Optimization of a microfluidic system for automated detection of Cholangiocarcinoma cells in bile

Cholangiocarcinoma (CCA) remains a clinical challenge despite ongoing advancements in its characterization and diagnostics. Early diagnosis of CCA is particularly difficult due to its anatomically deep location; this often results in advanced-stage diagnoses and consequently poor prognoses. Our prior work demonstrated that analysis of bile can provide useful diagnostic information. However, identifying the tumor origin and obtaining bile is technically demanding, requiring specialized equipment, skilled personnel, and cooperative patients. In this work, we therefore sought to improve our prior integrated microfluidic system approach for CCA diagnosis and prognosis. Thirty-three clinical samples were first analyzed by a benchtop assay, and the accuracy, sensitivity, and specificity were 97.0, 96.6, and 100 %, respectively, in 5 h. Our optimized, automatic system was associated with similar values of 93.8, 92.9, and 100 %, respectively, with a subset of sixteen samples, within 180 min. Given its accuracy and sensitivity, this device could detect CCA at earlier stages and could monitor prognosis, thereby improving overall clinical utility.