Colorectal cancer is the third most common type of cancer diagnosed in the United States and is the third most common cause of cancer-related death.
Colorectal cancer is the third most common type of cancer diagnosed in the United States and is the third most common cause of cancer-related death. The majority of cases are sporadic, with hereditary colon cancer contributing up to 15% of all colon cancer diagnoses. Treatment consists of surgery for early-stage disease and the combination of surgery and adjuvant chemotherapy for advanced-stage disease. Management of metastatic disease has evolved from primary chemotherapeutic treatment to include resection of single liver and lung metastases in addition to resection of the primary disease and chemotherapy.
In the United States, colorectal cancer (CRC) is the third most common type of cancer diagnosed and the third most common cause of cancer-related death in men and women. In 2010, an estimated 102,900 new cases of colon cancer were diagnosed (49,470 male, 53,430 female) and 51,370 patients (26,580 male, 24,790 female) died from CRC.1 The death rate from colon cancer decreased over the preceding decade, from 30.77 per 100,000 people to 20.51 per 100,000 people.1 The lifetime risk of developing colon cancer in industrialized nations is 5% and is stable or decreasing.2 In contrast, the incidence in developing countries continues to rise, hypothesized to be due to increased exposure to risk factors.3 It has been estimated that 1.5 million people in the United States will be living with CRC by 2020.4 The financial burden of caring for this population is significant: $4.5 to $9.6 billion per year.5 Disparities among different racial and ethnic groups are well known and attributed to a variety of factors, including environmental exposures, tumor biology, lack of access to preventive health services, and inequities in treatment.6 Both incidence and mortality rates are higher in black patients than in white patients.7
Hormonal factors: nulliparity, early menopause
Diet (rich in meat, poor in fiber)
Occupational exposure (asbestos)
High alcohol intake
Personal history of sporadic tumors
History of colorectal polyps
History of colorectal cancer
The majority of CRC cases are sporadic, with hereditary syndromes contributing approximately 5% to 15% of the incidence. There are many known risk factors for sporadic CRC, including nonmodifiable and modifiable variables (Table 1). Preventive measures should be targeted at tobacco use, dietary habits, and weight control.8 The inflammatory bowel disease (IBD) population is the second major category of patients at increased risk of CRC.9
The 2 main syndromes accounting for 5% of the inherited cases are hereditary nonpolyposis colon cancer (HNPCC) and familial adenomatous polyposis (FAP)10 The remaining 15% to 20% of inherited CRC cases do not have a clearly defined mechanism and likely result from a combination of gene polymorphisms, alterations in multiple susceptibility loci, and environmental influences.11 These cases are grouped into familial X CRC syndrome.
The prevalence of HNPCC, also known as Lynch syndrome, is estimated to be 2% to 5%. The syndrome is caused by a germline mutation in 1 of 6 currently identified DNA mismatch repair (MMR) genes: hMSH2, hMLH1, hPMS1, hPMS2, hMSH3, and hMSH6. Inactivation of these genes leads to the development of short repeats of DNA, known as microsatellites; 90% of the mutations in the MMR genes are found specifically in hMSH2 and hMLH1.11 Patients with HNPCC have an 80% lifetime risk of developing CRC.11,12 HNPCC is differentiated from sporadic colon cancer by a distinctive clinical picture. The average age of cancer diagnosis is much earlier (ie, 47 y vs 63 y), and there is a pattern of both metachronous and synchronous colon cancers, in addition to a high association with other primary tumors (eg, endometrial, ovarian, gastric, small bowel).11,13
FAP was first identified in 1861 and is the best characterized of the hereditary CRC syndromes, although it contributes only 1% to the inherited CRC syndromes.14 The disease is the result of a mutation of the APC gene (tumor suppressor gene), which is located on the long arm of chromosome 5q21. Classic FAP is characterized by autosomal dominant inheritance and an aggressive clinical course. Penetration is nearly 100%, although there are several variants with different clinical syndromes.11 Patients develop numerous adenomas throughout the colon and rectum at an early age, typically in their teens. Early polyps are classically located in the distal colon, so it is recommended that the onset of screening with flexible sigmoidoscopy begin at age 10 to 12 years.15
Once the colon can no longer be adequately surveyed due to the degree of polyposis, prophylactic colectomy is advised. In classic FAP, the most common age of cancer diagnosis is 39, with malignant degeneration occurring by age 40 to 50.10,16 The clinical features of FAP include duodenal, small bowel, and gastric adenomas, requiring annual endoscopy for the surveillance of FAP patients. Approximately 95% of FAP patients will have duodenal polyps identified on upper endoscopy; however, only 5% to 10% of patients go on to develop periampullary cancers.16,17 Currently, duodenal and desmoid tumors represent the most common causes of death in the FAP population.15 Attenuated FAP has a less aggressive clinical course compared to classic FAP; 90% of patients do not have an identifiable APC mutation and develop fewer than 100 polyps.
CRC has long been associated with ulcerative colitis (UC), and more recently also associated with Crohn’s disease. It is estimated that 1% of the CRC cases seen in the general population occur in UC patients.18 Duration of disease has long been considered to be a factor in the development of CRC in IBD. In 2001, a meta-analysis combining data from 116 studies found cumulative probabilities of developing CRC in UC patients of 2% by age 10, 8% by age 20, and 18% by age 30.18 Comparison of the CRC risk in Crohn’s disease and UC populations has demonstrated similar risk profiles, with a similar incidence rate and similar rate ratios.19
with FOBT or
FOBT annually or
every 10 y
Flexible sigmoidoscopy every 5 y with FOBT every 3 y or
Stool DNA, uncertain interval, or
Alternative: flexible sigmoidoscopy every 5 y
Screening colonoscopy every 10 y
Flexible sigmoidoscopy every 5 y or
Alternative: CT colonography every 5 y
Colonoscopy every 10 y or
Alternative: annual FOBT
Double-contrast barium enema every 5 y or
Alternative: fecal DNA every 3 y
CT colonography every 5 y
ACS indiates American Cancer Society; ACG, American College of Gastroenterology; CT, computed tomography; DNA, deoxyribonucleic acid; FOBT, fecal occult blood test; USPSTF; US Preventive Services Task Force; y, years.
Initiation of colon cancer screening in the average-risk patient is indicated at age 50; however, current screening guidelines do not clearly define the optimal modality to perform screening (Table 2).3,20-22 Sporadic colon cancer is believed to develop from benign lesions that deteriorate into carcinoma over a period of time, thus providing a window for early detection and treatment with the goal of lowering mortality.22 Between 1975 and 2000, the incidence of colon cancer decreased by 22%, with half of that volume attributed to screening and half to risk-factor modification and improved treatment.3 The screening methods for CRC are differentiated between detection and prevention. Fecal occult blood testing (FOBT) and stool DNA testing are methods that detect malignant disease, whereas computed tomography (CT) colonography, sigmoidoscopy, and colonoscopy can detect premalignant lesions.
FOBT is an inexpensive and safe test that has been used for many years and has been shown in 4 major, long-term, randomized, controlled trials to reduce mortality (Table 3).23-26 In a recently published UK study, sigmoidoscopy has also been shown to reduce mortality.27 Results from 2 major ongoing, randomized, controlled trials—the US-based Prostate, Lung, Colorectal, and Ovarian (PLCO) cancer screening trial and the Italian once-only sigmoidoscopy SCORE trial—are awaited.28,29
Colonoscopy has traditionally been used specifically to further evaluate patients with a positive finding on FOBT or sigmoidoscopy. However, it is increasingly being used as a primary screening tool, predominantly in the United States, where it has become the principal screening modality, despite the lack of level 1 evidence substantiating an increase in early cancer diagnosis or decrease in cancer-related mortality.2,20 Ongoing, randomized, controlled trials are evaluating colonoscopy as a primary screening tool (Table 3).
Click to enlarge
Newer, less invasive testing modalities such as CT colonography and stool DNA testing are 2 screening options.30 Current stool DNA testing has demonstrated higher sensitivity and specificity than FOBT and is not limited by lesion location; however, it is not clear how to best integrate this test into the catalog of screening tests.31 Although the American College of Gastroenterology21 and the American Cancer Society22 have incorporated CT colonography into their screening guidelines (Table 2), the US Preventive Services Task Force has yet to follow suit, choosing to wait for further data prior to making their recommendations.32
Sporadic CRC is postulated to follow the adenoma—carcinoma sequence, precipitated by cumulative genetic mutations.33-35 In 1993, the landmark National Polyp Study Workgroup demonstrated that polypectomy dramatically decreased the incidence of CRC in a cohort of 1418 patients followed for 8401 person-years.36 Point mutations, altered DNA methylation, gene rearrangements, amplifications, and deletions comprised the most common mutational events that led to 3 described pathways leading to tumorigenesis: (1) gain of function (oncogene activation); (2) loss of function (tumor suppressors/apoptotic pathways); and (3) epigenetic alterations (DNA methylation patterns).35,37
CRC is diagnosed either after routine screening or prompted by the onset of new symptoms. Symptoms in CRC are nonspecific and vague, and may include a change in bowel habits, weight loss, abdominal pain, and fatigue. Rarely, more specific symptoms such as obstruction, bleeding, or perforation may occur, prompting an urgent surgery.38 The goal of preoperative imaging is to accurately stage patients. This usually entails, at minimum, a CT scan of the abdomen and pelvis. Chest imaging to evaluate for pulmonary metastases may consist of either plain films or CT scan, depending on physician preference and regional standards.34 Positron emission tomography (PET) has excellent sensitivity and specificity in detecting metastatic disease in CRC, but limited sensitivity for lymph node involvement, although significantly higher than CT.39,40 The combined PET/CT is an emerging test that combines into 1 study the anatomical data from the standard CT scan with the functional data obtained from PET. It has been suggested that PET/CT has improved sensitivity for identification of extrahepatic metastases.41,42 However, there is currently no algorithm for preoperative imaging evaluation of CRC.
Primary tumor (T)
Tumor cannot be assessed
Carcinoma in situ
Tumor invades submucosa
Tumor invades muscularis propria
Tumor invades subserosa
Tumor directly invades adjacent organs/structures or through the visceral peritoneum
Regional lymph nodes (N)
Lymph nodes cannot be assessed
No lymph node metastases
Metastases in 1-3 lymph nodes
Metastases in ≥4 lymph nodes
Distant metastases (M)
Metastases cannot be assessed
No distant metastases
Distant metastases present
5-year survival (%)
T1, T2, N1
T3, T4, N1, N2
Any T, M1
The critical component that determines prognosis in colon cancer remains the pathologic stage (Table 4).43 The variation in survival between early- and late-stage colon cancer underscores the importance of screening and early diagnosis. One well-known biomarker, carcinoembryonic antigen (CEA), is traditionally used postoperatively to monitor for recurrence. A retrospective study evaluating survival of 506 patients following surgery demonstrated a significant association between survival and liver metastases, age, lymph node involvement, T stage, and postoperative CEA levels.44 It has been suggested that preoperative CEA be incorporated into the TNM staging system for CRC.45
Pharmacogenomics is an emerging field focused on studying drug efficacy in relation to specific genetic polymorphisms seen in various cancers. Understanding tumor biology and predicting response to therapy will allow physicians to create ideal therapies that maximize efficacy and minimize toxicity. The following biomarkers have recently gained notoriety for their use in predicting tumor aggressiveness and response to therapy: loss of heterozygosity of 18q, microsatellite instability, polymorphisms in thymidylate synthase (TS), IL-8 (interleukin-8), IL8R (interleukin-8 receptor), vascular endothelial growth factor (VEGF), epidermal growth factor (EGF), and KRAS mutations.46 Vascular endothelial growth factor receptor is the target of the drug bevacizumab and is correlated with risk of tumor recurrence and prognosis in CRC.47 EGF is overexpressed in many malignancies and is targeted by chimeric (cetuximab) and monoclonal (panitumumab) antibodies. KRAS is a guanosine triphosphate coupled protein responsible for receptor tyrosine signaling; the activation of specific pathways culminates in cell proliferation.33 Mutations are known to be associated with CRC development.48 Ideally, these markers would be used in conjunction with the TNM staging system to individualize treatment and improve outcomes.
Prognostic factors directly related to the technical aspects of surgery include the resection margin and the number of lymph nodes contained in the lymphadenectomy. It is accepted that the number of total lymph nodes retrieved does not reflect the quality of the surgical resection, although it could suggest the biological aggressiveness of the disease. Still, the significance of the total number of nodes retrieved in a colon cancer specimen is controversial. Risk factors for recurrence after surgery include poor differentiation, lymphovascular invasion, perineural invasion, T4 tumor stage, bowel perforation, clinical bowel obstruction, and elevated preoperative plasma CEA antigen levels.45 One study of 1320 patients derived a nomogram that predicts recurrence after surgery with a concordance index of 0.77.49 The variables included patient age, tumor location, preoperative CEA, T stage, N stage, lymphovascular invasion, perineural invasion, and use of adjuvant chemotherapy.
As laparoscopic surgery gains prominence in all surgical fields, its role in oncologic resections has been questioned. The Clinical Outcomes of Surgical Therapy (COST) study group trial randomized 872 patients and found that after 3 years, rates of recurrence were similar (18% vs 16%, P = .32) as well as overall survival (OS) at 3 years (85% vs 86%, P = .51).50 The Colon Cancer Laparoscopic or Open Resection study group (COLOR) similarly randomized 1248 patients to open or laparoscopic surgery and also reported comparable 3-year disease-free survival (DFS; 74% vs 76%, P = .7) and 3-year OS (81% vs 83%, P = .45) rates.51 The Cochrane review in 2008 assessed 33 randomized controlled trials (RCTs), 12 of which contained long-term data on outcomes following laparoscopic versus open surgery. Similar to the aforementioned studies, they reported no significant difference in recurrence rates between laparoscopic and open surgery.52 They concluded that laparoscopic colon resection for CRC is feasible and equivalent to open surgery in terms of oncologic outcomes.
Management of locally advanced disease has achieved major milestones in the last decade, reflected in longer progression-free survival (PFS) and OS rates. Adjuvant treatment for CRC is indicated for curative-intent resections of stage IV disease and for stage III disease. The utility of adjuvant therapy in stage II disease remains unclear and controversial. In the absence of evidence, the risks and benefits should be considered closely when offering treatment to this patient population. Patients who might be considered for such treatment are those who are considered high-risk (ie, with T4 tumors, obstructing presentation, poor differentiation, extramural venous invasion, fewer than 10 to 12 harvested lymph nodes, and indeterminate or positive resection margins). Major trials of adjuvant chemotherapy are summarized in Table 5.
Click to enlarge
Traditionally, the mainstay of chemotherapeutic treatment has been intravenous (IV) 5-fluorouracil (5-FU) in conjunction with leucovorin (LV). LV, also known as folinic acid, is known to improve tumor response rates (RRs) when combined with 5-FU, translating into longer DFS and OS.50,53 Starting in 2000, several new and exciting agents have been introduced into the battle against metastatic colon cancer. Capecitabine, an oral derivative of 5-FU, was introduced in 2005 after displaying equivalent efficacy compared to IV 5-FU.54 Irinotecan, an inhibitor of topoisomerase I, was introduced in 1996 and is currently approved as both primary therapy (ie, FOLFIRI [LV/5-FU/irinotecan]) and monotherapy after 5-FU failure.55-57 Investigators have begun evaluating the impact of substituting oral capecitabine for IV 5-FU (XELIRI [capecitabine/irinotecan] regimen). This regimen remains in phase II trials and has yet to be compared head to head with FOLFIRI.58
Oxaliplatin, a platinum derivative, was introduced in 2002. Oxaliplatin is not indicated for monotherapy, but has dramatically improved PFS and OS in combination with the FOLFOX (LV/5-FU/oxaliplatin) regimen, which is currently first-line treatment in the United States for advanced and metastatic colon cancer.59,60 In comparison to 5-FU/leucovorin, patients with stage IV disease treated with FOLFOX showed a significantly longer PFS (9 vs 6 mo, P = .001) and superior RR (50% vs 22%, P = .001).61 All chemotherapeutic agents have associated toxicity, notably oxaliplatin, which has a high incidence of sensory neuropathy that usually occurs after a cumulative dose of 640 mg/m2.62 The assessment of XELOX (capecitabine/oxaliplatin) versus FOLFOX has shown no significant difference between the 2 arms and concludes that XELOX may be indicated as first-line treatment in appropriate patients.63 A direct comparison of FOLFOX and FOLFIRI was mandated as the next chapter in the adjuvant treatment of colon cancer. Two large multicenter trials found no significant difference in PFS or OS between the 2 regimens.64 The main disparity was the toxicity profile, with FOLFOX affecting the nervous system and FOLFIRI affecting the gastrointestinal system.65
The introduction of bevacizumab, a monoclonal antibody against VEGF, generated considerable anticipation when Hurwitz and colleagues66 demonstrated significant improvement in PFS and OS in patients treated with IFL (irinotecan/bolus 5-FU/LV) after the addition of bevacizumab. Since then, the FOLFOX and XELOX combinations have also been shown to benefit from concurrent administration of bevacizumab.67 However, the recently published NSABP C-08 study showed no significant improvement in DFS in stage II or III patients treated with FOLFOX and bevacizumab.68 Two agents, cetuximab and panitumumab, currently target epidermal growth factor receptor. Cetuximab is approved for use in the United States and Europe for wild-type KRAS tumors. In 2009, the CRYSTAL and OPUS trials evaluated the efficacy of cetuximab administered in combination with FOLFIRI and FOLFOX, respectively, and found a significant improvement in PFS in the wild-type KRAS population.69,70 Panitumumab is currently used following chemotherapy in medically refractory wild-type KRAS tumors.71 It has a higher rate of toxicity documented in preliminary studies and, to date, has not been compared to cetuximab or bevacizumab.
The treatment of metastatic disease to the liver encompasses portal vein embolization, radiofrequency ablation (RFA), and liver resection in addition to chemotherapeutic treatment following curative resection of the primary disease. Of these options, the largest survival advantage is seen following liver resection (the 5-year survival is 25%-50%), which is the treatment of choice if the patient is an appropriate surgical candidate.72 The liver is the most common site of metastasis and recurrence in CRC, and up to 50% of patients with CRC will develop liver metastases.73 Patients with unresectable liver metastases have a 5-year survival of 6 to 18 months.73
The criteria for surgical resection of liver metastases have historically been strict, limiting surgical intervention to those patients who had 4 or fewer metastases, no involvement of portal lymph nodes or extrahepatic disease, and the ability to achieve 1-inch margins.74 New studies are pushing the boundaries of who can be a surgical candidate, and the pendulum has swung in favor of more aggressive surgical management. Although only a minority of patients with stage IV disease will fulfill the criteria delineated by the American Hepato-Pancreato-Biliary Association, Society of Surgical Oncology, and Society for Surgery of the Alimentary Tract for hepatic metastasectomy (ie, adequate postoperative functional reserve, complete resection of intra- and extrahepatic disease feasibility, and ability to perform complete resection with negative margins), surgical resection is the only modality that offers an opportunity for cure and should be pursued.74 Five-year survival rates for patients following hepatic metastasectomy range from 10% to 70% with a median survival of 30% compared with a 5-year survival of 0% in patients who do not undergo resection.75
In 2008, the European Organisation for Research and Treatment of Cancer (EORTC) study reported DFS in patients treated with and without perioperative chemotherapy (FOLFOX4) plus surgery in patients with resectable liver metastases. The study revealed a nonsignificant PFS increase of 7.3% in patients who received perioperative chemotherapy. Therefore, in the absence of level I evidence, the use of perioperative chemotherapy in patients with resectable metastatic colorectal liver metastases should be tailored to the needs of individual patients, accounting for side effects that result from its use.
Neoadjuvant chemotherapy and RFA are also used as conversion therapy. Conversion therapy can result in downstaging of unresectable liver metastases in 12% to 33% of patients, enabling an R0 resection.76 Masi and colleagues77 followed a cohort of 196 patients with unresectable metastatic colon cancer treated with neoadjuvant FOLFOXIRI (LV/5-FU/oxaliplatin/irinotecan), 19% of whom became eligible for surgical metastasectomy and report an impressive 5-year survival of 42%. Another prospective study comparing operative mortality, OS, and DFS between patients with resectable liver metastases and patients who were initially unresectable and became operative candidates after downsizing with neoadjuvant chemotherapy (FOLFOX4) found no significant difference in operative mortality or OS, but a shorter DFS (31% vs 58%, P = .04) in patients who were initially unresectable.73 FOLFOX4 resulted in a clinical response in 60% of patients in the North Central Cancer Treatment Group study, with an R0 resection rate of 33% and median DFS of 19 months.78
With the increasing use of neoadjuvant chemotherapy, a relationship between specific chemotherapeutic regimens and hepatic histopathologic change has been recognized. Wicherts and colleagues79 retrospectively studied 146 patients treated with neoadjuvant chemotherapy followed by surgery and 85 patients treated with surgery alone. They found an association between oxaliplatin, regenerative nodular hyperplasia, and postoperative hepatic morbidity, but no association with mortality. These results mirror what other studies have reported; this confirms that until additional data are available to delineate specific chemotherapeutic regimens that best combine with surgical resection, current standards should continue.
It is not clear where the new monoclonal antibodies (ie, bevacizumab and cetuximab) fit into the perioperative treatment algorithm, and definitive clinical data reports are awaited. The use of RFA in metastatic colon cancer is also not well-defined, as there are no RCTs examining ideal use, outcomes, and comparison to resection.80 Recently, the use of hyperthermic intraperitoneal chemotherapy (HIPEC) for peritoneal carcinomatosis has been introduced for CRC. To date, small series of patients have reported varying results, suggesting that there may be a benefit in select patients.81 HIPEC remains a last-resort treatment for patients who would otherwise have no alternatives.
After surgery with curative intent, 30% to 50% of patients will develop recurrence.82 The majority of recurrences occur within the first 5 years; thus, early and intensive follow-up is vital. A meta-analysis of 2923 patients found that patients undergoing intensive follow-up versus less-intensive follow-up identified surgically amenable recurrences earlier, allowing more patients to undergo reresection (10.7% vs 5.7%, P = .0002).82 Current recommendations from the National Comprehensive Cancer Network (NCCN) include a history and physical exam every 3 to 6 months for 2 years, then every 6 months for 5 years; CEA levels every 3 to 6 months for 2 years, then every 6 months for 5 years; and colonoscopy after 1 year, then at 3 and 5 years, then every 5 years, if normal. A CT scan should also be added every year for 3 years.83
The last decade has witnessed exciting new strategies for the diagnosis and treatment of colon cancer, enabling improved patient survival. With the advent of molecular modeling and new tools that predict recurrence, the future will bring more individualized treatment, which ideally will result in improved outcomes.
Sarah Popek, MD, is a fifth-year resident in the Department of Surgery at the University of Arizona in Tucson. Vassiliki Liana Tsikitis, MD, is the assistant professor of surgery in the Division of Surgical Oncology at the University of Arizona College of Medicine and University Physicians Healthcare in Tucson, Arizona.
The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.
Address correspondence to:
Vassiliki Liana Tsikitis, MD, Department of Surgery, Oregon Health & Sciences University, 3181 SW Sam Jackson Pk Rd, L223A, Portland, OR 97239. E-mail: firstname.lastname@example.org.