by Robert Walgate
What can developing countries do about pandemic flu? A vaccine, in 1.5 million doses, would be the answer – but how ready are the major vaccine companies to make it?
Mexican swine flu is creating a global response – and not without good reason. The virus is an influenza A strain, with H1N1-type surface proteins (the ‘studs’ in the picture), which makes it very similar to the virus of the 1918 pandemic, which killed 20-40 million people worldwide, more than died in the First World War.
Even regular seasonal flu is believe to kill 250,000 – 500,000 people a year, in 3 – 5 million cases, with a mortality of already about 10%. So a new pandemic strain, to which most people may not be immune, is something to prepare for – but remember, the real mortality rate of the Mexican virus, and its potential for spread, is not yet clear. And ‘pandemic’ to WHO just means a globally spreading human disease, not necessarily a fatal one.
Nevertheless the fundamental answer to a viral disease like influenza, whether seasonal or pandemic, is a good vaccine. Smallpox was eliminated with a vaccine. But flu is flexible – the virus frequently recombines pieces of human, pig and bird flu, and drifts in genetic content even within the human population. So the vaccines need to keep changing, and production capacity is the key.
“Developing countries have recently become acutely aware of the need to develop [influenza vaccine] production capacity in their own countries” a WHO expert told RealHealthNews. “We know that maximum present production capacity for a monovalent vaccine is 1.5 billion doses. That’s nothing compared to the world population.”
Moreover there are three billion people in urban slums, which means no clean water, sanitation or easy access to health care.
“In slum areas as everywhere else, the answer is social distancing – finding places to go to keep apart” said the expert. “In Hong Kong in the SARS epidemic, when they had an outbreak in an apartment building, they were able to move all those people to areas where they’d be relatively safe.
“But it’s a very difficult issue while there is not enough vaccine supply. There’ll be a lag, should there be a pandemic, as the virus is identified and the vaccine made.”
So said David Heymann, now WHO’s Assistant Director-General for Health Security and Environment, speaking to RealHealthNews in September 2007, but we are facing exactly that lag now, as the Mexican H1Ni swine flu epidemic grows –
So where does the world stand for pandemic flu vaccine production now, in May 2009? Sanofi Pasteur, based in Paris, is the vaccine division of the Sanofi Aventis group, and claims to produce around 40% of the world’s seasonal influenza vaccines. Novartis, the Swiss company, produces another 30-40%. And the international GSK (GlaxoSmithKline) mops up most of what’s left over. So how are these three key companies placed for making large volumes of a vaccine against the potential coming pandemic?
For an answer, RealHealthNews went straight to the companoes. We began by asking them about the prospects for making pandemic flu vaccines in cell culture, like beer in a brewery. Cell culture vacccines were signalled just a few years ago as the answer to pandemics – because they could be produced faster, and allow the production of greater, in theory unlimited, volumes of vaccine.
RealHealthNews talked first to Sanofi Pasteur:
> Robert Walgate, RealHealthNews: Where is Sanofi Pasteur at the moment with cell-based flu vaccines?
Albert Garcia, spokesman for Sanofi Pasteur: It’s a complex question. Today, in fact, and for the next ten years let’s say, we at Sanofi Pasteur think that we will be relying totally on the egg based technology. We already produce around 40% of the world influenza vaccine in the world, in that way.
We are of course involved in cell technology, in flu cell technology, in research and development. We use cell lines that we ourselves developed, in fact, for other vaccines like polio several years ago. But at this moment Sanofi is not licensed to produce a cell-based flu vaccine.
We consider ourselves pioneers in developing cell based vaccines, but we know today that we can obtain million of doses of influenza, just using eggs, not cells. And we think that in fact cell technology will not allow us to gain a lot of time in vaccine production.
>RHN: Why is that, do you think?
AG: We have all year round access to millions and millions of eggs, special eggs from health poultry, in a secure process that no-one can enter. The eggs are delivered to us, and then virus grows very well in three days, and then you collect the virus from the egg.
>RHN: So that’s very quick.
AG: Yes, but then you have a long process of around 12 weeks mostly taken by quality assurance, testing and so on. With a cell-based technology, we would gain maybe one week, maybe ten days – not so long. The only thing that cell-based technology will allow us to do is to get rid of the eggs in the process, but we manage very well with them.
So if there is a pandemic in the next coming weeks, yes, we are prepared for that. In fact, we have been preparing for several years because of the H5N1 bird flu issue, so we have been preparing our production plans to produce a pandemic strain as soon as the authorities and the World Health Organization ask us to produce a pandemic strain.
>RHN: Supposing that you were given one of the new Mexican H1N1 swine flu strains, from which to make a vaccine, how long would it take before it was in production?
AG: For a seasonal vaccine, it takes three months.
>RHN: Three months?
AG: Three months for a seasonal vaccine, which is three strains – you know that it’s three vaccines in one.
>RHN: For the trivalent vaccine, yes.
AG: If it will be monovalent, we would need three months, only because of the quality control and testing etc, which you cannot shorten. But obviously, we would be working and have been working since last week, very closely with the authorities, to try and shorten the time as much as we can.
>RHN: Three months already sounds quite quick. People talk about six months to create a new flu vaccine rather than three months.
AG: You need the extra three months because that’s the usual time for WHO, when we’re making seasonal vaccines, to select and pick up the strains, and from these strains create a working seed lot for the vaccine to be produced.
>RHN: But there can be hiccups in the whole process, I understand.
AG: Sometimes for example you have strain that grows very quickly and another that is harder to grow. It’s a living organism so each strain is different.
So then you have this parameter over which we have no control. But no producer in the world has a control on this parameter, but we have a great deal of experience in this going back decades. So our vaccine production is totally dedicated to producing influenza vaccine, whatever the strains are, seasonal trivalent, all kinds of mixed strains, or we can switch from seasonal to pandemic at the instant the authorities ask us to produce a pandemic strain.
>RHN: But roughly speaking, if the virus grows well enough, you are saying that three months after being given the seed strain from WHO, you could be producing global quantities of vaccine?
AG: This quantity is firstly shaped to the quantity of doses of seasonal influenza that we can produce worldwide. We at Sanofi Pasteur produce more or less 40% of the worldwide production in influenza, so in our two plants, in the US and Normandy, France. Last year we produced 170 million doses of trivalent seasonal vaccine, so for the monovalent you can multiply by three.
But our production is adjusted to the seasonal demand; and then we have another parameter that we assess with authorities, which is the number of micrograms of the virus proteins that we need to protect human beings. For seasonal, it’s 15 microgram per strain, so in each dose of seasonal flu vaccine you have 45 micrograms, 15 of each strain. Then, for the whole new strain, this may change, or not, so this is to be decided very quickly, in relationship with the authorities and the WHO obviously.
Usually we identify the right amount by a clinical trial, so in case of emergencies, it will be a question of discussion between authorities and producers. It will not be a completely unknown disease, so we can capitalise on decades of knowledge on influenza vaccine production and protection.
And sometimes you have some slight differences between the strain which is included in the vaccine and the virus which is circulating in the population. The strains are separating because influenza strains always may change a little bit.
>RHN: They evolve, yes.
AG: Yes, but we try to be as close to the circulating strains as possible, and this is done in collaboration with WHO and the other authorities.
>RHN: One last question, you explained very clearly at the beginning about why cell culture only gives you an extra week or so, but you mentioned that most of the time, most of the three months was taken up with this quality assessment. What exactly has to be done there?
AG: Well, you know that the vaccines should be of the utmost quality. At Sanofi, or any other producer, we have hundreds of tests that we apply. And the authorities also need to be certain about the final product before it can be released as a public vaccination. So sometimes the testing is done twice. And in some countries even a third testing is also performed on these vaccines.
>RHN: Three times the 170 million doses of the trivalent vaccine, comes to about 500 million on monovalent; so you’re saying roughly speaking that three months after being given the seed dose from WHO, you could have 500 million doses of vaccine for this Mexican H1N1 strain.
AG: We would not have 500 million doses after the first week, but from the fourth month, we will be delivering hundreds of doses, million of doses in the next weeks, and producing continuously, seven days a week.
I cannot say how many doses we will be able to produce per week or per month, it’s impossible because of the parameters that we discussed, but what is sure is that we will be producing as many doses as possible, in the short term, but we cannot anticipate how many million doses we would be able to produce.
>RHN: Just one last question, are you actually working on cell culture for influenza or not?
>RHN: You are. But in this case you don’t think it’s going to be particularly important.
AG: Definitely not.
RealHealthNews then talked to Novartis:
>Robert Walgate, RealHealthNews: What does Novartis do for influenza vaccine production?
EA: We have three vaccines plants in Europe: Siena in Italy, Marburg, Germany and then Liverpool in the UK. The Liverpool plant mainly manufactures the vaccine for the US market. But the other plants manufacture also many other vaccines and also flu vaccines for the European market. So it’s an interesting heritage.
>RHN: Of acquisitions, I imagine.
EA: Right. So we have different manufacturing processes for the flu vaccine.
>RHN: Oh, you do. Well, let me come to that in a minute. For a rough measure of capacity, what proportion, globally speaking, do you have of the seasonal flu vaccine market?
EA: We’re the second largest supplier – manufacturer would be more appropriate, depending on how many people buy. I think we have somewhere between 30% and 40%.
>RHN: So it’s close to the Sanofi Pasteur figure, which is about 40%.
EA: Right, so it’s a little less than that and then you have GlaxoSmithKline, which is also one of the key suppliers, and numerous small ones.
>RHN: Right, so that’s the proportion. Now, tell me about the technologies. I thought you would probably be using eggs, but you say you have more than one technology.
EA: Well even for the egg manufacturing, there are different processes that go with it, so to speak. So for egg-based, we have the process in Liverpool is slightly different from the one we use in Siena, though both are egg based.
>RHN: So these give slightly different products.
EA: Exactly. And then in Marburg we have two technologies right now. There’s an egg-based technology in Marburg but that’s ‘running out’. And then we have the cell culture technology that we manufacture out of Marburg.
>RHN: That’s Optaflu, I believe?
EA: Exactly. And for Optaflu, if I can add just one more thing, we’re building also an additional manufacturing site in the US.
>RHN: So tell me how Optaflu is going, because I saw some quotes from Novartis, I think it must have been in 2007, that you hope to be supplying the whole of the European Union with that in the 2008/09 season. But you didn’t.
EA: No, we didn’t. And there were numerous reasons for that. I mean we still have our egg-based capacity and, to be quite honest, the predictions that had been done about the demand for flu vaccines didn’t quite live up to expectations.
>RHN: Oh, I see, so that was market driven rather than anything to do with the technology?
EA: Right. A large part was market driven, yes.
>RHN: So it’s going well, that technology?
EA: Oh, it’s a new technology so you still have challenges, but yes, it’s producing.
>RHN: All right. Let me just turn to this potentially pandemic Mexican swine flu. You told me earlier that depending on how the virus grows, etc, you’d make a choice of technologies to use. There’s also a choice to be made in the trade-off in production capacity between seasonal and the Mexican swine flu, but I guess that’s made on advice from WHO.
EA: Right, yes. So there’ll be trade-offs on everything basically.
>RHN: Because you haven’t got the capacity to continue the same amount of seasonal and add a pandemic vaccine, I imagine?
EA: No, and I don’t believe that any of the manufacturers would supply seasonal. Part of the premise for a manufacturer of these kind of things is that you don’t want the manufacturing capacity to sit idle, so you produce seasonal. Now if there is need for a pandemic vaccine, then the decision will have to be made, do you switch, do you produce less of one or the other?
>RHN: Which of course depends upon the progression of the pandemic and the mortality, because seasonal flu also kills: according to WHO, annual epidemics result in between three and five million cases of severe illness and between 250 000 and 500 000 deaths every year around the world.
EA: Right, so we respond to what the health authorities ask for.
>RHN: A difficult call. Okay. So you could use Optaflu for that, could you? I mean you could adjust the technology to use the H1N1 virus?
>RHN: Why only theoretically?
EA: Because nobody has the seed virus yet to see in which medium it grows best.
>RHN: I see. So if you had the seed virus, if one had been prepared by the US Centers for Disease Control in Atlanta (CDC) or wherever, then you would test it out to see which medium it grew best in? And it might potentially grow best in your cell culture?
EA: That’s the theory, it could.
>RHN: In theory.
EA: Well, there’s still many unknowns around it right now. That’s why I’m being careful.
>RHN: But the cell-based flu vaccine was originally hyped as being the fast way of producing vaccines, responding rapidly to epidemics, and I wondered if that was actually held up by the reality of the technology.
EA: Not really. When you think about it for egg-based, you still need to get the eggs to grow the vaccine in, which is the limiting factor because you need to get the eggs. So that’s the challenge, the longer timeframe for the egg supply, whereas with cells basically you can start up very quickly.
>RHN: Okay. You can start up very quickly. What Sanofi told me was that they thought it might give you an extra week to ten days. But of course they’re defending their egg-based technology. So I wondered what you would say.
EA: I’m not going to play tit for tat! So here’s the reality: the manufacturing process, since you’re growing basically the antigen, or you’re growing the virus, is a biological process. So put it in an egg or a cell culture, and the timeline is fairly similar. The limiting thing for egg-based is the egg supply because the hens can only lay so fast. That’s very cheeky but that is the limiting factor.
>RHN: I see. But potentially with a bioreactor of cells you can create as much as you like.
>RHN: So any advantage has to do with the capacity rather than the speed at which you can respond to a change in the viral envelope, or whatever?
EA: Right, it’s the capacity and the ability to sustain. A lot of concern before a week ago, especially when it was around the H5N1 avian flu, because that affected poultry, was how would that actually affect the egg supply? And then when you have no egg supply you actually can’t manufacture any vaccines; it was a catch 22 at that point. We have to have sustained supply of eggs which we are all still dependent on, including us at Novartis. We have some more freedom with the cell culture to sustain the production [in the absence eggs].
>RHN: Right, although of course you need enough bioreactors and all the rest of the production chain.
>RHN: At the moment you can produce enough Optaflu for a small fraction of your market for seasonal flu, but not your whole market.
EA: Correct. But then again, when you look at it in terms of seasonality, okay, you basically only produce one big batch and then you’re idle for the rest of the year until you pick up again. In the case of a pandemic you would sustain production throughout, so it would be continuous.
>RHN: Let’s say it proves to be less of a problem, the Mexican swine flu, than it might appear at first, because the mortality is not so high, at least outside Mexico. So let’s hope it’s not a serious problem. But this whole experience might be seen as another learning experience, along the curve towards the killer pandemic when it comes some time in the future. So let’s say, looking to the future of your production either for epidemic or for seasonal flu, do you expect to be shifting towards cell-based technology? Does it have advantages for you as a company?
EA: I do believe it has advantages and, from what I understand from the scientists, is that you get a purer product. And you don’t have this dependence on the eggs, as I said before, which creates also more flexibility. So I mentioned earlier the site we’re building in the US, in Holly Springs, North Carolina – that is going to be only for cell culture vaccines.
>RHN: So maybe that’s a signal for the future…
GlaxoSmithKline refused to answer us directly, but referred us to an undeniably informative backgrounder, recently prepared. They say:
“Influenza vaccines have been available for over 50 years. Generally, they are trivalent, i.e. they contain three different, inactivated virus strains, either whole viruses or parts or subunits of them. Because the flu virus is constantly changing its antigenic shape (either in minor drifts or in major shifts), the composition of the flu vaccine needs to be adapted to these changes regularly.
“The monitoring of these changes is done by the Global Influenza Surveillance Network of the World Health Organization (WHO). At the beginning of each year, the WHO makes a recommendation for the strains to be included into the vaccine for the coming influenza season.
“Traditionally, influenza vaccines are produced in fertilized chicken eggs. Eleven days after fertilization, the influenza virus – each strain is grown separately – is injected into the eggs and accumulates in the fluid surrounding the embryo. A high-yielding donor strain is co-injected. The embryo becomes infected so that the virus can multiply. After several days of incubation, machines open the eggs and harvest the virus. Then the virus is carefully purified, chemically inactivated and used to produce the vaccine.
“On average, between one and two eggs are needed to produce one dose of vaccine. The entire production process lasts at least six months.
“An alternative way of producing flu vaccine is based on cell or tissue cultures. This method of production was first described in the mid-nineties and is still in its experimental stage, yet all major players in the vaccines industry have embarked on development.
“Mammalian kidney cells are preferably used for these cell cultures. The virus is injected into these cells, which multiply as the virus does in them, before the the cells’ outer walls are removed, harvested, purified and inactivated.
“This process resembles a biotechnological fermentation, in which you move from small litre jars to huge fermenters during production. It is not unusual to produce vaccines in cell cultures; polio vaccine, for example, is made this way.
“But the development of such a cell culture -based production is a long and arduous process, namely in terms of its efficiency, standardization and validation.
“The egg-based production of flu vaccines is well established and cost-effective. It works well – though with obvious disadvantages: the extensive planning (procurement of many millions of eggs, long timeline) necessary limits flexibility in case of exponentially increasing demand.
“A flu pandemic could probably not be contained and defeated on egg-based production, because the production takes too long and eggs don’t grow on demand. Cell culture based systems, however, could be rapidly expanded and scaled up in times of emergency – which also points to one potential downside of tissue culture production: the up-front costs for operational readiness of such plants (involving huge fermenters). These costs are much higher than the costs for egg -based systems, and the yield may be slightly slower.
“But mammalian cell-based flu vaccine production may have other benefits: potential impurities in eggs (antibiotics, other viruses) may cause sterility problems. There is also a risk of allergies against egg albumin. None of these risks exists in cells. Last but not least, the strains grown in cell cultures equal the original clinical isolates,
while the growth of epidemic viruses in eggs result in variants that are antigenically distinct from the original viruses.
“Moreover emerging endemic viruses sometimes do not grow
at all in eggs. Virus grown in mammalian cell culture is therefore more representative of the circulating wild type virus than that grown in eggs.
“While it is not yet proven that cell culture based flu vaccine products will be superior to egg-based products, they are expected to progressively capture an increasing market share starting in the future.”
Please write your comments below, click preview, and then submit