Cancer-on-a-chip for precision cancer medicine

Abstract

Many cancer therapies fail in clinical trials despite showing potent efficacy in preclinical studies. One of the key reasons is the adopted preclinical models cannot recapitulate the complex tumor microenvironment (TME) and reflect the heterogeneity and patient specificity in human cancer. Cancer-on-a-chip (CoC) microphysiological systems can closely mimic the complex anatomical features and microenvironment interactions in an actual tumor, enabling more accurate disease modeling and therapy testing. This review article concisely summarizes and highlights the state-of-the-art progresses in CoC development for modeling critical TME compartments including the tumor vasculature, stromal and immune niche, as well as its applications in therapying screening. Current dilemma in cancer therapy development demonstrates that future preclinical models should reflect patient specific pathophysiology and heterogeneity with high accuracy and enable high-throughput screening for anticancer drug discovery and development. Therefore, CoC should be evolved as well. We explore future directions and discuss the pathway to develop the next generation of CoC models for precision cancer medicine, such as patient-derived chip, organoids-on-a-chip, and multi-organs-on-a-chip with high fidelity. We also discuss how the integration of sensors and microenvironmental control modules can provide a more comprehensive investigation of disease mechanisms and therapies. Next, we outline the roadmap of future standardization and translation of CoC technology toward real-world applications in pharmaceutical development and clinical settings for precision cancer medicine and the practical challenges and ethical concerns. Finally, we overview how applying advanced artificial intelligence tools and computational models could exploit CoC-derived data and augment the analytical ability of CoC.

Fig. 1. Comparisons among different types of current preclinical models for cancer study.

Fig. 2. Cancer-on-a-chip microphysiological systems for tumor microenvironment modeling and precision cancer medicine screening.

Fig. 3. Representative cancer-on-a-chip models for tumor microenvironment modeling and therapy screening. (A) A tumor vasculature CoC model mimics the extravasation of tumor cells from vasculature. Reproduced from ref. with permission from Springer Nature, copyright 2017. (B) A tumor stromal niche model studies CAFs activated by tumor cells and induced the over deposition of ECM components collagen, fibronectin and hyaluronic acid. Reproduced from ref. with permission from Wiley, copyright 2016. (C) An GBM immune niche model studies TAM associated immunosuppression and promoted angiogenesis. Reproduced from ref. with permission from Elsevier, copyright 2018. (D) A CoC with vascularized micro tumors screened effective chemo drugs with high reproducibility and biomimicry. Reproduced from ref. with permission from Royal Society of Chemistry, copyright 2021. (E) A leukemia chip modeled the in vivo leukemic bone marrow niche and CAR T cell therapy on chip. Reproduced from ref. , CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). (F) An automatic microfluidic CoC platform enabled personalized drug screening of for different patients. Reproduced from ref. , CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).

Fig. 4. The path to build the next generation CoC model for precision cancer medicine.

Fig. 5. Strategies to evolve CoC with higher accuracy, analytical ability and translational applications. (A) A patient-derived colorectal cancer (CRC) chip reproduced in vivo pathophysiology and anatomical structure built with autologous patient cancer cells, CAFs, TILs, and colon mucosa components including colonocytes (CCs), transit-amplifying cells (TAs) and intestinal stem cells (ISCs). Reproduced from ref. with permission from Springer Nature, copyright 2024. (B) A vascularization of colon organoids-on-a-chip showed enhanced growth under perfusable culture on chip comparing with conventional static condition. Reproduced from ref. with permission from Wiley, copyright 2020. (C) A heart and liver cancer multi-organs-on-chip model built with human iPSC-derived cardiomyocytes and hepatocellular carcinoma cells for investigating the acute toxicity induced by anti-tumor drugs. Reproduced from ref. with permission from National Academy of Sciences, copyright 2017. (D) Programmable flow control on CoC chip. Reproduced from ref. , CC BY-NC 3.0 (https://creativecommons.org/licenses/by-nc/3.0/). (E) A multi-sensor integrated chip system with electrode-based O₂, pH sensors and lactate and glucose biosensors. Reproduced from ref. with permission from Royal Society of Chemistry, copyright 2014. (F) A standardized high-throughput microfluidic platform in 384-well plate format containing 40 units of colorectal cancer tubes for studying drug-induced toxicity on epithelial barriers. Reproduced from ref. , CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). (G) A deep-learning model trained with clinical data and integrated with on-chip readouts to predict patient survival and identify drug candidates based on T cell infiltration in tumor sites. Reproduced from ref. with permission from the authors, copyright 2022. (H) An ODE-based computational model calibrated with GBM CoC-derived data depicted the interactions between T cells, TAMs and GBM cells in TMEs of different GBM subtypes (proneural, classical, mesenchymal) and tested combinational immunotherapies to enhance treatment efficacy. Reproduced from ref. with permission from Wiley, copyright 2021.

Lunan Liu

Katsuo Kurabayashi

Weiqiang Chen