26 August 2010

Chronic Fatigue Syndrome Virus Saga Continues

In a paper published today by "The Proceedings of the National Academy of Sciences", scientists found traces (in the form of gene sequences) of murine leukemia virus- (MLV) related viruses in 32 of 37 chronic fatigue syndrome patients; but in only 3 of 44 healthy blood donors. The researchers did not find XMRV. . . . A paper published by the CDC at that time found no XMRV or other MLV-related viruses in patients with the syndrome. . . .

MLV are known to cause cancer and neurological problems in mice, but it is not known whether they cause any diseases in human beings. XMRV is among several different members of the MLV family; and its conspicuous absence from the results in this study means that there remains no resolution as to the role XMRV may play in chronic fatigue syndrome.

It remains unclear why only two research teams have found evidence of these retroviruses. It is possible that different research teams are using significantly different methods of detection; and federal health officials are attempting to standardize the process. Of concern is the fact that the head of the federal tissue safety laboratory has been unable to isolate XMRV, whereas the group in Reno, Nevada claims to have done so.


From here.

This is very odd, because a previous study, late last year, "found that a retrovirus [XMRV] that originated in mice and attacks the human immune system is found in 68% of people diagnosed with chronic fatigue syndrome, but only 4% of healthy control subjects."

By themselves, either of the two studies would seem to be powerful evidence that a particular mouse source virus was the primary cause of chronic fatigue syndrome.

So, different high powered research labs test people with and without chronic fatigue syndrome. One lab finds definitive evidence of a link to XMRV. One finds a definitive evidence of a link to different MLV virus but not XMRV. A third finds neither virus.

This is very odd. Most experimental outcomes that can't be replicated involve weak signals from the evidence that could be statistical flukes, not strong yet contradictory signals. Yet, the odds of either of the studies that showed a viral link to chronic fatigue syndrome being statistical flukes, even with small sample sizes, is vanishingly small because the strength of the effect is so great.

One possibility is that there are subtle, but critical, differences in methodology between the studies that are generating either false positives or false negatives at one or more of the labs. Another possibility is that the actual cause may a germ which often is transmitted with a mouse vector that was not detected in any of the studies, and that the XMRV or other MLV infections found were merely opportunistic co-morbid condition that vary from place to place depending on the kind of viruses circulating in the local mouse vector population.

I can't think of enough examples of anything like this happening in medical research to have much of a sense of which resolution is most likely. The fact that two independent labs are finding some kind of strong MLV link to Chronic Fatigue Syndrome, suggests very strongly that even if MLV is not the cause of Chronic Fatigue Syndrome that its presence is a powerful diagnostic sign of the disease in many cases, and that any other true cause is also implicated in the MLV infection.

It also seems to me that even if we don't have a definitive determination of the cause of Chronic Fatigue Syndrome, that it would be entirely appropriate to begin treating patients with anti-viral regimes that could affect MLV type viruses, on the theory that MLV or another virus vulnerable to an anti-viral treatment regime is more likely to be a cause than any other known cause of a notoriously hard to diagnose and treat syndrome. If anti-viral therapies do work in a significant number of cases, in the end, it really doesn't matter that much how they work, or which particular gene is being shut down. The goal is to help people with illnesses get better, and in the absence of seriously side effects the possibility that a particular experimental treatment might not work for a particular patient is par for the course.

No comments: