Ninety percent of clinical trials fail because - despite prior testing with animals - new drugs are often not effective or they turn out to be too toxic. We are developing microphysiological systems that can better predict which drug candidates will be successful in first-in-human trials. Microphysiological systems (MPS) are microfluidic cell-culture devices that house mimics of human organs made from primary or stem cell sources. Testing drugs with MPS instead of animals could potentially save billions of dollars, because data from MPS might uncover those compounds that won’t be successful earlier in the drug development process. However, despite years of development, current MPS are not ready to realize their potential role in drug development. The devices can be difficult to assemble and use, and they also do not yet fully capture the complexity of the human body. Our goal is to build an MPS device that mimics the body more fully (i.e a device that contains only physiological amounts of blood surrogate) and that is easy to operate.
We are developing tissue chips and multi-tissue microphysiological devices that support the culture of multiple tissues with physiologically-relevant connections and streams of cell-culture medium. The goal in this project is to construct a device that realistically mimics human metabolism as a means to reliably predict the efficacy or toxicity of drugs. Our focus is on a design that will reduce the amount of liquid (blood surrogate) needed to operate the device to near-physiological levels. We are achieving that goal by developing pumpless systems, reducing the volumes of interconnecting microfluidic channels, as well as by reducing the volumes of any valve components. Using those strategies, we have developed a pumpless system for the culture of healthy and activated endothelial cell layers, and a body cube that is capable of the co-culture of four tissues (Fig. 1). We are currently testing the devices to determine the effects of low liquid levels on the long-term culture of tissues. We are also developing mathematical models of drug metabolism inside the devices and ways to translate the data obtained with the devices to predictions for humans. The impact of MPS devices with near-physiological liquid levels is that they will predict drug action in patients with much higher reliability because drug metabolites in the system will be produced at near-physiologic concentrations.