A Quarter of Participants With Advanced Neoplasia Have Discordant Results From 2-Sample Fecal Immunochemical Tests for Colorectal Cancer Screening.

ABSTRACT

BACKGROUND & AIMS: Some colorectal cancer (CRC) screening programs use 2-sample fecal immunochemical tests (FITs). We aimed to assess advanced neoplasia (AN) yield of 2 different FIT assays performed on the same bowel movement and have discordant results. METHODS: We conducted a large prospective comparative accuracy study within the Dutch national CRC screening program to evaluate 2 quantitative FIT assays (FOB-Gold, Sentinel, Italy and OC-Sensor, Eiken, Japan) with comparable performance characteristics. We asked 42,179 screening-naïve individuals, 55-75 years old, to perform both tests on the same bowel movement, from May 2016 through March 2017. Participants with ≥1 positive test result (≥15 µg hemoglobin/gram feces) were invited for colonoscopy examination. Detection of AN by colonoscopy was the reference standard. RESULTS: A total of 21,078 participants (50% participation rate) were included. FIT results were both negative for 19,032 participants (90%), both positive for 1163 participants (5.5%), and discordant for 883 participants (4.2%). AN was detected in 500 participants with 2 positive FIT results (43%) compared to 187 with discordant FIT results (21%) (p < .001). Of the 687 participants found to have AN by colonoscopy, 187 had only 1 positive FIT result (27%). CONCLUSION: In a large 2-sample FIT-based CRC screening study, more than a quarter of participants in whom AN was detected by colonoscopy in the first screening round had discordant FIT results. AN was detected in one-fifth of those with FIT discordance. Participants with discordant results from 2 FITs should undergo colonoscopy. (www.trialregister.nl; no. NTR5874).