The MPS World Summit brings together the scientists, pharmaceutical teams, regulators, and technology developers working to advance the use of human-relevant complex in vitro models (CIVMs) in drug discovery and testing. The tone this year has shifted. Earlier summits were often centred on technology explainers, model development, characterisation studies, and making the case for how microphysiological systems (MPS) could be used in principle. This year, case studies featured more prominently, showing how MPS is being applied within real-world drug discovery and testing programmes. This shift reflects a technology that is moving beyond demonstration and becoming an increasingly important part of workflow modernisation, helping teams increase human relevance and make better decisions earlier in development.
As MPS adoption broadens, the same next-phase challenges came up across talks, posters, satellite sessions, and the conversations between them: improving the usability of complex models, getting results you can reproduce, scaling up without compromising the biology, and producing data that better reflects human clinical outcomes. Across all of this was a steady focus on whether MPS can help teams make better decisions before they reach the clinic.
Automation and high-throughput workflows
One message landed harder than the rest at MPS World SUmmit: complex biology only matters if it can be integrated into existing drug discovery and testing workflows. Several presentations highlighted the obstacles that have held CIVMs back, including reproducibility, maintenance burden, specialist handling, lengthy sample readouts, and limited endpoint generation per experiment.
This is where commercialised, workflow-ready systems can add real value. As our Director of Sales, Joe Parisi, noted, we heard the same feedback from customer after customer: one of the biggest benefits of the PhysioMimix® Core platform is the broad range of endpoints that can be generated from each sample. This is supported by the platform’s ability to generate high volumes of media for sampling at every media change, alongside a large volume of recoverable tissue at the end of the assay.
Patient-derived CIVMs and translational relevance
Patient-derived CIVMs and donor-derived models were the other dominant theme at MPS World Summit. They came up repeatedly in connection with translatability in predicting treatment response, making sense of patient heterogeneity, and getting preclinical data closer to the clinic.
The appeal is that these models maintain features of the tissue they were obtained from, including aspects of disease biology and, in certain cases, epigenetic memory. That matters when you are trying to work out why two patients respond differently to the same drug, or when you are testing drug combinations to find the right strategy for a given patient.
Patient-derived organoids are being used for anticancer drug testing, including in hospital-linked settings, with human-derived 3D cancer models presented alongside genomic, transcriptomic, histopathological, and clinical data. The aim is to build models that capture how one patient differs from the next, rather than ones that flatten everyone into an average.
Presenters returned to questions around how patient-derived models should be compared with matched healthy tissue, what turnaround time is acceptable for treatment-guiding results, and how model outputs can be connected to survival or real therapeutic response. These considerations explain why patient-derived biology is so promising, but also why it requires careful validation before it can support treatment-relevant decisions.
Richer readouts for more complex models
As the biology gets more sophisticated, so do expectations for the data. A theme that came up repeatedly was the case for measuring early and often, and across more than one drug modality type. Cytokine profiling, multiomics, flow cytometry, live-cell fluorescence imaging , ELISA, multiplexed ELISA, and RNA sequencing all featured.
Endpoint assays capture a single moment, so they can miss responses that rise, fall, or change direction over time. Time-resolved and multimodal readouts help show how a response develops, when it peaks, and whether biology is stabilising, adapting, or deteriorating. This is also where the PhysioMimix approach is valuable. Media can be sampled at each media change to track markers such as liver function and cytochrome P450 (CYP) activity over time, while recoverable tissue at the end of the assay supports deeper endpoint analysis, including RNA sequencing. In the PhysioMimix Liver-12 plate, the volume of recoverable tissue also enables each scaffold to be split, for example with one half used for fluorescence imaging and the other for -omics analysis. Looking ahead, real-time sensing technologies represent the next natural step, giving teams an even clearer view of dynamic biological responses as they happen.
Find out more about how the versatility of the PhysioMimix platform for data collection
Hepatic models highlighted the point well. Hepatic identity, albumin expression, albumin-to-alpha-fetoprotein ratios, and CYP3A4 activity were all used to confirm and characterise liver function, with RNA sequencing and bioinformatics used to identify differentially expressed genes and group biological processes. The functional performance of the same cells is enhanced when they are perfused and cultured in a more tissue-like structure, which is why physiological relevance matters. Metabolic capacity and assay sensitivity can be improved, while culture longevity is extended, enabling endpoints that have historically been challenging to determine in vitro, as well as studies of repeat dosing.
For anyone running these systems, the value of a model now depends as much on the data it can produce as on the cells, samples, and architecture it is built from.
For anyone running these systems, the value of a model now rests as much on the quality of the question it can answer as on the volume of data it can produce. More readouts earn their place only when they sharpen a decision: catching risk earlier, building confidence in a candidate, or redirecting a programme before an expensive late-stage failure.
Broadening MPS use with new tissue models and applications
The range of applications on show was striking and encouraging. Cardiotoxicity and blood-brain barrier models were notably prominent, alongside sessions covering nose-to-brain delivery, intestinal absorption, PROTAC permeability, psoriasis, neurotoxicity, intestinal toxicity, vascular injury, lymph node models, and blood-retina barrier models.
Recreating the tissue environment: ECM, mechanics, and multicellular models
A theme running throughout the whole meeting was the push to recreate the tissue environment more accurately, not just the cells. Extracellular matrix (ECM) modeling was prominent, along with multicellular models, vascularization, fluidic flow, barrier function, and mechanical stress.
The underlying question has changed. It is no longer whether cells can be grown in perfused 3D culture, but whether a model captures enough of the tissue environment to generate evidence that supports real decisions. Depending on the application, that context can include cell-cell interactions, matrix composition, fluid flow, mechanical stretch, oxygen gradients, immune cells, and circulating components.
Examples included organoids combined with organ-on-a-chip systems, cyclic stretching, barrier function measurements, and side-by-side comparison of RNA sequencing profiles from patient samples and chip-based models. Intestinal explants and primary intestinal epithelial systems came up in relation to oral absorption, intestinal metabolism, and permeability. Biological relevance is not a feature you switch on, but the result of several connected choices: cell source, tissue architecture, matrix, mechanical input, and how the results are analyzed.
At the same time, adding complexity does not automatically make a model more useful. Greater biological sophistication can strengthen translational relevance, but only when it helps answer a clear question. For these models to be adopted more widely in pharmaceutical workflows, they need to balance biological relevance with practical usability, reproducibility, cost, timelines, and data outputs that scientists can interpret and act on.
Application evidence beats platform novelty
This was reflected across the exhibition floor, posters, and satellite sessions. The examples that landed most strongly were the ones tied to an actual drug discovery program or translational decision. MPS performance specifications on their own did not hold attention in the same way. Collaborations with pharmaceutical companies, academic groups, and cell model providers were used to show how a model answered a specific biological question. As these systems mature, evidence that a model solves a real-world problem counts for more than platform novelty alone.
What Novo Nordisk and Bristol Myers Squibb spoke to in our exhibitor hosted session
Our exhibitor hosted session gave researchers from Novo Nordisk and Bristol Myers Squibb the floor to talk about how our PhysioMimix platform facilitates their drug discovery and testing programs in the real-world. A few points came up more than once.
Ease of use was one of them. Users were getting productive without a specialist standing over every run. Closely tied to that was how the systems fitted into their existing lab processes, so teams could add them to current workflows instead of rebuilding around new equipment. Both teams also spoke highly of the support they got from CN Bio.
The data side carried just as much weight. Researchers were able to obtain several readouts from a single experiment, and this changes the math on running more complex biology, as one experiment can answer more questions. Scale was the other recurring point. The same system that suits small exploratory studies can also be adopted and scaled for higher-throughput screening, so a project does not have to move onto a different platform as it grows.
Related products and services
Contract research services
Discover how to utilize our cross-species models to inform next-step decision making via our DILI in vitro Contract Research Services here.
Add PhysioMimix Core in your lab
To develop your own cross-species Liver MPS models, you will need:
