestine were associated with an underlying cancer; most of those malignancies were metastatic. Others have also recognized the association between colonic intussusception and the presence of underlying tumors in the adult.¹¹⁷^,²⁹³^,⁷³⁴^,⁸⁸⁷^,¹⁰⁷⁹ The importance of resection without reduction is discussed later.

FIGURE 23-1. Ileocecal intussusception from cecal carcinoma producing intestinal obstruction. (Courtesy of Rudolf Garret, MD.)

FIGURE 23-2. Partial colonic obstruction secondary to carcinoma of the sigmoid with intussusception. Note the characteristic coiled-spring appearance of the intussusceptum.

Duration of Symptoms

Unfortunately, one continues to be impressed by the long history of symptoms reported by many patients who come to surgery for colorectal cancer. However, those with a short history of symptoms do not have a better prognosis. Individuals with symptoms of less than 5 months’ duration have a higher incidence of resection for cure, but the actual long-term survival has not been shown to be improved.⁴⁷³^,⁸⁹⁷

EVALUATION

Practice Parameters for the Detection of Colorectal Neoplasms

Colorectal cancer screening is relatively inexpensive compared with screening for breast and cervical cancer, with cost estimates suggesting that the amount necessary to prevent one cancer is essentially equivalent to that required to treat a symptomatic patient.⁴²⁵ Although the most cost-effective approach has yet to be identified, screening can decrease mortality by making possible the identification of tumors at an earlier stage and the removal of benign lesions before they become malignant, thus preventing the subsequent development of cancer.¹⁰²⁷ Several protocols have been established by a number of organizations for the detection of colorectal neoplasms at the earliest possible stage.

Clinical guidelines and a rationale for colorectal screening have been endorsed by the American Cancer Society, the American College of Gastroenterology, the American Gastroenterological Association, the American Society of Colon and Rectal Surgeons, the American Society for Gastrointestinal Endoscopy, the Crohn’s and Colitis Foundation of America, the Oncology Nursing Society, and the Society of American Gastrointestinal and Endoscopic Surgeons (SAGES). Guidelines from the United States Multi-Society Task Force were published in 2003 and updated in 2008.⁵⁷²^,¹¹⁰³ The following general recommendations have been proposed:

People with symptoms that suggest the presence of colorectal cancer or polyps should have appropriate diagnostic evaluation and fall outside of screening recommendations.
Personal and familial risk factors need to be evaluated when screening is being considered.
Screening for colorectal cancer and adenomatous polyps should be offered to all men and women without risk factors beginning at age 50.
Physicians should recommend a diagnostic evaluation of the colon to follow up a positive result of a screening test.
Follow-up surveillance should be considered after treatment of colorectal cancer or removal of adenomatous polyps or in the presence of underlying premalignant conditions such as inflammatory bowel disease.
Health care providers who perform the tests should have appropriate proficiency, and the tests should be performed correctly.
Screening should be accompanied by efforts to optimize the participation of both patients and health care providers in screening tests and appropriate diagnostic follow-up.
People who are candidates for screening should be given adequate information on the risks and benefits of the various screening procedures.

In 2006, the American Society of Colon and Rectal Surgeons published guidelines for the detection of colorectal neoplasms based on the recommendations described earlier.⁵²⁸

Determination of Occult Blood

The stool guaiac or orthotoluidine test has been the subject of a number of reports (see Chapter 5). In fact, fecal occult blood tests are the only colorectal screening tests with supporting evidence from prospective, randomized trials, with demonstrated reductions in colorectal cancer mortality of 15% to 33%. Greegor studied patients with known asymptomatic colorectal cancers and found the presence of blood in at least one of three stool specimens.³⁷⁰ Because of the relatively high false-positive rate, he recommended a special diet that succeeded in reducing this rate to approximately 1%. The diet is free of meat, fish, and chicken and is relatively high in roughage (fiber) to stimulate bleeding from an existing lesion. Norfleet’s study, however, failed to demonstrate any benefit with respect to sensitivity or specificity from this diet.⁷⁴⁷ Ostrow and associates studied healthy volunteers and found that the test slide preparation gave consistently positive results after the administration of 25 mL of blood and usually gave positive results with only 10 mL.⁷⁶⁹

A study from the United Kingdom by Tate and colleagues compared three fecal occult tests—Hemoccult, Fecatwin, and E-Z Detect—to determine which is best suited for use in asymptomatic patients.¹⁰⁰⁴ The test most sensitive for blood was Fecatwin; it found 93% of cancers and 69% of other mucosal diseases, but the incidence of false-positive results was three times that of Hemoccult. Home testing methods have thus far demonstrated no advantage through increased compliance to outweigh the lower sensitivity.⁸¹⁴

In Gilbertsen’s study, guaiac testing revealed that cancer was responsible for positive test results in 5.1% of patients, and benign tumors in 24%, with no evidence of gastrointestinal neoplasms in 68%.³³³ A later report from this center evaluated 48,000 asymptomatic patients during a period of approximately 4 years.⁷⁴⁶ Invasive carcinoma was found in 113; more than one-half of the tumors had not breached the seromuscular surface of the bowel wall. Hardcastle and Pye believe that the predictive value of a positive test result for invasive cancer is 11% to 17%, and for adenomas, 36% to 41%.⁴¹³ In another report from the Minnesota Colon Cancer Control Study, annual fecal occult blood testing with rehydration of the samples decreased the 13-year cumulative mortality from colorectal cancer by 33%.⁶³¹

In a prospective randomized trial, Kewenter and coworkers investigated a number of new colorectal neoplasms that developed during the first 7 years after the end of a rescreening program by means of occult blood determination.⁵¹⁶ One hundred one carcinomas were diagnosed in the screened group and 128 in the control group during the follow-up period. The results indicated that screening and rescreening of a population had little influence on the stage of the cancers in the test group compared with controls during this 7-year period.

Robinson and colleagues reported the use of an immunologic fecal occult blood test called Hemeselect, comparing this with standard guaiac testing in almost 1,500 patients who completed both evaluations.⁸⁵⁰ Hemeselect had a much higher positive predictive value for cancer and adenomas than did Hemoccult.

Evaluation of a patient with a positive occult blood determination can be expensive, but other pathologic entities that may be of significance are worth identifying.²⁷² Although the cost versus the benefit of a massive screening program is debatable, no one doubts the value of early diagnosis, especially before a malignancy supervenes.¹⁸^,³³³^,⁵¹⁷^,⁶²¹^,⁶³⁰^,¹¹⁰⁰^,¹¹⁰⁵ Scudamore showed that when a patient has no gastrointestinal symptoms, a 100% possibility of curative resection can be expected, with an 88% 5-year survival.⁹⁰⁹ Mapp and coworkers found that screening by means of occult blood determination improved survival in a randomized, controlled study.⁶³⁴ Others challenge the concept of unsupervised mass screening from the point of view of cost-effectiveness, but because of voluntary services and supplies, such projects are probably useful in educating the public about colon and rectal cancer and the value of early detection.¹⁶³^,⁵²⁰ Lieberman suggests that screening with fecal occult blood testing and sigmoidoscopy, as recommended by the American Cancer Society, may not be as cost effective as screening with colonoscopy.⁵⁷⁶ The guidelines previously mentioned, as published under the auspices of multiple societies, recommend fecal occult blood test screening on an annual basis.¹¹⁰³ Testing of two samples from each of three consecutive stools for the presence of occult blood, followed by colonoscopy, has been demonstrated to reduce the risk of death from colorectal cancer. Average-risk people with an abnormal screening test result by fecal occult blood testing
(i.e., a trace- positive or positive test result from any sample) require an accurate examination of the entire colon and rectum, ideally by colonoscopy.¹¹⁰³ The other option would be to perform a double-contrast barium enema, preferably with flexible sigmoidoscopy (see later and Chapter 5). Newer approaches to screening and additional data are also discussed in Chapter 5.

Digital Rectal Examination

William J. Mayo remarked, “The physician often hesitates to make the necessary examination because it involves soiling the finger.”⁶⁵³ Or, simply stated, if you do not put your finger in, you’ll put your foot in. The index finger has also been termed “God’s bioprobe.” However, the efficacy of digital examination today for identifying cancer of the rectum is less compelling than previously thought. Only 10% of colorectal cancers are potentially within reach of the examiner’s finger. Even when the cancer is palpable, the physician may not be sufficiently cautious and diligent to permit discovery of a lesion. The risks of the examination, however, are nonexistent, and no one can argue the cost versus the benefit.

Digital examination will identify the location of the tumor, anterior or posterior, and whether it occupies part or the whole of the circumference. The tumor may be fixed or movable, ulcerated or scirrhous, exophytic or invasive. Careful palpation of the presacral space may reveal hard lymph nodes suggestive of tumor metastases; this may be a valuable prognostic sign. The fact that the tumor is palpable will often suggest the type of operation possible or whether the lesion is indeed resectable. Fixity may indicate a need for supplemental treatment, such as neoadjuvant therapy. Therefore, despite the numerous, often esoteric studies available to evaluate today’s patient, digital examination of the rectum is still a very important adjunct.

Proctosigmoidoscopy

The rigid sigmoidoscope is one of the most valuable diagnostic tools used (see also Chapter 5). Examination with this instrument may reveal mucosal excrescences, polyps, polypoid lesions, cancer, inflammatory changes, strictures, vascular malformation, and anatomic distortion from extraluminal masses. It may also detect numerous anal conditions, such as fistulas, hemorrhoids, fissures, and abscesses. When the instrument is passed to its full length of 25 cm, perhaps one-half to two-thirds of all cancers of the colon and rectum may be identified. Unfortunately, with the rigid instrument, insertion to its full length is possible in only about 50% of patients, the average penetration being approximately 20 cm.

Many investigators have advocated routine proctosigmoidoscopy for early diagnosis,¹⁵⁰^,¹⁸³^,³³²^,⁴⁴¹^,⁴⁹³^,⁸⁰⁸^,⁹²¹^,⁹⁸⁹ but others have questioned the need for this procedure on an annual basis.²⁵⁹^,²⁶⁰^,⁷⁰⁰ The American Cancer Society encourages physicians to search for colorectal cancer before the onset of symptoms. Optimally, an annual proctosigmoidoscopic examination for all patients 40 years of age and older would be recommended. However, with more than 90 million such persons in the United States, this is indeed an awesome task. It was estimated in 1980 that annual sigmoidoscopic examination on all people more than 40 years old in the United States would cost approximately $2.75 billion.²⁷² Corman and colleagues studied 2,500 consecutive asymptomatic patients with the rigid sigmoidoscope as part of a general examination. Excluded from the study were symptomatic patients and those with a prior history of colorectal disorders. The proctoscope was inserted to a mean length of 20 cm. Adenocarcinoma was found in eight patients, and in two of these, carcinoma developed in polypoid adenomas. A total of 432 benign polypoid lesions were found in 228 patients (9.1%).¹⁹⁹

In this study, all lesions found in patients younger than 50 years old were benign; no cancers were detected before the sixth decade of life. This is not surprising because only 5% of colorectal cancers occur in patients younger than 45 years old.⁴³ Although bowel cancer can develop in a person at any age, because of the limitations of time, space, and personnel, they suggested that routine proctosigmoidoscopic examination be performed in those age groups most likely to benefit from the procedure: patients 50 years old and older.¹⁹⁹ If this criterion were met, 30% fewer examinations would need to be performed.

Selby and colleagues provided the strongest possible evidence of the value of screening rigid sigmoidoscopy.⁹¹⁴ In a case-control study, they determined that individuals who had undergone one or more screening sigmoidoscopic examinations in the preceding 10 years had a 60% to 70% reduction in the risk of death from rectal or distal colon cancer in comparison with those who had not undergone such an examination. Furthermore, their finding that the risk for death from these cancers was markedly reduced for 10 years following a single examination is of considerable interest.⁵⁷¹ The authors concluded that screening once every 10 years may be nearly as efficacious as more frequent examinations.⁹¹⁴

The flexible sigmoidoscope has virtually replaced the rigid instrument for screening purposes (see later and Chapter 6). Flexible sigmoidoscopy every 5 years is one of the recommended options for average risk colorectal cancer screening.⁵⁷² Although no one can argue against the value of screening a greater colonic surface, it is a sad commentary that many surgical residents coming out of training programs today do not know how to use the rigid instrument. There is no question that the rigid instrument is far superior to the flexible one for determining the level of the lesion when a cancer is identified in the rectum or rectosigmoid. It is also generally superior for evaluating the state of the anastomosis as part of a follow-up protocol or when the patient experiences symptoms following resective surgery.

Biopsy

Obviously when a patient has symptoms, a proctosigmoidoscopic examination, at least, is mandatory. When a tumor is identified, a biopsy of the lesion should be performed. This is usually a simple office procedure requiring no anesthetic and the very minimum of special tools. For polypoid, exophytic lesions, appreciable bleeding is rarely a concern after a biopsy. If electrocoagulation equipment is not readily available and bleeding is encountered, pressure with a long, cotton-tip applicator soaked in a topical solution of adrenaline will usually suffice.

The sample for biopsy should be taken from the edge of the lesion at the junction of the tumor and the normal-appearing bowel, and placed in a fixative solution. Notation should be made of the distance from the anal verge to the lower level of the lesion. The size, macroscopic appearance (ulcerated or polypoid), and location should also be recorded.

Flexible Sigmoidoscopy and Colonoscopy

The flexible sigmoidoscope has been recommended as the preferred initial screening tool for colorectal cancer (see also Chapter 6).¹⁰²⁷^,¹⁰⁵²^,¹¹⁰⁶ Its primary advantage is that it allows more proximal evaluation of the bowel. It remains to be seen whether the relatively high cost of the instrument and the time spent in examination are justified by the increased yield. It must be remembered, however, that the rectum is evaluated better by the rigid sigmoidoscope than by the flexible instrument.

The examiner must be wary of performing such procedures as biopsy-cautery and snare excision with the flexible instrument in an inadequately prepared colon. With the limited cleansing regimen commonly employed for this instrument examination, there is a serious risk for explosion when electrical equipment is used for tumor biopsy or removal.

The place of colonoscopy as a screening tool in the asymptomatic, low-risk population has been somewhat controversial. Lieberman and coworkers reported the Veterans Cooperative Study Group experience with 3,197 patients who had been enrolled.⁵⁷⁷ In this group of almost exclusively men, colonoscopic examination demonstrated one or more neoplasms in 37.5% of the patients. As a consequence of this and other studies, the United States government insurance program, Medicare, will pay for colonoscopy as a screening test every decade beginning at age 50. Furthermore, colonoscopic screening for neoplasms in asymptomatic first-degree relatives of patients with colon cancer is strongly recommended.³⁸⁸^,⁶⁶⁹^,⁷⁷³ The aforementioned multiorganizational cancer screening guideline protocol recommends that close relatives (e.g., siblings, parents, and children) of a person who has had colorectal cancer or an adenomatous polyp should be offered the same options as average-risk people, but beginning at the age of 40 years.¹¹⁰³ If colorectal cancer has been diagnosed in the close relative before the age of 55 years, or an adenomatous polyp before the age of 60, special effort should be made to ensure that screening takes place. Interestingly, transplant patients have the same risk for the development of colorectal neoplasms as the general population.⁷⁸⁰ In the absence of risk factors, consideration should be given to offering a colonoscopy every 10 years.

Estimates have indicated that up to 14 million colonoscopies are performed in the United States each year, most often for colorectal cancer screening. The rationale for screening colonoscopy is that identification and removal of adenomatous polyps will prevent later colorectal cancer (or at least detect malignancies at an early stage). Yet, there are no prospective randomized trials attesting to a reduction in either the incidence or mortality associated with colorectal cancer. In the National Polyp Study, the incidence of colorectal cancer after a clearing colonoscopy was reduced by 76% to 90%, as compared to three reference populations.¹¹⁰⁷ However, subsequent studies have not suggested a level of protection anywhere close to this magnitude.¹⁴^,⁸⁴⁹^,⁸⁹³

Colonoscopy has been of demonstrable value for the patient with a known neoplasm by identifying a synchronous tumor (Figure 23-3). Herbst and associates, in a retrospective study of 55 patients, found that 9 (16%) harbored another lesion.⁴³⁷ This discovery caused the operation to be modified. In the report of Reilly and associates, 7.6% of 92 patients had a synchronous cancer.⁸³⁸ All were missed on barium enema examination, and none was more invasive than the index lesion. Other studies have demonstrated that 2% to 8% of patients will have a synchronous carcinoma elsewhere in the colon.⁴⁷⁴^,⁵⁴⁶^,⁷⁷¹^,⁷⁷⁴^,⁸³⁸^,¹⁰¹⁹^,¹⁰⁷⁴ When polyps are present, the risk is even higher. Conversely, perhaps as many as 50% of patients with colon cancer harbor one or more colon polyps.⁷⁷¹

FIGURE 23-3. Colonoscopic examination reveals an ulcerating tumor of the sigmoid colon. The tumor encompasses approximately one-third of the bowel circumference. There is a small polyp proximal to the tumor that should emphasize the importance of total colonoscopic evaluation.

Because many symptomatic patients undergo proctosigmoidoscopy and barium enema sequentially, Winawer suggests that the colonoscope be used if small-to-moderate tumors are present because the instrument can usually slip past the lesion.¹¹⁰¹ He believes that the procedure is more valuable in evaluating patients with known right-sided tumors of the colon. The yield is higher, and the distal bowel can be more easily examined. These studies confirm what has long been suspected: often, the metachronous tumor actually represents the missed synchronous one.

Gilbertsen and coworkers believe that barium enema examination is not sufficiently reliable in evaluating patients for suspected colorectal cancer and have abandoned the routine use of this modality in favor of colonoscopy for patients in whom a colorectal lesion is suspected.³³⁴ They hold that barium enema examination often need not be performed if colonoscopy is diagnostic. Others have recommended the technique as a valuable screening tool for those who are at an increased risk for development of colon cancer (e.g., with a positive family history).³⁸⁶^,⁵⁹³ Most concur, however, that the detection rate of colon cancer is about the same with both studies, provided that they are competently performed and bowel preparation is adequate. Colonoscopy, though, is believed to be superior to barium enema in detecting rectal carcinoma.⁸³⁵ One must remember that even colonoscopy is imperfect. Byrd and colleagues noted a 97% correlation with the resected specimen, but in 3%, lesions were missed.¹⁴⁴ Blind areas in the colon plus misjudgment that the instrument had been inserted to the perceived level were responsible for the majority of colonoscopic errors. Recognizing the limits of this procedure, colonoscopy should be the initial examination that is performed on anyone who presents with signs and symptoms suggestive of a large bowel problem.⁹⁴⁴

Complete visualization of the colon must be accomplished within a reasonable time following identification of a bowel neoplasm. Optimally, this should be performed preoperatively because it may alter the type of operation.⁴⁵⁴ Many patients are submitted to this investigation the day before or the day of colon resection in order to obviate the need for a second bowel preparation. Preoperative barium enema is used only when the size of the lesion precludes passage of the instrument. If total colonoscopy is not undertaken before surgery, it should be carried out by the sixth postoperative week.

Comment

From the economic and manpower perspective, frequent colonoscopy as a screening test cannot be justified in an asymptomatic individual who is not in a high-risk group. The procedure often requires sedation and is not without complication (e.g., a perforation rate of 0.1% to 0.5%). In any event, the reality is that there are insufficient resources for performing total colonoscopy on everyone older than 50 years old, even if this were accomplished every 5 years. The real question is how much is a human life worth, or putting it in another way, how much “good medicine” can one afford!

Cytology

Establishing the diagnosis of carcinoma of the colon by means of cytologic evaluation, washing out the colon with saline solution, has been advocated by some authors (see Chapter 5).¹⁷⁴^,⁸¹⁹ Winawer and colleagues performed brush cytology and lavage on selected patients with colonic neoplasms.¹¹⁰⁴ They and others believe that brush cytology i mproves the yield of tissue diagnosis when combined with biopsy, but lavage cytology alone does not seem to be as useful.⁷¹⁴^,¹¹⁰⁴ Chen suggests that cell brushings may be of particular value when colonic stricture and obstruction prevent the colonoscope from reaching the lesion for biopsy.¹⁷⁴

Stool DNA

Because of the utility of identification of mutations in oncogenes and tumor suppressor genes, it has been suggested that this observation may have a demonstrable advantage over such indirect studies of the stool, such as occult blood determination.¹²^,⁵⁷⁰^,¹⁰³² Some studies have been reported which use purified DNA from stool samples in individuals known to harbor colorectal cancer and have detected these mutations.⁵³⁸^,¹⁰³² Additional work is needed to determine the specificity of these genetic tests in asymptomatic patients and to define more precisely the prevalence of the mutations and the sensitivity of the assay.²⁵⁷ A commercially available, stoolbased, DNA colorectal cancer screening test, Pre-Gen, can be obtained in the United States through EXACT Sciences (http://www.exactsciences.com). Testing stool for molecular markers is a promising and rapidly evolving area.

Barium Enema: Air-Contrast Enema Examination

There are no studies evaluating whether screening doublecontrast (air-contrast) barium enema alone reduces the incidence or mortality of colorectal cancer in individuals who are at average risk for development of the disease.¹¹⁰³ The previously mentioned cooperative recommendation on colorectal cancer screening suggests that an individual be offered this radiologic study every 5 to 10 years.

The barium enema has traditionally been the most commonly employed investigative study for evaluation of carcinoma of the large bowel, but for screening and preoperative assessment, it has been replaced by colonoscopy. Most people believe that despite meticulous double-contrast technique, the examination cannot be performed with the accuracy of colonoscopy (see Chapter 5). Even the most careful and competent of radiologists can overlook a colon carcinoma through failure to observe such subtle points as missing haustral folds; disharmony of interhaustral fold patterns; small, radiolucent filling defects; local contractions; and residuelike masses.³⁰⁷ Also, the absence of any therapeutic potential relegates this examination to a second choice.

When barium enema is performed in the presence of a known rectal carcinoma, it should be accomplished with great care (Figure 23-4). After the procedure, the rectum must be cleansed vigorously by multiple enemas to avoid the possibility of obstruction from inspissated barium.

Figure 23-5 demonstrates a typical apple-core lesion at the rectosigmoid juncture. This segment is fairly long; the normal mucosal pattern is lost in the involved bowel, and characteristically the mucosa overhangs the lesion. Another frequent appearance of carcinoma of the sigmoid is shown in Figure 23-6.

The presence of a pedicle (Figure 23-7) does not rule out the diagnosis of carcinoma. In fact, the illustration shows a polypoid carcinoma on a stalk measuring more than 2 cm in length. However, the radiologist is not in a position to comment whether a polyp is benign or malignant.²⁰⁶

Barium enema study (Figure 23-8) reveals complete retrograde obstruction to the flow of barium at the level of the mid-sigmoid colon. Retrograde obstruction, although an impressive radiologic finding, is not necessarily indicative

of antegrade obstruction. Occasionally, patients will report minimal change in bowel habits, even with this radiologic picture. The resected specimen is shown in Figure 23-9.

FIGURE 23-4. Barium enema study of a patient with known rectal carcinoma invites the hazard of inspissated barium precipitating colonic obstruction. The physician must weigh the value of screening the proximal bowel against this risk. (From Corman ML, Veidenheimer MC, Swinton NW. Diseases of the Anus, Rectum, and Colon. Part I: Neoplasms. New York, NY: Medcom; 1972, with permission.)

FIGURE 23-5. Apple-core carcinoma at the rectosigmoid juncture. (From Corman ML, Veidenheimer MC, Swinton NW. Diseases of the Anus, Rectum, and Colon. Part I: Neoplasms. New York, NY: Medcom; 1972, with permission.)

FIGURE 23-7. The presence of a pedicle (arrow) does not exclude carcinoma. This lesion was entirely malignant. (From Corman ML, Veidenheimer MC, Swinton NW. Diseases of the Anus, Rectum, and Colon. Part I: Neoplasms. New York, NY: Medcom; 1972, with permission.)

FIGURE 23-6. Carcinoma of the sigmoid. An irregularly marginated mass projecting into the lumen of the colon shows the characteristic shoulders of a malignancy.

FIGURE 23-8. Retrograde obstruction to the flow of barium from a carcinoma may or may not be associated with significant obstructive symptoms clinically. (From Corman ML, Veidenheimer MC, Swinton NW. Diseases of the Anus, Rectum, and Colon. Part I: Neoplasms. New York, NY: Medcom; 1972, with permission.)

FIGURE 23-9. String stricture. Resected specimen of the carcinoma causing obstruction shown in Figure 23-8. (From Corman ML, Veidenheimer MC, Swinton NW. Diseases of the Anus, Rectum, and Colon. Part I: Neoplasms. New York, NY: Medcom; 1972, with permission.)

The barium enema study shown in Figure 23-10 demonstrates carcinoma of the sigmoid colon with associated diverticular disease. Differentiating between these two conditions is often difficult. It is sometimes said that the presence of diverticula, as seen in this patient, excludes the diagnosis of carcinoma. This is certainly not true. The most important radiologic distinction is that the mucosal pattern usually is maintained in diverticular disease, whereas in carcinoma the mucosa is destroyed or the pattern is lost. In some patients, however, it is impossible to distinguish between the two conditions, and without further information a resection must be performed.

FIGURE 23-10. Carcinoma of the sigmoid with diverticular disease. Because of the relative frequency of both conditions, this is not an uncommon picture. (From Corman ML, Veidenheimer MC, Swinton NW. Diseases of the Anus, Rectum, and Colon. Part I: Neoplasms. New York, NY: Medcom; 1972, with permission.)

Carcinoma of the sigmoid colon with perforation is shown in Figure 23-11. Under such circumstances, differentiation between carcinoma and diverticular disease is virtually impossible. Another complication of colon cancer is fistulization. Figure 23-12 demonstrates reflux of barium into the upper intestinal tract from a carcinoma near the hepatic flexure.

Figure 23-12 demonstrates carcinoma involving one wall of the cecum in a patient who presented with anemia. Careful bowel preparation is necessary for evaluation of cecal tumors because fecal matter frequently obscures this area.

It is important to remember that not all tumors of the colon originate within the bowel. Metastases of cancers from an ovary, breast, or other sites, as well as direct extension from adjacent organs, can produce a radiologic picture virtually identical to that of an intrinsic lesion. For example, the transverse colon is involved in 8% of pancreatic malignancies (Figure 23-13).²⁷⁷ This is another area in which colonoscopy with biopsy has a particular advantage. Despite the plethora of illustrative radiologic material shown, colonoscopy has indeed become the standard for evaluation and interpretation of colonic tumors. The use of barium enema as a diagnostic tool, particularly for screening, continues to decline rather markedly.²⁸⁰^,⁸⁴⁹ In fact, a barium enema for neoplastic disease is almost a historical curiosity, so effectively has colonoscopy replaced this examination. There is so little interest in gastrointestinal radiology on the part of radiologists today that it is virtually impossible to recruit such an individual with this expertise to any radiology department. Although there is no denying that colonoscopy is a superior study for the evaluation of mucosal disease, barium enema is much preferred for understanding colonic anatomy, extrinsic compression, and identification of intramural lesions. Even in this area, however, barium enema has been relegated to a secondary role because of the advent of computed tomography (CT).

FIGURE 23-11. Perforated carcinoma may be indistinguishable from diverticulitis. (From Corman ML, Veidenheimer MC, Swinton NW. Diseases of the Anus, Rectum, and Colon. Part I: Neoplasms. New York, NY: Medcom; 1972, with permission.)

FIGURE 23-12. Carcinoma with fistula. Note the reflux of barium into the duodenum and stomach through a fistula from a hepatic flexure lesion.

FIGURE 23-13. Partial obstruction to flow of barium at hepatic flexure. The lesion was caused by an invasive carcinoma of the head of the pancreas.

Urologic Evaluation

For many surgeons in the past, an integral part of the preoperative evaluation of a patient about to undergo bowel surgery was intravenous pyelography (IVP).¹⁰⁵⁷ Unexpected findings are often identified through its use. In the experience of Prager and coworkers, ureteral duplication was seen in 2.2% of patients, in addition to a number of other congenital anomalies and serendipitous findings.⁸¹⁰ Moreover, the course of the ureters and the presence of ureteral obstruction can be ascertained, and the postvoiding residual may be estimated. However, in a retrospective study of more than 500 individuals, the incidence of complications was the same in patients with normal and abnormal IVP results as well as in patients who did not undergo IVP.¹⁰⁰² Because most injuries occur after low anterior resection or abdominoperineal resection, it seems reasonable to apply the technique to the preoperative evaluation of those whose surgical procedure predisposes them to an increased risk. However, with the ubiquitous application of CT with intravenous contrast in the preoperative assessment of individuals with colorectal cancer, IVP is unnecessary. The urinary tract is well visualized by means of this investigation.

FIGURE 23-14. Carcinoma of the cecum presenting with bowel obstruction. A: Note retrograde obstruction to the flow of barium by the rounded mass. The radiologic picture is that of an intussusception. B: Following reduction by the enema, the polypoid tumor can be seen to fill the cecum. C: Computed tomography reveals the transverse position of the right colon, with a soft tissue mass in the cecum.

Virtual Colonoscopy

See Chapters 5 and 6.

Computed Tomography

Sophisticated radiologic studies may be worthwhile if they can help in therapeutic decision making before the operation. However, it has been said that CT is not useful in this respect because the presence of metastases does not, in and of itself, contraindicate palliative surgery.⁷⁶⁷ Still, although useful information may be obtained, it rarely causes the proposed colon surgery to be altered (Figures 23-14 and 23-15). Kerner and colleagues studied 158 consecutive patients who underwent CT as part of the preoperative evaluation for
primary colorectal carcinoma.⁵¹⁵ Fifty-six percent had unsuspected findings, of which 35% were considered clinically important. These observations caused the surgeon to alter the proposed operative procedure, or they added additional technical information that was meaningful in the preoperative assessment. Still, the single most common finding was liver metastasis, with abdominal wall or contiguous organ invasion as the second most common finding.

FIGURE 23-15. Computed tomography demonstrates profound colonic dilatation secondary to an obstructing carcinoma of the descending colon (arrow).

The issue of performing closed liver biopsy with the potential risk of dissemination of tumor was addressed by Rodgers and colleagues.⁸⁵² The investigators undertook a multicenter, retrospective review that involved 43 individuals who underwent preoperative biopsy. The authors concluded that there is a significant risk of local dissemination of the tumor, but there was no demonstrated effect on resectability or survival.⁸⁵² Liver surgeons uniformly recommend against percutaneous biopsy if surgery is contemplated.

FIGURE 23-16. Computed tomography demonstrates polypoid mass filling part of the ascending colon, consistent with the subsequently demonstrated apple-core lesion observed on barium enema study (arrows).

The accuracy of CT scan for liver evaluation has improved considerably with the incorporation of methods that can assess volume—that is, injection of contrast medium in order to permit visualization of the parenchyma in the arterial, portal, and delayed phases.²⁴ CT has also been used to assess the stage of the primary tumor, but studies have failed to demonstrate that the technique is sufficiently accurate (Figure 23-16).³¹⁶ Hypodense contrast media (air or 1,000 to 1,200 mL of water) by retrograde administration (CT enteroclysis) immediately before scanning allow the entire
large bowel to be visualized for such lesions.²⁴ The density of the tumor with respect to the adjacent wall markedly increases following injection of the contrast. CT scans have a reported sensitivity of 19% to 67% for the detection of

FIGURE 23-17. Computed tomography demonstrates multiple defects in the liver parenchyma, consistent with metastases.

regional lymphadenopathy⁴⁶⁴^,⁶⁵⁴ and 90% to 95% for liver lesions more than 1 cm in diameter.¹⁰⁶⁷

The main advantage of preoperative CT is that it provides a means for comparison should the patient require subsequent evaluation for the possibility of recurrent tumor.³⁴¹ Depending on the timing of planned surgery, one should always try to obtain a preoperative CT scan. This study is a requisite, however, if one plans to remove the bowel by a laparoscopic technique (Figure 23-17; see Chapter 19). The place of CT and ultrasonographic evaluation of an individual with known or suspected metastatic disease is discussed later in this chapter.

FIGURE 23-18. Carcinoma of the colon. A: Barium enema study reveals classic napkin ring tumor with destruction of the mucosa. B: Ultrasonography demonstrates so-called pseudokidney or target sign. A hypoechoic mass can be seen (arrows) with an echogenic center. The echogenic portion represents the “bull’s-eye” of the target. This is the narrowed lumen. The thickened bowel wall, a consequence of tumor, can be clearly seen. (Courtesy of Cynthia Withers, MD.)

Ultrasonography

Ultrasonographic examination has become well established for the evaluation of a host of conditions within the abdominal cavity—most commonly, assessment of gallstones or gallbladder disease. With the use of sonography alone, however, it has been considered impossible to detect lesions of the colon with any reliability. Figure 23-18 illustrates a classic ultrasonographic finding in a patient with carcinoma of the colon, the “pseudokidney,” or “target” sign. Limberg studied this modality and compared it with hydrocolonic sonography in the diagnosis and staging of colonic tumors.⁵⁷⁹ The latter method consisted of instillation of up to 1,500 mL of water into the colon following the injection of a bowel relaxant. Continuous transabdominal sonographic examination of the large intestine was then carried out beginning at the time of water instillation. In every case, colonoscopy was performed following the examination. With the instillation of water into the colon, it was possible to display the bowel sonographically from the rectosigmoid to the cecum in 97% of the patients examined.⁵⁷⁹ A further advantage was the fact that the layers of the colonic wall could be seen in detail so that the depth of invasion by tumor could be determined. The authors confirmed the lack of value of conventional abdominal sonography. Hydrocolonic sonography is certainly an interesting diagnostic procedure that merits further evaluation.

Transcolorectal endosonography has been suggested to offer an important advantage. The application of this modality for assessing rectal cancers is well established and is discussed in Chapters 5, 7, and 24. Tio and colleagues performed echoendoscopy in some cases with newer instruments that were prototypes being developed.¹⁰²⁴ The accuracy of staging rectal and colonic carcinomas was 81% and 93%, respectively. Although the potential for benefit in managing patients with rec tal cancer is genuine (there may be other surgical options besides resection), the alternatives to colon resection constitute a less valid proposition.

Carcinoembryonic Antigen

Attention to the immunologic aspects of bowel cancer has been stimulated by the findings of Gold and Freedman, who identified an antigen in extracts from colon cancer tissue.³⁴³^,³⁴⁴ This antigen is a glycoprotein absent from normal adult intestinal mucosa but present in primitive endoderm. It was therefore called carcinoembryonic antigen (CEA). Thomson and associates described a radioimmunoassay for CEA in the serum and reported positive results in 97% of patients with colon cancer.¹⁰¹⁵ However, the high accuracy of CEA as a diagnostic test for bowel cancer reported in earlier articles apparently resulted from the fact that most of the patients studied had advanced disease with extensive metastases. In such individuals, CEA is not only frequently detected but also is present at very high levels in the blood, especially when the liver is involved. In cancer localized to the mucosa and submucosa, without invasion into the muscularis propria, the percentage of patients with an elevated test result falls to between 30% and 40%. Even when recurrence is confined to the bowel wall, the results of the test are usually negative. Therefore, the use of CEA as a screening technique for the asymptomatic population cannot be justified.³⁴²^,⁴³⁶^,⁸⁴⁴

Despite this admonition, levels of CEA can be applied usefully in assessing the prognosis of individuals with colorectal cancer. If the tumor has been completely excised, any elevated level preoperatively should return to normal within a few days. A limited fall to an intermediate, albeit elevated, level is indicative of incomplete excision. Subsequent elevation after return to normal implies recurrence of tumor (see later). There have been many articles attesting to the fact that preoperative and postoperative assessment of the serum CEA level is extremely helpful in determining whether tumor has been left behind after operation. This is true not only after resection of the primary cancer but also after resection for recurrent tumor.⁴⁴⁸ Specifically, by determining preoperative and postoperative CEA levels, one can identify patients in a poorer prognostic group who may benefit from early introduction of adjuvant therapy.¹⁷⁸

PHIL GOLD (1936-PRESENT)

Phil Gold was born in Montreal, Canada and attended McGill University in that city, where he obtained four degrees. He received a bachelor of science in physiology in 1957 and his doctorate in medicine as well as master of science in 1961. Medical graduation was accompanied by the Wood Gold Medal, the J. Francis Williams Scholarship in Medicine and Clinical Medicine, the Women’s Pavilion Prize in Obstetrics and Gynecology, and the Prize of the College of Physicians and Surgeons of the Province of Quebec in Pathology and Medicine. After a year of rotating internship and another of residency in internal medicine, the next 2 years were spent in the laboratories of the Montreal General Hospital Research Institute; he obtained a PhD for his thesis Carcinoembryonic Antigens of the Human Digestive System. Gold’s discovery of the carcinoembryonic antigen (CEA), along with the description of α-fetoprotein at about the same time, ushered in the modern era of human tumor marker research. The radioimmunoassay for circulating serum CEA has become the most frequently used marker for human cancer. During training at the Public Health Research Institute of New York City in 1967 and 1968 as a Centennial Fellow of the Medical Research Council of Canada, Gold acquired the concepts and technology for electromicroscopy, tissue culture, virology, and cell biology. He then returned to the Montreal General Hospital, its Research Institute, and McGill University as an assistant professor of medicine and a medical research council career investigator. In 1972, he was appointed professor of medicine. In 1977, he became Director of the Division of Clinical Immunology and Allergy at the Montreal General Hospital, and the following year he assumed the position as the first Director of the McGill Cancer Centre, now the Department of Oncology of McGill’s Faculty of Medicine. In 1980, Gold returned to the Montreal General Hospital as Physician-in-Chief and served as the Chairman of the Department of Medicine at McGill University for the statutory 5-year period between 1985 and 1990. In 1995, he became Executive Director of the Clinical Research Centre of the Research Institute of the McGill University Health Centre. The other positions he holds include Douglas G. Cameron Professor of Medicine and Professor, Departments of Physiology and Oncology, McGill University. In recognition of the scientific contributions made by Phil Gold, he has been the recipient of numerous international awards and honors and has been elected to a wide variety of scientific organizations. These include Companion of the Order of Canada, Officer of the L’Ordre Nationale du Québec, and recipient of the Gold Medal Award of Merit of the Graduate Society of McGill University. (Courtesy of Faisal Aziz.)

The preoperative CEA level in and of itself also has some prognostic significance. Patients with localized disease (as evaluated by clinical methods) have a higher recurrence rate when a high preoperative CEA level is noted than when the preoperative level is low.⁴²⁴ Such an elevation may be suggestive of inapparent spread of the tumor. Wiratkapun and coworkers noted that CEA levels about 15 ng/mL predicted an increased risk of metastatic recurrence in potentially curative colonic cancer.¹¹¹¹ This suggests undetected disseminated disease. Ashton and colleagues demonstrated a statistically significant association between survival and a high preoperative CEA level.³⁵ Others have demonstrated a relationship between the preoperative CEA level and the depth of invasion according to Dukes’ classification (see later).¹¹¹⁹ Tumor fixation has also been correlated with the level of CEA elevation.²⁶⁸ Sener and colleagues showed, through the use of a cancer registry system in a retrospective analysis, that the preoperative serum CEA level can be an indicator of survival that is independent of the stage of disease at diagnosis.⁹¹⁶

CEA is of limited value in the search for a primary site if metastatic carcinoma is noted. The antigen is detected
in about 50% of tumors of the breast, stomach, lung, and in other solid tumors. Levels higher than normal have also been found in heavy smokers and in persons with cirrhosis, pancreatitis, uremia, peptic ulcer, intestinal metaplasia of the stomach, as well as ulcerative colitis. The antigen has been reported in tissue of intestinal polyps, colonic inflammatory mucosa, and normal intestinal mucosa of children. CEA has also been found in cancerous tissue from the breast, liver, and lung, as well as in body fluids exposed to cancer. In colonic washings, high levels of CEA have been found in patients with colon cancer and colon polyps, intermediate levels in those with ulcerative colitis, and lower levels in normal subjects.¹⁰⁹⁹ Yeatman and colleagues demonstrated an elevated CEA in gallbladder bile in some individuals who showed no evidence of hepatic metastasis at the time of surgery for the primary tumor.¹¹³³ They suggested that those with slight increases should be followed closely for the possible subsequent appearance of such lesions.

Nuclear Medicine Studies

Positron Emission Tomography

Positron emission tomography (PET) scan is a functional imaging technique that uses short-lived radioisotopes attached to tracers to examine abnormal biochemical processes associated with disease. PET scans typically use fluorine-18- labelled deoxyglucose (FDG) to identify tissues with increased glucose metabolism and transport, such as cancer cells. PET scans are particularly useful in distinguishing indeterminate lesions identified on conventional imaging studies such as CT scan.⁹⁵² Such distinctions are especially valuable for excluding occult metastases in patients being considered for resection of metastatic disease or in the detection of disease recurrence.¹¹⁶

Liver Scan

The liver scan has been abandoned for the evaluation of this organ in patients with colorectal cancer, either preoperatively or postoperatively. The study has been supplanted by CT.

Pelvic Lymphoscintigraphy

Pelvic lymphoscintigraphy has been used to try to discriminate between normal and diseased large bowel and to determine the extent of nodal uptake, but it has no demonstrable value in the diagnosis or staging of colorectal cancer.⁸²⁹ Angiography of the mesenteric arteries likewise has not been proved useful in the preoperative evaluation of colon carcinoma.⁵⁰⁹

Radiolabeled Antibody Imaging

Anti-Carcinoembryonic Antigen

A modification of the CEA assay involves the use of radiolabeled antibodies to the antigen followed by an external photoscan.³⁴⁷,³⁴⁸ and ³⁴⁹ Goldenberg and coworkers evaluated 50 patients by injecting antibody against CEA labeled with ¹³¹I and performing total body scans.³⁴⁸ In this study, 83% of the preoperative group and approximately 90% of the postoperative group were found to have the tumors correctly localized. The authors conclude that among other potential benefits, this technique can help to stage the tumor preoperatively and complement other methods used to assess tumor response to therapy.

Beatty and colleagues studied 100 patients with known or suspected colorectal cancer by radioimmunoscintigraphy, using murine monoclonal anti-CEA antigen labeled with indium.⁶² Sensitivity was 76% for primary tumors, 44% for hepatic metastases, 38% for extrahepatic abdominal metastases, and 78% for extra-abdominal metastases.

Lechner and coworkers submitted 47 patients to radioimmunoscintigraphic investigations for primary or recurrent colorectal cancer by means of technetium-99m (^99mTc)-fab’ fragment (Immu 4).⁵⁵⁹ The advantage of this particular commercially available product is that it has a short half-life with high photon abundance. It can be administered at high doses and permits early imaging with a gamma camera. The overall accuracy of this study was 93.75% in primary and 91.60% in recurrent colorectal cancer. The investigators concluded that immunoscintigraphy had a decisive influence on treatment planning in one-third of patients with primary colorectal cancer and was superior to CT in the detection of early recurrences (see the later discussion under follow-up evaluation).⁵⁵⁹

Gastrointestinal Cancer Antigen

Gastrointestinal cancer antigen (CA 19-9) is a monoclonal antibody produced against human colorectal carcinoma cell line SW1116.⁴⁷⁶^,⁸²² However, this antigen is not specific for cancer and has been found in certain normal tissues, most notably pancreas, gallbladder, and gastric mucosa.⁴⁷⁶ There is no evidence to suggest that CA 19-9 is superior to CEA in predicting recurrence, although there may be some advantage in screening high-risk patients for colorectal cancer.⁸²²

Other Antibody Imaging Options

Yiu and colleagues used the same concept through the application of monoclonal antibodies that react with epithelial membrane antigen.¹¹³⁴ Indium-111 (¹¹¹In)-M8 detected 13 of 16 tumor sites, whereas In¹¹¹-77-1 detected 10 of 15 tumor sites. The implication again is that these monoclonal antibodies may have a role in the preoperative immunolocalization of colorectal cancers. Additionally, others have shown unsuspected tumor sites using a similar technique, a method that directly affected treatment in 18% of patients.²⁵⁰

Yamaguchi and colleagues used another tumor marker, NCC-ST 439, in the preoperative and postoperative evaluation of individuals with colorectal cancer.¹¹³¹ This is a monoclonal antibody obtained using ST-4, a cell line derived from poorly differentiated stomach cancer. The authors demonstrated results comparable to those obtained with the other markers, especially when it was used in combination with CEA.

Shibata and coworkers studied circulating anti-p53 antibodies in patients with known colorectal carcinoma to evaluate the clinical application of this technique.⁹³² Circulating anti-p53 antibodies were detected in 68% of their patients.⁹³² The authors concluded that this particular analysis may become an important diagnostic indicator of colorectal malignancies (see earlier discussion).

Carpelan-Holmström and associates investigated whether there are differences in serum levels of CA 242, a tumor marker demonstrated to have a high preoperative sensitivity for colorectal cancer and CEA.¹⁵⁷ The authors concluded that the two studies supplement each other, in that the two markers together show a higher sensitivity than either one alone.

van Kamp and coworkers assessed CA M43, a serum marker for colorectal cancer, to determine its clinical utility.¹⁰⁴⁹ This monoclonal antibody study identifies tumorassociated mucin. The investigators concluded that this
marker shows a positivity rate equivalent to that of CEA, and as was demonstrated with CA 242, it appears to complement CEA. Together, they reached 87% positivity in the presence of metastatic disease.¹⁰⁴⁹

Vasoactive Intestinal Peptide Receptor Imaging

Vasoactive intestinal peptide (VIP) is a major regulator of water and electrolyte secretion in the gut. Various endocrine tumors, as well as intestinal adenocarcinomas, express large numbers of high-affinity receptors for VIP. Virgolini and associates evaluated the usefulness of scanning with VIPlabeled ¹²³I to localize gastrointestinal tumors.¹⁰⁶¹ Among those with colorectal cancer, primary recurring tumors were visualized in all 10 patients, liver metastases were seen in 15 of 18, and lung metastases were detected in 2 of 3. All four patients with lymph node metastases were identified with this technique. It appears that scanning with radiolabeled VIP permits visualization of intestinal tumors and metastases, provided, of course, that they express receptors for VIP.

Comment

The advent of monoclonal antibody technology has permitted the development of a number of radiolabeled tumor-reactive probes. These can be used in conjunction with gamma camera imaging equipment to identify the anatomic distribution of colorectal cancer within an individual. The reader is referred to the discussion on the postoperative application of these techniques in the section on follow-up evaluation. The real excitement will occur when it will be possible to attach a therapeutic agent to the monoclonal antibody to treat the primary or metastatic cancer.

Serum Gastrin

Kameyama and colleagues evaluated the relationship between serum gastrin levels and liver metastases in colorectal cancer to determine whether serum gastrin can be used as a predictor of liver metastases, independent of other prognostic variables.⁵⁰⁵ In a series of 140 patients who underwent surgery for colorectal cancer, the fasting serum gastrin level was determined preoperatively. The incidence of liver metastases was statistically significantly higher in those patients with a serum gastrin level of 150 pg/mL or greater than in those with a serum gastrin level less than this number.⁵⁰⁵ These results suggest that serum gastrin serves as a useful predictor of liver metastases that correlates well with the pathologic determination of venous invasion.

Leukocyte Adherence Inhibition

The leukocyte adherence inhibition assay is an in vitro test based on the observation that, following incubation with tumor extracts from the same organ, leukocytes from cancer patients lose their ability to adhere to glass surfaces.⁴⁴ Reports have demonstrated relatively consistent identification of malignant processes.³⁸⁴^,¹⁰⁰³^,¹⁰¹⁶ Theoretically, it may be possible to use this phenomenon to improve the detection rate of colorectal cancer.⁵⁸³ However, the lack of specificity for localization of the growth and inconsistencies in the performance of the technique relegate the procedure at this time to the status of a research tool.

▶ PATHOLOGY

By far the most common malignant lesion affecting the colon and rectum is adenocarcinoma. The tumor arises from the glandular epithelium, and it can invade microscopic blood vessels as well as metastasize to distant organs, most commonly the liver. It can spread by way of the lymphatics to regional lymph nodes and ultimately pass into the systemic circulation. The tumor may also extend locally into adjacent organs (e.g., posterior vaginal wall, uterus, bladder, small bowel, stomach, and retroperitoneal structures).

Microscopic Appearance

Histologically, the cancer may appear well differentiated (Figure 23-19), moderately differentiated (Figure 23-20), or poorly differentiated (Figure 23-21). The tumor may produce so much mucin that the nucleus is pushed to one side of the cell, creating a signet ring appearance (Figure 23-22). This last type has been the subject of some debate with respect to its prognostic implications. Minsky has observed that the incidence of this manifestation in patients with colorectal cancer is approximately 17%.⁶⁸⁶ He found that colloid carcinoma was not an independent prognostic factor for survival, but believed that it should be reported separately from other histologic patterns in order to achieve a better understanding of its natural history. Generally, the more poorly differentiated tumors are more invasive at the time of diagnosis, and the more invasive the tumor, the poorer the prognosis.

Macroscopic Appearance

In addition to degree of differentiation, tumor morphology— whether polypoid, infiltrative, or ulcerated—has been found to be an important prognostic variable. It, too, should be reported as part of a comprehensive pathologic evaluation of the patient.⁹⁷² Macroscopically, the tumor can display a number of forms (Figures 23-23,23-24,23-25,23-26,23-27,23-28,23-29,23-30,23-31 and 23-32).

Factors Affecting Rates of Growth and Spread of Tumor

Carcinomas of the colon and rectum are relatively slowgrowing tumors. Symptoms usually appear early in the development of the disease, and metastases occur relatively late. Tumor growth and spread display considerable variation, depending partly on histologic grade (based on cellular

arrangement and differentiation), increased ameboid action of some cancer cells, enzymes such as hyaluronidase, decreased adhesiveness of the tumor cells, size of the lesion at the primary site, and length of time the tumor has been present.³⁷⁹^,⁹⁹⁰ Daneker and colleagues showed the interaction of the tumor with the basement membrane to be an important factor in predicting spread.²²² They noted that in general, more poorly differentiated cancers tend to adhere and therefore to invade much more readily. Additional variables include location of the tumor, indeterminate host
factors, manipulation at surgery, and the age and sex of the person.⁴⁰^,¹⁸⁹^,²⁷⁰^,⁴⁴⁰^,⁸⁸⁹

FIGURE 23-19. Well-differentiated adenocarcinoma. The neoplastic glands display somewhat oriented epithelium and resemble crypts in their overall architecture. (Original magnification × 250; courtesy of Rudolf Garret, MD.)

FIGURE 23-20. Moderately differentiated adenocarcinoma. The glands are more irregular and exhibit less orientation of the epithelium. (Original magnification × 250; courtesy of Rudolf Garret, MD.)

FIGURE 23-21. Poorly differentiated (undifferentiated) adenocarcinoma. There is no definite formation of glands. (Original magnification × 250; courtesy of Rudolf Garret, MD.)

FIGURE 23-22. Signet ring carcinoma. In this malignant variant, mucin displaces the nucleus to one side. This is occasionally seen in lesions of the right side and when carcinoma arises in ulcerative colitis. (Original magnification × 600; from Corman ML, Veidenheimer MC, Swinton NW. Diseases of the Anus, Rectum, and Colon. Part I: Neoplasms. New York, NY: Medcom; 1972, with permission.)

FIGURE 23-23. Relatively small polypoid carcinoma. (From Corman ML, Veidenheimer MC, Swinton NW. Diseases of the Anus, Rectum, and Colon. Part I: Neoplasms. New York, NY: Medcom; 1972, with permission.)

FIGURE 23-24. Ulcerating carcinoma. (From Corman ML, Veidenheimer MC, Swinton NW. Diseases of the Anus, Rectum, and Colon. Part I: Neoplasms. New York, NY: Medcom; 1972, with permission.)

FIGURE 23-25. Polypoid carcinoma with ulceration. (From Corman ML, Veidenheimer MC, Swinton NW. Diseases of the Anus, Rectum, and Colon. Part I: Neoplasms. New York, NY: Medcom; 1972, with permission.)

FIGURE 23-26. Polypoid carcinoma with mucosal hyperplasia (arrow). This association is commonly seen. (From Corman ML, Veidenheimer MC, Swinton NW. Diseases of the Anus, Rectum, and Colon. Part I: Neoplasms. New York, NY: Medcom; 1972, with permission.)

FIGURE 23-27. Large polypoid carcinoma. Despite its size, the prognosis is better with this tumor than with a smaller, invasive lesion. (From Corman ML, Veidenheimer MC, Swinton NW. Diseases of the Anus, Rectum, and Colon. Part I: Neoplasms. New York, NY: Medcom; 1972, with permission.)

FIGURE 23-28. Scirrhous or infiltrating carcinoma is associated with a poorer prognosis. (From Corman ML, Veidenheimer MC, Swinton NW. Diseases of the Anus, Rectum, and Colon. Part I: Neoplasms. New York, NY: Medcom; 1972, with permission.)

Histochemical studies assess those factors that contribute to tumor growth and spread. These include changes in sialomucin in the distal resection margin and the presence of human chorionic gonadotropin.²²⁹^,⁹³⁸ For example, Dawson and colleagues showed that sialomucin adjacent to a primary colorectal cancer provides a crude assessment of tumor invasiveness and therefore the risk for local recurrence.²²⁸ Ideally, as more information is developed about prognostic factors, more meaningful recommendations with respect to supplementary therapy will be made.

FIGURE 23-29. Carcinoma (light area, center) arising in villous adenoma. (From Corman ML, Veidenheimer MC, Swinton NW. Diseases of the Anus, Rectum, and Colon. Part I: Neoplasms. New York, NY: Medcom; 1972, with permission.)

FIGURE 23-30. Colloid carcinoma. Prognosis has generally been believed to be poorer with this variant of carcinoma. Note the gelatinous appearance. (From Corman ML, Veidenheimer MC, Swinton NW. Diseases of the Anus, Rectum, and Colon. Part I: Neoplasms. New York, NY: Medcom; 1972, with permission.)

The first opportunity to measure the growth of colonic cancer at its site of origin was reported in 1961 by Spratt and Ackerman.⁹⁵⁵ In this study, nine air-contrast enemas were performed during a period of 7.5 years.⁹⁵⁴ Radiographic measurements of the tumor were taken, and by appropriate plotting on a graph, the growth curve was shown to conform to an exponential increase in volume. The tumor was calculated to have a doubling time of 636.5 days. When desquamation from the surface was taken into account, it was postulated that a net increase of only one cell per 1,000 cells per day was adequate to account for the observed rate of growth. Another study demonstrated the mean doubling time of tumor volume to be only 130 days.¹⁰³ The authors believed that this high rate of growth could be attributed to the large
size of the tumors and therefore their greater likelihood of being malignant at the initial examination.

FIGURE 23-31. Perforated cecal carcinoma.

FIGURE 23-32. A polypoid adenoma with carcinoma at the tip. An ulcer is shown (arrow). (Courtesy of Rudolph Garret, MD.)

Doubling time of pulmonary metastases from colon and rectal carcinomas has been calculated radiographically and found to be 109 days.⁹⁵³ Metastatic tumors increase their cellular complement six times faster than primary cancers.⁹⁵⁴ It is theorized that the absence of desquamation in the metastatic site accounts for this observed difference.

Finlay and colleagues studied the growth rate of hepatic metastases by means of serial CT.²⁹⁹ They discovered that the mean doubling time for obvious metastases (those discovered at the time of laparotomy) was 155 days, compared with 86 days for occult metastases. By means of extrapolation, the authors concluded that the mean age of the former lesions was 3.7 years, and the latter was 2.3 years.

The metastatic behavior of neoplasms varies. According to Spratt, examination of the cancer-host interface is helpful in determining the behavioral pattern of the tumor.⁹⁵³ One that is less likely to metastasize exhibits a well- circumscribed, intact margin. The tumor that is more likely to metastasize to lymph nodes has a loose or infiltrating margin with little inflammatory reaction. Sacchi and coworkers looked at the border between normal tissue and tumor in colorectal cancer and opined that it is the mast cell, probably influenced by the inflammatory infiltrate and/or colorectal cancer cells themselves, which destroy lymphatic vessels, thereby preventing spread into the lymphatic system.⁸⁷⁴

CUTHBERT ESQUIRE DUKES (1890-1977)

Cuthbert Dukes was born in Bridgwater, Somerset, England and graduated in medicine from Edinburgh University in 1914. He served in World War I in the Royal Army Medical Corps, receiving the Order of the British Empire. After the war, he became a demonstrator in bacteriology at University College, London, and he joined the staff of St. Mark’s Hospital in 1922 as its first pathologist. It was there that he produced his classification of cancer of the rectum that became invaluable as a guide to prognosis. His many writings include the pioneering work on familial polyposis (and the Polyposis Registry at St. Mark’s Hospital) as well as several books. He became successively presidents of the Royal Society of Medicine (1944), the section of urology (1957), and the section of the history of medicine (1959), and was elected to honorary fellowship in the Royal College of Surgeons and the American Society of Colon and Rectal Surgeons.

Some investigators have suggested that the concentration of circulating C-reactive protein may play a role in predicting recurrence and survival in colorectal cancer. McMillan and colleagues studied the presence of a systemic inflammatory response as measured by circulating C-reactive protein in 174 patients who were believed to have undergone a curative resection.⁶⁶⁶ C-reactive protein concentration was found to be an independent predictor for survival—that is, the presence of elevated concentrations predicts a poor outcome.

Delay in Diagnosis: Legal Implication

The one article by Spratt and Ackerman has been responsible for more confusion, misinterpretation, and misrepresentation in the sphere of medical litigation with respect to the accusation of delay in diagnosis than all other variables combined. A question asked by attorneys in medicolegal actions is when an earlier diagnosis will make a difference with respect to prognosis (see Chapter 34). Clearly, the issue of tumor doubling time does not provide the answer. In an unofficial poll of the Chicago Society of Colon and Rectal Surgeons (1997), all but 2 of the more than 50 surgeons who were in attendance responded “6 months.” For the record, the concept of cancer doubling time is without merit in considering the likelihood of survival and its applicability to earlier diagnosis and prognosis in the individual patient.

Staging and Prognosis—History and Current Status

Classifications

The importance of tumor invasion and its prognostic implications was postulated by Dukes in 1930 and was subsequently revised by him in 1932.²⁶²^,²⁶³ This has come to be known as Dukes’ classification (not Duke’s). The classification was originally directed toward rectal cancer (Table 23-1). After performing numerous meticulous dissections of resected specimens to identify metastases to lymph nodes, Dukes
further modified his classification in 1944.²⁶⁴ Those tumors with lymph node involvement but with a negative node at the ligation of the inferior mesenteric artery he called C₁. Lesions classified as C₂ had metastases to the node at the level of the ligature (Table 23-2). The addition of the D category has been generally accepted as representative of tumor spread beyond the reach of potential surgical cure but was not included in his classification (Table 23-3).

TABLE 23-1 Dukes Classification of Rectal Cancer (1932)

STAGE	CLASSIFICATION
A	Carcinoma limited to wall of rectum
B	Carcinoma spread by direct continuity to extrarectal tissues; no lymph node metastases
C	Metastases present in regional lymph nodes

Others have introduced their own classifications, expanding and subdividing Dukes’ system and broadening it to include colon cancer and disseminated metastases; degree of differentiation; tumor morphology; and histogram pattern, among others.³⁶^,²⁹⁵^,⁵²³^,⁷⁸⁵^,¹⁰³⁶^,¹¹⁴⁰ Jass and colleagues suggested a classification based on prognosis⁴⁸⁵:

Excellent prognosis
Good
Fair
Poor

A scoring system was developed based on several factors that appear to influence survival—number of lymph nodes with metastatic tumor, character of the invasive margin, presence of peritumoral lymphocytic infiltration, and local spread. Most observers believe that the classification of Dukes is of greater prognostic value and more reproducible than that of Jass.²³³

Figure 23-33 illustrates other staging systems. In theory, the more one “substages,” the greater the refinement potential to predict tumor behavior, and therefore to gain more information about the possible outcome.⁷³⁸ There is even a Japanese classification that subdivides vertical invasion of the submucosa into Sm1 (upper third of the submucosa), Sm2 (middle third), and Sm3 (invasion near the “inner surface of the muscularis propria”). The question, however, is whether this is really important, especially in the absence of adjuvant therapy specific to disease-invasion subsets. The TNM system has become the most popular classification in use in the United States and indeed in the world (Tumor-Node-Metastasis). The American Joint Committee on Cancer (AJCC) seventh edition staging system is shown in Table 23-4.

TABLE 23-2 Dukes’ Classification of Rectal Cancer (1944)

STAGE	CLASSIFICATION
A	Carcinoma confined to wall of rectum
B	Carcinoma spread by direct continuity to perirectal tissue; no lymph node metastasis
C₁	Metastasis present in nodes but not to ligature
C₂	Metastasis present in nodes to level of ligature

TABLE 23-3 Modified Dukes’ Clinicopathologic

STAGE	CLASSIFICATION
A	Carcinoma confined to wall of bowel
B	Carcinoma spread by direct continuity to perirectal or pericolonic tissue; no lymph node metastasis
C	Metastasis present in regional lymph node
D	Omental implant; peritoneal seeding; metastasis beyond the confines of surgical resection

Fisher and colleagues compared the relative prognostic value of the Dukes, Astler-Coller, and TNM staging systems in 745 pathologically evaluable individuals with rectal cancer.³⁰¹ They concluded that the Dukes method was the simplest and most consistent algorithm related to prognosis. They observed, however, that the Astler-Coller C1 and C2 designations were a uniquely valuable contribution to prognostic discrimination. The same is probably true of the TNM classification.

Goligher’s Commentary

John Cedric Goligher (see Biography, Chapter 21) wrote that the various classifications proposed can create only confusion and mislead the reader interpreting the surgical results from one institution to another. He observed,

It is the unalienable right of every pathologist and surgeon to evolve his own system for categorizing the extent of spread of cancers of the colon and rectum … [but] I suggest that it would be more useful to restrain his urge to classify and accept Dukes’ categorization exactly as it was defined by him.³⁵¹

Despite this suggestion, since 1991 the Council of the American Society of Colon and Rectal Surgeons endorsed the TNM staging classification. More’s the pity, for I believe that the clinicopathologic classification of Dukes fulfills all reasonable criteria for prognostication. Other studies comparing the various staging classifications concur with this conclusion.¹⁶⁶^,²³³^,³⁰⁰^,³⁰¹^,⁵⁸¹^,⁷²⁸ Sadly, the plea for romance and historical precedent has inevitably fallen on deaf ears—the TNM classification has become the benchmark that we all must accept and adopt.

Lymphatic Invasion

The importance of lymph node involvement by tumor has been well established through providing important prognostic
information.³³¹^,³³⁵^,³⁶⁸^,³⁸² It seems reasonable, therefore, that if one were to use special clearing techniques, a greater number of lymph nodes will be identified, thereby improving the ability to prognosticate.³⁸²^,⁴⁶⁸^,⁷⁹⁶ Koren and colleagues proposed the use of a lymph node-revealing solution composed of various traditional fixatives and fatty solvents to clear the mesentery and identify more lymph nodes.⁵³⁵ In 30 problematic cases, in which an unsatisfactory number of lymph nodes were found by the traditional method, this approach resulted in upstaging a number of tumors from N0 to N1. The number of positive nodes has been shown in one study to be a more accurate predictor of survival than depth of tumor penetration.¹¹¹⁸ Very few pathology laboratories employ a clearing technique, but if all pathologists vigorously pursued lymph node identification, a greater consistency in reporting the results of treatment would be achieved.

FIGURE 23-33. Comparison of staging classifications for colorectal carcinoma.

Tsakraklides and colleagues evaluated the histologic morphology of lymph nodes in an attempt to improve prognostic ability.¹⁰³⁵ They found a higher rate of survival (which was not statistically significant) in patients whose nodes showed germinal center predominance compared with those whose nodes showed lymphocyte predominance, the “unstimulated pattern.”

Cutait and coworkers demonstrated that by using immunoperoxidase staining of CEA and cytokeratins of lymph nodes previously considered free of disease, restaging of 22 nodes became necessary.²¹⁷ However, follow-up at 5 years failed to show any statistically significant difference in survival.

Numerous studies have appeared in the literature in recent years attesting to the impressive survival advantage for patients with more lymph nodes identified in their surgical specimens.¹⁷¹^,⁴⁹⁹^,⁵⁷⁴ The implication is that low node counts are markers for inadequate surgical technique and/or suboptimal pathologic examination, often falsely “downstaging” patients who harbor undetected lymph node metastases. The need to find 12 or more nodes has even been specified as a quality indicator for colon cancer surgery. However, it should be clear that this is a marked oversimplification of the relationship between lymph node counts and prognosis because studies using node clearing techniques to identify dozens of additional nodes have shown only a minimal frequency of upstaging.¹²³^,⁹⁰⁶ Indeed, even when surgical and pathologic factors are controlled for, the number of lymph nodes identified in the specimen continues to be strongly associated with improved prognosis.⁷⁰⁵ It may be that a greater number of nodes is a surrogate for a more immunogenic tumor or a more immunocompetent host.

Blood Vessel Invasion

Compared with lymph node involvement, the importance of blood vessel invasion (Figure 23-34) has been emphasized to a much lesser extent.¹¹⁵^,¹²²^,³⁰² However, the prognostic implication of blood vessel invasion has been well established,²⁰³^,⁹³⁶ a fact that has stimulated different approaches to the technique of surgical removal of the tumor (see later).⁵³^,³⁰²^,⁸⁸²^,¹⁰³⁶ It should be axiomatic that the pathologist seek to identify and report invasion of blood vessels with the same concern applied to identifying involvement of lymph nodes.

Neural Invasion

As with blood vessel invasion, neural invasion has prognostic import—that is, the presence of perineural invasion implies a more ominous prognosis than if such invasion were not present.

DNA Content or DNA Ploidy

Some studies have suggested that tumor DNA content, as determined by flow cytometry, can provide valuable information about the biologic behavior of neoplastic cells and is therefore an important prognostic factor in determining survival.²⁷^,⁵¹^,⁵³¹^,⁹⁰⁸^,¹⁰⁹³ Aneuploid (nondiploid) tumors contain a population of cells that exhibit a DNA content distinctly different from the DNA noted in “normal” (diploid) malignant cells. For example, Scott and colleagues reported that DNA nondiploid rectal carcinomas were associated with a statistically significant increased incidence of vascular invasion, tumor fibrosis, and advanced Dukes’ stage.⁹⁰⁷ In another study from the same unit, these cancers also were associated with a significantly poorer prognosis in patients with unresectable disease.⁹⁰⁸ Kokal and colleagues demonstrated that in comparison with diploid tumors, tumors with abnormal DNA content tended to be less well differentiated, to invade the serosa or extend beyond, and to have lymph node metastases.⁵³² It should be remembered, however, that tumor cell DNA content, although it can be an independent prognostic factor, is not as accurate as Dukes’ classification in this respect.¹⁰⁸^,⁴³³^,⁹⁰² Deans and associates performed flow cytometry on 312 patients with adenocarcinoma of the colon and rectum and found by univariate survival analysis that no flow cytometric variable was statistically significantly related to survival.²³⁴ They concluded that Dukes’ stage, patient age, and tumor differentiation are the variables most closely related to survival, and that conventional histologic variables remain the best predictors of prognosis for this condition.²³⁴ Others believe, however, that tumor DNA content
is the single most important prognostic factor among all the clinical and pathologic variables available.²³²^,⁵³²^,⁵³³ Although there may be some controversy in this respect, there is no disagreement that nondiploid lesions are generally associated with a greater level of invasion and therefore a more advanced Dukes’ stage.

TABLE 23-4 Tumor Staging with Tumor-Node-Metastasis (TNM) System

STAGE	DESCRIPTION
Tumor (T)
TX	Primary tumor cannot be assessed
T0	No evidence of primary tumor
Tis	Carcinoma in situ: intraepithelial or invasion of lamina propria
T1	Tumor invades submucosa
T2	Tumor invades muscularis propria
T3	Tumor invades through the muscularis propria into pericolorectal tissues
T4a	Tumor penetrates to the surface of the visceral peritoneum
T4b	Tumor directly invades or is adherent to other organs or structures
Regional Lymph Nodes (N)
NX	Regional lymph nodes cannot be assessed
N0	No regional lymph node metastasis
N1	Metastasis in one to three regional lymph nodes
N1a	Metastasis in one regional lymph node
N1b	Metastasis in two to three regional lymph nodes
N1c	Tumor deposit(s) in the subserosa, mesentery, or nonperitonealized pericolic or perirectal tissues without regional nodal metastasis
N2	Metastasis in four or more regional lymph nodes
N2a	Metastasis in four to six regional lymph nodes
N2b	Metastasis in seven or more regional lymph nodes
Distant Metastasis (M)
M0	No distant metastasis
M1	Distant metastasis
M1a	Metastasis confined to one organ or site (e.g., liver, lung, ovary, nonregional node)
M1b	Metastases in more than one organ/site or the peritoneum

Oncogene Analysis

Rowley and colleagues compared DNA ploidy and nuclearexpressed p62 c-myc oncogene in the prognosis of colorectal cancer.⁸⁶³ As discussed earlier, oncogenes exist in the normal human genome as proto-oncogenes but may be activated by various means in tumors, including overexpression and mutation (see also Chapter 22).⁸⁶³ The authors could not suggest a replacement for Dukes’ staging system for prognostication but observed that the combination of ploidy status and oncogene expression predicted survival better than ploidy alone.

Nucleolar Organizer Regions

Nucleolar organizer regions are loops of ribosomal DNA in nuclei that direct ribosome and protein formation.⁷⁰⁸ Moran and colleagues studied the prognostic implication of these regions, as well as ploidy, in individuals with advanced colorectal cancer.⁷⁰⁸ They concluded that nucleolar organizer regions are the most important individual variable for predicting survival, whereas ploidy values are equivalent to histologic differentiation.

Nuclear Morphology

Mitmaker and coworkers studied nuclear shape as a possible factor in determining prognosis.⁶⁹³ The nuclear shape factor was defined as the degree of circularity of the nucleus, with a perfect circle recorded as 1.0. A shape value greater than 0.84 was associated with a poor outcome and indeed was the most significant predictor of survival even when corrections were made for sex, age, histologic grade, and Dukes’ classification.

Doppler Perfusion Index

Leen and associates assessed the relative value of Dukes’ staging and the Doppler perfusion index (DPI) as prognostic indices in individuals who underwent apparently curative surgery for colorectal cancer.⁵⁶¹ This index, the ratio of hepatic arterial to total liver blood flow, was measured before resection by means of duplex/color Doppler sonography. The authors observed that the DPI identified two groups of individuals—78% of those with an abnormally elevated DPI value had recurrent disease or died, whereas 97% of those with a normal DPI value survived. The implications concerning the suitability for adjuvant therapy following apparent curative resection are clear.

Tumor Budding

Tumor “budding” refers to the presence of microscopic clusters of undifferentiated cancer cells ahead of the invasive front of the lesion. This has been believed to have prognostic significance with respect to cure rates following resection for colon and rectal cancer. Hase and colleagues analyzed all surgical specimens for this phenomenon in 663 patients who underwent curative resection.⁴²⁷ The presence of severe budding was associated with a poorer prognosis and paralleled a ‘more invasive Dukes’ classification.

FIGURE 23-34. Blood vessel invasion by colonic adenocarcinoma. Note the extensive growth of tumor inside the vein that has thin, black elastica in its wall and fibrous thickening of the intima in the artery (left). (Original magnification × 80; from Corman ML, Swinton NW Sr, O’Keefe DD, et al. Colorectal carcinoma at the Lahey Clinic, 1962-1966. Am J Surg. 1973;125:424.)

Miscellaneous

Lindmark and coworkers, in an attempt to identify better prognostic predictors than Dukes’ classification, studied 212 patients by means of clinical, histopathologic, cellular, and serologic tumor characteristics.⁵⁸¹ Dukes’ stage was the most powerful variable. Others found significant survival associations with erythrocyte sedimentation rate, leukocyte blood count, alkaline phosphatase level, aspartate aminotransferase level, number of small blood vessels, age of the patient, and six different serum tumor markers.

Proliferating Cell Nuclear Antigen Expression

Cellular proliferative activity has been shown to be a useful indicator of biologic aggressiveness in colorectal carcinoma. Choi and associates investigated the correlation between proliferative activity and malignancy potential in colorectal cancers to determine whether the proliferative index of cancer cells has prognostic significance.¹⁷⁶ Using immunohistochemical methods involving a monoclonal antibody to proliferating cell nuclear antigen (PCNA), they obtained 86 pathologic sections and compared PCNA with conventional clinicopathologic factors as well as other prognostic parameters. The authors determined that PCNA at the invasive tumor margin was a valuable predictor for determining those individuals with a higher potential for metastasis or recurrence.¹⁷⁶

Anatomic Distribution

Numerous reports have demonstrated a change in the distribution of colorectal carcinoma.¹^,¹⁴⁸^,²¹¹^,³⁵⁶^,³⁷⁴^,⁷¹⁰^,⁹⁴³ It had long been thought that 75% of colorectal cancers were within the range of the rigid proctosigmoidoscope. More recent data suggest that the figure should be closer to 60%.

Cady and associates reviewed almost 6,000 resected specimens of the large bowel that had been removed during a 40-year period (1928 to 1967).¹⁴⁸ During this time, the incidence of cancer of the right side of the colon increased from 7% to 22%, and the incidence of rectosigmoid, sigmoid, and rectal carcinomas fell from 80% to 62%. These changes represented statistically significant differences. A trend to smaller tumors was also evident, with less frequent lymph node involvement of the distal lesions, possibly reflecting an improvement in early detection. Beart and colleagues reviewed the Mayo Clinic experience by studying the new cases diagnosed among residents of Rochester, Minnesota, between 1940 and 1979.⁶⁰ They found an increase in the incidence of proximal lesions (from 15.1/100,000 person-years in 1940 to 1959 to 17.3/100,000 in 1960 to 1979). This coincided with a fall in distal lesions (from 35.5/100,000 person-years to 28.2/100,000 person-years). In a 2002 report from the Netherlands, the incidence of colorectal cancer has almost doubled from 1981 to 1996, whereas the proportion of proximal cancers increased from 25% to 37%.⁶⁷⁸ Gonzalez and colleagues found that increased age, female gender, Black non-Hispanic race, and the presence of comorbid illnesses were factors associated with a greater likelihood of developing colorectal cancer in a proximal location.³⁵⁶ In a Veterans Administration study, black race was likewise shown to be associated with a higher incidence of right-sided tumors than that of whites.²¹¹

A possible explanation for the observed increase in the proportion of more proximal colon cancers is the “search and destroy” concept for the management of colorectal polyps. By clearing the rectum through electrocoagulation, biopsy excision, and snare excision, one essentially prevents the subsequent development of a distal malignancy. Another obvious implication is the limitation of proctoscopy or flexible sigmoidoscopy as a screening tool.

▶ SURGICAL TREATMENT Historical Perspective

Resection of the bowel with intestinal anastomosis is a relatively recent operation. In 1818, Zang declared that “every
intestinal suture is a mighty procedure in a highly vulnerable organ, and therefore a dangerous, yes, a very dangerous undertaking.”³¹⁴ Exteriorization was the method of treatment for intestinal disease and injury, a technique that had remained essentially unchanged for more than two millennia.

Reybard of Lyons in 1833 was the first to perform a successful resection of the sigmoid colon; for this he was criticized by the Paris Academy of Medicine.²²¹^,⁸⁴³ Colostomy was advocated as a palliative procedure as early as 1839. Before this time, an occasional obstructing carcinoma of the colon was relieved by the spontaneous formation of a fecal fistula, when it was relieved at all. Before 1889, the mortality rate for colonic resection was 60%, but this figure had been reduced to 37% by 1900.⁷⁰⁹ Because of the high mortality for intra-abdominal resection and anastomosis, the staged extraperitoneal operations of Paul,⁷⁸⁴ von Mikulicz,⁶⁸³ and Bloch⁹⁷ (exteriorization-resection) were usually employed.

JEAN-FRANÇOIS REYBARD (1790-1863)

Reybard was born in Cloisia (Jura) in France near the Swiss border. His parents were farmers. His father was also the mayor of the community and was therefore able to send his son to college at Nantes and then to Lyons for the completion of his studies. Reybard developed an interest in medicine while at the university and became an assistant at the Hôtel-Dieu in 1813 when he was 18 years old. While at the Charity Hospital, he had the opportunity of seeing the horrible consequences of war injuries, the result of Napoleon’s campaigns. When the city of Lyons was occupied by the Austrians, Reybard treated the wounds of both sides. In 1816, he was appointed to the Faculty of Medicine at Lyons. In 1827, he reported the results of his experiments on intestinal suturing. In May 1833, he successfully performed a resection of the sigmoid colon with a primary anastomosis on a 28-year-old man, although the tumor recurred 6 months later. Because of skepticism by the academy, he offered to repeat the procedure on dogs, but all seven died. While performing an operation for anal fistula, he stuck himself, developed septicemia, and succumbed 5 days later.

FRANK THOMAS PAUL (1851-1941)

Paul was born at Pentney, Norfolk, England and was educated at the Yarmouth grammar school. In 1869, he entered Guy’s Hospital and qualified as a member of the Royal College of Surgeons in 1873. After a year as resident house surgeon, he left London for Liverpool, where he practiced for the rest of his professional life. In 1875, he was made the first resident medical officer at the Liverpool Royal Infirmary. He became successively pathologist, lecturer in dental surgery, professor of medical jurisprudence, and surgeon to the Royal Infirmary. His 1891 article on exteriorization-resection, using “Paul’s tube,” antedated the modification described by von Mikulicz by a dozen years. Although he wrote little else, he was considered a masterful technician. Lord Moynihan of Leeds often visited his theater, observing that Paul was “the neatest operator he had ever seen.” Another visitor stated, “Paul, operating in the heyday of his manual efficiency, always made me think that he did with his hands what Pavlova did with her feet, only Paul’s work was much more useful.” He retired to grow orchids and died in his 90th year. (Photo from Paul FT. Frank T. Paul—his life. Br J Surg. 1951;39:195, with permission.)

JOHANN VON MIKULICZ-RADECKI (1850-1905)

Mikulicz was born in Cernowicz in the Balkan Peninsula, then a part of the Austro-Hungarian Empire. He was educated at Hermannstadt and earned his way through the university by giving piano lessons and playing the organ. He achieved the MD degree in Vienna in 1875 and became an assistant to Billroth. At the age of 32, he was appointed director and professor of surgery at Kraków, Poland. While there, he performed the first esophageal resection and introduced lateral pharyngotomy for excision of malignant tumors of the tonsillar region. In 1887, Mikulicz became director of the clinic and professor of surgery at Konigsburg. At the age of 40 (1890), he moved to Breslau as professor of surgery, a position he held until his death. One of the great innovative minds of 19th-century surgery, Mikulicz’s name has been used eponymously for various conditions and procedures, including Mikulicz’s cells, disease, drain, line, mask, ointment, and at least 10 operations (e.g., exteriorization-resection of the colon).

The initial attempts to coapt ends of intestine involved insertion of various objects into the lumina of the cut ends of the bowel, with or without the addition of sutures. These procedures involved the use of a reed pipe, goose trachea, cardboard smeared with sweet oil, a cylinder
of fish glue, a wax ring, and a silver ring, to name a few.³¹⁴^,⁵⁴⁵ Balfour suggested the use of a tube stent.⁴⁹ Probably the most famous internal stent was the Murphy “button” (Figure 23-35).⁷²⁰ Introduced in 1892, it quickly became the primary method of anastomosing bowel.

Until the 19th century, most surgeons regarded wounds of the intestine as a noli me tangere (do not touch), believing that nature would be more successful in the healing process than any artificial attempt to effect closure.⁹¹⁸ Travers (1812) is believed to be the pioneer in the use of sutures to perform intestinal anastomoses.¹⁰³¹ Lembert (1826) is eponymously associated with a type of intestinal suturing that produced a serosa-to-serosa apposition.⁵⁶⁶ As he was with so many other operative innovations, Billroth was also one of the leaders of intestinal anastomotic surgery. In 1887, Halsted reported an experimental study demonstrating the submucosa to be the primary layer responsible for a safe and secure anastomosis and emphasized the importance of inversion of the intestinal suture line.⁴⁰⁶ This was confirmed in a subsequent report.⁴⁰⁷ Later, Connell described his continuous inverting suture.¹⁹⁴ Allis simplified the technique of suturing intestine when he introduced his tenaculum forceps in 1901.¹⁷ For an excellent, comprehensive review of the history and evolution of the diverse anastomotic techniques developed prior to the 20th century, one should read Senn’s classic article on the subject.⁹¹⁸ A wonderful, contemporary perspective can be gleaned from the article by Steichen and Ravitch.⁹⁶⁸

DONALD CHURCH BALFOUR (1882-1963)

Donald Balfour was born in Toronto, Canada and obtained his undergraduate education at the Hamilton Collegiate Institute. He graduated from the University of Toronto Medical School in 1906. In 1907, Balfour became an assistant in pathology at the Mayo Clinic in Rochester, Minnesota and was appointed successively a junior surgeon and, in 1912, head of a section of general surgery. In 1937, he became director of the department of surgery. In 1910, he married Carrie Mayo, daughter of William J. Mayo. Balfour wrote on varied subjects, including tonsillectomy and thyroid and biliary procedures, but he gradually focused his research and writings on gastrointestinal surgery. He devised a number of instruments—an abdominal retractor, an operating table, and an operating room mirror. Among his many offices and honorary fellowships, he was one of the founders of the World Medical Organization and a charter member of the World Health Organization.

JOHN B. MURPHY (1857-1916)

John Murphy was born in Appleton, Wisconsin, the son of Irish immigrant pioneers. He entered the Rush Medical College in Chicago and graduated in 1879, winning the “first place” position as intern at the Cook County Hospital. In 1882, following 2 years of medical practice, he traveled for 18 months throughout Europe and came under the direction of Billroth in Vienna. On his return to practice in Chicago, he became known as a bold, brilliant surgeon and teacher, an indefatigable worker, and an innovator. Although he was often the subject of controversy (“the stormy petrel of surgery”), he was always greatly respected. In succession, he became Chief Surgeon at Chicago’s Mercy Hospital; Professor of Surgery at the Northwestern University Medical School; Chief of the Editorial Staff of the Journal of Surgery, Gynecology and Obstetrics; and President of the American Medical Association. He was a founder and regent of the American College of Surgeons and a recipient of numerous honorary titles. His writings were diverse and numerous and included treatises on the following: gunshot wounds of the abdomen; appendicitis; ileus; and vascular, pulmonary, neurologic, and orthopedic surgery. His interests in experimental surgery inspired the anastomotic button. Although controversial even in his day, it is still appreciated and indeed has been imitated as an ingenious method of possibly circumventing the risks of conventional suture technique.

In the 20th century, resection of the colon and primary anastomosis did not become generally used until the antibiotic era. In fact, in many centers, resection of the sigmoid colon was not attempted without a diversionary colostomy until the 1950s. Dixon, among others, is regarded as one of the advocates of primary anastomosis without colostomy.²⁴⁷ Numerous articles have been written that address the means to avoid fecal contamination with bowel resection (closed anastomosis with noncrushing clamps; rubber-shod clamps) and to avoid inverting too much tissue (e.g., the Gambee suture).³²² To complicate the issue further, Getzen and associates advocated the use of eversion, concluding that it produces a more secure anastomosis than inversion.³²⁹ Although this contention was supported by some studies,¹⁵²^,⁴⁰⁸^,⁴³¹^,⁵⁸⁷ other reports refuted the observation.³⁵⁵^,⁸⁷³ The hand-sewn eversion technique is in disfavor and should never be used. Suturing has historically been the true surgeons’ art form. But today, surgeons-in-training have only a limited opportunity to sew. Maintaining this skill, however, is critically important because there are times that one cannot employ a stapling technique.

de Petz was not the first person to describe a stapling device, but he is generally credited with having produced the initial practical instrument, employing it for gastrectomy.²⁴⁴ Stapling techniques, however, were developed as early as 1908 by Hültl and Fischer.⁹⁶⁹ However, the real credit must be given to the Russians for producing the contemporary

stapling instruments.³¹⁴ They, as well as others, performed stapled anastomoses that were sometimes everting and sometimes inverting.²³^,⁸²³^,⁸²⁴ Ultimately, the Russian SPTU gun was described. It produced an inverting end-to-end anastomosis by means of a circular row of staples placed within the lumen of the bowel. The United States Surgical Corporation (Tyco; Covidien, Inc., Norwalk, CT) was the initial developer of stapling instruments in this country, and other American companies produced further modifications. Reports of their successful application are myriad.³⁵²^,³⁵⁴^,⁴²⁹^,⁸²⁵^,⁸³⁶ The latest major innovations involve the application to laparoscopic bowel surgery (see Chapter 19).

FIGURE 23-35. Murphy button (A) with and without spring cup attachment (B). (From Murphy JB. Cholecysto-intestinal, gastro-intestinal and entero-intestinal anastomosis, and approximation without sutures. Med Rec N Y. 1892;42:665.)

BENJAMIN TRAVERS (1783-1858)

Benjamin Travers was born in London, the second of 10 children of Joseph Travers, sugar broker. After receiving a classical education at the Grammar School of Cheshunt, Hertfordshire, he was taught privately until he was put into his father’s countinghouse at the age of 16. He evinced a decided dislike for commercial life, and since his father frequently attended the surgical lectures of Henry Cline and Astley Cooper, he was articled to Cooper in 1800 for a term of 6 years. During the last year of his apprenticeship, Travers gave occasional lectures on anatomy to his fellow students and established a Clinical Society, meeting weekly. Travers spent most of the year 1807 at Edinburgh, and on his return began to practice. He was appointed Demonstrator of Anatomy at Guy’s Hospital and was elected in 1809 to the lucrative office of surgeon to the East India Company’s warehouses and brigade. The following year he was appointed surgeon to the London Infirmary for Diseases of the Eye. He is justly described as the first general hospital surgeon in England to devote himself specially to the treatment of diseases of the eye. Travers became a fellow of the Royal Society in 1813, and on May 1, 1815, was elected a surgeon to St. Thomas’ Hospital. He became president of the Hunterian Society in 1827 and in the same year was elected president of the Royal Medico-Chirurgical Society. Travers was a good pathologist, inheriting the best traditions of the Hunterian School. He was a man of cultivated mind and of a strong personality. As an operator, he was considered nervous and clumsy. Tradition assigns to him an exquisite polish of manners and states that he took off his hat and acknowledged salutes more elegantly than any contemporary dandy. He died at his home in Grosvenor Square, London, March 6, 1858. (From Power D, Spencer WG, Gask GE. Plarr’s Lives of the Fellows of the Royal College of Surgeons of England. London, United Kingdom: John Wright & Sons Ltd., Simpkin Marshall Ltd.; 1930:2 and with appreciation to Muath M. Bishawi, MD.)

ANTOINE LEMBERT (1802-1851)

Lembert was born April 19, 1802 in Nancy, France. Very little is known about his life, and no rendering of his likeness has been found. He studied under Guillaume Dupuytren in Paris, where he was epidemiologist to the Seine-Départment. In 1828, he won a 5,000-franc prize from the Académie des Sciences for his essay, la Méthode Endermique. He is particularly remembered for his intestinal suture, an inverting technique used in circular enterorrhaphy that was published in 1826 (Mémoire sur l’entéroraphie avec la description d’un procédé nouveau pour pratiquer cette opération chirurgicale). What has come to be known as Lembert’s suture ensures serosa-to-serosal apposition, the cornerstone of contemporary gastric and intestinal suturing. Johann Dieffenbach is credited with being the first person to employ his method successfully. Lembert died of stomach cancer at the age of 49.

CHRISTIAN ALBERT THEODOR BILLROTH (1829-1894)

Billroth was born in Bergen, on the Baltic Sea island of Rügen. When he was 5 years old, his father, an impoverished preacher, died, leaving a young widow with five sons, the youngest being Theodor. Billroth was considered a poor student, possibly, by today’s terminology, dyslexic. He wished to pursue a musical career; however, his mother and physician uncle encouraged him to enter medical school at the University of Greifswald in 1848. Although he was an indifferent student at that institution, he became inspired when he followed Professor Baum to the University of Göttingen. After 3 years of physiologic and pathologic investigations in a number of areas, he completed his studies in Berlin under von Langenbeck. After a year and a half of military service, travel, and failure to establish a general practice, Billroth fortunately obtained a position as assistant in the Langenbeck Clinic, his entrée into surgery. In 1859, he was appointed professor of surgery in Zurich. It was there that he wrote his famous textbook, General Surgical Pathology and Therapy, which ran through 11 editions. He also wrote a number of musical pieces. In 1867, Billroth accepted a professorship of surgery in Vienna. He was the first surgeon to resect the stomach for cancer, the first to resect the esophagus, and the first to perform a laryngectomy. Billroth declined many offers of other positions, including the chair in Berlin, preferring to remain in his adopted Austrian homeland with his close musical companions, Johannes Brahms and Johann Strauss. (Concerning enterorrhaphy. Dis Colon Rectum. 1986;29:284-287 [Über enteroraphie. Wien Med Wochenschr. 1879;1:1-6]).

WILLIAM STEWART HALSTED (1852-1922)

William Halsted was born in New York City to a prominent, affluent family. He attended Phillips Academy in Andover, Massachusetts and received his AB degree from Yale University in 1874. He entered medical school at the Columbia University College of Physicians and Surgeons in New York and graduated first in his class in 1877. After a time at Bellevue Hospital in New York, he became the first house physician at New York Hospital. In 1878, he went to Europe to study at the great institutions of the day. It was through his association with the German, Austrian, and Swiss masters that he developed his passionate interest in clinical and laboratory investigation. Returning to New York in the fall of 1880, he began his practice of surgery. Within a period of 6 years, he held appointments, including chief surgeon, in at least eight hospitals. His clinical duties were awesome, but he still managed to be actively involved in teaching and research. When an infected finger became complicated by painful neuritis, Halsted sought relief by the use of cocaine and became habituated. As an invalid unable to carry on his practice, he came to Baltimore in 1887, having been attracted to the university under the leadership of William Welch. It was there that his genius for investigation flourished. In 1899, Halsted was elected surgeon-in-chief of the Johns Hopkins Hospital, subsequently leading that department for 33 years. He laid the foundation for a model training school for surgeons and introduced numerous instruments and techniques: rubber gloves, hemostat, “mosquito” forceps, and “cigarette” drain, to mention only a few. He was an advocate of fine silk sutures, as opposed to catgut. His study of a comparison of suture techniques to effect intestinal anastomosis is generally regarded as the best example of the importance of laboratory investigation to the clinical setting. Halsted can with justification be called the father of American surgery.

OSCAR HUNTINGTON ALLIS (1836-1921)

Oscar Allis was born in Holley, New York, a direct descendant of a Puritan settler of the first colony at Salem, Massachusetts. He attended Lafayette College, receiving an MA degree in 1864, and graduated from Jefferson Medical College in 1866. Following an internship at the Philadelphia General Hospital, he ultimately became one of the original surgeons at the Presbyterian Hospital of Philadelphia. Allis acquired a special interest in orthopedic surgery and was recognized as an authority on fractures and dislocations. Relaxation of the fascia between the iliac crest and the greater trochanter, a sign of fracture of the neck of the femur, is Allis’ sign. Among his many valuable contributions to surgery were his dissector, his ether inhaler, and his splint—all called by his name. Today, Allis forceps continues to be one of the standard instruments in any general kit.

NICHOLAS SENN (1844-1908)

Nicholas Senn was born in the canton of St. Gall in Switzerland. In 1852, he came with his parents to the United States and settled in the frontier town of Ashford, Wisconsin. Following 2 years at a local country school, he attended Northwestern University. In 1865, he entered the Chicago Medical College, and on completing the course of study there, he won the competitive examination for residency at the Cook County Hospital. He practiced for 5 years in the small town of Elmore, Wisconsin, and in 1877 he traveled to Munich, where he spent a year studying bacteriology. Returning to Milwaukee, he was called to the chair in surgery at the College of Physicians and Surgeons in Chicago. In 1888, he became professor of surgical pathology at the Rush Medical College. Senn was a prolific writer, his manuscripts comprising 160 volumes. His personal library was believed to be the best private medical collection in America. Senn was one of the first surgeons in the Western hemisphere to recognize the importance of the animal laboratory as a resource for improving and modifying surgical technique. In association with Connell, it was said that he “constructed the arch and placed the keystone in the structure of intestinal anastomosis.”

CLAUDE F. DIXON (1893-1968)

Claude Dixon was born in Piedmont, Kansas and received his doctor of science degree from the University of Kansas in 1919 and his medical degree from the same institution in 1921. Following an internship at the University of Kansas Hospital, he entered the Mayo Graduate School in Rochester, Minnesota as a resident in surgery. In 1928, he became a member of the Mayo Clinic staff and head of the section of general surgery, a post he held until he retired in 1957. With a special interest in abdominal surgery, particularly surgery of the colon, Dixon became one of the leaders in the development of anterior resection. He was also a recognized authority in surgery of the head and neck and contributed more than 300 articles to the surgical literature.

Preoperative Preparation

For elective colon resection in the United States, patients are usually admitted to the hospital the day of surgery, depending on the likelihood of achieving an adequate bowel cleansing if the preparation is undertaken on an outpatient basis. A more prolonged hospitalization to prepare the bowel or to perform preoperative testing is generally not believed to be necessary and, in the era of cost containment, can rarely be justified to insurance carriers, even when the indications are reasonable.

Several studies have been published comparing inpatient and outpatient preparation for elective colorectal surgery.³¹⁵^,⁴¹⁰^,⁵⁶⁰^,⁷⁹¹ Results of both approaches are certainly comparable with respect to surgical complications. However, individuals with medical problems may not tolerate extensive fluid shifts and may require intravenous supplementation and monitoring.⁵⁶⁰ Certainly, if the preoperative preparation cannot be performed with safety because of medical conditions (cardiac, renal, hepatic) or because of certain logistic concerns, the procedure should be undertaken in a hospital setting.

Mechanical Preparation

A bowel preparation regimen consists of appropriate dietary restriction and mechanical cleansing. Generally, the patient is placed on a clear liquid diet 24 hours before the operation. Ideally, the mechanical preparation begins at noon the day before surgery. A vigorous cathartic is administered at this time so that its effect will have dissipated in time for the patient to have a reasonable night of sleep. The choice of laxative is a matter of the surgeon’s personal experience or prejudice because when adjusted for dosage, most laxatives have a comparable effect.⁶⁵,⁶⁶ and ⁶⁷^,²⁶¹^,³⁰⁴^,⁴²⁰^,⁴⁷⁹ Whole gut

irrigation has become (at least for surgeons if not patients) the most commonly used option. Another alternative that has achieved some converts is the use of Visicol tablets (sodium phosphate monobasic monohydrate, USP, sodium phosphate dibasic anhydrous, USP). This has the advantage of the patient consuming virtually tasteless tablets with comparable cleansing efficacy in clinical trials. At least of equal or perhaps greater importance than the administration of the cathartic itself, according to many surgeons, is the use of enemas until the returns are clear on the morning of the operation. The fact is that most surgeons base their protocol in this area on their subjective experience and habit.¹¹⁴³

Wolters and colleagues undertook a prospective, randomized study of preoperative bowel cleansing using three different methods—gut irrigation with Ringer’s lactate, Prepacol (a combined preparation comprising bisacodyl tablets and sodium phosphate solution), and polyethylene glycol (PEG).¹¹²⁰ All were equally effective in cleansing the bowel. However, the postoperative complication rate was significantly increased in the Prepacol group despite its being tolerated better. Overall, the authors observed that PEG is the recommended bowel preparation for individuals undergoing elective colorectal surgery. Oliveira and coworkers also performed a randomized, controlled study in which sodium phosphate and PEG-based oral lavage solutions were compared.⁷⁶² Although the two were equally effective and safe, patient tolerance was much greater with the sodium phosphate.

The question has been posed whether mechanical bowel preparation is truly necessary. This seems counterintuitive, suggesting that one turn back the clock to those bad old days of a high incidence of septic complications following elective colon surgery. But intraoperative spillage of bowel contents has been reported to occur in 14% of cases, increasing the rate of postoperative infectious complications. In one study, spillage was more likely to occur in patients receiving a mechanical bowel preparation owing to the liquid content.⁶²³ All surgeons agree that it is preferable to deal with formed stool in the bowel rather than to have watery feces spilling into the peritoneal cavity as a consequence of an incomplete preparation.

van Geldere and colleagues prospectively analyzed 250 consecutive operations without a mechanical bowel preparation, with no deleterious consequences observed.¹⁰⁴⁸ Burke and coworkers also reported whether a mechanical bowel preparation for elective colorectal surgery was indeed required.¹³² Patients were randomized to receive a standard mechanical preparation or none. The overall morbidity rate (18%) was similar in the two groups. The two deaths that occurred were both in individuals who had received bowel preparations. The authors concluded that bowel preparation does not affect the outcome after elective colorectal surgery.¹³² Zmora and associates randomized patients undergoing elective colon surgery into two groups: ethylene glycol bowel preparation and no mechanical preparation.¹¹⁴² Almost 200 patients were allocated to each group. There were no statistically significant differences between the two with respect to overall septic complications (10.2% vs. 8.8%), wound infection (6.4% vs. 5.7%), anastomotic leak (3.7% vs. 2.1%), and abdominal abscess (1.1% vs. 1.0%). In a meta-analysis of randomized trials, Wille-Jørgensen and colleagues found no evidence in the literature for any beneficial effects from the use of bowel cleansing.¹⁰⁹¹

Antibiotics

The nonabsorbable antibiotic regimen advocated by Nichols and coworkers has been believed to reduce the incidence of infectious complications.⁷³⁹ This consists of neomycin (1 g) and an erythromycin base (1 g) at 1 p.m., 2 p.m., and 11 p.m. the day before surgery. The use of a broad-spectrum systemic antibiotic immediately preoperatively, intraoperatively, and for one or two doses postoperatively has been suggested in numerous experimental regimens to reduce the incidence of infection following elective colon resection.²⁴⁰^,²⁸⁶^,³⁶⁰^,⁶⁶⁰^,⁶⁷⁶^,¹⁰⁷³^,¹⁰⁹⁶ In a survey of 352 colon and rectal surgeons, all favored some antibiotic preparation.⁹⁴⁸ Eightyeight percent preferred a combined regimen with oral and systemic antibiotics. Condon and associates, as well as others, have demonstrated no discernible benefit from adding parenteral antibiotic prophylaxis in elective colon surgery if mechanical cleansing and neomycin and erythromycin are employed.¹⁹³^,⁹⁷⁴ However, other investigators have found that the addition of perioperative parenteral cefoxitin, in comparison with oral antimicrobial agents alone, greatly reduces the incidence of wound infections in patients undergoing elective colorectal surgery.⁵⁵³ Corman and coworkers showed that single preoperative doses of cefuroxime and metronidazole are as effective in the prevention of postoperative infections associated with colorectal surgery as standard therapy with four doses of cefoxitin, but single doses of either cefuroxime or cefoxitin were less effective than the combination regimen.²⁰² Others reached a similar conclusion with the use of a single 2-g dose of cefotetan, given preoperatively and 12 hours postoperatively.⁴⁸⁰^,⁷¹⁵ Favorable results were reported with a single intravenous dose of Timentin (ticarcillin-clavulanic acid),⁹⁴^,²¹⁹^,¹⁰⁴¹ metronidazole-netilmicin,⁹⁴ ceftizoxime,¹¹⁰⁸ and mezlocillin.¹⁰⁴¹ Still others have found that oral ciprofloxacin offers advantages in efficacy and ease of administration and is as effective as other parenteral antibiotics.⁶⁵⁶ Jensen and colleagues concluded, based on their prospective, randomized trial, that a single dose of an appropriate antibiotic in acute or elective colorectal surgery is as protective for infection as a tripledose regimen.⁴⁹²

Song and Glenny used several databases and found 147 relevant trials in order to assess the relative efficacy of antimicrobial prophylaxis for the prevention of postoperative wound infection in patients undergoing elective colorectal surgery.⁹⁴⁹ Although they confirmed that antibiotic prophylaxis is indeed effective, they could not demonstrate statistically significant benefit with one or another regimen. Still, they were able to show that certain regimens are inadequate. These include metronidazole alone, doxycycline alone, piperacillin alone, and oral neomycin and erythromycin the day before surgery (note the earlier discussion). Furthermore, they concluded that a single dose administered immediately (or within 1 hour) before the operation is as effective as long-term postoperative antimicrobial prophylaxis.⁹⁴⁹

Opinion and Recommendations

It is clear that systematic antibiotic prophylaxis is indicated for elective colon surgery. Surgeons by now should have accepted the favorable results reported in many studies by using preoperative intravenous antibiotics within 1 hour of making the incision. Furthermore, there is no good evidence to suggest that the oral preparations administered the day before add any benefit. Therefore, the so-called Nichols-Condon preparation should be abandoned. All studies that determine effective blood levels with systemically administered antibiotics indicate that the drug should be circulating within 1 hour before making the incision, ideally less. The likely pathogens when performing colorectal surgery are enteric gram-negative bacilli, anaerobes, and enterococci. Therefore, the most beneficial and cost-effective drugs for elective, uncomplicated colorectal surgery are cefotetan, cefmetazole, or cefoxitin. Another alternative is cefazolin plus metronidazole. In the individual who is allergic to penicillin,
fluoroquinolone plus clindamycin is preferred (Medicare Quality Improvement Project, 2002). When there has been gross fecal contamination, in the presence of obstruction, perforation, abscess, and when there has been a prolonged operative time or excessive blood loss, antibiotics should be reasonably continued beyond the prophylactic regimen. A patient with valvular heart disease and immunocompromised individuals may require special consideration with respect to antibiotics (see Chapter 5).

Urinary Catheter

An indwelling urinary catheter should be placed for all bowel operations in my opinion. It may be removed the day after surgery. For an abdominoperineal resection or low pelvic operation, the urinary catheter should remain in place for a longer time in order to minimize the likelihood of subsequent urinary retention.

Ureteral Catheters

The routine placement of ureteral catheters before colonic surgery has been advised by some surgeons. However, most surgeons limit their use selectively, especially to those individuals who are to have reoperative pelvic surgery, who underwent radiation, and when the dissection is anticipated to be difficult. Indications are generally broadened for laparoscopic surgery (see Chapter 19). One must remember, however, that although they may aid in identifying the ureters (by palpation), care must be taken not to rely too much on them. The precaution of seeking and carefully dissecting the ureters away from the area of the surgical dissection should be taken, regardless of the ease or difficulty of the operation or the presence or absence of ureteral catheters. Lighted catheters, of course, permit easier visualization without the need to palpate, but they are quite expensive. Darkening the operating room to facilitate identification by these means is a nuisance. Ureteral catheters are not harmless—ureteral edema, oliguria, and anuria, as well as ureteral injury itself, are potential consequences.⁹²⁶ However, one unquestioned benefit is that there is no doubt as to the location or nature of the injury when the catheter is divided. It is suggested that the reader peruse an excellent article on ureteral stents by Saltzman.⁸⁸⁶

It is wise to remove one catheter at a time. Generally, the first may be taken out at the end of the operation. The second can usually be removed the following day. Still, prudence suggests that one use the services of the urologist in determining the timing of withdrawal—after all, he or she is the individual responsible for inserting the catheter in the first place and should be given the opportunity to provide an opinion.

Hydration

Overnight intravenous hydration before operation is advisable for any patient who may be prone to adverse cardiac or renal consequences of excessive fluid loss. Oral hydration during catharsis is also helpful and is one of the reasons why a PEG preparation may be the better mechanical cleansing for high-risk individuals.

Hyperalimentation and Nutritional Assessment

Hyperalimentation has been advocated before operation for the nutritionally depleted patient in order to prepare for the assault on the patient’s metabolic processes. Many individuals with colorectal cancer have lost weight, are anemic and hypoalbuminemic, or have a variety of fluid and electrolyte problems. With the exception of weight loss, however, laboratory abnormalities can usually be corrected by 2 or 3 days of appropriate therapy. Delaying surgical intervention for a week or more to administer hyperalimentation is an expensive, time-consuming extravagance that has in itself an associated morbidity.

Nutritional assessment has also been advised because of its potential in anticipating postoperative morbidity and mortality. In the experience of Thompson and colleagues, subnormal findings for triceps skinfold thickness and percentage of ideal body weight were associated with a higher complication rate, but these occurred so infrequently that the routine use of such tests was not recommended.¹⁰¹⁴ Ondrula and coworkers identified a number of preoperative factors that increased the risk of colon resection: emergency operation, age older than 75 years, congestive heart failure, prior abdominal or pelvic radiation therapy, corticosteroid use, serum albumin less than 2.7 g/dL, chronic obstructive pulmonary disease, previous myocardial infarction, diabetes, cirrhosis, and renal insufficiency.⁷⁶⁴

Opinion

One is impressed by the salutary effects of surgery and the consequent improvement in the nutritional status after the expeditious removal of colonic pathology. The patient can usually commence a diet rather quickly. Therefore, I am not an advocate of the use of hyperalimentation in the preoperative management of patients who need surgery for malignant disease.

Nasogastric Intubation

The routine use of a nasogastric tube is discouraged for colon resections. It does not protect the anastomosis and merely causes patient discomfort. In a prospective study by Colvin and colleagues comparing the preoperative use of a long intestinal tube (Cantor), a nasogastric tube, and no tube, there were no significant differences in the lengths of hospital stay, duration of postoperative ileus, adequacy of intraoperative intestinal decompression, gastric dilatation, and postoperative complications.¹⁹¹ Wolff and coworkers counsel that even though there is an increase in the rate of minor symptoms of nausea, vomiting, and abdominal distension when nasogastric decompression is not employed, nasogastric intubation still is not warranted prophylactically.¹¹¹⁵ The only caveat with respect to advocating the prophylactic use of a nasogastric tube is the patient who is obtunded or who is at serious risk for aspiration. Obviously, on a therapeutic basis (intestinal obstruction, postoperative vomiting), gastric drainage is indicated.

Thrombophlebitis Prophylaxis

Only gold members can continue reading. Log In or Register to continue