Optimizing Clinical Performance of Fecal Immunochemical Tests

Kaiser Foundation Research Institute, Oakland CA

Investigators

Linked publications & trials

Abstract

Project Summary The primary goal of Project 1 Is to characterize remediable failures of fecal immunochemical tests (FIT) to detect colorectal neoplasms, with the aim of optimizing FIT's effectiveness for community-based screening. FIT is recommended by all current national colorectal cancer (CRC) screening recommendations for screening persons at average risk for CRC. Potential advantages of FIT include that it is non-invasive, has greater sensitivity than older fecal blood tests, does not require a bowel preparation, does not require restrictions of diet or medications, has no direct associated risks, and can be used in mailed outreach programs; however, FIT may sometimes result in false negative and false positive test results. Kaiser Permanente in Northern and Southern California (KPNC/KPSC) has utilized FIT extensively to reach unscreened members; this multi-modality strategy increased screening rates among the KPNC/KPSC screening-eligible populations of >2 million members to 75%, In the top quartile of all US health plans. Despite widespread use, FIT's performance characteristics for population screening are not well characterized. A single FIT evaluation has a CRC sensitivity of 65%-80%; the effects of repeated screening have been modeled, but not reported from real life populations. In addition, FIT performance characteristics may vary among patients, by use of anticoagulants and non-steroidal anti-inflammatory medications (which may increase tumor blood loss) and by temperature and weather conditions. In addition, DNA methylation increases as people age and this may accelerate the transformation of adenomas to carcinoma, decreasing the likelihood of detecting pre-invasive CRCs through screening. Finally, even with 100% adherence, FIT will fail to detect some cancers, although reasons for these failures have not been thoroughly explored. The current study proposes to use the large, diverse patient populations and electronic record systems of KPNC/KPSC to characterize patterns of FIT use and predictors of FIT failures, including: failures to screen, failures to detect (false negative) and failures to follow-up a positive FIT. Candidate predictors include age, sex, race/ethnicity, socioeconomic status and medical history, as well as environmental factors. A candidate next generation stool DNA test will Include a set of biomarkers (e.g. methylated vimentin, other methylation markers and k-ras) which are promising options for stool-based CRC screening. We will therefore measure these biomarkers among FIT negative vs. FIT positive cancers, to assess the potential magnitude by which a methylation marker panel might increase the detection of CRC beyond that detected by FIT alone.