Cancer for Laymen

Colon Cancer

Prof. Dr. Giancarlo Marra


One of the best known processes of tumorigenesis(1) in humans is that which occurs in the colon (or large intestine). Thanks to the major advances achieved in the last three decades in the fields of endoscopy(2), histology(3), and molecular pathology(4), cancer of the large intestine is no longer viewed as a single disease entity: several distinct phenotypes(5) have been identified. And this phenotypic variability is already evident in the precancerous lesions(6) that develop in the gut mucosa(7) . Even today, these lesions are often referred to collectively as colorectal polyps(8) . However, although most of these premalignant lesions are raised, polyp-like growths, more recent research has revealed that there are others that are only slightly elevated above the mucosal surface, or flat, or even depressed like a crater. Polyps are much easier to see during routine colonoscopy, and that is one reason they have received so much attention. But the nonpolypoid lesions are now being identified with increasing frequency, in part because clinicians are becoming more aware of their existence and importance and in part because of the development of more sensitive endoscopic techniques.

Precancerous colorectal lesions are also collectively referred to as adenomas(9) . This term refers to the pattern of cellular dysplasia(10) seen by the pathologist who examines the lesion under a microscope. The adenomatous pattern is very common in precancerous colorectal lesions, but it is not the only pattern. Some benign lesions have cells that are arranged in a saw-toothed or serrated pattern, and they seem to give rise to a peculiar colorectal cancer phenotype(11) (see below).

The phenotype of a tumor is the outward expression of the specific genetic and epigenetic alterations(12) found in the tumor cells. Some of these somatic alterations have already been well defined; others have been partially characterized, and many have yet to be identified. Changes affecting the genes have wide-ranging effects that are not limited to the appearance of the tumor and the arrangement of its cells: they also determine how the tumor behaves, its aggressiveness and responsiveness to anti-cancer drugs. Thanks to the availability of high-throughput analytical tools (e.g., genomics, transcriptomics, proteomics, etc.(13) ), we can now identify, in each colon tumor tissue, a huge number of molecular characteristics that produce these phenotypic features– and this is an essential step toward individualized (and hopefully more effective) treatment regimens.

(1) the process by which tumors – benign and malignant -- are formed
(2) visual examination of the interior of a hollow body organ by means of a long slender tube-like instrument known as an endoscope
(3) the branch of biology that studies the microscopic structure of animal or plant tissues
(4) a branch of pathology that focuses on the study and diagnosis of disease through the examination of molecules within cells, organs, tissues or bodily fluids
(5) the characteristics of a tumor that reflect interaction between its genetic characteristics and the environment. It includes macroscopic features like shape and size, microscopic characteristics (e.g., the appearance of the cells and their arrangement), submicroscopic or molecular characteristics (e.g., proteins that are missing, others that are produced in excess), and behavioral features like aggressiveness or susceptibility to different types of treatment
(6) areas presenting nonmalignant abnormalities that are likely to become invasive cancers if they are not removed
(7) the mucus-secreting membrane lining all body cavities or passages that communicate with the exterior
(8) growths that rise above the surface of the surrounding tissue
(9) a histological category of benign tumors characterized by a glandular structure or glandular origin
(10) abnormal growth or development of cells or tissues
(11) the characteristics of a tumor that reflect interaction between its genetic characteristics and the environment. It includes macroscopic features like shape and size, microscopic characteristics (e.g., the appearance of the cells and their arrangement), and submicroscopic or molecular characteristics (e.g., proteins that are missing, others that are produced in excess)
(12) Genetic changes are mutations that alter the sequence of the nucleo
tide units that make up the DNA. Epigenetic changes also alter the expression of a gene, but they do not involve sequence alterations and they are also potentially reversible
(13) large-scale studies of genes, RNAs and proteins in a cell or organism


Endoscopic view of a nonpolypoid adenoma
of the colon (stained with indigo carmine)
Adenomatous cells stained (brown color) with
anti pH2AX antibodies (histologic view)

The incidence(1) of colon cancer―and many other cancers as well― increases with age. This means that in countries where life expectancy is high and the mean age of the population is on the rise, improved strategies for prevention and early diagnosis are unlikely to lead to a rapid decrease in the number of new cases of colon cancer. The widespread use of colonoscopic screening has been shown to reduce colon cancer incidence and mortality, but neither has decreased substantially over the last 20 years. And the decreases that have been observed are not uniformly distributed across the general population. The variable success of these prevention / early diagnosis programs is probably related in part to the characteristics of the populations being screened, i.e., education levels, economic status, and access to health care.

The risk of colon cancer is indisputably linked to several behavioral or lifestyle factors. The protective effects of a diet that favors fruits, vegetables, and fish over meat and high-fat foods (especially when it is associated with regular physical activity) are supported by solid evidence. Numerous studies have been conducted to identify the specific nutritional components that might confer protection against colon cancer. The problem with these studies is that each focuses on a single type of dietary intervention. But given the complexities of our body and its metabolism, it now seems likely that appreciable preventive effects will be achieved only with multiple dietary and lifestyle changes (regular physical activity, for example). For this reason, isolated dietary modifications (increased intake of fiber, calcium/vitamin D, or folic acid) have not been shown to reduce the risk for colorectal adenomas or cancer in clinical trials.

The prophylactic(2) effects of various drugs have been also extensively investigated, and positive results have been demonstrated for aspirin, the anti-inflammatory drug celecoxib, and the cholesterol-lowering statins. But the concept of using drugs to prevent a common disease like colorectal tumors has also raised concerns. About 30% of asymptomatic persons in their fifties are found to have one or two colorectal precancerous lesions(3) at screening colonoscopy. Should one of the prophylactic drugs mentioned above be prescribed for all of these individuals? Should all individuals who have reached middle age be started on prophylactic statin or anti-inflammatory drug therapy? There are two important facts to remember here. The first is that probably only around 5% of these precancerous lesions will actually lead to cancer within the next 5-10 years if they are not removed. The second is that all drugs have contraindications and potential adverse effects that can be more or less serious. Clinical trials on the use of celecoxib, for example, were recently halted because there was an increase in serious cardiovascular events among the study participants being treated with this drug. These caveats are not intended to downplay the importance of precancerous colorectal lesions or to exaggerate the risks associated with aspirin, statins, or anti-inflammatory drugs. The point is that preventing colon cancer with drugs is a highly complex question that involves careful weighing of the pros and cons in each case, and it is certainly not a decision that can be made by patients themselves!

We now know that around one out of ten colorectal cancers is associated by epigenetic alteration(4) of a gene called MLH1. This alteration seems to occur frequently in a fairly well defined subset of cancers with the following features: 1) They are usually found in persons over 70 years of age; 2) they are particularly common in women, and 3) they are located almost exclusively in the proximal colon(5) . The epigenetic alteration “silences” MLH1 expression, which means that the protein it encodes is no longer produced. The protein in question is involved in an important cell process known as DNA mismatch repair(6) . When it is not produced, DNA mismatch repair goes awry, and the benign precancerous lesion suddenly begins to accumulate mutations in its DNA. These mutations fuel the transformation process, and the benign lesion is rapidly transformed into an invasive cancer. (This process frequently occurs in the precancerous lesions with serrated histology mentioned in the Introduction section.) The incidence of this type of colon cancer peaks among 70-75-year-olds, which means that it occurs more than 5 years later than other sporadic colon cancers(7) . This information has important implications for colonoscopy screening in the elderly. In persons over 70 years of age, it is especially important to explore the entire colon, including the proximal segment.

(1) the rate of occurrence of new cases of a particular disease in a population being studied
(2) guarding from or preventing the spread or occurrence of disease or infection
(5) the initial segment of the large intestine and therefore the one that is farthest from the rectum. It is sometimes referred to as the right colon because it lies in the right side of the abdominal cavity
(6) a process involving multiple proteins that correct certain types of errors―base pair mismatches―that can occur when genes are copied prior to cell division
(7) those that occur in a seemingly random fashion involving single patients as opposed to those that develop in families where 2 or more members have the same kind of tumor. Sporadic cancers are much less likely to be inherited.

Early Detection and Diagnosis

High-resolution endoscopic image of the surface of normal mucosa
(each dot is the pit of a gland, which is 35-50 microns large)
Abnormal pit pattern of an adenoma

These are areas in which we can expect to see a tremendous amount of innovation and improvement within the next few years. There are basically two lines of active research. The first involves diagnostic biomarkers(1) that can be used to identify patients who are likely to have colorectal tumors (which―regardless of whether they are benign or malignant―are usually asymptomatic in the early stages). The goal of current research is to identify substances whose presence or abundance in a blood or stool sample (both of which can be easily collected with no discomfort to the patient) will provide an accurate estimate of the likelihood of colon cancer. If the biomarker is positive, the patient can be scheduled for colonoscopy to find out if he/she really does have a tumor. This type of selection is quite important, since it restricts colonoscopy to those cases where there is a substantial probability of finding a lesion. The benefits in terms of screening costs are obvious. In addition, however, patients are likely to be more willing to have their blood or stool analyzed regularly than to undergo a colonoscopy.

Today the fecal occult blood test(2) is widely used for early detection of colon cancer. It is easy to perform and relatively low-cost, but its accuracy is also fairly low. (Small amounts of blood in the stool may also be caused by conditions that have nothing to do with cancer – hemorrhoids, for example, or nose bleeds, or even a meal of under-cooked meat!) More advanced versions of this test based on immunochemical detection(3 ) of hemoglobin are producing promising results. Numerous laboratories are working to characterize the molecular alterations(4) that characterize the cells of colorectal adenomas or cancers. The hope is that these alterations can then be detected in the feces (which also contain dead cells that have been shed from the normal intestinal mucosa and from any tumors that might be present in the gut) or even in the serum (due to passage of specific molecules from the tumor into the blood vessels that drain it). Efforts are being made to improve methods for extraction of DNA, RNA, and proteins from fecal material. And recent advances in the field of proteomics(5) (in particular, a technique called mass spectrometry(6) ) are allowing the detection of tumor-specific peptides (proteins or pieces of proteins) with high sensitivity and specificity (two requisites for accurate and reliable early diagnosis).

It is becoming increasingly clear that one biomarker is probably not going to be able to alert us to the presence of all colorectal tumors. For this reason, many groups are assessing the diagnostic power of different panels of genetic and epigenetic alterations that are frequently found in the tumor DNA shed in the feces. Certain alterations are found in precancerous lesions; others are characteristic of advanced cancers; and some are present in both phases of the disease. To detect all of these tumors, fecal DNA is analyzed for the presence of around 10 different alterations. The specific changes that are sought are likely to change frequently over the next years. The goal is to find the combination that allows the highest diagnostic accuracy. A similar approach is being pursued with protein biomarkers.

If the results of fecal and/or serum testing are positive, the next step is endoscopy, and this is where improvements can be expected from the second main line of research. Today 5-10% of all colon lesions are missed during endoscopy, and this is especially true of those that are nonpolypoid. Advanced endoscopic technologies (high resolution endoscopes, robotic endoscopes, chromoendoscopy, confocal endomicroscopy, etc.(7) ) are not only improving the accuracy of these examinations, they also reduce the discomfort and potential side effects associated with the procedure. They are particularly useful for examining patients with ulcerative colitis(8) . The gut mucosa is much more vulnerable in these cases, but even more important is the fact that neoplastic colorectal lesions in these patients are frequently small and nonpolypoid and therefore easily missed. Standard and high-quality colonoscopy are both being used more and more for treatment as well as diagnosis. This means that, for example, it is also possible to safely and reliably eliminate large colorectal precancerous lesions without resorting to surgery.

In the future, other approaches might also be available for pre-colonoscopy screening. One possibility is “virtual colonoscopy,” which involves a completely noninvasive examination of the colon with imaging techniques like computed tomography (CT) or magnetic resonance imaging (MRI). These examinations (CT and MR colonography), which are now in the final stages of development, will one day be used to identify persons who need therapeutic colonoscopy and to plan in advance how this latter procedure will be carried out.

Approximately 20% of all colorectal cancers are thought to be the result of an inherited predisposition (sometimes to more than one type of cancer). These cancers strike multiple members of a family, so the relatives of a patient diagnosed with this type of cancer had to undergo yearly colonoscopic examinations. Today the predisposing germ-line gene mutations(9 ) have already been identified for many of these familial cancer syndromes, and this has vastly improved the early diagnosis and prevention of this type of colorectal cancer.

Around 3% of all colorectal cancers are caused by an inherited defect in one of 4 genes that play important roles in the DNA mismatch repair system(10) (like MLH1, which was discussed in the Prevention section). Families with this condition, which is known as Lynch syndrome, usually include several members who develop colorectal cancer fairly early in life (age 35-55 years). The inherited defect facilitates the development of other cancers as well (endometrial, ovarian, gastric, duodenal, and small intestinal). There are now two relatively simple tests to determine whether a malignant tumor is caused by defective DNA mismatch repair. The first involves the analysis of DNA extracted from tumor cells for microsatellite instability(11) , a phenomenon that is specifically associated with mismatch repair-deficiency. The second method involves immunostaining(12) of tumor sections with antibodies against the 4 major mismatch repair proteins (MSH2, MSH6, MLH1, and PMS2). Absent or weak staining means that one of these proteins is absent, and DNA mismatch repair is therefore deficient. The advantage of this method is that it tells us which of the 4 proteins is not being produced. The gene that encodes that protein can then be examined for mutations. Once the mutation has been identified, blood samples from the patient’s first-degree relatives(13) can be tested for the same mutation with a simple polymerase chain reaction(14) -based assay. Not all family members will harbor the mutation―each child born to a parent who carries the mutation has a 50-50 chance of inheriting the defect―but for those who do, intensive endoscopic surveillance is offered that can significantly reduce the risk of developing colorectal cancer. The major breakthrough involves those who do not carry the mutation. Before the era of genetic testing, there was no way to distinguish these individuals from their mutation-harboring relatives: one had to assume that they, too, were predisposed to cancer. Now if the mutation is found in a specific family and genetic tests are negative in a specific individual in that family, this family member can be reassured that his/her risk of cancer is no greater than that of the general population, and there is no reason for him/her to submit to yearly colonoscopy.

Similar diagnostic and clinical strategies are used to identify and prevent colorectal cancers associated with other mendelian(15) cancer syndromes, such as familial polyposis(16) , which is caused by germ-line mutations in the APC and MYH genes and accounts for 1% of all colorectal cancers.

(1) a biological substance or phenomenon whose presence or abundance is indicative of a process, event, or condition (e.g., aging, exposure to a toxic substance, or disease)
(2) a check for blood present in the feces which is not visibly apparent
(3) a laboratory technique that identifies and quantifies (usually in minute amounts) a protein (e.g., a hormone or an enzyme) based on its ability to react with a specific antibody
(4) alterations of the properties of molecules, such as proteins or RNAs, in a cell, with important repercussions on their interactions
(5) the large-scale study of the entire complement of proteins –the proteome- in a cell or organism
(6) an analytical technique for the determination of the elemental composition of a molecule, such as a protein whose amount in a cell can thus also be quantified
(7) recently developed technologies and techniques that increase the diagnostic accuracy of colonoscopy and improve the clinical management of patients
(8) a chronic inflammatory bowel disease that is associated with an increased risk of colorectal cancer.
(9) mutations that are inherited from the parents and are therefore present in all cells of the body, not just tumor cells
(11) changes of the length of microsatellites that are DNA stretches of repeating sequences of 1,2 or more nucleotides, the DNA building blocks
(13) parents, siblings, and offspring
(14) aroutine technique in molecular biology to amplify a piece of DNA across several orders of magnitude
(15) Mendelian inheritance is a set of principles relating to the transmission of hereditary characteristics from parents to offspring which were derived from the work of Gregor Mendel
(16) an inherited predisposition of persons to a large number of precancerous polypoid lesions in the colon and the rectum. Some of these lesions develop into cancer

Management, Treatment, and Follow-up

Endoscopic resection of a big adenoma with a broad base Endoscopic view after the resection

The past 2 decades have also witnessed substantial improvements in the treatment of colon and rectal cancers. Improved survival has been achieved with more effective combinations of surgery and chemotherapy. Liver metastases are a common complication of colorectal cancer, and they were once considered untreatable. Today, liver metastases in a large portion of patients are being effectively treated with surgical resection or nonsurgical methods like radiofrequency thermal ablation(1) . As far as chemotherapy is concerned, a new drug regimen based on 3 different drugs (5-FU/leucovorin/oxaliplatin) is providing better results than those achieved with combination of only 2 of these drugs. This is a clear example of the synergistic / additive effects displayed by many anti-cancer drugs. Combination regimens are not only proving to be more effective in terms of improved survival, they also reduce side effects and delay the onset of drug resistance(2) .

High-throughput methods based on comprehensive analysis of tumor cells’ genetic (DNA), transcriptional (mRNA), and translational (protein) profiles are highlighting the complexities of the molecular homeostasis of colorectal cancers, and it is becoming increasingly clear that effective, long-lasting control of these tumors can best be achieved by drug regimens that simultaneously target several cellular proteins and/or signaling pathways(3) that are crucial for tumor development and progression. One such process is angiogenesis(4) . Drugs that block this activity include bevacizumab, a monoclonal antibody(5) directed against vascular endothelial growth factor (VEGF). Other monoclonal antibodies (panitumumab, for example) target the epithelial growth factor (EGF) Receptor. One of the signaling pathways being explored as a potential therapeutic target is the one that causes the cells to undergo a form of programmed self-destruction known as apoptosis(6) .

One of the newest fields of research revolves around a tiny subpopulation of tumor cells that display phenotypic features suggestive of “stemness”(7 ). Like stem cells that reside in normal body tissues, these cells appear to play progenitor roles, that is, they can give rise to all the other tumor cell types. When these cells are injected into the skin of a laboratory mouse (for example), they give rise to a tumor that resembles the original one from which these cells were isolated. Likewise, if these cells are transported by the blood from the colon to the liver, they are likely to produce a metastatic lesion in that organ. By identifying the molecular characteristics of these cells and developing drugs that can act against them, researchers hope to strike a major blow against the growth and spread of colorectal cancers.

How and whether these different types of drugs might be combined to combat colon cancer will probably depend more and more on the pharmacogenetic(8) characteristics of the individual case, i.e., the genetic and epigenetic variations detectable in the tumor or in the host germ-line which might be associated with different sensitivity to a given drug. As mentioned before, colorectal tumors are characterized by biologically distinct types. Certain cellular processes may be more important in the growth of one type of tumor than in another; other characteristics (mismatch repair deficiency, for example) may be associated with early-onset drug resistance. The genetic background of the host can also influence how the tumor responds to certain drugs. For example, certain polymorphisms(9) of the gene encoding thymidylate synthase affect responses to the drug 5FU and the toxic effects it causes. Thus, the drug regimen prescribed for a given patient will be probably dictated by multiple factors, including the molecular phenotype of the tumor and the genetic background of the host. This “personalized” approach to cancer treatment will require an interdisciplinary approach with input from various fields of medicine.

Research is also being focused on the development of biomarkers(10) that are prognostic(11) rather than diagnostic (see section on Prevention). These markers (serum proteins for the most part) will be useful for fine-tuning the treatment regimen and follow-up. Other groups are working on the development of biomarkers that will allow early identification (and treatment) of disease recurrence.

(1) the use of electrodes to generate heat and destroy abnormal tissue
(2) the ability of tumor cells to withstand a drug to which they were once sensitive
(3) a series of chemical reactions occurring within a cell in response to a stimulus, such as the binding of a growth factor to the cells
(4) the process of developing new blood vessels that support tumor growth
(5) an antibody produced by a single clone of cells and therefore a single, homogeneous type of antibody
(6) a form of cell death in which a programmed sequence of events leads to the elimination of cells
(7) an essential characteristic of stem cells which distinguish them from the other cells of a tissue or organism
(8) the study of genetic or epigenetic variations that gives rise to differing response to drugs
(9) a variation in the DNA that is too common in the population to be considered a new, detrimental mutation. However, some polymorphisms may be associated with predisposition to a given disease or adverse drug reaction
(10) a biochemical feature that can be used to measure the progress of disease or the effect of treatment
(11) relating to the prognosis of a disease

Pictures : Dr. Federico Buffoli, Chief of the Endoscopy and Gastroenterology Unit, Istituti Ospitalieri di Cremona (Italy)