Strong subsite-specific variation in detecting advanced adenomas by fecal immunochemical testing for hemoglobin.

Abstract

Fecal immunochemical tests (FITs) for hemoglobin are increasingly recommended and used in colorectal cancer (CRC) screening. We aimed to provide a detailed assessment of the sensitivity of FIT according to type and subsite of neoplasms in a true screening setting. A quantitative FIT (FOB Gold, Sentinel Diagnostics, Milano, Italy) was applied prior to colonoscopy by 3,466 participants of the German screening colonoscopy program. Subsite specific sensitivity for various types of colorectal neoplasms was derived by comparing FIT results with findings at screening colonoscopy. The most advanced finding at colonoscopy was CRC, advanced adenoma, and nonadvanced adenoma in 29, 354 and 686 cases, respectively. Per-adenoma sensitivity for large advanced adenomas (>1 cm) strongly varied by location (p < 0.001): cecum: 0/14 (0%), ascending colon and right flexure: 11/43 (26%), transverse colon and left flexure: 2/14 (14%), descending colon: 7/12 (58%), sigmoid colon: 47/92 (51%), rectum: 14/39 (36%). By contrast, the FIT detected all of 5 proximal CRC and 23 out of 24 (96%) distal CRCs, whereas per-adenoma sensitivity of both proximal (17/259, 7%) and distal nonadvanced adenomas (20/237, 8%) essentially equaled the false positivity rate among those without neoplasms (152/2,397, 6%). In conclusion, we found a very large gradient of subsite specific FIT sensitivity for detecting large advanced adenomas ranging from 0% for advanced adenomas located in the cecum to >50% for those located in the descending or sigmoid colon. By contrast, FIT sensitivity was uniformly excellent for CRC and uniformly poor for nonadvanced adenomas, regardless of their location.