Why would you need a pilot plant?
I was asked this question at a local industry happy hour recently, as we were discussing some of the available biotech manufacturing capacity here in the greater Seattle area. A former company in the area had an amazing pilot plant, which was later purchased by a different company for conversion to GMP, which I feel was a big loss for both the original and current owners of the facility. Not to mention it turned out to be an incredibly difficult conversion, because the original engineers designed the building in such a way that it was intended to be impossible to convert to GMP! Those engineers knew the value in having a pilot plant, and also knew that if it could be converted to commercial space it would be, so they did their best to ensure that it would remain a development asset for as long as possible.
Pilot plants allow for training, they allow for mistakes, they allow for technology development, and so much more.
1) A pilot plant is somewhere that new technicians can learn how to follow batch records for solution compounding and unit operations, without the threat of a lot loss hanging over their head if they make a mistake. It’s a particularly great option for training operators on column packing, which despite the proliferation of single-use membranes and prepacked columns is still an important skill for many processes.
2) A pilot plant allows for re-use of materials for teaching or testing purposes, or even allows “GMP-Expired” material to be used while it’s still in the actual product expiry period, which allows for these paid-for goods to be used and not just discarded.
3) A pilot plant allows for a process scale-up to happen and for any corrections or changes to be made on the fly during production, without generating deviations or impacting a critical clinical lot release. Even with decades of experience in scaling up projects, biologics are still biologics and they may behave a little differently once they’re at large scale – so being able to adjust processes to accommodate those issues quickly is critical for success.
4) A pilot plant is where new technologies can be tested with the goal of eventual implementation in the GMP facility. Try out that new vendor’s column hardware, see if that new filter housing design is truly scalable from the small-scale units, figure out the best assembly process for the new viral filter cartridge, and do mixing tests on the redesigned SUM bags – all without impacting an actual product run.
5) A pilot plant allows for documentation to be developed, new systems to be trialed, and staff to be trained on how to run deviations and implement CAPA’s, again without impacting any actual products. Many minor errors, typos, or improvements are found the first time a batch record is used in-plant, and having this first use be in a non-GMP setting allows for these changes to be made prior to the GMP run, where any mistakes or changes will result in a deviation that may impact lot release.
But running a pilot plant isn’t free!
Of course it’s not, and there does need to be a certain size and budget for a company to be able to support one as part of their research and development arm. But the benefits of having one, which are not limited to what I listed above, are worth the investment and will pay back dividends in better trained operators, fewer deviations, better implementation of new processes when scaled up, faster integration of new technology, and more.
Companies that don’t have their own facilities can reap some of these benefits by using the “Engineering Run” option at their CDMO. Engineering runs are performed in the GMP facility, using GMP batch records and materials, but they are not intended to produce/release a lot for use in humans, and as such there is a lot of flexibility to monitor and adjust the process in real-time. This is an especially great option for low-endotoxin toxicology study lots, as the GMP facility is much better equipped (even if only considering the HVAC system) to produce such material than the typical development lab. I believe that every new product should have at least one Engineering run, prior to the first clinical GMP batches, because it will catch any issues in batch records or scaleup and fix them without causing deviations in that critical first lot. Fewer deviations means a better chance of a rapid lot release, which means keeping to the scheduled timeline for product fill and clinical study initiation.
Non-GMP runs are a valuable resource that is vastly underutilized by many companies. The rush to be “first in human” and the push to meet the expectations of Wall Street often results in cutting corners wherever possible. Most of the time this works out, but it’s always risky, and sometimes it ends up in significant errors or deviations that then result in a longer delay than it would have been to do it “properly” the first time! Projects going into the clinic should consider carefully whether saving the few weeks in the timeline to go straight to GMP is worth the risk – and what will be impacted if that lot is delayed. The choice is there, but I will always choose to run a pilot run for a new process whenever possible.
#pilotplant #gmp #manufacturing #processtransfer #techdev #training