Cancer is a leading cause of death globally. Traditional in vitro models, such as monolayer cultures, fail to accurately represent the complexity of the tumor microenvironment (TME) and the in vivo behavior of cancer. Animal models, while widely used, suffer from low translational success rates due to significant physiological differences between animals and humans. This necessitates the development of more physiologically relevant human in vitro models for studying cancer biology and developing therapeutics. Two-dimensional (2D) Transwell-based platforms allow co-culture studies but lack the dynamic nutrient exchange and tissue architecture of in vivo settings. Three-dimensional (3D) organoid models offer improved tissue mimicry but still lack vascularization and the flow dynamics crucial for tumor growth and metastasis. The emergence of organ-on-chips (OOCs) or microphysiological analysis platforms (MAPs) addresses these limitations by integrating microfluidics to create perfused microchannels populated by differentiated cells. These systems offer spatial separation of tissues, tissue-tissue interfaces, controlled fluid flow, and integrated biosensors for real-time monitoring, paving the way for more accurate preclinical cancer research and drug development.

This review extensively examines existing literature on various in vitro cancer models, highlighting the limitations of traditional 2D and 3D models. It then delves into the evolution of organ-on-chips (OOCs) and their application in cancer research. The review systematically explores how different cancer types (breast, brain, gastrointestinal, lung, liver, pancreatic, and urinary tract) have been modeled using OOCs, emphasizing the unique characteristics of each cancer model and the specific aspects of the TME being recreated. Numerous studies demonstrating the use of OOCs for studying key biological hallmarks of different cancers, such as stroma invasion, angiogenesis, metastasis, and drug resistance, are discussed. The review also analyzes the use of integrated biosensors for real-time monitoring of cellular responses and drug efficacy within these models.

The authors employed a systematic review methodology. They conducted an extensive literature search across various databases to identify relevant publications on in vitro tumor models on chips and integrated microphysiological analysis platforms (MAPs). The search strategy likely involved specific keywords related to organ-on-a-chip, cancer-on-a-chip, microfluidic devices, tumor microenvironment, and specific cancer types. Inclusion and exclusion criteria were applied to select relevant papers. These criteria likely focused on studies that described the development and application of microfluidic devices for modeling cancer, including details on cell types used, TME components recreated, and applications in drug screening. The selected publications were critically analyzed to extract information on the design and functionality of the various microfluidic devices, the biological features of the cancers being modeled, and the key findings from each study. Data were likely extracted using standardized forms to ensure consistency and accuracy. The extracted data were then synthesized to provide a comprehensive overview of the current state of cancer-on-chip technology.

The review highlights significant advancements in cancer-on-chip technology. Different cancer types have been modeled, with unique focuses depending on the cancer's characteristics. For breast cancer, the models focused on recapitulating various stages of the invasion-metastasis cascade, including stroma invasion, intravasation, and extravasation to different organs (bone, muscle). Brain cancer (glioblastoma) models concentrated on recreating the diverse tumor niches (perivascular, invasive, hypoxic) to study cancer stem cell behavior and drug response. Gastrointestinal cancer models investigated tumor-stroma interactions and the impact of biophysical factors (flow, stiffness) on cancer progression. Lung cancer models integrated the air-liquid interface to mimic alveolar function and studied metastasis. Liver and pancreatic cancer models focused on tumor-vasculature and tumor-stroma interactions. Urinary tract cancer models explored the TME's impact on cancer growth. The review emphasizes the importance of fluid flow in these models, highlighting its role in mimicking in vivo conditions and improving cell viability. A shorter culture period in OOCs compared to in vivo tumor development was highlighted as a key advantage. The use of integrated biosensors for real-time monitoring of cellular responses and drug efficacy was also discussed. Finally, the review explored synergistic approaches combining organoids and OOCs and the use of vascularized human-on-a-chip systems.

The findings of this review demonstrate the significant potential of cancer-on-a-chip technology to revolutionize cancer research and drug development. The ability to recapitulate the complex TME in vitro offers a more accurate and physiologically relevant model compared to traditional methods. The incorporation of fluid flow and integrated biosensors enhances the precision and throughput of drug screening and testing. This technology allows for the modeling of different cancer types, providing insights into cancer biology and mechanisms of metastasis and drug resistance. The use of patient-derived cells and organoids further personalizes the models, improving the prediction of individual patient responses to therapies. The review also highlighted the need to address remaining challenges such as the complexity of fully recapitulating the TME, the choice of appropriate materials for chip fabrication, and the availability of patient-specific cells. These limitations require further investigation and refinement of the technology.

Cancer-on-a-chip technology holds immense promise for advancing cancer research and personalized medicine. The ability to create sophisticated in vitro models that accurately reflect the complexities of the tumor microenvironment offers significant advantages over traditional methods, particularly in drug screening and efficacy testing. While challenges remain, ongoing research efforts towards improving the complexity and precision of these models, along with incorporating advanced biosensors, are poised to significantly impact cancer biology and drug discovery.

The review focuses on published literature and may not include all relevant studies due to publication bias or limitations in database searches. The interpretation of results from cancer-on-a-chip models needs to be carefully considered, as these are simplified representations of the complex in vivo TME. The heterogeneity of cancers and individual patient responses may not be fully captured by all models. Furthermore, the review's breadth might lead to a less in-depth examination of individual cancer types.