Objectives: Quantitative and qualitative immunochemical fecal occult blood tests (FOBTs) have been proposed for noninvasive colorectal cancer screening, but comparative evaluation is lacking. The aim of this study was to determine the diagnostic accuracy of two (quantitative) enzyme-linked immunosorbent assay (ELISA)-based immunochemical FOBTs for identifying colorectal adenomas in the target population of screening and to compare the results with six (qualitative) immunochromatographic FOBTs, previously evaluated in the same study participants using the same stool samples.
Methods: A total of 1,319 participants of screening colonoscopy at average risk for colorectal neoplasia (mean age 63 years; age range 31-86 years; 50% men) were recruited prospectively from January 2006 to December 2007 in collaboration with 20 gastroenterological practices in Germany. Fecal hemoglobin and hemoglobin-haptoglobin levels were measured using an automated ELISA (RIDASCREEN). Test performance characteristics at different cutoff values were derived by comparing the results of stool testing with the results of colonoscopy in a blinded manner.
Results: A total of 130 participants (10%) had an advanced adenoma. The area under the receiver-operating characteristic curve with regard to advanced adenomas was 0.68 (0.65-0.71) for hemoglobin and 0.64 (0.61-0.67) for hemoglobin-haptoglobin (P=0.034). At a specificity of approximately 95%, the sensitivity (95% confidence interval) for advanced adenomas was 33% (25-42%) for hemoglobin and 24% (17-32%) for hemoglobin-haptoglobin, respectively. The sensitivity for hemoglobin was very close to sensitivities of the six qualitative FOBTs at (strongly divergent) levels of specificity observed for the latter.
Conclusions: ELISA-based measurement of hemoglobin was superior to hemoglobin-haptoglobin, but showed a similar sensitivity for advanced adenomas compared with (qualitative) immunochromatographic FOBTs at defined levels of specificity. Compared with the latter, its quantitative nature offers advantages in terms of transparency and flexibility regarding the positivity threshold (e.g., specificity can be oriented toward available colonoscopy resources or personal risk profiles) and in terms of a higher level of standardization regarding test analysis and interpretation.