Bioprinted Organoids Predict Tumor Shape, Prognosis

Abstract

Heterogeneity and the absence of a tumor microenvironment (TME) in traditional patient-derived organoid (PDO) cultures limit their effectiveness for clinical use. Here, Embedded Bioprinting-enabled Arrayed PDOs (Eba-PDOs) featuring uniformly arrayed colorectal cancer (CRC) PDOs within a recreated TME is presented. This model faithfully reproduces critical TME attributes, including elevated matrix stiffness (≈7.5 kPa) and hypoxic conditions found in CRC. Transcriptomic and immunofluorescence microscopy analysis reveal that Eba-PDOs more accurately represent actual tissues compared to traditional PDOs. Furthermore, Eba-PDO effectively capture the variability of CEACAM5 expression-a critical CRC marker-across different patients, correlating with patient classification and differential responses to 5-fluorouracil treatment. This method achieves an uniform size and shape within PDOs from the same patient while preserving distinct morphological features among those from different individuals. These features of Eba-PDO enable the efficient development of a label-free, morphology-based predictive model using supervised learning, enhancing its suitability for clinical applications. Thus, this approach to PDO bioprinting is a promising tool for generating personalized tumor models and advancing precision medicine.

A joint research team, affiliated with UNIST has developed a groundbreaking 3D-printed artificial tumor tissue capable of replicating the in vivo conditions of patient-derived cancer cells. This innovative model not only simulates the tumor microenvironment but also integrates artificial intelligence (AI) technology that can predict patient prognosis solely from images of tumor growth.

Led by Professors Tae-Eun Park and Hyun-Wook Kang from the Department of Biomedical Engineering at UNIST, in collaboration with Professor Seung-Jae Myung from Seoul Asan Medical Center, the team has developed an artificial tumor tissue called the Embedded Bioprinting-enabled Arrayed PDOs (Eba-PDOs). This model accurately reproduces the high matrix stiffness and hypoxic conditions characteristic of real tumor tissues, particularly in colorectal cancer (CRC). By analyzing the morphology of these artificial tumors with AI, the team achieved 99% accuracy in predicting the expression of key prognostic gene markers, such as CEACAM5, in colorectal cancer.

Cancer cells proliferate rapidly, leading to dense, firm tissues with low oxygen levels-conditions that influence tumor growth and drug response. Traditional artificial tumor models, even those derived from patient cells, have struggled to faithfully mimic these environmental factors, resulting in distortions in observed growth patterns and treatment responses.

To overcome this, the researchers cultured patient-derived cancer cells in three dimensions to create tumor organoids, which were then combined with bioink-composed of gelatin and extracellular matrix components-and printed in bead-like arrays. This method allowed for the precise fabrication of artificial tumors that replicate the dense, hypoxic environment of actual tumors. Each tumor, while maintaining a consistent shape within a single patient's samples, varied in size and morphology across different individuals.

Capitalizing on these morphological features, the team developed an AI model capable of predicting CEACAM5 gene expression from microscopic images. CEACAM5 is a protein frequently overexpressed in colorectal and other solid tumors, associated with increased metastatic potential and resistance to chemotherapy. When overexpressed, it weakens cell-cell adhesion, resulting in less dense tumor structures. The AI was trained to recognize these morphological cues and accurately estimate gene expression levels.

"Figure 1. Schematic illustration of the creation of tumor microenvironment (TME)-inspired embedded bioprinting-enabled arrayed patient-derived organoids (Eba-PDOs).

The artificial tumor model demonstrated a 29% improvement in gene expression similarity compared to conventional models, achieving a correlation of 90% with actual patient tumor tissues-up from approximately 70%. Furthermore, the model successfully replicated individual patient responses to 5-fluorouracil (5-FU) chemotherapy, highlighting its potential for personalized treatment planning.

This research was led by first authors Jeong Hyejin and Han Jonghyuk from UNIST. The team envisions that this approach, which faithfully mimics tumor growth and microenvironment conditions in vitro, will facilitate more precise, personalized cancer therapies. Future developments aim to incorporate immune cells and vascular structures into the models, further enhancing their fidelity.

The findings of this research were published online in the journal Advanced Science on March 28, 2025. This study has been supported by the Korean ARPA-H initiative under the Ministry of Health and Welfare (MOHW), the Industry Technology Alchemist Project under the Ministry of Trade, Industry & Energy (MOTIE), the Bio·Medical Technology Development Program and the Advanced Bio Tech·Talent Exchange Support Project under the Ministry of Science and ICT (MSIT), as well as the Glocal University 30 Project of the University of Ulsan funded by the Ministry of Education, and the COMPaaS joint research initiative.

Journal Reference

Jonghyeuk Han, Hye-Jin Jeong, Jeonghan Choi, et al., "Bioprinted Patient-Derived Organoid Arrays Capture Intrinsic and Extrinsic Tumor Features for Advanced Personalized Medicine," Adv. Sci., (2025).