Why your drug-device evidence strategy will fail the review

Written by HATEM RABEH, MD, MSc Ing

Your Clinical Evaluation Expert And Partner

in
S

The MDR does not allow you to split what is designed to work as one. When a device incorporates a medicinal substance as an integral part, the clinical evidence must demonstrate the combination performs safely and effectively together. Not the device alone. Not the drug alone. The system.

This is where most evidence generation strategies break down.

What MDR Article 1(8) Actually Requires

MDR Article 1(8) defines devices incorporating a medicinal substance as an integral part. The substance must have an action ancillary to that of the device. The device and substance must not be reusable separately.

This definition forces a specific regulatory path. The device part follows MDR. The substance part requires consultation with a medicines agency under Article 117. But the clinical evidence must address the combination.

Most manufacturers read this and assume they can generate device evidence through their usual pathways, then add a separate dossier for the drug component. This assumption creates a gap that becomes visible only during Notified Body review.

The reviewer sees two excellent reports that never answer the core question: does this integrated system work as claimed?

The Evidence Gap Between Components and Combination

Here is what happens in practice.

A manufacturer develops a drug-eluting device. They have strong clinical data on the delivery mechanism. They have pharmacokinetic studies on the drug substance. They submit both sets of evidence.

The Notified Body asks: where is the evidence that this specific drug, delivered by this specific device, at this dose and release profile, achieves the intended therapeutic effect in the target population?

The manufacturer points to the device data and the drug data. The Notified Body repeats the question.

This is not bureaucratic formality. This is the core of clinical evaluation for integral combinations. The performance of the system is not the sum of its parts. It is the interaction of its parts under real use conditions.

You must generate evidence that shows the integration works. Not that the components work.

Four Evidence Generation Strategies That Actually Close the Gap

The challenge is real. Generating clinical evidence for integral combinations requires more than standard device pathways. It requires strategic planning before the first patient is enrolled.

I have reviewed submissions where manufacturers tried to retrofit evidence after realizing the gap. It never works cleanly. The best strategies start with the combination in mind.

Strategy One: Design Clinical Investigations for the Integrated System

If you are running a clinical investigation, design it to evaluate the combination. Not the device with any drug. Not the drug with any delivery system. The specific integrated product.

Your primary endpoints must reflect the intended therapeutic effect of the combination. Your secondary endpoints should capture the device performance and the drug effect separately, but the primary outcome must be system-level.

This means your clinical investigation plan must be written with both device and drug expertise. Your protocol must address pharmacokinetics, tissue interaction, release kinetics, and clinical outcomes together.

I see protocols written by device engineers who add a section on drug monitoring. Or protocols written by clinical pharmacologists who treat the device as a black box. Both approaches fail.

The protocol must be designed by people who understand how the integration creates the therapeutic effect.

Strategy Two: Use Equivalence Only When the Integration Is Truly Equivalent

Equivalence is powerful for medical devices. It allows you to leverage existing clinical data when technical and biological characteristics are similar.

For integral drug-device combinations, equivalence becomes complex. You must demonstrate equivalence of the device platform and equivalence of the drug component and equivalence of the integration.

Many manufacturers claim equivalence based on device similarity. They argue that the delivery mechanism is equivalent to a predicate, and the drug is already approved in another form.

This is not sufficient. The integration creates a new biological interaction. The release profile may differ. The local concentration may differ. The tissue response may differ.

If you are using equivalence, your justification must explicitly address why the integration does not change the safety and performance profile. If you cannot make that case with strong technical and biological reasoning, equivalence will not hold.

Strategy Three: Generate Literature Evidence That Addresses the Combination

Literature can support clinical evaluation for integral combinations, but the literature must be relevant to the integrated system.

I review clinical evaluation reports where manufacturers include literature on the drug’s therapeutic effect from systemic administration, then literature on the device platform from previous generations without the drug, then claim this constitutes sufficient evidence for the combination.

It does not.

The literature search must target studies that evaluate similar integrations. Drug-device combinations with comparable release profiles. Localized drug delivery in the same anatomical site. Studies that measure outcomes reflecting the combined therapeutic mechanism.

If no such literature exists, that is not a problem with the search strategy. That is a gap in available evidence that must be acknowledged and addressed through other means.

You cannot create evidence by juxtaposition. Literature that sits side by side is not literature that demonstrates integration.

Strategy Four: Structure Your PMCF to Monitor the Integration

Post-market clinical follow-up for integral drug-device combinations must monitor the system in real use.

Your PMCF plan must include endpoints that capture device performance, drug effect, and integration outcomes. This means long-term safety monitoring for both components. It means effectiveness measures that reflect the combined therapeutic goal.

Many PMCF plans I review focus on device-related adverse events and miss the drug-related signals. Or they monitor drug levels without connecting those levels to clinical outcomes achieved by the device’s delivery mechanism.

Your PMCF strategy must be designed with the understanding that real-world use may reveal interaction effects not visible in controlled investigations. The plan must be sensitive to those signals.

This also means your PMCF must continue long enough to detect delayed effects. Drug-device interactions can produce late-emerging safety concerns. A one-year follow-up may not be sufficient.

The Regulatory Consultation Requirement

MDR Article 117 requires consultation with a medicines agency when a device incorporates a medicinal substance as an integral part. This is not optional.

The medicines agency provides a scientific opinion on the quality, safety, and usefulness of the substance, taking into account the intended purpose of the device.

This opinion becomes part of your technical documentation. The Notified Body will review it. If the opinion raises concerns, those concerns must be addressed in your clinical evaluation.

I see manufacturers treat this consultation as a separate track, managed by regulatory affairs without input from clinical evaluation. This creates disconnects.

The medicines agency opinion often includes recommendations for clinical evidence generation. Those recommendations must be integrated into your clinical evaluation strategy. If the opinion requests specific studies or data, your CER must explain how you addressed those requests.

What Reviewers Look for in Your Clinical Evaluation Report

When a Notified Body reviews a CER for an integral drug-device combination, they are looking for a clear demonstration that you understand what you are evaluating.

They expect to see clinical evidence that directly addresses the integrated system. They expect to see justification for how the drug and device work together to achieve the intended effect. They expect to see risk analysis that considers interaction effects, not just component risks.

If your CER reads like two separate documents merged into one file, the review will stop. If your benefit-risk analysis treats the device and drug independently, it will be rejected.

The reviewers want to see that your clinical evaluation strategy was designed for integration from the start. That your evidence generation plan anticipated the need to demonstrate combined performance. That your literature search targeted relevant integrations. That your clinical investigations were designed with the combination in mind.

This is visible in how you structure your appraisal. It is visible in how you synthesize data. It is visible in your benefit-risk conclusions.

A strong CER for an integral combination makes the integration the center of the evaluation, not an afterthought.

The Cost of Getting This Wrong

The consequences of a weak evidence generation strategy are not just regulatory delay. They are fundamental gaps that cannot be fixed without new data.

If you reach the Notified Body review stage and realize your evidence does not demonstrate the combination’s performance, you cannot solve this with literature updates or equivalence arguments. You need new clinical data.

That means new investigations. New timelines. New costs.

I have seen manufacturers three years into the review process, facing a request for clinical data they cannot provide because the evidence was never designed to answer the right question.

The only way to avoid this is to design your evidence generation strategy with the integration in mind from the first planning meeting. Before the first protocol is written. Before the first equivalence claim is drafted.

Moving Forward

If you are working on an integral drug-device combination, your evidence generation strategy must reflect the reality of integration. Your clinical evaluation must demonstrate that the system works as intended.

This is not theoretical. This is what the Notified Body will assess. This is what the medicines agency will evaluate. This is what your technical documentation must support.

Plan for integration. Design for integration. Evaluate integration.

Everything else follows from that foundation.

In the next part of this series, I will address ancillary medicinal substances in devices—where the drug plays a supporting role, not an integral one. The evidence strategy shifts, but the principle remains: you are still evaluating a system, not isolated components.

Peace,
Hatem
Clinical Evaluation Expert for Medical Devices
Follow me for more insights and practical advice.