The Emerging Value of Microphysiological Systems and New Approach Methodologies

Why most drug candidates never reach patients

The pharmaceutical and biotechnology industries face a persistent challenge: despite significant advances in molecular biology, genomics, and drug discovery, most therapeutic candidates continue to fail during clinical development. The main reason is straightforward: traditional preclinical models, especially animal studies and conventional two-dimensional cell cultures, struggle to predict how humans will actually respond. This translational gap has created an urgent need for more human-relevant models capable of improving the prediction of safety, efficacy, and patient outcomes earlier in the development process.^1-3

New Approach Methodologies: a new paradigm

New Approach Methodologies (NAMs) have emerged as a transformative framework to address this challenge. NAMs encompass a broad suite of technologies including three-dimensional organoids, microphysiological systems (MPS), organ-on-chip platforms, advanced computational modeling, machine learning, and artificial intelligence. Together, these approaches seek to create more predictive, mechanistically informative, and human-centered models of biology while reducing reliance on animal testing. The growing interest in NAMs is being reinforced by regulatory agencies worldwide, including the U.S. Food and Drug Administration (FDA) and the National Institutes of Health (NIH), which have increasingly recognized the potential of human-based technologies to complement and, in specific applications, replace traditional animal studies.

Microphysiological systems: the leading edge

At the forefront of this evolution are microphysiological systems, which combine advances in tissue engineering, stem cell biology, biomaterials, and microfluidics to recreate key structural and functional features of human organs in vitro. Unlike conventional cell culture systems, MPS platforms can incorporate multiple cell types, physiologically relevant tissue architecture, fluid flow, biomechanical forces, and, in some cases, inter-organ communication. These capabilities enable researchers to capture aspects of human physiology and disease that are difficult or impossible to model using traditional approaches. For toxicology and drug development applications, MPS technologies offer the potential to evaluate dose-dependent responses, characterize mechanisms of toxicity, assess long-term tissue function, and investigate complex biological interactions in ways that more closely reflect human biology.^1-3

More than a replacement for animal models

But the value of MPS goes well beyond replacing animal models. These systems provide an opportunity to generate mechanistic insights, identify biomarkers, explore patient-specific responses, and improve decision-making throughout the drug development pipeline. Human-derived organoids and organ-on-chip systems can be linked together to model organ-organ interactions, enabling the study of absorption, metabolism, distribution, and toxicity within integrated human-relevant platforms. When combined with computational modeling and AI-enabled analytics, these data-rich systems have the potential to substantially improve prediction accuracy while reducing development timelines and costs.

While challenges remain regarding standardization, scalability, validation, and regulatory acceptance, the field is increasingly converging on integrated testing strategies that leverage the complementary strengths of NAMs and traditional approaches. Rather than representing a single technology, NAMs constitute a new paradigm for biomedical research and translational science, one that prioritizes human biology, mechanistic understanding, and predictive performance. As these technologies continue to mature, microphysiological systems are expected to play a central role in modernizing drug discovery, safety assessment, and precision medicine, ultimately fostering a more efficient, ethical, and patient-centered future for biomedical innovation.

ATCC's work: from access to application

The landscape of nonclinical research is rapidly evolving. As the scientific community continues to embrace NAMs, researchers are increasingly turning to advanced in vitro models that provide more human-relevant data while reducing reliance on animal studies. ATCC's ongoing work developing, authenticating, and characterizing three-dimensional (3-D) cell cultures, patient-derived organoids, and microphysiological systems (MPS) is helping transform toxicology, disease modeling, and drug discovery by enabling researchers to interrogate human biology in more clinically relevant contexts.

The real barriers to adopting 3-D biology

As the ecosystem of advanced in vitro platforms expands from patient-derived organoids and spheroids to perfused organ-on-chip systems, the central challenge has shifted from simply accessing advanced models to determining how these models should be produced, qualified, selected, and applied. Several recurring barriers still limit broader adoption of 3-D biology, including heterogeneous starting materials, poorly characterized cell sources, variable media and culture conditions, batch-to-batch variability, limited cross-lab reproducibility, and the absence of universal reference standards. These issues are further complicated by inconsistent quality control metrics, non-standardized protocols, limited shared metadata frameworks, and evolving expectations for regulatory qualification of organoids and MPS as NAMs.

Infrastructure matters as much as biology

The key message: advancing these platforms takes more than sophisticated biology; it takes enabling infrastructure. The importance of standardized production workflows that span model acquisition, initial characterization, R&D optimization, model traceability, biobanking and preservation, accessibility, and end-user support. This infrastructure supports the development of reproducible protocols, lot-level records, defined acceptance criteria, cryopreservation and post-thaw quality control, transparent media and reagent documentation, and training resources that help researchers implement models consistently across laboratories.

A real-world example: the HCMI portfolio

The Human Cancer Models Initiative (HCMI) portfolio provides a practical example of this approach. Through ATCC’s long-standing collaboration with NIH and NCI, the HCMI collection now includes over 800 patient-derived cancer models, with more than 75% represented as 3-D models across 28 tissue sites. These clinically relevant models, supported by associated molecular and clinical data resources, illustrate how authenticated materials, traceable workflows, searchable catalogs, and open data access can help move advanced models from specialized research settings toward broader translational use.

Five questions researchers keep asking

Together, these themes underscore a central point: the value of MPS and related advanced models depends not only on biological complexity, but also on reproducibility, characterization, accessibility, and fit-for-purpose use. As a result, several practical questions consistently emerge for researchers, product developers, and regulatory stakeholders:

1. How do you select the right MPS platform for your application?
   It depends on your biological question, the endpoints you need, throughput, and how far along you are in development. Early discovery may benefit from scalable spheroids or organoids, while mechanistic toxicology, ADME, or chronic exposure studies may require perfused MPS or organ-on-chip systems that better capture flow, tissue architecture, and dynamic physiology. It is not unusual for a single platform to be unable to fully answer the question on its own, in which case additional technologies may be required; this makes harmonization of biologicals and protocols across platforms essential so that complementary datasets can be paired to support a single conclusion.
2. Are methods across MPS platforms standardized?
   Standardization is improving but remains application- and platform-dependent. Confidence in MPS data requires clear quality controls, defined acceptance criteria, reproducible protocols, benchmark compounds, validated endpoints, and transparent reporting. Because MPS systems vary in design and complexity, fit-for-purpose standardization is often more realistic than one universal protocol.
3. How do we justify the cost of MPS systems?
   Weigh the value of MPS against the quality and decision-impact of the data it generates. Although advanced systems can be more expensive and lower throughput than conventional animal or 2-D models, they may reduce downstream risk by improving human relevance, identifying liabilities earlier, supporting better candidate selection, and decreasing reliance on less predictive models.
4. What regulatory expectations should be considered when using MPS data?
   Regulatory acceptance depends on whether the MPS method is scientifically justified for its intended use. Key considerations include context of use, human biological relevance, technical characterization, reproducibility, assay performance, and transparent documentation. MPS data are most powerful when they address a defined decision point and are supported by appropriate validation or qualification evidence.
5. How do you align MPS selection with the intended context of use?
   Context of use should be defined before platform selection. Researchers should clarify what decision the model is meant to support, what biological features must be represented, what level of confidence is required, and how the resulting data will be interpreted. The best MPS platform is not necessarily the most complex system; it is the one that is fit for purpose and capable of answering the specific question with sufficient reliability.

A practical framework for platform selection

A practical selection framework should begin with the intended context of use and then weigh biological relevance, throughput, cost, technical complexity, endpoint requirements, and the level of evidence needed for decision-making. ^4-6

Recommendation on building an integrated MPS strategy

The future of drug discovery is unlikely to rely on a single model system. Instead, researchers are increasingly adopting a tiered strategy:

This tiered framework emphasizes that each model class contributes different value to the development pipeline. Simpler systems enable rapid screening and prioritization, while progressively complex platforms provide deeper mechanistic insight, stronger human relevance, and greater confidence for later-stage decisions. This approach balances throughput, cost, and physiological relevance while generating increasingly predictive datasets throughout development. 

ATCC's recent work consistently reinforced this concept: advanced models should be viewed as complementary tools rather than competing technologies. The goal is selecting the right model for the right question at the right stage of research.

Conclusion

As toxicology and drug development move toward human-relevant testing strategies, researchers now have access to a rich ecosystem of advanced model platforms spanning conventional 2-D systems, spheroids, patient-derived organoids, microphysiological systems, and organ-on-chip technologies. Each platform occupies a valuable position along the continuum of biological complexity, throughput, cost, and translational relevance. Rather than viewing these systems as competing technologies, the more practical approach is to deploy them as complementary tools within an integrated testing strategy. 

The challenge is no longer simply finding an advanced model; it is selecting and qualifying the model that most effectively answers the question at hand. This requires a clear context of use, defined performance expectations, appropriate biological relevance, reproducible methods, and transparent documentation of quality controls and acceptance criteria. As reflected in current regulatory discussions around NAMs, confidence in these systems depends not only on technical sophistication, but also on fit-for-purpose validation, assay robustness, and the ability to generate interpretable data that support specific scientific or regulatory decisions. 

Broad adoption of advanced models will depend on the infrastructure surrounding the biology. Standardized inputs, authenticated patient-derived materials, reproducible production workflows, traceable lot records, post-thaw quality criteria, accessible protocols, shared metadata, and end-user support are essential for moving 3-D and MPS technologies from specialized research environments into scalable translational applications. The HCMI portfolio illustrates how centralized infrastructure, clinically annotated models, open data resources, and distribution-ready materials can help democratize access to high-quality patient-derived systems while supporting reproducibility across laboratories.

Ultimately, the future of preclinical research will be shaped by thoughtful model integration rather than reliance on any single platform. A tiered strategy allows researchers to begin with scalable systems for broad screening, then progress to spheroids, organoids, MPS, or organ-on-chip models when greater biological complexity, functional endpoints, or mechanistic insight are needed. By matching experimental goals with the appropriate level of model complexity, scientists can generate more predictive datasets, reduce development risk, improve candidate selection, and accelerate the transition toward a more efficient, ethical, and human-centered drug development paradigm. 

Did you know?

ATCC's organoid collection includes 230+ clinically annotated, patient-derived organoid models supported by Organoid Growth Kits and detailed, step-by-step protocols.

Explore now

References

1. Jaeschke H, Ramachandran A. Are New Approach Methodologies (NAMs) the Holy Grail of toxicology? Toxicol Sci 208(1): 1–8, 2025. PubMed: 40795217
2. Mehta K, et al. Modernizing Preclinical Drug Development: The Role of New Approach Methodologies. ACS Pharmacol Transl Sci 8(6): 1513–1525, 2025. PubMed: 40567279
3. Wu X, et al. Reimagining human-centric drug development with new approach methodologies. Science 392(6796): 371–378, 2026. PubMed: 41990130
4. Pamies D, et al. Recommendations on fit-for-purpose criteria to establish a quality management for microphysiological systems (MPS) and for monitoring of their reproducibility. Stem Cell Reports 19(7):1041, 2024. PubMed: 38959889
5. Tomlinso, L, et al. Considerations from an International Regulatory and Pharmaceutical Industry (IQ MPS Affiliate) Workshop on the Standardization of Complex In Vitro Models in Drug Development. Adv Biol (Weinh) 8(8): e2300131, 2024. PubMed: 37814378
6. Reyes Hernandez D, et al. From animal testing to in vitro systems: advancing standardization in microphysiological systems. Lab Chip 24(5): 1076-1087, 2024. PubMed: 38372151