10 Things to Note About Manufacturing Viral Vectors
Blog Sep 17, 2019
Global interest in cell and gene therapies is on the rise, and continued growth is expected for the coming years. To discuss the challenges and solutions associated with manufacturing viral vectors at the scales and quality required for these therapeutics to be approved by the regulatory authorities, representatives from Pall and Molmed jointly hosted a seminar in Tokyo earlier this year.
Pall representatives focused on the equipment they have available, while their colleagues at Molmed gave perspectives as contract manufacturers using the equipment for production.
We had the opportunity to be a fly on the wall while the team (below) discussed the challenges involved:
- Clive Glover, Director, Cell & Gene Therapy, Pall Biotech (CG)
- Luca Alberici: Chief Business Officer, Molmed (LA)
- Giuliana Vallanti, Development & Quality Control Director, Molmed (GV)
Here are a few snippets from the discussion which highlight key challenges and progress in the field:
CG: When it comes to viral vectors, you need to put in a minimum of 4-5 genes in order to get the virus made. Another big problem is that some of the proteins associated with viruses are actually toxic to the cells that produce them, and it means that you have a very short window of production before all of your producer cells start to die. So there's an enormous amount of complexity that starts to come through as a result of some of that basic biology that's needed in order to put the virus together.
- Vectors are temperature-sensitive
GV: Vectors are sensitive to temperature – so vector production has to be done at a certain temperature to keep the vector stable and viable.
- Quality of the analytics
CG: At the moment, the set of analytics that is used to measure how much virus you have, and whether that virus is actually functional… they're very slow. They are low throughput, and they lack precision. Even as you are making your way through the manufacturing process, you often have to do the whole process blind, because there simply isn't the time to get the analytics happening so that you can really measure how things are going.
For this reason, a lot of that processing must be done in as quick a time as possible, but also you have to somehow figure out a way to make sure you do it exactly the same way each time. Because if there's any kind of deviation in the process parameters that you are using, you simply don't know what the outcome is going to be.
Overall, it is a very challenging process - largely from the fact that you have a very complicated molecule that you're trying to produce, and there is poor analytics associated with it all.
All in all, you need a very specialized skill set in order to be able to do it effectively.
- Cell and gene therapies are usually developed by academia, polished by industry
LA: The field is really transitioning and maturing. Previously, it was more academia that was developing and bringing this therapy into clinical trials. The processes that we had were really basic, manual and open systems.
Now what all of us as an industry are trying to do to, thanks to tool providers, CDMOs (contract development and manufacturing organizations) and sponsors, is to take over from academia and industrialize this process to allow the treatment of not just 10 to 15 patients for an ultra-rare indication.
Instead we want to target a large, oncological indication where you need to produce not liters, but hundreds of thousands of litres of vectors, and then treat several thousand patients.
- Upscaling efficiently is crucial
LA: Most, if not all, the therapies authorized so far, and a big chunk of those that are in clinical development are autologous therapies. And that has a logistic cost that is pretty large, but the production cost is much higher.
We are manufacturing a batch for every single patient, and that has a big impact in the system that you're using.
In the end, that's the reason why this therapy has such a high cost on the reimbursement side.
$300,000 to $400,000 seems very expensive, but the overall margin is way less than every other therapy that has been approved in the past, just because the cost of manufacturing these therapies is absolutely huge.
So in this sense, there are two big things that we are working on; one, is of course upscaling vector production, i.e. going from cell factories towards large-scale disposable bioreactors
For cell manipulation, it's really about closing the system, automating the process, and then running in parallel, multiple batches in the same facility in order to at least to minimize the fixed cost of maintaining the GMP facility and the entire quality system behind each batch that is manufactured.
- Bioreactors can reduce errors and operational costs
CG: What we have found is that by moving to a bioreactor-type system, you reduce the operational expenditures by about 40 to 50%.
Bioreactors offer a well-contained, single-use entity which can be scaled linearly. For example, the iCELLis 500 bioreactor that we offer at Pall, in a fairly small footprint, is the equivalent of almost 800 cell factories.
So there's dramatic space-saving, and a dramatic reduction in the amount of money that you have to spend to scale up that kind of process.
Also, the amount of labor associated with working on a bioreactor is shrunk almost five-fold. So rather than having lots of people trying to do these open manipulations you end up with one closed system, you can get away with one or two operators, more-or-less.
And this has huge advantages, not only in terms of cost savings, but also in terms of removing the possibility of human error from the process.
- The bioreactor market is just getting started
CG: Of the current therapies on the market, maybe one at most is produced using bioreactors. But I think what is true to say is that the target patient population for many of the existing therapies is quite small. So many of them are targeting quite rare diseases or those with fairly low dosages required.
So for something like Duchenne muscular dystrophy, at the moment, there are very, very high viral vector quantities being tested. And those therapies would simply just not be possible without the use of bioreactors to make the viral vectors. And we expect that there will be several of those kinds of therapies on the market in the next two to three years, treating things like haemophilia, spinal muscular atrophy, Duchenne muscular dystrophy… all of those kinds of therapies are coming to market soon and they all require very high cell doses and have relatively large patient populations. And it simply wouldn't be possible to bring them to market without the use of more automated, more sophisticated pieces of equipment like a bioreactor.
It's an exciting space for everybody I think, for tools providers have the same great opportunity. And the therapy providers like Molmed are hopefully starting to see the benefits in terms of the availability of tools to enable the manufacturing.
- Bioreactors are not necessarily out-of-reach for academics
CG: With our iCELLis bioreactor for example, the full-scale manufacturing device is really quite a large piece of equipment. Because it is capable of growing adherent cells, we generally measure it in meters squared. So our full-scale manufacturing bioreactor comes in the smallest size of 66 m2, and the largest size is 500 m2.
But what we also provide is what we call the iCELLis Nano, and the iCELLis Nano is really put out there for the major reason is that it's obviously very expensive to do process development at full scale. It costs hundreds of thousands of euros or pounds or dollars to run those, mainly in terms of plasmid DNA costs and cell culture media. So clearly doing that kind of process development at full scale... it just can’t be done.
So we also have the iCELLis Nano available, which allows you to run close to the same process at 0.53 - 4 m2. And so what it means is that you're able to do it at a dramatically smaller scale.
And what we also see, is that those kinds of bioreactors are being used by the likes of some of the more clinically-minded academic units.
So for example, one of the first publications that we had on the iCELLis came out of Memorial Sloan Kettering in New York, which is an academic facility, but they saw the potential of the iCELLis, started using the iCELLis Nano and showed that it could be used for this particular production.
So certainly, we are seeing the uptake of the smaller bioreactor by many of these academic units, which I think will only help the speed the movement of those kinds of therapies from academia more into the industrial domain where they have to go in order to become commercial products.
- There’s a need to shorten timelines
LA: Considering timelines as a whole, you have to remember the quality control tests that occur before the project is released. Those can take at least as much time as the manufacturing itself, in most cases, even longer
And that's why in most cases, we are going towards a frozen product that can be shipped all over the world. But nevertheless, this autologous therapy has a four to six-week lead time from the time of apheresis into the time of infusion into the patient.
There are potential other approaches, like allogeneic therapy, or in vivo gene therapy, where they are all off-the-shelf products. So you need to manufacture and release it, but then as soon as you enrol or identify the patient, you just treat them.
- Enormous growth = enormous demand on production and supply chains
CG: Another very important thing to think about is the supply chain of the tools with which you're manufacturing the drug. You know, there are sick patients out there, and if you can't get the single use manufacturing equipment on time, then that can significantly endanger patients’ lives. So it's very important that the supply chain of the manufacturing equipment also be considered in this.
Recently we've been investing quite heavily at Pall to try and make sure that the supply chain associated with our manufacturing equipment is robust enough, and that we have enough capacity in order to meet the ongoing big increase in demand.
We see this market growing at 30% plus per year, over the next few years which creates an enormous increase in the demand on the tool side of things.
So, ensuring that you've got good supply chains for the manufacturing side of things is incredibly important in order to get these drugs in a timely fashion.