Friday, February 19, 2010

Failure to replicate: The WPI's take

The Whittemore Peterson Institute has finally issued a statement about a recent UK study that failed to detect XMRV in ME/CFS patients.

By "finally," I mean three days after said UK study was published. Three days is probably a reasonable length of time, given the need to review the results, and given that they're probably busy with other things over there. Three days feels very long, though, when you've been clinging to the hope that maybe there will be a cure for ME/CFS in our lifetime after all. Three days feels very long you fear that hope might be snatched away.

(A summary for those just tuning in: ME/CFS is a debilitating disease that causes chronic pain, crushing exhaustion, and significantly shortened lifespan. In October of last year, the Whittemore Peterson Institute discovered a link between this disease and a retrovirus called XMRV. The results of the most recent study failed to confirm the WPI's results.)

WPI's statement begins, "WPI is aware of the recent UK study that was unable to detect the presence of XMRV in any CFS patient samples."

The wording made me smile. "WPI is aware ..." as if they could possibly have missed the news. Obviously I'm not the only person who's been impatiently awaiting their interpretation. They must have been getting hammered.

Anyway, their full response is here: Among other things, they point out that the new UK study didn't actually attempt replication. They used different techniques and technologies, which the WPI claims are insufficient to detect XMRV.

So, hope restored for the time being. It's strange, though. The UK lab almost certainly wasn't trying to manipulate the data to get negative results. Why did they not attempt a real replication study?


Jannie Funster said...

Hope is a good good thing to have restored again. I'm glad to hear that indeed! Maybe this will spur on a real replication study!

DeppityBob said...

There are other institutions and more grant money out there. Someone will replicate this, I'm sure.