CT-colonography (virtual colonoscopy) as triage technique in fecal occult blood test screening for prevention of colorectal cancer

Colorectal carcinoma (CRC) is one of the major causes of death in the Netherlands, accounting for 4383 deaths in 2002. It is the second leading cause of cancer related death. In the Netherlands prevention of CRC is addressed through the development of an invitation based population screening with fecal occult blood test (FOBT). A ZonMW granted FOBT-pilot study will start in Amsterdam and Nijmegen. An important problem of FOBT screening is that more than half of the positive FOBT results is false positive. This leads to an unnecessary use of (limited) colonoscopy capacity and other scarce resources. It also burdens participants, labeling them as positive screenees and referring then for further work-up that will identify them as false-positives. Triage, using computed tomography colonography with limited bowel preparation (CTC; virtual colonoscopy), is a directly available solution for this problem. If a substantial number of FOBT positive screening participants are identified by CTC as false positives, they do not have to undergo colonoscopy. So far, CTC has only been studied in symptomatic populations or low prevalence populations, not in FOBT positive screening participants, in which both the prevalence and spectrum of disease are likely to differ, affecting the test characteristics. We intend to study whether CTC is a suitable screening triage. In 300 consecutive FOBT-positive participants, sampled from the ZonMW granted FOBT-pilot population project, we will evaluate the test characteristics, acceptance and efficiency of CTC. CTC results will be prospectively compared to colonoscopy findings (segmental unblinding). Based on the literature a disease prevalence of approximately 40% (CRC and polyps larger or equal 10 mm) is expected. With inclusion of 120 participants with a positive CT result the one sided confidence interval will extend 2% from the expected negative predictive value of 98%. Given the disease prevalence and expected CTC test characteristics in total 300 participants will have to be included. The expected negative predictive value of CTC is certainly acceptable for triage, especially given the limitations of reference standard colonoscopy. For a provisional projection of the potential contribution of CTC to FOBT screening we have to assume that published results (high negative predictive value and sensitivity) in other populations are applicable. In this theoretical situation, CTC might prevent approximately 90% of the unnecessary colonoscopies performed in participants with a false positive FOBT result. The Dutch FOBT-pilot starting in March 2006 gives the possibility to perform this research at this moment, which has major advantages. Firstly, the results can directly be taken into account during the decision process on FOBT-screening for prevention of CRC in the Netherlands. Secondly, the use of the already ZonMW granted FOBT infrastructure is a substantial budgetary advantage.