Work

Modeling patient-specific adenocarcinoma as spheroids enhances the authenticity of cancer hallmarks and therapeutic susceptibility

Year: 2025

Type: article

Abstract: Introduction: To help understand and develop treatments for cancer, it is important to have patient-specific, reliable and physiologically-relevant human culture models. Cells from solid tumors have traditionally been grown in 2-dimensional (2D) monolayers, however, it is known that cancer cells can self-organize into 3-dimensional (3D) entities (“spheroids”), mimicking the intercellular relationships they presumably assume in situ in the patient. However, the cell molecular mechanisms underlying the differences between the same cancer cells grown in 2D compared to 3D have not been well established, particularly for adenocarcinomas with Epidermal Growth Factor Receptor (EGFR) mutations. Methods: In this study, using cells from an adenocarcinoma cell line bearing a mutation in the EGFR gene, we assessed the different cancer hallmarks (transcriptional profile, downstream signaling, epithelial-mesenchymal transition, invasiveness, therapeutic susceptibility) when the same cells were grown in 3D as opposed to 2D. Results: The 3D spheroids were enriched for genes critical to the cancer hallmark pathways Tumor Microenvironment Pathway, HIF1a Signaling, and Epithelial-Mesenchymal Transition (EMT) compared to the 2D cultures and expressed higher levels of phosphorylated ERK. They were also more radiosensitive to irradiation. Critically, the molecular constellation of the 3D “spheroids” more closely resembled that of the same type of tumor resected from a patient. Conclusion: The same tumor cells show different cancer-driving properties based on their spatial configuration in vitro; only when they are maintained in 3D is the actual primary tumor in situ accurately modeled. Therefore, we submit that patient-specific 3D spheroids are the best way to study cellular mechanisms driving disease and perhaps better plan a therapeutic regimen to which the patient is most likely to respond.

Source: Heliyon

DOI: 10.1016/j.heliyon.2025.e42977