Myths of Vaccine Manufacturing | In the Pipeline

In the last few days, the question of why more drug companies haven’t been enlisted for vaccine production has come up. It’s mostly due to this tweet:

The problem is, as far as I can see, this is simply wrong. There are not “dozens of other pharma companies” who “stand ready” to produce these mRNA vaccines. To me, this betrays a lack of knowledge about what these vaccines are and how they’re produced. Even though I’m not a pharma manufacturing person, I am indeed a pharma researcher in general. So I would be glad to fill in this gap, and here’s why it’s not possible to suddenly unleash dozens of companies to crank out the Pfizer/BioNTech and Moderna vaccines.

The first thing to understand is that these are not, of course, traditional vaccines. That’s why they came on so quickly. mRNA as a vaccine technology has been worked on for some twenty to twenty-five years now, from what I can see, and (as I never tire of mentioning) we’re very fortunate that it had worked out (and quite recently) several of its outstanding problems just before this pandemic hit. Five years ago we simply could not have gone from sequence to vaccine inside of a year. And I mean that “we” to mean both “we the biopharma industry” and “we the human race”.

At this point, let me briefly dispose of an even less well-founded take that’s been going around as well. I’ve seen a number of people say something like “We had the vaccine back in February! It only took until the end of the year to roll it out because of the FDA!” The main thing I’ll say about that idea is that no one who actually works on vaccines, in any capacity, has any time for that statement. Not all vaccine ideas work – we’re already seeing that with the current coronavirus, and if you’d like to talk to some folks about that, then I suggest you call up GlaxoSmithKline and Sanofi and ask them what happened to their initial candidate, and while you’re at it, call up Merck and ask them what happened to their two. Note that I have just named three of the largest, most experienced drug companies on the planet, all of whom have come up short. So no, we did not “have the vaccine” in February.

One of the other reasons we didn’t have it back then is the whole problem of figuring out how to make the stuff, and that brings us back to today’s discussion. How do you make the Moderna and Pfizer/BioNTech vaccines? And what’s stopping “dozens of other pharma companies” from doing the same? Let’s get into those details, stopping briefly again to imagine asking James Hamblin above to actually start naming “dozens” of pharma companies. Anyone have a good over/under on how many names would get rattled off?

OK, let’s look at the actual supply chains. The single most informative piece I have seen on this is from Jonas Neubert – I’ve recommended it before, and this is absolutely the time to recommend it again. I also have to mention this detailed article at the Washington Post, which focuses on the Pfizer/BioNTech vaccine, and this one at KHN about manufacturing bottlenecks in general. You should also read this Twitter thread from Rajeev Venkayya, who knows what he’s talking about when it comes to vaccine manufacturing, too. All of these will cover details that I’m not even going to get to today!

It’s not in my nature, since I’m an early-stage drug research person myself, but I’m going to totally sidestep all the R&D questions behind the various components and just treat this as a manufacturing process that fell from the sky in its final form. To distill a huge amount of background and detail down into the simplified steps, we have:

Step One: Produce the appropriate stretch of DNA, containing the sequence that you need to have transcribed into mRNA. This is generally done in bacterial culture.
Step Two: Produce that mRNA from your DNA template using enzymes in a bioreactor.
Step Three: Produce the lipids that you need for the formulation. Some of these are pretty common (such as cholesterol), but the key ones are very much not (more on this below).
Step Four: take your mRNA and your lipids and combine these into lipid nanoparticles (LNPs). I have just breezed past the single biggest technological hurdle in the whole process, and below you will learn why it's such a beast.
Step Five: combine the LNPs with the other components of the formulation (phosphate buffers, saline, sucrose and such) and fill those into vials.
Step Six: get those vials into trays, into packages, into boxes, into crates, and out the door into trucks and airplanes

OK, you have now produced the mRNA coronavirus vaccines and shipped them out into the world, so sit back and open a cold one. You will not reach that stage, though, without some significant challenges. Let’s take those step by step. The DNA production in Step One is not too bad. As the Neubert article details, Pfizer does this themselves in Saint Louis, and Moderna outsources this to the large and capable Swiss firm Lonza (update: a good part of the Lonza work is being done in Portsmouth, NH). DNA plasmid production on an industrial scale is pretty well worked out (and keep in mind that “industrial scale” for DNA means “a few grams”. It’s not something you can do in your garage – as with every step in this process there’s a lot of purification and quality control to make sure that you’re making exactly what you think you’re making and that it looks exactly within the same specs as the last time you made it. But that’s what biopharma manufacturing folks are good at, and there are a lot of people who can do it. That said, a goodly number of them are occupied doing that for just the vaccines, but if we needed more of this DNA, sure, we could produce more.

But we don’t. That’s not the rate-limiting step. Nor is Step Two, the transcription into mRNA. Pfizer and BioNTech do this in Andover, MA and at BioNTech facilities in Germany. They have manufacturing in Idar-Oberstein (a town I recall visiting in the cold rain one weekend in 1988 during my post-doc!) and last fall they bought another facility in Marburg which is just getting revved up for such production now. The Moderna mRNA step is also handled in Switzerland by Lonza. Now this is not so common as an industrial process, for sure, because it’s only relatively recently that people have been treating RNA species as actual drug substances themselves, worthy of scale-up manufacturing. If I had to ask someone else to make me some more bags of bespoke mRNA, I might turn to Alnylam (who have a manufacturing facility in Norton, MA although to be sure, they’re using it for their own drugs!), but doing so would not increase the number of vaccine vials coming out the other end of the process. RNA production is certainly closer to being rate-limiting than Step One, but it’s nothing compared to the real bottlenecks that are coming.

Now to the lipids in Step Three. This doesn’t have to be done in sequence like the DNA/RNA step, of course – the lipids needed for the formulation are an entirely different production process. As the Neubert article will show you, Pfizer and BioNTech are getting all of theirs from a UK company called Croda, with production likely going on in the town of Alabaster, Alabama, which (unlike Idar-Oberstein) I am certain that I have not visited. Now, each of these vaccines needs some odd lipids with positively charged groups on them; that’s a crucial part of the formulation. These are surely not trivial to make on scale, but they’re still small molecules with relatively straightforward structures. I’m sure that barrels of these things aren’t stacking up at the factory for lack of demand, but I don’t believe that they’re the limiting reagent in manufacturing, either. If you had to, you could surely get some other manufacturers up to speed on the process.

I’m going to skip ahead to Step Five and Step Six. These are surely running at a good clip, but they are more traditional functions of a drug company (or of any manufacturing company). It’s true that pharmaceutical vial fill-and-finish on this scale narrows you down to fewer players than would be involved in, say, canning tuna. But these folks are already involved. Pfizer is doing this in Kalamazoo and in Puurs, Belgium, and BioNTech is doing this in several locations in Germany and Switzerland, both at its own facilities and via at least two contract firms. Moderna, meanwhile, outsources this to some of the big players in the US and Europe: Catalent, Rovi, and Recipharm. Everyone in this part of the manufacturing business has known for months that a Big Vaccine Push has been coming, and has been cranking up vial manufacturing, bringing all available production lines up to speed, and signing deals all over the place with everyone who has any kind of advanced vaccine effort.

Ah, but now we get back to Step Four. As Neubert says, “Welcome to the bottleneck!” Turning a mixture of mRNA and a set of lipids into a well-defined mix of solid nanoparticles with consistent mRNA encapsulation, well, that’s the hard part. Moderna appears to be doing this step in-house, although details are scarce, and Pfizer/BioNTech seems to be doing this in Kalamazoo, MI and probably in Europe as well. Everyone is almost certainly having to use some sort of specially-built microfluidics device to get this to happen – I would be extremely surprised to find that it would be feasible without such technology. Microfluidics (a hot area of research for some years now) involves liquid flow through very small channels, allowing for precise mixing and timing on a very small scale. Liquids behave quite differently on that scale than they do when you pour them out of drums or pump them into reactors (which is what we’re used to in more traditional drug manufacturing). That’s the whole idea. My own guess as to what such a Vaccine Machine involves is a large number of very small reaction chambers, running in parallel, that have equally small and very precisely controlled flows of the mRNA and the various lipid components heading into them. You will have to control the flow rates, the concentrations, the temperature, and who knows what else, and you can be sure that the channel sizes and the size and shape of the mixing chambers are critical as well.

These will be special-purpose bespoke machines, and if you ask other drug companies if they have one sitting around, the answer will be “Of course not”. This is not anything close to a traditional drug manufacturing process. And this is the single biggest reason why you cannot simply call up those “dozens” of other companies and ask them to shift their existing production over to making the mRNA vaccines. There are not dozens of companies who make DNA templates on the needed scale. There are definitely not dozens of companies who can make enough RNA. But most importantly, I believe that you can count on one hand the number of facilities who can make the critical lipid nanoparticles. That doesn’t mean that you can’t build more of the machines, but I would assume that Pfizer, BioNTech, Moderna (and CureVac as well) have largely taken up the production capacity for that sort of expansion as well.

And let’s not forget: the rest of the drug industry is already mobilizing. Sanofi, one of the big vaccine players already (and one with their own interest in mRNA) has already announced that they’re going to help out Pfizer and BioNTech. But look at the timelines: here’s one of the largest, most well-prepared companies that could join in on a vaccine production effort, and they won’t have an impact until August. It’s not clear what stages Sanofi will be involved in, but bottling and packaging are definitely involved (and there are no details about whether LNP production is). And Novartis has announced a contract to use one of its Swiss location for fill-and-finish as well, with production by mid-year. Bayer is pitching in with CureVac’s candidate.

This is all good news, but it’s a long way from that tweet that started this whole post off. There are not “dozens of companies who stand ready” to produce vaccines and “end this pandemic”. It’s the same few big players you’ve already heard of, and they’re not sitting around and watching, either. To claim otherwise is a fantasy, and we’re better off with the facts.