What every physician needs to know:

Colorectal adenocarcinoma (CRC) is the third most common malignancy worldwide, and the fourth most common in the US. In the US, incidence of colorectal cancer in 2017 is estimated to be approximately 135,410, with a total mortality of 50,260, accounting for nearly 9% of all cancer related deaths. However, in the US, the overall rates of incidence and mortality have been decreasing by 2-3% each year since the mid-1980s. These improvements are attributed to effective screening, risk factor modification and improved treatments. Recent epidemiological studies in contrast suggest that the incidence and mortality from CRC in adults younger than 54 have been slowly but steadily rising (1-3% annually depending on site and age group) over the past few decades due to unclear reasons.

The lifetime risk of colorectal cancer is 5%. The median age of diagnosis is 72 years, with 90% of all new CRC occurring in individuals older than 50 years. Furthermore, it is estimated that it takes 10 years for the development of frank carcinoma from the earliest precursor lesions. Hence current guidelines recommend colonoscopy, the preferred method of screening is every 10 years after the age of 50 years in average risk individuals. Although the screening rate among US adults aged 50-75 years has been increasing, it remained under 65% in 2010.

Familial syndromes with increased risk and early onset of CRC include familial adenomatous polyposis (FAP) in 1-2% of all cases (characterized by early development of thousands of colonic polyps) and hereditary non-polyposis colorectal cancer (HNPCC) in 4-6% of cases. Both of these conditions are autosomal dominant and are also characterized by several extra-colonic tumors. Patients with these syndromes, with other conditions associated with increased risk of CRC [such as history of inflammatory bowel disease (IBD)], or with strong family history require more aggressive screening in addition to genetic counseling.

Continue Reading

Complete surgical resection of the primary tumor along with removal of associated loco-regional lymph nodes is the cornerstone of curative treatment and is performed in all stages I to III CRC patients. Following resection, current guidelines recommend adjuvant chemotherapy for 6 months in all stage III and high risk stage II colon cancer patients. Metastatic CRC is typically treated with sequential, combination chemotherapy with palliative intent. However, a small proportion of patients with metastatic CRC may undergo potentially curative resection. The liver is the most common site of metastasis and several liver directed therapies in addition to specialized surgical techniques are available for the treatment of select patients.

Rectal cancers are anatomically defined as those arising within 12cm of the anal verge by rigid proctoscopy (or intraoperatively, as those arising below the level of peritoneal reflection). Surgeries on tumors arising in the lower two thirds of the rectum are limited by the bony pelvis, by other anatomic structures in close proximity, and by the absence of a serosa around the rectum.

This is especially important because prognosis after resection of rectal cancer is dependent on the ability to obtain negative radial and distal margins, and quality of life depends on the ability to preserve adequate sphincter function. Therefore, patients with potentially resectable rectal cancers are typically treated with neoadjuvant chemoradiation (chemoRT) to downstage the tumor to increase the chance of resection and sphincter preservation and to decrease local recurrence. This is followed by en bloc total mesorectal excision (TME). Although there is no definitive evidence, it is standard practice to pursue further adjuvant chemotherapy after neoadjuvant chemoradiation in rectal cancer to complete at least 6 months of total perioperative therapy.

Are you sure your patient has colorectal adenocarcinoma? What should you expect to find?

CRC may have diverse manifestations related to the primary tumor causing gastrointestinal signs/symptoms due to distant metastases, or constitutional symptoms.

Gastrointestinal signs and symptoms
  • Abdominal pain and/or discomfort which may be associated with nausea and vomiting.

  • Altered bowel movements or even bowel obstruction (with or without bowel perforation), more commonly with left sided tumors.

  • Gastrointestinal bleeding – ranging from occult blood loss to melena and/or hematochezia. Hematochezia is more often associated with rectal cancer while melena and occult blood loss are typically seen with more proximal tumors.

  • Features of metastatic disease and constitutional symptoms: CRC can spread by contiguous, lymphatic, hematogenous or peritoneal routes. Since venous drainage from most of the colon is to the liver through the portal system, this is the first and most common site of hematogenous spread, followed by lung, bone and occasionally the brain. Liver metastases may cause anorexia, RUQ pain and jaundice. In contrast, tumors of the distal rectum may spread through the inferior rectal vein and the inferior vena cava to the lungs first. Constitutional symptoms are nonspecific and include weight loss, fatigue, anemia and occasionally fever of unknown origin.

Atypical presentations
  • Local invasion occasionally may cause malignant fistulas into the bladder or small bowel as in cecal and sigmoid cancers. In addition, rectal cancers may spread to involve sciatic or obdurator nerves causing neuropathic pain.

  • Streptococcus bovis and clostridium septicum sepsis are characteristically associated with underlying undiagnosed CRC.

  • According to some series, adenocarcinoma of unknown primary is colorectal in origin in 6% of all cases in whom a primary is eventually identified.

  • Synchronous CRCs (2 or more distinct primary tumors) occur in 3-5% and metachronous CRCs (new tumors at least 6 months after initial diagnosis) are seen in 1.5–3% of patients within the first 5 years after resection of primary, and up to 9% after several decades.

Patients with symptoms at diagnosis appear to have worse prognosis (5-year survival 49% vs 71% in one report) possibly since symptomatic cancers are diagnosed at a later stage. Obstruction and/or perforation are associated with poorer prognosis irrespective of stage.

Should I screen patients for colorectal adenocarcinoma?

The goal of CRC screening is to reduce incidence, morbidity and mortality by early diagnosis and adenomatous polypectomy. Current guidelines recommend CRC risk assessment by age 40 years (in patients without known family history) to determine the age to initiate screening. Patients with average risk are advised to start screening at age 50, using one of several available options: colonoscopy every 10 years, annual fecal-based tests, flexible sigmoidoscopy every 5 years, a combination of annual fecal tests and sigmoidoscopy every 5 years or CT colonography every 5 years.


Colonoscopy is the screening test preferred by most experts, since polyps can be removed at the time of the procedure, and the full colon is examined. Findings from non-randomized studies suggest that colonoscopy decreases the incidence of CRC by over 50%. A large Canadian population study suggested a decrease in risk of death by 3% for every 1% increase in colonoscopy rate. It is felt that this decrease in mortality may be mostly from decreased left-sided than right-sided CRC. The general consensus is that a 10-year interval is appropriate for most individuals with average risk since studies suggest that a negative colonoscopy is associated with standardized incidence ratio of 0.28% (95% CI, 0.09-0.65) at 10 years compared to the general population.

Adenomatous polyps more than 1 cm in size and villous adenomatous polyps in particular are associated with increased risk of CRC in contrast to hyperplastic polyps that are typically not associated with increased risk (except in the rare hyperplastic polyposis syndrome). Sessile serrated adenomatous polyps and flat adenomas are rare and can be often missed; guidelines for management of these lesions are not established yet but in the interim should be managed as other adenomatous polyps.

Flexible sigmoidoscopy

A randomized screening study comparing flexible sigmoidoscopy done once between ages 55-64 years to no screening suggested a 33% reduction in incidence and 43% reduction in mortality for CRC. Another large, randomized clinical trial also confirmed these findings. Although flexible sigmoidoscopy requires no sedation and less bowel preparation, it is limited to the lower half of the colon. Patients with lesions larger than 1 cm should be referred to colonoscopy since they are associated with a risk of more proximal colonic neoplasms.

Double-contrast barium enema

This procedure is no longer widely used and is typically reserved for patients who cannot undergo colonoscopy.

Computed tomographic colonography (virtual colonoscopy)

The National CT Colonography Trial with over 2500 patients suggested a sensitivity of 90-100% and a specificity of 87-89% for polyps measuring at least 1 cm depending on the type of bowel preparation used for this technique. For polyps larger than 6 mm, the corresponding rates were 80-96% and 89-90% respectively. However, this is an evolving technique and several issues need to be considered: a) positive findings require subsequent colonoscopy; b) about 16% of studies reveal extracolonic findings of unclear significance that potentially require further tests; c) technical and procedural aspects have not been standardized yet; d) although the precise risk cannot be determined yet, there is radiation exposure during this procedure.

Fecal occult blood test

The guaiac fecal occult blood test (FOBT), the most common stool test used is based on the pseudoperoxidase activity of heme in human blood. The Hemoccult TM II SENSA is the most sensitive and widely used assay. A meta-analysis of over 300,000 patients from 4 randomized studies suggested a 16% reduction (95% CI, 10-22%) in relative risk for CRC mortality with guaiac testing. FOBT of a single specimen (e.g. from a rectal exam) is not recommended for screening due to low sensitivity. Instead, FOBT should be performed on three successive specimens while on a prescribed diet.

Fecal immunochemical test (FIT)

FIT directly detects globin and thus does not require dietary restrictions or multiple samples. However, sensitivities and specificities for both FOBT and FIT vary widely and both are less reliable than endoscopy.

Stool DNA test

This is an emerging tool that detects the presence of known DNA alterations associated with CRC carcinogenesis. The stool DNA test has not yet been approved by the FDA and its use is currently limited to the research setting.

A recent trial compared a multi-target stool DNA test (analysing for KRAS mutations, aberrant NDRG4 and BMP3 methylation, and β-actin) with FIT. This suggested that although the stool DNA test was more sensitive for detecting colorectal cancer as well as precancerous lesions, it also suffered from lower specificity leading to higher rates of false positive results.


Multiple randomized studies and meta-analyses provide robust evidence for the association between long-term use of selective and non-selective non-steroidal anti-inflammatory drugs (NSAIDs) including aspirin and decreased incidence of colorectal adenomas and cancer. However, gastrointestinal and/or cardiovascular side effects of the long term use of these drugs due to prostaglandin inhibition need to be considered.

NSAIDs are currently not recommended for primary chemoprevention of CRC in individuals with average risk, including those with a family history of CRC. However, celecoxib (selective cyclooxygenase 2, COX-2 inhibitor) 400 mg PO bid was approved by the FDA for reducing polyp burden in patients with FAP as an adjunct to endoscopy and surgery. The role of celecoxib in adjuvant therapy of colon cancer is being explored in a major phase III clinical trial.

Data from two retrospective studies suggested that aspirin may significantly reduce the risk of recurrence in patients with resected colon cancer whose tumors have PIK3CA mutations. However, a third study did not support these findings. Currently it is recommended to consider aspirin for secondary chemoprevention after treatment for colon cancer, although this should be weighed against the long-term risk of GI bleeding.

Beware of other conditions that can mimic colorectal adenocarcinoma:

Because of the nonspecific presenting symptoms of CRC, it may be confused with:

  • Irritable Bowel Syndrome

  • Diverticulosis

  • Inflammatory Bowel Disease

  • Gastrointestinal infections

  • Hemorrhoids

Therefore, it is important to rule out the presence of CRC in any patient with persistent symptoms, especially in those older than 40 years.

Which individuals are most at risk for developing colorectal adenocarcinoma:

Hereditary CRC syndromes, personal or familial history of CRC or polyps, history of ulcerative colitis (UC), and acromegaly cause the greatest increase in risk for CRC, sufficient to alter recommendations for CRC screening in these populations.

Hereditary syndromes and family history

Hereditary CRC syndromes include familial adenomatous polyposis (FAP), hereditary non-polyposis coli (HNPCC or Lynch syndrome), and MYH-associated polyposis. Individuals diagnosed with these syndromes should be offered genetic counseling and information about colonic and extra-colonic cancer screening. Carriers of FAP and MUTYH gene mutations or at-risk family members should undergo colonic surveillance with a flexible sigmoidoscopy and/or colonoscopy every year starting at the age of 10 years, and continuing until age 35 – 40 years if negative.

Patients diagnosed with or strongly suspected to have Lynch syndrome should be offered colonoscopy starting at the age of 20-25 years (except for those with MSH6 mutations, who may start at age 30 years) or 10 years prior to the earliest age of colon cancer diagnosis in the family (whichever is earlier). Patients with high-risk findings must also be considered for prophylactic colectomy as appropriate.

Family history of CRC in one first-degree relative increases the risk about two-fold, which is further increased if CRC was diagnosed below the age of 50, or if more than one first-degree relative were affected. Family history of a large adenoma also appears to confer increased risk, especially if diagnosed at less than age 60 years in the index case. Such patients with family history of early CRC or advanced adenoma in a first degree relative below 60 years or with more than one first-degree relative at any age should be offered screening colonoscopies at the age of 40 years, or 10 years younger than the earliest diagnosis in the family.

Prior history of colorectal adenocarcinoma or polyps

Approximately 1.5–3% of patients with history of CRC develop metachronous primary cancers in the first 5 years after surgery. Patients with adenomatous polyps greater than 1 cm are at 3.5–6.5 times higher risk of CRC.

  • Patients with small, hyperplastic rectal polyps should have a follow up colonoscopy in 10 years.

  • Patients with 1-2, low grade, small (< 1 cm) polyps should have a follow up colonoscopy in 5-10 years.

  • Patients with 3-10 adenomas or adenomas with villous/high-grade dysplasia, or larger than 1cm, should have a follow up colonoscopy in 3 years.

  • Patients with over 10 adenomas should be evaluated for a possible underlying familial syndrome and be offered colonoscopy in under 3 years.

  • Patients with piecemeal removal of sessile polyps should be re-evaluated in 2-6 months for verification of complete removal.

Inflammatory bowel disease (IBD)

The risk of CRC in patients with ulcerative colitis (UC) varies depending on the duration and extent of involvement. It is estimated that the annual incidence is about 0.5% between 10-20 years after diagnosis, increasing to 1% annually thereafter. The risk appears to begin about 8–10 years after diagnosis in patients with pancolitis (5–15 fold increase in risk) and at 15–20 years (3-fold) in those with left-sided colitis.

The data on Crohn’s disease is less robust, but risk appears to be increased in patients with extensive involvement. Surveillance is recommended in all patients with IBD except those with UC limited to the rectum. The American Gastroenterological Association recommends patients with pancolitis should have surveillance after 8 years of disease and those with left-sided colitis after 15 years of disease. Patients with long standing colitis must be considered for prophylactic colectomy.


The incidence and mortality of CRC appears to be higher in African Americans. CRC mortality appears to be higher in men. Although some organizations recommend stratification of screening based on these factors, such an approach is not widely accepted. Multiple environmental factors have been associated with a small increased risk of CRC, mainly in observational studies with unclear causal relationship.

Diabetes mellitus, androgen deprivation, and acromegaly

Insulin resistance as occurs in diabetes mellitus appears to be associated with increased risk of CRC (relative risk [RR] 1.30, 95% CI 1.20–1.40) possibly due to increased levels of insulin, which serves as an important growth factor. Long-term androgen deprivation therapy, either through a gonadotropin-releasing hormone agonist or orchiectomy, also increases risk of CRC (hazard ratio [HR] 1.31, 95% CI 1.12–1.53, and HR 1.37, 95% CI 1.14–1.66, respectively) This association may reflect insulin resistance, although the exact mechanism is unclear. Small observational studies also suggest an increased rate of multiple colonic polyps proximal to the splenic flexure, possibly related to higher insulin-like growth factor-I levels, in patients with acromegaly.


Alcohol consumption, especially if greater than 45gm/day (approximately 2-3 drinks/day) may increase risk of CRC modestly (RR 1.41, 95% CI 1.16–1.72), possibly through decreased folate intake and absorption. Smoking has been associated with increased risk of developing polyps, cancer incidence (RR 1.18, 95% CI 1.11 – 1.25), and mortality (RR 1.25, 95% CI 1.14 – 1.37). Dietary factors, such as long term consumption of red or processed meats and caffeine, have inconsistently been shown to be associated with an increased risk of CRC.

What laboratory and imaging studies should you order to characterize this patient's tumor (i.e., stage, grade, CT/MRI vs PET/CT, cellular and molecular markers, immunophenotyping, etc.) How should you interpret the results and use them to establish prognosis and plan initial therapy?

Obtaining a tissue diagnosis

CRC may be suspected in patients presenting with suggestive symptoms and signs, or may be diagnosed by routine screening. Since the vast majority of colon and rectal cancers are endoluminal, colonoscopy is the best test for diagnosis. It is also used to evaluate for synchronous tumors and the presence of additional polyps. Approximately 20% of patients have metastatic disease at the time of presentation, most commonly in the liver, and can be diagnosed by ultrasound- or CT-guided liver biopsy.

Blood tests

All patients with CRC should also be evaluated with a complete blood count, blood chemistries and liver function tests. Carcinoembryonic antigen (CEA) and carbohydrate antigen (CA) 19-9 are associated with CRC but have low diagnostic ability due to overlap with several benign conditions. These markers are not recommended for screening. However, serum CEA levels should be obtained in patients with known CRC to aid in staging, treatment planning, and assessment.

Pathologic analysis

According to the internationally accepted World Health Organization (WHO) histological classification of CRC (Table I) the majority of CRCs are adenocarcinomas of no specific type. Other less common histological subtypes include the subtypes with adverse prognostic significance independent of stage are signet-ring carcinoma and small cell carcinoma (high-grade neuroendocrine carcinoma). Medullary carcinoma associated with tumor infiltrating lymphocytes and high levels of microsatellite instability may occur sporadically or in association with hereditary nonpolyposis colon cancer (HNPCC). Although immunohistochemical staining is not pathognomonic, CRC is typically cytokeratin (CK) 20 positive, CK 7 negative, CDX 2 positive, and thyroid transcription factor-1 (TTF-1) negative.

Table I.

Who Histological Classification of CRC

See Table I. WHO histological classification of colorectal adenocarcinoma.


The 7th edition of the American Joint Committee on Cancer’s (AJCC) Cancer Staging Manual includes a number of modifications to the TNM staging (Table II) from the previous edition. Due to similar outcomes, rectal cancer and colon cancer continue to have the same staging system. The prefixes “c”, “p” and “yp” are used to denote clinical staging (c), pathologic staging following surgery (p), and pathologic staging following neoadjuvant treatment (yp).

The pathology report should report the grade of the cancer, TNM staging, status of resection margins, lymphovascular/perineural invasion, and extra nodal tumor deposits if any. Histological grading systems vary but most stratify tumors into 3-4 grades. Studies using a 2-tiered system (low-grade versus high-grade) have shown a stage-independent negative prognostic effect for high grade tumors. Other studies have shown negative prognostic effects for both lymphatic and perineural invasion. In fact, several studies have shown that perineural invasion is a high-risk factor for systemic recurrence in resected stage II (29% versus 82%; P=0.0005) and stage III CRC.

Other poor prognostic factors in CRC include:

  • Tumor perforation (2.6–9% of cases)

  • Extra-nodal tumor deposits (discrete tumor deposits within the lymphatic drainage of the tumor in the peri-intestinal fat; classified as pN1c).

  • Extent of residual disease after resection (classified R0-R2).

  • Poor response to neoadjuvant therapy as measured by tumor regression grade in the primary tumor.

In rectal cancer, a positive radial margin is a powerful negative prognostic marker associated with increased local recurrence (10.0% versus 38.2% P < 0.001) and decreased overall survival (29% versus 72% P < 0.001). The number of regional lymph nodes retrieved is dependent on multiple factors but a minimum of 12 lymph nodes are required to accurately identify stage II CRC. In fact, numerous studies have shown that examination of increased number of lymph nodes is itself associated with increased survival.

Molecular profiling

Accumulating evidence suggests that CRCs may have mutations in several oncogenes (including RAS, RAF, PTEN, PIK3CA) and tumor suppressor genes (APC, p53, SMAD2, SMAD4, TGFBR2), chromosomal defects such as microsatellite instability (MSI), 18q imbalance, and over-expression of proteins such as epiregulin and amphiregulin, all of which have potential prognostic and/or predictive implications. Based on currently available therapies, testing for KRAS, BRAF, and MSI may guide management decisions for CRC.


Activating mutations in exons 2, 3, and 4 of the KRAS, or NRAS GTPases immediately downstream to the epidermal growth factor receptor (EGFR) in the MAPK pathway are seen in 50-60% of colorectal cancers. Currently, it is recommended that tumors from metastatic CRC patients being considered for anti-EGFR monoclonal antibody (MoAb) therapy be tested for mutations in KRAS and NRAS in a CLIA-accredited laboratory. Since RAS mutations are early events in CRC carcinogenesis, either primary or metastatic tumors can be tested with a concordance rate of > 95%. In addition to codons 12 and 13, codons 61 and 146 should be tested in the KRAS and NRAS genes; other codons may be recommended depending on ongoing studies.


BRAF is a serine-threonine kinase immediately downstream to KRAS and is mutated in 7-10% of CRC. The V600E mutation in exon 15 of the kinase domain accounts for over 90% of all mutations. Interestingly, KRAS and BRAF mutations are almost always mutually exclusive. BRAF mutations also seem to be associated with right-sided tumors, microsatellite instability – high (MSI-H), and CpG island methylator phenotype (CIMP).

Similar to RAS mutations, BRAF mutations can be tested in either primary or metastatic tumors. In contrast to RAS mutations, BRAF mutations are not predictive (although this is a controversial area) but have strong negative prognostic effects. Although the effect of BRAF mutations on progression free survival (PFS) in resected CRC is unclear, it has a strong negative prognostic effect in metastatic and recurrent CRC, with some studies showing OS half of that in BRAF wild type patients.

Microsatellite instability (MSI)

MSI tumors are more commonly right-sided, and on histological exam reveal poorly differentiated tumors with a mucinous component and extensive lymphocytic infiltration. MSI can be diagnosed by PCR for microsatellites and/or by IHC for loss of mismatch repair proteins. Further germ line mutation testing in the affected gene can be performed, as in the case of HNPCC.

Tumors may be classified into MSI-high (MSI-H), MSI-low (MSI-L), or microsatellite-stable (MSS). Current guidelines recommend initial clinical screening for HNPCC using the Amsterdam criteria or the Bethesda guidelines. Identified cases should then be referred for genetic counseling and also screened aggressively for extracolonic tumors per established guidelines.

Revised Bethesda Guidelines for testing CRC tumors for MSI based on HNPCC workshop 2002:

Tumors from individuals should be tested for MSI in the following situations:

  • Colorectal cancer diagnosed in a patient who is less than 50 years of age.

  • Presence of synchronous or metachronous colorectal, or other HNPCC-associated tumors,* regardless of age.

  • Colorectal cancer with the MSI-H† histology‡ diagnosed in a patient who is less than 60 years of age.§

  • Colorectal cancer diagnosed in one or more first-degree relatives with an HNPCC-related tumor, with one of the cancers being diagnosed under the age of 50 years.

  • Colorectal cancer diagnosed in two or more first- or second-degree relatives with HNPCC-related tumors, regardless of age.

Revised Amsterdam criteria by the International Collaborative Group on HNPCC:

  • There should be at least three relatives with a HNPCC-associated cancer (colorectal, endometrial, small bowel, ureter, or renal pelvis).

  • One relative should be a first-degree relative of the other two.

  • At least two successive generations should be affected.

  • At least one relative should be diagnosed before the age of 50 years.

  • Familial adenomatous polyposis should be excluded in the colorectal cancer case(s), if any.

  • Tumors should be verified by pathological examination.

  • MSI testing for HNPCC should be strongly considered for all CRC patients aged less than 50 years. It can also be considered in patients with resected stage II disease planned for adjuvant chemotherapy with a fluoropyrimidine alone; retrospective pooled analyses from adjuvant trials have suggested favorable prognosis in patients with resected sporadic MSI-H tumors and, in fact, detriment with 5-FU based therapy in stage II patients.

Molecular profiles

Recently, gene expression testing, such as the Oncotype DX Colon Cancer Assay and the ColoPrint, have identified several signatures with prognostic implications in stages II and III colon cancer. It is unclear, however, whether adjuvant treatment in “high risk” tumors improves outcomes, so the clinical utility of such testing to identify which patients should be treated is unproven. Current National Comprehensive Cancer Network (NCCN) guidelines do not recommend routine use of these tests to guide treatment decisions.

Staging the patient

In addition to a complete physical exam with particular attention to ascites, hepatomegaly and lymphadenopathy, the following staging work up is commonly performed:

CT scan

In the US, standard practice at most institutions is to obtain CT of the chest, abdomen and pelvis in all patients with stages II–IV CRC. The sensitivity of CT scans for detecting distant metastasis is 75-87%. (However, sensitivity for detecting peritoneal implants is low ranging from 11% for <0.5 cm size to 37% for lesions up to 5 cm.) Sensitivity for detecting nodal involvement is 45-73% and for detecting transmural invasion is approximately 50%. In rectal cancers, any perirectal adenopathy is presumed to be malignant.


Contrast-enhanced MRI of the liver identifies more hepatic lesions than are visualized by CT and can be useful in patients undergoing evaluation for liver resection. MRI is also useful in rectal cancer patients to identify involvement of perirectal modes with sensitivity ranging from 50–95%.


For rectal cancer, transrectal or endorectal ultrasound (EUS) can accurately predict T stage in 80–95% and nodal status in 70-75% of cases. This can further be improved by fine needle aspiration biopsy of suspicious nodes. Typically, EUS, MRI, or both are obtained preoperatively in patients with rectal cancer to provide information about T and N staging as well as the likelihood of obtaining a negative circumferential margin.


PET scans do not have a significant role in routine preoperative staging of CRC. However, PET or combined PET-CT scan may be considered for localizing site of occult disease in patients with rising CEA and non-diagnostic scans and in patients undergoing evaluation for resection of oligometastatic disease.

Finally, intra-operative staging is essential in all patients undergoing surgery to rule out low volume tumor, especially on peritoneal surfaces.

What therapies should you initiate immediately i.e., emergently?

Most cases of CRC do not require urgent initiation of treatment. However, a number of disease-related complications may need to be addressed emergently.

Bowel obstruction and/or perforation

Although uncommon, these events are seen more frequently with cecal or sigmoid carcinomas requiring emergent surgery. Rectal cancer patients presenting with obstructive and/or symptomatic tumors can undergo either resection followed by adjuvant therapy or intestinal diversion with chemoradiation to the primary tumor. Palliative placement of colonic stents by experienced endoscopists is an option in patients with malignant obstruction who are not operative candidates.

The role of stenting in patients who are operative candidates is not clear but may be considered for optimization of pre-operative medical status or in patients strongly opposed to an ostomy. Electrofulguration or laser ablation may also be effective in relieving luminal obstruction. However, all endoluminal procedures are associated with a risk of perforation and re-obstruction maybe seen.

Gastrointestinal (GI) bleeding

Melena is more often seen with colon cancer and hematochezia with rectal cancer. This is rarely acutely life threatening but may require emergent management including surgery in such cases.

Liver failure

Extensive liver metastases may cause liver failure due to infiltration and/or biliary obstruction. Biliary stenting and/or or urgent initiation of chemotherapy may be indicated in these cases.

Brain metastases

Brain metastases from CRC are less common than in any other tumor types such as breast, lung, and kidney cancer. When brain metastases are diagnosed, patients should be initiated on steroids and treated with whole-brain radiation therapy, stereotactic radiation, surgical resection, or a combination of these modalities.

What should the initial definitive therapy for the cancer be?

Malignant polyp

Polyps with carcinoma in situ (pTis, stage 0) are tumors that have not penetrated the submucosa are not capable of regional nodal metastasis, and require no further surgery beyond endoscopic resection. Malignant polyps are defined as having tumor invading the submucosa (pT1). It is recommended that the malignant polyp site be marked at the time of resection or within 2 weeks of polypectomy. If the polyp has favorable features, (complete resection, negative margins, grade 1 or 2 histology, no angiolymphatic invasion), no additional surgery is required.

Surgery, however, may be considered in patients with sessile malignant polyps because approximately 10% of these patients have regional lymph node metastases. In patients with unfavorable features or fragmented specimens with margins that cannot be assessed, colectomy with en bloc lymph node resection is recommended. All patients with resected polyps should have full colonoscopy and appropriate follow-up surveillance endoscopy.

Stage I

Stage I tumours account for approximately 20% of CRC diagnoses. Stage I encompasses TNM stages T1 and T2 N0M0. These are small tumors with no lymph node involvement. Surgical resection is the standard treatment. Adjuvant (post-operative) chemotherapy is not recommended for stage I.

Stage II

Stage II accounts for approximately 30% of CRC diagnoses. These are larger tumors invading through the muscularis propria into pericolorectal tissues (T3) or penetrating through visceral peritoneum with (T4b) or without (T4a) adjacent organ involvement. There is no evidence of locoregional lymph node involvement or distant metastases.

Surgical resection is the mainstay of treatment for stage II colon cancer. To decrease the risk of distant recurrences, adjuvant chemotherapy for 6 months should be considered for stage II colon cancer patients with high risk features, such as:

  • T4 tumors

  • Poorly differentiated histology

  • Perineural and/or lymphovascular invasion

  • Tumor presenting with bowel obstruction

  • Localized perforation

  • Inadequately sampled nodes

However, it should be noted that recent emerging evidence suggests improved outcomes and lack of incremental benefit with fluoropyrimidine-based adjuvant therapy in stage II MSI-H patients.

In patients with cT3-T4, N0-1 rectal cancer, neoadjuvant (preoperative) therapy is standard in most countries. Consensus guidelines recommend perioperative chemotherapy for 3-6 months including neoadjuvant therapy for rectal cancer patients.

Stage III

Stage III accounts for approximately 30% of CRC diagnoses. Stage III encompasses any T stage tumor with locoregional lymph node involvement but without distant metastases. These stages include T1-4b, N1-N2b, M0. Surgical resection is the mainstay of treatment for stage III colon cancer. Stage III rectal cancer patients are routinely treated with preoperative therapy. Stage III CRC patients should be treated with further adjuvant chemotherapy following surgery.

Neoadjuvant therapy in rectal cancer

The determination of an optimal treatment plan for rectal cancer patients is a complex process that incorporates the intent of surgery (palliative vs curative), stage of disease, probability of obtaining adequate margins and the chances of maintaining or restoring normal bowel and genitourinary function post-surgery. Management of rectal cancer requires a multidisciplinary approach.

In contrast to colon tumors, rectal tumors are associated with a higher risk of locoregional recurrence. Therefore combined modality therapy consisting of surgery, radiation therapy (RT) and chemotherapy is recommended for the majority of patients with stages II and III rectal cancer. Radiation with or without concurrent fluoropyrimidine-based chemotherapy is often given preoperatively due to several potential advantages to this approach:

  • Reduced tumor volume increasing likelihood of negative margins and sparing sphincters

  • Improved oxygenation of surgery-naïve tissue resulting in increased sensitivity to therapy

  • Avoidance of radiation-induced injury to small bowel trapped in the pelvis by post-surgical adhesions

  • Anastamosis remaining unaffected by the effects of radiation therapy because irradiated tissue will be resected

The addition of concurrent chemotherapy to RT is thought to increase local RT sensitization, thereby increasing rates of pathological complete response (pCR) and decreasing risk of local recurrence. A potential disadvantage of the use of pre-operative RT, is the possibility of over-treatment of early-stage tumors that would not have required adjuvant radiation.

The current US approach is to administer concurrent neoadjuvant chemoradiation over 5-6 weeks followed by surgery after 5-10 weeks in resectable rectal cancer (Table III). This approach has been shown to decrease local relapse rates (6% versus 13%) and increase downgrading of tumor stage and rates of sphincter-sparing operations in patients with low-lying tumors (39% vs 19%). It has also been convincingly shown in another major trial that the addition of chemotherapy to neoadjuvant radiation improves pCR rates. However, these trials have failed to show improvement in PFS or OS with neoadjuvant therapy.

Table III.

Commonly used neoadjuvant chemoradiation regimens for rectal cancer.

Initial trials evaluating short-course RT (25 Gy in 5 fractions) suggested increased rates of gastrointestinal complications. However, more recent trials have not supported these findings. This approach, although popular in some European countries, has been slow in getting acceptance in the US, where 50.4 Gy in 28 fractions with concomitant fluoropyrimidine chemotherapy is the preferred choice by most providers.

Although the concomitant chemotherapy regimen used in these trials was infusional 5-FU, this has largely been replaced by capecitabine (administered orally), which has been shown to be non-inferior to 5-FU in large randomized trials. In contrast to the metastatic setting, no benefit is derived from oxaliplatin or irinotecan to fluoropyrimidines. Also, there is no role for adding cetuximab or bevacizumab in this setting. In select cases, after multi-disciplinary discussion, rectal cancer patients may be treated with neoadjuvant chemotherapy and chemoradiation in either sequence.

Surgical resection

Colon cancer

For resectable, non-metastatic colon cancer, the surgical procedure of choice is colectomy including a segment of colon of at least 5cm on either side of the tumor, along with en bloc removal of the regional lymph nodes. The extent of the resection is based on tumor location and the necessity to remove the arterial supply containing the regional lymph nodes. Resection needs to be complete with negative margins. At least 12 lymph nodes need to be examined for optimal nodal staging. Studies comparing laparoscopic versus open colectomy have shown equivalent long-term survival outcomes with the two surgical approaches and some studies suggesting faster recovery and shorter hospital stays with the laparoscopic approach. However, this approach is not recommended for tumors with obstruction, perforation or invasion into surrounding structures, or in patients with extensive adhesions.

Rectal cancer

For rectal tumors, a variety of surgical approaches are used depending on the location and extent of disease. Transanal excision and transanal endoscopic microsurgery may be considered for early, well differentiated tumors with a proximal location. For the majority of patients who are not eligible for these procedures, transabdominal surgery with total mesorectal excision (TME) is recommended to allow adequate lymphadenectomy and improve chances of achieving negative circumferential margins. A TME involves an en bloc removal of the mesorectum, through sharp dissection including associated vascular and lymphatic structures, fatty tissue and mesorectal fascia while sparing the autonomic nerves.

For lesions allowing a margin of 4-5 cm below the distal edge of the tumor, low anterior resection (LAR) with a colorectal or coloanal anastomosis is typically done. An abdominoperineal resection (APR) with resection of the rectosigmoid, rectum and the anus with TME is performed when the tumors involve the anal sphincter and/or the levator muscles, or a margin-negative resection of the tumor would result in loss of anal sphincter function or incontinence.

Accumulating data suggest that in the hands of experienced surgeons, the quality and outcomes with laparoscopic LAR are equivalent to those with open resection. In addition, laparoscopic resection is associated with faster post-operative recovery. Robot-assisted approaches represent a new technology that is still being evaluated and is limited currently by costs and longer operative times.

Adjuvant therapy

High-risk stage II and all stage III colon cancer patients should be considered for fluoropyrimidine-based adjuvant chemotherapy (Table IV) for 3-6 months starting 4-8 weeks post-resection to decrease the risk of recurrent disease and mortality. Current guidelines recommend combining oxaliplatin with infusional 5-FU (mFOLFOX6) or capecitabine in patients who can tolerate combination therapy. Data from the IDEA trial pooled analysis suggest that 3 months of adjuvant oxaliplatin based chemotherapy maybe sufficient in patients with low risk disease stage III cancers (T1-3N1) while those with high risk stage III cancers (T4 and/or N2) may require 6 months of such therapy. Additionally, if 6 months of therapy is chosen, FOLFOX is preferred over CAPOX, pending further study.

Table 4.

Commonly used adjuvant chemotherapy regimens for resected CRC.

In patients who cannot tolerate intensive therapy or oxaliplatin (e.g., due to neuropathy) 5-FU (Roswell Park, Mayo Clinic or de Gramont regimen may be used although toxicity is less with the infusional de Gramont regimen than with the bolus Roswell Park or Mayo Clinic regimens) or capecitabine therapy may be considered. Although the benefit of 5-FU based adjuvant therapy is established in patients older than 70 years, the benefit of adding oxaliplatin in this population is less clear and should be limited only to patients who are fit enough to tolerate it. There is no role for irinotecan, cetuximab, or bevacizumab in the adjuvant setting.

For rectal cancer patients receiving neoadjuvant chemoRT, further adjuvant chemotherapy is recommended, regardless of the pathology results largely based on extrapolation of data from adjuvant colon cancer treatment. FOLFOX, XELOX, 5-FU or capecitabine alone may be used in this setting. Although the optimal duration of adjuvant therapy in this setting is not established, a total of 4-6 months of perioperative chemotherapy (including chemoRT) is preferred.

For rectal cancer patients not receiving neoadjuvant chemoRT, adjuvant therapy typically involves a “sandwich” approach with a few cycles of systemic chemotherapy, then concurrent chemoRT, followed by further systemic chemotherapy to total 6 months of therapy (Table IV).

In contrast to rectal cancer, where the role of adjuvant chemoRT is well defined, RT is not typically a component of care in colon cancer after complete resection. The limited data available suggest possible benefit in patients with (1) a primary lesion fixed to adjacent anatomic structures or in close proximity to the retroperitoneum, which may limit the attainment of adequate margins, (2) a T4 lesion, or (3) positive margins may benefit from adjuvant radiation therapy.

Stage IV

Stage IV CRC accounts for 20% of all patients at diagnosis, and almost 50% of patients with CRC will eventually develop metastases. A small proportion of patients have oligometastatic disease that can potentially undergo curative resection, even if initially deemed unresectable. Therefore, patients with potentially resectable metastatic disease must undergo evaluation every 2 months to see if they are eligible for such resection. In contrast, the majority of stage IV patients are treated with a palliative intent to prolong survival by stopping tumor progression, improve tumor-related symptoms and/or maintain quality of life, with the following general principles:

  • The median OS for patients with unresectable metastatic CRC on best supportive care is approximately 6 months

  • Systemic chemotherapy produces improvements in median OS, with current regimens achieving median OS of 30 months in some large randomized trials.

  • Chemotherapy is indicated only for patients with adequate performance status (PS) of ECOG ≤ 2.

  • Doses must be adjusted for organ dysfunction per guidelines.

  • It is standard practice to start chemotherapy at diagnosis even if patients are asymptomatic and is typically continued until disease progression or treatment-related toxicity.

Cytotoxic agents

The backbone of palliative chemotherapy consists of biomodulated 5-FU (i.e. in combination with leucovorin (LV)) or its oral analogue, capecitabine (Xeloda), in various combinations with oxaliplatin, irinotecan, bevacizumab, ziv-aflibercept, cetuximab, and panitumumab.

  • Fluoropyrimidines impair DNA synthesis via inhibition of thymidylate synthase (TS) and possibly via inhibition of RNA synthesis with bolus administration of 5-FU. LV enhances the cytotoxicity of 5-FU by forming a stable ternary complex with TS prolonging the inhibition of TS by 5-FU.

  • Irinotecan is a topoisomerase inhibitor and has activity as a single agent.

  • In contrast, oxaliplatin, a platinum derivative that inhibits DNA synthesis does not have any activity in CRC unless combined with a fluoropyrimidine.

The exposure to all three cytotoxics in various sequences results in the longest survival. Combination chemotherapy with 5-FU/LV, oxaliplatin (FOLFOX) or with 5-FU/LV, irinotecan (FOLFIRI) provides better outcomes than with a fluoropyrimidine alone. Scheduled treatment interruptions or less intensive cytotoxic treatment may be considered if cumulative toxicity occurs, if disease stability is reached and the metastases are not resectable. Reintroduction of intensive therapy is indicated at recurrence of symptoms and/or disease progression.

FOLFOX and FOLFIRI have similar first-line activity but different toxicity profiles: more alopecia and diarrhea for irinotecan; more neuropathy for oxaliplatin. Both regimens consist of a 46-hour infusion of 5-FU every 2 weeks. Capecitabine can be substituted for 5-FU/LV in FOLFOX (CAPOX, every 3 weeks) with similar efficacy but more hand-foot syndrome and diarrhea, less myelosuppression than 5-FU. However, capecitabine cannot be substituted for 5-FU/LV in FOLFIRI, given the risk of intolerable diarrhea when combined with irinotecan.

Monoclonal antibodies

Monoclonal antibodies against vascular endothelial growth factor (VEGF) (bevacizumab), VEGF and placental growth factor (PlGF) (ziv-aflibercept), or epidermal growth factor receptor (EGFR) (cetuximab, panitumumab) improve outcomes in patients with metastatic CRC (Table V). Bevacizumab increases PFS, and in some settings OS, in combination with cytotoxic chemotherapy. Continuing an anti-angiogenic agent after progression on first-line therapy with or without bevacizumab has also been shown to improve OS modestly and is now routinely done. Ziv-aflibercept is currently approved for this indication (i.e. in the second-line setting in combination with FOLFIRI chemotherapy).

Table V.

Commonly used chemotherapy regimens for metastatic CRC

Therapy with bevacizumab and ziv-aflibercept is associated with hypertension and proteinuria, as well as rare but serious adverse events such as thrombosis, bleeding, gastrointestinal perforation, and impaired wound healing. In addition, ziv-aflibercept appears to be associated with increased rates of diarrhea, mucositis, neutropenic fever, and fatigue.

Activity of anti-EGFR antibodies is limited to patients with RAS wild type tumors. Both anti-EGFR antibodies improve RR and PFS when combined with cytotoxic chemotherapy; as single agents, they improve OS (in studies without crossover) or PFS (in studies with crossover). These agents cause an acneiform rash, hypomagnesemia, and allergic reactions; cetuximab, a chimeric antibody, may cause allergic reactions more often than panitumumab, which is a fully human monoclonal antibody.

Data from a large phase III trial, CALGB/SWOG 80405, suggests that bevacizumab and cetuximab may be equivalent when combined with chemotherapy in the first-line setting with chemotherapy in patients with RAS wild type colorectal cancer. However, unplanned, post-hoc analysis of data from this trial strongly suggested that patients with right sided primary tumors do not respond to anti-EGFR antibodies in the first line setting even if RAS WT – these data were validated by other retrospective studies. Therefore, patients with right sided tumors (ascending colon) should not be treated with anti-EGFR antibodies at least in the first line setting – although accumulating data suggest that similar negative predictive effects of right sidedness to anti-EGFR antibodies may persist in subsequent lines of therapy as well, definitive guidelines and expert recommendations are awaited.

In patients with metastatic disease, resection of the primary tumor is controversial and not routinely recommended unless associated with significant symptoms.

Targeted agents

Regorafenib (160 mg PO daily 3 weeks on/1 week off) is an oral multi- kinase inhibitor that has been shown to have an incremental benefit in chemotherapy refractory CRC over best supportive care. However, toxicities may be substantial.

Another recently approved agent, TAS 102 (Lonsurf), a combination of the nucleoside analogue trifluridine and the thymidine phosphorylase inhibitor tipiracil, demonstrated improvement in progression free survival over best supportive care in patients with chemorefractory colon cancer. The toxicity profile of Lonsurf appears to be slightly more favorable than that of regorafenib, but there is no established standard for the sequencing of these agents.


Patients deficient in DNA mismatch-repair (dMMR), corresponding to high microsatellite instability (MSI-H) (approximately 15% of all CRC and 3-5% of metastatic CRC), may respond to immunotherapy. These tumors demonstrate an increased burden of somatic mutations as well as an increase in tumor infiltrating lymphocytes. In a small phase 2 trial, the immune checkpoint inhibitor pembrolizumab, a humanized monoclonal antibody against programmed death 1 (PD1), has shown promising results in patients with mismatch repair-deficient tumors. In contrast, mismatch-repair proficient tumors, which have low mutation burden, did not appear to benefit. Based on these data, pembrolizumab was granted accelerated approval for all MSI-H tumors (including MSI-H metastatic colorectal cancer), agnostic to the site of origin, for which no satisfactory alternative treatments are available. Based on the promising results of a separate trial evaluating another immune checkpoint inhibitor nivolumab, this agent was also approved by the FDA for patients with MSI-H metastatic CRC after progression on prior fluoropyrimidine, oxaliplatin and irinotecan.

What other therapies are helpful for reducing complications?

Neurotoxicity with oxaliplatin

Oxaliplatin is associated with neurotoxicity manifesting acutely as paresthesias and dysesthesias of hands, feet and perioral region, which are often induced or aggravated by exposure to cold. Chronic, distal sensory neuropathy is related to the cumulative dose of oxaliplatin. Patients should be assessed regularly for symptoms, especially in the metastatic setting after 3-4 months. Infusions of Ca and Mg do not have benefit for prevention of neuropathy.

Toxicity with monoclonal antibodies

Cetuximab is a chimeric mouse-human monoclonal antibody. Premedication with anti-histamines is recommended. Incidence of infusion reactions with cetuximab have geographical variation with the highest incidence in the southeastern US that has been linked to pre-treatment exposure and formation of IgE antibodies against galactose-alpha-1,3 galactose, an oligosaccharide present on the Fab portion of the cetuximab heavy chain.

Therapy with EGFR-targeting monoclonal antibodies is also associated with an acneiform rash typically beginning within a few days or weeks of initiating therapy. In addition to symptomatic treatment of pruritus, oral tetracyclines (minocycline, doxycycline), topical hydrocortisone cream, moisturizing cream and avoiding sun are approaches used to treat and possibly prevent this rash. A phase II trial showed grade 2 skin reactions were decreased by prophylactically treating with a combination of approaches.

Bevacizumab and ziv-aflibercept are associated with hypertension that may require anti-hypertensive therapy, and proteinuria requiring monitoring by urine dipstick analysis. Both these agents should also be discontinued within 6-8 weeks of any peri-operative period, and in patients with serious bleeding or clotting issues.

Toxicity with other agents

Irinotecan has marked interpatient pharmacokinetic variability possibly related to biliary excretion. Its principal side effects are gastrointestinal (mainly diarrhea) and neutropenia. Immediate-onset irinotecan-associated diarrhea is thought to be cholinergic and can be managed with an anti-cholinergic such as atropine. Delayed diarrhea can be life-threatening, especially in elderly neutropenic patients.

Fluoropyrimidines are associated with diarrhea, skin toxicities (palmar-plantar erythrodysesthesia or hand-foot syndrome, photosensitivity, hyperpigmentation and nail changes) and neutropenia. Toxicities with fluoropyrimidines are most marked with bolus regimens and hence, these have mostly been replaced with better tolerated infusional 5-FU and capecitabine.

Regorafenib is associated with multiple toxicities including fatigue, rash, hand-foot syndrome, myelosuppression, and hypertension.

Evaluation of a patient with chemotherapy-induced diarrhea should include determination of severity, symptoms suggestive of complications such as infection, obstruction or volume depletion and neutropenia.

Non pharmacological treatment includes:

  • Aggressive hydration

  • Avoidance of foods that may aggravate symptoms such as dairy products, alcohol, caffeine and those high in fiber and fat

  • Avoidance of promotility and bulking agents

The mainstays of pharmacological treatment include:

  • Loperamide (Imodium) (4 mg initially followed by 2-4 mg every 4 hours or after each bowel movement until diarrhea free for 12 hours)

  • Diphenoxylate (Lomotil) (5 mg 3-4 times/day until control is achieved)

  • Octreotide may be used as second line therapy in patients whose diarrhea is not controlled with the above agents. Although the optimal dose has not been defined, an initial dose of 100 mcg SC three times a day with titration upwards as necessary may be administered

  • Fluoroquinolones are the antibiotics of choice in those with a suspicion of infection or neutropenia

Patients on fluoropyrimidines should be counseled about photosensitivity. Patients with hand-foot syndrome should be managed with topical application of aloe vera, moisturizing cream or petroleum. All patients with drug toxicities should have dose interruption and future dose reduction as necessary. In addition to toxicities, capecitabine also has clinically significant drug-drug interactions with agents such as warfarin and phenytoin.

Follow-up of patients on these trials also suggests that most patients have resolution of symptoms, especially above, and equal to, grade 3 within 12-18 months after discontinuation of therapy. Furthermore, data from other randomized trials support the “stop and go” approach in palliative chemotherapy with interspersion of non-oxaliplatin containing maintenance chemotherapy (e.g., 5-FU/LV intermittently with FOLFOX) to minimize oxaliplatin neurotoxicity.

What should you tell the patient and the family about prognosis?

The most important predictor of outcome is the stage of disease at diagnosis (Table VI). In general surgical resection is the only potential curative modality and is employed in patients with stages I-III and select IV patients with oligometastatic disease. In patients with resected stage III colon cancer, adjuvant FOLFOX improves OS (6-year OS 73% vs 69%) and DFS (5-year DFS 66% vs 59%).

Table VI.

Stage specific 5-yr survival of colon and rectal cancer.

Patients with earlier stages of disease, including stage I (5-year survival 93%), stage II (5-year survival 72-85%) do not require adjuvant therapy except in those with high risk stage II disease. Patients with resectable rectal cancer are at a higher risk of locoregional relapse and require chemoRT in addition to adjuvant chemotherapy to decrease this risk. Outcomes are similar to those of corresponding stages in colon cancer. The majority of patients with metastatic colorectal cancer are treated with a palliative intent to prolong life while preserving quality of life.

With modern chemotherapy regimens, the median OS of patients with metastatic CRC is 25-30 months (5-year survival approximately 10%). These numbers indicate clear improvements in outcomes for patients with CRC, and many promising novel therapies remain under development.

Recently, prognostic molecular biomarkers have been described. BRAF mutations occur in 7-10% of patients and are associated with poor outcomes, especially in the metastatic setting. Data from the 80405 study suggest that patients with primary tumors arising in the right side of the colon (ascending colon) have significantly worse overall survival compared to those with primary tumors arising in the left side of the colon (descending colon including rectum) irrespective of the type of therapy used. MSI-H, found in 22% of stage II and 11% of stage III patients, is associated with better outcomes in the adjuvant setting.

What if scenarios.

Oligometastatic disease

The liver is the most common site of metastatic spread. Twenty to thirty-four percent of patients with metastatic disease at presentation and as many as 50-60% of all CRC patients have liver involvement. Approximately half of patients developing liver metastases have this as the only site of metastatic disease, and approximately 20% of such patients may be eligible for resection. Patients who undergo metastatectomy have a median survival of 36-40 months and a 10-year survival/cure of 15-20%. These patients with oligometastatic disease to the liver or lung require an initial multi-disciplinary evaluation with input from an experienced surgical oncologist to assess resectability.

Modern, aggressive approaches to liver resection for CRC metastatectomy define resectable tumors as those that can be resected completely (R0) per initial staging with an adequate future liver remnant (FLR). In patients who may not have adequate FLR, a staged resection may be performed where one lobe of the liver is cleared of tumor following a period of recovery before the other side is addressed. Another technique is portal vein embolization of the lobe planned for resection, resulting in hypertrophy of FLR. In practice, these techniques are often used in conjunction.

In a small portion of patients with liver-limited disease that is thought to be unresectable due to involvement of critical structures, neoadjuvant chemotherapy with a chemotherapy regimen active in the metastatic setting may be used to downsize the tumors. In these patients, it is important to limit neoadjuvant therapy to 2-3 months and to perform surgery as soon as possible to avoid steatohepatitis and sinusoidal injury from irinotecan and oxaliplatin, respectively. A total of 6 months of perioperative systemic chemotherapy is recommended in patients undergoing resection.

Patients with multiple metastatic sites are unlikely to be resectable based solely on a favorable radiographic response, as these sites have residual micrometastatic deposits in most cases. Radiotherapy (stereotactic body radiotherapy [SBRT] or intensity modulated radiotherapy [IMRT]) and radiofrequency ablation (RFA) (if oligometastatic disease with 3 or fewer lesions, less than 5 cm in size, and not located near a major vascular structure that may act as a “heat sink’) are modalities that may be used for local control of oligometastatic disease in patients who are not ideal surgical candidates.

Patients with borderline resectable disease may benefit from neoadjuvant chemotherapy with a high objective response rate to downsize the tumor and these patients must undergo surgery as soon as the metastases become resectable. This is important because the duration of chemotherapy appears to be associated with liver damage (chemotherapy-associated steatohepatitis, or CASH) and subsequent risk of higher morbidity and mortality. Despite the paucity of data, consensus guidelines recommend a total of six months of perioperative therapy for patients who have undergone resection of hepatic metastases, typically with an oxaliplatin-based regimen given the data in other adjuvant settings.

A smaller proportion of patients presenting with extra-hepatic oligometastatic disease, such as in the lung, can be treated with similar approaches. In addition, patients with oligometastatic disease who are not surgical candidates can be treated with local approaches including radiofrequency ablation, chemoembolization, radioembolization (if limited to liver) or by stereotactic radiation (either intra- or extra-hepatic oligometastatic disease) either alone or in conjunction with resection. The role of these organ-directed therapies in prolonging survival is unknown.

Follow-up surveillance and therapy/management of recurrences.

In spite of optimal primary therapy, 30-50% of patients with resected CRC will have disease relapse. The main goals of surveillance are detecting relapses and second primaries that can be cured by surgical intervention. The majority of CRC recurrences are within 2 years of surgical resection; hence, patients are followed closely during this period. There is significant controversy in the optimal follow-up post primary therapy but national guidelines recommend:

  • History and physical exam every 3-6 months for first two years, every 6 months during years 3-5 and annually thereafter are recommended. CEA is recommended at these intervals for patients with T2 or greater lesions if the patient is a potential candidate for aggressive curative therapy. 3 month intervals may be more appropriate in stage III patients at higher risk of recurrent disease.

  • Colonoscopy is recommended approximately 1 year post-resection, typically repeated at 3 years and then every 5 years thereafter unless colonoscopy findings require shorter follow up.

  • CT scans are recommended annually for the first 3-5 years in stage II and III patients.

  • In addition, patients should have a survivorship care plan with regular health maintenance, regular exercise and smoking cessation.


Most CRCs (85%) arise from adenomatous polyps over a period of years to decades by accumulation of serial somatic mutations due to underlying inherited or acquired chromosome instability (CIN). According to the adenoma-carcinoma sequence model, the initiating mutation is in the APC gene. Subsequent mutations include KRAS (35-40% of all CRC), BRAF (7-10% of all CRC) with implications for treatment and prognosis respectively. Other events include p53 alterations and loss of chromosome 18q. FAP is characterized by germline mutations in APC leading to activation of the Wnt pathway.

HNPCC and about 15% of sporadic cases are characterized by germ line or somatic DNA repair defects or methylation changes in the mismatch repair (MMR) genes, which may lead to genomic instability due to impaired ability to correct DNA replication errors such as nucleotide based mis-pairing. This leads to mutations in cancer related genes thus driving carcinogenesis.

In addition, characteristically, expansion or contraction of repetitive nucleotide sequences called microsatellites compared to normal tissue is also noted in these cases. The hypermethylation phenotype (CIMP+) is characterized by DNA methylation of CpG islands of many genes including those involved in MMR resulting in silencing of gene expression typically causing serrated adenomas. A small portion of patients with multiple colorectal adenomas and family history of CRC have mutations in the base excision repair gene, MYH.

What other additional laboratory studies may be ordered?

Although consensus guidelines are lacking, pharmacogenomics testing for CRC chemotherapeutic agents including fluoropyrimidines and irinotecan are commercially available.

DPD deficiency

Dihydropyrimidine dehydrogenase (DPD) is the rate limiting enzyme in 5-FU metabolism and patients with partial or complete deficiency of DPD have increased toxicities with fluoropyrimidines. Although complete absence of DPD activity is rare, African-American women have the highest prevalence of DPD deficiency (12.3%), followed by African-American men (4.0%), Caucasian women (3.5%), and Caucasian men (1.9%). Although commercially available tests can detect DPD deficiency or gene mutations, their clinical utility is currently limited due to multiple factors.

UGT1A1*28 polymorphisms

The active metabolite of irinotecan, SN38 is glucuronidated by UGT1A, a member of the UDP-glucuronosyltransferase family. The activity of this enzyme is determined by polymorphisms in the promoter region of its gene, with the most common variant being UGT1A1*28, associated with Gilbert’s syndrome and can be detected by an FDA approved test.

The frequency of this allele has geographical variation. In North America, approximately 10% of the population is homozygous and 40% is heterozygous. This allele results in decreased enzymatic activity. Irinotecan dose reduction is recommended in UGT1A1*28 homozygotes.

Although the FDA approved genetic test is available for detecting this allele, significant confusion about this test still exists as homozygosity may not be a marker for toxicity from low dose irinotecan (50-180 mg/m2). Furthermore, there is no consensus on the initial dose reduction for UGT1A1*28 homozygotes, and there is evidence that this polymorphism may account only for a fraction of the variability in irinotecan pharmacokinetics.

Jump to Section