Over the last decade, organ-on-chip (OoC) technology has become increasingly popular in the field of biomedical research. These microfluidic devices, adorned with human cells, aim to mimic the architecture and functional subunits of real human organs. Scientists are excited about OoCs because they offer a more biomimetic way to study human biology relative to legacy 2D cell culture platforms, while reducing the need for animal models. But like any emerging technology, OoC systems come with limitations and challenges. In this article, we’ll explore both sides: why researchers are enthusiastic about OoCs, and where the technology still falls short.

This table offers a quick overview of the advantages and disadvantages of organ-on-chip technology, for more detailed insights keep on reading.

Advantages: Why Scientists Love Chips

Closing the Translation Gap in Drug Development

The promise of OoC systems is that they offer human biology, at scale. Instead of relying on animal experiments, researchers can work with living human tissues in a controlled microenvironment. These devices more accurately mimic human organ function and physiology than many animal models. Species specific differences undermine the value of data from animal models, and this lack of human relevance is thought to be a major reason underlying the failure of so many drugs in clinical trials, despite promising animal data (around 92% by some estimates).

Building Tissues, Not Just Cell Layers

Another reason scientists love OoCs is their ability to support complex co-cultures that mimic the architecture of native tissue. Traditional cell culture often involves growing one cell type at a time or mixing several types together in a dish or cell culture plate. Organ-on-chip devices, by contrast, typically have separate chambers or channels that allow multiple cell types to grow in close but controlled proximity, just as they do in real tissues.

This enables interactions and crosstalk between cells and tissues, analogous to those found in the body. For instance, Dynamic42’s human lung models can accommodate epithelial cells, microvascular endothelial cells, and macrophages, co-cultured together under continuous perfusion in their respective compartments. Despite the complexity, each cell type organizes in much the same way as these cells would in a human body. This organized co-culture allows researchers to observe realistic interactions (e.g. immune cells communicating with both tissue layers) that would be hard to maintain in a standard dish.

Real‑Time Windows into Tissue Function

Organ-on-chip systems can provide real-time readouts of biological processes, a feature that excites scientists and drug developers alike. Because these devices are small and optically accessible, they can be equipped with sensors or monitored continuously under a microscope. This means we aren’t limited to taking end-point measurements; instead, you can watch tissue physiology unfold over time. A great example is integrating electronic sensors to measure transepithelial or transendothelial electrical resistance (TEER) on-chip. TEER is an indicator of barrier integrity, essentially showing how “leaky” or tight a cellular barrier is. Traditionally, measuring TEER in a cell culture required removing the plate from the incubator and using handheld electrodes, a manual process done at intervals. Recently, Kaden et al. (2025) reported a plug-and-play organ-chip with built-in TEER electrodes³. This enabled continuous recording of the biochip’s intestinal barrier, while inside the incubator, without any human intervention. This real-time monitoring allowed them to capture rapid, transient drops in barrier function when an inflammatory stimulus was introduced, changes that a once-a-day end-point assay might completely miss. Continuous readouts like these give researchers a wealth of dynamic information. How quickly does a drug disrupt a tissue barrier? How long does it take the cells to recover? By providing higher sensitivity and time-resolution, TEER assays can reveal subtle effects that traditional assays might overlook. Beyond TEER, biochips are amenable to real-time imaging (e.g. using time-lapse fluorescence microscopy to track cell behaviors or pathogen growth) and even real-time biochemical sensing such as biosensors for oxygen or metabolites.

Bug Me – Host-Pathogen Interactions Under Physiological Flow

Yet another beloved feature of organ chips is the ability to study host–pathogen interactions under controlled flow conditions. In living tissues, cells are constantly bathed in fluids that deliver nutrients and microbes, apply shear stress, and remove waste. Pathogen infections in vivo often involve circulating organisms or immune cells moving through the bloodstream. Traditional cell cultures, however, are static: pathogens and cells just sit together in a dish, which may not replicate how infections develop in real life. Organ-on-chip devices incorporate microfluidic flow to better mimic the circulation and fluid dynamics of an organ. This is particularly useful for infection models.

For example, a human lung-alveolus-on-a-chip was used to study influenza virus and bacterial co-infection. In this model, the alveolar epithelial side was exposed to air after initial culture to induce differentiation, while the capillary side was perfused with culture medium to simulate blood flow. Researchers infected the biochip with influenza A virus (through the airway channel), then later introduced Staphylococcus aureus bacteria, mimicking a secondary bacterial infection on an inflamed lung. Under these controlled conditions, they could observe the sequence of events: the virus damaged the lung epithelium, which in turn allowed S. aureus to adhere to the epithelial surface (as visualized by electron microscopy), and eventually the co-infection led to endothelial injury and barrier breakdown, with subsequent bacterial translocation into the endothelial channel. Because the biochip maintained a flowing “bloodstream,” it also enabled immune cells (the added macrophages) to react in a physiologically relevant way, producing inflammatory signals like TNF-α in response to the bacteria. Such insights would be difficult to obtain in a static culture, but with OoCs, scientists can recapitulate conditions analogous to subunits of the body, leading to more authentic host–pathogen insights.

Modeling multi organ crosstalk and pharmacokinetics on chip

Organ-on-chip technology also shines in studying multi-organ crosstalk & pharmacokinetics, an area where conventional models often fall short. In our bodies, organs don’t operate in isolation, especially when it comes to how a drug works. A medication taken by mouth will be absorbed by the gut, altered by the liver, circulated via blood, and perhaps act on multiple organs, all before being excreted. Reproducing this systemic complexity in the lab has been extremely challenging. Enter multi-organ on-chip systems (MoOCs), sometimes called “body-on-chip” or interconnected microphysiological systems. These setups link two or more organ-specific biochips with microfluidic channels so that media circulates between them. One recent example tackled the question of drug pharmacokinetics in pregnant women.

Researchers connected three biochips, a gut-on-chip, a liver-on-chip, and a placenta-on-chip, to simulate what happens when a pregnant woman takes the anti-inflammatory drug prednisone orally. In this integrated platform, an ingested drug could be absorbed through the gut lining, then pumped into the liver chip where it gets metabolized, and the resulting compounds then flowed into the placental barrier. Using this system, Graf et al. (2025) simulated the dosing of the anti-inflammatory drug prednisone and tracked its conversion to prednisolone (an active metabolite) and the transfer of both across the placenta. This effectively replicated first-pass metabolism in the liver and the selective filtration of the placenta, demonstrating that the placental tissue greatly limited fetal exposure to the active drug. In fact, the measured drug levels on the “fetal” side of the placenta chip were well below toxic thresholds and closely matched known clinical data from pregnant patients. This kind of multi-organ crosstalk experiment is extremely difficult in traditional cell culture, and unethical or impossible to do in humans.

Curious to learn more about this 3-organ model? Watch the webinar below:

Ethical, efficient alternatives to animal testing

Finally, a compelling advantage of organ-on-chip technology is the promotion of ethics and efficiency. From an ethical standpoint, any method that can reduce reliance on animal testing is a welcome advancement. OoCs have the potential to replace certain animal experiments by providing data on human cells that regulators and researchers consider sufficient for safety and efficacy evaluation. The FDA Modernization Act 2.0 (signed in 2022) removed the federal mandate for animal testing, explicitly allowing methods like organ chips to be used in drug approval processes. More recently, legislators introduced the FDA Modernization Act 3.0 to ensure the FDA actively updates its regulations to reflect this new mandate. The impetus for Act 3.0 came from frustration that, years after Act 2.0, many FDA guidelines still defaulted to animal data. Old habits die hard, and the newer bill essentially pushes the agency to fully embrace 21st-century toxicology and pharmacology methods. The significance of these acts for OoC is huge. It signals official encouragement for using human-relevant platforms in safety and efficacy testing. However, in practice, regulators need to be confident that OoC assays are validated, reproducible, and truly predictive of human outcomes. It’s important to note that OoC systems are complements to other methods, not outright replacements (at least not yet). Furthermore, regulatory frameworks for organ-on-chip technologies remain nascent, with less well-defined data acceptance criteria compared to established animal or in vitro models, underscoring the urgent need for standardized validation guidelines to build regulatory confidence and support broader adoption.

Limitations of organ-on-chip and opportunities for growth

Why organ‑on‑chip is powerful but still only part of the picture

With so many advantages, one might wonder why organ-on-chip technology isn’t everywhere already. The truth is that engineering biology is hard, and this fundamental fact underlies many of the current limitations of OoCs. These devices are engineered approximations of living organs, and no matter how clever the design is, they inevitably simplify or omit aspects of true physiology. A biochip recapitulates some organ functions, but not the full complexity of a human body. This reductionist nature of OoCs means that results obtained on-chip may not translate perfectly to an intact organism. In practice, translating organ-chip data to clinical decision-making often requires additional modeling or assumptions. For instance, computational frameworks like physiologically-based pharmacokinetic (PBPK) modeling are being used alongside biochips to bridge the divide between the microscale model and the whole human. The need for such “digital twin” approaches underscores that an organ-on-chip is still an isolated system, a piece of the puzzle rather than the whole picture. It also demonstrates the potential to integrate OoC platforms with powerful new AI systems. Enabling real word calibration of AI, with relevant human data. One can envision a future of fully automated, ‘lights-out’ OoC laboratories, sans humans, continuously feeding data into sophisticated algorithms around the clock.

Cell Quality Drives Predictivity

Another hurdle for OoCs lies in sourcing appropriate cells. Many OoC models still rely on cell lines or easily available cells that may not perfectly represent the real tissue. Primary human cells are often considered the gold standard for physiological relevance, but they come with challenges: limited supply, donor-to-donor variability, and finite lifespan in culture. In addition, some primary cells quickly lose their specialized functions once taken out of the body. Cell lines offer robustness, low cost, and wide availability, but because they’re immortalized, they often acquire genetic and phenotypic changes that limit how well they resemble normal tissue.

Researchers are addressing this by exploring induced pluripotent stem cell (iPSC)-derived cells and by developing immortalized primary-like cells. One successful example comes from a microphysiological liver model. Early iterations of the model used generic endothelial cells (like HUVECs, human umbilical vein cells) to represent the liver’s blood vessels, HUVECs are not liver-specific and lack some functions of true liver sinusoidal cells. In a 2023 study, the team switched to using upcyte human liver sinusoidal endothelial cells, endothelial cells that retain many primary liver-specific traits but can be expanded in culture over many more passages, and more closely maintain their in vivo phenotype under in vitro conditions. By sourcing a more relevant cell type, they improved the model’s fidelity. To implement OoCs widely (in industry screening or personalized medicine), further work is needed to ensure a reliable supply of quality cells.

Materials Matter – Why what a chip is made of can shape your data

Materials and fabrication issues present another practical challenge of current OoC systems, in short, materials matter. Many organ chips are made from polymers like PDMS (polydimethylsiloxane), which was adopted from the microfluidics field. PDMS is convenient: it’s transparent, oxygen-permeable, and easy to mold into microchannels. But a notable issue is that PDMS tends to absorb small hydrophobic molecules. If you put a drug compound into a PDMS-based biochip, a significant fraction might soak into the device’s walls instead of reaching the cells. This can lead to underestimating a drug’s effect or requiring higher, supraphysiological doses to achieve a response.  At Dynamic42 we provide biochips made of a unique medical-grade plastics ensuring biocompatibility and low drug adsorption even of very hydrophobic compounds. In a three-organ Dynamic42 chip, Graf et al. (2025) showed that prednisone exhibited no significant loss to adsorption over 24 hours of continuous perfusion, indicating essentially complete recovery and enabling truly quantitative pharmacokinetics.

Throughput versus complexity

A practical constraint that many labs discover is the trade-off between throughput versus complexity in OoC experiments. Simply put, while OoCs can recapitulate complex biology, they typically do so at low throughput. Each organ models is a bespoke, miniaturised system that requires individual handling, fluidic connections, and careful monitoring. To increase throughput, it’s often required to simplify the model. Conversely, the more complex and “realistic” you make the model the more challenging it becomes to scale up the experiments. For industrial adoption, this is a real issue: drug companies need to test thousands of compounds, and OoC systems just aren’t at that level of throughput yet.

Cost, learning curves, and real world adoption

Finally, we must address the practical reality of upfront cost and complexity when it comes to OoC adoption. Building and running these systems is more complex than running conventional cell culture approaches, as each OoC device is the product of sophisticated engineering. The learning curve can be steep for traditional biology labs using legacy workflows. That said, establishing a novel organ model is a process that can be broken down into individual steps where a scientist can start small with only a few tissue layers or using cell lines instead of primary cells. Increasing the complexity can happen in a step-wise approach, which makes the development of such a model a lot more feasible.

In this webinar our CEO and co-founder Dr. Martin Raasch nicely breaks down the steps required to develop a new model from scratch.

Another hurdle can be the upfront or running cost of the required equipment such as pumps, biochips or flow controllers. Here, it’s advisable to pay close attention to how well a system integrates within existing workflows or equipment in the lab. Some systems do not require capital investment to get started or work with existing instruments such as incubators or perfusion pumps. At Dynamic42 we pride ourselves in offering a system that fits seamlessly into existing lab environments by working with a broad range of pressure supplies, using standard incubators and having low consumable cost. We furthermore offer specialist training on-site or online, helping researchers get started more swiftly.

The Road Ahead – From clever devices to indispensable research tools

OoC technology is still in its relative infancy, and there is broad consensus that it will improve significantly in the coming years. While today’s OoC systems have limitations, the trajectory is toward more robust, user-friendly, and impactful versions. The road ahead involves bridging engineering, biology and AI even more tightly, scaling up what works, and proving the value of biochips in practical applications. If successful, OoCs could become a staple of labs and industry, a once cutting-edge idea turned indispensable tool.

Take-Home Messages

• Organs-on-chips offer human realism that traditional models lack
• They enable complex biology in vitro
• Real-time, high-content readouts are a key strength
• OoC complement excisiting methodologys but are not a full replacement for animals yet
• Multiple organs can be connected but we can’t replicate a body-on-chip just yet
• Careful considerations should be made when establishing a novel OoC System in the lab considering time and investment
• The field is rapidly evolving and systems improving

Would you like to listen to some researchers actually working with organ models? You can find their stories on our case studies page.

More interesting articles: