Abstract CN14-06: Screening for colorectal cancer: 2009

Abstract

Abstract The objectives of this session are to: 1) review the new guidelines for average-risk colorectal cancer (CRC) screening; 2) review recent selected studies of CRC screening for currently recommended tests, and; 3) review recent decision analyses and a cost-effectiveness analysis on screening. Guidelines: Less than 2 years ago, colorectal cancer (CRC) screening guidelines were highly concordant and flexible. New guidelines issued by the American Cancer Society/American College of Radiology/Multisociety Task Force in March 2008 (1), the U.S. Preventive Services Task Force in November 2008 (2) and the American College of Gastroenterology in March 2009 (3) are now disparate, with two of three organizations implicitly or explicitly preferring colonoscopy despite the absence of published evidence supporting this preference. Two decision analyses, commissioned by the USPSTF, show four screening strategies with equivalence in CRC mortality reduction and life-years gained: colonoscopy every 10 years; a sensitive guaiac-based fecal occult blood test (FOBT) annually; fecal immunochemical testing (FIT) annually, and; the combination of flexible sigmoidoscopy every 5 years with a mid-interval sensitive FOBT or FIT. (4) Screening tests: For fecal DNA, a large screening study comparing a multi-component panel to Hemoccult II with colonoscopy as the reference standard showed that fecal DNA detected 52% of CRCs compared to 13% for Hemoccult II; the two tests had comparable specificity. Recent case-control studies of fecal DNA using somewhat different markers show a single application cancer sensitivity of 83–88% and specificity of 82%.(5) Yet another version of fecal DNA showed cancer sensitivity and specificity of 58% and 84%, respectively; sensitivity for adenomas 1 cm or larger was 46%.(6) The main issues limiting wider use of fecal DNA are its uncertain test interval and greater cost compared to other non-invasive tests. Computer tomographic colonography (CTC or “virtual colonoscopy”) continues to demonstrate good-to-excellent test characteristics for CRC and large (1 cm or larger) adenomas. The ACRIN trial reaffirmed CTC's test characteristics, and may have enhanced prospects for generalizability of test performance. (7) Ongoing issues for CTC include deciding on which CTC colonic findings should be reported and which should lead to colonoscopy (polyp size, number); radiation dose; the clinical and economic impact of extracolonic findings; cost-effectiveness of CTC; and whether it would increase population adherence to CRC screening. Immunochemical FOBTs (or FITs) were developed to enhance the test characteristics of guaiac-based tests. FITs use monoclonal or polyclonal antibodies to detect the intact portion of human hemoglobin and are specific for occult bleeding from a lower GI source. Several FITs are currently available for CRC screening as qualitative tests. Few studies of FITs use colonoscopy as a reference standard; the few that do show widely varying test characteristics for cancer and advanced adenomas, although these are generally better than those of guaiac-based FOBTs. (8) Studies comparing guaiac-based FOBTs with FITs show higher acceptance rates as well as higher detection rates for cancer and advanced adenomas. (9) The real potential of FITs may be in their use as quantitative tests, as data from Levi and colleagues (10) suggest, although more investigation is required to determine the optimal number of single-application tests and the threshold for a positive test. A recent cost-effectiveness analysis by Parekh and colleagues (11) suggests that, as CRC treatment costs increase, screening with FIT may be cost-saving. This analysis also showed that annually FIT screening dominated colonoscopy screening every ten years, meaning that it was both less costly and more effective, suggesting that annual FIT may be “better than” colonoscopy when FIT adherence is high. Until just a few months ago, the effectiveness of sigmoidoscopy was supported only by high-quality case-control studies. Seven-year interim findings from NORCCAP, one of 4 ongoing trials of sigmoidoscopy, were published earlier this year. (12) Study findings included no difference in cancer incidence (not unexpected given the relatively short follow-up); no difference in CRC mortality by intention-to-treat; and a 60% per protocol reduction in overall CRC mortality, including a 75% mortality reduction from recto-sigmoid cancer. Colonoscopy is currently the most “popular” CRC screening test, despite an absence of data demonstrating its superiority over other tests. Within gastroenterology, much attention is focused on monitoring the performance of colonoscopy through parameters such as the extent of examination (to the cecum and by which landmarks?); withdrawal time spent examining the mucosa; and adenoma detection rate. Evidence for the effectiveness of colonoscopy in reducing CRC incidence and mortality is indirect. Follow-up of the National Polyp Study cohort suggests a 76–90% reduction in CRC incidence when compared to 3 reference populations. (13) Kahi and colleagues compared the observed CRC rate in a cohort of screened persons with a median of 7 years follow-up and found a significant reduction in CRC incidence of 67% (95% CI, 38–90%) when compared with SEER data; CRC mortality was reduced by 65%, though this finding was not statistically significant. (14) The degree and duration of protection of colonoscopy from CRC has been questioned in two recent studies. In a population-based case-control study, Baxter and colleagues found that colonoscopy reduced CRC mortality by 37% overall. Colonoscopy reduced left-sided CRC mortality by 67%, but reduced right-sided mortality by just 1%, a non-significant result. (15) In a population-based retrospective cohort study, Lakoff and colleagues looked at risk of CRC after a “negative” colonoscopy, finding that colonoscopy was protective later and less consistently for the proximal colon than for the distal colon. (16) Both technical and biological factors may explain the apparent and relative ineffectiveness of colonoscopy in the right colon in the Baxter and Lakoff studies, respectively. What's coming? Future studies are expected on how colonoscopy can improve detection of adenomas; on FIT test performance, both qualitative and quantitative; on CT colonography performance and acceptance, the latter of which would be enhanced by use of a “virtual prep”; on test performance and logistics of fecal DNA; and on blood-based biomarkers for CRC/advanced adenomas. Finally, we might also expect tailoring of both screening and surveillance based on improved risk stratification. Citation Information: Cancer Prev Res 2010;3(1 Suppl):CN14-06.