Similar to CS, in which an optimal examination requires a clean colon, bowel cleansing is critical for CTC. Residual stools may mimic cancer or a polyp or, alternatively, may hide a real colonic lesion; the same is true for fluid residues. Bowel preparation is an evolving area of research since the ultimate goal is to reduce or abolish the use of laxative agents in order to improve patient compliance compared with CS, but not at the expense of diagnostic quality. The ideal colonic preparation is still being debated, but general agreement exists regarding the need for dietary restriction, fecal tagging, and bowel purgation[7]. Dietary restriction (e.g., a low fiber diet) for one to three days before CTC examination is aimed at reducing fecal volume and stool inhomogeneity, thus favoring optimal tagging[8]. Bowel purgation should normally be included in CTC preparation; however, in order to improve patient compliance, aggressive catharsis should be restricted to 24 h or less and a reduced volume of laxative agent should be preferred[9-12]. The choice of the agent reflects local experiences, with no clear preference for any agent.

In order to improve polyp detection and to reduce the number of false-positive examinations, “labelling” or “tagging” of residual stools and fluids is strongly recommended and is considered mandatory unless contraindicated. Fecal tagging consists of oral administration of either water-soluble iodinated contrast medium or, alternatively, a diluted barium sulfate suspension. Because such oral contrast agents are hyperattenuating to X-rays, the tagged residual bowel content appears hyperdense and can be readily distinguished from true colonic lesions. The use of electronic stool subtraction software allows the colon to be cleansed of tagged residues and virtual navigation to be performed[13] (Figure 1). Although robust data are not available in the literature comparing the two tagging agents[14-16], iodinated contrasts offer more homogeneous tagging and the possibility for same-day CS; the risk of allergic reactions, due to iodine absorption through the gastrointestinal tract, is extremely rare, or even anecdotal[17].

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Pedunculated polyp submerged by fluid residues. A: Polyp displaying soft-tissue density (arrow) is partially submerged by tagged fluid (asterisk) on an axial prone computed tomography image; B: On electronically cleansed endoluminal view, a polyp with a pedunculated morphology (arrow) is clearly observed; C: On colonoscopy, a pedunculated polyp is detected before resection.

A further step to improve patient compliance is the utilization of reduced bowel preparations[18] using only high doses of hyperosmolar contrast agents, without the need for any additional laxative agent[19,20]. This approach, which is associated with better patient compliance and with reduced cathartic effects compared with a conventional colon cleansing for CS, can now be considered routine practice, particularly in screening[5] and in elderly and frail patients[21,22].

The colon is inflated with air or carbon dioxide by using a thin and flexible rectal catheter by a radiologist, resident, technician, or nurse specifically trained in the technique[23]. Although automatic insufflation with carbon dioxide is the method of choice to optimize colonic distention and to maximize patient comfort, a higher cost is a drawback of this technique[24]. No sedation is required. The use of a spasmolytic agent, and in particular hyoscine butylbromide (Buscopan^®)[25], may provide better distension, especially of the sigmoid colon in the case of diverticular disease or severely stenosing cancer, although a clear improvement in the accuracy of polyp detection has not been demonstrated[26,27]. The intravenous injection of iodinated contrast medium is reserved for the following situations: in patients with known CRC, to improve staging, and in symptomatic patients based on the clinical indication, and for the need to investigate the extracolonic organs fully[28].

The use of multidetector row CT (MDCT) scanners (≥ 16 rows) is considered an essential prerequisite to achieve high-quality (max collimation ≤ 2.5 mm; single breath-hold) and low-dose examinations. The use of dose modulation devices[29], which reduce does exposure by 30%-35% (except in obese patients) is strongly recommended; if available, adaptive statistical iterative reconstruction and model-based iterative reconstructions[30], which reduce dose exposure up to 50%, are preferred.

Image post-processing is performed on dedicated off-line workstations suitable for 3D data management and reconstruction. CTC interpretation should be based on both 2D and 3D images, and the use of either a primary 3D or 2D approach depends on personal preference and availability[31]. Computer-aided detection (CAD) software is useful to reduce perceptual errors during CTC interpretation, if employed in a second-reader paradigm. Particularly, CAD improves sensitivity for small (6-9 mm) polyps, even in expert readers[32]. However, the use of CAD should not preclude adequate reader training, since the interpretation of CAD markings is under the control of the radiologist.

Recent meta-analyses[4,33] and randomized clinical trials have reported that the diagnostic performance of CTC for the detection of CRC and clinically significant adenomatous polyps (≥ 10 mm) both in symptomatic[2,3] and asymptomatic subjects[5] is similar to CS and is largely superior to BE. In fact, the sensitivity of CTC and CS for CRC were shown to be 96% and 95%, respectively, in a recent meta-analysis[33]. The importance of this study is that it showed that sensitivity is maintained, despite a wide variation in technique, demonstrating potential generalizability and widespread implementation of CTC. In the first randomized trial[2], which compared CTC and CS in symptomatic patients, the overall performances were not statistically significant different for either CRC or large (≥ 10 mm) polyps. Given the relatively low prevalence of CRC, even among symptomatic cohorts, primary CTC may be more suitable than CS for the initial investigation of suspected CRC. In fact, the cancer miss rate of CTC is low. In a study on 3800 patients who were followed-up by using the National Cancer Registry database after CTC[34], seven cancers were missed (five because of technical limitations and two because of perceptive errors; systems errors and severe patient co-morbidity contributed to three of the cases), with an overall missed rate of around 5.3%. These data are similar to those collected for CS (miss rate, 5.9%)[35] and were better than those collected for BE (missed rate, 6.7%)[36].

Several studies have demonstrated the accuracy of CTC in identifying and characterizing colonic polyps (Figure 2). Two randomized clinical trials[2,3,5], two multicenter trials[37,38], two single-center trials[39,40], and eight different meta-analyses[4,33,41-46] came to similar conclusions: CTC sensitivity for clinically significant polyps (larger than 10 mm) is high and is similar to that of CS; the sensitivity for small (6-9 mm) polyps is slightly lower than that of CS; the specificity and negative predictive value, even for small polyps, are good, especially if fecal/fluid tagging techniques are used; and the variability of the results among different series is mostly due to perceptual errors and consequently to reader inexperience[47]. The importance of reader experience was also confirmed by a multicenter study[48] designed to assess the accuracy of CTC in detecting polyps or cancers after a preliminary training and qualifying program for radiologists. The most relevant finding of the study was that radiologist’s polyp detection rate during training was the only significant factor in predicting the accuracy of CTC for detecting polyps.

Open in New Tab   Full Size Figure   Download Figure

Figure 2 Patient with incomplete colonoscopy due to severe angulation and stricture secondary to diverticular disease of the sigmoid colon. A: On a volume-rendered colon map, a stricture of the sigmoid colon (arrow) and a large filling defect (asterisk) on the medial wall of the descending colon are evident; B: On a coronal image, a large polyp (arrow) of the descending colon is observed.

This result leads to the issue of training and the learning curve. Although the debate is on-going, one study[49] attempted to determine how many CTC training datasets had to be evaluated by a novice reader to reach an adequate level of diagnostic confidence, which was defined as a sensitivity of 95% for lesions 10 mm or larger, a sensitivity of 90% for lesions 6 mm or larger, and a per-patient specificity of 80% for lesions 6 mm or larger. This study showed that at least 175 CTC studies with colonoscopic verification are necessary for most trainees, although a sub-group may require additional training because of the inability to reach the predefined threshold levels. However, once adequately trained, radiologists can obtain consistent performance for adenoma and advanced neoplasia detection, as well as other clinically relevant endpoints, in a CTC screening program[50].

Non-polypoid lesions represent a sub-group of colonic neoplasms, which are classified into the following types: slightly elevated (type 0-IIa), completely flat (type 0-IIb), and depressed (type 0-IIc)[51]. Type 0-IIa lesions are sometimes misclassified as sessile polyps because they are slightly elevated from the mucosal surface. In order to avoid this confusion, the height of a flat lesion should not exceed ½ of its diameter[52]. In CTC, a lesion is defined as flat when the vertical elevation above the surrounding mucosa is less than 3 mm[53] (Figure 3). Another sub-type of non-polypoid tumors is the so-called “carpet” lesion, which is defined as a flat, laterally spreading colorectal mass measuring 3 cm or greater in size[54].

Open in New Tab   Full Size Figure   Download Figure

Figure 3 Non-polypoid lesion (type II-A). A: Non-polypoid lesion (arrow) is observed on an axial image; B: Lesion (arrow) is confirmed on an endoluminal computed tomographic colonography image; C: Colonoscopy confirming the presence of the non-polypoid lesion (arrow).

Although the epidemiology is not completely clear, recent data suggest that the prevalences of non-polypoid lesions in Europe[55] and the United States[56] are close to that of Japan[57]. The likelihood of malignancy is directly related to morphology rather than to size: types 0-IIa and 0-IIb have a risk of harboring high-grade dysplasia or cancer similar to polyps, whereas type 0-IIc has a definitely higher likelihood of malignancy[57]. CTC can potentially detect flat and depressed adenomas, which result in focal thickening of the colonic wall, but it is probably not suitable for evaluating the presence of completely flat lesions[58]; however, the latter are exceedingly rare[56,57]. At the time of this paper, data regarding CTC sensitivity for flat lesions are sparse, and the patient series are small for the available studies. The largest studies report sensitivity in the range of 80%-90% for flat adenocarcinomas[59,60], and excellent results were also described for lateral spreading tumors[61].

A possible advocated advantage of CTC is the detection of extracolonic findings (ECF). ECF are common in symptomatic patients as well as in screening patients referred for CTC, with a prevalence ranging between 25% to more than 50%[62]. They increase with age and are not clinically significant in most cases; however, they are reported as important in 10% of patients, requiring further work-up[63]. Important ECF include extracolonic cancers (with the most common being renal and lung cancers and lymphoma), abdominal aortic aneurysms, and adrenal lesions. The work-up for ECF produces additional costs that should be considered when calculating the cost-effectiveness of CTC, particularly for screening. In a mathematic model, the detection of an ECF such as abdominal aortic aneurysm and extracolonic cancers in addition to CRC makes CTC a dominant screening strategy over both CS and CS with one-time ultrasonography[64].

Incomplete or failed CS: Incomplete or failed CS is not an uncommon event. The caecal intubation rate, a quality metric recorded by the endoscopists, is extremely variable in daily practice and it can be far from high-quality standards[65].

CS can be incomplete for a variety of reasons other than poor bowel preparation: obstructing masses, neoplastic or inflammatory strictures, redundant colon, patient discomfort, colonic spasm, severe diverticulosis, and adherences from previous surgery. Further, incomplete CS is more common in elderly patients and in women[66]. The experience of the endoscopist is another extremely important variable because up to 10% of CS procedures are considered technically challenging, even for experts[67].

In the case of incomplete/failed CS, examination of the entire colon is recommended in order to avoid missing an advanced lesion[68]. Because the performance of CTC is superior to BE[3,46,69], CTC is the examination of choice[70]. In the case of neoplastic obstruction, CTC can be used for excluding synchronous lesions in the remaining colon and for whole body staging of primary colonic neoplasm, if contrast-enhanced CTC is performed[71,72]. Moreover, CTC can define predictive factors for a potentially difficult CS, such as colonic looping, acute angle flexures, and tortuosity, because it reveals colonic anatomy[73].

Elderly and frail patients: CS, apart from being incomplete or failed, can be unfeasible because of the patient’s condition, including advanced age, poor compliance to bowel preparation, or associated severe co-morbidities. CS has an increased risk of perforation and bleeding in elderly patients[74] and in those undergoing anti-coagulant therapy[75], respectively. Because of the progressive aging of the population, this sub-group of patients is likely to increase in the future, with consequently higher frequencies of colonic cancer and diverticular disease[76,77]. CTC has the advantage of being technically feasible, well tolerated, and safe. Only patients with a positive finding will be referred for more invasive and risky examinations because of the high positive and negative predictive values for cancer and large polyps[78,79].

Investigation of diverticular disease: Diverticular disease is the most common colonic disorder in the Western world, with a prevalence of over 60% at 80 years[77]. The clinical diagnosis is challenging because patient symptoms and laboratory findings are unspecific and overlap with other gastroenterological entities (e.g., irritable bowel syndrome) and in young women (< 40 years of age) with gynecological disorders[80]. Thus, imaging tests are necessary for the diagnosis. The choice of either CTC or CS as the first-line imaging examination depends mostly on the patient’s age, risk factors, clinical status, and preference, as well as on imaging availability and local expertise[81]. In elderly individuals, especially those who are frail and have potential contraindications to CS and sedation, CTC is preferred because it is minimally invasive. On the other hand, CS should be the first choice in younger patients in whom symptoms might be related to colic inflammatory changes because the diagnosis would be challenging with CTC. The use of BE should be discouraged because of poorer patient compliance[82], longer examination time, higher risk of complications[83,84], and radiation exposure[85].

In patients with established diverticular disease, CTC can provide a balanced view of the disease by incorporating visual analysis with quantitative analysis by using a CTC-based diverticular disease severity score; this score appears to correlate with relevant coexisting lesions and can potentially influence therapeutic decision-making[86].

Symptomatic patients: Gastrointestinal alarm symptoms potentially suggestive of CRC such as abdominal pain or discomfort, rectal bleeding, iron-deficiency anemia, and unintended weight loss are highly non-specific; further, these symptoms are, unfortunately, common in the general population, particularly in the elderly[87]. The accuracy for identifying patients with underlying structural disease is disappointing, and many of these patients who are referred for CS will ultimately prove to be normal[88]. Elderly patients are often frail and more likely to have an incomplete or difficult CS, with a greater risk of adverse events[65,66] and an increased workload for endoscopy centers. For these reasons, CTC might be used as a triage technique, with only patients with positive findings being referred for CS. Practically, proposing CS as the best indication for symptomatic patients presenting with bleeding and diarrhea and proposing CTC for those with pain or weight loss may be reasonable.

Other indications: New and emerging indications have been explored, with some of them being entered de facto in clinical use: CTC for tumor localization before laparoscopic surgery because of the inaccuracy of CS in determining the precise localization and extent of the lesion[89]; investigation of patients with colonic stoma[90]; and assessment of colonic involvement in patients affected by deep pelvic endometriosis[91].

CRC screening: CRC screening is complex, and this is particularly true in Europe, where population-based programs using fecal occult blood tests are already organized. Within the framework of an established screening program, the role of CTC is to replace BE in the case of incomplete CS[92] and to evaluate patients with a positive fecal occult blood test who refuse CS[93].

Despite the current situation, research is committed to exploring the potential use of CTC as a screening modality alternative to the available tests. Before implementing CTC, data on patient adherence to CTC-based screening programs and cost-effectiveness are necessary. The recent results of a Dutch trial[5] showed that the participation rate for the screening program was increased with CTC compared with CS (34% vs 22%; P < 0.0001). Although these results are extremely interesting, they must be confirmed by other on-going trials[94,95].

The available data on the cost-effectiveness of CTC for screening based on mathematical models are discordant: in the majority of the models, CTC is dominated by either CS or a combination of flexible sigmoidoscopy and fecal occult blood test[96]. However, the costs of CT-screening in the Dutch trial[97] were substantially lower than the cost assumptions used in published cost-effectiveness analyses, indicating the need for re-evaluation of those data.

Apart from population screening, on an individual basis, CTC can be suggested as a CRC screening test provided the individual is adequately informed about the test characteristics, benefits, and possible risks[98].

Surveillance: In patients who have undergone previous surgery for CRC, colonic and extracolonic surveillance are needed because more than 50% of recurrent tumors will present as extracolonic metastatic disease (particularly in the liver parenchyma), and many local recurrences lack an intraluminal colonic component[99]. The current surveillance guidelines include a combination of clinical assessment, serum carcinoembryonic antigen (CEA) testing, CS, and contrast-enhanced CT[100]. CTC, if performed by using a different technical strategy (i.e., during the intravenous administration of iodinated contrast medium), combines the ability of detecting polyps and cancer with an accuracy similar to CS, while simultaneously offering an evaluation of extracolonic findings[101]. However, this approach is presently recommended only if a patient is unable or unwilling to undergo CS because of the lack of robust and evidence-based data.

Patients who have previously undergone endoscopic resection of colonic adenoma are likely to develop a metachronous lesion; therefore, these patients must be included in a surveillance program using CS. The frequency intervals for follow-up remain controversial[102]; these are based on the findings at index CS (size, number, and histology of the removed polyp/s). Unfortunately, despite the official recommendations, patient adherence to follow-up is extremely variable and is generally poor in clinical practice[103]. For those patients unwilling to undergo CS, follow-up CTC can be suggested as an alternative option.

Acute abdominal conditions, like diverticulitis or the acute phase of IBDs, are contraindications to CTC because of the high risk of complications (i.e., perforations)[84].

Further, precautions should be taken when performing CTC after endoscopic resection. A two-week delay is suggested[6], although clear scientific evidence is lacking to confirm this. In fact, a study on patients who underwent CTC within 24 h following CS with either polypectomy or biopsy sampling did not find colonic perforation[27]. On the other hand, endoscopic biopsy is not considered a contraindication and same-day CTC can be safely performed.

CTC should be also avoided as a surveillance test in patients with a long-standing history of ulcerative colitis or Crohn’s disease and in those with hereditary non-polypoid colorectal cancer (HNPCC), Lynch syndrome, and APC-associated polyposis conditions. In all of these cases, CS is the preferred diagnostic option at a timing and interval different from screening in average- or higher-than-average risk individuals because of the highly increased risk of developing CRC[104,105].