Two Products, One File—Why Your Evidence Strategy Fails

Written by HATEM RABEH, MD, MSc Ing

Your Clinical Evaluation Expert And Partner

in
S

Combination products create a unique regulatory burden. You are not simply adding a substance to a device. You are integrating two regulated entities—each with its own evidence requirements, its own benefit-risk profile, and its own clinical performance expectations.

Under MDR Article 117 and Annex I Section 9, the requirements for the device remain in full force. The substance must comply with pharmaceutical legislation. And the clinical evaluation must address both—separately and as a system.

This is not about satisfying two checklists. It is about constructing an evidence strategy that demonstrates clinical performance and safety for the device, clinical performance and safety for the substance, and the absence of negative interaction when they function as one system.

Most deficiencies I see in combination product submissions come from a failure to recognize this dual structure at the planning stage. Manufacturers treat the substance as an accessory. They assume biocompatibility testing covers the safety profile. They write clinical evaluation sections that merge device and substance data without clear boundaries.

The result is a file that looks complete but collapses under structured review.

What the Regulation Actually Requires

MDR Article 61 establishes the clinical evidence requirement for all devices. For combination products, the substance is considered an integral part of the device. That means your clinical evaluation must demonstrate safety and performance for the complete system.

But Section 9 of Annex I adds a second layer. The substance must comply with the requirements of Regulation (EC) No 1394/2007 or Directive 2001/83/EC, depending on its classification. This introduces pharmaceutical standards into your device evaluation.

The practical consequence: You need two evidence streams. One addresses the device component according to MDR standards. One addresses the substance according to pharmaceutical standards. Both must be integrated into a unified clinical evaluation that assesses the combination as a functional system.

This separation is not bureaucratic. It reflects the fact that reviewers trained in pharmaceutical evaluation will assess the substance section. They expect dose-response data, pharmacokinetic profiles, toxicological studies, and clinical trial evidence showing efficacy at the proposed concentration and exposure duration.

If your CER does not structure the evidence this way, the review stops.

The Device Evidence Stream

For the device component, the clinical evaluation follows standard MDR pathways. You demonstrate performance using bench testing, biocompatibility data, clinical investigations, or equivalence to predicates.

The key challenge: The device performance must be demonstrated with the substance integrated. You cannot claim equivalence to a device that uses a different substance or no substance at all unless you isolate the device mechanism from the substance effect.

This is where planning becomes critical. If your device relies on a physical mechanism—such as structural support, fluid management, or mechanical retention—you may argue that the substance does not alter this primary function. You then demonstrate equivalence for the device mechanism separately and address the substance contribution as an additional benefit.

But if the substance modifies device performance—if it reduces friction, enhances tissue integration, or alters biological response—then the equivalence claim breaks. You are now evaluating a combination system, not a device with an inert addition.

In practice, this means your literature search must isolate device-specific outcomes. Complication rates, primary effectiveness endpoints, and long-term performance metrics must be attributable to the device function, not confounded by substance variation across studies.

Reviewers look for this separation. If they cannot distinguish device performance from substance contribution in your evidence tables, they assume the data does not support your claim.

The Substance Evidence Stream

The substance component requires pharmaceutical-grade evidence. This is not a material biocompatibility assessment. It is a clinical demonstration that the substance, at the dose and exposure route used in your device, is safe and effective.

MDCG 2022-5 clarifies this expectation. The substance must be evaluated according to the applicable pharmaceutical regulation. That means you need to present:

• Pharmacological characterization of the active or ancillary substance
• Toxicological data covering acute, subchronic, and chronic exposure where relevant
• Clinical trial data or published studies demonstrating efficacy and safety at the proposed dose
• Justification for the route of administration and exposure duration

Many manufacturers assume that ISO 10993 biocompatibility testing satisfies this requirement. It does not. Biocompatibility tests assess material safety in contact with tissue. They do not evaluate pharmacological action, dose-response relationships, or systemic exposure effects.

If your substance has a pharmacological effect—if it reduces inflammation, promotes healing, or alters cellular behavior—you need clinical pharmacology evidence. If it is released systemically, you need pharmacokinetic data showing clearance and exposure levels.

The practical burden: You may need to conduct clinical studies specifically evaluating the substance component. If published data exists, it must match your dose, route, and patient population. Generic studies on the substance in other applications do not transfer without justification.

Integration Evidence: The Third Layer

Even when you have strong device evidence and strong substance evidence, you still need to address the combination. Does the device alter substance release? Does the substance affect device mechanical properties? Does the integrated system introduce risks not present in either component alone?

This is the integration analysis. It requires you to evaluate the physical and biological interactions between device and substance under clinical use conditions.

For example: A substance incorporated into a polymer matrix may degrade the polymer over time, altering device strength. A drug-eluting coating may create local tissue reactions that affect device fixation. An antimicrobial agent may leach unevenly, creating concentration gradients that impact both efficacy and safety.

These interactions are not theoretical. They represent real failure modes that occur when manufacturers assume compatibility without testing the integrated system.

Your clinical evaluation must include evidence addressing these interactions. That means:

• Stability studies showing the substance does not degrade the device material
• Release kinetics demonstrating controlled and predictable substance delivery
• Biocompatibility testing of the combination, not just the individual components
• Clinical data showing the integrated system performs as intended without unexpected adverse effects

When I review CERs for combination products, I see this section missing more often than any other. Manufacturers present device performance, present substance safety, then assume the combination is covered. It is not.

Reviewers will ask: What happens when the substance degrades? What is the worst-case release scenario? What if the substance distribution is uneven? If you have not addressed these questions with data, your file is incomplete.

Structuring the CER for Dual Evidence

A well-structured CER for a combination product separates these three evidence streams clearly. I recommend the following structure:

Section 1: Device Evidence
Present the clinical data supporting device performance. Include equivalence arguments if applicable, isolating the device mechanism from substance contribution. Show that the physical or mechanical function of the device meets MDR requirements independently.

Section 2: Substance Evidence
Present the pharmacological, toxicological, and clinical evidence for the substance. Address dose justification, route of administration, and safety at the proposed exposure level. Reference pharmaceutical standards and ensure the evidence matches the concentration and duration used in your device.

Section 3: Integration Evidence
Present the data showing the combination is safe and effective. Include stability studies, release kinetics, interaction testing, and clinical outcomes specific to the integrated system. Address failure modes that could arise from device-substance interaction.

Section 4: Unified Benefit-Risk
Synthesize the evidence into a single benefit-risk assessment for the combination product. Show that the benefits of the integrated system outweigh the risks, and that the combination provides clinical value beyond what either component would achieve alone.

This structure makes it possible for reviewers to assess each evidence stream independently while understanding how they integrate into a complete clinical evaluation.

When you structure the evidence this way, you also make it easier to respond to questions. If a Notified Body challenges the substance safety data, you can address that section without disrupting the device performance arguments. If they question equivalence, you can revise the device section without rewriting the substance justification.

Planning Implications

The dual evidence burden for combination products changes how you plan clinical development. You cannot rely solely on device development pathways. You need to involve pharmaceutical expertise early.

That means:

• Including toxicologists and pharmacologists in the evidence planning stage
• Identifying whether existing literature covers your substance at the relevant dose and route
• Determining whether you need dedicated clinical studies for the substance component
• Testing the integrated system for stability, release kinetics, and interaction effects
• Structuring your CER to address three distinct evidence streams from the beginning

I see manufacturers realize these requirements late in development, after bench testing is complete and clinical studies are designed. By that point, the cost of addressing gaps is significant.

If you plan for the dual burden at the concept stage, you integrate the necessary studies into your development timeline. You avoid surprises during Notified Body review. And you build a file that can withstand structured regulatory scrutiny.

Combination products are not high-risk because of the device. They are high-risk because the evidence strategy is more complex, the regulatory expectations span two frameworks, and the integration analysis requires data that most manufacturers do not generate by default.

When you recognize this structure early, you can build the evidence base systematically. When you do not, you face repeated deficiency rounds, study delays, and submission failures that could have been avoided.

The regulation does not treat combination products as enhanced devices. It treats them as integrated systems requiring dual demonstration. Your clinical evaluation must reflect that reality.