This Technical Blog was written with input from Endpoint’s Regulatory and Medical Device expert Dr. Sam Gilbert, VMD.

Content has been informed by the presentation and discussion following the recent webinar “Regulatory Strategy for Medical Device Preclinical Programs”.

“A well-written protocol does not guarantee a well-designed study.”

By the time a medical device reaches preclinical evaluation, months, or even years, of work have already shaped the development program. The intended clinical application has been defined, engineering teams have refined the design through bench testing and prototype iterations, and discussions around study models, endpoints, timelines, and regulatory strategy are already underway.

The protocol itself represents only one milestone in a much longer planning process. Regulatory and product goals have begun to take shape, along with the evidence requirements needed to move the program forward. The protocol doesn’t create that strategy; it documents one step within it.

At this stage, the challenge is making sure the proposed study is suited to what the program needs to accomplish next. Some work belongs on the benchtop, while other objectives require cadaveric work, animal studies, or clinical evaluation. Understanding where a study fits within the broader development roadmap can help a team determine whether the timing, model, endpoints, and expected outputs are aligned before the protocol becomes the plan by default.

That evaluation usually comes back to a few fundamental questions: Is the device ready for an in vivo study? What evidence is still needed before moving forward? Does the proposed study support the next regulatory milestone? Are the planned endpoints aligned with the decisions the team will need to make?

Answering those questions early helps clarify the role of each study within the development program and creates a stronger foundation for the decisions that follow.

Moving from benchtop testing into an animal study is one of the most significant transitions in a medical device development program. Engineering performance has often been demonstrated by this point, prototype iterations may be well advanced, and confidence in the device is growing. The momentum to begin in vivo work is understandable.

Before making that transition, however, it is worth asking:

“What are you trying to learn that you can’t learn any other way?”

The answer helps determine whether the program has reached the point where an animal study will provide information that cannot be obtained elsewhere. Biological response, tissue interaction, and other aspects of device performance may require a living system, while questions around deployment, procedural technique, handling characteristics, or user interaction can often be explored more efficiently through additional benchtop, cadaveric, or ex vivo work.

These stages serve different purposes within the development program. Benchtop testing can establish functional performance, while cadaveric and ex vivo work allow teams to refine techniques and better understand how a device will be used. Early in vivo studies can then focus on the biological evidence that requires a living system, with later definitive studies building on that work to support broader development and regulatory decisions.

Resolving what can be learned earlier, before progressing into increasingly complex studies, allows the in vivo work to remain focused on the evidence it is uniquely positioned to generate.

Once a program moves into animal testing, there can be a natural temptation to maximize the value of each study. If the resources are already committed and the study is already being conducted, adding another endpoint or objective can appear to be an efficient way to gather more information at once.

A feasibility study may gradually take on questions about long-term performance, while safety endpoints become combined with effectiveness, usability, biocompatibility, or toxicity assessments. Each addition can be reasonable on its own, but together they can create a study that is expected to accomplish too much and produces evidence that is harder to interpret.

A more effective approach begins with the outputs the study is expected to produce. Once those are defined, the team can determine which endpoints belong in the study, which model is appropriate, and which objectives are better addressed elsewhere in the development program. Separating that work may initially appear less efficient, but a deliberate sequence of focused studies can ultimately provide clearer evidence without asking a single study to answer every remaining question.

Once the purpose of a study is clear, attention can turn to model selection, where discussions often center on identifying the most clinically relevant animal model. Clinical relevance is important, but no model reproduces the clinical environment in every respect. Anatomy and physiology differ between species, recreation of a specific pathology or disease process may not be practical or even possible, and procedural workflows can differ from those ultimately encountered in patients.

The important consideration is which of those differences matter for the evidence a particular study is expected to generate.

“The right model isn’t the one that’s closest to humans. It’s the one that’s best suited to the evidence you’re trying to generate.”

A model selected to evaluate device deployment, for example, may not be the most appropriate choice for assessing long-term tissue response. A feasibility study intended to refine a procedure may likewise have different model requirements than a definitive study safety intended to support a regulatory submission.

Early studies can also provide valuable insight into how physicians actually interact with a device. Delivery techniques evolve with use, procedural workflows are refined, and clinicians may handle a device differently than the development team anticipated. Those observations can inform procedural refinements, instructions for use, and the design of subsequent studies.

Model selection therefore depends on the role the study plays within the development program. Feasibility, usability, and definitive GLP studies may require different approaches because the evidence expected from each is different. The goal is to select a model that allows the study to accomplish what it needs to at that stage without expecting it to resolve questions that belong elsewhere.

As development programs progress, the evidence expected from individual studies often becomes more demanding. Teams may add endpoints to capture safety observations, device performance, procedural outcomes, and exploratory findings, but the value of those data ultimately depends on whether they support what the study is intended to demonstrate.

“Collecting an endpoint isn’t the same as supporting a claim.”

That distinction is why study planning is most effective when it begins with the intended outputs rather than a list of measurements that could be collected. The evidence requirements inform the endpoints, the endpoints need meaningful success criteria, and the protocol should bring those elements together in a design capable of supporting the intended scientific and regulatory conclusions. Simply adding more endpoints does not strengthen that relationship.

By the time a definitive GLP study is underway, much of this work should already have occurred. Earlier feasibility studies should have refined the procedure, supported the selected model, and provided enough information to justify the next stage of development. The pivotal study can then build on that foundation as part of a broader evidence-generation strategy rather than becoming the point at which that strategy is first defined.

Ultimately, the value of a preclinical study is determined by more than whether it produces high-quality data. A study may confirm that a program is ready to move forward, identify a technical limitation that needs to be addressed, refine a procedure, or show that additional work is warranted before progressing. Each of those outcomes can be useful when the study has been designed with a clear understanding of what the program needs to learn and what decisions will follow.

That is why effective study planning begins well before the protocol is finalized. Understanding where the program needs to go, defining the evidence required to get there, and working backward from those requirements helps determine the appropriate sequence of studies, models, endpoints, and success criteria. It also gives teams an opportunity to resolve questions at the stage where they can be addressed most effectively rather than carrying them forward into increasingly complex studies.

Taken together, those decisions create a development program in which each study has a defined role and builds on what came before it. The objective is not to generate the greatest possible amount of data at every stage, but to develop the evidence needed to make informed decisions and continue moving the program toward its end goal.

“The most valuable preclinical studies aren’t defined by how much data they generate. They’re defined by how effectively they move a development program forward.”