A patient‐derived organ‐on‐a‐chip platform for predicting chemotherapy response in locally advanced colorectal cancer

Abstract

Colorectal cancer (CRC) remains a leading cause of mortality, underscoring the urgent need for predictive precision medicine tools. While patient‐derived organoids have advanced drug screening, they often lack the dynamic physical microenvironment found in vivo. Organ‐on‐a‐chip (OOC) systems offer superior physiological relevance; however, their translation from bench to bedside has been stalled by a lack of validation against clinical patient outcomes. Here, we bridge this critical translational gap by demonstrating the feasibility and predictive utility of a scalable, patient‐derived OOC (PDOC) platform. From a cohort of 14 primary locally advanced CRC cases, we engineered 8 patient‐specific PDOCs that maintained high‐fidelity histological recapitulation of the parental tumor tissue. Leveraging an automated, high‐content imaging workflow alongside metabolic assays, we profiled responses to standard‐of‐care chemotherapies (irinotecan, FOLFOX, and FOLFIRI). The platform successfully stratified patients into distinct phenotypic response groups, revealing specific sensitivities to FOLFOX versus FOLFIRI. Crucially, we present preliminary clinical evidence supporting these ex vivo predictions: in a 7‐month follow‐up, all patients identified as FOLFOX‐sensitive by our platform remained disease‐free post‐surgery. These findings support the PDOC platform not just as a complex in vitro model, but as a viable, predictive technology for functional precision oncology. This work represents an important step toward integrating microphysiological systems into prospective clinical decision‐making for CRC and other solid tumors.