Monday, May 01, 2006

"Benign" bird flu in the UK

Three poultry farms in eastern England (Norfolk county) have been afflicted with avian influenza, subtype H7N3. Since this is not the H5N1 subtype now rampant in birds in some 40 countries and over 200 cases and 100+ deaths in people, it is being described as "less dangerous to humans." Possibly. The fact is that H7 subtypes H7N7, H7N2 and (as here) H7N3 have infected human beings in the past and there is currently one confirmed infection in this outbreak. Three workers suspected of being infected are reported to have tested negative, but the three cases all had conjunctivitis ("pink eye"), a known presentation of H7 infection in humans, strongly suggesting the "negatives" are false negatives.

Since the consequences of infection are mainly benign, self-limiting cases of red eyes, tearing and itching one would think there is nothing much to worry about in any case. But there is more to it. In the largest outbreak to date, the H7N7 episode among poultry cullers in The Netherlands in 2003, the virus showed itself readily transmissible from person to person and fatal in one case, a 57 year old veterinarian who succumbed to the kind of viral pneumonia typically associated with H5N1. While the predominant symptoms in the 89 symptomatic Dutch worker cases was conjunctivitis, there were also seven cases of influenza-like-illness (ILI), defined as sudden onset of fever, muscle aches and pains and cough, runny nose or sore throat (see paper by Fouchier et al. here). Mask and goggle use by cullers seemed to have no effect, but Tamiflu did. Using antibody evidence of infection, moreover, an estimated 1000 people were infected and those infected passed on the infection to 59% of household contacts (Science 306:590, 2004).

The Dutch H7N7 was genetically similar to a low pathogenic H7N3 and H10N7 in Dutch ducks but had an HA cleavage site characteristic of high pathogenic strains. The suspicion was that the virus might have come from wild birds and mutated to a high path form in chickens. The virus from the fatal case was isolated as well and showed 14 separate amino acid changes in various genes when comparing it to the virus isolated from the chickens. This included the E627K mutation in PB2, characteristic of many H5N1 strains, as well as four other PB2 changes (see Table 1 in Foucier paper). By contrast, the virus isolated from the conjunctivitis cases has only one amino acid change (in the NS gene). Thus the H7N7 mutated significantly even during this outbreak but dead-ended in the fatal case.

An interesting paper by Olafsson et al. in 2005 (abstract) speculates that the conjunctivitis in the reported outbreaks of H7N7 and H7N3 might be related to the fact that the tissues covering the eyeball (the conjunctivae) have α2, 3 linked sialic acid receptors on them (the kind that bird viruses latch on to), but the tears and fluids bathing the eye have α2, 6 receptors, thus protecting the eye from human but not bird viruses. The situation in the human upper respiratory tract is just the reverse. The cells have α2, 6 receptors but are covered with a protective mucin layer with α2, 3 linked sialic acids. Olafsson et al. also wonder if the eye might not be an efficient way for bird viruses to make their way to the nose and throat through the nasolacrimal duct (tear duct), possibly allowing either adaptation or reassortment there with co-infecting human influenza viruses. Thus I don't feel especially comforted by the Dutch and UK examples of a supposedly benign bird flu virus. The H7 subtype seems readily transmissible and also has the potential to change to a highly virulent form in humans, although it didn't have both characteristics at once.

The UK cases also carry another warning. Because of the recent H5N1 cases in a Scottish swan UK poultry farms are on high alert and presumably practicing good biosecurity. Despite this, avian influenza struck one poultry farm and spread to two neighboring farms. The mode of spread is currently under investigation, but claims by the US poultry industry that they are unlikely to be affected even if H5N1 arrives here on wild birds because of their biosecurity measures sound rather less reassuring.

Influenza finds its natural host in birds. We are now seeing it spread from birds to humans and other species, where it is changing character. The reasons for this are not known, but suspicion must certainly fall on the huge factory poultry farms that jam birds together under unhygienic conditions made to order for epidemic disease in birds.

And from them to us.