H5N1 influenza: some troubling preliminary signals
A few data sources to keep an eye on
Sometimes three data points can tell us a lot. Back in early May 2021, the REACT study in the UK reported the results of genomic sequencing of three newly detected infections in London. Unlike routine COVID case data, these infections had been identified among people randomly tested in the community. And two of the three turned out to be the Delta variant.
At the time, it was still unclear how much Delta was spreading in the community, as opposed to just reflecting imported cases from India. But thanks to REACT, we knew two out of three randomly detected infections were Delta in London. It wasn’t a definitive signal, but it was an important one.
Which brings us to the current H5N1 avian influenza situation in North America. In recent weeks, there have been three unusual cases reported. In September, there was a hospitalised case in Missouri who had no known contact with animals. Then, earlier this month, a severely ill case in British Columbia without a clear source of infection. And finally, an infected child in California reported last week, again with the source of infection unknown.
When assessing potential epidemic threats, unusual cases like these may mean little, or they may mean a lot. So it’s important to also look at other emerging evidence as well. Here are a couple of recent findings that have also stood out:
Scott Hensley had a useful thread on Bluesky that discusses evidence of a potential mutational change at residue 226 of the HA surface protein of the virus from British Columbia. This is notable, because a study earlier this month showed that mutations at this location can change the preference of H5N1 to bind to the α2-6 sialic acids on cells in the upper airway (as it does with human adapted human viruses), rather than α2-3 sialic acids found on cells deeper in the lungs (as avian viruses historically do). In short, such a mutation could mean the virus is becoming better adapted to humans. In the comments, Jesse Bloom also points to a potential change at residue 190, which is also known to be important for which types of acids viruses bind to.
Sam Scarpino also had an interesting thread on Bluesky looking at wastewater surveillance in California. He noted that several urban areas are seeing an increase in influenza virus concentration in wastewater alongside an increase in the frequency of H5 positive tests among the water samples. This could be evidence of undetected human infections (which we know can happen from antibody studies among dairy workers). H5N1 has recently been found in raw milk on sale in California, which may be one possible route of exposure.
These are just early data points, of course. But the challenge is that outbreaks that don’t go anywhere can generate similar early signals to outbreaks that lead to major epidemics. As with every emerging disease threat, we therefore need a better understanding of what we’re facing, and a clearer plan for what to do about it.
Cover image: CDC via Wikipedia.
If you want to read more on global (lack of) pandemic preparedness:
