Facilitating early access whilst maintaining quality
Highlights from a recent joint EMA-FDA workshop
The European Medicines Agency (EMA) and the US Food and Drug Administration (FDA) both have schemes to fast-track and support development of medicines which address unmet medical needs. Their underlying aim is to help patients benefit from such therapies as soon as possible. However, due to time constraints and novel situations, applicants for these schemes often face challenges in relation to manufacturing development and meeting associated quality and data requirements. With this in mind, the EMA and FDA held a joint workshop with stakeholders in November 2018 to identify where the main challenges lie and what further support can be given in related areas.
The workshop focused on process validation, control strategies, good manufacturing practice (GMP) compliance, comparability, stability, and currently available regulatory tools. A detailed report of this meeting has now been published which summarises the associated discussions and main conclusions. Much has been proposed for further work and elaboration, and the salient points are outlined herein after a brief overview of the schemes currently offered by the FDA and the EMA.
Within the US, two programs provide expedited development and review of drugs for serious or life-threatening disease, or conditions where preliminary clinical evidence indicates that the drug may be a substantial improvement over existing therapies. These are known as the Breakthrough Therapy and Regenerative Medicines Advanced Therapy programs. Similarly, in the EU, the PRIME scheme is offered for medicines which offer a major therapeutic advantage over existing treatments or benefit patients without treatment options.
Drugs which meet these designations are given several benefits, including an expedited review clock, enhanced agency communication, and advice throughout development. However, this results in an accelerated clinical and manufacturing development program, which reduces the amount of information available at the time of submission. A risk-benefit assessment is performed on the regulators’ side, balancing risk of less manufacturing information versus patient benefit, whilst still maintaining product quality as a priority. In cases of urgent clinical need, the FDA also offers rolling assessment of data, in which complete modules are submitted for assessment at different times.
The workshop report shows that regulators are willing to be flexible and consider requirements on a case-by-case basis under these schemes. When discussing process validation, it was in fact emphasised that “not one size fits all”. There was general agreement between regulators and industry that for products of high unmet clinical need, it would be acceptable to defer submission of certain process validation data to the post-approval phase, but the extent of this deferral and mechanism for delayed submission remains to be further agreed.
Overall, it was emphasised by regulators that process validation is considered a life-cycle activity. They are open to exploring new avenues for the submission of confirmatory validation data, as long as there is a clear lifecycle validation plan in place and evidence that manufacturers know how to appropriately control their manufacturing process. A combination of currently available process validation approaches may also be necessary, including the use of prior knowledge, concurrent validation, ongoing/continued process verification protocols, and post-approval change management protocols (PACMPs). In relation to PACMPs, an acceptable strategy may be to register a constrained manufacturing process, with tight process parameter ranges, which are then eased post-approval once additional data has been gathered. Other options, which have already been used successfully by applicants, are to decouple active substance and finished product process validation activities, or pursue these in an overlapping parallel fashion.
Comparisons were made to Investigational Medicinal Product Dossiers (IMPDs) when it comes to control strategies. IMPDs are continuously reviewed and refreshed as learning and control evolves, and a similar approach is needed for drugs on accelerated pathways. Related avenues that were explored in relation to biological products include using pre-agreed batch analysis trending approaches, supplemented by product understanding and statistical elements to set and confirm appropriate specifications. It was also agreed that, in certain cases, it may be possible to register specifications which are wider than the batch data available at the time of approval. However, here regulators would need to see a clear plan for the establishment of appropriate specifications as the product knowledge increases post-approval. An associated option that was discussed was the use of continuous process verification to identify any out-of-trend batches which are still within specifications, followed by documentation of a root-cause analysis within the regulatory file. However, the most appropriate means of including such trending and statistical control is still to be elaborated.
In a separate breakout session on chemical products, the use of purge factors when setting specifications was discussed. Purge factor calculation is a tool used to predict the likelihood of carryover of impurities from one manufacturing step to another, or to the final drug substance. It is already detailed in ICH M7 as an acceptable tool to justify a substance’s control strategy in relation to mutagenic impurities. The acceptance and use of in-silico models and purge factor calculations has been identified as an area for joint follow up by the FDA and EMA, particularly regarding related information that would be required in a marketing authorisation application.
Stability studies for both biological and chemical products were highlighted by industry as a critical factor in many development and supply situations. A joint presentation by several drug companies made the case for accelerated stability studies, which would allow stability data generated under real time conditions to become “confirmatory rather than pivotal in developing product understanding”. However, representatives from regulators were divided on the acceptability of this approach, though the suggestion was made that these could be combined with predictive stability models and prior knowledge where available, to produce a more robust basis for a shelf-life claim. The development of guidance on the use and acceptance of accelerated stability data has been proposed as another area for joint EMA-FDA follow up.
Managing changes during development brings further challenges, and issues related to process comparability for biological products were also discussed. One area highlighted is that, for some products, the use of patient material for demonstrating comparability is not possible due to ethical or availability issues. In such cases, scientific knowledge on the difference between surrogate non-patient material (such as that from a healthy donor) and patient materials in relevant manufacturing processes would be needed. This would allow determination of whether comparability exercises can be carried out using the surrogate material, followed by concurrent validation using patient material. Risk-based approaches can also be used to narrow any comparability studies needed, by identifying critical quality attributes impacted by manufacturing changes. As manufacturing site transfer often requires extensive comparability data, another area that has been highlighted for follow up is the feasibility of starting early commercial supply from a clinical development site, rather than switching to a commercial manufacturing site.
In fact, GMP issues tend to be a stumbling block in many applications, so regulators advised industry to raise any relevant questions at kick-off meetings or scientific advice procedures. Also advised was strict coordination between the applicant and contract manufacturing organisations, as communication breakdowns have been encountered, in which manufacturing facilities were not ready or aware of the development program. Out-of-specification results, with specific reference to advanced therapy medicinal products (ATMPs) were also discussed, and the importance to lean on “the side of caution and strongly justify any decision taken” was highlighted. Identified for follow-up was the acceptability of using comparability as the basis for accepting clinical trial data which has been generated using product manufactured in a facility that is not fully compliant with GMP requirements, such as an academic laboratory or research hospital.
Summing up, regulators detailed the tools that are already available to help applicants navigate early access schemes but emphasised that the most important tool available is communication. This area will remain one to watch in the regulatory world, as both US and EU representatives stated that they will be reflecting on, and responding to, the challenges highlighted. In the meantime, the organising committee has proposed the development of a “Toolbox guidance” for PRIME products, which will summarise regulatory tools and related scientific elements that are already available in the EU to facilitate early access applications.