EDITORIAL EDITORIALS Fecal DNA Biomarkers for ...

Fecal DNA Biomarkers for the Detection of Colorectal Neoplasia: Attractive, but Is It Feasible?

EDITORAL S

Dean E. Brenner, Gad Rennert Because of the high mortality associated with advanced disease, accessibility to direct visual examination, extensive research, and recognition of the genetic mechanisms associated with its carcinogenesis, the colon and rectum have emerged as an important model for the introduction of innovative and effective cancer screening and early detection tools for large populations. Detection of Colorectal Neoplasia: Need for Improved Methods

Fecal Based DNA Tests for the Detection of Colorectal Neoplasia

Affiliations of authors: Departments of Internal Medicine and Pharmacology, University of Michigan Medical Center, Ann Arbor (DEB); VA Medical Center, Ann Arbor, MI (DEB); Clalit Health Service National Israeli Cancer Control Center, Haifa (GR); Department of Community Medicine and Epidemiology, Carmel Medical Center and B. Rappaport Faculty of Medicine, Technion, Haifa, Israel (GR). Correspondence to: Dean E. Brenner, MD, 2150 CCGC, University of Michigan Cancer Center, Ann Arbor, MI 48109-0930 (e-mail: dbrenner@umich. edu). See “Note” following “References.”

Fecal sample testing using molecular diagnostic tests is emerging as a potentially important new approach that has the

DOI: 10.1093/jnci/dji244 © The Author 2005. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: [email protected].

Journal of the National Cancer Institute, Vol. 97, No. 15, August 3, 2005

EDITORIALS

1107

Downloaded from http://jnci.oxfordjournals.org/ by guest on June 8, 2013

Wide-scale screening using fecal occult blood testing (FOBT) results in 15%–33% reduction in colorectal adenocarcinoma mortality, but at the expense of many unneeded colonoscopies (1–6). Despite specificity of approximately 95% and reasonable costeffectiveness, FOBT’s test sensitivity ranges from 15% to 30% (7), leaving room for substantial improvement. Traditionally low FOBT sensitivities, reflecting standard office practice (8,9), may be improved when employed in an organized, structured program (10,11). The addition of sigmoidoscopy to FOBT increases the detection of adenomas by approximately twofold over FOBT alone, but there are no data demonstrating enhanced mortality reduction (7). Colonoscopy, with estimated sensitivity exceeding 90% for detection of adenocarcinoma and large adenomas and specificity exceeding 99%, requires a thorough bowel preparation and sedation, causes patient discomfort and small but nonnegligible risk of major complications (7,12–17). In a recent analysis, colonoscopy has a reported cost-effectiveness of $17 010 incremental costs per life-year gained (18). Although differences in model assumptions will result in different incremental costs per life-year gained, colonoscopy is considered to have borderline cost-effectiveness as a screening tool for detection of adenocarcinoma of the colon. Public health theory defines the optimal screening test as one that is easy to conduct, cheap, noninvasive, and acceptable to the population. Most screening procedures in large, otherwise healthy populations are tedious, low yield, high liability, financially unrewarding procedures to health care professionals (19). The cost, morbidity, and burdens on the medical care system if colonoscopic screening were to be implemented in large populations provide a strong rationale for developing early detection biomarkers that can be easily and cost-effectively deployed. Such biomarkers should detect transformed and pretransformed neoplasms, with high specificity, reducing the need for repetitive invasive procedures such as endoscopies. Optimally, screening should be achieved with a one-time test. In reality, the dynamic carcinogenesis process over time probably calls for periodic screening. Individualization based upon risk profiles may help develop costeffective sequences of multiple screening procedures (for example, FOBT, fecal DNA testing, and colonoscopy) in the future.

potential of providing accurate and cost-effective, early detection of colorectal neoplasia. Details of many of the currently employed and novel approaches have been recently reviewed (20). The most prominent of genetically based fecal tests exploits the concept of chromosomal instability with mutations progressively accumulating in the adenomatous polyposis coli, p53 tumor suppressor genes, and the K-ras oncogene (21). This test relies upon preservation of naked DNA in human stool samples. The test requires a large-volume fecal sample from which purified DNA is prepared using oligonucleotide-based hybrid capture. The genetic mutation sites are detected and quantified with real time polymerase chain reaction (21). Preliminary publications in small trials (16–65 subjects) reported test sensitivity ranging from 62% to 91% for adenocarcinoma detection and 27% to 82% for adenoma detection, with specificity ranging from 93% to 98% (20). Validation of these preliminary data in a large (4404 evaluated subjects), prospective colorectal cancer screening trial resulted in a sensitivity of 51.6% (95% confidence interval [CI] = 34.8% to 68.0%) for detection of adenocarcinoma; 15.1% (95% CI = 12.0% to 19.0%) for detection of adenomas ≥1 cm, with double the sensitivity for dysplastic adenomas. Specificity for the fecal DNA test was 95.2% (21). These data support the concept that fecal DNA tests can enhance single-test sensitivity to FOBT and serve as an intermediate, noninvasive screening tool for colorectal adenocarcinoma. The cost-effectiveness Markov model employed by Song et al. (18) assumed the diagnostic sensitivity of the multigene fecal DNA panel to be 65% for adenocarcinoma and 40% for colorectal adenomas and an estimated base cost of $695 per test from the commercial vendor. The model generated a cost of $47 700 per incremental life-year gained. For fecal DNA testing to become cost effective, the base cost would need to be reduced to $195 per test using sensitivity assumptions that were 10% higher than the recently published prospective screening trial (18). Despite its novel technologic and genetic approach, as currently priced, the increased sensitivity of the fecal DNA panel developed by Exact Sciences (21) compared with that for the FOBT test seems insufficiently cost-effective to merit inclusion in population-based screening for colorectal adenocarcinoma.

Improving Upon Current Fecal DNA Diagnostic Tools

1108

EDITORIALS

REFERENCES (1) Mandel JS, Bond JH, Bradley M, Snover DC, Church TR, Williams S, et al. Sensitivity, specificity, and positive predictivity of the Hemoccult test in screening for colorectal cancers. The University of Minnesota’s Colon Cancer Control Study. Gastroenterology 1989;97:597–600. (2) Mandel JS, Bond JH, Church TR, Snover DC, Bradley GM, Schuman LM, et al. Reducing mortality from colorectal cancer by screening for fecal occult blood. Minnesota Colon Cancer Control Study [published erratum appears in N Engl J Med 1993;329:672]. N Engl J Med 1993;328:1365–71. (3) Kronborg O. Screening for colorectal cancer with Hemoccult-II in the average risk population. Gastroenterologie Clinique et Biologique 1998;22 (3 Suppl):S44–8. (4) Kronborg O, Fenger C, Olsen J, Jorgensen O, Sondergaard O. Randomised study of screening for colorectal cancer with faecal-occult-blood test. Lancet 1996;348:1467–71. (5) Rennert G. Fecal occult blood screening—trial evidence, practice and beyond. Recent Results Cancer Res 2003;163:248–53; discussion 264–6. (6) Hardcastle JD, Chamberlain JO, Robinson MH, Moss SM, Amar SS, Balfour TW, et al. Randomised controlled trial of faecal-occult-blood screening for colorectal cancer. Lancet 1996;348:1472–7. (7) Pignone M, Rich M, Teutsch SM, Berg AO, Lohr KN. Screening for colorectal cancer in adults at average risk: a summary of the evidence for the U.S. Preventive Services Task Force. Ann Intern Med 2002;137:132–41. (8) Nadel MR, Shapiro JA, Klabunde CN, Seeff LC, Uhler R, Smith RA, et al. A national survey of primary care physicians’ methods for screening for fecal occult blood. Ann Intern Med 2005;142:86–94. (9) Collins JF, Lieberman DA, Durbin TE, Weiss DG. Accuracy of screening for fecal occult blood on a single stool sample obtained by digital rectal examination: a comparison with recommended sampling practice. Ann Intern Med 2005;142:81–5. (10) Rennert G, Rennert HS, Miron E, Peterburg Y. Population colorectal cancer screening with fecal occult blood test. Cancer Epidemiol Biomarkers Prev 2001;10:1165–8. (11) Church TR, Ederer F, Mandel JS. Fecal occult blood screening in the Minnesota study: sensitivity of the screening test. J Natl Cancer Inst 1997;89: 1440–8.



Enhancement of the current state of fecal DNA testing for biomarkers of colorectal neoplasia might follow several promising future directions designed to improve the sensitivity of current genetic panels. First, enhanced sensitivity may be achieved through improved stool sample quality. It was surprising that long DNA, thought to reflect disordered apoptosis and found in preliminary data to be the most sensitive portion of the Exact Sciences fecal DNA panel (22–24), was the weakest biomarker in the large, multicenter, prospective screening trial (21). A potential explanation for these results might rest in the handling procedures of stool DNA after defecation, which can be easily digested by DNAses present in the feces. Enhanced preservation after collection may reduce DNA degradation, although considerable degradation in formed and unformed fecal matter may have occurred prior to defecation. Sample quality might also be improved through the reduction of the volume of stool required for completion of the analytical procedures. Second, commercial methods of purifying, cleaning, and maintaining viable exfoliated colonocytes in the feces are available (25). The increase in cellular shedding from colorectal neoplasms compared with normal colorectal cellular turnover provides a greater percentage of intact neoplastic cells than flat colonic mucosa (26). The problem of DNA degradation is rendered moot by the use of intact, viable colonocytes exfoliated in human stool. Third, the sensitivity of any fecal DNA testing panel that relies on the detection of mutations in selected genes rests upon the frequency of the specific mutations in cells shed from neoplasias that will bind to hybridizing oligonucleotides. The limited number of hybridizing oligonucleotides in a fecal DNA test reduces the likelihood that a neoplasia-associated mutation might be found due to the large number and heterogeneity of mutational events occurring in human neoplasia. Fourth, a fecal DNA testing panel employing nucleotide probes will not detect important epigenetic events associated with human carcinogenesis. The manuscript of Chen et al. in this issue of the Journal (27) provides a novel solution to the detection of epigenetic events in fecal DNA that potentially transcends the problem of inter- and intratumoral mutational variation. In this remarkably complete biomarker discovery and prevalidation assessment, the authors report that a dense CpG dinucleotide region upstream of the first exon and continuing across the first two-thirds of the first exon of a normally unexpressed and unmethylated gene in colorectal epithelial crypt cells, vimentin, becomes methylated in colorectal cancers. The gene is methylated in between 53% and 83% of all colon cancer samples examined in approximately 150 samples reported. This assay can be reproduced in fecal DNA. In a well-designed training and test set sensitivity analysis, the authors report a diagnostic sensitivity of 46% and specificity of 90%. The biomarker appears to be equivalently sensitive for late stage (III or IV) as it is for early stage (I and II) colorectal adenocarcinomas. Moreover, the biomarker is equivalently sensitive for proximal or distal cancers. Although these preliminary data compare favorably with results from the recently published larger fecal DNA panel (21), the methylated vimentin biomarker requires a large cancer screening sample validation prior to implementation as part of a fecal DNA test panel. It is surprising that this methylation-based gene that is not generally associated with a role in driving colorectal cellular initiation and promotion of colorectal carcinogenesis is aberrantly

methylated in many such cancers. Could this information point to new mechanisms of colonic carcinogenesis in humans? In addition to the data published in this issue of the Journal, recently published data evaluating fecal methylated DNA detection of three genes known to drive colorectal cellular initiation and promotion (CDKN2A, MGMT, and MLH1) from individuals with adenomas detected 49%, 34%, and 7% of adenomas, respectively. No information on detection of adenocarcinomas was provided (28). The impact of adding methylation biomarkers to mutationbased fecal DNA biomarkers remains unknown. Nevertheless, we believe that genetic methylation based biomarker panel should be pursued as part of a fecal DNA screening test for colorectal neoplasia screening. Prior to the addition of markers into a fecal DNA panel, the problems of analytical scale-up, development of high-throughput, low-cost analytical methods must be addressed for methylation-based or other novel biomarkers that might interrogate the diverse signaling pathways associated with colorectal carcinogenesis can be assayed. The age, risk group, and testing frequency of a fecal DNA diagnostic panel still needs exploration. Although detection of early stage invasive adenocarcinoma presents an important screening goal, future efforts should be directed toward detection of dysplastic cells present in adenomas. After all, management of colorectal adenomas is less morbid, and of much lower psychological and physical impact on the patient and family members than the management of invasive adenocarcinoma.

(22) Ahlquist DA, Skoletsky JE, Boynton KA, Harrington JJ, Mahoney DW, Pierceall WE, et al. Colorectal cancer screening by detection of altered human DNA in stool: feasibility of a multitarget assay panel. Gastroenterology 2000;119:1219–27. (23) Syngal S, Chung D, Willet C, Schoetz D, Schroy P, Stoffel E, et al. Loss of stool DNA mutation abnormalities in colorectal neoplasia after treatment. Gastroenterology 2003;124:A32. (24) Tagore KS, Lawson MJ, Yucaitis JA, Gage R, Orr T, Shuber AP, et al. Sensitivity and specificity of a stool DNA multitarget assay panel for the detection of advanced colorectal neoplasia. Clin Colorectal Cancer 2003;3:47–53. (25) Nair P, Lagerholm S, Dutta S, Shami S, Davis K, Ma S, et al. Coprocytobiology: on the nature of cellular elements from stools in the pathophysiology of colonic disease. J Clin Gastroenterol 2003;36(5 Suppl):S84–93; discussion S94–6. (26) Ahlquist DA, Harrington JJ, Burgart LJ, Roche PC. Morphometric analysis of the “mucocellular layer” overlying colorectal cancer and normal mucosa: relevance to exfoliation and stool screening. Hum Pathol 2000;31:51–7. (27) Chen WD, Han ZJ, Skoletsky J, Olson J, Sah J, Myeroff L, et al. Detection in fecal DNA of colon cancer-specific methylation of the nonexpressed vimentin gene. J Natl Cancer Inst 2005;97:1124–32. (28) Petko Z, Ghiassi M, Shuber A, Gorham J, Smalley W, Washington MK, et al. Aberrantly methylated CDKN2A, MGMT, and MLH1 in colon polyps and in fecal DNA from patients with colorectal polyps. Clin Cancer Res 2005;11:1203–9.

NOTE Supported by Early Detection Research Network, CA86400. Dr. Brenner was supported by a fellowship from the Lady Davis Foundation, Jerusalem, Israel.


EDITORIALS

1109


(12) Winawer SJ. Appropriate intervals for surveillance. Gastrointest Endosc 1999;49(3 Pt 2):S63–6. (13) Winawer SJ, Fletcher RH, Miller L, Godlee F, Stolar MH, Mulrow CD, et al. Colorectal cancer screening: clinical guidelines and rationale [published errata appear in Gastroenterology 1997;112:1060 and 1998;114:625]. Gastroenterology 1997;112:594–642. (14) Winawer SJ, St. John DJ, Bond JH, Rozen P, Burt RW, Waye JD, et al. Guidelines for the prevention of colorectal cancer: update based on new data. World Health Organization Collaborating Center for the Prevention of Colorectal Cancer. Zeitschrift fur Gastroenterologie 1995;33:574–6. (15) Winawer SJ, Zauber AG, Ho MN, O’Brien MJ, Gottlieb LS, Sternberg SS, et al. The National Polyp Study. Eur J Cancer Prev 1993;2 Suppl 2:83–7. (16) Winawer SJ, Zauber AG, Ho MN, O’Brien MJ, Gottlieb LS, Sternberg SS, et al. Prevention of colorectal cancer by colonoscopic polypectomy. The National Polyp Study Workgroup. N Engl J Med 1993;329:1977–81. (17) Winawer SJ, Zauber AG, O’Brien MJ, Ho MN, Gottlieb L, Sternberg SS, et al. Randomized comparison of surveillance intervals after colonoscopic removal of newly diagnosed adenomatous polyps. The National Polyp Study Workgroup. N Engl J Med 1993;328:901–6. (18) Song K, Fendrick AM, Ladabaum U. Fecal DNA testing compared with conventional colorectal cancer screening methods: a decision analysis. Gastroenterology 2004;126:1270–9. (19) Washam C. Mammographers’ ranks shrink as demand for breast images grows. J Natl Cancer Inst 2005;97:792–3. (20) Osborn NK, Ahlquist DA. Stool screening for colorectal cancer: molecular approaches. Gastroenterology 2005;128:192–206. (21) Imperiale TF, Ransohoff DF, Itzkowitz SH, Turnbull BA, Ross ME. Fecal DNA versus fecal occult blood for colorectal-cancer screening in an average-risk population. N Engl J Med 2004;351:2704–14.