Friday, 11 October 2013

How mice lie....

In a nice 9:40 video presentation to TEDMED 2013, H. Shaw Warren brings us up to speed on some of the shortcomings of using mice to understand our complex human immune responses to bacterial infection.

Some key points he makes are:


  1. We use inbred mice. They "all look the same" because they are. That does not capture the diversity of responses to infection that we must deal with in trying to understand human infections
  2. Mice are resistant to infection and inflammation requiring lots of material o be injected. This is not what happens in the wild (100-100,000x more resistant than humans)
  3. Trauma and burns can be studied by examining gene responses in humans. These studies found poor correlation with mouse models.
  4. Mice look very different from us, why wouldn't their immune responses be equally different?
  5. Mice have evolved in environments rich in microbial exposures
  6. Mice have large litter sizes and short gestational periods which would increase adaptive evolutionary cycles - perhaps they have adapted to tolerate larger inocula than humans
  7. Mouse model success often determines whether a candidate drug proceeds to human trials. Some of these may work in humans
  8. All of the drugs studied in mice that have worked, have failed in humans (I did not know that)
  9. Mice still useful for gene, gene pathways, techniques and toxicity studies but extrapolating to complex human inflammatory disease networks may be a stretch
  10. The scientific community should raise the bar in justifying a link between human and mice responses before proceeding
Very nice talk.

Thanks to @MsWZ for tweeting link