Lament from a flu outsider

Dear CDC,
Oh, if only you wouldn’t spurn me. Don’t you know we both want the same thing? Like you, I want the flu vaccine to work. I want everyone to get their flu shots. I know I’m an outsider, but maybe you could at least listen to my ideas? Is it my fondness for sequencing over serotyping? Or maybe my insistence that you should share all your sequencing data? Why not invite everyone to the party? It’d be much more fun. And think of all the flu viruses we could keep out.
Fondly,
Steven
The news this week has been filled with reports of a new flu outbreak around the country. My own university has reported over 400 cases, the worst flu season in many years. Those of you who’ve followed this blog may remember that I predicted this would happen in a November blog post. Part of me wanted to be wrong – I didn’t want a bad flu season – but it turns out I was right.

The CDC had a press conference on Friday (15 February) acknowledging that the vaccine was a bust, but they didn’t explain why they’d chosen an old strain of H3N2 (from 2005!) for inclusion in the vaccine. As I explained in November, they already knew it was a bad choice – or they should have – last year.

So what is happening out there? Well, from the CDC website we can see results from the week of Feb 3-9 (the most recent data they have) summarized in this figure:
The red cases are H3N2, which are clearly dominant. The blue cases are H1N1 (the vaccine works against that strain!), and the yellow are influenza A but not typed as H3 or H1 – most of these are probably H3N2 as well.

The raw numbers for the season tell an even more compelling story of vaccine failure. Out of 65 H3N2 viruses that the CDC has genotyped, only 9 (14%) matched the vaccine strain. The remaining 91% are a new strain, for which the vaccine either limited or zero efficacy. (The CDC doesn't say when those 9 cases appeared, but I'm betting they were early in the season.) A newer, 2007 strain, which they have decided to put into the vaccine for the Southern hemisphere for the 2008 season, should provide far better protection. (The flu season down under occurs in their winter – our summer – so it peaks 6 months after ours, and they can use a newer vaccine.) We’ll get the newer strain in our vaccine next year too, assuming the FDA agrees when they meet next week to decide. Unfortunately it’s too late for this year.

Here’s a proposal to the CDC: release all the sequence data you have from this year’s strains right away, and put it in GenBank. And release any other data you use to pick the strains that go in the vaccine – and do this every year. Then everyone can look at it and double-check your decision about which strain should go in next year’s vaccine. Sure, it might be annoying to have other scientists checking your work. But wouldn’t it be great if researchers – including students – all over the world could look at the flu data and make their own predictions about what strain would dominate next season? The science would only get better.

3 comments:

  1. Steven,

    I do appreciate your ongoing effort to get the CDC to actually perform its mission, but the CDC (CORPORATION for Disease Control) has no interest in sharing information with an outsider like yourself. If they were to provide their "proprietary" information to an outsider that might cut into their "profits" aka budget. You will never see that information released save by an act of Congress or by the courts.

    I am speaking from direct experience. Over a year ago the CDC contracted with Zeptometrix Corp. to produce a Panel of inactivated influenza strains including H5N1 re-aassortants that could be used to develop and validate the next generation of flu diagnostics. This Panel was supposed to be made available to industry so that a viable commercial biothreat protection system could begin to be created. This program was a result of pressure from both Capitol Hill and by NIH. Our company has been waiting for this product to be made available since early 2007. Now I hear that the CDC contact with Zeptometrix is going to be quietly allowed to expire without any of the reagent ever being made available outside of the CDC and its LRN. So keep up the effort, but I would not hold your breath .

    Thanks! Mike mmoser@tds.net

    ReplyDelete
  2. Sequencing would of course be more accurate. SS is totally right on. Why base these things on serotype? Makes less sense-- makes more generic understanding--makes more resistant flu outbreaks.

    We can only surmise that those untyped strains in the histogram have the same distribution as the typed ones above, that there are three times as many H3 as H1. Then again, if you are serotyping, why not also include the neuraminidase? H3Nx seems kind of fat fingered. Of course there are probably many reasons its done the way it is. I just dont know what they are, because it seems like we can do much better right now.

    Maybe they don't have much sequence to release? Certainly haven't had much luck getting that data, either.

    cl

    ReplyDelete
  3. Critical.crab,
    From what I've heard - including sources inside the CDC - they have quite a lot of sequence data. They have a substantial sequencing operation of their own in-house. They have been asked over and over again to release their sequences, and they continue to refuse. Sometimes they do this very cleverly, by saying, for example, that "we would love to release the data, but the labs that provided it won't allow us to." They don't tell you who those other labs are, so you can't check the truth of this claim.

    Unfortunattely (as the first comment also reflects) the CDC has a long culture of keeping everything to themselves. It seems to me that this is primarily political on their part - they want to control the data so they can control everything else, from publications to public health measures.

    ReplyDelete

Markup Key:
- <b>bold</b> = bold
- <i>italic</i> = italic
- <a href="http://www.fieldofscience.com/">FoS</a> = FoS

Note: Only a member of this blog may post a comment.