4 Mar, 16 | by EBM
By Dr. Geoffrey Modest
The Canadian Task Force on Preventive Health just published a provocative guideline on colorectal cancer (CRC) screening in those not at high risk (see DOI:10.1503 /cmaj.151125).
- CRC is the 2nd most common cause of cancer-related death, with lifetime probability of 3.5% in men and 3.1% in women
- The incidence and mortality in both men and women increases dramatically in the 70-79 year age groups and continues to increase in the >80 year olds
- Adults aged 50-74:
- Screen those aged 50-59 using gFOBT or FIT (guaiac fecal occult blood test, or fecal immunochemical testing) every 2 years or flexible sigmoidoscopy every 10 years — weak recommendation: moderate-quality evidence
- Screen those 60-74 as with 50-59, but strong recommendation: moderate-quality evidence
- Do not screen adults >75yo– weak recommendation: low-quality evidence
- Rationale for recommendations:
- Meta-analysis of screening studies:
- Guaiacs (in those 45-80yo): 18% reduced CRC mortality [RR = 0.82 (0.73-0.92), with absolute reduction AR = 2.7 per 1000 screened, and number-needed-to-screen NNS= 377 over 18.2 years], as well as incidence of late stage CRC [RR = 0.92 (0.85-0.99)]
- Sigmoidoscopy (in those aged 55-74): 26% reduced CRC mortality [RR = 0.74 (0.67-0.82), with AR = 1.2 per 1000 screened, and NNS = 850 but over 11.3 years], as well as incidence of late stage CRC [RR = 0.73 (0.66-0.82)]
- No diff in all-cause mortality
- In terms of age: studies found no reduction in mortality in those <60yo, but significant reduction in those 60-69 for guaiacs; mortality reduction in those 65-74 with sigmoidoscopy. All studies underpowered in the <60 and >70 yo Which is why there was a weaker recommendation for those 50-60.
- A systematic review compared guaiacs with FIT and found: FIT had greater sensitivity and higher rates of detection for CRC and advanced adenomas, and also had greater participation rates than guaiac cards. The actual positive predictive value of FIT depends on the cut-off used for positivity. (The current USPSTF draft update also prefers FIT testing over guaiacs, and is proposing flex sig screening every 10 years with annual FIT but not guaiac testing — see http://www.uspreventiveservicestaskforce.org/Page/Document/draft-recommendation-statement38/colorectal-cancer-screening2 )
- The Canadian Task Force recommends FIT/guaiac screening every 2 years to decrease the burden to patients, in light of lacking data that yearly is better (though annual was better in a study of those >70yo)
- Flex sig screening was increased to 10 yrs based on finding that flex sig screening decreased CRC mortality and incidence until at least 11 years of follow-up.
- Harms of flex sig were rare (0.001% with perforations, 0.05% minor bleeding, 0.009% for major bleeding and 0.015% for death)
- Harms of screening colonoscopy were much higher (0.05% with perforations, 0.08% minor bleeding, 0.1% for major bleeding and 0.002% for death)
- Adults >=75 yo:
- Not screen (weak recommendation; low quality evidence, mostly perhaps because the studies were underpowered). Though they do comment that those >74 “who do not have illnesses that affect their quality of life and/or their life-span” might consider CRC screening and should discuss with clinician
- Do not use colonoscopy for screening (weak recommendation, low quality of evidence)
- No data looking at benefit of colonoscopy over FIT or other screening tests
- More intensive (need more highly trained personnel), more costly. And they therefore hold colonoscopy to a higher standard in terms of lack of direct evidence of superiority of colonoscopy screening
- Do not use other screening tests (CT colonography, barium enema, digital rectal exam, serologic tests, fecal DNA), since no RCT evidence for them. A recent study on multitarget stool DNA testing was very sensitive (better than FIT), but lots of false-positives
- Meta-analysis of screening studies:
These recommendations differ significantly from those of the US Preventive Services Task Force of grade A recommendation to screen 50-75 yo, grade C for 76-80 yo; with guaiac or FIT every year, flex sig every 10 yrs with FIT every year, or colonoscopy every 10 years. — Though as noted above there will be new recommendations in 2016.
So, a couple of points:
- They do not recommend screening for CRC in those >75. In a reasonably healthy 75-80 yo, both the incidence and mortality from CRC continues to increase dramatically, life expectancy overall as per prior blogs is increasing (though much more so in the US for those of higher income, and much less of an income-related difference in Canada — see http://blogs.bmj.com/ebm/2016/02/24/primary-care-corner-with-geoffrey-modest-md-increasing-disparities-in-life-expectancy/ ). And by minimizing colonoscopy as a screening test, its attendant harms in those over 75 are mitigated (g., harder to get good prep, less adequate exam, and increased risk of severe complications such as perforation — see http://blogs.bmj.com/ebm/2015/04/10/primary-care-corner-with-geoffrey-modest-md-risks-of-colonoscopy-in-older-people/ ). So, my bar for continued screening with non-colonoscopy methods is much lower than with colonoscopy and probably should be considered more often in healthy older people.
- I think these guidelines could be a game-changer. The new draft USPSTF still lists colonoscopy screening as recommended, though also on the list are: guaiac/FIT annually, flex sig every 10 years with FIT annually, and colonoscopy every 10 years (still much more intense than the Canadian ones, with the questionable options of plain guaiacs every year, but sigmoidoscopy every 10 years with FIT every year!!??!!). But I think there are pretty compelling arguments against colonoscopy: noneed of a really terrible bowel prep, much decreased morbidity/mortality from this more aggressive screen, less need for conscious sedation, less need for an accompanier to drive the patient home after the test, much less need for highly trained personnel (there are many sites where nurse practitioners do flex sigs outside-of-the-hospital), and much decreased cost of the other tests. One argument is that with sigmoidoscopy, one sees only ½ the colon and therefore are missing ½ of the potential cancers. But a few studies have found that colonoscopy does not affect the mortality of right sided lesions, only the left sided ones (g., see Ann Intern Med. 2009;150(1):1, a case-controlled study finding no decreased mortality by colonoscopy for right-sided lesions). Reasons?? One is that there is more microsatellite instability in right sided lesions, and these have a better prognosis. But all of this together adds to the appeal of flexible sigmoidoscopy over colonoscopy.