Cover Image

PAPERBACK
$63.25



View/Hide Left Panel

Appendix P
Natural History of Colorectal Adenomas and Cancer T.R.Levin, M.D.

SLIDE 1

SLIDE 1 NOTES: No notes.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 263
Economic Models of Colorectal Cancer Screening in Average-Risk Adults: Workshop Summary Appendix P Natural History of Colorectal Adenomas and Cancer T.R.Levin, M.D. SLIDE 1 SLIDE 1 NOTES: No notes.

OCR for page 263
Economic Models of Colorectal Cancer Screening in Average-Risk Adults: Workshop Summary SLIDE 2 SLIDE 2 NOTES: No notes.

OCR for page 263
Economic Models of Colorectal Cancer Screening in Average-Risk Adults: Workshop Summary SLIDE 3 SLIDE 3 NOTES: The concept of de novo cancer differs according to the perspective of the individual using the term. To an endoscopist it may mean one thing; to a molecular biologist another. For the purpose of an outcome study or a CEA model of colorectal cancer screening, it may not matter. Any adenoma that cannot be found on any of the existing technologies can be referred to as de novo. However, over time, colonoscopy and other techniques may improve our ability to detect lesions that cannot be detected today. Before I re-reviewed the literature on this subject, I believed that the general consensus for the proportion of cancers that arise de novo—10 percent—was substantially correct. However, one case-cohort Markov model from Taiwan followed about 13,000 people who had colonoscopy and recorded the percent who developed colon cancer during followup (Chen et al., 2003). They estimated a very high rate of 27–32 percent, depending on assumptions such as the detectability of adenomas on colonoscopy. Although the population was Taiwanese, which could have a different natural history or different endoscopic techniques, I believe this is an upper bound and possibly an overestimate. Kaiser’s study found a 60 percent mortality reduction for distal cancer from people exposed to rigid sigmoidoscopy (Selby et al., 1992). Can that study be interpreted to suggest that 40 percent of distal cancers arise de novo? Of course that study reflected a series of examinations given to all comers, regardless of the quality of the examination, degree of penetration of the sigmoidoscope, etc. Therefore, many of the cancers do not meet the true definition of de novo. Rather, they reflect the existence of endoscopic failures.

OCR for page 263
Economic Models of Colorectal Cancer Screening in Average-Risk Adults: Workshop Summary SLIDE 4 SLIDE NOTES 4: The National Polyp Study found five cancers as part of surveillance colonoscopies (Winawer et al., 1993). The first one listed was quite small and was detected only by an expert colonoscopist in the sigmoid colon. The others were all beyond the reach of sigmoidoscopy and might be classified as left-sided cancers. When these findings are compared with the Kaiser sigmoidoscopy findings (previous slide) they raise the question whether the only cancers that can be prevented are ones in the distal colon.

OCR for page 263
Economic Models of Colorectal Cancer Screening in Average-Risk Adults: Workshop Summary SLIDE 5 SLIDE 5 NOTES: Here is a table from the PLCO study (Schoen et al., 2003). This paper suggests an incidence of de-novo distal cancer of 0.02 percent per year, at least in patients undergoing sigmoidoscopy. This result is not much different from the findings of the National Polyp Study.

OCR for page 263
Economic Models of Colorectal Cancer Screening in Average-Risk Adults: Workshop Summary SLIDE 6 SLIDE 6 NOTES: Paul Doria-Rose has been working on a study at Kaiser Permanente that looks at early cohorts from our sigmoidoscopy screening program (Doria-Rose et al., 2004). Focusing on the distal cancers only, about 30 cases occurred over 302,424 person years of follow-up, or approximately 0.01 percent per person year. That rate is somewhat lower than was reported in the PLCO trial, probably because the individuals in this study developed diagnosed cancer, whereas in PLCO some cancers were found through rescreening.

OCR for page 263
Economic Models of Colorectal Cancer Screening in Average-Risk Adults: Workshop Summary SLIDE 7 SLIDE 7 NOTES: We compared the experience in the Kaiser sigmoidoscopy program with the incidence of cancer reported in the SEER registry for our region. The comparison implies about an 85 percent reduction in incidence of CRC during those years of follow up (Doria-Rose et al., 2004). Would that imply that roughly 15 percent of the cancers are de novo?

OCR for page 263
Economic Models of Colorectal Cancer Screening in Average-Risk Adults: Workshop Summary SLIDE 8 SLIDE 8 NOTES: The Telemark randomized trial compared sigmoidoscopy vs. no screening, and then polypectomy at colonoscopy and colonoscopy surveillance (Thiis-Evensen et al., 1999). Two interesting results from that study may be relevant. One is the 80 percent reduction in colorectal cancer incidence over the follow-up period. Does that mean that we cannot prevent 20 percent of cancers because they are effectively de novo at present? Or does it mean that sigmoidoscopy cannot detect many cancers? Another interesting finding of this study is that there were actually more deaths in the screening group than there were in the control group. Any mortality reduction that was achieved with colorectal cancer was completely swamped by excess cardiovascular death. One theory is that somehow screening is actually contributing to cardiovascular disease and death. But this is just hypothesis. Given the theories about the relationship between infectious disease and cardiovascular disease, perhaps we are liberating some microbes from the colon at the time we insert a scope. Such a hypothesis would be a worst-case scenario, and at present there is simply no evidence to explain these findings. Whether or not colorectal cancer screening increases or simply does not decrease all-cause mortality, we must ask whether by screening we simply save people from colorectal cancer to die of other things.

OCR for page 263
Economic Models of Colorectal Cancer Screening in Average-Risk Adults: Workshop Summary SLIDE 9 SLIDE 9 NOTES: Other sources of data on de novo rates are expected soon, from David Lieberman’s cooperative study of follow-up from colonoscopy screening, and the consensus estimates obtained by Reid Ness. Finally, the true proportion of de novo cancers probably depends on the age and gender of the person. They may be more common in women, especially as we are learning that proximal cancers appear to be more common in women and de novo cancers are more likely to be proximal cancers. African Americans may also have a higher rate of de novo cancers than do Caucasians.

OCR for page 263
Economic Models of Colorectal Cancer Screening in Average-Risk Adults: Workshop Summary SLIDE 10 SLIDE 10 NOTES: The next issue is what is the best estimate of the length of time it takes for adenomas to transition into cancer. One problem in attempting to estimate this is the fact that there may be two kinds of adenomas—those that are never going to progress and those that will progress. We are just beginning to scratch the surface in understanding what influences the progression rate. The “10 year” assumption comes from studies of biopsied polyps that were left in place (Morson, 1984). Those very few studies found that it took about 10 to 15 years before cancer developed. In a study of cancers in patients with polyps removed, Atkin found an average 14 year lapse (Atkin et al., 1992). Finally, case control studies, including a recent one by Polly Newcomb, show that the protection persists up to 14 years (Selby et al., 1992; Newcomb et al., 2003). So, the evidence from different studies is consistent and points to a protective period of about 14 years.

OCR for page 263
Economic Models of Colorectal Cancer Screening in Average-Risk Adults: Workshop Summary SLIDE 11 SLIDE 11 NOTES: I found no recent information on the distribution of dwelling times. The Stryker study is the most frequently cited (Stryker et al., 1987). That Mayo Clinic report suggests that about 75 percent of large polyps found on barium enema and left unresected for various reasons never progressed to cancer in the 20-year period of observation. A study of small adenomas showed that about 2.5 adenomas per 1,000 per year progressed to cancer (Eide, 1986).

OCR for page 263
Economic Models of Colorectal Cancer Screening in Average-Risk Adults: Workshop Summary SLIDE 12 SLIDE 12 NOTES: The guidelines promulgated by the GI Consortium led by Dr. Winawer concluded that adenomas of any size were present in 25 percent of people by age 50, large adenomas in 4.6 percent by age 54, and large adenomas in 15.6 percent of patients at age 75 (Winawer et al., 1997). These conclusions were based on a reading of the autopsy literature, which has been referred to by other presenters at this Workshop. Probably the three largest studies that have looked at this in living patients (with endoscopy) have been the VA Cooperative Study, the Lilly Colonoscopy Study by Tom Imperiale, and our study at Kaiser (Levin et al., 1999; Imperiale et al., 2000; Lieberman et al., 2000).

OCR for page 263
Economic Models of Colorectal Cancer Screening in Average-Risk Adults: Workshop Summary SLIDE 13 SLIDE 13 NOTES: Regarding the question of the frequency with which advanced proximal neoplasia—both adenomas and cancer—are accompanied by a synchronous adenoma in the distal colon, there are a number of studies that suggest the rate is about 4 to 10 percent of people with a small distal tubular adenoma, and 1 to 5 percent in people without an adenoma in the distal colon. The Kaiser CoCaP study found 5.3 percent of individuals with no adenomas on sigmoidoscopy had at least one advanced proximal adenoma, but most of these were not cancer (Levin et al., 1999). Our study was prospective, so we were not dependent on retrospective review of electronic and paper medical records, as some studies were.

OCR for page 263
Economic Models of Colorectal Cancer Screening in Average-Risk Adults: Workshop Summary SLIDE 14 SLIDE 14 NOTE: As part of the Kaiser study we analyzed (through classification and regression analysis) the risk factors associated with advanced proximal neoplasia (adenomas and cancer) by type in close to 3,000 people who had colonoscopy (Levin et al., 1999). The most important risk factor was the presence of a villous architecture in the polyp; 12.1 percent of those individuals were found to have advanced proximal neoplasia. The next most important factor was age. Notice, also, that the number of tubular adenomas is not as important as the other two. Indeed, age alone is a stronger predictor than the presence of a distal tubular adenoma. Distal adenoma size did not enter into the CART analysis at any level. These findings suggest the potential for mixed strategies—using different screening tests at different ages—to optimize a screening program. For example, perhaps it would be better to screen younger people with sigmoidoscopy and switch to a total colon examination after 60 or 65 years of age. This is where modeling could be especially useful, because it would take many years to answer the question via a clinical trial of such strategies.

OCR for page 263
Economic Models of Colorectal Cancer Screening in Average-Risk Adults: Workshop Summary SLIDE 15 SLIDE 15 NOTES: This chart shows recent findings on the sensitivity of sigmoidoscopy for advanced proximal polyps and cancers (Imperiale et al., 2000; Lieberman et al., 2000). The two studies indicate that about 60–70 percent of clinically important lesions in the proximal colon can be found through following patients with lesions in the distal colon. That finding may vary with the reach of the sigmoidoscope into the colon, and it may also vary with age of the patient. Sensitivity could be lower in older patients, because of the higher prevalence of proximal lesions in that group.

OCR for page 263
Economic Models of Colorectal Cancer Screening in Average-Risk Adults: Workshop Summary SLIDE 16 SLIDE 16 NOTES: Lewis did a meta-analysis of studies that examined the relationship between distal and proximal advanced adenomas and cancers. From the three studies, it appears that roughly 3 percent of all screened individuals have an isolated proximal adenoma that would not be detected by sigmoidoscopy. That implies that colonoscopy provides an incremental benefit over sigmoidoscopy to only 3 or so percent of the screened population (Lewis et al., 2003).

OCR for page 263
Economic Models of Colorectal Cancer Screening in Average-Risk Adults: Workshop Summary SLIDE 17 SLIDE 17 NOTES: What is the prognostic value of finding adenomas? In Atkin’s study of follow-up after sigmoidoscopy, the presence of a distal advanced adenoma increases the risk of subsequent cancer 3-fold; and multiple advanced adenomas increased it by 6-fold (Atkin et al., 1992). However, finding the presence of a single polyp or multiple small polyps did not carry an increased risk. (The relatively young age of the people in that study, however, may have an effect on these findings.) The dietary intervention studies, which involve periodic surveillance with colonoscopy, also show that individuals with one or two small adenomas rarely get subsequent cancers.

OCR for page 263
Economic Models of Colorectal Cancer Screening in Average-Risk Adults: Workshop Summary SLIDE 18 SLIDE 18 NOTES: Is there any new information that can inform modelers about the stage-specific dwelling times for colon cancer? Koretz’s 1993 analysis showed that the time from the development of malignancy until clinical detection is close to five years. That analysis is based on a comparison of autopsy data and clinical reports of prevalence of cancers at different stages (Koretz, 1993).

OCR for page 263
Economic Models of Colorectal Cancer Screening in Average-Risk Adults: Workshop Summary SLIDE 19 SLIDE 19 NOTES: This slide includes a list of tumor and patient characteristics that appear to be correlated with progression of CRC. Note, however, that this list is by no means exhaustive. Other prognostic factors exist. The list shown here is from a single search of the medical literature (via Medline) conducted on 1–16–2004. The molecular characteristics of tumors that affect both progression and response to therapy are not well understood. For example, we do not yet know what proportion of different cancers carry the various tumor markers that have been identified so far. These markers probably interact with stage at detection, location, histology, race, gender, and other more obvious and easily measured things. For example, cancers in African Americans may be more likely to have a specific molecular characteristic. However, at present we do not have enough information on how the molecular characteristics map into more easily measured characteristics. I would propose that the method of detection of cancers and polyps may also interact with the genotype or phenotype or prognosis. It may be that cancers detected with stool marker tests that search for a certain molecular panel of fecal DNA may be different in their molecular distribution from those detected by more traditional fecal occult blood tests. And, those differences may mean that the progression of cancers found on such tests differs from those found with fecal occult blood.

OCR for page 263
Economic Models of Colorectal Cancer Screening in Average-Risk Adults: Workshop Summary SLIDE 20 SLIDE 20 NOTES: Similarly, the cancers that can be seen with colonoscopy alone may be different in terms of progression from those that can be seen with both colonoscopy and virtual colonoscopy. Perhaps tumors located “behind a fold” (and therefore more difficult for virtual colonoscopy) may have a different biologic behavior. We simply do not know whether such is the case at present (Shibata et al., 1996; Carethers et al., 1998; Liefers et al., 1998; Watanabe et al., 2001; Marcella and Miller, 2001; Makinen et al., 2001; Liang et al., 2002; Garcia et al., 2003; McArdle et al., 2003; Rabeneck et al., 2003; Jemal et al., 2003; Kama et al., 2003). We do know that most advanced adenomas will never develop into clinically apparent cancer. Those that do will take a long time. Some cancers will arise de novo, but we cannot know at present whether they are inherently unique biologically or whether they grew from adenomas that were missed on colonoscopy. This distinction may not matter unless there is one day a major advance in colonoscopic technology. Increasing age does predict proximal disease, which probably leads to more cancers developing in the interval between screens. In the National Polyp Study, all individuals with interval cancers were over 60. There may be a molecular correlation with this phenomenon: DNA methylation is one cancer pathway that seems to accumulate as people age. Finally, villous histology is such a strong predictor of cancer risk that there may be a molecular marker that correlates with it. However, to my knowledge no such marker has yet been identified.

OCR for page 263
Economic Models of Colorectal Cancer Screening in Average-Risk Adults: Workshop Summary REFERENCES Atkin WS, Morson BC, Cuzick J. 1992. Long-term risk of colorectal cancer after excision of rectosigmoid adenomas. N Engl J Med. 326(10):658–62. Carethers JM, Hawn MT, Greenson JK, Hitchcock CL, Boland CR. 1998. Prognostic significance of allelic lost at chromosome 18q21 for stage II colorectal cancer. Gastroenterology. 114(6): 1188–95. Chen CD, Yen MF, Wang WM, Wong JM, Chen TH. 2003. A case-cohort study for the disease natural history of adenoma-carcinoma and de novo carcinoma and surveillance of colon and rectum after polypectomy: implication for efficacy of colonoscopy. Br J Cancer. 88(12): 1866–73. Doria-Rose VP, Levin TR, Selby JV, Newcomb PA, Richert-Boe KE, Weiss NS. 2004. The incidence of colorectal cancer following a negative screening sigmoidoscopy: implications for screening interval. Gastroenterology. 127(3):714–22. Eide TJ. 1986. Risk of colorectal cancer in adenoma-bearing individuals within a defined population. Int J Cancer. 38(2):173–6. Garcia JM, Rodriguez R, Dominguez G, Silva JM, Provencio M, Silva J, Colmenarejo A, Millan I, Munoz C, Salas C, Coca S, Espana P, Bonilla F. 2003. Prognostic significance of the allelic loss of the BRCA1 gene in colorectal cancer. Gut. 52(12):1756–63. Imperiale TF, Wagner DR, Lin CY, Larkin GN, Rogge JD, Ransohoff DF. 2000. Risk of advanced proximal neoplasms in asymptomatic adults according to the distal colorectal findings. N Engl J Med. 343(3):169–74. Jemal A, Murray T, Samuels A, Ghafoor A, Ward E, Thun MJ. 2003. Cancer statistics, 2003. CA Cancer J Clin. 53(1):5–26. Kama NA, Kologlu M, Reis E, Doganay M, Atli M, Dolapci M. 2003. A prognostic score for colorectal cancer. Hepatogastroenterology. 50(53):1356–61. Koretz RL. 1993. Malignant polyps: are they sheep in wolves’ clothing? Ann Intern Med. 118(1): 63–8. Levin TR, Palitz A, Grossman S, Conell C, Finkler L, Ackerson L, Rumore G, Selby JV. 1999. Predicting advanced proximal colonic neoplasia with screening sigmoidoscopy. JAMA. 281(17):1611–7. Lewis JD, Ng K, Hung KE, Bilker WB, Berlin JA, Brensinger C, Rustgi AK. 2003. Detection of proximal adenomatous polyps with screening sigmoidoscopy: a systematic review and meta-analysis of screening colonoscopy. Arch Intern Med. 163(4):413–20. Liang JT, Huang KC, Lai HS, Lee PH, Cheng YM, Hsu HC, Cheng AL, Hsu CH, Yeh KH, Wang SM, Tang C, Chang KJ. 2002. High-frequency microsatellite instability predicts better chemosensitivity to high-dose 5-fluorouracil plus leucovorin chemotherapy for stage IV sporadic colorectal cancer after palliative bowel resection. Int J Cancer. 101(6):519–25. Lieberman DA, Weiss DG, Bond JH, Ahnen DJ, Garewal H, Chejfec G. 2000. Use of colonoscopy to screen asymptomatic adults for colorectal cancer. Veterans Affairs Cooperative Study Group 380. N Engl J Med. 343(3):162–8. Liefers GJ, Cleton-Jansen AM, van de Velde CJ, Hermans J, van Krieken JH, Cornelisse CJ, Tollenaar RA. 1998. Micrometastases and survival in stage II colorectal cancer. N Engl J Med. 339(4):223–8. Makinen MJ, George SM, Jernvall P, Makela J, Vihko P, Karttunen TJ. 2001. Colorectal carcinoma associated with serrated adenoma-prevalence, histological features, and prognosis. J Pathol. 193(3):286–94. Marcella S, Miller JE. 2001. Racial differences in colorectal cancer mortality. The importance of stage and socioeconomic status. J Clin Epidemiol. 54(4):359–66.

OCR for page 263
Economic Models of Colorectal Cancer Screening in Average-Risk Adults: Workshop Summary McArdle CS, McMillan DC, Hole DJ. 2003. Male gender adversely affects survival following surgery for colorectal cancer. Br J Surg. 90(6):711–5. Morson BC. 1984. The evolution of colorectal carcinoma. Clin Radiol. 35(6):425–31. Newcomb PA, Storer BE, Morimoto LM, Templeton A, Potter JD. 2003. Long-term efficacy of sigmoidoscopy in the reduction of colorectal cancer incidence. J Natl Cancer Inst. 95(8):622–5. Rabeneck L, Souchek J, El-Serag HB. 2003. Survival of colorectal cancer patients hospitalized in the Veterans Affairs Health Care System. Am J Gastroenterol. 98(5):1186–92. Schoen RE, Pinsky PF, Weissfeld JL, Bresalier RS, Church T, Prorok P, Gohagan JK. 2003. Results of repeat sigmoidoscopy 3 years after a negative examination. JAMA. 290(1):41–8. Selby JV, Friedman GD, Quesenberry CP Jr, Weiss NS. 1992. A case-control study of screening sigmoidoscopy and mortality from colorectal cancer. N Engl J Med. 326(10):653–7. Shibata D, Reale MA, Lavin P, Silverman M, Fearon ER, Steele G Jr, Jessup JM, Loda M, Summerhayes IC. 1996. The DCC protein and prognosis in colorectal cancer. N Engl J Med. 335(23):1727–32. Stryker SJ, Wolff BG, Gulp CE, Libbe SD, Ilstrup DM, MacCarty RL. 1987. Natural history of untreated colonic polyps. Gastroenterology. 93(5):1009–13. Thiis-Evensen E, Hoff GS, Sauar J, Langmark F, Majak BM, Vatn MH. 1999. Population-based surveillance by colonoscopy: effect on the incidence of colorectal cancer. Telemark Polyp Study I. Scand J Gastroenterol. 34(4):414–20. Watanabe T, Wu TT, Catalano PJ, Ueki T, Satriano R, Haller DG, Benson AB 3rd, Hamilton SR. 2001. Molecular predictors of survival after adjuvant chemotherapy for colon cancer. N Engl J Med. 344(16):1196–206. Winawer SJ, Fletcher RH, Miller L, Godlee F, Stolar MH, Mulrow CD, Woolf SH, Glick SN, Ganiats TG, Bond JH, Rosen L, Zapka JG, Olsen SJ, Giardiello FM, Sisk JE, Van Antwerp R, Brown-Davis C, Marciniak DA, Mayer RJ. 1997. Colorectal cancer screening: clinical guidelines and rationale. Gastroenterology. 112(2):594–642. Winawer SJ, Zauber AG, Ho MN, O’Brien MJ, Gottlieb LS, Sternberg SS, Waye JD, Schapiro M, Bond JH, Panish JF, et al. 1993. Prevention of colorectal cancer by colonoscopic polypectomy. The National Polyp Study Workgroup. N Engl J Med. 329(27):1977–81.