Organ-on-Chip Technologies in 2026 TissUse in the Spotlight: NIH, FDA and GAO Push and the Formation of Vendor Alliances

Download PDF here:

SL 011_HMBC_TissUse MPS in 2026_August2026.pdf

Download PDF • 2.53MB

An Interview between Ganakshi Gehlot, Vignesh Viswanathan, TissUse, and Henning Mann, PhD, HM.BioConsulting

TissUse Authors: Ganakshi Gehlot, Vignesh Viswanathan

HM.BioConsulting Authors: Desiree Goubert, Henning Mann

Over the years of Assay Development in the MPS field, we have seen fluctuating interest in combining organs in in-vitro models: liver and kidney, for example, to reproduce toxic effects. Over the past decade, organ-on-chip has evolved from proof-of-concept microfluidic tissue into robust, application-driven platforms, a journey we've tracked closely in prior HM.BioConsulting Spotlights.

Early single-organ models established human relevance; today the field is moving toward integrated multi-organ systems that reflect systemic biology and real drug behavior. Multi-organ systems can catch toxicity that isolated models miss. A compound may look benign until liver metabolism turns it nephrotoxic. Modern systems now extend further into ADME, exposure sequencing, and mechanistic insight.

That momentum is now also organizational: the recent formation of the Industry Alliance for Microphysiological Systems (IAMPS), a Brussels-based alliance bringing together MPS developers to coordinate with regulators and policymakers, marks a shift from individual platform validation toward field-wide standards. TissUse is a founding and leading member of IAMPS, positioning the company at the center of the effort to translate platform-level performance into an accepted regulatory evidence category, a theme we return to at the close of this conversation.

Henning Mann: From single to multi-organ: where could HUMIMIC outperform single-organ or organoid-only systems for translational decisions?

TissUse: The key advantage of HUMIMIC multi-organ chips is their ability to capture organ-organ communication and systemic effects, which are critical for translational decisions and impossible to see in single-organ models. This matters for predicting ADME, PK, and PD, since it simulates metabolism and systemic toxicity. Our two-organ configuration (Chip2) is a practical entry point for in-vivo researchers moving toward chip-based approaches who aren't yet ready for full multi-organ complexity: it lowers the operational barrier while still generating crosstalk data single-organ systems can't produce.

Henning Mann: You have three main chip designs, a Chip2, Chip3, and Chip4, with distinct design trade-offs. How do you choose the right organ pairings and configuration?

TissUse: Chip2 supports any two-organ combination; Chip3 handles up to three. Organs are selected based on the research question, and the platform integrates barrier models, spheroids, organoids, and hydrogels for flexibility. Each chip type is itself modular through different compartment formats, lid options, and insert compatibilities, so the circuit adapts to the experiment instead of following a fixed template. The real trade-off is simplicity and control versus physiological complexity and systems relevance. To reproduce at scale, our AutoLab platform runs up to 48 replicates  simultaneously with standardized incubation, dosing, and imaging.

Chip4 supports four organ models, is PBPK-compliant and ADME-specific, and fixes kidney, liver, and intestine, with a flexible fourth organ (e.g., BBB, cardiac, or neurospheres). It's built for pharmacokinetics and ADME studies, analyzing metabolites and excretome via LC/MS. Most groups start with Chip2 to explore organ interactions and core PK/PD assumptions, then graduate successful configurations to Chip4 once the added organs and dual-circuit design justify managing the added complexity.

In October 2016, NCATS announced roughly $6 million in FY2016 funding for three Tissue Chip Testing Centers under the Tissue Chip for Drug Screening program, supported by the Cures Acceleration Network. The goals were to enable independent validation of tissue-chip platforms, ensure broad availability for regulators and pharma, and accelerate adoption across the research community.

Nearly a decade later, that vision is taking shape with an entire community of Organ-on-Chip/ MPS/ NAM companies. TissUse is a technology vendor in the Liver Ring Trial, a cross-pharma evaluation of six liver MPS models across six organizations, focused on predicting Drug-Induced Liver Injury (DILI) and intrinsic clearance while building regulatory confidence. Momentum is reinforced by participation from big pharma and a recent webinar, ‘The 3Rs in Action: Reducing Reliance on Animals,’ signaling real big-pharma pull.

Henning Mann: How can we enable automation in experiments, provide accelerated and instant assay visibility, and convert that into fast, decision-ready value?

TissUse: Regulatory frameworks increasingly favor predictive, human-relevant models like Multi-Organ-Chips. The remaining challenge is running complex preclinical models reproducibly and at scale, since assay throughput on its own is no longer the constraint. Automation reduces manual intervention and inter-operator variability while standardizing execution, monitoring, and data collection. The HUMIMIC AutoLab integrates chip handling, environmental control, microscopy, and continuous data acquisition into one workflow.

The goal is to achieve earlier, more confident decisions, with speed as a secondary benefit: longitudinal, high-content datasets under standardized conditions surface efficacy signals, safety liabilities, and therapeutic windows earlier, thus supporting go/no-go calls before costly late-stage failures. Longer term, combining continuous data with AI-supported analysis and digital twins could push preclinical development toward earlier, higher-confidence, more physiologically relevant decisions, focusing resources on the most promising assets while cutting downstream attrition.

The AutoLab has the potential to turn liver MPS from data generator to decision enabler: mechanisms of action become visible while an experiment is still “alive,” well before the usual weeks-later readout. While this speed is incredibly helpful, it does have its limits, as the generated data needs to fit into a context of use. Only standardized assays, readouts, and interpretation let rapid data become trusted data, letting automated liver-chip results scale from internal learning to cross-site comparability and regulatory confidence.

Henning Mann: Given that need for immediate readout and standardization, how do you think about assay classes and KPIs today? Which qualification panels, such as mass balance, barrier integrity, and metabolite tracking, have become table-stakes for safety, DMPK, and efficacy work?

TissUse: Table-stakes metrics like mass balance and metabolite tracking are only as good as the biological signal behind them, which is why we use a scaffold-independent design with high-biomass integration: 100–200 spheroids per well. A single-spheroid setup often has poor signal-to-noise for metabolite tracking; boosting cell numbers and metabolic turnover keeps LC-MS detection robust enough for pharmaceutical decisions. These parameters still need to be read within the context of systemic multi-organ circulation and controlled microfluidic perfusion, since continuous flow supplies the shear stress and nutrient gradients that keep biomass functional and able to produce the reactive metabolites needed to capture human DILI risk.

 The chip architecture remains constant across applications, with interchangeable scaffolds allowing the integration of diverse tissue formats, including hydrogels, organoids, and barrier models, to meet the requirements of each specific use case. Moreover, decision-readiness needs more than one endpoint: because our systems generate substantial biological material, we combine viability with CYP3A4 activity, high-depth omics (scRNAseq), and morphology in the same experiment. These data packages support regulatory-aligned frameworks like the FDA ISTAND Qualification Program, which we're part of.

Multi-organ chips exist because single-organ readouts would miss the systemic story: first-pass metabolism, circulating metabolites, secondary organ injury, and organ-organ feedback. Once tissues are connected, physiological scaling and flow directionality (who's upstream vs. downstream, residence time, dilution, recirculation) become the experiment.

A clean example is the human liver→kidney MPS from Ed Kelly's group (Chang et al., 2017). Aristolochic acid looked “quiet” in isolated liver or kidney settings, but placing the kidney downstream of the liver revealed a liver-generated metabolite that was strongly nephrotoxic, showing that directionality can determine whether the real mechanism is visible at all. In practice, organ selection is pragmatic: pair organs that reflect ADME and known liabilities, such as gut-liver (oral uptake/first pass), liver-kidney (metabolism/clearance), and increasingly BBB-liver-kidney when CNS exposure and detox matter, and directionality is central to all of it.

Henning Mann: On inter-organ crosstalk: what has TissUse learned about directionality (gut→liver vs. liver→brain), and how does HUMIMIC preserve physiological ratios and flows?

TissUse: For us, directionality is the core biological question behind every design choice. In our Gut–Liver MPS, compounds like APAP are absorbed by the intestinal equivalent and transported via controlled flow to the liver; in overdose scenarios, that physiological order directly produced DILI in the liver compartment, showing that gut→liver directionality determines safety outcomes.

In our Skin–Liver Chip2 study of genistein, topically applied compound underwent immediate first-pass metabolism in the skin equivalent, sharply reducing the bioavailable parent compound and changing the metabolite profile reaching systemic circulation. As a result, the topical dose cleared without triggering the gene-expression changes we saw under direct basolateral exposure. Static, single-tissue cultures would miss this route-dependent effect entirely, and a standard in-vitro assay would misjudge the systemic liver hazard of a topical compound.

For liver–brain crosstalk, using the HUMIMIC Chip4, we focused on how liver metabolism shapes CNS exposure through the blood–brain barrier. The system showed that liver-generated glucuronidated propranolol was effectively restricted by the BBB, while lipophilic propranolol crossed readily and hydrophilic atenolol was largely excluded validating both directional flow and barrier function. Importantly, we also observed evidence of local brain metabolism, indicating that inter-organ communication is not purely passive but involves organ-specific activity layered onto systemic exposure.

Together, these models highlight why preserving physiological ratios, flow paths, and directionality is essential to making multi-organ chips predictive and mechanistically meaningful.

TISSUSE TECHNOLOGY SNAPSHOT

The HUMIMIC Chip4 uses integrated micropumps and channels to create continuous, controlled medium flow that simulates blood flow, connecting organ models in the body's natural sequence. It also controls shear stress to physiologically relevant levels, which is essential for cell health and function, including endothelial cells lining blood vessels.

More info: HUMIMIC-Chip4

With directionality, pairing, and scaling giving multi-organ models mechanistic meaning, regulator readiness and marketability become the next hurdle. Regulatory confidence needs structured artifacts, a defined context of use, and cross-site validation that turn innovation into accepted evidence. Recent FDA modernization efforts and NIH support for NAMs signal institutional momentum, and expectations are tightening: reproducibility, qualification standards, and documentation rigor, while industry users want ready-to-deploy SOPs, validated workflows, and clear acceptance criteria.

Henning Mann: On the regulatory path: what concrete regulatory-facing artifacts (SOPs, acceptance criteria, ring trials) does TissUse supply or recommend, and how do FDA interactions inform them?

TissUse: Our regulatory strategy rests on SOPs, GLP compliance, and robust qualification data aligned with FDA and EMA/ECVAM expectations. The goal is to show our HUMIMIC platforms are fit for purpose in drug development and regulatory submissions.

Henning Mann: Can you tell us more about the Ring Trials you are participating in? It seems like they would be a central part of this approach.

TissUse: The Liver Ring Trial is a critical component of our regulatory strategy. Regulatory bodies will not accept a new methodology based on single-site data, so to move from isolated performance to broadly accepted, decision-ready evidence, inter-laboratory validation is essential.

We are proud to be part of a first-of-its-kind consortium initiated by UCB, bringing together computational modeling experts from ESQlabs and leading pharmaceutical companies, including AstraZeneca, Sanofi, Orion Corporation, Servier, and Boehringer Ingelheim. Guided by input from the European Medicines Agency and EURL ECVAM, the trial rigorously evaluates the reproducibility and predictive performance of our HUMIMIC Liver MPS across independent sites.

Specifically, we assess its ability to predict drug-induced liver injury (DILI) and intrinsic clearance. In parallel, integrating ESQlabs' computational models with our MPS platforms strengthens the link to clinical translation. The resulting Ring Trial Report is designed to serve as a central regulatory-grade evidence package.

The same chip setup is also being evaluated within the FDA's Innovative Science and Technology Approaches for New Drugs (ISTAND) Pilot Program, where the generated evidence contributes to the regulatory assessment of novel drug development tools.

Together, the ring trial and ISTAND activities aim to establish the platform's credibility, reproducibility, and translational relevance, supporting its broader acceptance in preclinical drug development.TissUse's platform has also been independently evaluated by the National Institutes for Food and Drug Control (NIFDC) of China, in a liver-kidney proximal tubule model with repeated-dose multi-drug toxicity testing. This serves as an indicator of growing international regulatory interest beyond the US and EU.

With regulatory structure in place, another adoption threshold emerges: complexity versus usability. As multi-organ systems scale, added organs and deeper analytics raise biological realism but also cost, variability, and timelines. The strategic question centers on how complex the model needs to be for the decision at hand.

Henning Mann: On throughput vs. complexity: where's the sweet spot between organ count, assay depth, and timeline, and when do you deliberately keep models simpler?

TissUse: It depends entirely on the question. Early toxicity screening calls for single-organ, high-throughput setups; ADME profiling needs two-to-four organ complexity at medium throughput. Higher complexity generally means lower throughput, and that trade-off is intentional. A model should only be as complex as the question requires, since added complexity without clear rationale raises cost, time, and variability without improving the answer. The field's baseline keeps shifting, though: two-organ chips were emerging a decade ago, four-organ configurations were considered hard, and ten-organ concepts looked impractical. What looks premature today can become routine as new questions drive new designs.

As complexity rises, the next question is where it creates the most immediate value. Most of the conversation so far has focused on healthy tissue models, but many research programs are disease-based, and not every disease area needs the same organ interplay, assay depth, or systemic integration. Some are closer to practical adoption than others.

Henning Mann: Starting with healthy models was the logical entry point. Looking at the applications roadmap now, which disease areas (respiratory, dermatology, metabolic) look most ready?

TissUse: We started with healthy organ models and have expanded into disease models. PDX and cancer models are available, developed with partners including the University of Greifswald, and our current emphasis is on metabolic-pathway models. The initial focus on healthy models gave us a straightforward entry point; now we're centered on increasing biological complexity and predictive power to better support preclinical decisions.

Henning Mann: How do you promote adoption? What does a 90-day migration plan from legacy 2D/organoid assays to HUMIMIC look like for a pharma program manager?

TissUse: The approach is data-driven throughout. 2D models have delivered for decades but have real limits; organoids add complexity but, without shear stress, crosstalk, or dynamic flow, still can't address certain mechanistic or translational questions. When those aspects become critical, HUMIMIC is the logical next step. We frame the pitch around cost, data quality, and translational relevance, benchmarked against the client's existing datasets and workflows.

A typical 90-day plan follows a stepwise de-risking approach: a proof-of-concept phase transfers the most relevant legacy model into HUMIMIC, with an early milestone of establishing a shared media environment compatible with both systems, which is often critical for integration and biological stability, followed by iterative optimization and benchmarking against existing 2D, organoid clinical data to demonstrate added physiological relevance and predictive value.

Henning Mann: What about the legacy of established animal models? A lot of comparisons still get made to the animal model, “closest to human that we have,” although this is increasingly debunked, and I've heard requests, from my time at Nortis, to have an animal model available to compare the new human models to.

TissUse: From TissUse's perspective, this comparison sits outside the central focus of our development strategy. Our R&D efforts are primarily directed toward human tissue-based models, because the core objective is to address biological and pharmacological questions in human-relevant systems.

That said, we operate within the framework of the 3Rs principle. Our aim is to integrate scientifically robust, human-relevant models that add value alongside existing approaches. In some cases, data generated from in vivo animal studies can provide important biological context and reference points to support model evaluation, interpretation of biological responses, and confidence-building for new testing strategies.

However, these comparisons are not intended as direct one-to-one replacements, as each model system provides different biological insights and has its own limitations. Overall, this represents a context-dependent shift toward improved predictivity where feasible, applied selectively and on a case-by-case basis.

For example, in our bone marrow-based models, there is evidence that clinically relevant safety thresholds, such as the lowest tolerable dosing ranges observed in Phase I and II trials, can be estimated earlier and more accurately using predictive New Approach Methodologies than with certain animal models.

More broadly, the strength of human-based systems lies in improving the translational link to clinical outcomes, beyond simply reproducing the historical benchmark of animal studies. The objective is to reduce uncertainty ahead of first-in-human studies by generating mechanistically relevant, human-specific evidence earlier in the development pipeline.

We've moved from foundational questions to regulatory artifacts, adoption playbooks, and the operational realities of replacing legacy 2D, organoid, and animal-based systems. We appreciate the insights TissUse shared. NAM/organ-on-chip technologies are now structured, data-driven propositions competing on cost, predictive value, and decision impact.

Following the recent formation of the Industry Alliance for Microphysiological Systems (IAMPS), a Brussels-based alliance bringing together TissUse, MIMETAS, InSphero, AlveoliX, and others, the conversation is scaling further into policy and regulatory coordination across Europe and the US. IAMPS positions itself at the interface of industry, regulators, and policymakers, accelerating innovation while shaping science-based regulation, which raises a final question:

Henning Mann: With IAMPS now coordinating industry voices and regulatory dialogue, how do you see collective action accelerating regulatory acceptance and broader adoption of MPS across Europe and beyond? And is there anything else you'd like to add?

TissUse: Collective action matters because regulatory qualification of a technology class isn't something one company can drive alone. IAMPS coordinates MPS developers on the standards and evidence frameworks regulators need before they can consistently accept MPS data. It's slow, often invisible work that converts individual platform performance into an accepted evidence category.

Companies still have to qualify their own platforms through formal regulatory pathways, but that success depends on a shared, aligned foundation. That's why consortium efforts like the EMA-guided Liver Ring Trial matter: they show individual platforms work and demonstrate that the field can align on standardized procedures and reproduce results across independent labs.

Regulators aren't waiting for a perfect chip; they're asking whether the data is trustworthy, the context of use is clearly defined, and results are reproducible across sites. Working through IAMPS and large-scale ring trials lets the field answer those questions in a structured, evidence-driven way instead of through isolated examples.

Conclusion

Thank you for taking the time for this conversation. We see, and are excited to do so, that what began as individual proof-of-concept systems has evolved into a coordinated, industry-level initiative.

Hopefully, structured validation, growing regulatory dialogue, and establishing alliances like IAMPS will align innovation with policy, and help moving microphysiological systems from promising alternatives to strategic infrastructure. The next chapter will be defined not just by better models but by collective action translating science into standard practice and without any doubt, it will be very exciting!