Colorectal cancer (CRC) is the second leading cause of cancer-related death in Europe and the USA . It is the most frequently diagnosed cancer in Europe for both genders combined, with more than 400,000 new cases and more than 200,000 deaths in 2008 . About 46,000 incident cases, 267,000 prevalent cases, and 16,000 deaths from CRC are estimated in Italy for the year 2005 . The lifetime risk of CRC in western countries is about 5%. Population screening in asymptomatic individuals at average risk for CRC reduces mortality, both through the detection of malignancies at earlier, more treatable stages, and through the identification and removal of adenomatous polyps (pre-cancerous lesions that may evolve to CRC) [3–9]. In addition, given the increase in the costs of CRC treatment, screening has actually become cost-saving . There are several tests currently available for screening in the general population, including stool tests, such as variants of the fecal occult blood test (FOBT), and structural examinations such as flexible sigmoidoscopy (FS), colonoscopy, and CT colonoscopy (CTC). Each screening strategy has its advantages and drawbacks.
The simplest and best-evaluated available screening method is the FOBT, which is relatively inexpensive and non-invasive, but less accurate than structural examinations. The guaiac-based FOBT (gFOBT) may reduce cancer mortality by up to 20% if offered biennially [4, 5, 11], and possibly more if offered annually . The newer immunochemical FOBT (iFOBT) is considered to perform better than gFOBT in detecting advanced neoplasia . The disadvantages of FOBT include its low sensitivity for adenoma and the requirement for frequent testing, which may limit compliance and thereby effectiveness. Furthermore, repeat testing leads to high positivity rates.
FS is an endoscopic procedure, in which the distal 40 to 60 cm of the colon is inspected. Total colonoscopy is advised in cases of positive findings. FS is less invasive than colonoscopy, and requires only a simple bowel preparation of a single enema within 2 hours prior to the procedure. Three European randomized control trials (RCTs) on FS have been performed [9, 13, 14]. In the UK, one-time screening with FS significantly reduced the incidence of CRC (by 26%) and associated mortality (by 36%) . In Italy, an 18% reduction in incidence and a non-significant 22% reduction in mortality were reported , whereas in Norway, no benefit was observed after 7 years of follow-up . A disadvantage of FS is that it does not examine the proximal colon, but distal findings are used to select a higher-risk group for colonoscopy. Another consideration is that its sensitivity depends on the varied experience of the examiners, that has a major impact on the adequacy of mucosal inspection.
Colonoscopy and CTC are structural examinations allowing inspection of the complete colon, and enabling early detection of advanced adenomas (that is,, adenomas ≥10 mm or with unfavorable histological features) and CRC . Although colonoscopy is the most complete endoscopic procedure available for CRC screening [17, 18], direct evidence about its effectiveness, complications, and acceptability among individuals at average risk of CRC is still not adequate to justify its use for population screening . Observational evidence suggests that colonoscopy might not be as effective in the proximal colon as in the other segments of the colon and rectum [20, 21]. Furthermore, colonoscopy capacity is a limiting factor for its widespread use as a primary screening test.
CTC may represent a reasonable alternative for colonoscopy . First, this technique has a very high sensitivity for already-developed CRC (96%) . Secondly, results from a large study of asymptomatic average-risk individuals published in 2003 showed the diagnostic performance of CTC for clinically relevant polyps to be equivalent to that of colonoscopy . Results from the largest screening study (over 2,500 participants)  showed 90% sensitivity of CTC for polyps 10 mm or larger and 86% specificity; the positive and negative predictive values were 23% and 99%, respectively. Important advantages of CTC over colonoscopy are its minimally invasive nature (only a small-caliber flexible rectal catheter is needed for colonic distension) and the use of limited bowel preparation [25, 26]. The risk of complications from CTC is extremely low, particularly in asymptomatic individuals [27–30]. CTC with limited bowel preparation has a lower burden and is preferred by patients compared with CTC with full cathartic preparation [31, 32]. Potentially, the addition of CTC to CRC screening options could have a marked effect on current low adherence rates, likely in a cost-effective manner [33–35]. Indeed, results from a recent trial  on the use of colonoscopy versus CTC for population-based screening for CRC showed that participation with CTC was substantially higher than with colonoscopy, and thus led to similar advanced neoplasia detection rates. Disadvantages of CTC include the exposure of individuals to ionizing radiation. However, the chances of radiation-induced malignancy are considered very low, especially when a low-dose protocol is used. Furthermore, screening with CTC requires subsequent colonoscopy if lesions are detected. A high referral rate to colonoscopy might increase examination costs. CTC displays the abdominal organs external to the colon. The prevalence of extracolonic diseases requiring further investigation is not negligible, occurring in approximately 6% of the individuals in asymptomatic average-risk populations [25, 36]. It is generally agreed that relevant extracolonic findings should be reported . However, although the detection of relevant extracolonic diseases at CTC could be beneficial , the risks and costs associated with false-positive results and inconsequential findings may be substantial .
Data concerning the neoplasia yield and the acceptability of CTC compared with colonoscopy have been recently reported from the CoCos trial , but no information is available concerning the performance of CTC compared with the recommended screening tests, namely, iFOBT and FS. In European countries where national screening programs are ongoing, national policy makers would require comparative data concerning the performance of CTC (as of any other new screening test) assessed against existing strategies before considering the introduction of this test as an available screen for average-risk individuals. A RCT, the SAVE Trial, has recently been designed and funded in Italy to compare CTC and iFOBT in a screening setting .
We are currently conducting a multicenter RCT to compare the performance of CTC and FS in the context of the population-based screening programs offering FS once in a person’s lifetime in the Piedmont region and the province of Verona in Italy. These organized programs were started in 2004, targeting all men and women, aged 58 years (in Piedmont region) or 60 years (in Verona) who are at average risk of CRC. Individuals who do not respond to the FS invitation are offered iFOBT as alternative.
Several issues need however to be addressed before the implementation of a program using CTC as a primary CRC screening test. First, the local availability and expertise of radiologists may affect the feasibility of CTC screening. The use of information technology (IT) infrastructures may allow subjects to undergo CTC at their nearest imaging center, while test interpretation could take place at a centralized level, requiring a smaller number of readers. Implementing telediagnosis should thus ensure high reporting quality, provided that highly qualified radiologists, certified to report CTC, are selected. The local availability of radiological centers could also favor greater participation of individuals in the screening programs.
Secondly, it is widely recognized that the interpretation of CTC is challenging, probably more so in a low-prevalence population. The need to view a large number of images to detect a small number of clinically significant lesions, the subtle nature of many radiological characteristics of colonic lesions, and radiologist fatigue or distraction may lead to an undesirable rate of false-negative findings [41, 42]. Computer-aided detection (CAD) has the potential to improve the cost-effectiveness of CTC by increasing detection of clinically significant lesions and/or reducing reporting times . CAD as a second reader (that is, CAD is applied after a complete and unaided assessment) has been shown to increase reading sensitivity, albeit at the cost of increasing reading times [44–47]. This increase in reading times is generally undesirable if a large number of cases must be read sequentially, as occurs for population screening. A potentially more time-efficient paradigm is first-reader CAD, in which the reader’s interpretation is restricted to CAD prompts alone . However, detection of colonic findings typically not targeted by CAD systems, such as masses or atypical lesions, poses challenges .
The limitation of CAD for mass detection provides motivation for double-reading first-reader CAD (DR FR-CAD paradigm in which first-reader CAD is followed by a rapid two-dimensional review of unprompted areas of the colon, searching for masses or larger lesions missed by CAD . The addition of the human component to first-reader CAD as a rapid control for eventual CAD errors may overcome the issues related to automatic detection. We hypothesized that DR FR-CAD could play a key role in CTC screening by improving detection sensitivity compared with unassisted reading, and by shortening interpretation time (while maintaining equally high sensitivity) compared with second-reader CAD. Thus we performed two preliminary studies [50, 51] to investigate the feasibility of using DR FR-CAD as a possible reading strategy for CTC screening and to compare the diagnostic performance and time-efficiency with those of second-reader CAD for CRC screening . Participants included in this preliminary study underwent limited bowel preparation with fecal tagging, as the available evidence showed that limited preparation was not associated with a decrease in the CTC diagnostic performance compared with full-cathartic preparation .
This multicenter RCT, assessing an average-risk population aged 58 to 60 years, will compare the detection rates of advanced neoplasia (advanced adenoma and CRC) of CT colonography and sigmoidoscopy, and the participation rates for the two programs. The data from our preliminary research have allowed us to define a CTC screening protocol (adopting an organizational model based on telediagnosis together with DR FR-CAD for CTC interpretation). The cost-effectiveness of this protocol will also be tested in the proposed RCT.