UCLA's AI platform tracks cancer treatment responses in tumor organoids at single-cell level

AI Platform Transforms Cancer Drug Testing

Researchers at UCLA Health Jonsson Comprehensive Cancer Center have built an AI platform that merges 3D bioprinting, advanced imaging, and artificial intelligence to track cancer treatment responses in tumor organoids with unprecedented detail 1

1

. Published in Nature Protocols, this AI-powered platform addresses a critical challenge in cancer research: combining biological accuracy with the speed and scale needed for clinical applications while testing personalized cancer therapy options 2

2

.

The technology creates lab-grown tumor replicas from patient-derived tumor cells and continuously monitors their responses to different drugs. By analyzing thousands of individual organoids simultaneously, the platform helps scientists evaluate hundreds of potential therapies and uncover drug responses that could inform treatment strategies for rare and hard-to-treat cancers 1

1

.

Deep Learning Enables Single-Organoid Resolution

The platform uses extrusion bioprinting to generate three-dimensional tumor organoids embedded in extracellular matrix constructs designed for high-throughput multiwell formats 2

2

. These organoids are monitored using high-speed, label-free quantitative phase imaging that tracks changes in biomass and growth dynamics to measure tumor fitness over time. Unlike traditional methods requiring dyes or destructive assays that alter cell behavior, this approach allows continuous observation without interference 1

1

.

The platform incorporates automated image reconstruction, deep learning-based segmentation, and machine learning-based tracking of individual organoid responses to therapy. This allows researchers to quantify drug responses at single-organoid resolution across thousands of samples, providing detailed views of tumor heterogeneity and how different tumors respond to treatment 2

2

.

Detecting Resistant Tumor Populations

Dr. Michael Teitell, director of the UCLA Health Jonsson Comprehensive Cancer Center and co-senior author, explained the platform's capabilities: "Instead of asking whether a drug works on average for a large number of tumor cells, we can now determine which specific organoids respond and which do not, and, ultimately, have an approach to determine the underlying reasons for unique response profiles" 1

1

. The system measures drug responses across thousands of individual organoids, detects rare resistant tumor populations, tracks growth and cancer treatment responses over time, and better predicts which therapies may work for particular patients 2

2

.

The platform successfully measured how tumor organoids responded to drug treatment over time in both established cancer cell lines and patient-derived tumor samples. This capability to identify resistant populations early could help clinicians avoid ineffective treatments and select alternatives more quickly 1

1

.

Implications for Personalized Cancer Therapy

The technology points toward a future where doctors could test cancer drugs on a patient's own tumor cells before treatment begins. By helping researchers identify which therapies are most likely to work for a particular tumor, the method could support more personalized treatment decisions, particularly for patients with rare and hard-to-treat cancers 2

2

. This matters because tumor organoids more closely resemble patient tumors than traditional laboratory models, yet many current systems struggle to combine biological accuracy with the consistency and scale needed for clinical use 1

1

.

The study was co-authored by Dr. Teitell and Alice Soragni of the University of Colorado School of Medicine, with first author Bowen Wang, a postdoctoral fellow in the Teitell Laboratory. The work received funding from the Air Force Office of Scientific Research, the Department of Defense, the National Science Foundation, and the National Institutes of Health 2

2

.