AI chip may enable same‑day therapy
Researchers unveiled μPharma—an AI‑enabled lab‑on‑a‑chip that rapidly predicts drug response and could enable same‑day treatment choices for pediatric T‑ALL patients. The platform shows how point‑of‑care AI chips might be adapted to cytology and FNA workflows for faster molecular triage. (healthcare.utah.edu)
A peer‑reviewed manuscript describing μPharma appears in Cell Medicine (Cell Press) and is the basis for the Huntsman Cancer Institute announcement. (cell.com) μPharma combines an automated digital microfluidic immunofluorescence assay tailored for suspension cells with machine‑learning models trained on comprehensive single‑cell morphological and protein‑expression features. (sciencedirect.com) The study reports specific pretreatment biomarker–drug links, notably phosphorylated BCL2 (pBCL2) as a predictor of venetoclax sensitivity and phosphorylated LCK (pLCK) as a marker for dasatinib response in T‑ALL samples. (innopsys.com) Authors state μPharma generates pharmacotyping predictions within hours by inferring drug sensitivity from biomarkers rather than exposing cells to drugs, shortening turnaround compared with conventional ex vivo drug assays that can take days to weeks. (digipharmalab.com) Platform testing used pediatric T‑ALL patient samples and single‑cell analyses that revealed intratumor heterogeneity relevant to mixed‑response clones, which the paper quantifies using spatial protein distribution and cell morphology features. (cell.com) The manuscript and institution note next steps include prospective clinical validation, workflow integration for point‑of‑care cytology/FNA specimens, and regulatory evaluation before routine clinical deployment. (healthcare.utah.edu)
