A Deep Learning Approach for Tracking Colorectal Cancer-Derived Extracellular Vesicles in Colon and Lung Models

Comprehensive Summary

This study by Chiabotto, et al., uses a deep learning approach to investigate the extracellular vesicle (EV) driven progression of colorectal cancer. EVs in colon and lung models were monitored and their migration was quantified. EV morphology and cytoxicity were also analyzed. A deep learning algorithm was used to track EV internalization. 3D models of organ tissue were also used to analyze EV uptake. It was shown that EVs tend to act as regulators for the tumor microenvironment and do contribute to tumor initiation, displaying a preference to target healthy cells as shown in the 3D models. In the cell lines utilized in the study (including healthy colon cells and the CRC-EV parental line), high levels of colorectal EV internalization were observed with nearly 100% of cells being EV-positive.

Outcomes and Implications

With colorectal cancer being the third most common cancer in the world, understanding tumor initiation in the context of EVs could prove to be helpful in the development of therapies. EVs could be used as a way to identify colorectal cancer earlier in its progression. The tropism towards healthy cells displayed by the EVs suggests that they could be used as biomarkers for diagnostic purposes. The use of deep learning technology also proved to be useful in identifying EV metastatic potential. The ability to track EV behavior with artificial intelligence provides an approach that could be translated into clinical monitoring.