resistant hypertension: are diuretics harmful?

 stop colonoscopy age 75??



Modest, Geoffrey (HMFP - Medicine)

​Modest, Geoffrey (HMFP - Medicine)​

 

A recent article assessed people who had had adenomas on prior colonoscopy but then had their follow-up colonoscopy after age 75, finding that there was very small benefit for doing a follow-up colonoscopy: see colon ca stop at age 75 JAMA2026 or doi:10.1001/jama.2026.3414. Though there is a reasonable argument to the contrary!!!

Details:

-- 91,952 individuals underwent colonoscopy between January 1,2006, and December 31,2019 on or after age 65, with 25,538 individuals (28%) having had an adenoma resected, compared with 66,414 individuals (72%) who were without adenomas

-- this study was done to estimate retrospectively the cumulative colorectal cancer (CRC) risks, non-CRC mortality, and all-cause mortality comparing these two groups

-- the researchers accessed the VA oncology domain cancer registry data, the CRC deaths form the National Death Index-linked primary cause of death data, and the non-CRC death and all-cause mortality from the National Death Index data

    -- the point of the study was to see if those with known earlier adenomas were in fact at higher risk of a bad event subsequently after age 75 than those without adenomas in their pre-75yo colonoscopies

-- 98% male, 78% non-Hispanic white/12% non-Hispanic Black/ 7% Hispanic

-- BMI <18.5 in 1%/18.5-25 in 19%/25-30 in 41%/>30 in 40%

-- VA Frailty index: non-frail 50%/prefrail 24%/mild frailty 14%/moderate frailty 7%/severe frailty 5%

-- Charlson Comorbidity Index (a score of at least 2 reflects significant comorbidities): 0 in 38%/1 in 23%/2 in 15%/at least 3 in 23%

-- aspirin use 50% (aspirin use in several prior studies decreases the development of more severe adenomas and CRC)

-- adenomas were found in 25,538 individuals, 27.8%; adenoma types:

    -- advanced adenomas in 2766 individuals (11%)

        -- an advanced adenoma was defined as 10 mm or larger in size, any size with villous or tubulovillous features, and any size with high-grade dysplasia

    --nonadvanced adenomas in 22,772 individuals (89%)

        -- nonadvanced adenoma was defined as any adenoma without the advanced adenoma features

-- age at index colonoscopy 71, period between index colonoscopy and their 75th birthday was 3-4 years

-- length of followup for CRC: 

    -- 77% of those with adenomas, 68% without had <5 years of followup; 20% with adenomas and 27% without had 5-10 years of followup; and 3% (809 individuals) with adenoma and 6% without had at least 10 years of followup

-- for their statistical analysis, they appropriately used the "Gray test" instead of the "p-test" to assess the statistical significance of the different outcomes

    -- the p-test assesses the probability of obtaining a result by random chance, assuming the null hypothesis is true

    -- the Gray test compares the cumulative incidence curves across 2 or more groups when there are competing risks of survival.

        --both have a P<0.05 as being statistically significant

-- main outcomes: estimated cumulative incidence of CRC, CRC death, non-CRC death, and all-cuase mortality

Results:

-- as mentioned above, adenomas were found in 25,538 individuals, 27.8% of the group, with advanced adenomas in 2766 individuals (11%) and nonadvanced adenomas in 22,772 individuals (89%)

-- those with adenomas were overall older, there were fewer females, more Black and Hispanic individuals, and more who were nonfrail

-- overall results: 

-- overall:

    -- 10-year follow-up revealed cumulative CRC incident of 1.1% (0.8%-1.3%) in those with prior adenomas vs 0.75 (0.5%-0.8%) if no adenoma, Gray P<0.001

    -- cumulative incidence of CRC death was 0.5% (0.3%-0.7%) in those with prior adenomas vs 0.4% (0.3%-0.5%) if no adenoma, Gray P=0.005

        -- both of these were highly statistically significant

--it is clear from the above, that for people with prior adenomas, the incidence of non-CRC deaths is much more than CRC deaths at 5 years (22.4% vs 0.2%) and at 10 years (48.4% vs 0.5%)

-- and for people without prior adenomas, the incidence of non-CRC deaths is much more than CRC deaths at 5 years (21.0% vs 0.1%) and at 10 years (47% vs 0.4%)

-- graph of colorectal cancer cumulative incidence below:

    -- the other graphs of the incidence of CRC death had a less impressive difference, and the cumulative all-cause mortality was no different between the groups with and without prior adenomas, as also noted in the table above

--it is clear from the above graph that those with prior adenomas did have a higher subsequence incidence of colorectal cancer, and this differential continued to increase over time of follow-up

-- the cumulative incidence of non-CRC deaths at 10-year follow-up substantially exceeded the incidence of CRC across all levels of frailty (from 34.2% in the nonfrail group to 82.0% in the severe frailty group)

-- and there was not much difference in the cumulative risk for incident CRC according to adenoma risk

Commentary:

-- cancer is the second leading cause of death in those >75yo, after heart disease; and the cancer risk increases with age

    -- there are, of course, many other causes of death with increasing age, as chronic noncancer diseases also increase with age

-- clinical guidelines are pretty consistent that in average risk patients with prior normal colonoscopies, there is an overall upper age of 75yo; this is largely because of increased morbidity/mortality in older individuals and small net benefit by colonoscopy. all of the guidelines do include, however, the comment that personal preference, life expectancy, overall health and prior screening history could lead to later colonoscopies:

    -- USPSTF https://gmodestmedblogs.blogspot.com/2021/05/colorectal-cancer-screening-new-recs.html, recommending screening up to 75 years old, with caveat that net benefit in older people is small, but the decision should be individualized

    -- American Cancer Society: https://www.cancer.org/cancer/types/colon-rectal-cancer/detection-diagnosis-staging/acs-recommendations.html

    -- US Multi-society Task Force: https://gastro.org/clinical-guidance/updates-on-age-to-start-and-stop-colorectal-cancer-screening-recommendations-from-the-u-s-multi-society-task-force-on-colorectal-cancer/ which reinforces the 50-75yo limit in those not already screened, but then notes "for individuals 76-85, the decision to start or continue screening should be individualized and based on prior screening history, comorbidity, life expectancy,  CRC risk, and personal preference", and "screening is not recommended after age 85"

-- colonoscopy screening rates in the US are in the 75% range in those 65yo and older 

-- overall adenoma rates on colonoscopies are 38% (similar to the 27.8% in this study)

-- the researchers in the above study conclude:

   -- there was a higher risk of subsequent incident CRC and CRC death in those with adenomas in prior colonoscopy vs those without adenomas

   -- but these numbers were low in both those with and without prior adenomas

   -- and these risks paled in comparison to non-CRC deaths

   -- so, the benefits of screening people after age 75 is quite small and "may be too low to justify surveillance colonoscopy, particularly in light of well-established evidence that risks associated with colonoscopy increase with age and fraility"

        -- and one important characteristic of the current study is that it included frailty, not found in most other studies

-- a 2021 systematic review and meta-analysis of 19 studies of colonoscopy in older adults with a history of polyps (https://pubmed.ncbi.nlm.nih.gov/34406584/) found:

    -- the risk of detecting CRC (11 individuals) was higher than in those >70yo vs 50-70yo with risk ratio of 1.5 (1.1-2.2)

    -- the risk of detecting advanced polyps (8 individuals) was higher than in those >70yo vs 50-70yo with risk ratio of 1.3 (1.2-1.3)

        -- most of these studies had follow-up in less than 5 years, and very few individuals developed CRC

-- one major concern with the above VA study is that they assessed the 10-year follow-up of their outcomes, but:

    -- a 10-year outcome is insufficient in healthy adults aged 75yo who may have many years of life ahead of them:

        -- the eprognosis website (https://eprognosis.ucsf.edu/leeschonberg.php) calculates that a male 75-79yo, if in excellent health, has a 37-44% risk of 10-year mortality and a 42%-52% risk of 14-year mortality (ie, one in two are expected to live >14 years)

        -- the same input for women into this calculator produces a 15-21% 10-year mortality and a 27-36% 14-year mortality  (ie, 2 of 3 are expected to live >14 years)

        -- the Johns Hopkins calculator (https://hub.jhu.edu/2014/07/23/life-expectancy-gains-threatened/) estimated in a "study, supported by the American Insurance Group, found that, on average, a 75-year-old American woman with no chronic conditions will live 17.3 additional years (that's to more than 92 years old). But a 75-year-old woman with five chronic conditions will only live, on average, to the age of 87, and a 75-year-old woman with 10 or more chronic conditions will only live to the age of 80

    -- so, the 10-year horizon, as in this study, misses the mark on the preponderance of healthy individuals who are expected to live significantly longer than that...

-- this all brings up the following concerns with the current study:

    -- they only assessed the 5- and 10-year CRC risks

    -- notably, 77% of those with adenomas and 68% of the patients without had <5 years of follow-up (ie, the significant majority); 20% with adenomas and 27% without had 5-10 years of follow-up; and 3% (809 individuals) with adenoma and 6% without had at least 10 years of follow-up (a significant minority)

        -- so actually not the follow-up they advertised

    -- not surprisingly, the non-CRC death rate far exceeded the CRC death rate:

        -- cardiovascular disease is known to be more common than cancer disease

              -- and limiting mortality to CRC (a smaller subset of those who would ultimately die of cancer and would be considered to have a non-CRC cause of death) would pretty obviously tilt the results to non-CRC mortality

        -- and we have no information about comorbidities, medications taken, nonpharmacologic interventions that would deal with the big causes of non-CRC mortality (cardiovascular, all cancers, other common events)

    -- the graph above confirms a pretty significant increased risk of CRC incidence in those with prior adenomas

        -- and, it seems reasonable to assess CRC incidence and not CRC mortality in this 10-year study since:

            -- the incidence will be more common to find in a short study than deaths, which take longer

            -- there is a high likelihood that someone who is nonfrail and a healthy 75-80yo would have a much higher incidence of CRC incidence if followed longer (and, as noted above, the risk of a cancer diagnosis increases with age)

        -- having a diagnosis of CRC is hugely traumatic, both in terms of the morbidity/mortality of therapy especially in elderly people as well as quality of life and psychological health for the individuals and those their caretakers

            -- and this would likely tip the scale to earlier detection and treatment

Limitations:

-- this study assessed mortality outcomes, though it is really important to assess morbidity, since many of the most relevant outcomes to elderly (and others) are morbidity-related dysfunction leading to medical disabilities and the array of psychosocial issues (loss of self-esteem, loss of independence and dependence on others, inability to enjoy prior interests, depression, fragility,...)

-- as a study of US veterans, this involved a pretty unique group of people that limit the generalizability to other groups: vastly more men, likely different prior exposures that might affect longevity (environmental, psychological)

-- as an observational study, we have little knowledge of the reasons that the clinicians chose to prescribe colonoscopy in those over 75yo, either in those with or without prior adenomas. did these people even reflect the VA population accurately?

--the databases accessed for their study may not have been complete, potentially affecting the quality of their analysis

so, an interesting study assessing the utility of colonoscopy screening in those individuals >75yo, finding no big difference in CRC mortality in those with prior adenomas vs normal colonoscopies

    -- this suggested that the presence of even advanced adenomas did not really affect longterm outcomes at the 5- and 10-year time-periods

    -- their conclusions do run counter to the guideline recommendations, which suggest an individualized approach, taking into account the overall-health/comorbidities/life-expectancy of the individuals

    -- and the longevity calculators noted above do note that a 75yo in good health has a significant life-expectancy, well exceeding the 10 years in this study

    -- and, i think the above critique of this VA study really undercuts the validity of their conclusions

    -- of course, it really does not make sense to do screening of any sort, let alone a somewhat invasive one like colonoscopy, without a reasonable assessment of benefit.

        -- which all means that we really do need long-term intervention studies with a diverse group of individuals with varying medical and psychosocial conditions to confirm or reject extending colonoscopy to later than 75yo in healthy adults

geoff

-----------------------------------

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu

to get access to all of the blogs:  go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

to get access to the dropbox: https://www.dropbox.com/scl/fo/pcf4vymtpo77f49wd2ljr/ANyJqB40CuBd-DMEr67XDoI?rlkey=5dsalmuhf8ekd0svsgvn93xm1&dl=0

or you can just click on the magnifying glass on top right, then type in a name in the search box and get all the blogs with that name in them