WSJ's blog on health and the business of health.
December 20, 2010, 5:24 PM ET
XMRV: Raising the Issue of Contamination
By Amy Dockser Marcus
Four papers published today in the journal Retrovirology—and a fifth
one commenting on the papers—demonstrated how easy it is for mouse
contamination to skew lab experiments involving the virus XMRV.
But they are unlikely to resolve the debate over whether XMRV is
linked to diseases like chronic fatigue syndrome or prostate cancer,
especially since the authors of the papers disagree on the
interpretation of their data.
XMRV has been the topic of much debate since a paper was published
last year in Science, linking the virus to chronic fatigue syndrome.
The virus was also found in smaller numbers of healthy controls,
raising the possibility that the blood supply might be infected. XMRV
had previously been linked to prostate cancer. But some other groups
have not been able to find XMRV in either CFS or prostate cancer
patients or in healthy people.
Greg J. Towers from University College London, a senior author of one
of the papers, told the Health Blog that his group's findings indicate
that "XMRV is not a human pathogen." Tests used to detect XMRV are
also able to detect mouse DNA, and even if a tiny bit of mouse DNA
gets in a lab sample, the test can be positive even if the patient is
not, he explained. He added that he was not intending to criticize the
work of other scientists. "They published their observations in good
faith and we have simply reexamined their findings and made new
observations and come to a more likely conclusion."
But John M. Coffin, a retrovirologist and a co-author of another of
today's Retrovirology papers, told Health Blog that while his group's
study demonstrated that mouse DNA is everywhere in labs, none of
today's published papers "definitively show that any prior study is
Robert A. Smith, a research assistant professor at University of
Washington in Seattle who wrote a commentary in Retrovirology
summarizing the studies, told Health Blog that the possibility of
contamination means that future studies must be done very carefully
before conclusions about disease association are made. But he said he
is unwilling to state that the reported link between XMRV and CFS or
prostate cancer is no longer viable.
The papers focus on various problems associated with a specific kind
of test used to detect XMRV but does not examine every method used to
detect XMRV. Smith pointed out that some of the previous papers on
prostate cancer found XMRV integrated into the patients' DNA and "I
can't come up with a mechanism where there would be contamination
Judy Mikovits, who led the team of researchers that published the link
between CFS and XMRV in Science last year, said her team was able to
culture XMRV from the patients' blood and show antibody responses
indicating they had been exposed to XMRV at some point. "You will not
make an immune response to a lab contaminant," she told Health Blog.
Dr. Coffin said the debate over XMRV will continue. "This is not the
end of XMRV," Coffin said, "but it is a warning we have to be very,