Colon cancer is the most common, and the most preventable, form of gastrointestinal cancer. Survival rates have improved markedly, and long-term survival can now be achieved in patients with advanced-stage cancers at diagnosis. See the image below.
See Benign or Malignant: Can You Identify These Colonic Lesions?, a Critical Images slideshow, to help identify the features of benign lesions as well as those with malignant potential.
An estimated 101,340 cases of colon and 39,870 cases of rectal cancer are expected to occur in the United States in 2011. Colorectal cancer is the third most common cancer in both men and women. Median age at diagnosis is 69 years. More than 50,000 people die from the disease annually. Incidence rates have been declining over the past 20 years, attributed to more effective screening, in addition to a continuing reduction in mortality.  Survival rates are related to stage at diagnosis, declining to less than 10% for patients with distant metastatic disease.
Bowel cancer is the fourth most common cause of cancer death worldwide, estimated to be responsible for 8% of the total deaths in 2008. Bowel cancer mortality rates are lowest in Middle Africa and South-Central Asia and highest in Central and Eastern Europe, with a 6-fold variation in male mortality rates between the regions of the world and a 5-fold variation in female rates. 
Subsets of the population, such as those with hereditary polyposis, family history of colon cancer, and inflammatory bowel disease, have been found to be at higher risk. 
Bowel cancer mortality is strongly related to age. In the United Kingdom between 2008 and 2010, an average 80% of bowel cancer deaths were in people aged 65 years and older. Mortality rates increase sharply from around age 50 years in men and around age 65 years in women, reaching a peak at age 85+ years in both sexes. The male-to-female ratio increases with age, from 13:10 at age 50-54 years to 14:10 at age 85+ years. 
The prevalence of right-sided colonic adenocarcinoma increases with patient age. In one study of 2942 patients aged 11-95 years (mean age, 61 years), adenocarcinoma was found on endoscopy in 191 patients. In patients younger than 50 years, the prevalence was 2 (15%) of 13; for those aged 50-59 years, 8 (21%) of 39; for patients aged 60-69 years, 18 (32%) of 57; aged 70-79 years, 25 (42%) of 49; and for those 80 years or older, 16 (57%) of 28. 
Most colonic adenocarcinomas (80%) arise in preexisting adenomatous polyps, and progression from polypoid adenomas to carcinoma occurs over several years.
The examples below show a patient referred for screening CT colonoscopy who returned 5 years later with no interim follow up.
Most cancers are asymptomatic until the disease is advanced. Blood may be noted in the stool, but occult bleeding is more common; patients may present with iron deficiency anemia. Increasing obstruction of the large bowel may cause symptoms such as constipation, change in bowel habits, and, in advanced cases, feculent vomiting. The tumor may hemorrhage acutely, perforate, or cause pain by invasion of adjacent organs.
Limitations of techniques
No single technique is 100% accurate in the detection of bowel carcinoma, or its precursor, adenomatous polyps. The following screening guidelines are designed to stratify patients by risk.
Separate guidelines have been issued for those at average and higher risk.
Screening is recommended to begin at age 50 years, with annual fecal occult blood tests or stool testing for exfoliated DNA (the frequency for this test is not yet established). Options then include flexible sigmoidoscopy, imaging with double-contrast barium enema every 5 years, or CT colonography every 5 years.  Flexible colonoscopy is recommended every 10 years. Clinicians are advised to discuss the risks and benefits of these recommendations on an individual basis with their patients.
In addition to the diagnostic accuracy of each test, the risk of complications and individual tolerance of invasive procedures needs to be weighed. Although a double-contrast barium enema study is included in the guidelines for screening for colon tumors, a study by Ferrucci et al  emphasized the low yield of this procedure when used to evaluate patients after incomplete colonoscopy, and the technique has fallen into disfavor in the past 6 years as more patients undergo endoscopy or CT colonoscopy.
A subset of colon cancers (2-5%) are associated with inherited syndromes, including Lynch syndrome, familial adenomatous polyposis, MUTYH-associated polyposis, and certain hamartomatous polyposis conditions, each condition associated with a far higher risk of developing colon cancer. Up to one third of colon cancers exhibit increased familial risk.  For these groups, full colonoscopy is recommended, commencing at an earlier age.
Clinical staging of colorectal cancer
Stage is the strongest predictor of survival for patients with colorectal cancer.  The TNM (tumor, nodes, metastasis) staging system proposed by the American Joint Committee on Cancer/Union Internationale Contre le Cancer (AJCC/UICC) is the recommended staging system for colorectal cancer. 
One of the primary advantages of the TNM system is that the system is data driven and susceptible to continuous improvement.  Another advantage of a single, united system is a common language regarding tumor burden that is understood by clinicians throughout the United States and internationally.
The TNM system is derived from surgical resection specimens. The T refers to local extent of the untreated primary tumor. The N refers to tumor involvement of regional lymph nodes and the lymphatic system. The M refers to metastatic disease.
Tables 1-4 below are adapted from the AJCC Colon and Rectum Cancer Staging.
Table 1. Primary Tumor (T) (Open Table in a new window)
|Primary Tumor (T)|
|TX||Primary tumor cannot be assessed|
|T0||No evidence of primary tumor|
|Tis||Carcinoma in situ; intraepithelial or invasion of lamina propriaa|
|T1||Tumor invades submucosa|
|T2||Tumor invades muscularis propria|
|T3||Tumor invades through the muscularis propria into the pericolorectal tissues|
|T4a||Tumor penetrates to the surface of the visceral peritoneumb|
|T4b||Tumor directly invades or is adherent to other organs or structuresb,c|
a This includes cancer cells confined within the glandular basement membrane (intraepithelial) or mucosal lamina propria (intramucosal) with no extension through the muscularis mucosae into the submucosa.
b Direct invasion in T4 includes other organs or other segments of the colorectum as a result of direct extension through the serosa, as confirmed on microscopic examination (eg, invasion of the sigmoid colon by a carcinoma of the cecum) or, for cancers in a retroperitoneal or subperitoneal location, direct invasion of other organs or structures by virtue of extension beyond the muscularis propria (ie, a tumor on the posterior wall of the descending colon invading the left kidney or lateral abdominal wall, or a mid or distal rectal cancer with invasion of the prostate, seminal vesicles, cervix, or vagina).
c Tumor that is adherent to other organs or structures, grossly, is classified cT4b. However, if no tumor is present in the adhesion, microscopically, the classification should be pT1-4a depending on the anatomical depth of wall invasion. The V and L classifications should be used to identify the presence or absence of vascular or lymphatic invasion, whereas the PN site-specific factor should be used for perineural invasion.
Table 2. Regional Lymph Nodes (N) (Open Table in a new window)
|Regional Lymph Nodes (N) a|
|NX||Regional lymph nodes cannot be assessed|
|N0||No regional lymph node metastasis|
|N1||Metastasis in 1-3 regional lymph nodes|
|N1a||Metastasis in one regional lymph node|
|N1b||Metastasis in 2-3 regional lymph nodes|
|N1c||Tumor deposit(s) in the subserosa, mesentery, or nonperitonealized pericolic or perirectal tissues without regional nodal metastasis|
|N2||Metastasis in 4 or more regional lymph nodes|
|N2a||Metastasis in 4-6 regional lymph nodes|
|N2b||Metastasis in 7 or more regional lymph nodes|
|a A satellite peritumoral nodule in the pericolorectal adipose tissue of a primary carcinoma without histologic evidence of residual lymph node in the nodule may represent discontinuous spread, venous invasion with extravascular spread (V1/2), or a totally replaced lymph node (N1/2). Replaced nodes should be counted separately as positive nodes in the N category, whereas discontinuous spread or venous invasion should be classified and counted in the Site-Specific Factor category Tumor Deposits (TD).|
Table 3. Distant Metastasis (M) (Open Table in a new window)
|Distant Metastasis (M)|
|M0||No distant metastasis|
|M1a||Metastasis confined to one organ site (eg, liver, lung, ovary, nonregional node)|
|M1b||Metastases in more than one organ/site or the peritoneum|
Table 4. Anatomic Stages and Prognostic Groups (Open Table in a new window)
|Anatomic Stage/Prognostic Groups a|
|IVA||Any T||Any N||M1a||-||-|
|IVB||Any T||Any N||M1b||-||-|
a cTNM is the clinical classification and pTNM is the pathologic classification. The y prefix is used for those cancers classified after neoadjuvant pretreatment (eg, ypTNM). Patients who have a complete pathologic response are ypT0N0cM0 that may be similar to Stage Group 0 or I. The r prefix is to be used for cancers that have recurred after a disease-free interval (rTNM).
b Dukes B is a composite of better (T3 N0 M0) and worse (T4 N0 M0) prognostic groups, as is Dukes C (Any TN1 M0 and Any T N2 M0). MAC is the modified Astler-Coller classification.
For patients with UICC stage 1 and 2 colon cancer, 5-year survival rates are on the order of 95% for those treated with surgery and no adjuvant chemotherapy.  Adjuvant therapy is not routinely used; however, in a subgroup (12% of the sample), certain clinicopathologic characteristics (not assessed in the TNM classification systems) have been found to be associated with a significantly increased risk of tumor recurrence and tumor-related death. These include lymphatic vessel invasion (P =.034), poor tumor grading (G3/G4) (P =.020), and extended tumor length (6 cm) (P =.042).
The efficacy of CT scanning to diagnose a primary tumor may be limited by tumor size or location. Despite no existing recommendations for the use of CT scanning as a screening modality, other than CT colonography, there may be cases in which a primary tumor may be evident on a scan obtained for another purpose and, if colon or rectal wall thickening, mesenteric infiltration, adenopathy, or liver lesions are not noted, could represent a reason for litigation.
In the interpretation of CT colonography, diagnosis of lesions of less than 5 mm, and limitation in detection of flat lesions, is not accurate. Staging inaccuracies of imaging techniques are a limitation that is understood, and other validation by endoscopy, pathological node and tumor analysis, tumor markers, and surgical techniques are used in the definitive clinical staging of colorectal carcinoma.
Optical colonoscopy remains the criterion standard against which the accuracy of other screening tests is measured.  A major advantage of colonoscopy is that polyps can be excised during the procedure. However, it has a relatively higher cost and requires conscious sedation. Not all patients may be able to tolerate sedation, or it may be too risky to discontinue anticoagulation in persons with heart valves, for example. It may not be as cost effective to use colonoscopy in a population with low prevalence of disease.
Alternatives such as screening CT colonoscopy may an advantage in being less invasive and perhaps more suitable to screen patients with a low prior probability of disease. Radiography, CT (including virtual CT colonoscopy), MRI, transrectal ultrasound, and positron emission tomography (PET) have roles in the diagnosis and management of colon carcinoma.
Plain abdominal radiographs are a useful first step in the evaluation of patients presenting with complications of colon cancer, such as large bowel obstruction or perforation. Free air below the diaphragm may be detected by plain erect chest radiograph. Rarely, mucin-producing colon cancers show calcification in the primary tumor and in hepatic and peritoneal secondary deposits visible on radiographs.
Double-contrast barium enema is one of the modalities approved for colon cancer screening, but the frequency of its use has markedly decreased over the past decade. As Ferrucci et al note,  “DCBE [double-contrast barium enema] is a low-yield procedure for detecting polyps, with a high false-positive rate, and is not likely to be performed by experienced practitioners in the future.”
Although barium enema is no longer among common screening tools, the following examples offer an overview of its utility in diagnosis of polyps and carcinomas, which can be synchronous, and gross tumor morphology, as shown by double-contrast barium enema.
Approximately 5% of patients with colon cancer have more than 1 cancer at diagnosis (see the image below).
Approximately 35% of patients with colon cancer have an associated adenomatous polyp (see the image below).
The appearance of colonic tumors on double-contrast barium enema reflects diverse morphologic types: polypoid, annular, or flat.
Polypoid lesions vary from small, smooth tumors to larger lobulated masses with an irregular surface and an associated contour deformity along 1 margin of the bowel wall (see the image below). The incidence of carcinoma in an adenomatous polyp is related to its size and surface features; larger, more irregular ulcerated lesions are more likely to contain carcinoma.
Annular lesions result from irregular, circumferential masses that severely constrict the bowel lumen. The margins of the carcinoma show overhanging edges, the tumor shelf or shoulder (termed "apple-core" lesion). The mucosal folds in the narrowed segment are destroyed; ulceration may be present (see the image below).
Flat lesions, which are rare, are visualized as a unilateral, broad-based, contour defect. Ulceration may be present (see the image below). Flat lesions may infiltrate the bowel wall and, if extensive, cause areas of nondistensibility.
Small carcinomas usually present as a polypoid mass with a smooth outline; they may be indistinguishable from a benign polyp. Rarely, they may present as a small, flat lesion.
Radiologically, a polypoid mass is visualized either as a filling defect in the barium column (single-contrast study) or, more commonly, as a barium-coated soft tissue mass protruding into the air-filled lumen (double-contrast study).
A sessile polyp may be visualized as a crescent (or ring) shadow on the bowel wall (see the image below).
Pedunculated polyps have stalks that may be identified easily on profile (see the image below). When the stalk is seen through the polyp itself, this results in a target appearance. Malignant change may occur in the head of a stalked polyp. A long (2 cm or more), thin (5 mm or less) stalk may hinder the spread of carcinoma from the head of the polyp into the wall.
Degree of confidence
The accuracy of double-contrast barium enema is under 80%. 
The false-negative rate for double-contrast barium enema is 22.4%. 
CT scanning (including multidetector computed tomography [MDCT] and CT colonography) is used as an adjunct in screen i ng for colon carcinoma, in staging colon cancer before surgery, for assessing and staging recurrent disease, and for detecting the presence of distant metastases.
Colon tumors may be diagnosed on a CT scan as an incidental finding (see images below) or in patients presenting with acute symptoms related to complications of a colonic tumor, such as perforation.
Adenocarcinoma of the colon presenting as an acute abdominal condition
Elderly patients are more likely to present with symptoms associated with complications from late-stage colonic tumors. Two examples are demonstrated in the images below of patients previously undiagnosed as having colon carcinoma, presenting with signs and symptoms of acute abdominal perforation.
Spontaneous perforation may be associated with slightly higher postoperative mortality, but long-term survival in such patients is a function of the stage at presentation, which tends to be more advanced.  Other complications that may present acutely include obstruction and fistula formation, as shown in the image below.
CT in screening for colorectal carcinoma
CT colonography, better known as virtual colonoscopy, has been introduced as an alternative method for screening for colorectal carcinoma and is gaining in acceptance as an alternative or complementary diagnostic test to conventional colonoscopy. Its efficacy was established in the multicenter trial, the American College of Radiology Imaging Network (ACRIN) study. 
CT colonography is a minimally invasive examination of the colon and rectum designed to evaluate for polyps and neoplasms.  It is similar to diagnostic colonoscopy in that it requires 24 hours of bowel preparation with a combination of cathartics, contact laxatives, dietary restriction, and hydration.  However, new developments in fecal tagging may be able to eliminate the bowel preparation process.  Once the patient has prepared for the examination, a rectal tube is inserted and room air or carbon dioxide is insufflated into the colon under controlled pressure to provide full colon distention. 
CT images are typically obtained through the colon in at least 2 positions (usually supine and prone).  Additional imaging is obtained on an as-needed basis after repositioning and reimaging to ensure that all colonic segments are imaged adequately. Once images are obtained, they are sent to an integrated computer system for image review and further manipulation. Recent advances in technology and software have allowed the conversion of 2-dimensional (2D) axial, coronal, and sagittal slices into 3-dimensional (3D) endoluminal rendering or “fly-through” techniques.  See the images below.
Indications/guidelines for ordering
Indications for CT colonography are similar to those for colonoscopy. Individuals undergoing the examination fall into 1 of 3 categories designated as screening, surveillance, or diagnostic. 
Screening seeks to detect individuals with polyps or colon cancer without signs or symptoms of the disease. For disease screening, the US Multi-Society Task Force on Colorectal Cancer, American Cancer Society, and the American College of Radiology recommend a screening CT colonography every 5 years. 
Diagnostic CT colonography encompasses symptomatic individuals or individuals in whom a screening examination detected an abnormality and further imaging is required. A common indication is for “completion colonography” following an incomplete optical colonoscopy.
In contrast to screening and diagnostic CT colonography, surveillance CT colonography includes monitoring people with known colorectal cancer. Repeat examinations are tailored to a patient’s individual case and treatment plan.
CT colonography can also be offered to patients who are at increased risk for complications from conventional colonoscopy, such as increased anesthesia risk. Contraindications for CT colonography include the following  :
Symptomatic acute colitis
Recent acute diverticulitis
Recent colorectal surgery
Symptomatic colon containing abdominal wall hernia
Recent deep endoscopic biopsy
Suspected colonic perforation
Symptomatic/high-grade small bowel obstruction
In addition, there are few indications for which CT colonography has not been approved, including routine follow-up of inflammatory bowel disease, hereditary polyposis/ nonpolyposis syndromes, evaluation of anal canal disease, or in pregnant patients. 
Lesions are categorized using the CT Colonography Reporting and Data (C-RAD) system  as C0 if the study was inadequate and as C1 if the study was normal. Polyps of 6-9 mm and fewer than 3 in number are classified as C2 (indeterminate), and continued surveillance or colonoscopy is recommended. C3 lesions include those larger than 10 mm in diameter or if more than 3 lesions of 6-9 mm are present, for which colonoscopy is recommended. C4 is used to describe a colonic mass with associated luminal narrowing or extracolonic extension, for which urgent referral for consideration of surgery is recommended. The system also recommends categorization of significant extracolonic findings.
The accuracy of CT colonography compared with traditional colonoscopy has been greatly debated over the past few years. The first large CT colonography–based screening trial in the United States in 2005 estimated a sensitivity of greater than 90% for the detection of polyps larger than 1 cm.  Subsequent trials in the mid 2000s reported lower sensitivities of 55% and 59%, respectively, [26, 27] but a large scale study sponsored by ACRIN in 2008 reported a specificity of greater than 90% and a sensitivity of 90% for polyps larger than 9 mm.  Moreover, sensitivity for lesions measuring 5-9 mm was reported with a sensitivity of 85-89%.
CT colonography has several disadvantages over traditional colonoscopy. The first is radiation exposure possibly leading to a long-term carcinogenic effect, although with newer image acquisition systems, radiation exposure has been significantly decreased.  Another disadvantage is that if a significant lesion is found on CT colonography, then the patient would likely have to be brought back for a second study with a second bowel preparation (ie, colonoscopy) for confirmation, surveillance, or removal of the lesion if graded C2 or higher.  Yet another disadvantage is interinterpreter, intersystem variability. The reported results may not be replicable without extensive training in the technique.
CT colonography carries a small risk of procedural complication, including bowel perforation.  Risk of perforation is lower than for optical colonoscopy. Recent retrospective analyses of almost 29,000 mostly symptomatic patients screened in the United Kingdom and Israel reported a perforation rate of 0.06% and 0.08%, respectively,  lower than that of optical colonoscopy but not negligible.
The sensitivity of CT colonography for the detection of flat lesions is has not been established; such lesions may constitute 40% of all adenomas.  See the images below.
CT in staging adenocarcinoma of the colon.
CT scanning plays an integral role in the staging of colon cancer. CT scanning can help identify the tumor site, the size of the tumor, and locoregional and metastatic spread, which can help guide surgical planning.  The degree of wall invasion, tumor extension into the mesentery, and the presence of lymph node and distant metastases are major factors that influence patient prognosis.   33] .
The current role of CT in patients with known colon cancer is controversial; overall accuracy rates for preoperative staging of colon cancer based on early evaluations of CT, prior to the introduction of multidetector technology, ranged from 48-77%.  Determination of depth of tumor invasion through the colonic wall has not been accurate. 
The hope is that the introduction of MDCT scanning and improved processing software can improve accuracy. One 2012 study found that using MDCT, the accuracy rates for identifying T1 and T2 tumors were as high as 90.4% and 73.9%, respectively,  although the accuracy of correctly identifying tumor involvement of lymph nodes was lower for N0 and N1 patients, at 61.6%. This limitation is related to size criteria for evaluation of lymph nodes, with an inability to identify malignant infiltration of nonenlarged lymph nodes.
An earlier small study using axial MDCT in conjunction with multiplanar reformatted images found similar results with tumor staging accuracy, at 83%, and slightly improved nodal staging, at 80%. 
CT colonography also offers some promise as a staging modality. CT colonography is increasingly being performed after incomplete colonoscopy to assess for synchronous lesions and metastasis.  One study evaluated CT colonograms from 246 patients to assess tumor staging. Overall accuracy for tumor stage was 79%. 
As technology advances, it is likely that MDCT scanning and CT colonography will play increasingly important roles, along with MRI, in the initial staging of colorectal cancer.
The primary tumor may be inapparent on CT staging, particularly if the lesion is relatively small, or flat, and the bowel is filled with feces. Lesions appear as asymmetric, wall-thickening focal intraluminal masses, or they may be annular. Occasionally, a colon carcinoma presents as a stricture with large bowel obstruction. Annular carcinomas are detected by a thickening of the bowel wall and narrowing of the lumen. Use of multiplanar images to view the colon in the orthogonal plane facilitates assessment of tumor thickness and lesion length. Adenomatous polyps and adenocarcinomas are enhanced by iodinated contrast. As tumors progress, vascularity is recruited into the tumor and adjacent mesentery. See the images below.
Extracolonic tumor spread is indicated by a loss of tissue fat planes between the colon and surrounding structures. Advanced colonic tumors may directly invade the anterior abdominal wall, retroperitoneum, liver, pancreas, spleen, or stomach. See the images below.
The accuracy of correctly identifying tumor involvement of lymph nodes is lower, with accuracy rates for N0 and N1 patients at 61.6%,  although a recent small study using axial MDCT in conjunction with multiplanar reformatted images found similar results, with tumor staging accuracy of 83% and slightly improved nodal staging, with an accuracy rate of 80%. 
Nodes greater than 10 mm in diameter are considered abnormal. No CT tissue characteristics have been identified that enable discrimination between enlarged benign nodes and enlarged malignant nodes. Furthermore, malignant foci may be present in nodes less than 1 cm in diameter.
Enlarged nodes may be detected in the mesentery and retroperitoneum (see the first image below). Enlarged nodes may also be observed around the porta hepatis (see the other images below).
Rectosigmoid tumors may metastasize to external iliac nodes.
Despite limitations in accuracy of staging, there is an association with survival  ; in a study of more than 500 patients who underwent curative resection for colonic cancer, radiological (CT) nodal staging (not using subclassifications) was associated with a survival rate at 5 years of 83% (N0), 76% (N1; 1-3 regional lymph nodes), and 54% (N2; 4 or more regional lymph nodes).
Hepatic metastases are the most common site of distant spread and develop in 50% of cases. Following injection of intravenous contrast medium (see the images below), hepatic metastases are demonstrated as well-defined areas of low density (compared with normal liver parenchyma) best seen (ie with maximum difference in tissue attenuation between unaffected liver and lesion) in the portal venous phase. In the earlier arterial phase, hepatic metastases may show rim enhancement or become hyperdense or isodense in relation to normal liver.
Even extensive hepatic metastases may be suitable for surgical resection, provided the primary tumor has been resected for cure (R0) and that complete resection is achievable based on anatomic criteria and maintenance of hepatic function. Increasingly sophisticated techniques may include staged resection, a combination with preoperative portal vein embolization, or radiofrequency ablation prior to, or in conjunction with, resection. 
Many clinical trials are in process to evaluate the timing and optimal regimens of preoperative and adjuvant chemotherapy, and a review of surgical consensus conferences by Masi et al  summarizes that the only rules when performing surgical resection of colorectal liver metastases, in addition to a complete (R0) resection, are the preservation of 2 contiguous liver segments (with adequate vascular inflow and outflow) and adequate future liver remnant (>20% of the total volume in a healthy liver). Limited portal adenopathy does not exclude surgical resection.
Kanas et al  performed a review and meta-analysis of studies published from 1999-2010 and determined that median 5-year survival rate after resection of colorectal liver metastasis was highly variable (ranging from 16-74%) and depended on a number of factors. Median survival was better in patients with a carcinoembryonic antigen (CEA) level of less than 200, with a negative tumor margin, and fewer than 3 liver metastases. Patients with higher tumor grade had the worst median survival, and those with negative nodes did better. Survival chances were also improved by higher annual clinic volume. The more recent studies did not show improved survival compared with the earliest studies. The overall median survival following colorectal liver metastasis liver resection was 3.6 years. 
Given the continuing controversy regarding the benefit of hepatic resection, CT evaluation for intrahepatic metastatic disease should focus on accurate reporting of intrahepatic location and size of metastases, bearing in mind the need for 2 contiguous segments with adequate vascular supply and drainage for resection. Radiologists may be requested to make assessments of liver volume in patients being considered for hepatic resections. See the image below.
The concept of unresectability is changing, and patients with disease previously thought unresectable may still be candidates for surgery with curative intent, but there is as yet no consensus on the selection of this subset. 
Common sites of metastatic involvement include the lungs, adrenal glands, peritoneum, and omentum. In females, the ovary may be involved.
Although pulmonary metastases may be detected by chest radiography, CT scanning has a higher sensitivity for small pulmonary metastases (< 10 mm). See the image below.
Adrenal metastases may occur in as many as 14% of patients with colon cancer. They manifest with enlargement (>2 cm), asymmetry, and heterogeneity. Adrenal metastases are fluorodeoxyglucose (FDG) avid and may be further characterized by MRI evaluation. See the images below.
Bony and cerebral metastases are uncommon.
Reporting of metastatic disease should concentrate on careful reporting of location and size of lesions, which will be the continued focus of attention as the patient undergoes further treatment. At this point, use of Response Evaluation Criteria in Solid Tumors (RECIST) Group criteria  is not generally applied in clinical practice, but is a requirement in clinical trials for standardization.
An example of a postoperative complication is shown below.
CT in surveillance after colon cancer therapy
Surveillance guidelines recommend annual CT surveillance of the chest, abdomen, and pelvis,  in addition to frequent clinical evaluation, CEA testing, and follow-up colonoscopy. The American Society of Clinical Oncology (ASCO) guidelines have been updated to recommend CT surveillance based on meta-analyses, showing CT scanning or liver imaging is associated with a survival benefit. The mortality rate for patients who undergo liver imaging was found to be 25% lower than that for patients who do not undergo such imaging. Rectal cancer patients with poor prognostic factors may also benefit from pelvic CT scanning, especially if they have not been treated with radiation.
Chest CT scanning is beneficial in the ASCO panel’s conclusions in that the highest proportion of resectable lung metastases, particularly in patients with rectal cancer, is found by using chest CT and is often not associated with elevation of CEA markers.  See the images below.
Contrast-enhanced CT colonography has been evaluated for efficacy of colorectal cancer in surveillance after curative resection,  and the authors concluded that the technique matched colonography CT screening trials in accuracy, could detect extracolonic metastatic disease, and may eliminate the need for more invasive surveillance colonoscopy during routine postoperative surveillance. However, sensitivity for adenomas was more limited at 80% versus 100% for carcinoma, and the authors conceded that alternating techniques may be a more prudent approach
Degree of confidence
Early evaluations of the accuracy of CT scanning were published in the late 1980s through early 2000. The introduction of MDCT scanning, which enables rapid acquisition of images in multiple phases and multiplanar reformatting of images based on thin section slices as low as 0.5 mm, has provided radiologists with images of high quality and improved spatial resolution. The accuracy of tumor staging has improved, but prediction of tumor involvement within nonenlarged lymph nodes is still limited. Based on a recent meta-analysis of 19 series, Dighe et al  concluded that MDCT has the potential to make a considerable impact in improving the accuracy of staging of colon cancer, with a specificity of 93% and a sensitivity of 86% for detecting tumor invasion and a specificity of 87% and a sensitivity of 70% for detecting lymph node metastasis.
Huh et al,  who reviewed more than 500 cases (excluding stage IV) scanned from 1997-2007, noted radiological staging was a significant independent predictor of long-term survival. Five-year survival progressively fell in patients staged by CT as T1 (96%), T2 (89%), T3 (75%), and T4 (79%), as well as in those TNM-staged by CT, from 90% (stage I) to 81% at stage II, and 70% in stage III disease. CT scanning cannot replace pathologic staging, but it provides useful insight into disease manifestations and locoregional and distant spread, which can help in operative planning and management.
In some settings, for example elevation of CEA following therapy for colon cancer, CT was found less accurate than positron emission tomography (PET) scanning; a 2010 study showed a sensitivity of MDCT of 70.3%, as compared to PET scanning, with a sensitivity of 97%. Limitations of CT were in detection of recurrence in the presacral space, in metastatic subcentimeter lymph nodes, in peritoneal deposits, at the periphery of ablated liver metastases, and in the uterine cervix. 
Inadequate bowel preparation and/or distention, flat lesions, and small polyps are causes for missed lesions (false negatives) at multidetector row CT colonography. In one study, 3 of 3 flat lesions larger than 1 cm were not distinguished. Other reasons for error included classification of polyps as feces. 
CT scanning signs for colon cancer are not specific and may be caused by any disease associated with focal thickening of the colon wall, including diverticulitis, Crohn disease, ischemic colitis, and tuberculous colitis. A paracolic collection may be seen in diverticulitis, as well as in local perforation of a carcinoma.
In cachectic patients, the absence of fat planes is a result of nutritional status and may limit evaluation for tumor extension.
Chronic radiation changes in the pelvis may mimic recurrent colon tumors and require PET scanning correlation and/or biopsy for differentiation.
Tumors in the transverse colon and colon flexures may be visualized incompletely. A primary gastric carcinoma with extension into the colon may be indistinguishable from a colon tumor invading the stomach.
Hypodense hepatic lesions may be caused by simple cysts rather than metastases. Hemangiomas also may cause confusion.
Magnetic Resonance Imaging
MRI has roles in staging of rectal tumors and in assessment of liver metastases in colorectal carcinoma. Before the introduction of high field–strength magnets and endorectal or phase array coils, CT was considered superior to MRI for local staging of rectal cancer.  Although no advantage has been yet found in the use of 3-Tesla field-strength imaging in comparison to 1.5-Tesla in tumor staging of rectal cancer,  MRI has now been shown to be able to reliably exclude T3 tumors with a negative predictive value of 97%. 
One notable pitfall is the difficulty with differentiating a desmoplastic reaction in the perirectal fat (T2) and true tumor infiltration (T3), although in practice it may not change therapy.
Preoperative MRI is also used to identify tumor distance from the mesorectal fascia, which is a predictor of tumor positivity at the circumferential resection margin (CRM) after a total mesorectal excision.  A positive resection margin is defined as tumor extension to within 1 mm of the mesorectal fascia. See the image below.
Recent advances in MR colonography have been spurred on by improvements in equipment and techniques. MR colonography is similar to CT colonography in that the patient is imaged prone and supine. Fecal tagging techniques have also been developed to make the preexamination preparation more tolerable. Contrast between the colonic lumen and potential polyps is created through either a bright lumen (gadolinium enema) or dark lumen (water/air/carbon dioxide) technique.  MR colonography can accurately identify polyps 10 mm or greater while not exposing the patient to ionizing radiation.  While the lifetime risk of radiation-induced cancer in the average patient aged 50 years and older is low, it is significant when one considers using CT colonography as a widespread screening technique. Additionally, MR colonography may play a role in screening for younger patients with genetically increased risks for colon cancer. 
MRI provides greater contrast between soft tissues than CT scan. Although both T1- and T2-weighted sequences are recommended, T2-weighted sequences appear the most useful, combined with high-resolution matrices, thin-section imaging, and small field of view. Fat suppression (used to improve detection of perirectal fat invasion) and diffusion-weighted imaging may also be helpful. [56, 57] Gadolinium administration for evaluation of primary rectal tumor extension has not been found to be helpful. 
Tumor enhancement can be achieved by paramagnetic agents such as gadolinium. Gadolinium-based contrast agents include gadopentetate dimeglumine (Magnevist), gadobenate dimeglumine (MultiHance), gadodiamide (Omniscan), gadoversetamide (OptiMARK), and gadoteridol (ProHance). However, these agents have been linked to the development of nephrogenic systemic fibrosis (NSF) or nephrogenic fibrosing dermopathy (NFD) in patients with renal impairment.  Patients with an estimated glomerular filtration rate (GFR) of less than 30 mL/min/1.73 m2 should not receive gadolinium-based contrast agents unless the benefits significantly outweigh the risks. The disease has occurred in patients with moderate-to–end-stage renal disease after being given a gadolinium-based contrast agent to enhance MRI or MR angiography scans.
NSF/NFD is a debilitating and sometimes fatal disease. Characteristics include red or dark patches on the skin; burning, itching, swelling, hardening, and tightening of the skin; yellow spots on the whites of the eyes; joint stiffness with trouble moving or straightening the arms, hands, legs, or feet; pain deep in the hip bones or ribs; and muscle weakness.
The recently developed technique of MRI colonography (during which no intravenous contrast is used) can detect colon polyps greater than 1 cm in diameter and is likely to compete with CT colonography in screening programs.
MRI has limited accuracy in determining nodal involvement, when solely relying on anatomical criteria. The addition of MR diffusion-weighted imaging has increased the sensitivity and specificity up to 97% and 81%, respectively. 
MRI may be used in symptom-based evaluations of other organs systems, if, for example, there is concern for brain metastasis or to characterize an adrenal mass in the setting of a patient with newly diagnosed colon carcinoma. See the image below.
MRI has been advocated as the most sensitive liver imaging in patients with colorectal cancer with liver metastases, who need evaluation for suitability for partial hepatectomy.  In a meta-analysis of prospective studies of 3391 previously untreated patients (39 articles; 1990-2010), it was determined that on a per-patient basis, the sensitivities of CT, MRI, and fluorodeoxyglucose (FDG) positron emission tomography (PET) were 83.6%, 88.2%, and 94.1%, respectively, with comparable specificity. For lesions smaller than 10 mm, the sensitivity estimates for MRI were significantly higher than those for CT. The sensitivity of MRI increased significantly after January 2004. The use of liver-specific contrast material and multisection CT scanners did not provide improved results in this study. Data on FDG PET/CT were too limited for comparisons with other modalities.
See the image below.
Typically, metastases from adenocarcinoma of the colon appear hypointense on T1-weighted images (not shown) and hyperintense on T2- and diffusion-weighted sequences, with peripheral enhancement following administration of gadolinium.
Use of agents that undergo biliary excretion in addition to renal excretion, such as gadolinium-ethoxybenzyl-diethylenetriamine pentaacetic acid (Gd-EOB-DTPA; Eovist or Primovist; Bayer Healthcare Pharmaceuticals; Wayne, NJ) and gadobenate dimeglumine (Gd-BOPTA; MultiHance; Bracco Diagnostics; Princeton, NJ) is recommended.  With these agents, delayed T1-weighted images demonstrate normal parenchyma as hyperintense and metastatic lesions appear hypointense.
Degree of confidence
MRI has lower sensitivity and higher specificity than CT scanning in tumor staging. The techniques have similar overall accuracy in tumor staging, as well as similar overall accuracy (approximately 60%) in the detection of enlarged lymph nodes (nodal staging). In detecting local recurrence, MRI has a higher sensitivity (91%) than CT scan (82%) and a higher specificity (100%) than CT scan (69%).
MRI, while useful for specific problem solving (eg, staging of rectal carcinoma, evaluation of liver metastases), is not a cost-effective screening modality for the entire abdomen and pelvis.
However, MRI is recommended as the preferred first-line modality for evaluating colorectal liver metastases in patients who have not previously undergone therapy, as it has the highest sensitivity for lesions smaller than 1 cm. 
Limitations of MRI are similar to those of CT scanning. Colon cancer may be indistinguishable from a large benign tumor and from metastasis to the colon (usually from an ovarian primary). MRI signs for colon cancer are not specific and may be caused by any disease associated with focal thickening of the colon wall. These diseases include diverticulitis, Crohn disease, ischemic colitis, and tuberculous colitis.
A paracolic collection may be seen in diverticulitis, as well as in local perforation of a carcinoma. In addition, chronic radiation changes in the pelvis may mimic recurrent colon tumors and require biopsy for differentiation. Enlarged lymph nodes may result from inflammation rather than tumor, and lymph nodes of normal size may contain a tumor.
Transrectal ultrasound (TRUS) may be used to distinguish between layers of the rectal wall and thus detect depth of tumor penetration and perirectal spread. The technique requires no ionizing radiation and is relatively inexpensive to perform.  Tumor staging accuracy for TRUS (84.6%) is superior to that of CT (70.5%), but both underperform similarly in nodal staging (64.1% vs 61.5%, respectively).  T2 lesions with peritumoral inflammation, which obscures differentiation of the rectal wall layers, remain a typical overstaging pitfall for TRUS. 
Although not primarily used as a primary staging modality, hepatic metastases from a colon primary tumor may be detected as hyperechoic masses (increased echogenicity in relation to normal liver) on transabdominal real-time ultrasound, but their appearances can be variable.  See the image below.
Ultrasonic contrast agents, although approved for use in Europe and Asia, are of limited availability in the United States and remain FDA unapproved for noncardiac indications. These agents generate microbubbles of high acoustic amplitude when activated, which recirculate, and may be viewed over intervals of several minutes by real-time or Doppler ultrasonography. Lesion detectability is enhanced and Doppler is more sensitive in detecting tumor vascularity. See the images below.
Intraoperative ultrasound (IOUS) is routinely used during partial hepatectomy to confirm and localize lesions in guiding surgical planning, and it is common to find additional lesions not demonstrated by other imaging techniques. Contrast enhancement is now also being used. 
In a study of 60 patients already selected for liver resection, who had undergone preoperative staging with CT or MRI, IOUS using an HDI-5000 scanner (Philips) and a finger probe with pulse inversion harmonic capability (low MI: 0.02–0.04) was performed after intravenous injection of 3-4 mL of SonoVue (Bracco SPA, Milan) and detected liver metastases with higher sensitivity than CT/MRI or IOUS alone (96.3% vs 76.7% and 81.5%, respectively; P < .05). Surgical management was also altered in 30% of cases, owing to additional (19.3%) or fewer (3.5%) metastases or characterization of lesions as benign (5.3%) or rarely, vascular proximity (1 case), and CE-IOUS altered combined IOUS/CT/MR staging in 35%.
A primary colon tumor typically appears as an echo-poor mass with a hyperechoic center, which is known as the target sign (see the image below). Other findings include localized irregular colon wall thickening, an irregular contour, lack of normal peristalsis, and an absence of the normal layered appearance of the colon wall.
Intussuscepting colon tumors have a characteristic targetlike appearance from concentric rings of soft tissue and mesenteric-fat density (see the image below).
The most commonly used radiotracer is fluorodeoxyglucose (FDG), a glucose analog that is taken up and trapped in cells. To differentiate tumor from background, FDG positron emission tomography (PET) relies on the increased glucose metabolism of tumor cells and calculates a semiquantitative measure of glucose uptake, also known as standard uptake value (SUV). With the advent of simultaneous CT scanning, PET scanning also improves both spatial resolution and anatomic localization of findings.
FDG-PET scanning has become more popular in the imaging of colorectal cancer and is used in pretreatment staging, posttreatment follow-up, and surveillance. It achieves similar sensitivity and specificity compared with contrast-enhanced CT scanning. For certain indications, such as monitoring response to therapy,  assessing for recurrence in patients with elevated carcinoembryonic antigen (CEA),  assessing for hepatic metastases post partial hepatectomy,  and differentiating local recurrence versus scar,  small studies have shown incremental benefit over conventional diagnostic imaging. For now, the lack of larger studies and the higher costs of FDG-PET scanning prohibit more widespread adoption. [70, 71, 72]
See the images below.
Degree of confidence
A study by Meta et al evaluated the impact of FDG-PET on the management of patients with colorectal carcinoma and noted a change in the clinical stage and major management decisions in approximately 40% of patients.  Of the changes in clinical stages in 25 patients, the disease was up-staged in 20 patients (80%) and down-staged in 5 patients (20%). As a result of PET findings, physicians avoided major surgery in 41% of patients for whom surgery was the intended treatment.
False-positive results may occur with FDG from nonspecific inflammatory reactions following radiotherapy or in patients with abscesses. Infectious/inflammatory conditions also demonstrate FDG uptake, which limits the ability to differentiate from tumor. For example, it is not able to differentiate diverticulitis from tumor, a pitfall PET scanning shares with conventional CT scanning. FDG-PET also has decreased sensitivity for mucinous (58%) versus nonmucinous (92%) adenocarcinomas. 
Angiography has a role in the administration of chemotherapy via the hepatic artery, an invasive approach based on the rationale that higher doses of chemotherapy may be delivered to the tumors with reduced systemic toxicity and reduced toxicity to the unaffected liver, since normal supply is predominantly via the portal venous system. As yet, consensus has been reached on the ideal combination of agents. 
A recent study evaluated a protocol of intra-arterial therapy with mitomycin C, alone or in combination with irinotecan or gemcitabine, in 463 patients who had not responded to systemic chemotherapy.  The infusions were followed by embolization with lipiodol and starch microspheres in order to occlude small tumor vessels and obstruct the vascular bed of the liver tumors. Although 37% had evidence of progressive disease, partial response was obtained in 15% and 48% remained stable. There was no difference between the protocols, and median survival from date of diagnosis of liver metastases was 38 months and from the start of chemoembolization treatment was 14 months. Survival for those with progressive disease was 13 months, in comparison to 18.2 months in those with partial response, a statistically significant difference.
Results for patients with unresectable, chemorefractory liver metastases using radioembolization with yttrium 90 (90 Y) microspheres are also encouraging. Response rates of 30-60% are being achieved, and, in one study, median survival of 457 days for patients with colorectal tumors was achieved. 
Minor complications of90 Y therapy include fatigue, pain, nausea, bilirubinemia, and reduction in lymphocyte count. Major complications occurred in less than 5%, including gastrointestinal ulceration, radiation-induced cholecystitis, biloma, and hepatic abscess, all requiring further intervention.