Following her session at Medtech Summit, Medtech Insight interview Maetrics’ Dr Amie Smirthwaite for a practical overview of how to meet the new clinical evidence requirements.
Amanda Maxwell, European Regulatory Affairs Editor, Medtech Insight
Clinical evidence requirements are among the biggest challenges for manufacturers meeting the requirements of the EU’s new Medical Device Regulation. Maetrics’ Amie Smirthwaite explains how manufacturers should transition from the one to the other in this interview with Medtech Insight following her session at MedTech Summit.
When it comes to clinical evidence requirements under the Medical Device Regulation, the requirements are lengthy, and the wording may seem complicated to follow. It is an area constantly highlighted by industry, regulators and notified bodies as being both challenging and difficult to prepare for.
One key question that has generated a lot of attention is: what defines “sufficient” clinical evidence?
Despite the perception that the bar for clinical evidence is higher under the MDR than under the Medical Devices Directive (MDD), the definition of “sufficient clinical evidence” hasn’t really changed under the MDR; it’s just that more people are understanding what it means, Amie Smirthwaite, senior director, global clinical practice at Maetrics consultancy told Medtech Insight. The main thing that is genuinely tougher is that many of the equivalence routes will be gone under the MDR, she explained. The greater depth and frequency of technical documentation assessment also make the requirements more demanding, the Maetrics consultant added.
Amie Smirthwaite, senior director, global clinical practice at Maetrics consultancy
Medtech Insight caught up with Dr Smirthwaite to follow up on her session at Medtech Summit.
Medtech Insight: How can medtech manufacturers who are compliant with the MDD or Active Implantable Medical Devices Directive (AIMDD) prepare to collect sufficient clinical evidence as they start to think about complying with the MDR? AMIE SMIRTHWAITE: First they should review their existing clinical evidence, removing any sources which are no longer valid under MDR (e.g. equivalence routes no longer valid under MDR, clinical evidence which does not meet the new definition of clinical data under Article 2(48), etc). You can then analyse this with a bottom-up approach – what does it demonstrate, for which indications and patient populations and/or anatomical locations, which treatment variants and device combinations, over what period of time? How well does this compare with outcomes or performances achievable with other state-of-the-art treatment or diagnostic options for the intended patient population?
What should manufacturers do if there are gaps or areas where clinical evidence is weak?
Smirthwaite: If there are gaps or areas where the clinical evidence is weak, it may be possible to make a justification for the applicability of one data set to an area where a gap exists. If not, they should use the remaining time under their MDD or AIMDD certification to undertake appropriate post-market clinical follow-up (PMCF) address those gaps prior to application for certification under the MDR.
Should manufacturers declare this to the notified body?
Smirthwaite: If it’s a genuine gap that cannot be justified, the manufacturer should consider reducing the scope or indications or removing the device from the market.
If the overall benefit-risk conclusion of the device can be justified despite gaps (for all treatment indications, patient populations, device variants, etc) then the clinical data should be presented in a clear and transparent way, and a justification based on objective evidence for why the manufacturer believes the devices are safe and effective in the areas where there are apparent gaps.
Are there any other options for manufacturers who find themselves with gaps or weak clinical evidence?
Smirthwaite: It may also be possible to explore alternative equivalence pathways. Particularly for non-implantable class IIa and IIb devices, it may also be possible find additional equivalent devices.
However, it is important to be aware that there is generally an expectation that PMCF should have been undertaken for devices where initial clinical evaluation relied on equivalence; Meddev 2.12/12 says that “the NB shall as part of its assessment of a specific medical device: … verify that PMCF is conducted when clinical evaluation was based exclusively on clinical data from equivalent devices for initial conformity assessment and that PMCF addresses the residual risks identified for the equivalent devices.”
Despite the word “exclusively” in this sentence, the expectation is normally that the manufacturer should gather sufficient clinical evidence such that use of equivalence data is no longer required in future certifications.
Therefore, if legacy devices are relying on equivalence data in an MDR application, this should be carefully justified and backed with an appropriate PMCF plan.
Where in the MDR, does it define “sufficient” clinical data?
Smirthwaite: The MDR does not explicitly define “sufficient” clinical data, but clinical evidence is defined in Article 2(51) as: “clinical data and clinical evaluation results pertaining to a device of a sufficient amount and quality to allow a qualified assessment of whether the device is safe and achieves the intended clinical benefit(s), when used as intended by the manufacturer.”
Is there a definition of sufficient clinical evidence in the EU’s guidance documents?
Smirthwaite: Meddev 2.7/1 rev 4, the clinical evaluation guidance document written in the context of the MDD but updated in the knowledge of what would be coming under the MDR, defined “sufficient clinical evidence” as “an amount and quality of clinical evidence to guarantee the scientific validity of the conclusions.”
Although MDCG 2020-6 – written in the context of the MDR – indicates that the meddev guidance definitions in general cannot be assumed to be applicable to MDR (via Appendix I), there is nothing obviously inconsistent between this definition and implicit requirements for quantity and quality of data in the MDR.
What sources might manufacturers use for clinical evidence under the MDR?
Smirthwaite: Article 2(48) defines acceptable sources of clinical data. These include:
What about the use of non-clinical evidence?
Smirthwaite: Article 61(10) also allows the use of non-clinical evidence where justified in terms of the intended purpose, clinical benefits, risks to patient, etc. MDCG 2020-6 highlights that Article 61(10) should only be applied in exceptional circumstances and further clarifies that direct clinical benefits require clinical evidence.
So, for example, it may be possible to demonstrate that a guide catheter can navigate tortuous vessels through simulated use and other cumulative evidence as indicated in the MDCG 2020-6 Appendix III table, but you could not demonstrate claims that a balloon catheter is a suitable option for treatment of peripheral vascular disease without clinical data.
Can you explain more about the importance of Article 61 (10) on non-clinical evidence and when it can and cannot be used, since the phrasing is ambiguous?
Smirthwaite: Article 61(10) allows the use of non-clinical evidence for demonstration of aspects of clinical safety and performance for some devices, where justified, based on the intended purpose and risk evaluation.
The wording “without prejudice to paragraph 4” means that you can’t use Article 61(10) to bypass the requirements of Article 61(4) – i.e., if the devices are class III or implantable, you can’t say that it’s not appropriate to use clinical data to demonstrate clinical safety and performance.
What is indirect clinical evidence and when can this be used?
Smirthwaite: Indirect clinical evidence is evidence which supports safety or performance conclusions, but which either does not come directly from experience with patients or comes from data sources which are not robust in terms of scientific validity. Examples include:
Indirect clinical evidence can only be used to support safety or performance requirements which are not linked to a direct clinical benefit. So, for example, you might be able to demonstrate that a stent has sufficient radial force to maintain vessel patency through animal or simulated use testing. But to demonstrate that stents are a successful treatment for atherosclerosis you will require good quality patient data (i.e., direct clinical evidence).
What about sufficient clinical evidence for legacy devices?
Smirthwaite: It will really depend on the intended purpose, state of the art, the risks associated with a device, and the clinical unknowns, which can be illustrated by comparing clinical evidence required for legacy hip replacements with that for a central venous catheter.
The expectation for a hip replacement would be that it would have robust clinical evidence addressing all device variants. This is because: there are too many clinical unknowns associated with small design variations of implants; there is a history of poorly performing devices – it’s not a given that every hip replacement will performance well in relation to the state of the art; and the lack of overt safety signals does not guarantee that a device will perform well in relation to the state of the art, even if a device has been on the market a long time and has a good sales volume.
As registries began to expand and gather more complete data sets, it became apparent that many very well-established hip replacements were not performing as well as other devices with the same indications for use (and with no other specific advantages). The expectations from the EU have been that devices like hip stems should already have their own data, of sufficient quality and quantity, to demonstrate that they perform at least as well as other devices with the same intended purpose and patient population.
On the other hand, a central venous catheter might require less clinical evidence, unless it has novel design features compared to the state of the art or is making unusual claims.
In general, the design and materials are well-established, there are no moving parts, the clinical risks and complications are well-understood, etc, so they might fall into the category of legacy “well-established technology.” In this case, cumulative evidence from levels five-12 from Appendix III of the MDCG 2020-6 guidance can be used. This might include overall evaluation of the state of the art, simulated use testing, compliance with design specifications defined by common specifications or standards, complaints and vigilance trending, etc.
The idea here would be to specify the intended performance and associated clinical risks and benefits, and provide a narrative that helps to explain from a common-sense based perspective why the evidence available demonstrates that the benefits will be achieved and the risks adequately mitigated in relation to the state of the art.
