OoCs are specifically engineered to replicate even the most complex physiological functions of human organs or tissues in a controlled laboratory environment. This advanced technology enables researchers to study the functionality or disease modeling of entire organs or tissues, providing valuable insights into their behavior under various conditions.⁴ These small, bioengineered platforms are designed to simulate the dynamic structural and functional properties of organs by integrating cellular biology, engineering, and microfluidics. The central principle of an OoC involves creating a microenvironment that closely mimics the cellular and mechanical characteristics of specific organs.^1,3 This is achieved by seeding living cells—such as epithelial, endothelial, or stromal cells—onto a chip made from flexible, transparent materials like polydimethylsiloxane (PDMS). The chip is typically equipped with microchannels and chambers that facilitate fluid flow, akin to blood circulation, supplying the cells with nutrients. This fluid flow mimics key physiological conditions, such as shear stress, tissue deformation, and chemical gradients, allowing the cells to behave in a manner similar to that in a natural organ.²

Mathematical modeling is instrumental in optimizing the design and functionality of these systems. Fluid dynamics models, including the Navier-Stokes equations, simulate blood flow through the microchannels, while diffusion models based on Fick’s law predict the transport of gases and nutrients.¹³ Additionally, tissue mechanics models are used to simulate the mechanical stretching or contraction of tissues, such as stretching or pulsatile pressure, particularly relevant for organs such as the heart or lungs. Shear flow is applied through laminar, pulsatile, or interstitial flow within microchannels, designed to mimic blood and other fluid flows experienced by tissues. This model is essential for studying cellular responses to varying flow conditions. Next, compression forces are implemented using devices like pistons or diaphragms to apply pressure on cells, simulating conditions found in bone or cardiac tissues where compressive stress is common. Finally, stretch or strain is achieved by cyclically applying vacuum pressure, which recreates the stretching forces experienced by tissues such as lung cells during breathing. These models together enable OOC platforms to replicate realistic cellular environments, facilitating controlled studies of cellular responses to physical forces and enhancing the relevance of in vitro tissue models.¹⁴ For example, a lung-on-chip may incorporate membranes that stretch to simulate breathing, while a heart-on-chip could replicate contractions.¹⁵ Cellular interaction models and pharmacokinetics/pharmacodynamics models further enhance the biological relevance of these systems by predicting cellular responses, drug interactions, and nutrient dynamics. These models shift from traditional PK-PD approaches to physiologically-based PK models, which assess drug distribution across interconnected organ systems. MOCs, leveraging microfluidics and bioengineered tissues, enable close replication of in vivo conditions by connecting organ models like the liver and heart to capture drug interactions. Such models improve the prediction of drug efficacy, toxicity, and metabolism, ultimately enhancing in vitro-to-in vivo extrapolations for drug testing and disease modeling.¹⁶ OOC technology enables the detailed study of organ-level functions,¹⁷ disease mechanisms,¹⁸ drug responses and toxicity,¹⁶ and reducing reliance on animal models.²⁰ These systems can integrate sensors to monitor parameters like oxygen levels, fluid flow, and cellular responses in real time.²¹ OoC technology, which stems from advancements in microfluidics, offers a unique capability to interact with multiple cell types simultaneously. This makes it an ideal candidate for studying near-human physiological interactions, presenting advantages over conventional 2D or 3D cell cultures. However, despite its potential, significant improvements are still required to achieve the same level of efficacy as in vivo testing.

The fabrication of OoC devices is a multidisciplinary process that blends microfluidics, biomaterials, tissue engineering, and cell biology to create functional microenvironments that mimic human organs.²² This process involves several key fabrication techniques and material considerations that vary depending on the type of organ being modeled. Typically, the fabrication techniques used in OoC can be divided into three categories: subtractive manufacturing, formative manufacturing, and additive manufacturing.²³

The choice of material is a critical aspect of OoC design. The materials used must be biocompatible, mechanically robust, and often optically transparent to facilitate real-time imaging and analysis. Common materials used in OoC systems are discussed in Table 1 below.

A critical challenge in OoC fabrication is designing systems that accurately replicate the physiological environment of an organ. This involves controlling fluid flow, mechanical stress, and biochemical gradients within the device.³⁶ For instance, flow systems in microfluidic channels are designed to replicate blood flow or nutrient supply, which is essential in maintaining the viability of cells in systems like blood vessel-on-a-chip and kidney-on-a-chip.³⁷ MOC devices, like lung-on-a-chip, mimic the physical forces on cells by using cyclic stretching to recreate the lung’s expansion and contraction during breathing.³⁸ Similarly, kidney-on-a-chip systems use shear stress to simulate the filtration dynamics of renal epithelial cells. These forces are key to maintaining cellular functions in an in vivo-like environment.³⁸ A major advancement in OoC technology is the integration of multiple cell types within a single chip. A design challenge is multi-organ integration, where several OoCs are connected to replicate inter-organ communication. This “body-on-a-chip” approach is especially useful for drug toxicity studies, as it mimics the interaction between different organs, such as the liver and kidney.³⁹ For systems like pancreas-on-a-chip, which require intricate interactions between different cell types (e.g., insulin-secreting beta cells and glucagon-secreting alpha cells), microfluidic devices can simulate the dynamic glucose regulation in the body. This setup is valuable for studying diseases like diabetes.⁴⁰ Similarly, liver-on-a-chip models aim to replicate hepatocyte interactions within a 3D matrix to assess liver functions like detoxification.¹⁰

Several OoC models have been developed over the past few years, particularly since the introduction of the first lung-on-a-chip model by Huh et al in 2007.⁴¹ Today, nearly all human organs have been replicated in microfluidic models to study their complex interactions. Table 2 below provides a comprehensive list of OoCs developed to date.

Several OoC models have been developed over the past two decades, with the most extensively researched models being those for the lungs, liver, and tumors.⁸⁰ This focus is primarily due to the essential functions of these organs, their disease prevalence in clinical settings, their diverse roles in drug metabolism, and the high incidence of cancer in these organs, making them critical for therapeutic testing and development.^80,81 The limitations of traditional in-vitro models in replicating diseased states have driven the development of these OoC models. However, with the need to accelerate disease modeling and preclinical testing, an increasing number of models are now being explored.⁸² With advancements in technology, OoCs are now able to create near-accurate microenvironments for several organs, incorporating dynamic circulation and achieving a degree of actual tissue complexity within the laboratory.^2,83 Below are various engineering approaches demonstrated by research teams worldwide.

In a study presented by Pauline et al, a second-generation lung-on-a-chip model replicates the structure and function of the human lung alveoli more accurately than previous models, as depicted in Figure 3. The primary innovation is the use of a collagen-elastin membrane that mimics the extracellular matrix of lung tissue, replacing synthetic materials like PDMS. The goal was to better recreate the complex mechanical and biochemical environment of lung alveoli in vitro. The biological membrane, made from collagen and elastin, is created by pipetting a CE solution onto a gold mesh, where it forms a thin, stretchable membrane through surface tension and evaporation. The mesh provides a structural scaffold for an array of 40 alveoli with dimensions similar to those found in vivo. The researchers cultured human lung alveolar epithelial cells and endothelial cells on both sides of the membrane to model the air-blood barrier. They observed that the membrane not only supported cell growth but also allowed for mechanical stretching that mimicked breathing motions, with cells forming a functional barrier over several weeks.⁸⁴

Compared to PDMS membranes, the CE membrane showed superior properties, such as better permeability, optical clarity, biodegradability, and lower absorption of small molecules. The cells maintained their phenotypes, expressed typical alveolar markers, and demonstrated robust tight junction formation, crucial for mimicking the lung’s barrier function. Additionally, the stretchable CE membrane showed mechanical resilience under cyclic negative pressure, closely simulating physiological strain during breathing. This second-generation lung-on-a-chip platform addresses several limitations of earlier models by better mimicking the alveolar environment in terms of geometry, mechanical properties, and cell-matrix interactions. This system has potential applications in drug screening, disease modeling, and personalized medicine, providing a more accurate representation of the lung’s responses to treatments or pathological conditions.⁸⁴

Another model developed by Kwon et al, as in Figure 4, presents a liver acinus dynamic (LADY) chip designed to model the liver’s zonation and assess drug-induced zonal hepatotoxicity. This innovative chip mimics the structure of a liver acinus, recapitulating key features such as zonal expression patterns and metabolic functions of the liver. The LADY chip consists of HepG2 cells cocultured with human umbilical vein endothelial cells (HUVECs) to more accurately reflect the liver’s microenvironment. In this study, the chip was developed to simulate the flow of oxygen and nutrients, as seen in vivo, from periportal zone 1 to perivenous zone 3. The result was the formation of liver zonation, with different zones showing varied metabolic activity—zone 1 hepatocytes expressed higher levels of the enzyme phosphoenolpyruvate carboxykinase, while zone 3 hepatocytes showed elevated levels of cytochrome P450 2E1. This zonal distinction is critical in assessing the liver’s response to drugs. The researchers tested acetaminophen (APAP)-induced hepatotoxicity and found that zone 3 cells were more susceptible to cell death due to higher metabolic activation of APAP in that region, correlating with the natural function of CYP2E1 in drug metabolism. Interestingly, coculturing HepG2 cells with HUVECs resulted in enhanced resistance to APAP-induced toxicity, likely due to the supportive interactions between hepatocytes and endothelial cells. Overall, the LADY chip successfully mimics liver zonation and offers a promising tool for drug testing, especially in evaluating zonal hepatotoxicity, reducing reliance on traditional animal models that may not fully predict human liver responses. The study also suggests future applications in personalized medicine and the testing of drug candidates for liver toxicity.⁸⁵

Tumor-on-chip models are gaining widespread attention due to their efficacy in mimicking the tumor microenvironment. A great example is a study by Mehta et al.,⁸⁶ in which the team introduces a novel microfluidic device for personalized drug testing in oral cancer, as illustrated in Figure 5. This research addresses key challenges in cancer treatment, such as tumor heterogeneity and drug resistance, by developing a dynamic, patient-derived spheroid model for testing drug combinations. The device was designed to overcome the limitations of previous models, including lengthy fabrication, absence of cancer stem cells, and lack of clinical correlation. The major thrust of this study includes the development of a 3D-printed, mold-based microfluidic platform with serpentine loops that allow for the mixing and testing of seven combinations of three chemotherapeutic drugs (paclitaxel, 5-fluorouracil, and cisplatin). Patient-derived cancer stem-like spheroids exhibited significant differences in drug responses, correlating with tumor differentiation and clinical diagnoses. For instance, spheroids from patients with well-differentiated tumors responded differently compared to those from moderately differentiated tumors. The study also revealed varying drug resistance profiles, with patient 1’s spheroids showing the highest resistance, which was associated with hypoxia and reduced cell proliferation. Another key finding was the successful maintenance of patient-derived spheroids under clinically relevant oxygen levels (below 5% O₂), replicating the tumor microenvironment more accurately. This study demonstrated that tumor differentiation status significantly influences drug response, which has rarely been explored in prior research. The research also highlighted the importance of hypoxia in contributing to drug resistance, particularly in patient 1’s spheroids, which exhibited both high hypoxia levels and strong resistance to chemotherapy.⁸⁶