A new platform designed at the UCLA Health Jonsson Cancer Center combines 3D bioprinting, high-speed imaging, and artificial intelligence to monitor how cancer responds to medical treatment. This technology helps scientists identify effective therapies more rapidly by testing drugs directly on lab-grown replicas of a patient’s tumor cells.
The research, published in the journal Nature Protocols, details how scientists create tumor organoids—miniature, lab-grown versions of actual cancers—to observe their reactions to various drugs. Traditional laboratory models often lack the necessary speed or scale for large studies, but this system generates and analyzes high volumes of patient-derived samples simultaneously. The platform tracks how these clusters grow or shrink, allowing for a precise evaluation of drug efficacy.
The workflow utilizes extrusion bioprinting to embed these organoids in specific matrix structures. Rather than relying on traditional dyes, which can disrupt cell behavior, the system uses label-free quantitative phase imaging. This method captures biomass changes continuously, providing a clear view of how tumors change over time without damaging the specimens being studied.
High-throughput, non-destructive monitoring mirrors the evolution of genomic sequencing, which moved from slow, manual processes to the rapid, automated systems that currently underpin modern precision medicine. Just as high-speed sequencing allowed researchers to move past studying single genes toward understanding entire pathways, this platform enables a granular look at how individual tumor clusters behave. Such systems are often required to overcome the limitations of older, static models that frequently fail to capture the biological reality of a growing tumor.
Researchers are also exploring AI drug discovery to further accelerate these efforts.
Related: The Best Products for Diamond Ring Cleaning and Maintenance
Once the imaging process is complete, deep learning models and machine learning algorithms take over to process the massive datasets. The software segments images and tracks individual responses across thousands of samples, identifying subtle differences in how specific organoids react to various substances. This level of detail helps distinguish between sensitive tumor cells and resistant populations that might otherwise be overlooked in a bulk analysis.
The technical specifications of the platform include:
- Bioprinting for consistent organoid formation.
- Continuous, label-free quantitative phase imaging for biomass tracking.
- Deep learning-based segmentation for automated data processing.
- Single-organoid resolution to identify heterogeneity within samples.
The system is complex to calibrate, but it offers a way to see how tumors react to drugs in real time. Dr. Michael Teitell, the director of the center, explained that this allows for a move away from average results. He noted that the team can now determine which specific organoids respond to treatment and which do not, helping to reveal the reasons for unique response profiles.
The team successfully demonstrated the platform’s utility by measuring responses in both established cancer cell lines and a patient-derived tumor sample. This provides a framework for future applications where doctors might test drugs on a patient’s own cells before starting a regimen. By identifying which therapies show promise for a particular tumor, the method could eventually inform clinical decisions for patients facing difficult-to-treat cancers.
