Organoids and organ-on-a-chip devices fall under the umbrella of microphysiological systems (MPS), which, in general, can be defined as miniature cell cultures—usually three dimensional (3D) models—that recapitulate aspects of human physiology (Skardal et al., 2016; Clevers, 2019). Today, our field has made significant progress since the days where 2D cell cultures and animal models were our only options as preclinical and basic science experimental model systems. We are privileged to have access to countless cell lines that we can increase the utility of by implementing them in 3D (Prestwich et al., 2007). We have generated biomaterial approaches to create a variety of methods by which to support human patient–derived primary cell–based 3D organoids and tissue constructs. (Mazzocchi et al., 2017). We have merged microfluidic device technology with 3D cell cultures to generate tissue- and tumor-on-a-chip platforms (Bhise et al., 2014). This is a rapidly evolving field. However, adoption of these models—while growing—is still limited given the reliance of biomedical research on 2D cell cultures and animal models. (Maltman and Przyborski, 2010). In this Specialty Grand Challenge, we consider the benefits, the hurdles, and the current implementation and future directions of MPS.

Three dimensional approaches to cell cultures, such as organoids, hydrogel-based constructs, and tissue chips, allow for unprecedented support of cell types that previously were not easy to maintain in 2D cell cultures. As a prime example, only 10–15 years ago, primary human hepatocytes (PHH) were seen as the gold standard for pharmaceutical drug compound toxicity screening. However, the cultures used at the time were either 2D cultures on tissue culture plastic or collagen hydrogel sandwich cultures. Previously, cells would die in about 10 days, during which their functional capabilities (drug metabolism, albumin, and urea production, etc.) would deteriorate drastically within a day or two. In the latter, the collagen sandwich cultures further extended PHH viability, but after 2 or 3 weeks, viability and functionality would quickly deteriorate also. Circa 2012–2015 and onwards, numerous novel biomaterial and 3D culture techniques were invented and deployed, which have made PHH culture relatively straightforward. Examples include integration of decellularized liver extracellular matrix (ECM) into hydrogel biomaterials to provide supporting extracellular matrix (ECM) components and biochemical factors to PHHs (Skardal et al., 2012). Spheroid cultures—both PHH alone and PHHs combined with other liver cells—enabled crucial cell-cell interactions that dramatically improved PHH viability and function (Skardal et al., 2017; Rajan et al., 2020a; Skardal et al., 2020). Today, we have seen numerous liver-on-a-chip devices with impressive in vivo liver-like functionality, albeit appropriately volumetrically scaled down (Kang et al., 2015; Knowlton and Tasoglu, 2016; Bauer et al., 2017). These advancements were made by efforts to move past simplistic cell cultures, and to rather recapitulate the complexities of native tissue microenvironments. Today our MPS portfolio comprises a range of form factors, including relatively simple 3D models such as homogeneous cell spheroids, organoids maintained in ECM hydrogels, organ- and tumor-on-a-chip systems, and integrated multi-tissue “body-on-a-chip” platforms (Figure 1).

We have seen similar significant advances in other areas of biomedical research, but probably most impressively in the field of cancer research. The adoption of organoids, alternative 3D cultures, and organ-on-a-chip systems have rapidly accelerated in cancer research, perhaps more than in any other disease-specific research space (Shroyer, 2016; Devarasetty et al., 2018; Gr et al., 2019). In vitro cancer MPS occupy spaces in a wide range of cellular, physical, biomaterial, and device-based complexities. Simply moving a common cancer cell line from a 2D tissue culture plastic to a 3D ECM hydrogel matrix can do wonders. A cell line thought to be somewhat artificial can take on highly relevant disease-specific characteristics and serve as a useful research tool (Aleman and Skardal, 2018; Dedhia et al., 2023; Nairon and Skardal, 2023). This is particularly true in rare cancers where patient-derived biospecimen availability for research might be limited.

Utilization of patient-derived cell populations—inherently heterogeneous and more complex than cell lines—can further enhance MPS, in cancer and elsewhere (Vlachogiannis et al., 2018; Ooft et al., 2019). In cancer, the heterogeneity of patient-derived tumor samples can be embraced. The tumor microenvironment (TME) contains many types of cells that are not derived from the tumor itself. These include fibroblasts, immune cells, and many others (Votanopoulos et al., 2019a; Li et al., 2020). As a result, heterogeneous biospecimens from tumors and other tissues are a gift to researchers. They enable the generation of organoids and organ-on-a-chip systems where one can now query biological phenomena such as inflammation and fibrosis (Nairon et al., 2022), and how changes from these phenomena influence disease progression, such as metastasis in the case of cancer (Devarasetty et al., 2017; Devarasetty et al., 2020; Dominijanni et al., 2020). One challenging aspect of in vitro tumor models has been integrating immune cells. This has traditionally been difficult, as many of the immune cells would die off, while tumor cells and stromal cells would survive in cell culture conditions of years past. However, more advanced ECM hydrogel systems, new tissue culture media formulations, and increased cellular heterogeneity have provided increased support for various immune cell types. For example, our hyaluronic acid hydrogel platform has successfully supported T cell maintenance and function within melanoma, sarcoma, Merkel cell carcinoma, and appendiceal cancer patient-derived tumor organoids (Votanopoulos et al., 2019a; Votanopoulos et al., 2019b; Forsythe et al., 2022a; Forsythe et al., 2022b). With resident immune cells, immunotherapies such as immune checkpoint blockade therapies can be explored (Votanopoulos et al., 2019a; Forsythe et al., 2022a; Forsythe et al., 2022b). Likewise, cellular immunotherapies have been assessed using such in vitro model systems (Dominijanni et al., 2020; Yuki et al., 2020). Both chemotherapies and immunotherapies therapies can be tested in such systems, the results of which can potentially be used in the clinic to inform patient-specific therapeutic decisions, something that we have been fortunate to have performed (Votanopoulos et al., 2019a). Efforts are underway at numerous institutions and medical centers to integrate MPS-based diagnostics with clinical decision making.

Ultimately, we can also use these platforms to combine multiple organoids or 3D cultures into serial or parallel recirculating fluid circuits, enabling crosstalk by secreted factors between 3D constructs, or even cellular transport through circulation to different tissue types (Esch et al., 2014; Esch et al., 2016; Skardal et al., 2017; Aleman and Skardal, 2018; Skardal et al., 2020). These multi-tissue MPS have been relatively limited compared to single tissue type MPS, but are rapidly advancing. To date, there have been several interesting studies demonstrating the potential of these technologies. A clever “body-in-a-cube” system comprised of stacked planar tissue chips supported GI tract, liver, bone marrow, and kidney cell lines, with viability percentages ranging from 80% to 95% (Chen et al., 2020). Ronaldson-Bouchard et al. recently demonstrated the development of a liver, heart, bone, and skin body-on-a-chip connected by vascular flow. Importantly, this system moved away from cell lines, instead using a mix of induced pluripotent stem cell (iPSC)-derived cells and primary cells for the most part. This platform remained viable with appropriate tissue-specific phenotypes for 4 weeks and was used to evaluate pharmacokinetic and pharmacodynamic profiles of doxorubicin (Ronaldson-Bouchard et al., 2022). Our lab was fortunate to be part of a team of researchers that first produced a three-tissue liver, lung, and cardiac platform, in which we showed several examples of crosstalk between tissue types following the introduction of several drug compounds (Skardal et al., 2017). Later, we expanded to a six-tissue “body-on-a-chip” platform housing liver, lung, cardiac, vascular, testis, and brain organoids and tissue constructs. This represented a significant increase in biological complexity, which we initially anticipated to be a significant hurdle for keeping each tissue type viable in the integrated system. Interestingly, despite being comprised of over 15 distinct iPSC-derived or primary cell types, a relatively simple tissue culture media could be used. After some inflammatory biomarkers were observed in the initial days of culture, over time the system self-regulated itself through conditioning of the media, and the viability and functionality remained high for 4 weeks (Skardal et al., 2020). These multi-MPS highlight the advancements in complexity that have been made in the field of organ-on-a-chip technology.

Microphysiological systems are poised to make a significant contribution to how medicine and biomedical research are performed. In the United States, the FDA decision to remove the absolute requirement for animal model testing in the drug development pipeline serves as a loud signal that the world is open to new technologies and model systems to evaluate the toxicity and efficacy of therapeutics in development. Already, laboratories throughout the world have shown that, in organoids and organ-on-a-chip platforms, we can recapitulate toxicity outcomes observed caused by drug compounds that were removed from the market due to harming or killing human patients, as well as toxicity associated with environmental toxins (Forsythe et al., 2018; Skardal et al., 2020).

Personalized medicine is a clinical area where MPS can make a significant impact. For those of us actively working in this field, this is not a new concept, but for most, it is. We can readily see a clinical workflow where we utilize patient-derived cells to generate MPS specific to a patient. These patient-specific MPS will then be used to test multiple therapeutics to generate empirical data to inform clinicians on which therapeutic should be prescribed to achieve the most effective result with the least amount of side effects. We have already done this in select cases (Votanopoulos et al., 2019a), but there is much red tape to work through to make this standard practice at medical centers throughout the world.

Nevertheless, while organoid, organ-on-a-chip, and similar technologies have hurdles to overcome before universal acceptance and utilization, those of us working in-depth with these technologies stand poised to make a major impact in both biomedical research and clinical enterprises.