Can we predict screening colonoscopy results based on faecal immunochemical test value?

Aims To evaluate the correlation between faecal immunochemical test (FIT) value and screening colonoscopy results in the Hungarian population-based screening program that was introduced in 2019, and to optimize the FIT cut-off value, if possible.

Methods In this retrospective cohort study, individuals who underwent screening colonoscopy (based on the current cut-off value of 17.00 µg/g) at two endoscopic centers (University of Szeged and Bács-Kiskun County Teaching Hospital) between 2019 and 2022 were included. FIT values (SENTiFIT measuring range: 4.30 µg/g – 170.00 µg/g) and screening colonoscopy results (negative, non-neoplastic lesion, premalignant lesion, including their number, size, location, and histologic results, malignant lesion) were obtained from the Hungarian National Public Health Center registry, supplemented by additional data from local hospital information systems. The correlation of FIT values (maximum and mean FIT value) and screening colonoscopy results were determined by ROC analysis.

Results During the study period, data from 771 screened individuals were obtained. The average time between the faecal test and endoscopy was 1.91±1.53 months. In 137 cases (17.8%) the colonoscopy was negative; 190 patients (24.6%) had only non-neoplastic lesions; in 54.1% of the cases (N=421) there were premalignant lesions, and from these patients 174 had advanced adenoma (22.6%), furthermore in 23 individuals (3.0%) colorectal cancer was confirmed. The mean value of maximum FIT and mean FIT were 79.85±56.63 µg/g and 54.23±44.08 µg/g. During the ROC analysis we found that the area under the curve (AUC) was rather low when evaluating the correlation between neoplastic lesions (premalignant and malignant lesions), maximum (AUC=0.567), and mean FIT value (AUC=0.580). Examining the correlation between advanced neoplasias (advanced adenoma and carcinoma) and maximum FIT (AUC=0.600) we found that changing the currently used cut-off value to 20.00 µg/g, 30.00 µg/g or 40.00 µg/g would result in a sensitivity and specificity of 92.0% and 9.0%, 78.1% and 30.0%, 70.2% and 40.5% compared to the currently used one, and would decrease the number of screening colonoscopies by 7.3%, 27.9%, and 38.0%, respectively. On the other hand, optimal cut-off points for differentiating individuals with and without carcinoma could be determined as 129.70 µg/g maximum FIT value (AUC=0.766), and 91.09 µg/g mean FIT value (AUC=0.812).

Conclusions Although FIT can be an effective way of improving adherence to screening colonoscopy, it alone cannot be used to differentiate individuals with premalignant lesions from the ’healthy’ population. Increasing the cut-off point could increase specificity, and therefore decrease the number of excess colonoscopies, but it would simultaneously result in the loss of sensitivity. Nevertheless, determining a secondary, higher cut-off value might be useful to classify individuals with higher risk for malignancy. These high-risk patients could be referred semi-urgently to screening colonoscopy, decreasing the chance of disease progression.

 

Scroll to top

Conflicts of Interest

Authors do not have any conflict of interest to disclose.

Publication History

Article published online:
27 March 2025