Virtual Colonoscopy to Screen for Colorectal Cancer
http://www.100md.com
《新英格兰医药杂志》
To the Editor: The data presented by Pickhardt and colleagues (Dec. 4 issue)1 on the accuracy of virtual colonoscopy are attractive, and to be sure, virtual colonoscopy is a promising diagnostic approach. However, the public and the community of practicing physicians have been led to believe that virtual colonoscopy could or should be implemented as a screening tool for colorectal cancer in place of optical colonoscopy because it is now reported to be as accurate as or more accurate than optical colonoscopy. Unfortunately, this belief is not supported by the aggregate data.
First, multiple studies have shown much lower sensitivity for virtual colonoscopy than that in the study by Pickhardt et al. The use of more advanced technology in the current study (three-dimensional rendering) is unquestionably responsible in part, but in addition, it appears that the study population examined was unique. Indeed, this cohort seemed to be highly compliant and carefully prepared. For example, 24-hour preparation with stool tagging was used instead of a typical overnight preparation without stool tagging. Thus, the results of such a study may not be generalizable to other groups.
In addition, the accuracy of optical colonoscopy for detecting lesions that were 1 cm or more in diameter in the study by Pickhardt et al. is at odds with the findings in other studies. The sensitivity of optical colonoscopy for detecting such lesions was 87.5 percent (42 of 48 lesions), a rate substantially below the standard when optical colonoscopy is performed by experts (>95 percent sensitivity). Therefore, although virtual colonoscopy will almost certainly be a key component in colon imaging and perhaps in screening for colon cancer, much more study is required before it replaces optical colonoscopy or is widely implemented as a screening tool for colorectal cancer.
Don C. Rockey, M.D.
Duke University Medical Center
Durham, NC 27710
References
Pickhardt PJ, Choi JR, Hwang I, et al. Computed tomographic virtual colonoscopy to screen for colorectal neoplasia in asymptomatic adults. N Engl J Med 2003;349:2191-2200.
To the Editor: Pickhardt et al. conclude that diminutive polyps (<6 mm in diameter) should be regarded as clinically insignificant and therefore ignored on virtual colonoscopy. They state that "a 10-mm threshold would probably render the use of virtual colonoscopy for screening cost effective." In their study, the majority of adenomatous polyps (503 of 554, or 91 percent) were less than 10 mm. They found that the incidence of advanced neoplasia in diminutive polyps was 0.1 percent. The authors do not report the prevalence of villous changes in polyps that were 6 to 9 mm. In other colonoscopy series,1,2,3 the prevalence of villous changes or severe dysplasia in adenomas that were less than 10 mm ranged from 9 percent to 13 percent. In one of these series,1 there was a 4 percent prevalence of advanced changes in adenomas that were less than 6 mm. In these studies,1,2,3 small advanced adenomas (<10 mm) accounted for 19 to 58 percent of all advanced lesions. According to a previous report in the Journal, villous changes, regardless of polyp size, are important predictors of colon cancer.4 Hence, it is best to perform a biopsy in patients with small polyps, rather than to ignore these lesions.
Thomas M. Zarchy, M.D.
Shireen Pais, M.D.
University of Southern California Keck School of Medicine
Los Angeles, CA 90033
zarchy@usc.edu
References
Read TE, Read JD, Butterly LF. Importance of adenomas 5 mm or less in diameter that are detected by sigmoidoscopy. N Engl J Med 1997;336:8-12.
Lieberman DA, Weiss DG. One-time screening for colorectal cancer with combined fecal occult-blood testing and examination of the distal colon. N Engl J Med 2001;345:555-560.
Schoen RE, Corle D, Cranston L, et al. Is colonoscopy needed for the nonadvanced adenoma found on sigmoidoscopy? The Poly Prevention Trial. Gastroenterology 1998;115:533-541.
Atkin WS, Morson BC, Cuzick J. Long-term risk of colorectal cancer after excision of rectosigmoid adenomas. N Engl J Med 1992;326:658-662.
To the Editor: Pickhardt et al. compared the characteristics of computed tomographic (CT) virtual colonoscopy with those of screening colonoscopy in a population at average risk. The report on this well-designed study does not address morphologic characteristics — specifically, flat or depressed polyps. This point is particularly relevant to the authors' rather strong statements regarding the evidence in discussing the management of polyps that are less than 5 mm in diameter. There is substantial evidence1,2 that flat or depressed lesions, with an estimated incidence as high as 30 percent among polyps found during colonoscopy, are associated with an increased frequency of high-grade dysplasia or carcinoma in situ, irrespective of their size. The ability to detect these particular lesions should therefore be important in the evaluation of any screening approach. Current guidelines from the American College of Gastroenterology3 recommend biopsy of all polyps found during screening colonoscopy, regardless of their size and morphologic characteristics, since appearance to the naked eye is unreliable. Finally, the authors make no reference to problems during colonoscopy that are due to the barium administered for stool tagging, which would have an obvious effect on the performance of colonoscopy.
Jorge R. Uribe, M.D.
Charles Bongiorno, M.D.
Philip O. Katz, M.D.
Albert Einstein Medical Center
Philadelphia, PA 19140
uribej@einstein.edu
References
Rembacken BJ, Fujii T, Cairns A, et al. Flat and depressed colonic neoplasms: a prospective study of 1000 colonoscopies in the UK. Lancet 2000;355:1211-1214.
Saitoh Y, Waxman I, West AB, et al. Prevalence and distinctive biologic features of flat colorectal adenomas in a North American population. Gastroenterology 2001;120:1657-1665.
Bond JH. Polyp guideline: diagnosis, treatment, and surveillance for patients with colorectal polyps. Am J Gastroenterol 2000;95:3053-3063.
To the Editor: Pickhardt and colleagues' comparison of the use of virtual colonoscopy with that of optical colonoscopy for colon-cancer screening is a well-designed addition to the field. An unaddressed potential downside of virtual colonoscopy, however, is exposure to radiation. The implications of serial virtual colonoscopy in all persons in a population who are 50 years old or older magnify this potential. How much of the body was included in the field of view of the multislice CT used to perform virtual colonoscopy, and what was the typical effective radiation dose? What are the implications of this exposure for the population being screened?
Gregory S. Thomas, M.D., M.P.H.
Mission Internal Medical Group
Mission Viejo, CA 92691
gthomas@mimg.com
Dr. Pickhardt replies: Rockey raises several important issues that are worthy of consideration. First, it should be clarified that the intention is not to have virtual colonoscopy "replace" optical colonoscopy but, rather, to provide another effective complement for screening. As such, virtual colonoscopy could be initially targeted to adults of average risk who are reluctant to undergo optical colonoscopy for primary screening. This promising addition to the screening armamentarium should come as welcome news, given the disappointing rates of compliance with existing strategies.
Second, our study population was indeed "unique" in that we evaluated a group of patients for whom virtual colonoscopy screening is ideally suited: asymptomatic average-risk adults. Previous studies of virtual colonoscopy largely concentrated on symptomatic or high-risk patients, for whom optical colonoscopy would generally remain the preferred initial study. Furthermore, although "careful" colonic preparation is indeed essential to the success of virtual colonoscopy, our low-volume preparation should be readily "generalizable" to other healthy cohorts. Finally, previous estimates of the miss rate with optical colonoscopy are all hindered by the fact that optical colonoscopy has served as its own reference standard. This means that polyps located in relative blind spots for optical colonoscopy, such as behind a colonic fold, could be repeatedly missed. Therefore, for adenomas that were 10 mm or more in diameter, the 12 percent miss rate with optical colonoscopy we observed with the use of virtual colonoscopy as a separate reference standard would seem more applicable than the 6 percent rate derived from back-to-back optical colonoscopy.1
Zarchy and Pais question the appropriateness of following patients with lesions that are smaller than 1 cm. Not only has such surveillance proved safe, but polyps that are 5 to 9 mm have actually had an overall tendency to decrease in size.2,3 One could argue that the risk of invasively removing small polyps with the use of optical colonoscopy might outweigh their minuscule inherent neoplastic risk. The frequency of advanced lesions among adenomas that were less than 10 mm and those that were 6 to 9 mm was 1.6 percent (8 of 503) and 4.4 percent (7 of 159), respectively; none were malignant.
Uribe et al. broach the controversial topic of flat lesions. Of 1228 polyps that were less than 10 mm in our study, none of the 174 lesions prospectively labeled "flat" on optical colonoscopy showed high-grade dysplasia, and 122 (70.1 percent) were nonadenomatous. Our experience extends the conclusion reached by Bond4 (and others5) that "small flat adenomas with a high malignant potential seem to be rare in Western countries, and there is little evidence that they are being overlooked." Furthermore, in our study, the rate of detection of nondiminutive flat lesions with virtual colonoscopy was similar to that of polypoid lesions. Finally, the barium administered orally for stool tagging is a 2 percent suspension (as compared with the 58 percent concentration administered rectally for air-contrast enemas) and had no detrimental effect on the performance of optical colonoscopy.
Thomas expresses a legitimate concern about radiation exposure. The mean effective dose with our current CT protocol is approximately 5 to 6 mSv, which is roughly equivalent to the typical exposure from either an air-contrast barium enema or natural background radiation over a two-year period. When confined to adults who are 50 years of age or older, this additional exposure is probably not clinically significant.
As with other paradigm shifts in medicine, primary screening with virtual colonoscopy will require deliberate implementation, a fact-based evolution, and a willingness to accept change. With more than 50,000 Americans dying each year of a largely preventable disease, it behooves our profession to pursue all effective and feasible means of increasing the numbers of people who undergo screening.
Perry J. Pickhardt, M.D.
University of Wisconsin Medical School
Madison, WI 53792
ppickhardt@mail.radiology.wisc.edu
References
Rex DK, Cutler CS, Lemmel GT, et al. Colonoscopic miss rates of adenomas determined by back-to-back colonoscopies. Gastroenterology 1997;112:24-28.
Hofstad B, Vatn MH, Larsen S, Osnes M. Growth of colorectal polyps: recovery and evaluation of unresected polyps of less than 10 mm, 1 year after detection. Scand J Gastroenterol 1994;29:640-645.
Hofstad B, Vatn MH, Andersen SN, et al. Growth of colorectal polyps: redetection and evaluation of unresected polyps for a period of three years. Gut 1996;39:449-456.
Bond JH. Doubling time of flat and polypoid colorectal neoplasms: defining the adenoma-carcinoma sequence. Am J Gastroenterol 2000;95:1621-1623.
Zauber AG, O'Brien MJ, Winawer SJ. On finding flat adenomas: is the search worth the gain? Gastroenterology 2002;122:839-840.
First, multiple studies have shown much lower sensitivity for virtual colonoscopy than that in the study by Pickhardt et al. The use of more advanced technology in the current study (three-dimensional rendering) is unquestionably responsible in part, but in addition, it appears that the study population examined was unique. Indeed, this cohort seemed to be highly compliant and carefully prepared. For example, 24-hour preparation with stool tagging was used instead of a typical overnight preparation without stool tagging. Thus, the results of such a study may not be generalizable to other groups.
In addition, the accuracy of optical colonoscopy for detecting lesions that were 1 cm or more in diameter in the study by Pickhardt et al. is at odds with the findings in other studies. The sensitivity of optical colonoscopy for detecting such lesions was 87.5 percent (42 of 48 lesions), a rate substantially below the standard when optical colonoscopy is performed by experts (>95 percent sensitivity). Therefore, although virtual colonoscopy will almost certainly be a key component in colon imaging and perhaps in screening for colon cancer, much more study is required before it replaces optical colonoscopy or is widely implemented as a screening tool for colorectal cancer.
Don C. Rockey, M.D.
Duke University Medical Center
Durham, NC 27710
References
Pickhardt PJ, Choi JR, Hwang I, et al. Computed tomographic virtual colonoscopy to screen for colorectal neoplasia in asymptomatic adults. N Engl J Med 2003;349:2191-2200.
To the Editor: Pickhardt et al. conclude that diminutive polyps (<6 mm in diameter) should be regarded as clinically insignificant and therefore ignored on virtual colonoscopy. They state that "a 10-mm threshold would probably render the use of virtual colonoscopy for screening cost effective." In their study, the majority of adenomatous polyps (503 of 554, or 91 percent) were less than 10 mm. They found that the incidence of advanced neoplasia in diminutive polyps was 0.1 percent. The authors do not report the prevalence of villous changes in polyps that were 6 to 9 mm. In other colonoscopy series,1,2,3 the prevalence of villous changes or severe dysplasia in adenomas that were less than 10 mm ranged from 9 percent to 13 percent. In one of these series,1 there was a 4 percent prevalence of advanced changes in adenomas that were less than 6 mm. In these studies,1,2,3 small advanced adenomas (<10 mm) accounted for 19 to 58 percent of all advanced lesions. According to a previous report in the Journal, villous changes, regardless of polyp size, are important predictors of colon cancer.4 Hence, it is best to perform a biopsy in patients with small polyps, rather than to ignore these lesions.
Thomas M. Zarchy, M.D.
Shireen Pais, M.D.
University of Southern California Keck School of Medicine
Los Angeles, CA 90033
zarchy@usc.edu
References
Read TE, Read JD, Butterly LF. Importance of adenomas 5 mm or less in diameter that are detected by sigmoidoscopy. N Engl J Med 1997;336:8-12.
Lieberman DA, Weiss DG. One-time screening for colorectal cancer with combined fecal occult-blood testing and examination of the distal colon. N Engl J Med 2001;345:555-560.
Schoen RE, Corle D, Cranston L, et al. Is colonoscopy needed for the nonadvanced adenoma found on sigmoidoscopy? The Poly Prevention Trial. Gastroenterology 1998;115:533-541.
Atkin WS, Morson BC, Cuzick J. Long-term risk of colorectal cancer after excision of rectosigmoid adenomas. N Engl J Med 1992;326:658-662.
To the Editor: Pickhardt et al. compared the characteristics of computed tomographic (CT) virtual colonoscopy with those of screening colonoscopy in a population at average risk. The report on this well-designed study does not address morphologic characteristics — specifically, flat or depressed polyps. This point is particularly relevant to the authors' rather strong statements regarding the evidence in discussing the management of polyps that are less than 5 mm in diameter. There is substantial evidence1,2 that flat or depressed lesions, with an estimated incidence as high as 30 percent among polyps found during colonoscopy, are associated with an increased frequency of high-grade dysplasia or carcinoma in situ, irrespective of their size. The ability to detect these particular lesions should therefore be important in the evaluation of any screening approach. Current guidelines from the American College of Gastroenterology3 recommend biopsy of all polyps found during screening colonoscopy, regardless of their size and morphologic characteristics, since appearance to the naked eye is unreliable. Finally, the authors make no reference to problems during colonoscopy that are due to the barium administered for stool tagging, which would have an obvious effect on the performance of colonoscopy.
Jorge R. Uribe, M.D.
Charles Bongiorno, M.D.
Philip O. Katz, M.D.
Albert Einstein Medical Center
Philadelphia, PA 19140
uribej@einstein.edu
References
Rembacken BJ, Fujii T, Cairns A, et al. Flat and depressed colonic neoplasms: a prospective study of 1000 colonoscopies in the UK. Lancet 2000;355:1211-1214.
Saitoh Y, Waxman I, West AB, et al. Prevalence and distinctive biologic features of flat colorectal adenomas in a North American population. Gastroenterology 2001;120:1657-1665.
Bond JH. Polyp guideline: diagnosis, treatment, and surveillance for patients with colorectal polyps. Am J Gastroenterol 2000;95:3053-3063.
To the Editor: Pickhardt and colleagues' comparison of the use of virtual colonoscopy with that of optical colonoscopy for colon-cancer screening is a well-designed addition to the field. An unaddressed potential downside of virtual colonoscopy, however, is exposure to radiation. The implications of serial virtual colonoscopy in all persons in a population who are 50 years old or older magnify this potential. How much of the body was included in the field of view of the multislice CT used to perform virtual colonoscopy, and what was the typical effective radiation dose? What are the implications of this exposure for the population being screened?
Gregory S. Thomas, M.D., M.P.H.
Mission Internal Medical Group
Mission Viejo, CA 92691
gthomas@mimg.com
Dr. Pickhardt replies: Rockey raises several important issues that are worthy of consideration. First, it should be clarified that the intention is not to have virtual colonoscopy "replace" optical colonoscopy but, rather, to provide another effective complement for screening. As such, virtual colonoscopy could be initially targeted to adults of average risk who are reluctant to undergo optical colonoscopy for primary screening. This promising addition to the screening armamentarium should come as welcome news, given the disappointing rates of compliance with existing strategies.
Second, our study population was indeed "unique" in that we evaluated a group of patients for whom virtual colonoscopy screening is ideally suited: asymptomatic average-risk adults. Previous studies of virtual colonoscopy largely concentrated on symptomatic or high-risk patients, for whom optical colonoscopy would generally remain the preferred initial study. Furthermore, although "careful" colonic preparation is indeed essential to the success of virtual colonoscopy, our low-volume preparation should be readily "generalizable" to other healthy cohorts. Finally, previous estimates of the miss rate with optical colonoscopy are all hindered by the fact that optical colonoscopy has served as its own reference standard. This means that polyps located in relative blind spots for optical colonoscopy, such as behind a colonic fold, could be repeatedly missed. Therefore, for adenomas that were 10 mm or more in diameter, the 12 percent miss rate with optical colonoscopy we observed with the use of virtual colonoscopy as a separate reference standard would seem more applicable than the 6 percent rate derived from back-to-back optical colonoscopy.1
Zarchy and Pais question the appropriateness of following patients with lesions that are smaller than 1 cm. Not only has such surveillance proved safe, but polyps that are 5 to 9 mm have actually had an overall tendency to decrease in size.2,3 One could argue that the risk of invasively removing small polyps with the use of optical colonoscopy might outweigh their minuscule inherent neoplastic risk. The frequency of advanced lesions among adenomas that were less than 10 mm and those that were 6 to 9 mm was 1.6 percent (8 of 503) and 4.4 percent (7 of 159), respectively; none were malignant.
Uribe et al. broach the controversial topic of flat lesions. Of 1228 polyps that were less than 10 mm in our study, none of the 174 lesions prospectively labeled "flat" on optical colonoscopy showed high-grade dysplasia, and 122 (70.1 percent) were nonadenomatous. Our experience extends the conclusion reached by Bond4 (and others5) that "small flat adenomas with a high malignant potential seem to be rare in Western countries, and there is little evidence that they are being overlooked." Furthermore, in our study, the rate of detection of nondiminutive flat lesions with virtual colonoscopy was similar to that of polypoid lesions. Finally, the barium administered orally for stool tagging is a 2 percent suspension (as compared with the 58 percent concentration administered rectally for air-contrast enemas) and had no detrimental effect on the performance of optical colonoscopy.
Thomas expresses a legitimate concern about radiation exposure. The mean effective dose with our current CT protocol is approximately 5 to 6 mSv, which is roughly equivalent to the typical exposure from either an air-contrast barium enema or natural background radiation over a two-year period. When confined to adults who are 50 years of age or older, this additional exposure is probably not clinically significant.
As with other paradigm shifts in medicine, primary screening with virtual colonoscopy will require deliberate implementation, a fact-based evolution, and a willingness to accept change. With more than 50,000 Americans dying each year of a largely preventable disease, it behooves our profession to pursue all effective and feasible means of increasing the numbers of people who undergo screening.
Perry J. Pickhardt, M.D.
University of Wisconsin Medical School
Madison, WI 53792
ppickhardt@mail.radiology.wisc.edu
References
Rex DK, Cutler CS, Lemmel GT, et al. Colonoscopic miss rates of adenomas determined by back-to-back colonoscopies. Gastroenterology 1997;112:24-28.
Hofstad B, Vatn MH, Larsen S, Osnes M. Growth of colorectal polyps: recovery and evaluation of unresected polyps of less than 10 mm, 1 year after detection. Scand J Gastroenterol 1994;29:640-645.
Hofstad B, Vatn MH, Andersen SN, et al. Growth of colorectal polyps: redetection and evaluation of unresected polyps for a period of three years. Gut 1996;39:449-456.
Bond JH. Doubling time of flat and polypoid colorectal neoplasms: defining the adenoma-carcinoma sequence. Am J Gastroenterol 2000;95:1621-1623.
Zauber AG, O'Brien MJ, Winawer SJ. On finding flat adenomas: is the search worth the gain? Gastroenterology 2002;122:839-840.