39-15: Colorectal Cancer

Colorectal cancer is the second leading cause of death due to malignancy in the United States. Colorectal cancer will develop in approximately 4.2% of Americans and has a 5-year survival rate of 65%. In 2019, there were an estimated 145,600 new cases of colorectal cancer in the United States, with an estimated 51,020 deaths. Between 1996 and 2010, its mortality rate decreased by 46%. A 2015 National Health Interview Survey estimated that 62% of US adults have undergone recommended screening. On average, new cases have been falling 3.2% each year over the last 10 years.

Colorectal cancers are almost all adenocarcinomas, which tend to form bulky exophytic masses or annular constricting lesions(eFigure 39–7). The majority of colorectal cancers are thought to arise from malignant transformation of an adenomatous polyp (tubular, tubulovillous, or villous adenoma) or serrated polyp (hyperplastic polyp, traditional serrated adenoma, or sessile serrated adenoma). Polyps that are “advanced” (ie, polyps at least 1 cm in size, adenomas with villous features or high-grade dysplasia, or serrated polyps with dysplasia) are associated with a greater risk of cancer. Approximately 85% of sporadic colorectal cancers arise from adenomatous polyps. They have loss of function of one or more tumor suppressor genes (eg, p53, APC, or DCC) due to a combination of spontaneous mutation of one allele combined with chromosomal instability and aneuploidy (abnormal DNA content) that leads to deletion and loss of heterozygosity of the other allele (eg, 5q, 17q, or 18p deletion). Activation of oncogenes such as KRAS and BRAF is present in a subset of colorectal cancers with prognostic and therapeutic implications discussed further below.

Approximately 10–20% of colorectal cancers arise from serrated polyps, most of which have hypermethylation of CpG-rich promoter regions that leads to inactivation of the DNA mismatch repair gene MLH1, resulting in microsatellite instability, and activation of mutations of the BRAF gene. Serrated colon cancers have distinct clinical and pathologic characteristics, including diploid DNA content, predominance in the proximal colon, poor differentiation, and more favorable prognosis.

Up to 5% of colorectal cancers are caused by inherited germline mutations resulting in polyposis syndromes (eg, familial adenomatous polyposis) or hereditary nonpolyposis colorectal cancer (HNPCC or Lynch syndrome). These conditions are discussed further in Chapter 15.

A number of factors increase the risk of developing colorectal cancer. Some of these factors include smoking, consumption of red and processed meats, alcohol intake, diabetes mellitus, physical inactivity, obesity, and history of inflammatory bowel disease. Recognition of these factors has had an impact on screening strategies. However, 75% of all cases occur in people with no known predisposing factors.

A. Age

The incidence of colorectal cancer rises sharply after age 45 years, and 90% of cases occur in persons over the age of 50 years. The median age at diagnosis is 68 for men and 72 for women. Over the past two decades there has been a 50% increase in incidence among adults under age 50, possibly due to increases in obesity. At present, however, the incidence in 45- to 49-year-olds (34/100,000) remains substantially lower than in 50- to 54-year-olds (60/100,000).

B. Family History of Neoplasia

A family history of colorectal cancer is present in approximately 20% of patients with colon cancer. Hereditary factors are believed to contribute to 20–30% of colorectal cancers; however, the genes responsible for most of these cases have not yet been identified. (See Chapter 15 for discussion of inherited polyposis syndromes.) Approximately 6% of the Ashkenazi Jewish population has a missense mutation in the APC gene (APC I1307K) that confers a modestly increased lifetime risk of developing colorectal cancer (odds ratio [OR] 1.4–1.9) but phenotypically resembles sporadic colorectal cancer rather than familial adenomatous polyposis. Genetic screening is available, and patients harboring the mutation merit more intensive colorectal screening.

A family history of colorectal cancer or adenomatous polyps is one of the most important risk factors for colorectal cancer. The risk of colon cancer is proportionate to the number and age of affected first-degree family members with colon neoplasia. People with one first-degree family member with colorectal cancer have an increased risk approximately two times that of the general population; however, the risk is almost four times if the family member was younger than 45 years when the cancer was diagnosed. Patients with two first-degree relatives have a fourfold increase, or 25–30% lifetime, risk of developing colon cancer. First-degree relatives of patients with adenomatous polyps also have a twofold increased risk for colorectal neoplasia, especially if they were younger than 60 years when the polyp was detected or if the polyp was 10 mm or larger.

C. Inflammatory Bowel Disease

The risk of adenocarcinoma of the colon begins to rise 8 years after disease onset in patients with ulcerative colitis and Crohn colitis (see Chapter 15). For this reason, initiation of surveillance with colonoscopy is recommended at 8–10 years after onset of inflammatory bowel disease symptoms.

D. Dietary and Lifestyle Factors and Chemoprevention in Colorectal Cancer Risk

In epidemiologic studies, diets rich in fats and red meat are associated with an increased risk of colorectal adenomas and cancer, whereas diets high in fruits, vegetables, and fiber are associated with a decreased risk. However, prospective studies have not shown a reduction in colon cancer or recurrence of adenomatous polyps with diets that are low in fat; are high in fiber, fruits or vegetables; or that include calcium, folate, beta-carotene, or vitamin A, C, D, or E supplements.

Meta-analyses suggest that individuals with increased physical activity are up to 27% less likely to develop colon cancer. There also is a correlation between increasing body mass index and cancer risk, such that for each increase of 5 kg/m² in BMI, there is a 5% increased cancer risk. Patients with higher levels of pre- and post-diagnosis physical activity experience reduced colorectal cancer–specific mortality and all-cause mortality. Vitamin D deficiency has been suggested to reduce colorectal cancer risk in some prospective studies. However, a randomized, double-blind, placebo-controlled trial did not find it beneficial in reducing colorectal adenoma recurrence within 3–5 years after removal of one or more adenomas. Neither the Institute of Medicine nor NCCN recommends screening for vitamin D deficiency or supplementing with vitamin D to reduce the risk of colorectal cancer occurrence, although there is emerging data on its role in patients who already have a diagnosis of colorectal cancer. Maintaining a healthy body weight, a healthy diet, and a physically active lifestyle are recommended in colorectal cancer survivors. Patients with stage III colon cancer who reported a healthy lifestyle during and after adjuvant chemotherapy had a 42% lower chance of death and a trend toward lower probability of recurrence than those patients who reported less healthy lifestyles.

Low-dose aspirin has been associated with a reduced risk of colorectal adenomas and cancer in multiple studies. A 2016 USPSTF systematic review of controlled trials concluded that prolonged regular use of low-dose aspirin (81 mg/day) is associated with a 40% reduction in colorectal cancer incidence after 10 years and a 33% reduction in colorectal cancer mortality after 20 years. Because long-term aspirin use is associated with a low incidence of serious complications (gastrointestinal hemorrhage, stroke), low-dose aspirin should not be routinely administered as a chemopreventive agent without other medical indications. Specifically, the Task Force recommends initiation of low-dose aspirin in patients who are age 50–59 years, who have a greater than 10% 10-year risk of cardiovascular disease events, who are not at increased risk for bleeding, who have a life expectancy of at least 10 years, and who are willing to take the low-dose aspirin daily for at least 10 years (grade B recommendation). The Task Force states that the decision to initiate low-dose aspirin use for the primary prevention of colorectal cancer (and cardiovascular disease) in adults who are age 60–69 years and who have a 10% or greater 10-year cardiovascular disease risk should be an individual one. Persons who are not at increased risk for bleeding, have a life expectancy of at least 10 years, and are willing to take low-dose aspirin daily for at least 10 years are more likely to benefit; persons who place a higher value on its potential benefits than its potential harms may choose to take low-dose aspirin (grade C recommendation). For adults younger than 50 years or older than 70, the Task Force states that the current evidence is insufficient to assess the balance of benefits and harms of aspirin use for the primary prevention of colorectal cancer (and cardiovascular disease) (grade I recommendation). Low-dose aspirin may also be considered in patients with a personal or family history of colorectal cancer or advanced adenomas; however, its administration does not obviate the need for colonoscopy screening and surveillance.

E. Other Factors

The overall incidence of colorectal cancer is similar in men and women; however, similar incidence rates are reached in women about 4–6 years later than in men. A higher proportion of cancers are located in the proximal colon in women (46%) than men (37%). The incidence and mortality of colon adenocarcinoma is higher in blacks and Native Americans than in whites. It is unclear whether this is due to genetic or socioeconomic factors (eg, diet or reduced access to medical care).

A. Symptoms and Signs

Adenocarcinomas grow slowly and may be present for several years before symptoms appear. However, some asymptomatic tumors may be detected by the presence of fecal occult blood (see Screening for Colorectal Neoplasms, below). Symptoms depend on the location of the carcinoma. Chronic blood loss from right-sided colonic cancers may cause iron deficiency anemia, manifested by fatigue and weakness. Obstruction, however, is uncommon because of the large diameter of the right colon(eFigure 39–8) and the liquid consistency of the fecal material. Lesions of the left colon often involve the bowel circumferentially. Because the left colon has a smaller diameter and the fecal matter is solid, obstructive symptoms may develop with colicky abdominal pain and a change in bowel habits. Constipation may alternate with periods of increased frequency and loose stools. The stool may be streaked with blood, though marked bleeding is unusual. With rectal cancers, patients note tenesmus, urgency, and recurrent hematochezia. Physical examination is usually normal except in advanced disease. The liver should be examined for hepatomegaly, suggesting metastatic spread. For cancers of the distal rectum, digital examination is necessary to determine whether there is extension into the anal sphincter or fixation, suggesting extension to the pelvic floor.

B. Laboratory Findings

A CBC should be obtained to look for anemia. Elevated liver biochemical tests, particularly the serum alkaline phosphatase, raise suspicion of metastatic disease. The serum carcinoembryonic antigen (CEA) should be measured in all patients with proven colorectal cancer but is not appropriate for screening. The CEA is not elevated in many patients with confirmed colorectal cancer; conversely, the CEA may be elevated in active smokers and those with a variety of other nonmalignant conditions. A preoperative CEA level greater than 5 ng/mL is a poor prognostic indicator. After complete surgical resection, CEA levels should normalize; persistently elevated levels suggest the presence of persistent disease and warrant further evaluation. The sensitivity of CEA for detection of recurrence ranges from 68% to 82% and its specificity from 80% to 97%. For this reason, CEA is routinely monitored at the time of adjuvant therapy and during postoperative surveillance for patients who had elevated levels before resection.

C. Colonoscopy

Colonoscopy is the required diagnostic procedure in patients with a clinical history suggestive of colon cancer or in patients with an abnormality suspicious for cancer detected on radiographic imaging. It should be considered in all nonmenstruating adults with new-onset iron deficiency anemia. Colonoscopy permits biopsy for pathologic confirmation of malignancy.

D. Imaging

Chest, abdominal, and pelvic CT scans with contrast are required for preoperative staging. CT scans may demonstrate distant metastases but are less accurate in the determination of the level of local tumor extension (T stage) or lymphatic spread (N stage). Intraoperative assessment of the liver by direct palpation and ultrasonography can be performed to detect hepatic metastases (M stage). For rectal cancers (generally defined as tumors arising 12 cm or less proximal to the anal verge), pelvic MRI or endorectal ultrasonography is required to determine the depth of penetration of the cancer through the rectal wall (T stage) and perirectal lymph nodes (N stage), informing decisions about preoperative (neoadjuvant) chemoradiotherapy and operative management. PET is not routinely used for staging or surveillance in colorectal cancers.

The TNM system is the commonly used classification to stage colorectal cancer. Staging is important not only because it correlates with the patient’s long-term survival but also because it is used to determine which patients should receive adjuvant or neoadjuvant therapy.

The nonspecific symptoms of colorectal cancer may be confused with those of irritable bowel syndrome, diverticular disease, ischemic colitis, inflammatory bowel disease, infectious colitis, and hemorrhoids. Neoplasm must be excluded in any patient over age 40 years who reports a change in bowel habits or hematochezia or who has an unexplained iron deficiency anemia or occult blood in stool samples.

A. Surgery

Resection of the primary colonic or rectal cancer is the treatment of choice for almost all patients who have resectable lesions and can tolerate general anesthesia. For colon cancer, multiple studies demonstrate that minimally invasive, laparoscopically assisted colectomy results in similar outcomes and rates of recurrence to open colectomy. Regional dissection of at least 12 lymph nodes should be performed to determine staging, which guides decisions about adjuvant therapy.

For rectal carcinoma, preoperative (neoadjuvant) chemoradiation with 5-fluorouracil is recommended in all node-positive tumors, and in T3 and greater tumors, due to increased risk of local recurrence. After neoadjuvant therapy, the operative approach depends on the level of the tumor above the anal verge, the size and depth of penetration, and the patient’s overall condition. Clinical staging by endorectal ultrasound or MRI with endorectal coil is important in guiding the treatment approach. In carefully selected patients with small, mobile (less than 4 cm), well-differentiated T1 rectal tumors that are less than 8 cm from the anal verge, transanal endoscopic or surgical excision may be considered. This approach avoids laparotomy and spares the rectum and anal sphincter, preserving normal bowel continence. All other patients will require either a low anterior resection with a colorectal anastomosis or an abdominoperineal resection with a colostomy, depending on how far above the anal verge the tumor is located and the extent of local tumor spread. Careful dissection of the entire mesorectum by either open or laparoscopic surgery reduces local recurrence to 5%. With improvements in surgical techniques, it is generally possible to perform low anterior resection provided there is a margin of at least 2 cm of normal tissue below the tumor. Although low anterior resections obviate a colostomy, they are associated with increased immediate postsurgical complications (eg, leak, dehiscence, stricture) and long-term defecatory complaints (eg, increased stool frequency, and incontinence). With unresectable rectal cancer, the patient may be palliated with a diverting colostomy. Pathology review of colorectal cancers should include testing for mismatch repair proteins for all patients. Tumors of patients with metastatic colorectal cancer should also be tested for extended RAS and BRAF mutations.

B. Systemic Chemotherapy for Colon Cancer

Chemotherapy has been demonstrated to improve overall and tumor-free survival in select patients with colon cancer depending on stage (Table 39–2).

1. Stage I

Because of the excellent 5-year survival rate (approximately 92%), no adjuvant therapy is recommended for stage I colon cancer. The 5-year survival rate of stage I rectal cancer is approximately 87%.

2. Stage II (node-negative disease)

The 5-year survival rate is approximately 87% for stage IIA disease and 63% for stage IIB disease. A significant survival benefit from adjuvant chemotherapy has not been demonstrated in most randomized clinical trials for stage II colon cancer (see discussion for stage III disease). However, otherwise healthy patients with stage II disease who are at higher risk for recurrence (perforation; obstruction; close or indeterminate margins; poorly differentiated histology; lymphatic, vascular, or perineural invasion; T4 tumors; or fewer than 12 lymph nodes sampled) may benefit from adjuvant chemotherapy. Patients whose tumors reveal microsatellite instability have a more favorable prognosis and do not benefit from 5-fluorouracil-based adjuvant therapy. In two independent validation studies, a twelve-gene recurrence score (Oncotype DX® Colon) has shown association with recurrence risk though it does not predict which patients will benefit from adjuvant chemotherapy.

3. Stage III (node-positive disease)

With surgical resection alone, the expected 5-year survival rate is 30–50%. Postoperative adjuvant chemotherapy significantly increases disease-free survival as well as overall survival by up to 30% and is recommended for all fit patients (Table 39–2). Multiple large, well-designed studies of adjuvant therapy for stage III colorectal cancer reported a higher rate of disease-free survival at 5 years for patients treated for 6 months postoperatively with a combination of oxaliplatin, 5-fluorouracil, and leucovorin (FOLFOX) (73.3%) than with 5-fluorouracil and leucovorin (FL) alone (67.4%). Similar benefit was reported for patients treated with oxaliplatin and capecitabine (orally active fluoropyrimidine). The benefit of adding oxaliplatin to 5-fluorouracil has not been demonstrated in patients who are 70 years of age or older. A large international randomized controlled trial comparing 3 months with 6 months of adjuvant therapy found that 3 months of adjuvant therapy resulted in equivalent disease-free survival for patients with earlier stage T1, T2, or T3, and N1 disease but not with more advanced T4 or N2 disease. Additionally, similar disease-free survival was seen with use of 3 months vs 6 months of capecitabine and oxaliplatin but not when infusional 5-fluorouracil plus oxaliplatin was used. The reason for the difference between regimens is currently unknown; the shorter treatment duration did result in significantly less neuropathy across regimens. For patients with high-risk disease (T4 or N2), 6 months of adjuvant chemotherapy are still recommended. The addition of a biologic agent (bevacizumab or cetuximab) to adjuvant chemotherapy does not improve outcomes.

4. Stage IV (metastatic disease)

Approximately 20% of patients have metastatic disease at the time of initial diagnosis, and an additional 30% eventually develop metastasis. A subset of these patients has limited disease that is potentially curable with surgical resection. Resection of isolated liver metastases may result in long-term (over 5 years) survival in 35–55% of cases. For those with unresectable hepatic metastases, local ablative techniques (cryosurgery, radiofrequency or microwave coagulation, embolization, hepatic intra-arterial chemotherapy) or radiation may provide long-term tumor control. A subset of patients who have isolated pulmonary metastases may undergo resection or radiation with potential cure. The majority of patients with metastatic disease do not have resectable (curable) disease. In the absence of other treatment, the median survival is less than 12 months; however, with current therapies, median survival approaches 30 months. Tumor location has been found to have a potential prognostic importance: median survival rates are 33.3 months for patients with left-sided colon cancers compared to 19.4 months for those with right-sided cancers.

The goals of therapy for patients with metastatic colorectal cancer are to slow tumor progression while maintaining a reasonable quality of life for as long as possible. Currently, either FOLFOX (the addition of oxaliplatin to 5-fluorouracil and folinic acid) or FOLFIRI (the addition of irinotecan to 5-fluorouracil and folinic acid) is the preferred first-line treatment regimen for fit patients. For convenience, oral capecitabine (instead of intravenous 5-fluorouracil and leucovorin) can be used in combination with oxaliplatin since it has similar efficacy to 5-fluorouracil; however, combination with irinotecan is not recommended due to increased toxicity (diarrhea). Infusional 5-fluorouracil in combination with both oxaliplatin and irinotecan (FOLFOXIRI) has been shown to achieve higher response rates than FOLFIRI alone but is associated with increased toxicity. Addition of a biologic agent to combination chemotherapy improves response rates and overall survival and is recommended in the first-line of treatment in suitable patients. Bevacizumab is a monoclonal antibody targeting VEGF. Combination therapy with bevacizumab and FOLFOX or FOLFIRI prolongs mean survival by 2–5 months compared with either regimen alone. Cetuximab and panitumumab are monoclonal antibodies targeting EGFR. Activating K-ras gene mutations downstream of EGFR are present in approximately 35% of patients with metastatic colorectal cancer and are a biomarker for nonresponse to cetuximab and panitumumab, for which reason the use of these agents is restricted to patients with tumors wild-type for K-ras. Mutations in N-ras and BRAF are also predictive of nonresponse to EGFR inhibition alone. In stage IV patients with K-ras wild-type tumors, the addition of panitumumab or cetuximab to FOLFOX or FOLFIRI prolongs survival by approximately 4 months. Bevacizumab may cause serious side effects, including arterial thromboembolic events, bowel perforation, or serious bleeding, in up to 5% of patients. EGFR-targeted agents cause an acneiform rash in the majority of patients. Cetuximab is associated with severe infusion reactions in approximately 2–5%.

When disease progresses despite treatment either with FOLFOX or with FOLFIRI (often in conjunction with bevacizumab or an EGFR-targeted antibody), therapy is switched to the alternate regimen. Although the ideal sequence of agents is not known, patients do benefit from exposure to all available therapies. The novel anti-angiogenic agent, aflibercept, is FDA approved for second-line treatment of colorectal cancer in combination with FOLFIRI. Continuation of bevacizumab with second-line chemotherapy after progression on a bevacizumab-containing first-line regimen also improved outcomes compared to discontinuation of bevacizumab in the randomized, phase III ML18147 trial. Palliative therapy with cetuximab or panitumumab alone or in combination with irinotecan can benefit patients with RAS wild-type tumors whose disease has progressed after first-line and second-line chemotherapies. Both trifluridine/tipiracil and the multi-targeted kinase inhibitor regorafenib are FDA approved for patients with metastatic, refractory colorectal cancer after progression on standard regimens. Both agents offer limited improvement in overall survival and are associated with considerable toxicity. In the phase III RAISE trial, FOLFIRI combined with ramucirumab, a monoclonal antibody that inhibits binding of VEGF ligand to VEGF receptor-2, showed improved overall and progression-free survival compared to FOLFIRI alone in the second-line treatment of patients with metastatic colorectal cancer. Clinical trial participation should be considered for eligible patients who are intolerant of or ineligible for standard therapies or in whom disease has progressed.

The addition of an anti-EGFR (in K-ras/N-ras wild-type patients) or an anti-VEGF agent to first-line chemotherapy showed similar outcomes in the CALGB 80405 trial. However, the FIRE-3 trial showed better outcomes with FOLFIRI combined with cetuximab versus FOLFIRI combined with bevacizumab. This discrepancy is thought to be due to the differing number of patients with right-sided primary tumors between the two studies; right-sided primary tumors are much less likely to benefit from anti-EGFR therapy. The simultaneous combination of both anti-EGFR and anti-VEGF agents with chemotherapy has been associated with worse outcomes in multiple studies. This sobering finding demonstrates the complexity of the signaling pathways involved. Of note, it is unknown whether continuation of bevacizumab versus switching to aflibercept or an anti-EGFR agent (in K-ras wild-type patients) with second-line chemotherapy will achieve superior outcomes.

Patients with metastatic colorectal cancer whose tumors harbor a mutation in the BRAF oncogene (approximately 8% in metastatic disease) have a significantly worse prognosis than patients whose tumors are wild-type for BRAF. Unlike BRAF mutations in melanoma, significant responses to single agent BRAF inhibition are not typically seen in colorectal cancer. However, triplet combinations of a BRAF inhibitor, an anti-EGFR antibody, and a MEK inhibitor have shown encouraging activity and are now recommended by NCCN Guidelines. However, these responses are typically not long-lasting and patients should be referred for clinical trial participation, whenever feasible.

There are 3.5–6.5% of stage IV colorectal cancers that are dMMR or MSI-H. Tumors that are dMMR or MSI-H are susceptible to immune checkpoint blockade and have improved outcomes with the use of immunotherapy. Both pembrolizumab and nivolumab have been approved for the treatment of MSI-H or dMMR colorectal cancer that has progressed through standard lines of chemotherapy. Additionally, both TRK-fusions and HER2-amplification can be seen in metastatic colorectal cancers and such tumors have been shown to be responsive to targeted therapies.

Because the list of potentially actionable mutations continues to grow, extended molecular testing with a next generation sequencing panel should be considered in all new diagnoses of metastatic colorectal cancer.

C. Neoadjuvant and Adjuvant Therapy for Rectal Cancer

Compared with colon cancer, rectal cancer has lower long-term survival rates and significantly higher rates of local tumor recurrence with surgery alone (approximately 25%) attributed in part to the difficulty of achieving adequate surgical resection margins and to lack of serosal encasement of the rectum. When initial imaging studies suggest stage I disease, surgery may be performed first (see Surgery above). For stage II and III tumors by clinical staging (including endorectal ultrasound or pelvic MRI), preoperative chemoradiation with 5-fluorouracil or capecitabine as a radiation-sensitizing agent improves disease-free survival and decreases pelvic recurrence rate. Preoperative chemoradiation has been shown to be superior to postoperative chemoradiation with lower local recurrence and toxicity rates in the randomized, phase III CAO/ARO/AIO-94 trial. Following surgical resection with total mesorectal excision (see Surgery above), adjuvant 5-fluorouracil–based therapy (generally with the FOLFOX regimen extrapolating from its benefit in patients with similarly staged colon cancers) is recommended for a total of approximately 6 months of perioperative therapy inclusive of neoadjuvant chemoradiation. Due to its better tolerability and potentially decreased risk of distant metastatic spread, the majority (or even the entirety) of the adjuvant FOLFOX chemotherapy can be given up-front prior to chemoradiation as “total neoadjuvant therapy,” per NCCN guidelines. Many institutions are shifting to this approach. A subset of patients who receive total neoadjuvant therapy will be found to have a complete response on follow-up imaging and endoscopy. Whether or not these patients need to undergo surgical resection is not currently known; however, it is the subject of multiple ongoing studies. When patients elect for a nonsurgical approach they need to be monitored very closely with physical examination, cross sectional imaging, and endoscopic surveillance. Preliminary data suggests that when this is done, recurrences can be detected early and salvage surgery with curative intent is often feasible.

Colon cancer patients who have undergone resections for cure are monitored closely to look for evidence of symptomatic or asymptomatic tumor recurrence that may occasionally be amenable to curative resection. Patients should be evaluated every 3–6 months for 2 years and then every 6 months for a total of 5 years with history, physical examination, and laboratory surveillance, including serum CEA levels if baseline levels are elevated. The NCCN and ASCO guidelines recommend surveillance contrast CT scans of chest, abdomen, and pelvis up to every 6–12 months for up to 5 years post-resection in high-risk stage II and all stage III patients. Patients who had a complete preoperative colonoscopy should undergo another colonoscopy 1 year after surgical resection. Patients who did not undergo full colonoscopy preoperatively also should undergo a full colonoscopy after completion of all adjuvant therapy to exclude other synchronous colorectal neoplasms. If a colonoscopy does not detect new adenomatous polyps 1 year postoperatively, surveillance colonoscopy should be performed every 3–5 years thereafter to look for metachronous polyps or cancer. In patients with rectal cancer, because of the high incidence of local tumor recurrence, proctoscopy surveillance of the low anterior resection anastomotic site may also be performed periodically. New onset of symptoms or a rising CEA warrants investigation with chest, abdominal, and pelvic CT and colonoscopy to look for a new primary tumor or recurrence, or metachronous metastatic disease that may be amenable to curative or palliative therapy. The majority of colorectal cancer recurrences occur within 3 years of the conclusion of treatment, and almost all (greater than 90%) occur within 5 years. For patients with a rising CEA level with unrevealing CT imaging, a PET scan may be more sensitive for the detection of occult metastatic disease.

The stage of disease at presentation remains the most important determinant of 5-year survival in colon cancer, which is estimated in older registries as: stage I, greater than 90%; stage II, 70–85%; stage III with fewer than 4 positive lymph nodes, 67%; stage III with more than 4 positive lymph nodes, 33%; and stage IV, 5–7%. Long-term registry follow-up data from the modern chemotherapy era is not yet available. For each stage, rectal cancers have a worse prognosis. For those patients whose disease progresses despite therapy, meticulous efforts at palliative care are essential (see Chapter 5).

Colorectal cancer is ideal for screening because it is a common disease that is fatal in almost 50% of cases and yet is curable if detected at an earlier stage. Furthermore, most cases arise from benign adenomatous or serrated polyps that progress over many years to cancer, and removal of these polyps has been shown to prevent the majority of cancers. Colorectal cancer screening is endorsed by the USPSTF, the Agency for Health Care Policy and Research, the American Cancer Society, and every professional gastroenterology and colorectal surgery society. Although there is continued debate about the optimal cost-effective means of providing population screening, there is unanimous consent that screening of some kind should be offered to every patient over the age of 50 years. Several analyses suggest that screening is cost-effective. It is important for primary care providers to understand the relative merits of various options and to discuss them with their patients.

Less than 65% of adults over age 50 are up-to-date with colorectal screening. Discussion and encouragement by the primary care provider are the most important factors in achieving patient compliance with screening programs. A 2018 Kaiser Permanente study reported increased participation to 82.5% through an organized screening program. Currently in the United States, the two most widely used strategies are colonoscopy every 10 years or fecal immunochemical testing (FIT) annually.

Recommendations for screening from the 2016 USPSTF and 2008 US Multi-Society Task Force (USMSTF) are listed in Table 39–6. Screening is recommended for all men and women 50 through 75 years of age who are at average risk for cancer. Many guidelines recommend screening for African Americans and Alaska Natives beginning at age 45. In 2018, the American Cancer Society provided a qualified recommendation for starting screening at age 45 for average-risk adults; however, other agencies do not feel that available evidence yet supports this recommendation. The potential for harm from screening must be weighed against the likelihood of benefit, especially in elderly patients with comorbid illnesses and shorter life expectancy. Although routine screening is not recommended in adults above age 75, it may be considered on a case-by-case basis in adults age 76 through 85 years who have excellent health and functional status. In 2017, the USMSTF issued revised, simplified recommendations in which screening tests were placed into three tiers (Table 39–7). For average-risk patients, colonoscopy every 10 years or FIT testing is preferred (Tier 1). Tier 2 tests (CT colonography every 5 years, flexible sigmoidoscopy every 5 years, or fecal FIT-fecal DNA testing every 3 years) and Tier 3 tests (colon capsule every 5 years) may be offered but are considered less suitable due to various disadvantages (discussed below).

Patients with first-degree relatives with colorectal neoplasms (cancer or adenomatous polyps) are at increased risk. Therefore, most guidelines recommend initiating screening at age 40–50 years (or 10 years younger than the familial diagnosis) in individuals with first-degree relatives with colorectal cancer or with advanced adenomas. Recommendations for screening in families with inherited cancer syndromes or inflammatory bowel disease are provided in Chapter 15. For patients at average risk for colorectal cancer, the recommendations of the USPSTF and USMSTF are discussed below.

Screening tests may be classified into two broad categories: stool-based tests and examinations that visualize the structure of the colon by direct endoscopic inspection or radiographic imaging.

A. Stool-Based Tests

1. Fecal occult blood test

Most colorectal cancers and some large adenomas result in increased chronic blood loss. A variety of tests for fecal occult blood are commercially available that have varying sensitivities and specificities for colorectal neoplasia. These include guaiac-based fecal occult blood tests (gFOBT) (eg, Hemoccult I and II and Hemoccult SENSA) that detect the pseudoperoxidase activity of heme or hemoglobin and FITs that detect human globin. In clinical trials, FITs have proven superior to gFOBT in sensitivity for detection of colorectal cancer and advanced adenomas with similar specificity. Because FITs are not affected by diet or medications and have superior accuracy, the USMSTF now recommends their use instead of gFOBT.

When gFOBT is administered to the general population as part of a screening program, 2–6% of tests are positive. Of those with positive tests, 5–18% have colorectal cancer that is more likely to be at an earlier stage (stage I or II). For optimal detection, annual testing is required. In a meta-analysis of four large, prospective, longitudinal studies, annual or biennial screening with Hemoccult I or Hemoccult II reduced mortality from colorectal cancer by 25% among those who were compliant with regular testing.

Several FIT kits are commercially available. These tests are highly specific for detecting human globin and therefore eliminate the need for pretest dietary restrictions. In 19 clinical studies, the pooled sensitivity and specificity of FIT for colorectal cancer in average-risk patients were 79% and 94%, respectively. The optimal interval (yearly or every 2 years) and number of stool samples (one or two) required for optimal FIT testing is as yet undetermined. Three randomized controlled trials comparing one-time colonoscopy with FIT testing for colorectal cancer screening are ongoing. Currently, annual testing of a single stool sample is recommended.

FIT testing is the preferred option for population-based screening in various European and Australian programs. In the United States, it is offered as the preferred option by many healthcare plans. For healthcare systems in which screening colonoscopy is readily available, FIT is a suitable option for patients seeking a noninvasive screening test who are willing to undertake annual fecal testing. Patients with a positive FIT test must undergo further evaluation with colonoscopy.

2. Multitarget DNA assay

Stool DNA tests measure a variety of mutated genes and methylated gene markers from exfoliated tumor cells. A newer-generation assay (Cologuard) combines a fecal DNA panel with a FIT. In a prospective comparative trial conducted in persons at average risk for colorectal cancer undergoing colonoscopy, the sensitivity for colorectal cancer for Cologuard was 92.3% vs 73.8% for FIT alone and the sensitivity for adenomas larger than 1 cm or serrated polyps for Cologuard was 42.4% vs 23.8% for FIT alone. A positive stool DNA test requires colonoscopy evaluation. Compared with FIT testing alone, FIT-fecal DNA testing has disadvantages including higher cost, lower specificity, lower cost-effectiveness, and cumbersome requirements for stool collection and mailing. Hence, the 2017 USMSTF relegated it to Tier 2, noting that it may be recommended in patients age 50–65 years who seek a noninvasive test with high-sensitivity, due to acceptable specificity in this age group.

B. Endoscopic Examinations of the Colon

1. Colonoscopy

Colonoscopy permits examination of the entire colon. In addition to detecting early cancers, colonoscopy allows removal of adenomatous polyps by biopsy or polypectomy, which is believed to reduce the risk of subsequent cancer. Over the past decade, there has been a dramatic increase in screening colonoscopy, with over 60% of US adults screened in the past 10 years. In asymptomatic individuals between 50 and 75 years of age undergoing screening colonoscopy, the prevalence of advanced colonic neoplasia is 4–11% and of colon cancer is 0.1–1%.

Although colonoscopy is believed to be the most sensitive test for detecting adenomas and cancer, it has several disadvantages. To allow adequate visualization of the entire colonic mucosa, it requires thorough bowel cleansing the evening and morning prior to the examination. To alleviate discomfort during the procedure, intravenous sedation is used for most patients, necessitating a companion to transport the patient home post-procedure. Serious complications occur uncommonly; they include perforation (0.1%), bleeding (0.25%), and death (2.9/100,000).

The skill of the operator has a major impact upon the quality of the colonoscopy examination. In several studies, the rate of colorectal cancer within 3 years of a screening colonoscopy was 0.7–0.9%, ie, approximately 1 in 110 patients. This may be attributable to polyps and early cancers that were overlooked during the colonoscopy. Studies of back-to-back colonoscopies confirm that endoscopists overlook 6–12% of polyps greater than 1 cm in size and up to 25% of smaller adenomas. Polyps that are small, flat, or located behind folds are easily missed, especially if the bowel preparation is poor. Population-based case-control and cohort studies suggest that colonoscopy is associated with greater reduction in colorectal cancer incidence and mortality in the distal colon (80%) than the proximal colon (40–60%). This may be attributable to incomplete examination of the proximal colon, and differences between the proximal and distal colon that include worse bowel preparation, suboptimal colonoscopic technique, and a higher prevalence of serrated polyps and flat adenomas. The latter are more common than previously recognized, are more likely to contain advanced pathology, and are more difficult to identify than raised (sessile or pedunculated) polyps. To optimize diagnostic accuracy as well as patient safety and comfort, colonoscopy should be performed after optimal bowel preparation by a well-trained endoscopist who spends sufficient time (at least 7 minutes) carefully examining the colon (especially the proximal colon) while withdrawing the endoscope.

2. Flexible sigmoidoscopy

Use of a 60-cm flexible sigmoidoscope permits visualization of the rectosigmoid and descending colon. It requires no sedation and, in many centers, it is performed by a nurse-specialist or physician assistant. Adenomatous polyps are identified in 10–20% and colorectal cancers in 1% of patients. The finding at sigmoidoscopy of an adenomatous polyp in the distal colon increases the likelihood at least twofold that an advanced neoplasm is present in the proximal colon. Therefore, patients with an adenomatous polyp of any size found during screening sigmoidoscopy should subsequently undergo colonoscopy to evaluate the proximal colon. A 2012 US study of almost 155,000 participants randomized to sigmoidoscopy screening or usual care between 1993 and 2001 reported a 50% reduction in distal colorectal cancer mortality in the sigmoidoscopy group after 11 years median follow-up.

The chief disadvantage of screening with flexible sigmoidoscopy is that it requires some bowel cleansing, it may be associated with some discomfort (since intravenous sedation is not used), and it does not examine the proximal colon. The prevalence of proximal versus distal neoplasia is higher in persons older than age 65 years, in African Americans, and in women. For these reasons, the 2017 USMSTF recommendations have placed flexible sigmoidoscopy among Tier 2 tests. If chosen, flexible sigmoidoscopy should be performed every 5–10 years.

C. Radiographic and Other Imaging of the Colon

1. CT colonography

Using helical CT with computer-assisted image reconstruction, two- and three-dimensional views can be generated of the colon lumen that simulate the view of colonoscopy (virtual colonoscopy). CT colonography requires a similar bowel cleansing regimen as colonoscopy as well as insufflation of air into the colon through a rectal tube, which may be associated with discomfort. Nonetheless, this examination is performed rapidly and requires no sedation or intravenous contrast. Several large studies have compared the accuracy of virtual colonoscopy with colonoscopy for colorectal screening. Using current imaging software with multidetector helical scanners, the sensitivity is greater than 95% for the detection of cancer and greater than 84–92% for the detection of polyps 10 mm or larger. CT colonography is less sensitive than colonoscopy for the detection of polyps smaller than 1 cm, flat adenomas, and serrated polyps.

Patients undergoing screening with CT colonography should be managed appropriately. If no polyps are found, the interval for repeat screening examination is uncertain; however, 5 years may be reasonable. All patients with polyps 10 mm or larger should be referred for colonoscopy with polypectomy because of the high prevalence (30%) of advanced pathology (cancer, high-grade dysplasia, or villous features) within these polyps. The optimal management of patients with polyps less than 10 mm in size is controversial. The USMSTF currently recommends that colonoscopy with polypectomy be offered to patients with one or more 6–9 mm polyps. Patients who refuse or who have increased risk of carcinoma should undergo surveillance CT colonography in 3–5 years. At the present time, there is no consensus on the management of patients with polyps smaller than 6 mm; however, some radiologists choose not even to report these findings.

The chief disadvantages of CT colonography are the need for a bowel preparation, limited availability in many health care systems, a possible increased risk of neoplasia due to radiation exposure, and the potential for finding incidental extracolonic findings that may lead to further evaluations. The 2017 USMSTF has therefore classified this as a Tier 2 screening option (Table 39–7). CT colonography is an excellent screening option in patients who do not wish to undergo or are unsuitable for colonoscopy and in patients in whom colonoscopy could not be completed.

2. Capsule colonoscopy

Imaging of the colon can be accomplished by oral ingestion of a capsule that captures video images of the colon. Compared with colonoscopy, the colon capsule has reduced sensitivity for polyps greater than 6 mm (64% vs 84%) and for colorectal cancers (74% vs 100%). At present, it is approved by the FDA for evaluation in patients who are not suitable candidates for colonoscopy or in whom colonoscopy could not evaluate the proximal colon. In addition to its suboptimal sensitivity for neoplasia, the main disadvantages of capsule colonoscopy are its cost, need for extensive bowel preparation, lack of reimbursement by most insurance carriers, and small risk of small bowel obstruction. For these reasons, the 2017 USMSTF recommendations classified it a Tier 3 screening test, suitable for highly selected patients (Table 39–7).

3. Barium enema—Double-contrast barium enema was previously used as a screening technique because it was widely available, relatively inexpensive, and safe. However, compared to CT colonography, it is more time-consuming and difficult to perform, less comfortable, and less accurate. It can no longer be recommended for routine screening.

• Patients with symptoms (change in bowel habits, hematochezia), signs (mass on abdominal examination or digital rectal examination), or laboratory tests (iron deficiency anemia) suggestive of colorectal neoplasia should be referred for colonoscopy.

• Patients with suspected colorectal cancer or adenomatous polyps of any size should be referred for colonoscopy.

• Virtually all patients with proven colorectal cancer should be referred to a surgeon for resection. Patients with clinical stage T3 or node-positive rectal tumors (or both) also should be referred to medical and radiation oncologists preoperatively for neoadjuvant therapy. Patients with stage II, III, or IV colorectal tumors should be referred to a medical oncologist.

• Patients with complications of colorectal cancer (obstruction, acute bleeding) requiring urgent evaluation and intervention.

• Patients with severe complications of chemotherapy.

• Patients with advanced metastatic disease requiring palliative care.