MDR Article 117 – a look into the Crystal Ball (Part 5)
- July 10, 2018
- Posted by: confinis
- Category: Regulations
Many thanks to all of you who provided their feedback to Part 1, 2, 3 and 4 of this series of articles – highly appreciated and overall a lot of great information for all stakeholders!
Flashback: In May and June 2018, TOPRA (www.topra.org) published two highly interesting articles written by Mark Chipperfield and Tim Chesworth that list a number of concerns and open questions that require clarification. As in the previous parts, please find below the respective text form the TOPRA articles (italic) and my interpretation / suggestion:
Degree of applicability. The principle that the medicinal product device constituent of such a product is not to be treated fully as a CE medical device drives two alternate interpretations:
” Treat the medicinal product device as a medicinal product. Only apply MDR Annex 1, handle all other aspects of medicinal product device design, development, risk management, manufacturing, etc, under the Medicinal Product Directive (MPD) and pharma guidelines, versus;
” Treat the medicinal product device as a CE-medical device. Apply the MDR in full to the device constituent, except for the CE-marking requirement.
Under the Medical Devices Directive, a marketing authorisation application (MAA) typically includes a declaration that Annex 1 Essential Requirements (ERs) have been met, often supplemented by a simple ER checklist. Art.117 introduces enhanced oversight of medicinal product devices. However, it also has the potential to cause disruption to current approaches. Explanation of the extent to which the broader MDR should be applied to the medicinal product device is needed. If notified body (NB) opinion is to be based upon ‘would-be medical device’ risk class, this requires clarification.
According to my interpretation, only Annex I (GSPR) applies to device constituent parts that fall under article 117. The commission requested the Notified Bodies to apply the MDR “as it is written” and not to interpret it so this is the logic consequence. However, in order to provide objective evidence that the GSPRs are met, harmonized standards or common specifications may be used and there are tons of standards out there (harmonized and non-harmonized) which are considered to be state of the art (e.g. EN ISO 14971 regarding risk management, EN 60601-1 regarding electrical safety or the entire ISO 11608-x series regarding needle-based injection systems). I strongly believe that those standards still apply if you want to meet the GSPRs (unless an applicant has a better method).
The only question is how certain paragraphs which refer to the GSPR are interpreted. As an example, Article 61 (1) says “Confirmation of conformity with relevant general safety and performance requirements set out in Annex I …shall be based on clinical data…”. However, a clinical evaluation is no longer part of Annex I…
Quality system. Article 117 clearly states that the medicinal product device will be reviewed and regulated as a medicinal product. MAA applicants often leverage components of their pharmaceutical quality system (PQS) when developing or manufacturing medicinal product devices. Although PQS elements may meet the spirit of medical device quality systems, NB review approaches may not be designed to assess these effectively, were they to audit. The topic of quality systems audit requires guidance.
Article 117 only refers to the GSPR. How an applicant provides objective evidence regarding how to meet them is at his discretion as long as he can demonstrate that this leads to a safe and effective device according to the state of the art. Nothing in article 117 says that a certified quality management system is required. Although EN ISO 13485 will likely be harmonized with the MDR at some point of time, it’s still a standard and not the law.
Therefore, I believe that applicants can continue to leverage the GMP quality system for medicinal products when manufacturing integral DDCs, but depending on the specific type of device, need to complement it with relevant new documented procedures, e.g. if you develop a device containing software.
Where NB opinion is not needed. The complexity of medicinal product development is increased when integrating a medicinal product device. It is important to focus NB opinion on where it is needed. An Art.117 opinion is not needed before performance of clinical studies. Confirmation of the General Safety and Performance Requirements (GSPRs) that are relevant in the clinical study may be useful study risk mitigation and may support ethics committee approval. However, this is outside the remit of Art.117 and should not be an MDR requirement. It is recommended that further clarification be provided.
While I believe that a sponsor should have objective evidence that the device constituent part meets all applicable GSPRs as part of the clinical study application, I also think it’s outside of article 117. Article 117 only speaks about the marketing authorization dossier.
Timing for seeking an opinion. To allow for the finalisation of device-related information while minimising delay to the medicinal product MAA submission, some consideration needs to be given to options for parallel review processes. This is heavily dependent on the type of information regarding the drug and the drug-device interface that the NB will need (or want) to review – and its availability.
If the NB opinion could be submitted during MAA review or against a specific review timepoint (eg, no later than day 30 or 120, or similar), then the applicant could present a more final package to the NB, and the NB could form a more complete opinion.
It is recommended that a competent authority (CA) perspective on this aspect be sought with urgency and incorporated into any appropriate subsequent guidance.
There is indeed some room for interpretation. Article 117 only says that the dossier needs to contain the opinion but not when it needs to be available. Since the CA should (ideally) only check if the opinion is favorable, it seems acceptable to me that the opinion is made available by Day 80 or even by Day 120 (or equivalent, for a decentralized review). There is no risk for the target population by doing so but gives the industry a bit more time. Notified Bodies will need a final package to perform a thorough assessment and provide their opinion.
The applicant for NB opinion. It seems appropriate that the medicinal product developer is the direct applicant to the NB and any opinion certification or documentation carries their company name as (legal) manufacturer. This ensures clarity that the issued opinion relates to the specific medicinal product device. However, consideration needs to be given to an applicant leveraging an opinion across multiple products when the device is devices equivalent or similar. Where a device technology developer wishes to seek a NB opinion, careful consideration needs to be given to the potential for absence of information related to the final product.
We believe the applicant should be the (legal) manufacturer of the medicinal product while a device developer may play a key supporting role. It is recommended that expectations of who should be the applicant are clarified.
I also believe that the applicant should be the MAA and the device developer/manufacturer will play a key role.
If a device developer wishes to seek an NB opinion it will most probably be difficult to provide objective evidence that all applicable GSPRs are met, notably
o GSPR 5: Eliminating or reducing risks related to use error
o GSPR 10.3: Compatibility with medicinal products concerned ( stability, extractables, leachables)
o GSPR 22: Products intended for use by lay persons (usability / human factors); if applicable
o GSPR 23: Label and instructions for use
Since NBs are private organizations they may allow device manufacturers to file their device platform as kind of a “master file” which will be assessed regarding all GSPRs that can be met by the device platform alone (e.g. biocompatibility, electrical safety, dose accuracy). This will not lead to an “opinion” as defined in article 117 but rather a kind of a “pre-assessment report”. This would not only be helpful for the device manufacturer but also the medicinal product manufacturer who then has some degree of certainty that the device constituent underwent an assessment with a positive outcome. The downside of this approach is that the medicinal product manufacturer probably needs to work with the same NB as the device manufacturer. However, a final opinion can only be issued if the intended use / indications are defined including but not limited to the drug product, target users and labeling.
Submissible documentation. For a CE-medical device, the requirement is for technical documentation, typically presented in the summary technical documentation (STED) format. Some NBs have recently commented that their requirement for documentation for an Art.117 opinion would be the same as for a CE-marking model.
Setting an expectation for pharma companies to generate STED-format documentation specifically when information is available in an alternative form may be considered somewhat unreasonable. In extreme cases, there could be a large quantity of documents to restructure retrospectively, requiring significant time to adapt and giving no direct improvement to the safety and efficacy of the associated product.
There could be value in a parallel guidance to the Notified Body Operations Group (NBOG) (BPG 2009-1) “Guidance on design dossier examination and report content” specifically for these products and NB opinion.
It is recommended that NBs performing opinion assessments on these types of products be able to accommodate the differing existing forms of documentation structure. Clarification of the expected structure for documentation is needed. If major changes are needed to current practice, then industry must have time to adapt.
One proposal is for the 3.2.R module of the MAA common technical document to be the document file submitted to the NB. This would leave the main body of Module 3 for drug substance, drug product and drug-based elements of integration of the device while 3.2.R can capture the device and device-related elements of integration with the drug. With some limited overlap, both reviewers would see the information relevant to them.
An alternative could be for provision of a GSPR checklist with references to the relevant supporting documents. This could be submitted to the NB for an initial consultation before specific documents being requested to support the establishment of safety and performance.
While I’m aware of the fact that certain NBs require or at least encourage the use of specific formats I don’t believe there is a legal basis for it. The technical documentation of a product needs to provide objective evidence that the relevant GSPRs are met, nothing more, nothing less.
Annex II and Annex III of the MDR define the content of a technical documentation for devices to be CE marked. The intro in Annex II says “… shall be presented in a clear, organised, readily searchable and unambiguous manner and shall include in particular the elements listed in this Annex.” It clearly speaks about the content and not the structure!
Being a NB technical documentation reviewer myself and knowing many colleagues, it may be wise to provide a matrix that shows links from the technical documentation to the different elements listed in Annex II to facilitate a speedy review.
The proposal of the authors to use the information in 3.2.R module of the MAA common technical document may work for simple devices (e.g. pre-filled syringes) but for more complex devices (e.g. on-body injectors containing electronics and software) that level of detail will hardly be enough from the NB perspective to perform a thorough assessment as required to issue an opinion.
Review of the product as a system. As the review of constituents is divided and responsibility placed with different parties, there is potential to overlook or duplicate the review of topics related to interactions between the constituents. To illustrate this and the associated challenges, the example of a syringe-based autoinjector is described below.
The syringe must protect and contain the medicinal product formulation, minimise particulates, ensure integrity, maintain sterility and allow visual inspection. Barrel, stopper and tip-cap or needle shield materials and coatings interact directly with the medicinal product formulation. These components ensure integrity through life and would be reviewed by the medicine CA.
The syringe also has device functionality and interfaces with the autoinjector. Attributes such as accurate graduation, appropriate stiction (break-loose force) and friction (gliding force), mechanical strength, would logically be reviewed by the NB.
Needle protrusion length may be a clinical consideration requiring medicine CA review. However, the engineering aspects are more likely to require NB review. Dead volume and gas bubble impact on dose accuracy could also, in principle, be reviewed by either party, but perhaps they would have different focal points in their review.
Integrating the syringe with an autoinjector requires: dimensional compatibility; confirmation that assembly does not breach container closure integrity; controlled force application to the external surfaces of the syringe; capability to overcome stiction and friction forces of the syringe and the resultant injection time influence, etc. Outer body parts may partially cover the needle protrusion, may restrict the visibility of the syringe for inspection. Any restricted piston travel has implications for dose accuracy. Time and storage conditions may influence the performance or even the function. These, again, could be considered by either party, but possibly with different intent.
The responsibility for review of different parts of a system requires clarification for all parties involved.
Given the challenges nicely presented above, clarification is indeed required. While a certain overlap will be difficult to avoid, I’m more concerned that certain aspects may be overlooked and therefore not be covered.
More to follow in Part 6… Keep your eyes open!
Again, I strongly encourage other professionals involved in this field to provide their opinion and start a lively conversation on this platform – the whole industry can only benefit from this!