Organ-on-chip based dynamic 3D breast cancer model for investigating glucose restriction as an adjuvant to chemotherapy

Triple-negative breast cancer (TNBC) represents a significant clinical challenge due to its aggressive nature and lack of targeted therapies. Recent studies highlight a correlation between glucose metabolism and tumor progression, with emerging evidence suggesting that glucose restriction may serve as an effective adjuvant therapeutic strategy. To address the limitations of traditional in vitro and in vivo experimental models adopted for translational research and drug discovery, an advanced 3D in vitro TNBC model was developed utilizing MDA-MB-231 cells embedded in tunable stiffness alginate-based hydrogels integrated into an organ-on-chip platform (MIVO®) resembling dynamic culture conditions that better mimic the in vivo-like diffusion of nutrients, drugs, and biochemical signals. The breast cancer model was employed (i) to recapitulate the clinically observed impact of glucose intake on tumor cells viability by simulating normal glucose uptake (25 mM glucose), glucose restriction (5 mM glucose) and fasting (0 mM glucose) conditions. Additionally, (ii) the cross-interaction between fasting and chemotherapy was investigated by combining Cisplatin (10 uM) treatment with different fasting approaches. More specifically, distinct fasting windows, both before and/or during chemotherapeutic treatment, were considered to faithfully mimic different used clinical approaches. Both 3D and 2D in vitro tumor models were cultured in static and dynamic conditions using the organ on chip platform, and the effects of the different therapeutic strategies were evaluated in terms of cell viability and proliferation. In parallel, an in vitro healthy control model composed of human dermal fibroblasts (HDFs) was also conducted, to evaluate the potential toxic side effects on healthy tissues. Results demonstrated a marked decline in tumor cell viability and proliferation following glucose deprivation, whereas the healthy control showed minimal changes in response. Furthermore, groups treated with combined fasting-mimicking approach and cisplatin exhibited a reduction in cell viability and proliferation consistent with clinical findings, with differences in terms of effects of the treatment (drug and fasting) onto cells morphology. Interestingly, these results indicate that the combinatory approach of 3D tumor cells and fluid dynamic cultures enable to successfully demonstrate the tumor cells sensitivity to glucose deprivation, much more than healthy tissue, supporting the potential clinical application of glucose-restricted adjuvant therapies. Additionally, this dynamic organ-on-chip model exhibited a more permissive diffusion of drugs into the 3D matrix-based tumors, resulting in higher cell death than in the static culture, aligning more closely with clinical data. In conclusion, this combined platform represents a robust and scalable tool for investigating tumor cells-matrix interactions, screening drug efficacy, and exploring metabolic dependencies in a more controlled and predictive in vitro environment.