?This virus is very unusual in its high virulence for a broad spectrum of animals that occasionally includes humans?
Earl Brown, professor at the Faculty of Medicine of the University of Ottawa, Canada, reviews for Scitizen what we know and what we ignore about the Avian Flu.
During these last months, weíve assisted to the spread of the H5N1 virus across much of the globe. For the time being, itís primarily a bird disease. Whatís the situation exactly?
So, maybe I should backtrack a little bit just to put the H5N1 situation in prospective with other highly pathogenic avian influenza viruses. The first report of a highly pathogenic bird flu was in 1878 in Italy. This was followed several years later by the first isolation of a HPAI virus, A/Brescia/1902 (H7N7). Since that time thereís been outbreaks of highly pathogenic avian flu in many parts of the developed world, Europe, North America, Australia, and South America. This is the first time that we have highly pathogenic avian influenza in Asia, and this is probably related to changes in farm practices: the viruses that originate from the wild bird population world is not pathogenic - the reservoir of avian influenza in ducks doesnít cause disease - but when virus of the H5 and H7 serotypes get into a farm situation where there is a large number of susceptible hosts packed together closely then you can get the rapid evolution to high virulence.
This is a function of high dose transmission (large populations that contain all possible single nucleotide substitutions) that can be experimentally achieved by serial passage in mice where high dose transmissions generates highly virulent variants. This seems to have caught up with Asia recently. They started to have highly pathogenic avian influenza in the early-mid nineties. The Chinese experience is unusual because the previous remedies for the outbreaks was not effective- possibly because a window of opportunity had passed or alternatively this ďwestern solutionĒ was not applicable to their situation. In the past when highly pathogenic avian influenza showed up in a farm they killed all the poultry, cleaned up the facilities and started again. Although this process can take a period of months itís never failed to control the outbreak (although the Mexican trails with HPAI H5N2 avian influenza have been somewhat protracted since the mid 1990ís) whereas in Asia, since 2003, the virus has spread into new countries that are beyond Chinaís borders. Itís been very difficult to control and we are getting more new countries added to the list of those infected with the HPAI H5N1 Z- lineage as time goes on. Very few countries in Eurasia are now free of the virus. This is unprecedented as it hasnít been seen in history before, where highly pathogenic avian influenza has spread so extensively and is gotten back into the wild bird and is sort of cycling back and forth between farm poultry and wild birds. Itís quite unusual in that respect. Moreover, the H5N1 virus itself, the Z lineage that weíre talking about here, is unusual too because itís altering its genetics as time goes on, and in certain forms is killing ducks, in certain forms is not, in most of its forms it can kill chickens as well as all sorts of other avian species. This general ability to infect and kill a broad spectrum of avian and mammalian species is very unusual even in highly pathogenic avian influenza virus which can kill one or two species but not this broad range of poultries or wild birds. This H5N1 can also kill a lot of experimental as well as wild animals: itís killing some wild carnivores in nature, it can kill mice, ferrets but not all animals; H5N1 Z infects pigs and sickens but does not kill. So this virus is very unusual in its high virulence for a broad spectrum of animals that occasionally includes humans that have been dead-end hosts; up to this time.
The H5N1 virus can infect humans and other mammals, but with some difficulty. Why?
Thatís a good question. Experimentally you can infect some other animals quite easily like cats and mice. For humans we have no experimental data, obviously, but the natural data suggests that it is difficult for humans to be infected, but that being said, once theyíre infected thereís a high rate of death (>50%). So there appears to be some basic genetics barriers. From a virus stand point, a life cycle always begins with receptor binding to susceptible the cell, the specificity of the virus receptor interactions are a component of host range, and so itís known that human influenza viruses bind to receptors that are slightly different from those sialic acid containing receptors that are in birds. In humans this H5N1 Z virus binds better to the cells in a region deep in the lung instead of the normal site of human influenza virus infection that targets infection to the epithelial lining of the major and minor airways. This virus seems to be not suited for infecting the upper respiratory lining and trachea of humans, and this is something that will have to change for H5N1 Z to become a human pathogen. But there is some concern since mutant virus with enhanced human receptor specificity has been seen repeatedly, most recently in Turkey and also before in Vietnam and in Hong Kong in 2003. The virus has shown some sign of ability to mutate towards increased infection of humans epithelial linings but it hasnít done this in a stable fashion yet and all human evolutionary process have been aborted. The fear is that a chain of human infections may allow the evolutionary process to proceed to the point of selection a human adapted variant. The most recent reports form Indonesian chronicle 2 cycles of human to human transmission that is the first glimpse of a situation that may drive human evolution.
The virus has killed half of the nearly 200 people infected so far. Why is it so virulent?
Itís related to the receptor specificity issue. The good news is itís hard to get infected with this virus, the bad news is that itís infecting a bad part of your lung (with respect to induced pathology), which is likely to give you a severe pneumonia with a high consequence of death. Further to that is that this virus is not limited to lung. Normally influenza virus only infects the lung airways being associated with and the tracheobronchitis, it doesnít goes into the body. This H5N1 causes a systemic infection: itís been found in brain, travelling around nerves as well as in blood with targeting of the intestine to shed virus into feces (these pathologies parallel avian infection with HPAI influenza). So itís infecting a dangerous part of the lung as well as attacking key organs. Another feature of this virus is that it induces a cytokine storm possibly as a function of its resistance to IFN and its high replication abilities. The cytokine storm result in atoxid shock-like syndrome associate with sever infections. This virus is much more aggressive and pathogen, it infects more tissues and causes more damage, that your normally see for a typical influenza infection in the human. The high virulence is a function of enhanced ability to replicate in host tissues and furthermore this enhanced replication ability extends to a broad range of hosts; therefore the problem associated with this virus to infect new species.
The virus has a poor human-to-human transmissibility. Have we any idea of if and when a pandemic could strike?
Presumably, weíll have to go through some particular cycling of accumulating key mutations by some series of human-to-human infections, and right now, we donít know exactly how many mutations are needed to cross this barrier into human pathogen. We also donít know the exact nature of those mutations. However we are seeing adaptive events occurring in the virus in human infections. If a person dies and doesnít transmit the virus, then thatís the end of the evolution. But if you can have the virus evolving in the first people infected (as is being seen) then starting to spread to other susceptibles, the process presumably could continue. So what youíd need for a virus to become adaptive to humans is some cycle of human-to-human transmission for some period of cycles, I canít tell you the number. I can only give you an indication: when I take a human influenza virus and turn it into a mouse adapted variant I have to go through mice about 12 to 20 times and so it may take similarly 12 to 20 passages to turn this H5N1 virus into a human virus.
As we have never seen this process in action for humans this is at present a matter of speculation.
In an editorial of Science, Peter Lu stresses that we urgently need faster and more robust diagnostic test for field use. How far are we from that?
We have rapid diagnostic tests to characterize an influenza A virus infection and rapid diagnostic tests to tell you it is a H5, but what we really need is a diagnosis to tell you the strain. Whereas they have diagnosis that work in laboratory settings that are able to do this, theyíre a little bit away from developing something for the field. Just recently, a company (Combimatrix) developed a DNA chip technology for strain and serotyping and also for sequencing influenza. Whereas the sequencing is not 100 percent perfect yet and need to be improved, the hope is that youíll have a micro chip technology, that you could plug your isolate into and get detailed strain information for out in the field. Weíre not there yet.
Do we have efficient drugs against H5N1 virus?
No we donít. Thereís 2 parts to that question. We donít have enough of the drugs that we have identified right now. There are 2 inhibitors. The Adamantanes, developed and licensed for H2N2 viruses in 1968. Viruses however, become resistant to this class of drugs very quickly. If you treat a patient with Adamantanes HCl, 2 or 3 days after youíve started the treatment you already start to see resistance in the virus population. So thereís a serious problem of resistance to this drug, but this drug is very cheap and itís been used quite a lot in North America. The resistance of human strains is quite high in NA. The original H5N1 Z isolates in 1997 and 2003 were also resistant. This raises the question: were Adamantane derivatives used against the bird flu in China? The Chinese have never admitted to this, though they have stated that they will not use this class of drugs on poultry now or in the future. The other class of drugs are the neuraminidase inhibitors. Thatís Tamiflu and Relenza, the two main products. Whereas the resistance may be a little bit slower, it seems the virus becomes resistant to this drug quite rapidly but the preclinical data indicates that such variants are attenuated for growth. The general feature of influenza nativiral drugs is that we donít have enough of them. If you look at drugs used for viruses, HIV for example, you canít use one drug at a time because the viruses are able to escape the inhibition. They have to use 3 drugs at a time, and then it becomes mathematically impossible for the virus to accumulate 3 types of mutation simultaneously. So we need something like that for flu. Now weíve got 2 drugs that have already been used and thereís some resistance in the environment. So what we probably need is 3 new categories of drugs at least so we can have a multi drug therapy. I think weíre quite slow on this.
It seems that some of these drugs are expensive and slow to produceÖ
Exactly, both Relenza and Tamiflu are very expensive. Just recently a paper came out showing a cheaper organic starting point to be used for the chemical synthesis, so hopefully that will make Tamiflu cheaper but that actually hasnít happened on a production scale yet.
What about a vaccine?
You have the standard problem of strain matching. We have the current H5N1 Z strains on the globe and they could make a vaccine against any one of those. Some countries have. The US has made quite a large batch of vaccine against the strain that was present in Vietnam in 2004. But a vaccine against one strain wouldnít be very effective against another. So right now, people are wondering how many vaccines you would have to make in order to be ready for any of those if they become a human pathogen. The standard approach is you wait until you have a pandemic, you take the virus, you make the vaccine out of it so that you have a good strain match. The problem is that that process takes 6 to 8 months. A lot of companies are trying to make better vaccines. People would like to make a broadly reactive vaccine so you would not have to worry about exactly what strain it is ie a serotype specific vaccine.
The other problem of the vaccine is that we have enough vaccine production to make vaccines for 5% of the world population.
Another problem that has turned up with trials with H5 vaccines is that they are poorly immunogenic requiring 6 times the dose to perform even less well (50% seroconversion) that conventional vaccines that employs 15 ug.
So I think vaccine-wise, weíre not in great shape.
What is your feeling about the Avian flu threat? Do you think we all overreact?
Itís difficult to say, influenza is my speciality, I know some of the possibilities of influenza infection form the past and of course I remember that it caused one of the worst pandemics in history in 1918. Influenza is the respiratory pathogen that is the most likely to send you to the doctor or to the hospital each year when we have annual influenza outbreaks. It is the most serious of the respiratory pathogens. And when you put the H5N1 in context, itís a virus thatís broken so many rules. The problem is that our ignorance is too great to understand the whole biology of influenza. When we see H5N1 Z, a virus that seems to ignore the species barrier largely and is virulent for so many different birds and mammals, itís very unusual, so that makes us concerned. To the next step: whatís going to cause the next human pandemic? We canít say that. We donít know whatís going to happen. So are we overreacting? Itís hard to say. I think we were underreacting to infective diseases in general in the last several decades, so itís probably a good wake up call. If one steps back and look at the number one killers, cancer and heat disease data is accumulating that infections may be the primary causes of the affliction. It was a mistake to declare the battle against infection to be won in the 1960ís when vaccines and antibiotics were able to control most of the infectious disease. Since then the re-emergence of old foes like TB and new ones like HIV plus many others we have to redress or attack on infectious disease. As disease is a function of host density we are now in unprecedented levels of human and animal infection. We may learn bold lessons again and hopefully the lesson wonít be too hard?
We should be ready for influenza in particular but we should be ready for another new emerging virus, because weíre getting usually one or two new viruses coming out of nature each year; SARS didnít have staying power but it had features you donít want to see in a virus with staying power. The world is changing its environment and weíre getting the microbes with their rapid adaptability being able to rapidly evolve in this more crowded world weíre living in, so I think we should expect more new viruses and whereas some people may be criticised for overreacting to the H5N1, itís probably a good practise to get our preparedness plans in place so that if there were to be a new epidemic disease then we have plans ready and we can act in a coordinate fashion. So itís probably overdue to have this preparedness exercise for an unknown infectious threat and whether this H5N1 is going to come into people is still a big question, but it surely cannot be ignored as a demonstrated threat to birds and animals as a plausible although at this time a theoretical threat to humans..
Prof. Earl brown, thank you.
Earl brown works at the Department of Biochemistry, Microbiology and Immunology of the Faculty of Medicine of the University of Ottawa, Canada.Interview by Gilles Prigent