Current practices in drug development have led to therapeutic compounds being approved for widespread use in humans, only to be later withdrawn due to unanticipated toxicity. These occurrences are largely the result of erroneous data generated by in vivo and in vitro preclinical models that do not accurately recapitulate human physiology. Herein, a human primary cell-and stem cellderived 3D organoid technology is employed to screen a panel of drugs that were recalled from market by the FDA. The platform is comprised of multiple tissue organoid types that remain viable for at least 28 days, in vitro. For many of these compounds, the 3D organoid system was able to demonstrate toxicity. Furthermore, organoids exposed to non-toxic compounds remained viable at clinically relevant doses. Additional experiments were performed on integrated multi-organoid systems containing liver, cardiac, lung, vascular, testis, colon, and brain. These integrated systems proved to maintain viability and expressed functional biomarkers, long-term. Examples are provided that demonstrate how multi-organoid 'body-on-a-chip' systems may be used to model the interdependent metabolism and downstream effects of drugs across multiple tissues in a single platform. Such 3D in vitro systems represent a more physiologically relevant model for drug screening and will likely reduce the cost and failure rate associated with the approval of new drugs.

The current drug development pipeline takes approximately fifteen years and $2.6 billion to get a new drug to market. Typically, drugs are tested on two-dimensional (2D) cell cultures and animal models to estimate their efficacy before reaching human trials. However, these models are often not representative of the human body. The 2D culture changes the morphology and physiology of cells, and animal models often have a vastly different anatomy and physiology than humans. The use of bioengineered human cell-based organoids may increase the probability of success during human trials by providing human-specific preclinical data. They could also be deployed for personalized medicine diagnostics to optimize therapies in diseases such as cancer. However, one limitation in employing organoids in drug screening has been the difficulty in creating large numbers of homogeneous organoids in form factors compatible with high-throughput screening (e.g., 96-and 384-well plates). Bioprinting can be used to scale up deposition of such organoids and tissue constructs. Unfortunately, it has been challenging to 3D print hydrogel bioinks into small-sized wells due to well-bioink interactions that can result in bioinks spreading out and wetting the well surface instead of maintaining a spherical form. Here, we demonstrate an immersion printing technique to bioprint tissue organoids in 96-well plates to increase the throughput of 3D drug screening. A hydrogel bioink comprised of hyaluronic acid and collagen is bioprinted into a viscous gelatin bath, which blocks the bioink from interacting with the well walls and provides support to maintain a spherical form. This method was validated using several cancerous cell lines, and then applied to patient-derived glioblastoma (GBM) and sarcoma biospecimens for drug screening.

Variability in patient response to anti-cancer drugs is currently addressed by relating genetic mutations to chemotherapy through precision medicine. However, practical benefits of precision medicine to therapy design are less clear. Even after identification of mutations, oncologists are often left with several drug options, and for some patients there is no definitive treatment solution. There is a need for model systems to help predict personalized responses to chemotherapeutics. We have microengineered 3D tumor organoids directly from fresh tumor biopsies to provide patient-specific models with which treatment optimization can be performed before initiation of therapy. We demonstrate the initial implementation of this platform using tumor biospecimens surgically removed from two mesothelioma patients. First, we show the ability to biofabricate and maintain viable 3D tumor constructs within a tumor-on-a-chip microfluidic device. Second, we demonstrate that results of on-chip chemotherapy screening mimic those observed in subjects themselves. Finally, we demonstrate mutation-specific drug testing by considering the results of precision medicine genetic screening and confirming the effectiveness of the non-standard compound 3-deazaneplanocin A for an identified mutation. This patient-derived tumor organoid strategy is adaptable to a wide variety of cancers and may provide a framework with which to improve efforts in precision medicine oncology.