Clinical Gastroenterology and Hepatology 2014;-:-–-
Prevention, Early Detection, and Overdiagnosis of Colorectal Cancer Within 10 Years of Screening Colonoscopy in Germany Hermann Brenner,*,‡ Lutz Altenhofen,§ Christian Stock,*,k and Michael Hoffmeister* *Division of Clinical Epidemiology and Aging Research, German Cancer Research Center (DKFZ), Heidelberg, Germany; ‡ German Cancer Consortium (DKTK), Heidelberg, Germany; §Central Research Institute of Ambulatory Health Care in Germany, Berlin, Germany; and kInstitute of Medical Biometry and Informatics, University of Heidelberg, Heidelberg, Germany BACKGROUND & AIMS:
Screening colonoscopy was introduced in Germany in October 2002. We aimed to quantify its effects on prevention, early detection, and overdiagnosis of colorectal cancer (CRC) in the 10 years since its introduction.
METHODS:
We analyzed data from more than 4.4 million screening colonoscopies (conducted on individuals 55–79 years old from 2003 through 2012) available through the national screening colonoscopy registry. CRCs prevented, detected earlier than they would have been without screening, and overdiagnosed (cancers detected at screening colonoscopy that would not have become clinically manifest during the patient’s lifetime) were estimated by Markov models. Model parameters included sex-specific and age-specific findings at screening colonoscopy; mortality; rates of transition from nonadvanced to advanced adenoma, advanced adenoma to preclinical cancer, or preclinical cancer to clinically manifest cancer; and protection from screening colonoscopy.
RESULTS:
Overall, approximately 180,000 CRCs (1/28 screening colonoscopies) were estimated to have been prevented, and more than 40,000 CRCs (1/121 screening colonoscopies) were detected earlier than they would have been without screening, compared with approximately 4500 overdiagnoses (1/1089 screening colonoscopies). Almost all CRCs prevented or detected earlier than they would have been without screening resulted from screening colonoscopies performed on individuals up to 75 years old (97% and 89%, respectively), whereas 28% of overdiagnoses occurred from screening colonoscopies of individuals older than 75 years old.
CONCLUSIONS:
On the basis of a 10-year analysis of data from a national registry in Germany, screening colonoscopies have large potential for prevention and early detection of CRC, with low risk of overdiagnosis.
Keywords: Adenomas; Colon Cancer; Colorectal Neoplasms; Tumor.
olorectal cancer (CRC) is one of few cancers for which effective screening options are established. Randomized controlled trials have shown reduction of CRC mortality by annual or biennial fecal occult blood test screening1–3 and reduction of both CRC incidence and mortality by screening sigmoidoscopy.4 Although long-term results from randomized controlled trials for screening colonoscopy are not available yet, observational studies suggest even stronger reductions of CRC incidence and mortality,4 and screening colonoscopy has been recommended for CRC prevention and early detection by expert panels for more than 10 years.5–8 Germany was one of the first countries in the world to introduce colonoscopy as a primary screening offer. Timely estimation of its impact on prevention and early detection but also on potential overdiagnoses of CRC is crucial for decisions regarding maintenance or eventual adaptation and optimization of screening colonoscopy in Germany and for informing decisions on introduction,
C
maintenance, or adaptation of CRC screening offers in other countries. In this article, we aim for a comprehensive, joint, and detailed analysis of prevented, early detected, and overdiagnosed CRCs by screening colonoscopy according to sex and age in the initial 10 years of the German screening colonoscopy program.
Methods German Screening Colonoscopy Program The offer of screening colonoscopy was introduced in Germany in October 2002 as an alternative to fecal occult Abbreviations used in this paper: CRC, colorectal cancer; SHI, statutory health insurance. © 2014 by the AGA Institute 1542-3565/$36.00 http://dx.doi.org/10.1016/j.cgh.2014.08.036
2
Brenner et al
blood testing (which has been offered since 1977) for women and men aged 55 years or older covered by Statutory Health Insurance (SHI). If the first screening colonoscopy is conducted when younger than 65 years of age, a second screening colonoscopy is offered 10 years later. Close to 90% of older adults in Germany are covered by SHI, and the vast majority of those not included have private health insurance that provides equivalent screening offers in most cases.
National Screening Colonoscopy Registry Along with introduction of screening colonoscopy, a national registry was built up,9–11 to which all screening colonoscopies among people covered by SHI are reported anonymously on a standardized form (no such registry exists for privately insured people). Reporting is virtually complete, because it is a prerequisite for physicians’ reimbursement by SHI funds. The registry includes only primary screening examinations (ie, it does not include colonoscopies conducted for surveillance, work-up of symptoms, or other screening tests). Findings at colonoscopy are reported, including number, size, and histologic characteristics of polyps. In case of multiple neoplasms, only the most advanced finding (nonadvanced adenoma, advanced adenoma, or cancer) is recorded. Advanced adenomas are defined as at least 1 adenoma 1 cm or at least 1 adenoma with villous components or high-grade dysplasia. Within the initial 10 years of this screening offer, approximately 22% of eligible women and 20% of eligible men covered by SHI had a screening colonoscopy. For this analysis, we used data from 4,407,971 first-time screening colonoscopies in 2003–2012 among participants aged 55 years or older included in the national screening colonoscopy registry.
Statistical Analysis We first derived numbers of participants of screening colonoscopy aged 55–80 years in 2003–2012 by sex, single year of age, and most advanced finding at colonoscopy. Next, for each sex and each single year of age, we derived the probability of having a CRC prevented or early detected and for having a CRC overdiagnosis, denoted Probprev, Probearly, and Probover, respectively, as Probprev ¼ Pnon Cnon þ Padv Cadv RR; Probearly ¼ PCRC CCRC ; and Probover ¼ PCRC ð1 CCRC Þ; where Pnon, Padv, and PCRC denote the sex- and age-specific prevalences at screening colonoscopy of nonadvanced adenoma, advanced adenoma, and CRC, respectively, Cnon, Cadv, and CCRC denote the sex- and age-specific probabilities
Clinical Gastroenterology and Hepatology Vol.
-,
No.
-
of developing clinically manifest CRC during lifetime of carriers of nonadvanced adenoma, advanced adenoma, and CRC, respectively, and RR denotes the relative reduction of CRC risk by screening colonoscopy. In this approach, overdiagnoses are defined as cancers detected at screening colonoscopy that would not have become clinically manifest during lifetime without screening colonoscopy. Early detected cancers are defined as the complementary group of cancers detected at screening colonoscopy in which screening colonoscopy led to earlier detection of the disease that would otherwise become clinically manifest later in life. RR was assumed to be 69% on the basis of a recent meta-analysis.4 Pnon, Padv, and PCRC were directly obtained from the screening colonoscopy registry. Cnon, Cadv, and CCRC were derived by 4 state Markov models, with annual iterations starting at the specific age of colonoscopy up to a maximum age of 100 (variation of this upper age limit by 5 years had very little impact on the results). At each iteration, progression between states (carriage of nonadvanced adenoma, advanced adenoma, preclinical cancer, clinically manifest cancer) was modeled on the basis of previously derived sex- and age-specific annual transition rates.11,12 These estimates had been derived at very high levels of precision by birth cohort analyses by using the German national screening colonoscopy registry and by combining CRC prevalence estimates from the German national screening colonoscopy database and from national cancer incidence data (Table 1).11,12 For each iteration and each transition, we accounted for mortality that was obtained from general population life tables for the year 2010.13 Because estimates of sex- and age-specific transition rates were available up to age groups 75–79 (transition rates from nonadvanced adenoma to advanced adenoma and from advanced adenoma to preclinical cancer) or 80–84 (transition rates from preclinical to clinically manifest cancer) only, available transition rates for these age groups were assumed to also apply at older ages. To account for uncertainties in the assumed transition rates, sensitivity analyses were carried out in which all transition rates were varied between the lower and upper ends of the 95% confidence intervals that are shown in Table 1 for each sex and all age groups. Finally, we calculated cumulative numbers of prevented and early detected CRCs and of overdiagnoses from age 55 on up to various ages between 55 and 80 years for each sex. Sex- and age-specific numbers to be added up were obtained by multiplying sex- and agespecific numbers of screening colonoscopy participants with sex- and age-specific estimates of Probprev, Probearly, and Probover. To derive estimates for the total German population, the numbers of screening colonoscopy participants included in this analysis were determined by multiplying sex- and age-specific numbers of registered screening colonoscopy participants covered by SHI with the inverse values of the sex- and age-specific SHI coverage proportions during the period of investigation.
-
2014
10 Years of Screening Colonoscopy in Germany
3
Table 1. Annual Transition Rates Between the 4 States of Colorectal Neoplasms Used in This Analysis Annual transition rates, % (95% confidence intervals)a Nonadvanced adenoma to advanced adenoma11 Men (y) 55–59 60–64 65–69 70–74 75–79 80–84 85þ Women (y) 55–59 60–64 65–69 70–74 75–79 80–84 85þ
Advanced adenoma to preclinical cancer11
Preclinical cancer to clinically manifest cancer12
4.2 4.0 4.0 4.1 3.7 3.7 3.7
(3.8–4.6) (3.6–4.4) (3.6–4.3) (3.6–4.6) (2.9–4.6) (2.9–4.6)b (2.9–4.6)b
2.6 3.1 3.8 5.1 5.2 5.2 5.2
(2.4–2.9) (2.8–3.3) (3.5–4.1) (4.8–5.5) (4.6–5.8) (4.6–5.8)b (4.6–5.8)b
18.1 19.2 21.3 20.6 20.1 18.2 18.2
(16.7–19.5) (18.1–20.3) (20.3–22.4) (19.5–21.7) (18.9–21.4) (16.7–19.9) (16.7–19.9)c
4.0 3.6 3.7 4.7 3.7 3.7 3.7
(3.6–4.5) (3.2–4.1) (3.2–4.1) (4.1–5.3) (2.8–4.7) (2.8–4.7)b (2.8–4.7)b
2.5 2.7 3.8 5.0 5.6 5.6 5.6
(2.2–2.7) (2.4–3.0) (3.5–4.1) (4.5–5.4) (4.9–6.3) (4.9–6.3)b (4.9–6.3)b
21.3 22.5 21.9 20.8 19.2 17.3 17.3
(19.5–23.4) (20.9–24.2) (20.6–23.3) (19.4–22.2) (17.9–20.7) (16.0–18.8) (16.0–18.8)c
a
Probability of progression from one state to the next state within 1 year in the absence of death from other causes. Transition rates assumed to be the same as in age group 75–79. Transition rates assumed to be the same as in age group 80–84.
b c
Thus, our results pertain to the entire German population rather than the SHI covered population only. An implicit assumption of this approach is that participation in and findings at screening colonoscopy are comparable between SHI covered and privately insured women and men.
Results Table 2 shows the numbers of participants in the German screening colonoscopy program included in the national registry in 2003–2012 according to sex, age, and most advanced finding at colonoscopy. Overall, records from almost 2 million men and more than 2.4 million women were included in the database. Slightly more than half of them were between 55 and 64 years of age; only a minority of approximately 2.5% were 80 years or older. At least 1 neoplasm (adenoma or cancer) was found in 28.5% of men and 17.6% of women. The age distribution was shifted toward older ages in these men and women, with proportions younger than 65 years of age decreasing to close to 50%, 45%–46%, and 31%–32% among participants whose most advanced finding was a nonadvanced adenoma, an advanced adenoma, and cancer, respectively. The probabilities of having a CRC prevented or early detected and of having an overdiagnosis of CRC according to sex and age at screening colonoscopy are shown in Figure 1. All probabilities are higher among men than among women at all ages, with probabilities of having a CRC prevented being by far highest within both sexes. Probabilities of prevention are highest for screening
colonoscopies conducted around 60 years of age (close to 5% among men, close to 3.5% among women) and lowest at oldest age (below 2% at age 80 in both men and women). By contrast, the probability of having a CRC early detected increases with age at colonoscopy from below 0.5% at 55 years of age to slightly above 2% at 80 years among men and from less than 0.3% at 55 years of age to approximately 1.7% at 80 years among women. Probabilities of overdiagnosis are below 0.1% up to 65 years of age among men and up to 71 years of age among women, and they gradually increase to 0.9% and 0.6% at age 80 in men and women, respectively. Sensitivity analyses that use the lower end of the 95% confidence intervals for the transition rates shown in Table 1 yielded 11%–18% lower probabilities of potential CRC prevention, 0.3%–3.6% lower probabilities of CRC early detection, and 9%–16% higher probabilities of overdiagnosis. An opposite pattern was observed when using the upper end of the 95% confidence intervals of the transition rates. However, the overall patterns of sex- and age-specific probabilities remained essentially unchanged. Cumulative numbers of CRC cases prevented and early detected and of CRC overdiagnoses by screening colonoscopies up to various ages are shown in Figure 2. Overall, screening colonoscopies conducted in Germany in 2003–2012 are estimated to have prevented approximately 180,000 CRCs (a slight majority of which among men), which corresponds to 1 prevented CRC per 28 screening colonoscopies. The vast majority (97%) of these cases were prevented by screening colonoscopies conducted up to 75 years of age. In addition, more than 40,000 CRC cases that would otherwise have become
No.
567,036 468,912 464,297 299,299 139,834 50,432 1,989,810 769,572 570,926 528,910 327,492 157,038 64,223 2,418,161
28.5 23.6 23.3 15.0 7.0 2.5 100 31.8 23.6 21.9 13.5 6.5 2.7 100
100 100 100 100 100 100 100 100 100 100 100 100 100 100
428,428 338,257 324,037 202,552 94,193 34,320 1,421,787 659,272 475,155 428,847 257,852 122,170 49,196 1,992,492
30.1 23.8 22.8 14.2 6.6 2.4 100 33.1 23.8 21.5 12.9 6.1 2.5 100
75.6 72.1 69.8 67.7 67.4 68.1 71.5 85.7 83.2 81.1 78.7 77.8 76.6 82.4
97,406 86,484 90,502 60,682 27,242 9125 371,441 79,939 66,356 67,697 45,868 21,609 8445 289,914
26.2 23.3 24.4 16.3 7.3 2.5 100 27.6 22.9 23.4 15.8 7.5 2.9 100
17.2 18.4 19.5 20.3 19.5 18.1 18.7 10.4 11.6 12.8 14.0 13.8 13.1 12.0
37,801 39,465 43,467 30,197 14,780 5231 170,941 27,932 26,550 28,463 20,186 10,754 4829 118,714
22.1 23.1 25.4 17.7 8.6 3.1 100 23.5 22.4 24.0 17.0 9.1 4.1 100
6.7 8.4 9.4 10.1 10.6 10.4 8.6 3.6 4.7 5.4 6.2 6.8 7.5 4.9
3401 4706 6291 5868 3619 1756 25,641 2429 2865 3903 3586 2505 1753 17,041
13.3 18.4 24.5 22.9 14.1 6.8 100 14.3 16.8 22.9 21.0 14.7 10.3 100
0.6 1.0 1.4 2.0 2.6 3.5 1.3 0.3 0.5 0.7 1.1 1.6 2.7 0.7
-,
Women
55–59 60–64 65–69 70–74 75–79 80þ Total 55–59 60–64 65–69 70–74 75–79 80þ Total Men
Column % Row % N Column % Row % Sex
Age (y)
N
Column % Row %
N
Column % Row %
N
Column % Row %
N
Advanced adenoma Nonadvanced adenoma No neoplasm
This is a comprehensive, joint analysis of prevented and early detected CRCs and of overdiagnoses in a national screening colonoscopy program. On the basis of data from more than 4.4 million screening colonoscopies from the German national screening colonoscopy registry, we demonstrate that probabilities of prevention or early detection of CRC by far exceed probabilities of overdiagnosis in both sexes and all ages up to 80 years. Probabilities of prevention are highest for screening colonoscopies conducted around 60 years of age (close to 5% among men, close to 3.5% among women) and lowest at oldest age (below 2% at age 80 in both men and women). Probabilities of early detection increase with age up to approximately 2% for screening colonoscopies conducted at age 80. Probabilities of overdiagnosis also increase with age, albeit at much lower levels (below 0.1% at age 65 and below 1% at age 80 among both men and women). Overall, approximately 180,000 CRCs were prevented and more than 40,000 were early detected by screening colonoscopies among participants younger than age 80 in 2003–2012 (1 per 28 and 1 per 121 screening colonoscopies, respectively), compared with approximately 4500 overdiagnoses (1 per 1089 screening colonoscopies). The vast majority of CRCs (97%) were prevented by screening colonoscopies conducted up to 75 years of age. Lifetime risk of CRC in Germany has been estimated to be 7.5% among men and 6.1% among women in 2007/2008.14 These estimates reflect average risks of people who did and did not undergo colonoscopy, and because a relevant proportion of older adults meanwhile had a colonoscopy at least once in their life,15 lifetime risks with no previous colonoscopy are likely to be somewhat higher. Our analyses suggest that a screening colonoscopy between 55 and 70 years of age can take away a large proportion of this risk. The risk reduction becomes smaller at older ages, but even between 75 and 80 years of age, it is much larger than the risk of overdiagnosis in both men and women. The latter also applies to the chances of early detection of CRC compared with the risk of overdiagnosis. To our knowledge, quantitative estimates of overdiagnoses were not previously available for any national
All participants
Discussion
Most advanced finding at colonoscopy
clinically manifest during lifetime (1 per 121 screening colonoscopies) were early detected up to age 80 (approximately 25,000 and 16,000 among men and women, respectively), 89% of them by screening colonoscopies conducted up to age 75. Cumulative numbers of overdiagnoses are estimated to be slightly more than 4500 (1 per 1089 screening colonoscopies, 70% of them among men). A majority of 58% of overdiagnoses arise from screening colonoscopies conducted older than 70 years of age (28% from screening colonoscopies conducted older than 75 years).
Clinical Gastroenterology and Hepatology Vol.
CRC
Brenner et al
Table 2. Numbers of Participants of Screening Colonoscopy According to Sex, Age, and Most Advanced Finding at Colonoscopy From German Screening Colonoscopy Registry, 2003–2012
4
-
-
2014
10 Years of Screening Colonoscopy in Germany
5
Figure 1. Probability of having CRC prevented or early detected or having overdiagnosis of CRC, according to sex and age at screening colonoscopy.
CRC screening program. Whereas the potential of overdiagnoses is often used as an argument against screening offers, our analyses clearly indicate that the number of overdiagnoses is very small compared with the numbers of prevented and early detected CRCs. The proportion of overdiagnoses among screening colonoscopy detected cancers is also much lower compared with screening options for other cancers. For example, the proportion of overdiagnoses among breast cancers detected at screening mammography in women aged 50–70 has been estimated to be 19%.16 In our analysis, the proportion of overdiagnoses among screening colonoscopy detected CRCs was 11% and 8%, respectively, among men and women overall and 7% and 4%, respectively, for screening colonoscopies conducted younger than age 70. For prostate-specific antigen screening among men aged 50–74 years, the risk of prostate cancer overdiagnosis has been estimated to be in the order of 2.3%– 3.3%, depending on screening intervals and thresholds for biopsy referral.17 In our analysis, the risk of CRC
overdiagnoses was below 0.4% for screening colonoscopies conducted among men younger than 75 years. In contrast to other cancer screening programs, screening colonoscopy is furthermore expected to result in major reduction not only of mortality but also of incidence because the number of prevented cancers by detection and removal of adenomas is many-fold higher than the number of overdiagnoses. On the other hand, a major proportion of detected adenomas would never develop into clinically manifest cancer but would prompt 1 or more surveillance colonoscopies. Such “oversurveillance,” which also carries some additional risk of overdiagnoses at a later point of time, should also be kept in mind in a comprehensive evaluation of benefits and harms of screening colonoscopy. Time course of effects is another major issue in the evaluation of screening effects. Whereas early detection and overdiagnoses occur immediately and may lead to a transient apparent increase in incidence, a large proportion of prevented cancers would have occurred many
Figure 2. Estimated cumulative numbers of CRC cases prevented and early detected and of CRC overdiagnoses by screening colonoscopies up to various ages. Total German population, 2003–2012.
6
Brenner et al
years after colonoscopy, and the impact of prevention on CRC incidence may only be seen in the long run. Whereas substantial reduction of CRC incidence has been observed in the United States18 where screening by sigmoidoscopy and colonoscopy has reached larger proportions of the population already in the 1980s and 1990s,19 first reductions in CRC incidence have been observed in Germany only very recently.20 Because trends in CRC incidence are determined by many factors, including screening colonoscopy, screening by fecal occult blood test, and trends in risk and preventive factors unrelated to screening and because incidence reduction resulting from screening is expected only with substantial delay, it is uncertain how much of this decline can be ascribed to screening colonoscopy. However, our analyses suggest that screening colonoscopy will contribute to further substantial reductions in CRC incidence in the years to come, which could be even larger with more widespread acceptance of the screening offers. Our study has a number of strengths and limitations. A major strength is its reliance on a very large database from a national screening colonoscopy registry that minimizes potential selection effects and limitations in external validity that might be encountered in studies restricted to specialized academic centers. Nevertheless, screening colonoscopy data were not available for a minority of 10% mostly privately insured people in Germany for whom we assumed comparable sex- and age-specific participation rates in screening colonoscopy and prevalences of neoplasms. Even though this assumption may not hold because privately insured patients are typically better educated and more health conscious, even major differences in participation rates and neoplasm prevalences would affect overall results only to a very minor extent because of the small share of this population group. The achievable degree of quality and standardization of colonoscopies and their documentation might be lower in a nationwide program than in studies restricted to specialized centers, despite major efforts of quality assurance in the German screening colonoscopy program.21,22 For example, prevalences of adenomas, in particular nonadvanced adenomas, are most likely to be somewhat underestimated because of non-negligible adenoma miss rates.23 Missed neoplasms would not contribute to prevented cancers, early detection, or overdiagnosis. Therefore, our main results that reflect prevention, early detection, and overdiagnoses resulting from neoplasms actually detected in routine screening practice should not be biased by missed neoplasms. Obviously, the validity of our calculations depends on the validity of the assumed transition rates from nonadvanced to advanced adenomas, from advanced adenomas to preclinical CRC, and from preclinical to clinically manifest CRC. Sex- and age-specific estimates of these transition rates have previously been derived at very high levels of precision partly from the same database as previously described.11,12 Our sensitivity
Clinical Gastroenterology and Hepatology Vol.
-,
No.
-
analyses indicate that statistical uncertainties regarding the assumed transition rates would not materially alter the overall patterns of sex- and age-specific probabilities of potentially prevented and early detected CRCs and of overdiagnoses. Despite its limitations, our analysis underlines the large potential of screening colonoscopy for prevention and early detection of CRC that can be achieved at low risk of overdiagnosis, even at advanced ages. In addition, more detailed analyses should aim for more precisely defining the best upper age limit for first-time CRC screening in a country like Germany where life expectancy has become very high. Such analyses should also take additional factors into account, such as potential complications of colonoscopy, comorbidity, and costs.24,25 Further work should be expanded to include estimates of potentially prevented CRC deaths, which will require reasonably accurate measures of ageand sex-specific survival according to year after diagnosis for screening detected and other CRCs. Notwithstanding the need for further research regarding these important additional questions, efforts to enhance acceptance of screening colonoscopy should be made with priority, which could further substantially enhance reduction of CRC incidence and subsequently CRC mortality.
References 1. Hewitson P, Glasziou P, Watson E, et al. Cochrane systematic review of colorectal cancer screening using the fecal occult blood test (hemoccult): an update. Am J Gastroenterol 2008; 103:1541–1549. 2. Scholefield JH, Moss SM, Mangham CM, et al. Nottingham trial of faecal occult blood testing for colorectal cancer: a 20-year follow-up. Gut 2012;61:1036–1040. 3. Shaukat A, Mongin SJ, Geisser MS, et al. Long-term mortality after screening for colorectal cancer. N Engl J Med 2013; 369:1106–1114. 4. Brenner H, Stock C, Hoffmeister M. Effect of screening sigmoidoscopy and screening colonoscopy on colorectal cancer incidence and mortality: systematic review and metaanalysis of randomised controlled trials and observational studies. BMJ 2014. 5. Winawer S, Fletcher R, Rex D, et al. Colorectal cancer screening and surveillance: clinical guidelines and rationale—update based on new evidence. Gastroenterology 2003;124:544–560. 6. U.S. Preventive Services Task Force. Screening for colorectal cancer: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med 2008;149:627–637. 7. Levin B, Lieberman DA, McFarland B, 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. CA Cancer J Clin 2008;58:130–160. 8. Schmiegel W, Pox C, Adler G, et al. S3 Guidelines Colorectal Cancer 2004 [in German]. Z Gastroenterol 2004;42:1129–1177. 9. Brenner H, Altenhofen L, Hoffmeister M. Sex, age and birth cohort effects in colorectal neoplasms: a cohort analysis. Ann Intern Med 2010;102:1411–1414.
-
2014
10 Years of Screening Colonoscopy in Germany
7
10. Pox C, Altenhofen L, Brenner H, et al. Efficacy of a nationwide screening colonoscopy program for colorectal cancer. Gastroenterology 2012;142:1460–1467.
20. Association of Population-based Cancer Registries in Germany (GEKID). GEKID data tables. Available: http://www.gekid.de/. Accessed April 23, 2013.
11. Brenner H, Altenhofen L, Stock C, et al. Natural history of colorectal adenomas: birth cohort analysis among 3.6 million participants of screening colonoscopy. Cancer Epidemiol Biomarkers Prev 2013;22:1043–1051.
21. Adler A, Wegscheider K, Lieberman D, et al. Factors determining the quality of screening colonoscopy: a prospective study on adenoma detection rates, from 12,134 examinations (Berlin colonoscopy project 3, BECOP-3). Gut 2013;62:236–241.
12. Brenner H, Altenhofen L, Katalinic A, et al. Sojourn time of preclinical colorectal cancer by sex and age: estimates from the German national screening colonoscopy database. Am J Epidemiol 2011;174:1140–1146. 13. Statistisches Bundesamt. Bevölkerung und Erwerbstätigkeit: Sterbetafel Deutschland 2008/2010. Wiesbaden: Statistisches Bundesamt, 2011.
22. Adler A, Lieberman D, Aminalai A, et al. Data quality of the German screening colonoscopy registry. Endoscopy 2013; 45:813–818.
14. Robert Koch Institut, Gesellschaft der epidemiologischen Krebsregister in Deutschland (eds). Krebs in Deutschland 2007/2008. 8th ed. Berlin: Robert Koch Institut, 2012. 15. Stock C, Ihle P, Schubert I, et al. Colonoscopy and fecal occult blood test use in Germany: results from a large insurance-based cohort. Endoscopy 2011;43:771–779. 16. Independent UK Panel on Breast Cancer Screening. The benefits and harms of breast cancer screening: an independent review. Lancet 2012;380:1778–1786. 17. Gulati R, Gore JL, Etzioni R. Comparative effectiveness of alternative prostate-specific antigen-based prostate cancer screening strategies: model estimates of potential benefits and harms. Ann Intern Med 2013;158:145–153. 18. Stock C, Pulte D, Haug U, et al. Subsite-specific colorectal cancer risk in the colorectal endoscopy era. Gastrointest Endosc 2012;75:621–630. 19. Stock C, Haug U, Brenner H. Population-based prevalence estimates of history of colonoscopy or sigmoidoscopy: review and analysis of recent trends. Gastrointest Endosc 2010;71:366–381.
23. Van Rijn JC, Reitsma JB, Stoker J, et al. Polyp miss rate determined by tandem colonoscopy: a systematic review. Am J Gastroenterol 2006;101:343–350. 24. Stock C, Ihle P, Sieg A, et al. Adverse events requiring hospitalization within 30 days after outpatient screening and nonscreening colonoscopy. Gastrointest Endosc 2013;77:419–429. 25. Sharaf RN, Ladabaum U. Comparative effectiveness and costeffectiveness of screening colonoscopy vs sigmoidoscopy and alternative strategies. Am J Gastroenterol 2013;108:120–132.
Reprint requests Address correspondence to: Hermann Brenner, MD, MPH, Division of Clinical Epidemiology and Aging Research, German Cancer Research Center, Im Neuenheimer Feld 581, D-69120 Heidelberg, Germany. e-mail: h.brenner@ dkfz.de; fax: þ49-6221-421302. Conflicts of interest The authors disclose no conflicts. Funding The German screening colonoscopy registry is funded by the National Association of Statutory Health Insurance Physicians and the National Association of Statutory Health Insurance Funds. This study was further supported in part by grants from the German Research Council (Deutsche Forschungsgemeinschaft, grant number BR 1704/6-4) and the German Federal Ministry of Education and Research (grant number 01KH0404, 01ER0814).
Prevention, Early Detection, and Overdiagnosis of Colorectal Cancer Within 10 Years of Screening Colonoscopy in Germany Hermann Brenner,*,‡ Lutz Altenhofen,§ Christian Stock,*,k and Michael Hoffmeister* *Division of Clinical Epidemiology and Aging Research, German Cancer Research Center (DKFZ), Heidelberg, Germany; ‡ German Cancer Consortium (DKTK), Heidelberg, Germany; §Central Research Institute of Ambulatory Health Care in Germany, Berlin, Germany; and kInstitute of Medical Biometry and Informatics, University of Heidelberg, Heidelberg, Germany BACKGROUND & AIMS:
Screening colonoscopy was introduced in Germany in October 2002. We aimed to quantify its effects on prevention, early detection, and overdiagnosis of colorectal cancer (CRC) in the 10 years since its introduction.
METHODS:
We analyzed data from more than 4.4 million screening colonoscopies (conducted on individuals 55–79 years old from 2003 through 2012) available through the national screening colonoscopy registry. CRCs prevented, detected earlier than they would have been without screening, and overdiagnosed (cancers detected at screening colonoscopy that would not have become clinically manifest during the patient’s lifetime) were estimated by Markov models. Model parameters included sex-specific and age-specific findings at screening colonoscopy; mortality; rates of transition from nonadvanced to advanced adenoma, advanced adenoma to preclinical cancer, or preclinical cancer to clinically manifest cancer; and protection from screening colonoscopy.
RESULTS:
Overall, approximately 180,000 CRCs (1/28 screening colonoscopies) were estimated to have been prevented, and more than 40,000 CRCs (1/121 screening colonoscopies) were detected earlier than they would have been without screening, compared with approximately 4500 overdiagnoses (1/1089 screening colonoscopies). Almost all CRCs prevented or detected earlier than they would have been without screening resulted from screening colonoscopies performed on individuals up to 75 years old (97% and 89%, respectively), whereas 28% of overdiagnoses occurred from screening colonoscopies of individuals older than 75 years old.
CONCLUSIONS:
On the basis of a 10-year analysis of data from a national registry in Germany, screening colonoscopies have large potential for prevention and early detection of CRC, with low risk of overdiagnosis.
Keywords: Adenomas; Colon Cancer; Colorectal Neoplasms; Tumor.
olorectal cancer (CRC) is one of few cancers for which effective screening options are established. Randomized controlled trials have shown reduction of CRC mortality by annual or biennial fecal occult blood test screening1–3 and reduction of both CRC incidence and mortality by screening sigmoidoscopy.4 Although long-term results from randomized controlled trials for screening colonoscopy are not available yet, observational studies suggest even stronger reductions of CRC incidence and mortality,4 and screening colonoscopy has been recommended for CRC prevention and early detection by expert panels for more than 10 years.5–8 Germany was one of the first countries in the world to introduce colonoscopy as a primary screening offer. Timely estimation of its impact on prevention and early detection but also on potential overdiagnoses of CRC is crucial for decisions regarding maintenance or eventual adaptation and optimization of screening colonoscopy in Germany and for informing decisions on introduction,
C
maintenance, or adaptation of CRC screening offers in other countries. In this article, we aim for a comprehensive, joint, and detailed analysis of prevented, early detected, and overdiagnosed CRCs by screening colonoscopy according to sex and age in the initial 10 years of the German screening colonoscopy program.
Methods German Screening Colonoscopy Program The offer of screening colonoscopy was introduced in Germany in October 2002 as an alternative to fecal occult Abbreviations used in this paper: CRC, colorectal cancer; SHI, statutory health insurance. © 2014 by the AGA Institute 1542-3565/$36.00 http://dx.doi.org/10.1016/j.cgh.2014.08.036
2
Brenner et al
blood testing (which has been offered since 1977) for women and men aged 55 years or older covered by Statutory Health Insurance (SHI). If the first screening colonoscopy is conducted when younger than 65 years of age, a second screening colonoscopy is offered 10 years later. Close to 90% of older adults in Germany are covered by SHI, and the vast majority of those not included have private health insurance that provides equivalent screening offers in most cases.
National Screening Colonoscopy Registry Along with introduction of screening colonoscopy, a national registry was built up,9–11 to which all screening colonoscopies among people covered by SHI are reported anonymously on a standardized form (no such registry exists for privately insured people). Reporting is virtually complete, because it is a prerequisite for physicians’ reimbursement by SHI funds. The registry includes only primary screening examinations (ie, it does not include colonoscopies conducted for surveillance, work-up of symptoms, or other screening tests). Findings at colonoscopy are reported, including number, size, and histologic characteristics of polyps. In case of multiple neoplasms, only the most advanced finding (nonadvanced adenoma, advanced adenoma, or cancer) is recorded. Advanced adenomas are defined as at least 1 adenoma 1 cm or at least 1 adenoma with villous components or high-grade dysplasia. Within the initial 10 years of this screening offer, approximately 22% of eligible women and 20% of eligible men covered by SHI had a screening colonoscopy. For this analysis, we used data from 4,407,971 first-time screening colonoscopies in 2003–2012 among participants aged 55 years or older included in the national screening colonoscopy registry.
Statistical Analysis We first derived numbers of participants of screening colonoscopy aged 55–80 years in 2003–2012 by sex, single year of age, and most advanced finding at colonoscopy. Next, for each sex and each single year of age, we derived the probability of having a CRC prevented or early detected and for having a CRC overdiagnosis, denoted Probprev, Probearly, and Probover, respectively, as Probprev ¼ Pnon Cnon þ Padv Cadv RR; Probearly ¼ PCRC CCRC ; and Probover ¼ PCRC ð1 CCRC Þ; where Pnon, Padv, and PCRC denote the sex- and age-specific prevalences at screening colonoscopy of nonadvanced adenoma, advanced adenoma, and CRC, respectively, Cnon, Cadv, and CCRC denote the sex- and age-specific probabilities
Clinical Gastroenterology and Hepatology Vol.
-,
No.
-
of developing clinically manifest CRC during lifetime of carriers of nonadvanced adenoma, advanced adenoma, and CRC, respectively, and RR denotes the relative reduction of CRC risk by screening colonoscopy. In this approach, overdiagnoses are defined as cancers detected at screening colonoscopy that would not have become clinically manifest during lifetime without screening colonoscopy. Early detected cancers are defined as the complementary group of cancers detected at screening colonoscopy in which screening colonoscopy led to earlier detection of the disease that would otherwise become clinically manifest later in life. RR was assumed to be 69% on the basis of a recent meta-analysis.4 Pnon, Padv, and PCRC were directly obtained from the screening colonoscopy registry. Cnon, Cadv, and CCRC were derived by 4 state Markov models, with annual iterations starting at the specific age of colonoscopy up to a maximum age of 100 (variation of this upper age limit by 5 years had very little impact on the results). At each iteration, progression between states (carriage of nonadvanced adenoma, advanced adenoma, preclinical cancer, clinically manifest cancer) was modeled on the basis of previously derived sex- and age-specific annual transition rates.11,12 These estimates had been derived at very high levels of precision by birth cohort analyses by using the German national screening colonoscopy registry and by combining CRC prevalence estimates from the German national screening colonoscopy database and from national cancer incidence data (Table 1).11,12 For each iteration and each transition, we accounted for mortality that was obtained from general population life tables for the year 2010.13 Because estimates of sex- and age-specific transition rates were available up to age groups 75–79 (transition rates from nonadvanced adenoma to advanced adenoma and from advanced adenoma to preclinical cancer) or 80–84 (transition rates from preclinical to clinically manifest cancer) only, available transition rates for these age groups were assumed to also apply at older ages. To account for uncertainties in the assumed transition rates, sensitivity analyses were carried out in which all transition rates were varied between the lower and upper ends of the 95% confidence intervals that are shown in Table 1 for each sex and all age groups. Finally, we calculated cumulative numbers of prevented and early detected CRCs and of overdiagnoses from age 55 on up to various ages between 55 and 80 years for each sex. Sex- and age-specific numbers to be added up were obtained by multiplying sex- and agespecific numbers of screening colonoscopy participants with sex- and age-specific estimates of Probprev, Probearly, and Probover. To derive estimates for the total German population, the numbers of screening colonoscopy participants included in this analysis were determined by multiplying sex- and age-specific numbers of registered screening colonoscopy participants covered by SHI with the inverse values of the sex- and age-specific SHI coverage proportions during the period of investigation.
-
2014
10 Years of Screening Colonoscopy in Germany
3
Table 1. Annual Transition Rates Between the 4 States of Colorectal Neoplasms Used in This Analysis Annual transition rates, % (95% confidence intervals)a Nonadvanced adenoma to advanced adenoma11 Men (y) 55–59 60–64 65–69 70–74 75–79 80–84 85þ Women (y) 55–59 60–64 65–69 70–74 75–79 80–84 85þ
Advanced adenoma to preclinical cancer11
Preclinical cancer to clinically manifest cancer12
4.2 4.0 4.0 4.1 3.7 3.7 3.7
(3.8–4.6) (3.6–4.4) (3.6–4.3) (3.6–4.6) (2.9–4.6) (2.9–4.6)b (2.9–4.6)b
2.6 3.1 3.8 5.1 5.2 5.2 5.2
(2.4–2.9) (2.8–3.3) (3.5–4.1) (4.8–5.5) (4.6–5.8) (4.6–5.8)b (4.6–5.8)b
18.1 19.2 21.3 20.6 20.1 18.2 18.2
(16.7–19.5) (18.1–20.3) (20.3–22.4) (19.5–21.7) (18.9–21.4) (16.7–19.9) (16.7–19.9)c
4.0 3.6 3.7 4.7 3.7 3.7 3.7
(3.6–4.5) (3.2–4.1) (3.2–4.1) (4.1–5.3) (2.8–4.7) (2.8–4.7)b (2.8–4.7)b
2.5 2.7 3.8 5.0 5.6 5.6 5.6
(2.2–2.7) (2.4–3.0) (3.5–4.1) (4.5–5.4) (4.9–6.3) (4.9–6.3)b (4.9–6.3)b
21.3 22.5 21.9 20.8 19.2 17.3 17.3
(19.5–23.4) (20.9–24.2) (20.6–23.3) (19.4–22.2) (17.9–20.7) (16.0–18.8) (16.0–18.8)c
a
Probability of progression from one state to the next state within 1 year in the absence of death from other causes. Transition rates assumed to be the same as in age group 75–79. Transition rates assumed to be the same as in age group 80–84.
b c
Thus, our results pertain to the entire German population rather than the SHI covered population only. An implicit assumption of this approach is that participation in and findings at screening colonoscopy are comparable between SHI covered and privately insured women and men.
Results Table 2 shows the numbers of participants in the German screening colonoscopy program included in the national registry in 2003–2012 according to sex, age, and most advanced finding at colonoscopy. Overall, records from almost 2 million men and more than 2.4 million women were included in the database. Slightly more than half of them were between 55 and 64 years of age; only a minority of approximately 2.5% were 80 years or older. At least 1 neoplasm (adenoma or cancer) was found in 28.5% of men and 17.6% of women. The age distribution was shifted toward older ages in these men and women, with proportions younger than 65 years of age decreasing to close to 50%, 45%–46%, and 31%–32% among participants whose most advanced finding was a nonadvanced adenoma, an advanced adenoma, and cancer, respectively. The probabilities of having a CRC prevented or early detected and of having an overdiagnosis of CRC according to sex and age at screening colonoscopy are shown in Figure 1. All probabilities are higher among men than among women at all ages, with probabilities of having a CRC prevented being by far highest within both sexes. Probabilities of prevention are highest for screening
colonoscopies conducted around 60 years of age (close to 5% among men, close to 3.5% among women) and lowest at oldest age (below 2% at age 80 in both men and women). By contrast, the probability of having a CRC early detected increases with age at colonoscopy from below 0.5% at 55 years of age to slightly above 2% at 80 years among men and from less than 0.3% at 55 years of age to approximately 1.7% at 80 years among women. Probabilities of overdiagnosis are below 0.1% up to 65 years of age among men and up to 71 years of age among women, and they gradually increase to 0.9% and 0.6% at age 80 in men and women, respectively. Sensitivity analyses that use the lower end of the 95% confidence intervals for the transition rates shown in Table 1 yielded 11%–18% lower probabilities of potential CRC prevention, 0.3%–3.6% lower probabilities of CRC early detection, and 9%–16% higher probabilities of overdiagnosis. An opposite pattern was observed when using the upper end of the 95% confidence intervals of the transition rates. However, the overall patterns of sex- and age-specific probabilities remained essentially unchanged. Cumulative numbers of CRC cases prevented and early detected and of CRC overdiagnoses by screening colonoscopies up to various ages are shown in Figure 2. Overall, screening colonoscopies conducted in Germany in 2003–2012 are estimated to have prevented approximately 180,000 CRCs (a slight majority of which among men), which corresponds to 1 prevented CRC per 28 screening colonoscopies. The vast majority (97%) of these cases were prevented by screening colonoscopies conducted up to 75 years of age. In addition, more than 40,000 CRC cases that would otherwise have become
No.
567,036 468,912 464,297 299,299 139,834 50,432 1,989,810 769,572 570,926 528,910 327,492 157,038 64,223 2,418,161
28.5 23.6 23.3 15.0 7.0 2.5 100 31.8 23.6 21.9 13.5 6.5 2.7 100
100 100 100 100 100 100 100 100 100 100 100 100 100 100
428,428 338,257 324,037 202,552 94,193 34,320 1,421,787 659,272 475,155 428,847 257,852 122,170 49,196 1,992,492
30.1 23.8 22.8 14.2 6.6 2.4 100 33.1 23.8 21.5 12.9 6.1 2.5 100
75.6 72.1 69.8 67.7 67.4 68.1 71.5 85.7 83.2 81.1 78.7 77.8 76.6 82.4
97,406 86,484 90,502 60,682 27,242 9125 371,441 79,939 66,356 67,697 45,868 21,609 8445 289,914
26.2 23.3 24.4 16.3 7.3 2.5 100 27.6 22.9 23.4 15.8 7.5 2.9 100
17.2 18.4 19.5 20.3 19.5 18.1 18.7 10.4 11.6 12.8 14.0 13.8 13.1 12.0
37,801 39,465 43,467 30,197 14,780 5231 170,941 27,932 26,550 28,463 20,186 10,754 4829 118,714
22.1 23.1 25.4 17.7 8.6 3.1 100 23.5 22.4 24.0 17.0 9.1 4.1 100
6.7 8.4 9.4 10.1 10.6 10.4 8.6 3.6 4.7 5.4 6.2 6.8 7.5 4.9
3401 4706 6291 5868 3619 1756 25,641 2429 2865 3903 3586 2505 1753 17,041
13.3 18.4 24.5 22.9 14.1 6.8 100 14.3 16.8 22.9 21.0 14.7 10.3 100
0.6 1.0 1.4 2.0 2.6 3.5 1.3 0.3 0.5 0.7 1.1 1.6 2.7 0.7
-,
Women
55–59 60–64 65–69 70–74 75–79 80þ Total 55–59 60–64 65–69 70–74 75–79 80þ Total Men
Column % Row % N Column % Row % Sex
Age (y)
N
Column % Row %
N
Column % Row %
N
Column % Row %
N
Advanced adenoma Nonadvanced adenoma No neoplasm
This is a comprehensive, joint analysis of prevented and early detected CRCs and of overdiagnoses in a national screening colonoscopy program. On the basis of data from more than 4.4 million screening colonoscopies from the German national screening colonoscopy registry, we demonstrate that probabilities of prevention or early detection of CRC by far exceed probabilities of overdiagnosis in both sexes and all ages up to 80 years. Probabilities of prevention are highest for screening colonoscopies conducted around 60 years of age (close to 5% among men, close to 3.5% among women) and lowest at oldest age (below 2% at age 80 in both men and women). Probabilities of early detection increase with age up to approximately 2% for screening colonoscopies conducted at age 80. Probabilities of overdiagnosis also increase with age, albeit at much lower levels (below 0.1% at age 65 and below 1% at age 80 among both men and women). Overall, approximately 180,000 CRCs were prevented and more than 40,000 were early detected by screening colonoscopies among participants younger than age 80 in 2003–2012 (1 per 28 and 1 per 121 screening colonoscopies, respectively), compared with approximately 4500 overdiagnoses (1 per 1089 screening colonoscopies). The vast majority of CRCs (97%) were prevented by screening colonoscopies conducted up to 75 years of age. Lifetime risk of CRC in Germany has been estimated to be 7.5% among men and 6.1% among women in 2007/2008.14 These estimates reflect average risks of people who did and did not undergo colonoscopy, and because a relevant proportion of older adults meanwhile had a colonoscopy at least once in their life,15 lifetime risks with no previous colonoscopy are likely to be somewhat higher. Our analyses suggest that a screening colonoscopy between 55 and 70 years of age can take away a large proportion of this risk. The risk reduction becomes smaller at older ages, but even between 75 and 80 years of age, it is much larger than the risk of overdiagnosis in both men and women. The latter also applies to the chances of early detection of CRC compared with the risk of overdiagnosis. To our knowledge, quantitative estimates of overdiagnoses were not previously available for any national
All participants
Discussion
Most advanced finding at colonoscopy
clinically manifest during lifetime (1 per 121 screening colonoscopies) were early detected up to age 80 (approximately 25,000 and 16,000 among men and women, respectively), 89% of them by screening colonoscopies conducted up to age 75. Cumulative numbers of overdiagnoses are estimated to be slightly more than 4500 (1 per 1089 screening colonoscopies, 70% of them among men). A majority of 58% of overdiagnoses arise from screening colonoscopies conducted older than 70 years of age (28% from screening colonoscopies conducted older than 75 years).
Clinical Gastroenterology and Hepatology Vol.
CRC
Brenner et al
Table 2. Numbers of Participants of Screening Colonoscopy According to Sex, Age, and Most Advanced Finding at Colonoscopy From German Screening Colonoscopy Registry, 2003–2012
4
-
-
2014
10 Years of Screening Colonoscopy in Germany
5
Figure 1. Probability of having CRC prevented or early detected or having overdiagnosis of CRC, according to sex and age at screening colonoscopy.
CRC screening program. Whereas the potential of overdiagnoses is often used as an argument against screening offers, our analyses clearly indicate that the number of overdiagnoses is very small compared with the numbers of prevented and early detected CRCs. The proportion of overdiagnoses among screening colonoscopy detected cancers is also much lower compared with screening options for other cancers. For example, the proportion of overdiagnoses among breast cancers detected at screening mammography in women aged 50–70 has been estimated to be 19%.16 In our analysis, the proportion of overdiagnoses among screening colonoscopy detected CRCs was 11% and 8%, respectively, among men and women overall and 7% and 4%, respectively, for screening colonoscopies conducted younger than age 70. For prostate-specific antigen screening among men aged 50–74 years, the risk of prostate cancer overdiagnosis has been estimated to be in the order of 2.3%– 3.3%, depending on screening intervals and thresholds for biopsy referral.17 In our analysis, the risk of CRC
overdiagnoses was below 0.4% for screening colonoscopies conducted among men younger than 75 years. In contrast to other cancer screening programs, screening colonoscopy is furthermore expected to result in major reduction not only of mortality but also of incidence because the number of prevented cancers by detection and removal of adenomas is many-fold higher than the number of overdiagnoses. On the other hand, a major proportion of detected adenomas would never develop into clinically manifest cancer but would prompt 1 or more surveillance colonoscopies. Such “oversurveillance,” which also carries some additional risk of overdiagnoses at a later point of time, should also be kept in mind in a comprehensive evaluation of benefits and harms of screening colonoscopy. Time course of effects is another major issue in the evaluation of screening effects. Whereas early detection and overdiagnoses occur immediately and may lead to a transient apparent increase in incidence, a large proportion of prevented cancers would have occurred many
Figure 2. Estimated cumulative numbers of CRC cases prevented and early detected and of CRC overdiagnoses by screening colonoscopies up to various ages. Total German population, 2003–2012.
6
Brenner et al
years after colonoscopy, and the impact of prevention on CRC incidence may only be seen in the long run. Whereas substantial reduction of CRC incidence has been observed in the United States18 where screening by sigmoidoscopy and colonoscopy has reached larger proportions of the population already in the 1980s and 1990s,19 first reductions in CRC incidence have been observed in Germany only very recently.20 Because trends in CRC incidence are determined by many factors, including screening colonoscopy, screening by fecal occult blood test, and trends in risk and preventive factors unrelated to screening and because incidence reduction resulting from screening is expected only with substantial delay, it is uncertain how much of this decline can be ascribed to screening colonoscopy. However, our analyses suggest that screening colonoscopy will contribute to further substantial reductions in CRC incidence in the years to come, which could be even larger with more widespread acceptance of the screening offers. Our study has a number of strengths and limitations. A major strength is its reliance on a very large database from a national screening colonoscopy registry that minimizes potential selection effects and limitations in external validity that might be encountered in studies restricted to specialized academic centers. Nevertheless, screening colonoscopy data were not available for a minority of 10% mostly privately insured people in Germany for whom we assumed comparable sex- and age-specific participation rates in screening colonoscopy and prevalences of neoplasms. Even though this assumption may not hold because privately insured patients are typically better educated and more health conscious, even major differences in participation rates and neoplasm prevalences would affect overall results only to a very minor extent because of the small share of this population group. The achievable degree of quality and standardization of colonoscopies and their documentation might be lower in a nationwide program than in studies restricted to specialized centers, despite major efforts of quality assurance in the German screening colonoscopy program.21,22 For example, prevalences of adenomas, in particular nonadvanced adenomas, are most likely to be somewhat underestimated because of non-negligible adenoma miss rates.23 Missed neoplasms would not contribute to prevented cancers, early detection, or overdiagnosis. Therefore, our main results that reflect prevention, early detection, and overdiagnoses resulting from neoplasms actually detected in routine screening practice should not be biased by missed neoplasms. Obviously, the validity of our calculations depends on the validity of the assumed transition rates from nonadvanced to advanced adenomas, from advanced adenomas to preclinical CRC, and from preclinical to clinically manifest CRC. Sex- and age-specific estimates of these transition rates have previously been derived at very high levels of precision partly from the same database as previously described.11,12 Our sensitivity
Clinical Gastroenterology and Hepatology Vol.
-,
No.
-
analyses indicate that statistical uncertainties regarding the assumed transition rates would not materially alter the overall patterns of sex- and age-specific probabilities of potentially prevented and early detected CRCs and of overdiagnoses. Despite its limitations, our analysis underlines the large potential of screening colonoscopy for prevention and early detection of CRC that can be achieved at low risk of overdiagnosis, even at advanced ages. In addition, more detailed analyses should aim for more precisely defining the best upper age limit for first-time CRC screening in a country like Germany where life expectancy has become very high. Such analyses should also take additional factors into account, such as potential complications of colonoscopy, comorbidity, and costs.24,25 Further work should be expanded to include estimates of potentially prevented CRC deaths, which will require reasonably accurate measures of ageand sex-specific survival according to year after diagnosis for screening detected and other CRCs. Notwithstanding the need for further research regarding these important additional questions, efforts to enhance acceptance of screening colonoscopy should be made with priority, which could further substantially enhance reduction of CRC incidence and subsequently CRC mortality.
References 1. Hewitson P, Glasziou P, Watson E, et al. Cochrane systematic review of colorectal cancer screening using the fecal occult blood test (hemoccult): an update. Am J Gastroenterol 2008; 103:1541–1549. 2. Scholefield JH, Moss SM, Mangham CM, et al. Nottingham trial of faecal occult blood testing for colorectal cancer: a 20-year follow-up. Gut 2012;61:1036–1040. 3. Shaukat A, Mongin SJ, Geisser MS, et al. Long-term mortality after screening for colorectal cancer. N Engl J Med 2013; 369:1106–1114. 4. Brenner H, Stock C, Hoffmeister M. Effect of screening sigmoidoscopy and screening colonoscopy on colorectal cancer incidence and mortality: systematic review and metaanalysis of randomised controlled trials and observational studies. BMJ 2014. 5. Winawer S, Fletcher R, Rex D, et al. Colorectal cancer screening and surveillance: clinical guidelines and rationale—update based on new evidence. Gastroenterology 2003;124:544–560. 6. U.S. Preventive Services Task Force. Screening for colorectal cancer: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med 2008;149:627–637. 7. Levin B, Lieberman DA, McFarland B, 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. CA Cancer J Clin 2008;58:130–160. 8. Schmiegel W, Pox C, Adler G, et al. S3 Guidelines Colorectal Cancer 2004 [in German]. Z Gastroenterol 2004;42:1129–1177. 9. Brenner H, Altenhofen L, Hoffmeister M. Sex, age and birth cohort effects in colorectal neoplasms: a cohort analysis. Ann Intern Med 2010;102:1411–1414.
-
2014
10 Years of Screening Colonoscopy in Germany
7
10. Pox C, Altenhofen L, Brenner H, et al. Efficacy of a nationwide screening colonoscopy program for colorectal cancer. Gastroenterology 2012;142:1460–1467.
20. Association of Population-based Cancer Registries in Germany (GEKID). GEKID data tables. Available: http://www.gekid.de/. Accessed April 23, 2013.
11. Brenner H, Altenhofen L, Stock C, et al. Natural history of colorectal adenomas: birth cohort analysis among 3.6 million participants of screening colonoscopy. Cancer Epidemiol Biomarkers Prev 2013;22:1043–1051.
21. Adler A, Wegscheider K, Lieberman D, et al. Factors determining the quality of screening colonoscopy: a prospective study on adenoma detection rates, from 12,134 examinations (Berlin colonoscopy project 3, BECOP-3). Gut 2013;62:236–241.
12. Brenner H, Altenhofen L, Katalinic A, et al. Sojourn time of preclinical colorectal cancer by sex and age: estimates from the German national screening colonoscopy database. Am J Epidemiol 2011;174:1140–1146. 13. Statistisches Bundesamt. Bevölkerung und Erwerbstätigkeit: Sterbetafel Deutschland 2008/2010. Wiesbaden: Statistisches Bundesamt, 2011.
22. Adler A, Lieberman D, Aminalai A, et al. Data quality of the German screening colonoscopy registry. Endoscopy 2013; 45:813–818.
14. Robert Koch Institut, Gesellschaft der epidemiologischen Krebsregister in Deutschland (eds). Krebs in Deutschland 2007/2008. 8th ed. Berlin: Robert Koch Institut, 2012. 15. Stock C, Ihle P, Schubert I, et al. Colonoscopy and fecal occult blood test use in Germany: results from a large insurance-based cohort. Endoscopy 2011;43:771–779. 16. Independent UK Panel on Breast Cancer Screening. The benefits and harms of breast cancer screening: an independent review. Lancet 2012;380:1778–1786. 17. Gulati R, Gore JL, Etzioni R. Comparative effectiveness of alternative prostate-specific antigen-based prostate cancer screening strategies: model estimates of potential benefits and harms. Ann Intern Med 2013;158:145–153. 18. Stock C, Pulte D, Haug U, et al. Subsite-specific colorectal cancer risk in the colorectal endoscopy era. Gastrointest Endosc 2012;75:621–630. 19. Stock C, Haug U, Brenner H. Population-based prevalence estimates of history of colonoscopy or sigmoidoscopy: review and analysis of recent trends. Gastrointest Endosc 2010;71:366–381.
23. Van Rijn JC, Reitsma JB, Stoker J, et al. Polyp miss rate determined by tandem colonoscopy: a systematic review. Am J Gastroenterol 2006;101:343–350. 24. Stock C, Ihle P, Sieg A, et al. Adverse events requiring hospitalization within 30 days after outpatient screening and nonscreening colonoscopy. Gastrointest Endosc 2013;77:419–429. 25. Sharaf RN, Ladabaum U. Comparative effectiveness and costeffectiveness of screening colonoscopy vs sigmoidoscopy and alternative strategies. Am J Gastroenterol 2013;108:120–132.
Reprint requests Address correspondence to: Hermann Brenner, MD, MPH, Division of Clinical Epidemiology and Aging Research, German Cancer Research Center, Im Neuenheimer Feld 581, D-69120 Heidelberg, Germany. e-mail: h.brenner@ dkfz.de; fax: þ49-6221-421302. Conflicts of interest The authors disclose no conflicts. Funding The German screening colonoscopy registry is funded by the National Association of Statutory Health Insurance Physicians and the National Association of Statutory Health Insurance Funds. This study was further supported in part by grants from the German Research Council (Deutsche Forschungsgemeinschaft, grant number BR 1704/6-4) and the German Federal Ministry of Education and Research (grant number 01KH0404, 01ER0814).
