Monday, February 27, 2006

Taking stock, part I

Highly Pathogenic Avian Influenza (HPAI) A/H5N1 has now spread to over 30 countries (by the time you read this the current count of 31 may well have increased so I'm not going to use a false precision). For a decade it has caused loss, fear and trouble in southeast asia and China, but suddenly, in the space of four months, it has burst out and is threatening to gain a permanent foothold among wild and domestic birds in Europe, Eurasia, the Indian sub-continent, the Middle East and Africa. Along with the bird infections have come sporadic human cases in southeast asia, China, Indonesia, Iraq, Turkey and perhaps elsewhere.

So it is time to pause and take stock, not so much of what we know but what we remain unsure of. There's a lot. Given the long interest in the influenza A virus and the new powerful methods of modern molecular biology, there is a surprising amount of basic stuff we don't know. This is not an indepth treatment but more like looking at wildflowers from horseback. We won't dismount to examine them closely and many will go unidentified. Because there's a lot to cover we'll do it in two parts. Part II tomorrow.

The setting: Influenza A is predominantly a bird disease. Birds can be infected by most of the many combinations of the 16 H subtypes and 9 N subtypes. Other animals become infected less frequently, but pigs, horses, cats, dogs, ferrets, marine mammals and of course humans are all capable of being hosts to one or another of the subtypes (H and N combinations) [terminological note: there are three types of influenza virus, A, B and C. The influenza A type has a possible 144 subtypes, the 16 H and 9 N combinations. Within each subtype, genetic variations are strains. This terminology is frequently abused and the use of strain for subtype is distressingly common.] Since the immune system "sees" the H and N proteins, our immunity is to these factors. Human influenza virus has been almost exclusively to H subtypes H1, H2 and H3. With the emergence a few years ago of documented infections with other H subtypes (H5, H7, H9) concern was raised about what would happen if one of these "new" (to humans) versions became easily transmissible from person to person, since there would be no immunological memory to protect us. If the virus were also virulent, meaning capable of causing serious disease, the resulting pandemic could be devastating. Just how devastating is provided by the historical event of the 1918 H1N1 pandemic.

When the first human cases of the H5N1 subtype (which was also extremely virulent for terrestrial poultry) were described, the alarm bells went off that this might be the early warning of a possible pandemic virus. Not only was H5N1 extremely virulent for poultry, but also for the human cases, with mortality over 30% (6 of 18 cases). Draconian measures of bird culling -- essentially attempting to kill all poultry in Hong Kong -- seemed to snuff out the threat at its origin. In the years that followed, human infections with H7 and H9 subtypes were also described but none had the virulence of H5N1. Then in 2003 H5N1 reappeared in Hong Kong and southeast asia. Thus began the outbreak which today seems to be spreading globally.

Here are some obvious questions. How is this bird virus spread from locality to locality? Are there effective ways to halt its spread among birds? What determines when an influenza A subtype will jump species boundaries and start infecting a new host, i.e., what determines host range? In particular, what determines if humans will become infected, and if infected be able to pass it easily from person to person? Why is H5N1 so virulent in people? How is influenza passed from person to person? How do we make a vaccine against it? What about antivirals? Will drug resistance emerge?

These are obviously significant questions but they don't exhaust the list. There are many others, I am sure, but it's more than enough to make the point I'm after. There remains a great deal about some important things we don't know.

Geographic spread: It now seems that wild and migratory birds can be infected and that some at least can harbor the infection but not be too sick to travel long distances. Unlike the asian case where the virus was discovered by its high mortality in poultry, in Europe it is the wild bird populations that are showing the first signs of local infection. But how does this virus get around: local poultry movements (trading, smuggling, etc.)? wild bird migrations? longer distance movement of poultry via trains, as was recently suggested? some other way? Do wild birds give it to poultry or do poultry give it to wild birds? If spread is a combination of these factors, what is the predominant mode of spread and are there any critical points where we could intervene to stop or slow spread? Many people think they know the answers to one or more of these questions but they don't agree with each other. The spread to a commercial turkey farm in France last week is especially puzzling as allegedly there was no chance of contact with wild birds. There are underlying political and disciplinary issues involved, as well. (e.g., the bird conservancy scientists versus public health scientists). Related to this question is whether it is small backyard holdings or factory farms that are the true incubators, and perhaps origin, of the HPAI strains of H5N1. This has implications for the economy and for control measures. Everyone has an opinion.

Control of spread among birds: Several main methods are in use. Mass cullling over a large area. Quarantine combined with mass culling within an area. Selective culling. Vaccination of birds. It is an open question as to the effects of mass culling in slowing the spread but it obviously is unable to stop it. Even the attempt at total extermination on an island venue (Hong Kong) didn't do it. In my view the option that makes the most sense at this juncture is selective culling of flocks where there are infected birds as a way to prevent infected animals from getting into markets. But it is easy to think of the weakness of this method, too. Vaccination has the virtue of being able to protect the birds from dying, but there is substantial evidence it doesn't protect them from being infected, after which they remain infective but shed virus at a considerably lesser rate. Thus this is unlikely to stop the spread of disease and may make it more difficult to track it since the birds will not be obviously ill nor will we be able to detect past infection by looking for antibodies (since all the birds will have antibodies if vaccinated). That's the best case, and assumes the bird vaccines are effective and effectively administered. In a mass vaccination situation with quickly ramped up production and many untrained vaccinators, this is not a safe bet. The result could be ineffective protection, with the virus being trekked from farm to farm with the vaccinators. Thus we still don't know the best way to control this disease among poultry, even if it weren't a public health problem but "just" an economic one.

We'll finish our stock taking survey in Part II.