COLON CANCER, an excellent overview

Colorectal cancer represents an interaction between the genome of the colorectal epithelial cell and the host environment. Both factors are essential for the development of tumor. In 1997, there were 131,000 new cases of colorectal cancer, which resulted in 57,000 deaths. Also, in 1997 in the United States, overall, there were approximately half a million-cancer deaths, of which 10% were caused by colorectal cancers. With respect to the epidemiology of colon cancer, there is a wide geographic variation. Industrial nations have the highest incidence, except for Japan, in which the incidence of colon cancer has been low until recently. Interestingly, the incidence of rectal cancer in the United States versus Japan is almost equal. We know from population migration studies, that patients who come from low risk geographic regions who move to high-risk environments tend to increase their incidence for the development of colorectal neoplasms. Hence, this suggests environmental factors having a more important function for the development of colorectal cancer versus genetics. From 1974 until 1991, the incidence of colorectal cancer in Japan has increased three-fold, and now is closely approaching that of most other western nations. Again, this would suggest that changes in environmental factors do have a more important role in the development of colorectal neoplasms. Interestingly, from 1950 to 1990, the incidence rates of colon cancer in the United States per 100,000 population has slowly increased from approximately 49 to 60, but mortality has slightly gone down, suggesting that the impact of early detection and treatment has had a positive impact with respect to this disease.(1.)

At present, there are no consistent observations of high-risk industries or occupations associated with colorectal cancer. Smoking, which has clearly been shown to increase the risk of adenoma prevalence, has not been directly related to the development of colorectal cancers. Alcohol use, on the other hand, specifically beer, increases the risk for the development of colorectal cancer two- to three-fold.(2,3,4) Several theories as to why alcohol increases the risk of colorectal cancer have been put forth, but most suggest an associated folate or methionine deficiency. It is known that women who have undergone pelvic irradiation for gynecologic matters, have a two- to three-fold increased risk of colorectal neoplasm.(5) At present, asbestosis, either inhaled or in the water, has no associated increased risk for colorectal cancer.

Dietary patterns

Clearly, caloric intake has been associated with colon cancer. The estimated risk of colon cancer increases approximately 2.3% for each 100 calories per day consumed.(6) In a study of 80,000 women, intake of animal fat of the highest quintile, was associated with a significantly elevated risk of colon cancer.(7) In addition, elevations of cholesterol greater than 276 have been suggested to increase the relative risk by 1.65 of colorectal cancers.(8) In the Framingham heart study, there was a four-fold increased risk for colon cancer, especially in obese men who had low cholesterol.(9) This interesting development has been subsequently shown in several other studies to suggest that with the development of cancer, cholesterol levels actually may decline, but why this is so is unknown. The question arises then, why is fat content of diet associated with colon cancer. Clearly, the amount of fat consumed has some role of bile acid composition, and subsequent bile acid activity on colonic epithelial proliferation. By increasing animal fat from 62 to 152 grams per day, there is an increase in the total fecal bile acid and fatty acid excretion in humans, yet there is no change in fecal fat, frequency of stools or transit time.(10)

Dietary fiber

If all data written on the topic of fiber is reviewed, there seems to be a consistent protective effect of fiber.(11,12) Fiber has been suggested to decrease fecal transit time, hence, diminishing the time that stool with toxins, and toxic bile acid exposure to epithelial cells. In addition, fiber dilutes the overall concentration of other colonic constituents. There is some evidence to suggest that fiber may also bind toxins, and that the presence of fiber decreases fecal pH, and subsequently generates short chain fatty acid, which may in part protect isolated colonic epithelial cells from deoxycholic acid.

Calcium

To date, with respect to calcium, there have been no clinical trials to suggest that calcium has a supportive effect in protecting the colon against cancer.

Selenium

It has been shown in several population studies that low levels of selenium are associated with increased risk of colon cancer.(13) Selenium is a constituent of glutathione peroxidase, which prevents free radical damage to tissue, and by an independent mechanism, antagonizes toxicity caused by heavy metal.

Vitamins: A ,C & E

To date, vitamins A, C and E, which are all well known antioxidants, have not been conclusively shown as agents that reduce the risk of colon cancer, though several recent studies published in the last several years may suggest some component of a protective effect.(14,15,16,17)

NSAIDs & Aspirin

Recently, much interest has been directed towards Sulindac. This, a prodrug, which is used as a nonsteroidal, has been shown to cause regression in polyps in patients afflicted with familial adenomatous polyposis.(18,19,20) In addition, several long-term studies have suggested that users of aspirin have a significant diminished incidence of GI cancers in general.(21,22,23,24) Both sulindac and aspirin most likely have the protective benefits, not because of their prostaglandin production modulation in the GI tract, but rather because of program cell death [apoptosis]. This apoptosis probably occurs in adenomatous cells, but not in cancer cells, and hence, may suggest the observed regression of adenomatous tissue in patients afflicted with FAP.

Recommendations with respect to diet

Hence, at present, diet plays the greatest role probably in the prevention of colorectal cancers. I can suggest at present that high fat intake, or lower fiber intake most probably are the two single likely factors identifiable by present literature to be directly associated with increased incidence in the development of colorectal cancer. Hence, with out patients in whom we find polyps, we make the recommendation that they decrease their caloric intake, they decrease their dietary fat to less than 25% of their total caloric intake, we stress 25 g of fiber daily in their diet and we stress increasing their diet with fresh fruits and vegetables.

Etiology Colon Cancer

The etiology of colon cancer is a rapidly evolving area of genetic research. Much of what we known about the genetics of colon cancer is based on patients with familial adenomatous polyposis (FAP) and hereditary nonpolyposis colorectal cancer syndromes. The sequence of events in FAP, and very likely in sporadic colorectal neoplasms, is multi-stepped. First, there most probably needs to be a germline mutation in the adenomatous polyposis coli gene, known as APC. This gene is located on chromosome 5Q. This germline mutation then leads about to early adenoma formation, after which, a sequence of mutations in the K-Ras cellular proto-oncogene, P53 gene, and LOH occurs. Much has been written recently about K-Ras cellular proto-oncogenes. In adenocarcinoma, the mutation is located at critical positions in the gene, primarily in the Ras protein. These mutations are such, that signal transduction is in some way activated, leading to additional cell growth. Probably, the single most important gene disruption is with the P53 gene. The P53 gene serves as a cell cycle checkpoint regulator, which prevents nuclear replication after injuries that are likely to occur to DNA. In the presence of damaged DNA, P53 protein levels rise in the cell, and progression to cell cycle is prevented at the G-S arrest. In-patients with adenocarcinoma, there is an over expression of a defunct P53 protein, and hence, allowing unregulated growth.

Clinical manifestations and risk factors for colorectal neoplasms

In general, tumors of the colon grow slowly. Hence, more often than not, diagnosis is by happen chance, usually today, in a screening setting or in the patient being worked up for hemoccult positive stool. In general, as tumors grow, and become more circumferential, it will cause symptoms of partial, if not true obstruction. Though of note, remember that the diameter of the right side is significantly larger than that of the left, and hence, tumors which grow on the right may take a considerable amount of time before clinical symptoms can occur versus tumors occurring on the left. As with all neoplasms, once tumors have penetrated beyond the serosa, local invasion can occur, causing symptoms of locally invaded organs.

As stated above, almost, if note all, colon cancers arise from adenomas , and the majority of these being initially polyps. In general autopsy studies performed in the past, approximately 30-40% of colons contain some form of adenomatous polyps, though the vast majority of these are less than 1 cm in diameter. A traditionally sighted study reported that patients with adenomas had a 50% risk of developing a metachronous adenoma after 15 years of observation, with a risk of 1 in 12 for men, or 1 in 20 for women, for developing cancer in that time frame.(25) Because the prevalence of adenomas is high, attempts have been made to identify subsets at greater risk of developing cancer. It is now depreciated that not all adenomas carry the same risk of recurrence to their host. A pair of retrospective studies reported that patients who had small adenomas removed, those smaller than 1 cm, had a diminished risk of recurrent neoplasia, and their survival rate was not reduced compared with general risk estimates established for this community. On the other hand, those with adenomas greater than 1 cm had an increased risk of cancer 2.7 times that estimated for the general population. In a similar opinion, the relative risk for rectal cancer was found to be 0.6 for patients who had small less than 1 cm rectal adenomas removed in a British, with an average follow-up period of almost 14 years. In contrast, the removal of large adenomas was associated with the relative risk of rectal cancer 2.1- to 2.6-fold greater than lesions 1-2 cm in diameter, and lesions larger than 2-cm respectively.(26,27)

Clearly, it has been well established that patients with chronic inflammatory bowel disease (Crohn's Disease and Ulcerative Colitis) are at greater risk for developing colorectal neoplasms. Though, the exact risk of this has been of some controversy in the literature. Partially, this controversy is based on issues of tertiary care referral center bias. Irrespective though, when all studies are looked at, there clearly is a significantly increased risk of developing colorectal neoplasms in patients with ulcerative colitis, and this risk increases with duration of disease, and distribution of disease. In-patients who have pancolitis, that is, colitis that involved the entire colon, that individual’s risk will be much greater than in an individual whose colitis remains only on the left side. Historically, we have told our patients that if they had Crohn’s disease, that their risk for colon cancer was no greater than the general population. This, though, is wrong. Unfortunately, the studies to date have note been able to clearly delineate what the true incidences of colorectal cancer are in Crohn’s patients. Though, the association of colorectal cancer and the relative risk with ulcerative colitis has been well established.

In the past, it has been suggested that cholecystectomy (the removal of the gallbladder) could increase the risk of colon cancer. Several recent studies though, and one meta-analysis has demonstrated no excess in the risk of either adenomas or carcinoma after cholecystectomy.(28,29) Hence, it is reasonable to disregard neoplastic risk when making decisions regarding gallbladder surgery, and there is no need for extra surveillance in patients who have had this operation.

Clearly, acromegalic patients have a three-fold excess of colon cancers. The role of growth hormone has not been explored, though the association with elevated serum gastrin levels in these patients may have some explanation to this increased risk.

In the past, the presence of skin tags suggested in some literature the presence of subsequent colon polyps. Unfortunately, recently studies have not been able to prove this, and hence, their significance is of little.

Patients, who have had streptococcus bovis infections, have been suggested to have an increased risk of colon cancer. Hence, such patients who have such bacteremia probably should have complete colonoscopy.

Pathology

At present, the majority of colon cancers reside in the distal colon, that, a distance which is reachable with flexible sigmoidoscopes. Approximately 65% of all colorectal cancers occur within 65 cm from the dentate line. In addition, 66% of all adenomatous polyps may be found in this region. Only approximately 30% of colorectal neoplasms arise in the proximal colonic anatomy, extending from the cecum to the splenic flexure. Approximately 34% of all polyps are found in this region. Interestingly, in the last several decades, there has been gradual proximal migration of colorectal neoplasms, with the incidence of right-sided tumors increasing over the last several decades.

Historically, colon cancers were staged by the Dukes-Turnbull classification. This staging system comprised of carcinoma in-situ, that being a malignancy in the tip of a polyp, which does not extend into the stalk. These tumors tend to be noninvasive, and with polyp resection, the actuarial cancer free 5-year survival rate is 100%. Dukes A are tumors that invade to the level of the muscularis propria. With resection, their actuarial cancer free 5-year survival rate is estimated from 95-100%. Dukes B cancers penetrate the muscularis propria, and may extend through the serosa into the pericolic fat. Such tumors, when resected, have an actuarial cancer free 5-year survival rate of 80-85%. Dukes C tumors have regional lymph node metastasis, and actuarial cancer free 5-year survival rates diminish markedly to 50-70%. With distant metastasis at the time of diagnosis, actuarial cancer free 5-year survival rates probably are less than 10%. Of recent, staging has changed from utilizing the Dukes-Turnbull classification system and instead, T&M staging criteria have become more en vogue.

Screening for colorectal cancer

Because colorectal cancer produces few symptoms while the tumors are small and most readily curable, screening for asymptomatic patients has been abdicated. In this context, screening refers to testing patients in the absence of specific symptoms. Screening for colorectal cancer can reduce the mortality from this disease. The principal issues of discussion are, which screen approach is optimal, the cost of the program and how much screening can be afforded. It has long been appreciated that colorectal neoplasms lead early in their natural history. Hemoglobin contains peroxidase activity, which can be detected using the guaiac test. For the last several decades, the least invasive standard for screening colorectal neoplasms has been the fecal occult card test. Specificity for hemoccults has been suggested to be around 98% with a sensitivity of 32%. The positive predicted value, therefore, being around 23%. There are different types of reagents and different types of occult card tests.

At present, general recommendations for screening in-patients without symptoms, less than the age of 50, in whom there is no family history of colon cancer or adenomatous polyp in first-degree relatives, these patients are not screened. Over the age of 50, in patients who have no symptoms of colorectal cancer or change in stool habits, and in patients who do not have first-degree relatives with adenomatous polyps or family members with colorectal cancer, a recommendations of annual fecal occult blood tests versus flexible sigmoidoscopy every 5 years versus dual contrast barium enema every 5 to 10 years has been recommended. In addition, colonoscopy every 10 years in this patient population has also been suggested. At present, there have been several small series suggesting that colonoscopy at a discounted price, may be the true gold standard that we should be attempting to achieve for colorectal screening. In-patients without symptoms of colorectal cancer, but who have family members who are first-degree relatives who had previously been diagnosed with adenomatous polyps before the age of 60, more vigorous screening is recommended. In addition, in individuals with first-degree relatives with colorectal cancer, such screening should begin at the age of 40. Whether to perform colonoscopy or flexible sigmoidoscopy with fecal occult blood tests annually is still to be debated.

What is the definitive diagnostic test for colorectal neoplasia?

If the initial diagnostic procedure is a barium enema, a negative result may create a diagnostic dilemma. In a study of 97 patients who had persistent large bowel symptoms and had a negative air contrast barium enema, colonoscopy revealed 4 carcinoma and 24 adenomatous polyps not detected by the radiologic approach.(30) In a similar study of 76 patients with symptoms of colonic disease, the diagnostic evaluation began with a rigid sigmoidoscopy, after which all the patients underwent flexible sigmoidoscopy, double contrast barium enema and then colonoscopy. The double contrast barium enema alone reached a final diagnosis in 67% of patients, whereas colonoscopy was successful in 91%. The addition of flexible sigmoidoscopy to double contrast barium enema improved the diagnostic yield to 76%.(31) This suggests that colonoscopy is the preferred diagnostic procedure for the initial work up of a patient with symptoms of colonic disease, and moreover, indicates that colonoscopy should be used to evaluate colonic symptoms or bleeding, even if the air contrast barium enema is negative. A review of 31 colon cancers overlooked on double contrast barium enema revealed that half of the lesions were missed as a result of receptive errors. That is, the lesions were visualized on the x-ray film and recognized in retrospect, but they could not be identified as neoplasms on the initial reading. Another third of the lesions were missed because of perceptive error complicated by technical factors obscuring the lesion, tumor that was not visible even on respective viewing in 10%, and in 6% the lesions were seen, but misinterpreted. In addition, almost one-third of barium enemas in patients older than 65 years of age are technically inadequate because of inability of the patient to obtain a suitable preparation or to cooperate with the procedure. Clearly, the figures obtained are highly dependent on the scale and persistence of the radiologist and endoscopist. It must be emphasized that some neoplastic lesions may be missed by the endoscopist. In a prospective study of tandem colonoscopies, (two consecutive examinations done sequentially on 90 patients by different endoscopists), demonstrated that an experienced endoscopist missed about 15% of neoplastic lesions smaller than 10 mm, but no lesions larger than 10 mm.(33) It appears that at present, colonoscopy will remain the gold standard for the diagnosis of colorectal cancer, and it would appear unlikely that any indirect approach such as radiology or sonography can approach this technique for sensitivity. Though, of recent, has been the development of virtual colonoscopy. This technique utilizes a standard CT scanner with specialized software, and preliminary studies suggest that this technique may someday equal colonoscopy in its ability to detect polyps and cancers, yet still, in the removal of polyps, colonoscopy for the near future will be with us.

For other related sites on Colon Cancer, visit: Our Links Page!