FIT testing, a long-term study

An Italian study assessed the 12 year performance of biannual fecal immunochemical testing (FIT) testing, finding that patients seem to do as well as with primary colonoscopy screening (see 10.1136/gutjnl-2017-314753)

Details:

--123,347 individuals had FIT testing every 2 years for 12 years starting in 2002, where all individuals 50-69 were invited to participate in FIT every 2 years

--by protocol, those with positive FIT had colonoscopy. If false positive FIT (negative colonoscopy), FIT was repeated after 5 years

Results:

--over the study, they found 781 people with colorectal cancers (CRCs), 4713 advanced adenomas, and 2844 nonadvanced adenomas

--6419 had positive FIT in the first screening (5.7%), then positivity remained in the 3.3-3.8% range for the rest of the study

--CRC detection rate (per 1000 subjects screened) was 3.34 in the first round, 1.69 in the second, then <1 for subsequent rounds (p<0.001)

    --CRC detection rate decreased 75%, RR 0.25 (0.20-0.32) between the first and third screenings, then remained stable. statistically significant for both age and sex

--advanced adenoma detection rate (per 1000 subjects screened) was 15.9 in first round, then about 8.5-10 (stable) on subsequent rounds (p<0.001), NNS (number needed to screen to detect one case of advanced neoplasia): 51 in first round, then about 100 for subsequent rounds (p<0.001). about half as many men (vs women) needed to be screened to detect one advanced neoplasia

    --advanced adenomas decreased 49%, RR 0.51 (0.47-0.56), then remained stable. statistically significant for both age and sex

--nonadvance adenoma detection rate (per 1000 subjects screened)  was 7.5 in the first round, then 5.8-6.7 on subsequent rounds (p<0.001)

     --nonadvanced adenoma decreased 24%, RR 0.76 (0.68-0.84),

--cumulative FIT positivity rates over 5 consecutive rounds were (per 1000 subjects screened), by age and gender being:

        ​--14.5 in 50-54 yo women and increasing to 32.8 by age 60-64

        ​--20.8 in 50-54 yo men and increasing to 43.4 by age 60-64

--cumulative CRC rates over 5 consecutive rounds, by age and gender being:

        ​--4.3 in 50-54 yo women and increasing to 8.6 by age 60-64

        ​--7.4 in 50-54 yo men and increasing to 25.9 by age 60-64

--cumulative advanced adenoma rates over 5 consecutive rounds, by age and gender being:

        ​--29.5 in 50-54 yo women and increasing to 81.4 by age 60-64

        ​--62.3 in 50-54 yo men and increasing to 144.1 by age 60-64

--PPV (positive predictive value)  for FIT was 36.9% in the first round, then about 20% subsequent rounds (p<0.001)

--PPV for advanced neoplasia decreased 18% in the second round, RR 0.82 (0.77-0.89), with no further reduction over time, remaining at about 30%

Commentary:

--there have not been any RCTs on important clinical outcomes with FIT, though there has been on guaiac fecal occult blood tests (though FIT has greater sensitivity and specificity than guaiacs, eg see Brenner H. Eur J  Cancer 2013; 49:3049). FIT is less sensitive than colonoscopy for finding advanced adenomas. Given the lack of RCTs, the above large population study provides some insight into the clinical importance of FIT.

--this study is quite impressive that both the detection rate for nonadvanced adenomas, advanced adenomas and for CRC decreased dramatically from their first FIT study (the prevalence study, picking up new and old lesions) to the third one (which reflects the incidence of new pick-ups), and then was steady.

--CRC detection rate continued to drop over several screening rounds, maintaining about 2-fold difference over advanced adenomas

--one concern was that FIT, vs colonoscopy, only picks up more advanced and bleeding lesions, so potentially might not affect long-term outcomes as well as colonoscopy (ie, were there early tumors that were potentially resectable/curable by primary colonoscopy screening that did not bleed on the one-time FIT that then grew to an untreatable cancer, or even one which required chemotherapy etc??? and what about all of those tubular adenomas (found in about ½ of colonoscopies and begetting even more colonoscopies)?? It is reassuring both that the numbers of advanced adenomas and CRCs decreased so dramatically and remained low after the initial screen, and that the positive predictive value of a positive FIT did not increase (which might have been expected if smaller nonbleeding nondetected adenomas progressed to more advanced detectable ones or to CRC)

--as a perspective vs primary colonoscopy screening, the cumulative detection rates for FIT in those 50-64yo were: CRC 0.85/1000 people, and for advanced adenomas 5.9/1000 people; which was strikingly similar to colonoscopy numbers of CRC (for people aged 50-69) in Italians 0.8/1000 people and advanced adenomas 6.0/1000 people, and US numbers of CRC  0.7/1000 people and advanced adenomas/sessile serrated polyps >1cm or 7.6/1000 people (note this also includes the serrated polyps)

--their conclusion, in terms of talking with patients is: “For every 1000 persons in this age bracket screened biennially over 10 years, an estimated four to seven colorectal cancers and 42 to 48 advanced adenomas are diagnosed, 170 to 180 persons have at least one positive test result and 91 to 98 are recalled at least once for a colonoscopy that yields a negative result"​

--and, it is reassuring that these results were long-term: the cumulative FIT screening time-period in the study mirrored the time interval for normal primary colonoscopy screening of every 10 years.

so, though not a perfect study (ie, not an RCT of lots of people from different backgrounds with lots of comorbidities), this study does reinforce the utility of FIT as an adequate substitute for  primary colonoscopy, as suggested by the various society guidelines, and quells some unease (in me) that we were only really finding more advanced lesions that would bleed (and perhaps were letting the smaller ones grow and do bad things).  And FIT is so, so, so much easier for patients (no prep!!), had consistently >90% adherence rates by patients in the above study, and avoids some of the attendant morbidity/mortality/cost of colonoscopy (and the same stool sample can be used to check for H Pylori antigen, which may  be useful in populations at risk .…). but it does raise a few questions:

--should we be doing FIT every 2 years, as is done in Canada, instead of annually as is recommended in the US?  see http://gmodestmedblogs.blogspot.com/2016/02/colorectal-screening-canada.html 

--should we stratify screening by using different interval intensities depending on risk?  it seems that men overall have a much higher rate than women; and older is much higher than younger. so, for older men, should we be screening yearly for a few years, then every 2 years???  and, for younger women, screening every 3 years???  After all, there are risks to screening, and the PPV will change depending on baseline risk, so there would likely be higher probability for low risk patients to get colonoscopies for false positive FIT, along with the colonoscopy risks).

to receive these blogs regularly as emails, let me know at gmodest@uphams.org