Gastrointestinal Endoscopy
Volume 71, Issue 2 , Pages 342-345, February 2010

The best screening test for colorectal cancer is the one that gets done well

Clinical Professor of Medicine Emeritus, University of California, San Francisco, Department of Medicine, Division of Gastroenterology and Hepatology, San Francisco General Hospital, San Francisco, California, USA

Adjunct Investigator, Kaiser Division of Research

Article Outline

Abbreviations: CRC, colorectal cancer, FIT, fecal immunochemical test, FOBT, fecal occult blood test, gFOBT, guaiac FOBT, sGT, sensitive guaiac test, USPSTF, U.S. Preventive Services Task Force

 

When all patients eligible for screening are screened with colonoscopy, the fraction with no colorectal neoplasia is consistent, ranging from 75% to 83%; thus, most patients screened with colonoscopy will have neither adenomas nor cancers.

As long as Medicare and insurance companies pay for screening colonoscopies in covered patients, incentives will exist for GI societies to continue to promote colonoscopy as the “best” screening test, though the evidence of its superiority is based solely on observational studies and inferential evidence and despite recent studies that have raised concerns regarding the effectiveness of colonoscopic screening.1, 2, 3 The 2008 United States Preventive Services Task Force (USPSTF) Colorectal Cancer (CRC) Screening Guidelines4 do not promote any test as best or claim that structural examinations are superior to fecal tests. Five multiple-strategy analyses by this organization and the Institute of Medicine did not uniformly find that population screening with colonoscopy was the most effective or cost-effective screening strategy. Similar conclusions were reached by members of the Institute of Medicine and the National Research Council in a workshop conference, the Economic Models of Colorectal Cancer Screening in Average-Risk Adults.5, 6 In the latest version of the USPSTF guidelines, a decision analysis using two colorectal cancer microsimulation models from the Cancer Intervention and Surveillance Modeling Network supported colorectal cancer screening with annual screening by using a sensitive fecal occult blood test (FOBT), sensitive guaiac test (sGT), or fecal immunochemical test (FIT) for patients from age 50 to 75 years. The analysis showed that flexible sigmoidoscopy and sensitive FOBT or sensitive FOBT alone provided life years saved almost comparable with that of colonoscopy, assuming that there was equally high adherence to all aspects of such a screening program.7

Screening programs in most other countries are FOBT-based like the one discussed in this issue of Gastrointestinal Endoscopy.8 These countries have carefully evaluated their screening resources and have reached the conclusion that population screening with optical colonoscopy would overwhelm national colonoscopy resources and potentially lead to significant complications in patients who are otherwise healthy.9, 10 The concern about costs, capacity, and complications of tests chosen for screening programs should be as great in the United States as it is in other countries. In the United States, the recession, the cost of health care, and the large number of jobless and uninsured make it necessary for every citizen and clinician to consider the costs of a CRC screening program using colonoscopy as the preferred screening test. If large numbers of gastroenterologists are required to do colonoscopies on the tens of millions eligible for screening, what won't they be able to do? Is creating more endoscopy time for gastroenterologists a societal benefit? Are not patients with serious GI diseases such as inflammatory bowel disease, hepatitis B and C with cirrhosis, pancreatitis and pancreatic cancer, peptic ulcer disease, and GERD equally worthy of quality time with a gastroenterologist?

Evidence suggests that the manpower necessary to provide a skilled colonoscopic examination for all eligible U.S. citizens is inadequate.11, 12 In such a situation, unqualified examiners could absorb the overflow, and the increased inaccuracy and complications could undo the small incremental benefit that the test might offer over other tests.13 Good alternative tests, such as FOBT, are available and can target those in the population most likely to harbor advanced neoplasia and, therefore, benefit from colonoscopy. When all patients eligible for screening are screened with colonoscopy, the fraction with no colorectal neoplasia is consistent, ranging from 75% to 83%. Thus, most patients screened with colonoscopy will have neither adenomas nor cancers.14

In population screening for CRC, the two most important outcomes are mortality reduction by early detection and cancer prevention by identification and removal of advanced adenomas. Although the American Cancer Society/Multi-society Taskforce guidelines say that by using FOBTs the “opportunity for prevention is both limited and incidental and not the primary goal of CRC screening with these tests,” a search of the literature reveals that this assertion cannot be substantiated for the newly recommended FOBTs—the sGT, Hemoccult Sensa (Beckman Coulter, Inc., Brea, CA), and the FIT. Population cohort studies with endoscopic verification of test-negative participants have shown an application sensitivity of the sGT and FIT for advanced adenomas ranging from 20% to 40%.1, 15, 16 An application sensitivity ranging between 20% and 40% for advanced adenomas is neither limited nor incidental. Furthermore, it is unfair to judge FOBTs on their application sensitivity alone. Their use is intended to be in a program of repeat testing, either annually or biannually. Programmatic testing leads to programmatic sensitivity that, in those who comply, is superior to application sensitivity.

The positive outcomes achieved by screening for colorectal cancer with FOBTs cannot be realized if the quality of the colonoscopy for positive tests is suboptimal. Given the number eligible for screening, the number of colonoscopies completed is important but much less so than the quality of those performed. Bretagne et al8 in this issue of Gastrointestinal Endoscopy show how the endoscopist is an independent predictor of detecting adenomas in a large patient population screened biennially with a standard guaiac FOBT (gFOBT). The study highlights the potential problem of achieving desired outcomes when, for whatever reason, adenomas are missed.8

Studies like the Bretagne et al8 study are becoming more and more important as we come to understand that the mortality reduction predicted for screening colonoscopy is less than what had previously been predicted.17 Several previously published studies have shown that colonoscopy has a significant miss rate of advanced neoplasia,18, 19, 20, 21 prompting even colonoscopy advocates to question calling optical colonoscopy the criterion standard CRC screening test.22, 23, 24 More recently, evidence has been accumulating suggesting that protection against cancer afforded by having a negative colonoscopy is quite small (12%-33%) in the proximal colon as compared with the distal colon, where it is quite large (80%). These findings are consistent with trends in distal CRC rates in the United States that have been steadily decreasing since 1985, while rates for proximal colon cancers have remained largely unchanged.25, 26, 27, 28, 29, 30

Although the findings in these studies may be surprising and somewhat counterintuitive, there are reasonable explanations. First and foremost is the quality of the colonoscopic examinations. The criterion standard study for evaluation of appropriate surveillance intervals is the National Polyp Study.31 In this study, if the baseline colonoscopy did not clear the colon with high confidence, the examination was repeated before the patient was entered into the surveillance program. A high confidence examination was defined as one in which there was an excellent preparation, complete polypectomy, and slow withdrawal. These standards required repeat examinations in 13% of cases. It is unfortunate that the Bretagne et al8 study was unable to quantify withdrawal time, the quality of the bowel preparation, or even the number of repeat colonoscopies. As the authors point out, recording of these variables as well as feedback to the endoscopists on miss rates are essential for a successful population-based screening program.

Inherent limitations of colonoscopy include the difficulty of identifying hidden lesions behind folds or flat lesions. It is also likely that some cancers are rapidly growing tumors that will not be uncovered soon enough because the recommended surveillance interval after a normal examination is 10 years. Evidence is accumulating that indicates the biology of cancers in the right side of the colon, especially neoplasms characterized by inactivation of a mismatch repair gene, may make right-sided cancers grow more rapidly than left-sided ones. This biological difference and hard-to-detect flat lesions could also explain the difference in cancer reduction observed between the right and left side of the colon in programmatic colonoscopy screening every 10 years.32, 33 If more evidence confirms the limited effect of colonoscopy screening on death from right-sided CRC, the every-10-year recommendation for colonoscopy screening may have to be modified or bolstered by interval FOBT as recommended for sigmoidoscopy screening in the USPSTF guidelines.34

The methodology used to assess the endoscopists in the Bretagne et al8 study analyzed the numbers of CRCs and adenomas detected by each endoscopist, and the results point out a problem with detection of adenomas but not CRCs. Concern is offered about the reasons for this miss rate and, by implication, the possibility that the local French screening program will not be effective in decreasing cancer incidence. Likely, no data are or will be available on the size and histology of the presumed missed adenomas, although some information might be obtained by searching any pathology database the investigators might have of their colonoscoped cohort.

A discussion by the authors of how significant a miss of an adenoma might be would have been worthwhile. In the meta-analysis by van Rijn et al,35 colonoscopy miss rates were, as one would expect, greatest for the diminutive (<5 mm) adenomas and least for the large (>10 mm) ones, giving some assurance that adenomas most likely to go on to become fatal cancers are unlikely to be missed. The issue of the significance of a missed adenoma of <1 cm and the subsequent risk to the patient of death from colorectal cancer has been overemphasized in the screening literature and media36 and has unnecessarily frightened the primary care physician and the average citizen into believing that the only good polyp is the one in formalin and the only effective screening test is optical colonoscopy.

The definition of an advanced colonic neoplasm as used in the gastroenterology literature is very different from such a definition in a standard pathology textbook. As defined in the GI literature, an advanced neoplasm can be a range of lesions, from large tubular adenomas to early adenocarcinomas that vary widely in terms of the risk they pose for progression to fatal cancer. A large polyp (>1 cm) becomes colorectal cancer at a rate of roughly 1% per year and, if left in situ, has a cumulative risk of malignancy at 20 years of only 24%.37, 38 It is extremely unlikely for a small (<1 cm) adenoma to develop into invasive cancer in less than 5 years.39 The clinical course of an advanced adenoma is not known to be ominous. The term was created because researchers needed a surrogate outcome more common than colorectal cancer.40 Its use as an outcome measure might therefore be misleading if it is used in screening studies, because the natural history of this lesion is unknown.41 Because most polyps, even the advanced ones, do not directly lead to death from colon cancer, the most important value of one test over another is the incremental benefit of mortality reduction that such test confers on the patient being screened. A person at age 50 years has a 5% lifetime risk of being diagnosed with colorectal cancer and a 2.5% chance of dying from it.42, 43 The concern about missed advanced neoplasms in once-only testing with methods other than colonoscopy may not be as important as it has been portrayed. Tests that are used more often than once every 10 years, such as FOBT or flexible sigmoidoscopy (FSIG), leave the potential for discovery of a missed advanced neoplasm on subsequent screenings before it has become malignant or lethal.

The marketing of colonoscopy as the best test has been effective. In 2001, Congress voted to have Medicare cover optical colonoscopy as a screening test and, thus, bypassed a Center for Medicare and Medicaid Services assessment still required today for all other screening tests seeking Medicare reimbursement. Medicare data reveal that the use of flexible sigmoidoscopy has fallen by 54% from 1993 to 2003, with most of the fall occurring after Congress approved Medicare reimbursement for screening colonoscopy. The same can be shown for the use of FOBTs. The fact that the increase in screening seen since the 1980s has been entirely due to the use of screening colonoscopy is well documented in Centers for Disease Control (CDC) data and is correlated with the date Congress approved Medicare reimbursement for this screening test.44

Despite all efforts to raise screening rates since colonoscopy was promoted as the best test in 2000, only 50% of Americans of screening age were up to date with screening by 2005.45 Being up to date was true only for those people with insurance. Those without insurance coverage were found to have a rate of colorectal cancer screening of only 24%. In 2008, a publication from the CDC reported no progress in reducing most CRC screening disparities between 2000 and 2005 and emphasized a need to increase CRC screening in all subpopulations but, in particular, in Hispanic women and uninsured men and women.46 Proponents of screening the underserved with optical colonoscopy set up such a program in New York City. A patient navigator was used to increase compliance. From November 2003 until May 2006, the program was able to accommodate 140 patients a year.47 The New York State Department of Health Cancer Services Program's Colorectal Cancer Screening Initiative decided to screen their underserved population with FOBTs from August 1997 through September 2007. Approximately 8000 patients were screened per year, and 97 cancers were diagnosed in patients with positive tests. Sixty-six percent of these were in stage one or two. In addition to the cancers, 768 adenomatous polyps were found and removed in patients with a positive test. In Kaiser, Northern California, the decision to screen its eligible population for CRC by using FIT was made in 2007. In 2008, 419,000 kits were distributed, and 52% were returned. The test was positive in 5.4% of the participants returning tests, and this led to colonoscopic evaluation that, to date, has uncovered 403 cancers. The New York State Department of Health and the Kaiser Permanente screening programs are good examples of how the National Cancer Institute mandate of improving screening rates in the United States can be achieved.

The most important lesson to be learned from studies like that of Bretagne et al8 is that the quality of optical colonoscopies performed as screening or diagnostic tests after a positive FOBT must remain high for society to benefit, but this is also true for all the recommended screening tests. Issues such as testing and surveillance intervals, adequacy of preparation, withdrawal time, experience of the endoscopist, number of fecal tests for best sensitivity and specificity, qualitative versus quantitative FITs, laboratory technician or physician development, and interpretation of tests or automation all deserve scrutiny and evaluation for screening programs to obtain optimum results. Research efforts and funding for research should not be focused only on developing newer screening tests without first placing higher priority on quality implementation of the effective ones we already have. Attention only to new tests or test favorites in the past has been short sighted. More studies looking at the comparative effectiveness of screening tests already proven to be effective and at reasons for suboptimal testing are badly needed.

Back to Article Outline

Disclosure 

Dr. Allison disclosed the following conflicts relevant to this publication: He is a member of the Scientific Advisory Board regarding screening tests for colorectal cancer for IntelligeneScan (San Francisco, CA) and is a consultant regarding fecal immunochemical tests for Quidel (San Diego, CA).

Back to Article Outline

References 

  1. Levin B, Lieberman DA, McFarland , et al. Screening and surveillance for the early detection of colorectal cancer and adenomatous polyps, 2008: a joint guideline from the American Cancer Society, the US Multi-Society Task Force on Colorectal Cancer, and the American College of Radiology. Gastroenterology. 2008;134:1570–1595
  2. Rex DK, Johnson DA, Anderson JC, et al. American College of Gastroenterology. Am J Gastroenterol. 2009;104:739–750Epub 2009 Feb
  3. Dominitz JA, Robertson DJ. Selected Summary Gastroenterology. 2009;137:1529–1531
  4. Whitlock EP, Lin JS, Liles E, et al. Screening for colorectal cancer: a targeted, updated systematic review for the US Preventive Services Task Force. Ann Intern Med. 2008;149:638–658
  5. Pignone M, Saha S, Hoerger T, et al. Cost-effectiveness analyses of colorectal cancer screening: a systematic review for the U.S. Preventive Services Task Force. Ann Intern Med. 2002;137:96–104
  6. National Research Council on the Economic Models of Colorectal Cancer Screening in Average-Risk Adults. Available at: http://www.nap.edu/catalog/11228.html.
  7. Zauber AG, Lansdorp-Vogelaar I, Knudsen AB, et al. Evaluating test strategies for colorectal cancer screening: a decision analysis for the US Preventive Services Task Force. Ann Intern Med. 2008;149:659–669
  8. Bretagne JF, Harmonic S, Piette C, et al. Variations between endoscopists in rates of detection of colorectal neoplasia and their impact on a regional screening program based on colonoscopy after fecal occult blood testing. Gastrointest Endosc. 2010;71:335–341
  9. Benson VX, Patnick J, Davies AK, et al. Colorectal cancer screening: a comparison of 35 initiatives in 17 countries. Int J Cancer. 2008;122:1357–1367
  10. Macafee DAL, Scholefield JH. Population based endoscopic screening for colorectal cancer. Gut. 2003;52:323–326
  11. Seef LC, Manninen DL, Dong FB, et al. Is there endoscopic capacity to provide colorectal cancer screening to the unscreened population of the United States?. Gastroenterology. 2004;127:1661–1669
  12. Levin TR. Editorial colonoscopy capacity: can we build it? Will they come?. Gastroenterology. 2004;127:1841–1849
  13. Woolf SH. The best screening test for colorectal cancer: a personal choice. Editorial. N Engl J Med. 2000;343:1641–1643
  14. Kahi CJ, Rex DK, Imperiale TF. Screening, surveillance, and primary prevention for colorectal cancers: a review of the literature. Gastroenterology. 2008;135:380–399
  15. Morikawa T, Kato J, Yamaji Y, et al. A comparison of the immunochemical fecal occult blood test and total colonoscopy in the asymptomatic population. Gastroenterology. 2005;129:422–428
  16. Allison JE, Sakoda LC, Levin TR, et al. Screening for colorectal neoplasms with new fecal occult blood tests: update on performance characteristics. J Natl Cancer Inst. 2007;99:1–9
  17. Ransohoff DF. How much does colonoscopy reduce colon cancer mortality?. Ann Intern Med. 2009;150:50–52
  18. Rex DK, Rahmani EY, Haseman JH, et al. Relative sensitivity of colonoscopy and barium enema for detection of colorectal cancer in clinical practice. Gastroenterology. 1997;112:17–23
  19. Bressler B, Paszat LE, Vinden C, et al. Colonoscopic miss rates for right-sided colon cancer: a population-based analysis. Gastroenterology. 2004;127:452–456
  20. Pickhardt PJ, Nugent PA, Mysliwiec PA, et al. Location of adenomas missed by optical colonoscopy. Ann Intern Med. 2004;141:352–359
  21. Robertson DJ. Colorectal cancer in patients under close colonoscopic surveillance. Gastroenterology. 2005;129:34–41
  22. Lieberman D. Colonoscopy: as good as gold?. Ann Intern Med. 2004;141:401–403
  23. Rex D, Eid E. Considerations regarding the present and future roles of colonoscopy in colorectal cancer prevention. Clin Gastroenterol Hepatol. 2008;6:506–514
  24. Cooper GS. Colonoscopy: a tarnished gold standard?. Gastroenterology. 2007;132:2588–2604
  25. Pabby A, Schoen RE, Weissfeld JL, et al. Analysis of colorectal cancer occurrence during surveillance colonoscopy in the dietary Polyp Prevention Trial. Gastrointest Endosc. 2005;61:385–391
  26. Singh H, Turner D, et al. Risk of developing colorectal cancer following a negative colonoscopy examination. JAMA. 2006;296:2438
  27. Cotterchio M, Manno M, et al. Cancer Causes Control. 2005;16:865–875
  28. Cress R, Morris C, et al. Secular changes in colorectal cancer incidence by subsite, state at diagnosis, and race/ethnicity, 1992-2001. Cancer. 2006;107(suppl 5):1142–1152
  29. Lakoff J, Paszat LF, Saskin R, et al. Risk of developing proximal versus distal colorectal cancer after a negative colonoscopy: a population-based study. Clin Gastroenterol Hepatol. 2008;6:1117–1121
  30. Baxter NN, Goldwasser MA, Paszat LF, et al. Association of colonoscopy and death from colorectal cancer: a population-based case control study. Ann Intern Med. 2009;150:1–8
  31. Winawer SJ, Zauber AG, Ho MN, et al. Prevention of colorectal cancer by colonoscopic polypectomy. The National Polyp Study Workgroup. N Engl J Med. 1993;329:1977–1981
  32. Allison JE, Potter MB. New screening guidelines for colorectal cancer: a practical guide for the primary care physician. Prim Care. 2009;36:575–602
  33. Allison JE. Screening for colorectal cancer. CCO oncology in practice. 2009;Availabe at: http://www.clinicaloptions.com
  34. US Preventive Services Task Force. Screening for colorectal cancer: US Preventive Services Task Force recommendation statement. Ann Intern Med. 2008;149:627–637
  35. Van Rijn JC, Reitsma FB, Stoker J, et al. Polyp miss rate determined by tandem colonoscopy: a systematic review. Am J Gastroenterol. 2006;101:343–350
  36. Cram P, Fendrick MA, Inadomi J, et al. The impact of a celebrity promotional campaign on the use of colon cancer screening: the Katie Couric effect. Arch Intern Med. 2003;1601–1605
  37. Ransohoff DF. Lessons from the UK sigmoidoscopy screening trial. Editorial. Lancet. 2002;359:1266–1267
  38. Stryker S, Wolff B, Culp C, et al. Natural history of untreated colonic polyps. Gastroenterology. 1987;93:1009–1013
  39. Eide T. Risk of colorectal cancer in adenoma bearing individuals within a defined population. Int J Cancer. 1986;38:173–176
  40. Ransohoff DF. Virtual colonoscopy–what it can do vs what it will do. JAMA. 2004;291:1772–1774
  41. Imperiale TF, Wagner DR, Lin CY, et al. Results of screening colonoscopy among persons 40 to 49 years of age. N Engl J Med. 2002;346:1781–1785
  42. Burt RW. Colon cancer screening. Gastroenterology. 2000;119:837–853
  43. U.S. Department of Health and Human Services. Agency for Healthcare Research and Quality. U.S. Preventive Services Task Force. Available at: http://www.preventiveservices.ahrq.gov.
  44. Behavioral Risk Factor Surveillance System CD-ROM (1996-1997, 1999) and Public Use Data Tape (2001, 2002, 2004), National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, 1999, 2000, 2002, 2003, 2005.
  45. Shapiro JA, Seef LC, Thompson TD, et al. Colorectal cancer test use from the 2005 National Health Interview Survey. Cancer Epidemiol Biomarkers Prev. 2008;17:1623–1630
  46. Trivers KF, Shaw KM, Sabatino SA, et al. Trends in colorectal cancer disparities in people age 60–64, 2000–2005. Am J Prev Med. 2008;35:185–193Epub 2008 Jul 10
  47. Chen LA, Santos S, Jandorf L, et al. A program to enhance completion of screening colonoscopy among urban minorities. Clin Gastroenterol Hepatol. 2008;6:443–450Epub 2008 Mar 4

PII: S0016-5107(09)02686-8

doi:10.1016/j.gie.2009.10.032

Gastrointestinal Endoscopy
Volume 71, Issue 2 , Pages 342-345, February 2010

Article Tools