A three-dimensional (3D) printed artificial tumor tissue has emerged, capable of growing cancer cells derived from actual cancer patients in an environment that closely mimics in vivo conditions. When used in conjunction with artificial intelligence (AI) technology that predicts prognosis by analyzing images of the growth of the artificial tumor tissue, it is expected to aid in personalized treatment for patients.
Researchers, including Professors Tae-Eun Park and Hyun-Wook Kang from the Ulsan National Institute of Science and Technology (UNIST) Department of Biomedical Engineering and Professor Seung-Jae Myung from Asan Medical Center, have developed an artificial tumor tissue called 'Eba-PDO' that reproduces the high stiffness and hypoxic environment of real cancer tissue. The research findings were published in the online edition of the journal Advanced Science in March.
Cancer cells proliferate rapidly, leading to high density and resulting in a harder texture compared to normal tissue, and they thrive in environments lacking oxygen. Although existing artificial tumor tissues are made from cells taken from actual patients, they have failed to accurately recreate this environment, leading to distortions in the growth patterns of cancer cells and their responses to drugs.
The researchers mixed cancer organoids cultivated from cancer cells taken from patients in a three-dimensional format with bioink, aligning them in a bead form to print new artificial cancer tissue. The bioink is designed to reproduce the hard and hypoxic conditions that cancer cells require by mixing gelatin and extracellular matrix components.
The artificial tumor tissue grown using this method maintained a consistent shape for the same patient, but there were differences in size and shape among different patients. Additionally, the artificial tumor tissue exhibited a higher genetic expression similarity compared to cancer tissues taken from actual cancer patients, accurately reproducing differences in the responsiveness of patients to the chemotherapy drug 5-fluorouracil (5-FU).
The researchers also developed an AI capable of predicting the expression of the CEACAM5 gene solely from microscopic images. CEACAM5 is a protein commonly found in solid tumors, including colorectal cancer, and is known to increase metastatic potential and drug resistance.
When this protein is overexpressed in artificial tumor tissue, cell-to-cell adhesion weakens, resulting in a less dense and unbalanced form of the tumor tissue. The AI was trained to learn these morphological changes to predict gene expression levels. Experimental results showed that the AI could identify the expression of key marker genes for colorectal cancer prognosis with 99% accuracy.
The researchers noted, "This method of reproducing and analyzing the growth of actual cancer cells in vitro is expected to enable more precise patient-centered treatment, and by integrating immune cells and vascular structures in the future, it could be expanded into a more sophisticated artificial cancer model."
References
Advanced Science (2025), DOI: https://doi.org/10.1002/advs.202407871