In this installment of this week in influenza I am going to walk through a paper that came out this September in Vaccine ( Vaccine, Volume 23, Issue 38 , 7 September 2005, Pages 4678-4684) by Ghendon et. al. from Russia. For those wanting to read the article the PubMed abstract is available for free. They describe a cell culture procedure for the growth of influenza strains that can then be used to manufacture vaccine.
This story is cross-posted at the Microbiology Blog
Traditionally, flu vaccine has been grown in chicken embryos, a process developed in the 1950s. Scientists been aware that this process has several important limitations. First, because of the long time it takes to generate enough eggs, vaccine development can take up to 6 months. Second, lethal avian influenza strains (such as H5:N1) cannot be grown in chicken eggs. You cannot get enough virus out, because it ends up killing the egg. Third, chicken embryos can be contaminated with retroviruses and there is a small chance that they may find their way into the vaccine. Fourth, influenza strains passed through chicken embryos may loose one of the hemagglutinin domains that our immune systems recognize and raise an immune response against. This makes the vaccine less effective since our immune system does not see part of the real virus it may be challenged with later.
The equipment and medium that was used for these experiments are readily available, making it possible for any laboratory or pharmaceutical company that has cell culture facilities to become a supplier of the vaccine. In case of an emergency, facilities around the country could be used. One set of experiments looked at whether the virus changed during its growth in the fermenter. The concern was that the virus would mutate, either becoming a lethal strain, or changing enough from the pathogenic strain that it would no longer provide immunity. A number of tests clearly demonstrated that there was no genetic modification of the virus during growth.
The most important part of the paper IMHO was in the discussion. I produce an except here.
Usually, the immunizing dose should not exceed 106.5 to 107.0 EID50 per 0.2 ml for the vaccination with a live CA vaccine. The vaccine is introduced through the nose, 0.2 ml into each nostril. Therefore, some 108.0 EID50/ml is needed for a single vaccination. In our experiments, the highest output of CA reassortants in the fermenter during the titration of non-frozen samples was 1010 to 1010.5 EID50/ml. Since infection titer reduces by 1.0-1.5 lg during the lyophilization of live vaccines due to freezing, the number of doses of mono vaccine in a 10-l fermenter will be more than 100,000.
CA, means cold adapted. These are viruses that have been adapted to grow in cooler nasal passages. This type of influenza strain is what goes into the Flumist vaccines and allows the virus to replicate in the nose. The virus used for these vaccines is also attenuated such that it does not cause illness in the patient, but raises a strong immune response. Flumist vaccines developed in chicken embryos have shown effectiveness close to 90%. Now extrapolating these numbers, it should be possible to produce enough vaccine for the US population (300 million) with 3000, 10 liter batches. Enough to vaccinate the world would require 60,000 batches, a daunting amount. Of course the sensible thing is to scale up the volume of the fermenter.
The production of a live CA influenza vaccine may be rapidly increased in case of an influenza pandemic if MDCK cell lines cultivated in a fermenter are used as the substrate for the production of the vaccine. In our experiments, for example, with the cultivation of MDCK cell lines in a 10-l fermenter, the seeded 250-350 × 103 cells proliferated by a factor of 8-10 within 3-4 days, which is sufficient for seeding a 100-l fermenter. Thus, the amount of vaccine can be increased 10 times within less than a week.
MDCK is an acronym for a type of mammalian cells. The ones used in this study were harvested from the kidney of a dog. Here the authors show that a 10 liter fermenter can be used to inoculate a 100 liter fermenter and this can then be used for vaccine production. Fermenters of 1200 liters are available for cell culture. If the process could be scaled-up to that size, only 25 batches would need to be run for the entire U.S. population. If we assume a the total number of reactors available in the U.S. is 10, it would take 2.5 weeks to make enough vaccine for the entire U.S. population. However, this does not take into account the time needed to isolate potential candidate flu strains, attenuate them and create the cold adapted strains. While this process is well understood, it still takes time.
I also have no idea how many fermenters would really be available for this purpose, but I am sure in a pandemic situation it would be more than 10. The good news is that the basic science is already done and what is left is the scale-up of the process. This is never easy and hurdles will need to be overcome. Are there enough MDCK cells? Can the process remain free of contaminants, even at 1200 liters? Will we be able to create the candidate vaccine strain quickly enough? What we really need right now is time to move this process forward and to stockpile tamiflu. Lets all hope we have enough time.
Next installment I will cover the recent research on the reconstruction of the 1918 virus and what it means for the recent avian flu outbreaks.