The news service at Nature is reporting ("Bird flu vaccine not up to scratch" -- subscription required) that an egg-based whole virus Avian Flu vaccine, recently announced with fanfare as solving all our problems, is unlikely to be useful.  Fortunately, there is an alternative.  Alas, regulatory issues may prevent PowderMed from distributing it's DNA vaccine for the Avian Flu for some years yet.

The whole virus vaccine was announced with great fanfare just a few months ago.  But as I've written previously (here and here), egg-based vaccine production will never be sufficient for rapid responses to quickly evolving viruses like the flu.  Moreover, to produce a decent immune response the whole virus vaccine must be administered in 2 doses, each 6 times larger than an annual flu shot.  While this is in part because humans have never been exposed to an H5 virus (we are "immune naive"), it also appears that it just isn't a great vaccine.

While it is true that enthusiasm for DNA vaccines has gone through a bit of a boom and bust cycle, early results requiring high dose intramuscular injection are not representative of how the current technology works.  Genes coding for antigens for new viruses are slotted into a plasmid vector that has been proved safe in humans, the plasmids are loaded onto micron-sized gold particles, and the particles are injected into the skin using a high-pressure helium blast.  At Bio-ERA, we've been studying the vaccine and its utility, and it looks like the real deal.  An article in Red Herring quotes the CEO of PowderMed as saying;

What we really believe we’ve got is not just a vaccine; we actually have the ability to produce a capability for a country to cover anything really.  We have designed, with the help of contract manufacturers, a facility that would be able to produce 150 million doses in three months.

The key to the value of PowderMed technology is that the DNA vaccine is delivered directly in the nucleus of dendritic cells in the epidermis.  By getting dendritic cells to express coat proteins from pathogens and then present those proteins in complexes with MHC molecules, the vaccine directly stimulates a cellular immune response; T-cells are thereby primed to recognize and dispose of the virus and infected cells.

Vaccine production in chicken eggs or in cell culture requires at least 6 months to even begin cranking out doses, and requires significant infrastructure to do so.  PowderMed suggests that within three months of sequencing a new pathogen they can have vaccines ready to go.  But my estimates suggest it could be much faster than this.  Included in PowderMed's estimate is the time required to load the vaccine into their proprietary delivery system (a helium powered injector about the size of a flashlight).

My own estimate of the time required to fabricate the plasmids, followed by enzymatic amplification, is more like a week or two.  Injection of the vaccine does require particular technology (a "gene gun") but as it happens those have been used for ~10 years to genetically modify plants and animals.  There are gene guns scattered all across the developed and developing world.  If we had to, if the Avian Flu started to cause real problems in the human population, we could synthesize the vaccine in a widely distributed fashion (anywhere around the globe where people have access to large scale DNA synthesis) and deliver it using gene guns.  True, those instruments were intended for research use only, and were not designed for (or at least not marketed for) use on humans.  But if things start to go south, I'll be first in line.