In cancer research and pharmacological research, cancer organoids and cancer spheroids play an indispensable role in that they recapitulate tumor heterogeneity and pathophysiological processes in vitro, and can be generated in a time- and cost-effective manner. Imaging and studying cancer organoids and cancer spheroids, therefore, serve as the key to open challenging clinical and biological questions. Thus far, commercially available imaging solutions consist of mostly brightfield and fluorescence microscopy methods. These methods, however, have inherent limitations that prevent them to simultaneously offer non-invasive, high-content, label-free, and longitudinal imaging of 3D cancer models, especially considering the imaging-based monitoring of patient derived organoids.
In a new paper published in Light: Science & Applications, a team of scientists, led by Ap. Professor Mengyang Liu from the Center for Medical Physics and Biomedical Engineering, Medical University of Vienna, Austria, Associate Professor Kristen Meiburger from the Department of Electronics and Telecommunications, Politecnico di Torino, Italy, and co-workers have developed an AI-enhanced OC-PAM system for 3D cancer model imaging. Via three carefully designed experiments, it is demonstrated that the OC-PAM system can perform longitudinal tracking of organoids, evaluate cancer organoids' drug response to chemotherapy, indicate individual organoid's viability, and sort out drug tolerant persister (DTP) cell proxies. What is more, all these features are provided in a non-invasive and label-free manner.
More specifically, for the longitudinal imaging of cancer organoids, the optical coherence microscopy (OCM) mode was engaged to image breast cancer organoids after chemotherapy exposure via carboplatin administration. Automatic tracking of individual organoids and quantitative analysis of organoids' average volumes were performed to evaluate their response to carboplatin. Drug-treated organoids showed reduced growth rates, while a small subset displayed regrowth patterns consistent with drug-tolerant persister cells, highlighting the system's ability to capture rare and clinically relevant behaviors.
Beyond morphology, the study also introduces a radiomics-based analysis of OCM data to assess organoid viability. Using machine learning, the approach achieved high classification performance, demonstrating the potential for non-invasive and non-destructive monitoring of treatment response over time.
Finally, the team showed that OC-PAM can detect rare cells within dense 3D spheroids by imaging melanin-containing melanoma cells mixed with breast cancer cells. Even at very low concentrations, individual rare cells could be visualized, underscoring the sensitivity of the method.
Overall, the results establish OC-PAM as a powerful imaging platform for studying cancer organoids and spheroids. By combining high-resolution imaging with AI-based analysis, the technology opens new possibilities for investigating drug resistance, rare cell populations, and treatment response, with strong potential impact on cancer research, drug development, and precision oncology.
