Abstract & Commentary
Synopsis: Most neoplastic lesions are not detected either by Hemoccult screening or by multitarget analysis of fecal DNA. Nevertheless, the fecal DNA analysis was significantly more efficient than Hemoccult testing.
Source: Imperiale TF, et al. N Engl J Med. 2004; 351;26:2704-2714.
Colorectal cancer is a serious health problem, comprising the second leading cause of cancer death in adults, and over 55,000 deaths from colon cancer were anticipated in 2004. Despite strong recommendations, Imperiale et al state that less than 40% of the population undergoes colorectal cancer screening. Guaiac-based detection of occult blood is known to be an effective screening modality that reduces cancer incidence and risk of death. Nevertheless, even annual fecal occult blood testing detects only 25-50% of colon cancers and less than 10% of colon adenomas (known precursors of most colon cancers). Frequent false positive occult blood results also occur. Recent advances in the understanding of the molecular genetics of colon cancer have led to the discovery of multiple DNA markers associated with colonic neoplasia. Since exfoliated epithelial cells from cancers and adenomas will be contained in feces, the concept of fecal DNA testing has great theoretical appeal.
The current study involved 81 study sites. All these participating average risk subjects provided stool samples for DNA abnormalities (21 different targets) and also performed Hemoccult II tests prior to colonoscopy (done without knowledge of fecal DNA test results). Four thousand four hundred and four subjects were evaluable. Fecal DNA testing detected 16 of 31 invasive cancers vs 4 identified with Hemoccult testing (56% vs 12.9%). Thirteen cancers were found by DNA testing that had been missed by Hemoccult, and 1 cancer identified by Hemoccult testing had been missed by fecal DNA testing. Adenomas with dysplastic features were also detected more often by DNA testing (32.5% vs 15%). Specificities of tests were 92.4% for fecal DNA testing vs 95.2% for Hemoccult (ie, negative findings in patients who had normal colonoscopies).
Comment by Malcolm Robinson, MD, FACP, FACG
Although the sensitivity of the panel of fecal DNA markers was 4 times that of Hemoccult II for invasive cancer and twice as sensitive for adenomas with high grade dysplasia with no loss of specificity, this approach to colorectal cancer screening is still not ready for prime time use. As pointed out in an accompanying editorial, this test has a number of significant drawbacks. Previous data from Ahlquist et al1 indicated 90% sensitivity of fecal DNA testing for cancer and 82% for advanced adenomatous polyps. These somewhat better results may relate to the rate of cancer and advanced adenomas in the respective populations studied. There is no doubt that physicians and their patients need a dependable, inexpensive, and culturally acceptable technique for colorectal cancer screening. It isn't here yet. Virtual colonoscopy continues to provoke interest, but it remains far less accurate than colonoscopy and too expensive and unacceptably uncomfortable. Dr. Woolf points out that the fecal DNA data indicate sensitivity far lower than conventional colonoscopy (perhaps less efficient screening than flexible sigmoidoscopy).2 Low prevalence of colorectal cancer in asymptomatic individuals (240 cases/100,000 individuals 50 to 59 years of age) means that colorectal cancer will be diagnosed in only 2% of adults who have a positive fecal DNA test. The remaining 98% may remain terrified that the positive screening test indicates cancer that somehow was missed at colonoscopy. Fecal occult blood testing costs somewhere between $6000 and $18,000 per year of life gained. Use of the fecal DNA panel for screening would be at least 200 times more expensive. For some time to come, it will be hard to beat old fashioned expensive sedated colonoscopy as the best available test for screening and therapy that may avert cancer development altogether.
Dr. Robinson, Medical Director, Oklahoma Foundation for Digestive Research; Clinical Professor of Medicine, University of Oklahoma College of Medicine Oklahoma City, OK, is Associate Editor of Internal Medicine Alert.
1. Ahlquist DA, et al. Colorectal Cancer Screening By Detection of Altered DNA in Stool: Feasibility of a Multitarget Assay Panel. Gastroenterology 2000;119: 1219-1227.
2. Woolf SH. A Smarter Strategy? Reflections on Fecal DNA Screening For Colorectal Cancer. N Engl J Med. 2004;351:26:2755-2758.