By: Dr. Geoffrey Modest
4.4 million colonoscopies were performed in Germany since October 2002 as part of a national, government-sponsored initiative (and covered by their Statutory Health Insurance, with a federal registry of the findings), on people aged 55-80. They assessed mortality, the overall age- and sex-specific findings, and projected the numbers of colorectal cancers (CRCs) prevented, those detected earlier than would have been without screening, and the extent of overdiagnosis (see doi.org/10.1016/j.cgh.2014.08.036). Lifetime risk of CRC in Germany is approx 7.5% in men and 6.1% in women.
Results:
–22% of eligible women and 20% of eligible men had a screening colonoscopy in the initial 10 year period
–At least one neoplasm (adenoma or cancer) was found in 28.5% of men and 17.6% of women
–CRCs prevention: 180,000 (1 in 28 screening colonoscopies)
–CRCs detected earlier by screening: 40,000 (1 in 121 screening colonoscopies)
–CRC overdiagnosis: 4500 (1 in 1089 screening colonoscopies)
–97% of all CRCs prevented and 89% of those detected earlier by screening were in people up to 75 yo; 28% of overdiagnoses were in those >75 yo
–Probabilities of CRC prevention are highest in colonoscopies conducted around age 60 (5% among men and 3.5% among women), and lowest at age 80 (2% for men and women)
–Probabilities of early detection increase with age up to about 2% by age 80
–Probabilities of overdetection increase with age: <0.1% at age 65 to <1% by age 80 in both men and women
So, pretty impressive findings. a few caveats:
–We still do not have a real randomized controlled trial showing that colonoscopy is effective in reducing mortality (though RCTs do exist for fecal occult blood testing and sigmoidoscopies), though screening colonoscopies have been accepted as the gold standard in many places (including Boston)
–In terms of overdiagnosis, they state it is defined as “cancers detected as screening colonoscopy that would not have become clinically manifest during lifetime without screening”. unclear exactly where they get their data or its quality
–Transitions from adenoma to cancer were derived by mathematical modeling from previous trials, with presumed annual transition rates approx as follows.
–Nonadvanced adenoma to advanced: men approx 4.0%, women approx 3.7% with nonsignificant change by age
–Advanced adenoma to preclinical cancer: men 2.6% age 55-59, increasing to 5.2% if over 75yo; women 2.5% age 55-59, increasing to 5.6% if over 75yo
–Preclinical cancer to clinical cancer: men approx 19%; women approx 20% with no clear age trend
–Their conclusion is that screening between age 55-70 is the most effective in eliminating CRC risk (I would add that from older data I’ve seen, there is an increased risk of screening people >75 yo, with decreased ability to do good colonoscopy because of higher likelihood of poor prep and increased likelihood of perforation)
–They comment that the risk of overdiagnosis in CRC is much less than that found for screening mammography (19%). For PSA screening, they cite overdiagnosis of prostate cancer to be 2.3-3.3%.
–Bottom line: this study confirms and reinforces others that colorectal cancer screening is one of the most important screening tests we do. I should note that the age range suggested in the US is pretty variable depending on recommending agency. For those at average risk, for example, the USPSTF recommends any of 3 approaches, including colonoscopy every 10 years, from age 50-75. the Am College of Gastroenterology mostly agrees but suggests starting at age 45 in African-Americans, the Am Cancer Society/Multi-Society Task Force suggest starting at age 50 but continuing until the life-expectancy is <10 years. [As a perspective here, an 85 year old woman in excellent health has about a 10-year life expectancy.] Another issue with our not having a uniform and coherent health care system….