H3N2 in Australia – Australia’s unseasonal flu season rolls on

Posted onAuthorIan M Mackay, PhD (EIC)1 Comment

Australia’s unseasonal flu season is still with us, albeit a little quieter than it was. Some jurisdictions have already had a peak and have returned to almost normal numbers of flu cases. Some areas are still rising and others are higher than normal for this time of year, but stable. Australia is a big country. Influenza really highlights that a flu “season” is more a series of separate outbreaks that happen to co-occur. It just takes environmental, human and virological stars to align before a season takes off. So what’s new this week?

In 2019 we don’t know much about what the stars are that have aligned, just that they have. Nothing has described any likely contributing factors to our early flu surge. I’ve written about some things that probably aren’t drivers here. One thing that does really stands out though, is the lack of local research on the circulating flu viruses. My money is on them being different in some way. I‘m happy to be proven wrong, but that will take some data.

A harbinger of some H3N2 mismatch

Just this week a new preprint dropped. A preprint is a scientific paper that has not yet been peer-reviewed, but is open for everyone to read. It provided data suggesting Australia may have a difference between circulating and vaccine viruses. Some of the A/H3N2 viruses dominating Down Under in 2019 (members of a subgroup or “subclade” called 3c2.A A1b/131K) differ from the vaccine component virus chosen to protect us from them (a subclade A2/re virus, A/Switzerland/8060/2017).

We identified a potential antigenic mismatch of the H3N2 component of the 2019 Southern Hemisphere influenza vaccine. Nonetheless, it is important to continue to get vaccinated.@MackayIM@edwardcholmes@WHOCCFluMelb @sjturn
https://t.co/6N8vKlchrW— Hensley Lab (@SCOTTeHENSLEY) June 25, 2019

My great thanks to fluru (flu guru?) Scott Hensley, University of Pensylvania, for pinging me about his article on Twitter. And for doing this important work!

There were already signs locally that some wild A/H3N2 flu viruses have drifted away from the vaccine component expected to protect us from them. Future human studies will make this more clear.

No sign of recombination yet

What else do we know? I downloaded 23 Australian complete genomes found on the GISAID EpiFlu sequence database. These sequences come from flu viruses detected in humans from around Australia. There are none from Queensland, Western Australia or the Australian Capital Territory. These have had all 8 of their separate genetic segment’s sequences determined using a next-generation genetic sequencing technology.

A phylogenetic tree showing the groupings into which the hemagglutinin (HA) genes from all the complete genomes sequences from Australia between November 2018 and 25th June 2019, cluster. The clade and subclade names are shown at eth right-hand end of each branch and the State/Territory of origin of the infected person is shown using coloured dots. The relationships were inferred using a Neighbor-Joining method performed MEGA version 7. The optimal tree is shown. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (50 replicates) are shown next to the branches [2]. The tree is drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances were computed using the p-distance method [3] and are in the units of the number of base differences per site. The analysis involved 24 nucleotide sequences. All positions containing gaps and missing data were eliminated. There were a total of 859 positions in the final dataset. The tree image was copied into Inkscape 0.92.3 and prettied up.

This sample of genomes is less than 0.02% of all the flu cases recorded between November 2018 and June 25th. This means it’s not representative of anything specifically; there are too few and we don’t know why they were selected. But it’s interesting. It does give us a glimpse that there is lots of A/H3N2 viral diversity around. And this seems to be a growing feature of our flu seasons.

The main reason I created this tree was to compare it to trees made using the other 7 segments. I’m not showing all of that working here – just two examples below. I’ve compared the HA gene segments to the neuraminidase (NA) gene segments or to the nonstructural (NS) gene segments.

See more explanation about influenza genetic segments and genomes at An influenza virus is the sum of its parts

I can tell you that the same pattern resulted from all 7 comparisons. The same subclades within HA appeared in the trees made using the other 7 genes.

Not a tangled forest

There was no sign in these “tangle trees” that one of the H3N2 viruses had swapped a gene segment with another FluA virus. If they had it would have shown up as a line from the left (HA) side matching up to a different coloured partner on the right. This process, called recombination, could result in a child virus with different properties from either parent. Perhaps one that could survive or transmit more efficiently during summer, for example.

Neighbor-Joining phylogenetic trees were made using Geneious Tree Builder with the Tamura-Nei distance model in Geneious v8. Tree images were paired, aligned and lines added using Inkscape 0.92.3.

The hypotheses

Australia’s unseasonal flu season has a couple of main drivers in my opinion. One hypothesis is that a distinct virus arising from genetic recombination kicked off the early flu season; co-infecting viruses sharing their gene segments. It wasn’t surprising that this little experiment didn’t find any evidence to support that. We’ll need more sequences to be able to really test that.

Another hypothesis is that there have been mutations within one or more of the gene segments. This or these perhaps bestowed upon the mutants improved virus stability or transmission. The mutants might last longer on surfaces for us to pick up or be hardier in droplets, floating longer for us to inhale. Finding evidence of that will also take more full genome sequences. It’ll also need analysis and knowledge of the impact of changes to the viral proteins resulting from genetic mutations.

Reasons to conduct flu surveillance from 1999. The technology has changed dramatically since then potentially allowing us to detect virus earlier, faster and to know more about what we’re detecting. The reasons to use better methods remains as solid as ever.

Both hypotheses will need us to look more deeply at influenza than we have to date and to do that more regularly. And we should communicate what we find, as we find it instead of waiting for the high impact factor journal.

We need more viral genomics

We should consider working whole genome sequencing into flu surveillance with as much gusto as we cling to our beliefs that everything we need to know about seasonal influenza can be found in 12.5% of its genes.

There’s still lots of work to do in the field of seasonal influenza. Australia’s unseasonal flu season is just the latest reminder that we often settle for not knowing rather than seeking out precise answers.

1 thought on “H3N2 in Australia – Australia’s unseasonal flu season rolls on”

  1. This continue to be interesting an analysis, but as recent WHO summary (Update No. 344) pointed out, earlier and steep increases of flu activity are being recorded in various countries and region across Southern Hemisphere:

    ”In the temperate zones of the southern hemisphere, influenza detections continued to increase. The 2019 influenza season appeared to have started earlier than previous years in Australia, Chile, South Africa and New Zealand. Influenza A(H3N2) viruses predominated in Oceania and South Africa.Influenza A(H1N1)pdm09 viruses predominated in temperate South America. … Costa Rica where influenza A viruses activity was high.”

    Different subtypes but a similar pattern. Could a novel reassortant be behind this epidemiological behaviour? In 2009, when the H1pdm09 emerged, Mexico first epidemic had been followed by a global surge in flu activity within a couple of months. This is not that we are seeing in 2019.

    It remains to be clarified:

    1) SARI cases increased too?
    2) Excess mortality for all causes increased?
    3) Age-specific mortality rate has changed?
    4) A&E accesses for P&I are above level expected during severe seasons?
    5) Severe cases presented with primary viral pneumonia or with subsequent S. pneumoniae infections?
    6) ICU patients developed antibiotic-resistant super-infections? (A feature of H7N9 HPAI human cases in China for example)?

    We’ll see next weeks behaviour…

Comments are closed.