MIDD Is Changing the Way Preclinical Evidence Is Used
Developing a new drug is a long and uncertain process. Before a candidate enters clinical development, researchers must make decisions based on pharmacology, pharmacokinetics, biomarkers, toxicology, disease biology, and an evolving understanding of how the drug may behave in humans.
For many years, these decisions relied heavily on animal studies, including studies in rodents and nonhuman primates. Such studies remain important, but their results cannot simply be transferred from one species to another. Differences in metabolism, target expression, immune response, disease mechanisms, and drug exposure may all affect clinical translation.
This is the challenge that Model-Informed Drug Development, or MIDD, seeks to address. MIDD is not a single modeling technique, nor does it mean replacing experiments with computer simulations. It is a development approach that brings together different types of evidence and uses them to support practical decisions throughout the drug development process.
The central question is not how much data a program has generated, but whether the available evidence can help the development team make a better decision about dose, study design, patient selection, efficacy, or risk. A program can hold thousands of data points and still lack the one piece of evidence that would settle a dose question, which is why the framing of the question comes before the choice of method.
What Does MIDD Mean in Practice?
MIDD uses quantitative models and scientific knowledge to connect findings from different stages of drug development. Depending on the program, this may include pharmacokinetic and pharmacodynamic data, in vitro potency, biomarker results, disease models, historical clinical experience, and clinical or real-world data.
The models themselves may range from relatively simple exposure-response analyses to more detailed physiologically based pharmacokinetic or quantitative systems pharmacology models. The appropriate method depends on the question being asked. An exposure-response analysis describes how a measured effect changes with drug concentration or dose. A physiologically based pharmacokinetic model represents organs and tissues as connected compartments with their own blood flows and permeability, which is what allows it to be used when a program needs to reason about a species or a population that was never studied directly. Quantitative systems pharmacology models go further and represent the biology of a pathway or a disease process itself.
For example, a development team may need to estimate an active exposure range, select a starting dose for a first-in-human study, understand the relationship between target engagement and pharmacological response, or determine whether a biomarker is suitable for clinical use. These questions require different kinds of evidence and different levels of modeling.
MIDD therefore should not be judged by the complexity of its formulas. Its value lies in whether it improves a real development decision and makes the reasoning behind that decision more transparent.
The Role of NHP Efficacy Models in an Integrated Evidence Strategy
The growing interest in MIDD, organoids, organ-on-chip systems, and artificial intelligence has sometimes led to the impression that animal studies are no longer needed. This is an overly simple interpretation.
MIDD does not prescribe the use or removal of any particular experimental method. Instead, it asks which evidence is relevant to a defined scientific or development question. In some programs, a nonhuman primate efficacy model may provide information that is difficult to obtain from other systems.
This can be particularly relevant to biologics, immune-mediated diseases, central nervous system programs, cell and gene therapies, and drug candidates directed at targets with meaningful similarities between humans and nonhuman primates. In these settings, an NHP study may contribute to the understanding of target engagement, pharmacological activity, biomarker changes, exposure-response relationships, or the consistency of a mechanism across species. The reason is partly structural: antibody binding, Fc-mediated effector function, and the anatomy of the central nervous system are all areas where rodent biology departs from human biology in ways that a two-dimensional cell system cannot reproduce.
The purpose is not to treat NHP data as a reference that cannot be questioned. Like all preclinical evidence, NHP findings have strengths and limitations. Their contribution is greatest when the model is biologically relevant, the endpoints are well selected, and the study is designed around a clear translational objective.
In an MIDD framework, an NHP efficacy study is therefore one important node in a broader evidence network. Its results can be interpreted together with in vitro data, rodent studies, Pharmacokinetics (PK) and Pharmacodynamics (PD) analyses, biomarker findings, mechanistic understanding, and, when available, human data.
From NHP Findings to Clinical Hypotheses
The usefulness of an efficacy study depends on more than whether a treatment produces a measurable effect. A translationally valuable study should also help explain the conditions under which the effect occurs and how the findings may inform clinical development.
An NHP study can, for example, help characterize the relationship between dose, exposure, target engagement, biomarker response, and pharmacological effect. When these data are collected alongside appropriate PK and PD measurements, they may support exposure-response analyses and help distinguish a genuinely active exposure from a dose that produces an effect only under limited experimental conditions. The distinction matters at the point of dose selection, because an effect seen at a single high exposure tells a team very little about the exposure range that should be carried into patients.
The study may also help clarify the biological mechanism. Does the observed effect correspond to a relevant biomarker? Is the biomarker linked to target engagement or merely associated with a downstream change? Is the response consistent across animals and related to drug exposure? These questions are important when preclinical findings are being used to develop a clinical hypothesis.
None of this implies that an NHP result will perfectly predict the outcome in humans. Translational work always involves uncertainty. The objective is to describe the evidence carefully, identify the uncertainties, and use the combined evidence to reduce the likelihood of an inappropriate clinical decision.
Why Study Design Matters for MIDD
The quality of a model-informed strategy depends on the quality and relevance of the data used to build it. A study may be technically well conducted and still provide limited value for clinical translation if its endpoints, sampling schedule, or disease model are not aligned with the decision the sponsor needs to make.
This is why the development question should be defined before the study design is finalized. A study intended to demonstrate proof of mechanism may require a different model and different endpoints from a study intended to support dose selection or biomarker translation. In the first case the readout has to show that the drug reaches and acts on the target; in the second, the readout has to describe how response scales with exposure across a range of doses.
For NHP efficacy research, clinically relevant pharmacodynamic endpoints and biomarkers can be particularly valuable. PK and PD sampling should be considered together so that the relationship between exposure and effect can be evaluated. Where appropriate, the study design should also allow investigators to understand the timing of response, the durability of the effect, and the relationship between pharmacological activity and tolerability.
The resulting data should be sufficiently well characterized to support subsequent analysis and integration. Clear endpoint definitions, traceable methods, appropriate controls, and consistent documentation are not merely administrative details; they determine how confidently the findings can be used in a broader translational package.
Choosing the Right Level of Modeling
A common misconception is that a more sophisticated model will necessarily produce a better development decision. In practice, model complexity should reflect the question, the available data, and the consequences of being wrong.
A relatively simple exposure-response analysis may be appropriate for early candidate comparison. A more detailed PBPK or mechanistic PK-PD model may be justified when the team is considering first-in-human dose selection, multiple sources of variability, or important cross-species differences.
The same principle applies to data generated in NHP efficacy studies. Modeling should clarify the relationship between the study findings and the next development decision. It should not be added simply to make a report appear more advanced.
A fit-for-purpose approach may involve descriptive analysis, exposure-response modeling, biomarker analysis, cross-species comparison, or scenario analysis. The right combination depends on the drug, the disease model, the development stage, and the question that the sponsor needs to answer.
How a Specialized Preclinical CRO Can Contribute
MIDD is often discussed as a modeling discipline, but successful implementation also depends on how experimental studies are planned and conducted. Pharmacology, PK, PD, bioanalysis, biomarker development, data analysis, and regulatory strategy need to be connected from the beginning rather than handled as isolated activities.
A specialized NHP efficacy CRO can support this process by helping sponsors translate a development question into a practical study design. This may include selecting a relevant species and disease model, defining efficacy and biomarker endpoints, coordinating PK and PD sampling, and interpreting the resulting data in relation to the intended clinical strategy.
The most useful collaboration is not limited to delivering a final study report. It involves understanding how the study will be used, what decision it is expected to support, and which limitations should be considered when the findings are integrated with other evidence. A report that documents what happened, without a view of the decision it was meant to serve, is difficult to reuse later in a modeling exercise.
This is particularly important for programs in which an NHP efficacy study is expected to contribute to a broader translational package rather than serve as an isolated demonstration of activity.
The Future of NHP Research in Translational Drug Development
The future of preclinical development is unlikely to be defined by a choice between animal studies and newer technologies. In practice, drug development programs will continue to draw on different systems, each evaluated according to its biological relevance, technical reliability, intended use, and ability to address a specific source of uncertainty.
For some programs, human cell systems, organoids, or computational methods may provide highly informative evidence. For others, an appropriately designed NHP efficacy model may remain valuable because it captures aspects of pharmacology, physiology, or target biology that are not adequately represented elsewhere.
The important issue is how the evidence is connected. A useful NHP study should help the development team understand not only whether an effect was observed, but also why it occurred, what exposure was associated with it, which biomarkers support the interpretation, and how the findings may inform clinical research.
This is the direction in which MIDD is taking preclinical development: away from isolated results and toward a more integrated, transparent, and decision-oriented interpretation of evidence.
Conclusion
MIDD represents a shift from relying on individual experimental results to integrating multiple sources of evidence in support of drug development decisions.
This shift does not make NHP efficacy studies irrelevant. It places greater emphasis on their scientific purpose, biological relevance, study design, and connection to clinical questions. When appropriately selected and conducted, NHP models can contribute valuable information about mechanism, pharmacology, biomarkers, exposure-response relationships, and translational potential.
For drug developers, the objective is not to choose between experimental studies and computational models. It is to define the development question first and then assemble an evidence package in which each method is used for the purpose it can support most reliably.
As a specialized preclinical CRO, Prisys Biotech supports NHP efficacy studies designed around translational objectives and fit-for-purpose development questions. By integrating study design, pharmacology, PK/PD assessment, biomarkers, and data interpretation, Prisys Biotech helps sponsors generate evidence that can contribute to a more informed MIDD strategy.
