Occult blood detected by fecal immunochemical testing (FIT) could originate from upper gastrointestinal (GI) lesions. We aimed to assess whether individuals with a positive FIT in a colorectal cancer (CRC) screening program – even without colorectal advanced neoplasia (AN) – have increased risk of esophagogastroduodenoscopy-detectable upper GI cancers (EUGIC) within three years after screening.

Subjects aged 50–69 who underwent FIT-based CRC screening from 2005 to 2018 in Romagna were included. FIT positivity was defined as ≥20 µg hemoglobin/g feces. According to the last screening round result, subjects were classified as: FIT-positive with advanced neoplasia (FIT+/AN+; colorectal cancer or advanced adenoma), FIT-positive without AN (FIT+/AN−), or FIT-negative (FIT−). All esophageal, gastric, or duodenal cancers diagnosed within three years of the FIT were identified via linkage with the regional cancer registry. Incidence of EGD-detectable upper GI cancers was compared across FIT+/AN+, FIT+/AN−, and FIT− groups. In a sub-analysis, we evaluated a lower fecal hemoglobin cutoff (≥10 µg/g) for association with upper GI cancer. Multivariable logistic regression (adjusting for age, sex, and screening round) estimated odds ratios (OR) for EUGIC, and ROC analysis assessed the predictive performance of FIT. The number of screenees needed to scope (NNS) to find one EUGIC was computed.

A total of 281,370 individuals were included (FIT-positive: 7.8% [2.4% with AN]; FIT-negative: 92.2%). EUGIC were identified in 0.20% of FIT− participants, compared to 0.34% of FIT+/AN− and 0.38% of FIT+/AN+ participants (p<0.001). Lowering the FIT positivity threshold to ≥10 µg/g increased sensitivity with decrease in specificity. At multivariable analysis, FIT≥10 µg/g (OR 1.67, 95%CI 1.34-2.08), increasing age (OR 1.05, 95%CI 1.04-1.07), male sex (OR 1.74, 95%CI 1.46-2.08) and first-round screening (OR 1.46, 95%CI 1.20-1.77) were associated with EUGIC, with fair predictive performance (ROC 0.64). The predicted three-year EUGIC risk ranged from 0.09% (lowest-risk category) to 0.6% (highest-risk category). Corresponding NNS was 1111 for lowest-risk category and 167 for highest-risk category.

FIT-positive screening participants showed a 1.7-fold higher incidence of esophageal, gastric and duodenal GI cancers than FIT-negatives within three years from last screening round - although the absolute risk was low. The risk stratification based on a simple model including FIT, age, sex and screening round may represent a first step toward identifying subjects in whom targeted upper GI endoscopy could be clinically justified.