Here we go again. Late last year, scientists seemed to be homing in on the cause of chronic fatigue syndrome (CFS)—excessive tiredness and other symptoms that have no known biological cause--by finding a supposed viral link. But a new paper challenges that link, a development that may plunge the field back into the same confusion and acrimony that has characterized it for years.
Many CFS patients report that their symptoms began after an acute viral infection. Yet scientists have been unable to pin CFS on common viruses such as the Epstein-Barr virus. As a result, patients have faced skepticism for years that CFS might not be a real disease, or that it is perhaps a psychiatric disorder.
A team of American researchers thought it finally struck pay dirt last October when it reported in Science that it found DNA traces of a virus in the blood cells of two-thirds of 101 patients with CFS, compared with 4% of 218 healthy controls. XMRV is a rodent retrovirus also implicated in an aggressive prostate cancer, though why it might cause or be associated with CFS remains unclear.
Other scientists were dubious about the XMRV connection. They criticized the Americans for not explaining enough about the demographics of their patients and the procedures to control for contamination. Several virologists around the world practically sprinted to their labs to redo the experiments, and the discovery that a clinic associated with the Science paper was selling a $650 diagnostic test for XMRV made the issue more pressing. A British team already exploring the XMRV-prostate cancer link won the race, submitting a paper to debunk the claim on 1 December.
The team, led by Myra McClure, a professor of retrovirology at Imperial College London, examined DNA from the blood of 186 CFS patients ranging in age from 19 to 70, with an average age of 40. Most were markedly unwell. McClure's team used a PCR machine--which copies and amplifies scraps of DNA--to search for two viral sequences, one from XMRV and the other from a closely related virus. They discovered nothing. At a press conference discussing the results, published in PLoS ONE, McClure was blunt and confident: "If there was one copy of the virus in those samples, we would have detected it."
This null result prompts the question of what--if anything--was wrong with the original paper. In their own paper, the PLoS ONE authors seem to suggest that contamination was at fault, stating that they were careful to work in labs that had never handled XMRV and in PCR machines that analyze no mouse tissues. But McClure says her group merely wanted to make that explicit, not accuse anyone.
Regardless, the American team followed the same procedures, says Vincent Lombardi, a biochemist at the Whittemore Peterson Institute for Neuro-Immune Disease in Reno, Nevada, and co-author of the Science paper. He also expressed bewilderment that the McClure group didn't search its CFS samples for the same DNA sequence as his team had, raising the possibility that that's why the two groups came up with different results. McClure and colleagues, however, looked for not only an XMRV sequence but also a sequence in a closely related virus, MLV. That MLV sequence, highly conserved among viruses of its class, would presumably have been found if XMRV was present, they said.
One distinct possibility, says John Coffin, a microbiologist at Tufts University in Medford, Massachusetts, who studies retroviruses, is that both papers are right. He called the PLoS ONE paper too "preliminary" to settle the debate and said XMRV could show more genetic variety, and thus be harder to detect, than anyone assumed. It's also possible that distinct strains of XMRV appear in different parts of the world, like the retroviruses HIV and HTLV (a leukemia virus).
Coffin says one more possibility, raised by many different scientists, is that CFS is actually a suite of diseases that presents the same symptoms and so might have many causes. Lombardi seconds this point. "It's naïve to think that everyone with chronic fatigue has the same etiology. There's probably going to be a subset of people with CFS that have XMRV, and it will probably end up being classified as XMRV-related CFS."
All of this leaves doctors and patients in a muddle. There's no doubt they're hungry for information. Out of curiosity, Lombardi did a Google search on "XMRV" the day before the Science paper hit and found about 22,500 hits. Three months later, there are 400,000 hits.
But some scientists, including Coffin and McClure, fear that Lombardi's clinic took advantage of that hunger by offering the $650 diagnostic test, 300 of which have been administered so far. Lombardi's group never claimed XMRV caused CFS, so it's not clear what a patient could do with a positive result. Lombardi argues that patients can avoid infecting other people with XMRV and have their diagnoses validated, if nothing else. His test results also bolster the science in the original paper--he says 36% of tests have detected XMRV, including a few from the United Kingdom.
To resolve the dispute, both sides say they are willing to work with the other and possibly test each other's samples. In the meantime, more papers exploring the link are slated to appear in the next few months, and each side says it knows of work supporting its hypothesis. Meanwhile, the field will continue to churn. As McClure told Science, "We take no pleasure in finding colleagues wrong or dashing the hopes of patients, but it's imperative the truth gets out."