"Not unexpected" in swine flu is not unexpected

It seems swine flu is full of surprises that turn out not to be surprises. Or so it's claimed. Or not. Here is CDC's Dr. Anne Schuchat, the agency's chief health officer and spokesperson on swine flu, responding to NPR's Melissa Block's question about what has been her biggest surprise:

Dr. SCHUCHAT: I shouldn't have been surprised, but I have been surprised about this disproportionate toll that it's taking in pregnant women. I think I'd never lived before a pandemic before, and I actually hadn't seen the really sorry and just the tragic stories of healthy pregnant women coming down with such difficult diseases.

So that's been hard to take as a public health expert, and, of course, a reason that we strongly recommend pregnant women who have cough or fever, respiratory symptoms, to take that seriously and seek care and get antiviral medicines and, of course, to be vaccinated before they get such symptoms. (National Public Radio, interview November 25, 2009)

I like the way Schuchat answered this. Flu is notoriously unpredictable and it can do almost anything, including things we have seen in the past or expect it to do on the basis of reasoning. With so many plausible possibilities, almost any outcome will be deemed in hindsight to be unsurprising. In hindsight. Even so, we are (properly) surprised, and she (properly) admits to it.

Yet every time something vaguely worrisome happens -- like the emergence of antiviral resistance -- it's dubbed "not a surprise." For antiviral resistance, there's nothing surprising about it and a couple of dozen reports of Tamiflu resistance are already on record. Almost all, however, emerged during treatment and were not "epidemiologically linked," i.e., didn't show evidence of having been transmitted from one person to another. The resistant infections "dead ended." This is an important issue because the property of being resistant is separate from the property of being able to be passed easily from person to person. Resistance doesn't spread until you have both.

The virus is already highly transmissible, of course, so it would seem that only one mutation is needed, but as we've discussed here before, the eight genetic segments of the flu virus work together as a team, and when one team member changes character in a way that seems by itself to be beneficial to the virus it often happens that it degrades the overall objective of making copies of itself. That's one of many reasons why predicting what flu will do by just looking at the genetic sequence is a chancy affair. An apparent, although not conclusive, instance of a transmitted resistant strain was reported by CDC in July involving two young summer campers in North Carolina. Within the last week two further clusters indicating person to person transmission, one in hospitalized patients, again in North Carolina, and another in five hospitalized patients in the UK (Wales). So while there have a couple of dozen cases of resistant isolates previously, now it seems we are seeing clusters indicating there is a resistant form that is also transmissible. That's a change.

And in the last day or so a sixth person has been added to the UK cluster:

The sixth patient is linked to five people who last week became the world's first confirmed cases of person-to-person transmission of such a strain.

Another person in direct contact with the group at University Hospital of Wales, Cardiff, has been tested.

In the Tamiflu-related cases, three patients remain in hospital, with one in critical care and the other two treated in isolation.

[snip]

Dr Roland Salmon, director of the National Public Health Service for Wales' Communicable Disease Surveillance Centre, said it was "not unexpected" that more patients on a unit treating people with severe underlying health conditions would test positive for the Tamiflu-resistant strain. (BBC)

So it's not unexpected that it's "not unexpected" that a unit treating people with severe underlying conditions would test positive for a resistant strain, because "not unexpected" or "not surprised" is the norm -- after the fact. Presumably the patient's underlying medical condition allows the virus to replicate more quickly or more profusely, thus promoting the chances of a a sloppy copying mistake by the virus that accidentally makes it resistant. Presumably. And presumably the fact they may have been immunocompromised (some are reported to have suffered from blood cancers) suggests easier transmission because it lowers the barrier to being infected on the other end. So on reflection, not a surprise. The answer to everything.

After it happens.

More like this

The means by which these patients became infected seems to be the missing piece of information.

Presumably the campers in NC had direct contact. But what about the outbreaks among hospitalized patients in NC and Wales?

To me, this suggests spread by an infected hospital worker or at least poor infection-control practices.

Of course, because everything depends on everything else, a single point mutation can, by itself, both make a bug more pathogenic while making it less transmissible ...

I am hoping that you can help put some information that I've been pointed to from another site in context. - http://pf11.blogspot.com/2009/11/norway-sequence-with-225g-mixture-also…

The gist, if I understand it correctly, is this:

"225G on 5 sequences (Norway2924 and 4 flashfire specimens) all sharing a rare syn413K on 5 distinctly different backgrounds closes the door on spontaneous and / or random mutation in this dataset."

A separate uncommon mutation is, apparently, showing up with the 225G one against a variety of other backgrounds. So it seems to have traveled as a large chunk via recombination, it seems, widely. But then, to my read, it also must be not very virulent in most cases or it would be showing up more in more fatal cases all over. I think.

Maybe not surprising but my wee brain finds it all a bit confusing!

Don S: Alas you are wading into deep and contentious waters. It has been Niman's contention (and the basis of his commercial product) that the driving force in genetic variation in flu is neither point mutation nor reassortment but homologous recombination. I believe this was first broached by Gibbs in Australia but I haven't researched the history of the idea in detail. At the outset this seemed quite plausible to me. Why not? Recombination occurs with other viruses and there would seem to be plenty of opportunity when two different viruses co-infected a host cell (although with single stranded RNA there is no complementary matching). At any rate, there is good evidence that co-infection is quite common (see a recent paper by Ghedin et al., Mixed Infection and the Genesis of Influenza Virus Diversity, Elodie Ghedin, Adam Fitch, Alex Boyne, Sara Griesemer, Jay DePasse, Jayati Bera, Xu Zhang, Rebecca A. Halpin, Marita Smit, Lance Jennings, Kirsten St. George, Edward C. Holmes, and David J. Spiro: Journal of Virology, September 2009, p. 8832-8841, Vol. 83, No. 17, doi:10.1128/JVI.00773-09, open access here). So if this is true, point mutations, reassortment and homologous recombination could all be driving genetic variation and then it would be just a question of which one was the most powerful. Data like those cited in your link are compatible with all three although interpreted as recombination.

The problem is that the actual empirical evidence suggests that negative sense RNA viruses don't engage in homologous recombination to any significant extent (for a recent paper again confirming this, see Homologous Recombination Is Very Rare or Absent in Human Influenza A Virus, Maciej F. Boni, Yang Zhou, Jeffery K. Taubenberger, and Edward C. Holmes, open access here; note that these are among the most experienced and prominent flu scientists in the world, although that doesn't automatically make them right of course). Your earlier comment about the non-constructive aspect of "board wars" is related to Niman's small but ferocious contingent of followers (virtually none of whom are capable of understanding the science but at least one, a lawyer, has invested in his company). They believe on the basis of his contrarianism and oracular (and usually scientifically incomprehensible) style he must be a Great Man. The resulting invective and lack of collegiality has resulted in the demise of some sites as it underwent pro and anti-Niman binary fission or the outright banning of some commenters, including Niman. I recount this history with some trepidation as it will inevitably become the pretext for another round of attacks on us (the latest referred to by Jon Schultz in the comment threads). Right now the science is not on Niman's side (further proof to his acolytes that he is right, no doubt). I am satisfied to let the science of this play out in the peer reviewed scientific literature (to which Niman has yet to contribute on this subject; in fact his only peer-reviewed flu paper lists him as one of many authors -- in fact this paper is his only peer reviewed scientific paper on any subject since 1996 -- and it doesn't involve recombination). If it turns out he is right we'll know eventually and I will be more than happy to concede if that happens. I care about how the world works, not about who is right. But I won't let this site be degraded by uncivil, uncollegial and unsubstantiated claims that are related to a commercial product that so far has no support in the scientific literature. It is immensely destructive and I am trying to minimize even my discussion of it here by segregating it in a comment thread that few will read because of its length. No one else can comment on Niman's site, so this is a symmetrical relationship in his case. I'm doing this here, now, because you asked and others might also be wondering.

Aside from the Niman issue (and your link was to a Niman-promoting post), the latest paper by Ghedin et al. I cited above makes some important points that are worth emphasizing. Not only is co-infection of the same cells frequent (at least according to their data interpretation), but it points out that when we talk about a viral isolate from a single patient we are really speaking about a consensus sequence. Each infected individual has billions of viral particles and they don't all have the same sequence. I don't think most people appreciate this. Here is how Ghedin et al. put it:

The vast majority of studies of genetic diversity in acute RNA viruses have been conducted on an epidemiological scale, in which a single, consensus sequence is obtained from each infected individual by direct population sequencing. This sequence by necessity then describes the most common variant in the intrahost viral population, thereby masking a myriad of mutant sequences.

[snip]

Although consensus sequencing is valuable for many aspects of molecular epidemiology, it by necessity cannot shed light on evolutionary processes that take place within individually infected hosts. Our study has revealed a remarkably high level of mixed infection in human influenza virus, including diverse lineages of the same subtype that differ in both their propensity for drug resistance and antigenicity, and lineages of different virus types (A and B). A screening of the first 2,000 influenza virus samples published on GenBank for the IGSP shows that approximately 3% have some evidence of large-scale sequence polymorphism suggestive of mixed infection (unpublished observation). As published consensus sequences are necessarily skewed toward the dominant strain within isolates, this number is almost certainly a major underrepresentation of the true level of mixed infection. For example, we also found the clear signature of an H3N2/H1N1 mixed infection in an isolate (A/ Canterbury/247/2005) sampled in New Zealand in 2005, although the primary specimen was no longer available for analysis and we therefore could not confirm that this coinfection was present in the host.

In sum, we propose that mixed infection of diverse influenza viruses, a necessary precursor to reassortment, is a common occurrence during seasonal influenza in humans and will in turn accelerate the rate of adaptive evolution in this virus. In addition, intrahost populations of influenza virus will harbor genetic diversity generated by de novo mutation, which we have not assessed in the current study. As a consequence, we urge that intrahost sequencing be more routinely employed to assess the degree of genotypic and phenotypic diversity in populations of acute RNA viruses. With the advent of highthroughput next-generation sequencing platforms, viral variants are being much more explicitly revealed within specimens, and this type of data can be made available on a routine basis. [from first cite, Ghedin et al. in this comment]

While large scale polymorphisms could theoretically also be the result of recombination (a la Niman), a body of literature going back decades indicates that recombination isn't happening (I cite http://jvi.asm.org/cgi/content/full/82/10/4807?view=long&pmid=18353939-… as a recent example and readers can see it for themselves as it is open access and I have provided the link for both Boni et al. and Ghedin et al.). A blog or personal/company webstie is not the place to settle this nor is invective accusing someone else of spouting "nonsense." The place to settle this is in the peer reviewed scientific literature and that's where it is happening.

I regret having to recount this in this fashion because it brings nothing but grief to any blogger who raises this issue, but I wanted to answer you. The reveres have decided not to continue this discussion further in the comments or the front page, so I hope this is sufficient.

>

If trying to explain his to others, Is this process analogous to why a person must take their full course of antibiotics to kill off the stronger ones i.e. because they are different in small genetic way than the weaker first-to-be-killed-off bacteria??

BTW I cannot find any Ohio Department of Health, newspaper, or even forums that tally these things listing her as an H1N1 caused or related death, as in under-reporting -- perhaps.

Thanks.

I appreciate your placing this information in a meaningful context for me (and here I mean scientific context). I apologize in advance if my having asked for that incites any negative feedback from ... others.

Let me focus on the Ghedin et al work for a moment to make sure that I get it. They are saying that each individual is OFTEN infected with multiple influenza variants at the same time and that those variants no doubt are reassorting and developing de novo mutations within that individual before the most transmissible of the ones most successful at competing to reproduce within the host ones get passed onto the next hosts, correct?

Under that framework a mutation that leads somewhat lower transmissiblity (even if it succeeds in reproducing well deep within the host) may be included in small numbers in the next aliquot that infects a particular subsequent individual. Any many so called "rare" variants are actually not all that rarely present as part of the mix, even if they are not the majority member of the population so are left out of the consensus sequence. Also correct?

Interesting stuff.

Two additional comments -

In the paper that is cited as evidence against any significant role for homologous recombination - consensus sequences are used and the authors state that such might give a false sense of recombination as mixed infections are combined into one consensus read. But wouldn't the same averaged consensus read "average out" any evidence of homologous recombination, essentially reducing it to noise against the consensus signal?

That evidence each infected individual often has "a high level of mixed infection, including diverse lineages of the same influenza virus subtype, drug-resistant and -sensitive strains, those that are likely to differ in antigenicity, and even viruses of different influenza virus types (A and B)" makes me even more anxious as pandemic A/H1N1-SOIV (did I leave any aliases out?) travels to areas endemic for H5N1. Suddenly the swings at bat are multiplied by several additional orders of magnitude. Wow.

It also means that if this 225G mutation COULD (with the right supporting player genes) develop into a simultaneously pathogenic and transmissible bug, then it will have many many chances to do so, even within each infected individual.

Like the Grinch's heart my concern has grown two sizes this day.

Don: The question of detecting intrastrand recombination for these viruses is technically difficult. There are several lines of evidence that it doesn't happen in negative sense RNA viruses (not just flu) and wherever it is claimed there are more parsimonious explanations but like everything with flu, who knows? We know non-homologous recombination does occur, which makes more sense, as pairing can occur over a short stretch on another segment and that would seem to be harder with a single strand, but again, I never rule out anything with flu. With lots of co-infection and point mutation and reassortment going on there's plenty to drive genetic variation. You can look at reassortment as just another kind of recombination if you want since flu is unusual in the segments already being broken into pieces to allow the swapping that if it took place on one long strand of RNA would be called recombination. This is one reason I have never cared much about the Niman business because it is the variation that counts for me, not the mechanism in any particular case. I know there's a lot of variation and now we believe there is a lot of co-infection at the cellular level.

Each individual is really harboring a quasi-species and we are only seeing the consensus sequence, so there is a lot we are missing beneath the surface. That's been happening with flu since there was flu so that hasn't changed. But the number of combinations vast but the biologically realizable ones only a minute fraction of the total. Yes, the wheel spins many times, but the combinatorial possibilities far exceed the number of atoms in the universe so it isn't a given that the monkeys will type out Shakespeare.

I don't think we quite know how often co-infection occurs. This work is a first stab at getting a fix on that proportion, but since even when there is supposedly only one strain out there it varies quite a bit that I have to assume each of us can get infected with virus particles that differ from each other in various ways. Even if they are of the same subtype the reassortment is essentially a recombination and if you throw in point mutations that's a heady mix.

I'm not more worried than I was before. I've been worried about flu since the beginning. But for most people it's the devil they know and the new forms are the devil they don't know so it's more worrisome.

I'm not so worried about them typing out Shakespeare ... it's the typing out Stephen King I worry about!

Your points are well taken however.

In any case I still don't get (and apparently neither do you) what the proposed homologous recombination has to do with the alleged pathogenicity of this particular mutation.

It actually makes more sense to me that none of us are every infected with just a single virus and that any aliquot of viral population will actually more likely contain a variety of competing varieties and mutations traveling in the same droplet than a "pure culture". I just never thought about it that way before is all.

Thank you for taking the time to explain this.

I support the findings and discussions of Dr Niman. Are you going to BAN me from posting here.??

As a reader what I would hope the reveres would do is to allow reasoned discussion of ANY speculations or studies pertinent to a particular post so long as the posters remain respectful of each other and so long as it doesn't take over all discussions. And so long as the discussions stay focused on the ideas and the evidence and not the personalities involved.

I have participated in boards in which particular posters made threads function like those Onion columns, "Ask A {fill-in-the-blank}" - you may know them but if not here's an example - http://www.theonion.com/content/node/37376 - and it can ruin all discussions.

That sort of behavior should be a bannable offense (IMHO).

I can understand how they could want to erect a big wall if that has happened before. (And I have noticed other flu sites that also have that particular individual as one who shall not be named and who have what before hearing about this past have what seemed to be odd rules about not posting links to other blogs, etc.) You cannot let your site become taken over by a single issue.

But I'm just a reader who wants to able to read a variety of ideas and critical analyses on a variety of subjects posted by some smart folk. And I don't mean only the reveres - I mean many who post. I would hate to have those voices drowned out.