Prepare the way: Pilot manufacture for drug delivery devices
Bringing a drug delivery device to a clinical trial is a complex endeavor. You need to keep a handle on multiple moving parts, for example, the active pharmaceutical ingredient (API) development, the regulatory pathway, establishing the supply chain, and labeling. Developing a novel drug delivery device takes things to another level.
Many manufacturers shy away from the challenge, relying instead on proven technologies, so patients and clinicians don’t benefit from the most advanced user-centered design, and pharma companies can’t leverage the competitive advantage new technology delivers.
Here, I share some of the obstacles encountered conducting pilot builds in-house to help our clients bring devices to market – and give four pointers for ideal pilot manufacturing for clinical trials.
Develop your manufacturing process and architecture in tandem
Once a design is frozen, making changes is expensive. After it’s passed to a high-volume manufacturer, costs become exponentially higher. Understanding manufacturing processes – and how changes can impact a project’s timeline – is critical for successful delivery. You need to prepare for the supply-chain ‘whiplash effect’: a tiny change at the top of the chain can mean seismic shifts at the end of it. That knock-on is the reason your product development strategy should incorporate pilot manufacture. Pilot manufacture keeps this effect in check by minimizing the volumes involved.
It’s vital to consider the whole supply chain, not just the component manufacturer, but the process equipment partners, filling, packaging, sterilization, and logistics. Each step has requirements to be understood and communicated to relevant parties. By developing manufacturing and assembly processes in tandem with device design, we can be flexible to insights arriving from either direction.
Pick the right partners for success
One of my first jobs was for a major automotive company. In their heyday, they ran the foundries that made the ball bearings for their vehicles. Today, they wouldn’t dream of it. No company does everything anymore. Few organizations would claim to be experts in all areas of drug delivery. Even those that manufacture and fill their own devices rely on external partners to produce the plastic resin and packaging materials and often outsource activities such as sterilization.
Partnering with experts to contribute specific knowledge is a time-efficient way to overcome obstacles in the development pathway. It also unlocks access to cutting-edge equipment and facilities that are expensive to maintain. While developing a breath-actuated inhaler, we engaged an external test house to conduct bio-compatibility evaluations on the device. We may have the skills in-house to perform this testing but maintaining accreditation for an activity that isn’t core to our business doesn’t make financial sense.
Know the limits
When developing a device, it’s essential to explore sources of potential variation. The same goes for the manufacturing process. You can use various tools to do this, but we frequently return to the humble ‘process failure modes and effects analysis’ (pFMEA). The pFMEA is a structured way to consider all the process steps – and how they could go awry. Developing a robust pFMEA ensures the team focuses on the highest risk areas and starts thinking about implementing mitigations.
A key checkbox for each manufacturing process step is if the results can be verified or validated. The US Food & Drug Administration Code of Federal Regulations Title 21 defines verification as “confirmation by examination and provision of objective evidence that specified requirements have been fulfilled.” Many processes can be verified using in-process measurement systems. But several can’t, for example, the joining of two plastic parts by ultrasonic welding. You can’t determine the strength of this weld without destructive testing. The ultrasonic welding process needs to go through process validation to determine the limits within which the process should be operated.
When communicating with stakeholders, it’s crucial to know the volume limits and have a realistic plan for producing parts representative of the final production process. For example, how many parts can the mold tools make? There’s a trade-off between tool production speed, tool cost, and tool life. Low-cost soft aluminum tools might be ready in two weeks but only suitable for 2,000 shots, whereas a more expensive hardened steel version might take 16 weeks (without validation) but last for over 100,000 shots.
Validating injection mold tools can be a lengthy process. Exploring the process window needs planning and performing multiple molding and measurement runs and subsequent analysis. Companies only want to bear this cost once, so experienced development teams need to hold firm when encountering adverse test results. I know of an auto-injector that showed promise early on, albeit with an infrequent failure observed in testing during development, that was allowed to pass into design freeze. More thorough testing during design verification revealed results that triggered the regulatory application to be rejected. Cue months of tooling validation needing to be reassessed.
Combination products require the delivery devices to be filled or co-packaged with primary containers of the API. Clinical trials complicate this because they need devices filled with the API or safe and sterile placebo. The filling process can be complex, especially when the API is highly viscous or uses technologies such as microspheres to sustain the release of active components over time. You need to factor in time to explore the filling and develop the process settings. Thought needs to be given to the amount of API and placebo available and the lead times for new batches as this can limit the amount of filled and finished devices.
Not documented? You’re not done.
Understanding the controls needed to manage risk is essential for a manufacturer delivering high-quality, safe, and reliable products. ISO 14971 sets out a best practice framework for managing risk in the context of medical devices. We advise creating a quality control plan that summarizes the production risk mitigation controls identified through risk assessment in a clear, concise format. This control plan also blueprints the actions needed if a specific limit or check is breached.
Anyone who has experienced an audit by a notified body or regulatory agency will recognize their love of records. The mature management systems used by large manufacturers often aren’t available for the short-run low volumes involved at the scale-up stage. Building a bespoke database compliant with 21 CFR part 11 to handle records can be a lengthy activity, particularly when compared with the pace of setting up paper-based systems.
Managing paper records generated by the manufacturing process can be challenging, putting storage and recall burdens on a manufacturer. Companies scan these documents soon after completion to reduce this burden. But the destruction of originals is risky, and the recall and integrity of e-records must be checked before destruction.
Pilot manufacturing helps optimize the journey of a drug delivery device to clinical trial. It’s not without its own challenges, but synchronizing manufacturing process and device design development, partnering with experts, having a plan for producing components that’s representative of the final production process, and keeping a handle on records puts you in a position to maximize pilot manufacturing’s potential.
References
- CFR – Code of Federal Regulations Title 21 [Internet]. Accessdata.fda.gov. 2021 [cited 21 December 2021]. Available from: https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?fr=820.3
