Editorial: Understanding the Trade-Offs When Considering Positivity Threshold of Faecal Immunochemical Tests

Abstract

The faecal immunochemical test (FIT) is the most used colorectal cancer (CRC) screening test worldwide [1]. FIT is a family of tests, the performance of which varies by brand, positivity threshold and testing interval [2, 3]. Despite myriad studies, the optimal FIT threshold and interval remain elusive. Countries and large health care systems that screen primarily with FIT must decide on the threshold for a positive test, knowing that there are trade-offs.

The novel cross-sectional study [4] of 47,265 Norwegians who were first participants in a biennial FIT screening program in which the FIT threshold for positivity (and diagnostic colonoscopy) was 15 μg/g, was designed to quantify colonoscopies performed, yield for CRC, advanced adenomas (AA) and advanced serrated lesions (ASLs), and adverse events at different thresholds. The > 15 μg/g threshold was compared to thresholds of > 20, > 47, > 80, > 120 and > 150 μg/g. Outcomes include positivity rate, detection rates of CRC, AA and ASLs, positive predictive value (PPV) and severe adverse events. All outcomes were reported relative to the > 15 μg/g threshold.

As expected, as the threshold increased from 15 to 150 μg/g, the positivity (i.e. detection) rate and colonoscopy demand decreased, as did the yield for CRCs, AAs and ASLs, while the PPVs increased for CRCs and AAs, but not for ASLs. There was a small CRC stage shift, with detection of a lower proportion of Stage I CRCs and a higher proportion of advanced CRCs. Adverse events fell as the threshold increased, but the proportion of significant bleeding events increased due to a higher proportion of polypectomies for advanced polyps.

These data provide a quantitative framework for understanding the trade-offs among the various thresholds. Countries and large health care systems screening primarily with FIT—especially those with fixed or limited colonoscopy resources—need to know this to optimise those resources. Notably, the detection and PPV of advanced SSLs, both overall and by anatomical site, were unchanged at different thresholds, suggesting that detection may have been serendipitous, supporting the non-bleeding biology of these lesions. Trends by FIT threshold did not differ by age or sex.

Despite their public health utility, these data do not provide estimates of sensitivity and specificity because colonoscopy was not performed for a FIT value of ≤ 15 μg/g. Further, they do not tell us the ‘optimal’ threshold, as this determination requires judgement and because FIT screening is a longitudinal rather than one-time process. Lastly, these results may not generalise to other brands of FIT or to other populations.

Greater attention should be given to studying FIT's quantitative value, which may be useful for prioritising who requires colonoscopy sooner and for tailoring the re-testing interval. The Netherlands is conducting an observational study in which the re-testing interval among FIT-‘negative’ persons depends on the quantitative value. Several studies have used FIT's quantitative value along with other factors to stratify patient risk for advanced neoplasia [5-7]. There is likely untapped potential of FIT to improve the efficiency and cost-effectiveness of CRC screening.

Thomas F. Imperiale: conceptualization, writing – original draft, writing – review and editing, project administration, resources.

This article is linked to Randel et al paper. To view this article, visit https://doi.org/10.1111/apt.18314.