Industrializing Organ-on-a-Chip Technology

Quite simply, drug development productivity remains a major challenge. Historically, for every 10 drugs that enter phase 1 clinical trials, only about one succeeds. This low success rate highlights a gap in preclinical models’ ability to accurately predict human outcomes.

That was one of the inspirations behind organ-on-a-chip technology. It was developed to address this challenge by improving the translational relevance of preclinical testing, providing human-relevant data earlier in the pipeline to better inform decisions before clinical trials begin, and increase pharma productivity. 

Organ-on-a-chip is a technology that enables us to recreate the inner workings of biology on a culture device, by bringing cells together in a microenvironment where they believe they're inside the human body. As a result, they should function more like they do in the human body, thereby providing data with a higher translational value. 

This technology is well proven and has been widely published across multiple organ systems. The foundational lung-on-a-chip study, published in Science in 2010, demonstrated that the system could model a multistep immune response. Researchers recreated the human alveolar epithelial-endothelial interface and showed immune cell adhesion, migration from the vascular to epithelial compartment, and subsequent bacterial engulfment, mirroring the complex in vivo immune processes. This work established that organ-on-a-chip could reproduce sophisticated human biological functions in vitro.

Early organ-chip systems were highly customized research setups with extensive tubing and separate modules for the microfluidics and mechanical stretch. These systems were labor-intensive and had low throughput.

As Emulate spun out of the Wyss Institute, the technology was engineered into an integrated commercial platform called Zoë^®. This platform consolidated the microfluidics and vacuum stretch into a pressure-driven, self-contained instrument capable of running up to 12 chips per unit. With more than 600 instruments deployed globally, this marked a successful transition from academic proof-of-concept to a worthy industrial tool. 

In a landmark 2022 study, Emulate evaluated 870 human Liver-Chips to assess their ability to model DILI. The team tested 27 small molecules, 22 of which had known clinical DILI outcomes, across multiple concentrations.

Liver-Chip successfully identified hepatotoxic compounds with a sensitivity of 87% and a specificity of 100% while also distinguishing between structural analogs that were more or less hepatotoxic, including structurally similar drug analogs such as clozapine (hepatotoxic) and olanzapine (clinically safer). The platform also demonstrated superiority compared to hepatic spheroids or animal models.

While the biological value is well recognized, pharmaceutical companies prioritize reproducibility, reliability, throughput, and actionable decision-making. Users need statistically rigorous, scalable systems that can seamlessly integrate into existing workflows and automation pipelines.

The focus has shifted from proving biological relevance to demonstrating operational robustness and enabling confident decision-making for risk assessment and first-in-human transitions. 

AVA represents Emulate’s response to the industry’s demand for platforms that combine biological relevance with operational rigor. The benchtop, self-contained system integrates environmental control, automated imaging, and high-throughput capacity. AVA supports up to 96 chips (emulations) per run, while reducing hands-on time by approximately 60%. Automation-ready consumables enable seamless integration with robotic liquid handling.

Equally important, the platform has demonstrated strong technical and biological reproducibility. Microfluidic flow remains tightly controlled (<5% variation), 94% of chips pass quality control thresholds, hepatocyte function remains consistent, and results have been replicated across internal and external laboratories.

Together, these performance metrics suggest that AVA is engineered not only for scale but for the statistical confidence required in regulatory and industrial decision-making.

AVA generates thousands to millions of multimodal data points per experiment through longitudinal sampling, imaging, and endpoint molecular analyses. This rich dataset enables integration with AI and digital modeling approaches.

As datasets grow and qualification programs mature, organ-chips could contribute to next-generation risk assessment frameworks, potentially reducing reliance on animal testing and enhancing confidence in human translation. 

In the near term, organ-chips may coexist alongside traditional animal models to strengthen first-in-human decision-making. As validation progresses, there is potential to reduce from two animal species to one, and ultimately to replace animal testing in certain contexts.

The ultimate goal is to support a more predictive, human-relevant preclinical ecosystem, combining organ-chips, digital twins, and AI-enabled models to improve drug development outcomes.

The introduction to this interview includes text that has been created with the assistance of generative AI and has undergone editorial review before publishing. Technology Networks' AI policy can be found here.