Head-to-head comparison of the test performance of self-administered qualitative vs. laboratory-based quantitative fecal immunochemical tests in detecting colorectal neoplasm
Colorectal cancer (CRC) remains the third most commonly diagnosed cancer worldwide, with population-based screening through fecal occult blood tests (FOBTs) and endoscopy proving effective in reducing mortality. Among FOBTs, fecal immunochemical tests (FITs) have gained prominence due to their ease of use, cost-effectiveness, and superior diagnostic performance compared to traditional guaiac-based tests. FITs are categorized into qualitative (self-administered) and quantitative (laboratory-based) types, differing in result interpretation—dichotomous versus continuous measurements of fecal hemoglobin (Hb) concentration. While both types are widely implemented in CRC screening programs, direct comparisons of their test performance at manufacturer-preset thresholds remain limited. This study conducted a comprehensive head-to-head evaluation of three self-administered qualitative FITs and one laboratory-based quantitative FIT using prospectively collected fecal samples from a large CRC screening trial in China. The study was based on the TARGET-C trial, a multicenter randomized controlled trial comparing colonoscopy, annual FIT, and risk-adapted screening strategies for CRC detection. From 19,582 enrolled participants aged 50–74 years, 3,144 pre-colonoscopy fecal samples were included in the analysis. These samples comprised 24 CRCs, 230 advanced adenomas, 622 non-advanced adenomas, and 2,268 participants without significant colonoscopy findings. Advanced adenomas were defined as adenomas ≥10 mm in size, with villous architecture, high-grade dysplasia, or intramucosal carcinoma. The anatomical distribution of lesions showed that 70.8% of CRCs and 56.1% of advanced adenomas were located in the distal colon or rectum. Three qualitative FITs (Pupu tube, New Horizon Health Technology) with manufacturer-preset thresholds of 8.0, 14.4, and 20.8 μg Hb/g (equivalent to 100, 180, and 260 ng Hb/mL, respectively) were evaluated alongside one quantitative FIT (OC-Sensor, Eiken Chemical) with a recommended threshold of 20 μg Hb/g (100 ng Hb/mL). Stool samples were collected using standardized protocols, stored at −80°C, and tested by trained staff blinded to colonoscopy results. The qualitative FITs provided visual positive/negative results, while the quantitative FIT yielded numerical Hb concentrations. At manufacturer-preset thresholds, the qualitative FITs demonstrated higher positivity rates (12.1% for FIT-1, 6.6% for FIT-2, and 4.5% for FIT-3) compared to the quantitative FIT (2.8%). For CRC detection, sensitivities were comparable across all tests: 83.3% (95% CI: 65.8–94.1%) for FIT-1, 79.2% (61.1–91.4%) for FIT-2, 70.8% (52.1–85.4%) for FIT-3, and 58.3% (39.7–75.4%) for the quantitative FIT. However, for advanced adenoma detection, qualitative FIT-1 and FIT-2 showed significantly higher sensitivities (33.9% [28.7–39.4%] and 22.2% [17.7–27.2%], respectively) than the quantitative FIT (11.7% [8.4–15.8%]). This came at the cost of lower specificities (90.2% for FIT-1 vs. 98.4% for the quantitative FIT). No significant differences were observed between qualitative FIT-3 and the quantitative FIT for advanced adenoma detection (17.0% vs. 11.7%). Subgroup analyses revealed that qualitative FIT-1 outperformed the quantitative FIT in detecting distal (31.1% vs. 13.6%) and proximal (33.3% vs. 6.4%) advanced adenomas. For size-specific advanced adenomas, qualitative FIT-1 showed superior sensitivity for lesions measuring 10–15 mm (23.8% vs. 3.8%), 15–20 mm (39.1% vs. 17.2%), and ≥20 mm (53.3% vs. 22.2%). These findings suggest that lower Hb thresholds in qualitative FITs enhance detection of smaller or less bleeding lesions but increase false-positive rates. To address threshold-related heterogeneity, the study adjusted the quantitative FIT’s cutoff to match the specificities or positivity rates of the qualitative FITs. When thresholds were aligned to achieve comparable specificities (90%, 95%, and 97%), no significant differences in sensitivity for CRC or advanced adenoma were observed between the two test types. For example, at 90% specificity, qualitative FIT-1 and adjusted quantitative FIT showed similar sensitivities for advanced adenoma (33.9% vs. 25.2%). Similarly, when positivity rates were matched (12.1%, 6.6%, and 4.5%), the quantitative FIT’s performance metrics aligned closely with those of the qualitative FITs. ROC analysis further validated these findings. The quantitative FIT’s AUC was 0.900 (95% CI: 0.823–0.977) for CRC detection and 0.662 (0.624–0.699) for advanced adenoma. The qualitative FITs’ sensitivity-specificity pairs fell within the 95% CI of the quantitative FIT’s ROC curve, indicating equivalent discriminatory power when thresholds were appropriately calibrated. Multivariate regression analyses examined potential confounding factors affecting test performance. Sensitivities for advanced neoplasms did not vary significantly by sex, age, BMI, smoking status, alcohol consumption, NSAID/anticoagulant use, or family history of CRC. This consistency supports FIT’s robustness across diverse demographic and lifestyle subgroups. The study highlights critical considerations for CRC screening programs. While qualitative FITs offer simplicity and suitability for home use, their lower specificity at preset thresholds may lead to unnecessary colonoscopies. For instance, qualitative FIT-1’s PPV for advanced adenoma was 20.5% versus 31.0% for the quantitative FIT. Quantitative FITs, though requiring centralized processing, provide flexibility in threshold adjustment to balance sensitivity and specificity based on program goals. Manufacturers of qualitative FITs should consider higher thresholds to improve PPV and reduce colonoscopy burdens. In conclusion, this comprehensive comparison demonstrates that self-administered qualitative and laboratory-based quantitative FITs exhibit comparable diagnostic performance for colorectal neoplasms when thresholds are adjusted to yield similar specificities or positivity rates. The choice between test types should consider operational requirements, desired specificity levels, and manageable positivity rates in screening populations. These findings provide valuable evidence for optimizing FIT-based CRC screening strategies worldwide. doi.org/10.1097/CM9.0000000000001524
Was this helpful?
0 / 0