Risk for advanced neoplasia (AN) or colorectal cancer (CRC) could be identified by age, sex, and fecal hemoglobin (F-Hb) concentration, according to study results presented at Digestive Disease Week (DDW), held from May 21 to 24, 2022 in San Diego, California, and virtually.
CRC risk remains difficult to predict and available prediction tools require extensive information but provide limited accuracy, researchers stated. Recent reports have indicated that F-Hb concentrations were associated with future AN and CRC, however, the utility of using F-Hb concentration as a predictive biomarker for disease risk has not been validated.
To evaluate the potential clinical utility of F-Hb concentration for predicting risk for CRC, data were sourced from the Dutch biennial fecal immunochemical test (FIT)-based screening program. In total, 265,881 individuals who had completed 3 successive FIT screening rounds from 2014 to 2018 with negative results in the first 2 assessments were included. Risk for positivity during round 3 was evaluated using demographic and clinical factors. The generated risk score was validated using an independent cohort of 11,903 participants from a screening pilot study in the Netherlands.
The study population comprised 47.8% men, mean age, 69.0±1.9 years; 22.2% had F-Hb concentration greater than0 in round 1, and 8.9% had F-Hb greater than0 in round 2.
In round 3, 3.3% of the participants were FIT-positive, 1.2% went on to develop AN, and 0.2% had CRC.
Predictors for AN included male sex (adjusted odds ratio [aOR], 1.3; 95% CI, 1.3-1.4; P <.001). A positive relationship was observed between risk and F-Hb concentration in the first round from 0.1-9.9 mg Hg/g (aOR, 2.8; 95% CI, 2.6-3.1; P <.001) to 40.0-46.9 mg Hg/g (aOR, 9.4; 95% CI, 7.5-11.7; P <.001) and in the second round from 0.1-9.9 mg Hg/g (aOR, 4.8; 95% CI, 4.3-5.3; P <.001) to 40.0-46.9 mg Hg/g (aOR, 8.6; 95% CI, 7.0-10.5; P <.001).
For CRC, male sex (aOR, 1.2; 95% CI, 1-1.4; P =.05) and age (aOR, 1.8; 95% CI, 1.1-2.8; P =.01) were significant predictors. Also serving as a predictor were increasing F-Hb concentrations in the first round, from 0.1-9.9 mg Hg/g (aOR, 2.5; 95% CI, 2.0-3.2; P <.001) to 40.0-46.9 mg Hg/g (aOR, 6.3; 95% CI, 3.5-11.1; P <.001) and in the second round from 0.1-9.9 mg Hg/g (aOR, 3.0; 95% CI, 2.2-4.0; P <.001) to 40.0-46.9 mg Hg/g (aOR, 4.9; 95% CI, 2.8-8.4; P <.001).
Using these variables, the predicted risk score had a C-statistic of 0.78 for AN and 0.73 for CRC. Stratified by the risk score percentile, relative risk for AN ranged from 0.5 to 13.3-fold higher than average risk and “less predictive” for CRC at 0.2 to 9.4-fold higher than average risk, the investigators noted.
The risk score had a high predictive power for AN risk among the external validation cohort (C-statistic, 0.67; 95% CI, 0.61-0.73).
These findings may not be generalizable to a non-European population, the study authors noted.
This study data indicated to the researchers that age, gender, and F-Hb concentrations could be used to stratify patients as having low- or high-risk for AN and CRC, indicating that F-Hb may have clinical value.
Meester RGS, van de Schootbrugge-Vandermeer HJ, Toes-Zoutendijk E, et al. Fecal occult blood loss accurately predicts future detection of colorectal cancer–a prognostic model. Presented at: DDW 2022; May 21-24, 2022; San Diego, CA. Abstract/Poster Mo1105.