Study data support the use of quantitative fecal immunochemical tests (FITs) for the detection of colorectal cancer (CRC) and advanced adenomas, according to results published in Gastroenterology. FIT sensitivity for advanced neoplasia increased with the number of fecal samples per patient. Lower FIT thresholds for CRC positivity were also associated with greater CRC sensitivity and advanced adenoma.
Patients included in the study were enrolled in an Australian colonoscopy and interval FIT surveillance program between 2008 and 2019. Researchers restricted enrollment to patients with personal or family history of neoplasia that justified intermittent screening colonoscopies. Enrollees were instructed to complete 2-sample FITs provided by mail between scheduled colonoscopies. Patients were also instructed to provide 2 separate fecal samples to be collected at least 30 minutes apart. The analysis included patients who had undergone colonoscopy and who had returned a 1-sample or 2-sample FITs in the preceding year. Samples in the present study were assessed for hemoglobin (Hb) concentration, and results were expressed as μg of Hb/g of feces. Case phenotypes for each colonoscopy were defined as the most advanced pathology, with the hierarchy being CRC, advanced adenoma, non-advanced adenoma, and cases without neoplasia. FIT positivity for the cohort was assessed across a range of cut-off values (20 μg Hg/g to 150 μg Hg/g). For each cut-off value, the number of FIT-positive cases was divided by the number of cases detected with colonoscopy. This calculation gave the “effort” of study procedures, or the number of colonoscoped patients needed to detect a given lesion. The sensitivity of FIT for CRC and advanced adenomas was also calculated for each cut-off value.
The researchers observed the greatest sensitivity for advanced neoplasia in the smallest cut-off value (10 μg Hb/g). At this threshold, sensitivity for advanced neoplasia was 44.1% (95% CI, 40.2-48.1%) and 59.7% (95% CI, 55.8-63.6%) for 1-sample and 2-sample FIT kits, respectively. However, this lower cut-off value necessitated greater effort to detect neoplasia. The number needed to colonoscope for 2-sample kits was 9.29 (95% CI, 9.52-11.58), compared with 7.33 (95% CI, 6.12, 8.90) when the cut-off was set at 80 μg Hb/g. The most optimal balance between sensitivity and effort was found at a threshold value of 25 μg Hb/g, at which sensitivity and effort for 2-FIT were 46.6% (95% CI, 42.7-50.6%) and 6.59 (95% CI, 5.93-7.36), respectively. For 1-sample FIT kits, the sensitivity and effort at 25μg Hb/g were 29.6% (95% CI, 26.0-33.0%) and 6.51 (95% CI, 5.71-7.49), respectively.
Based on these results, a cut-off value of 25 μg Hb/g may provide best results for fecal screening procedures, with an appropriate balance between sensitivity for advanced neoplasia and number needed to colonoscope. Greater number of samples per FIT kit was associated with better sensitivity. Lower cut-off values for FIT positivity produced higher sensitivity but greater effort. These data may inform future programs which incorporate FIT kits into neoplasia screening.
Young GP, Woodman RJ, Ang FLI, Symonds EL. Both sample number and test positivity threshold determine colonoscopy efficiency in detection of colorectal cancer with quantitative fecal immunochemical tests [published online May 23, 2020]. Gastroenterology. doi: 10.1053/j.gastro.2020.05.008