M15 General Principles for Model-Informed Drug Development – What Does This Mean?
On 06/03/2026 the Food and Drug Administration (FDA) announced the final guidance “M15 General Principles for Model-Informed Drug Development.” This guidance explains how companies can use computer-based models, simulations, and quantitative evidence to support drug development decisions. Instead of relying on traditional studies, manufacturers may use well- planned models to help answer questions [1] . The guidance was developed through the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) [2] , which means regulators are aiming for a more consistent global approach. For manufacturers, models can be useful regulatory evidence, but only when the question is clear, the model is fit-for-purpose, and the uncertainty is openly explained.
Today’s focus will be for combination drug-HCT/P (Human Cell, Tissue, and Cellular and Tissue Based Product) manufacturers. What does this mean? Development is complex because the product may involve drug performance, tissue or cell characteristics, donor-related controls, manufacturing variability, storage conditions, and patient-specific factors. For these products, model-informed drug development (MIDD) can help manufacturers organize complicated evidence. A model might help connect release attributes to clinical performance, explore how the drug and tissue components behave together, predict how changes in manufacturing could affect safety or effectiveness, or support dose and regimen decisions when traditional clinical data are limited or difficult to collect.
Let’s look at the M15 idea in plain language:
Ask the right question.
Start with a specific “question of interest” such as whether a model can support dose selection, a manufacturing change, comparability, or a risk control strategy.
Define the context of use.
Be clear about exactly how the model will be used. A model used for internal learning may not need the same level of evidence as a model used to support a regulatory decision.
Check whether the model is fit-for-purpose (validation and applicability assessment).
Show that the data, assumptions, methods, and performance checks are appropriate for the decision being made.
Explain the uncertainty.
Do not hide weak spots. Identify where the model is strong, where it is uncertain, and how that uncertainty affects patient risk.
Document the story.
Regulators should be able to follow the logic from question to data, to model, to conclusion, to proposed action.
For combination drug-HCT/P manufacturers, here’s a basic roadmap for consideration:
Start with a decision, don’t build a model first and look for a use later. Define the regulatory or development decision the model is meant to support.
Map the product risks by including the drug component, the HCT/P component, the interaction between them, donor controls, manufacturing steps, storage, handling, and patient-use conditions.
Identify the evidence gap by asking what information is missing and whether modeling is a scientifically reasonable way to address it.
Choose the right model type. Match the model to the question. A simple model that is transparent and well supported may be better than a complex model that is hard to explain.
Use relevant data by combining nonclinical, clinical, manufacturing, release, stability, and real-world information when appropriate, but clearly describe data quality and limitations.
Stress-test assumptions by running sensitivity analyses and scenario testing so reviewers can see what happens when assumptions change.
Plan FDA interaction early. Use meetings and submissions to discuss the question of interest, context of use, model credibility, uncertainty, and proposed role of the model in decision making.
Keep the model alive by updating the model as product knowledge grows, especially after scaling up, process changes, or new safety signals.
What manufacturers should avoid:
Do not treat a model as a shortcut around required studies or controls. Modeling can support evidence, but it doesn’t automatically replace clinical, manufacturing, donor eligibility, or tissue practice requirements.
Don’t ignore the HCT/P component as key tissue related risks could be missed.
Don’t hide uncertainty. FDA expects uncertainty to be described and managed, not buried.
Don’t use overly technical explanations without a clear decision link. The model should help answer a practical question, not simply demonstrate mathematical sophistication.
Don’t wait until the final submission to discuss a high-impact model. Early alignment with FDA can reduce review surprises.
FDA’s M15 guidance is not just about sophisticated modeling, it’s about disciplined thinking. The guidance offers a useful way to connect complex product science with practical regulatory decisions. The strongest applications will be those that define the question clearly, use models that are fit for purpose, acknowledge uncertainty, and tie model results back to patient safety, product quality, and manufacturing control.
Thanks for joining us, keep checking back as we continue following industry trends!
Got questions? Reach out to us here!
References: