WPI PCR found 68 of 101 patients positive and 8 of 218 healthy controls. How can this be explained by lab mistakes and false positives? The probability of that is almost zero. Science wouldn't have published it if false results were as likely as you say they are.
Apparently you think that the chance of WPI being wrong and the chance of Cooperative Diagnostics being wrong are equal. Which makes no sense, logically speaking. WPI's test found an enormously significant difference between CFS patients and healthy controls, CD did not. WPI's results were published in Science, CD's have not been. CD has offered zero evidence in support of their test. CD has not established that their test can distinguish between CFS and healthy controls. CD tested 400 samples and did not get a single positive. I would say they have no credibility but since they offer no evidence, there is nothing to judge as credible or not. For all I know a Magic 8-ball would be as accurate as CD's test.
There are multiple possible reasons for a false positive result, I do not know them all. Testing artifacts are possible, for example. That means something in the test is wrong, in the reagents perhaps, or the way they combine with the DNA of the patient's own cells. Some combination of multiple problems can probably produce false positive results that look like a real result, including differences between test patients and controls.
Also, there is the 'wrong bug' type of false positive. That actually seems most probable to me. In other words, some other bug has part of the XMRV sequence. That could present exactly as their results showed, but not actually be XMRV.
Also the tested cohorts could be different. WPI used some biomarkers to select their test subjects, CD and other replication studies will probably just use a sample of ME/CFS patients.
If a lab is pursuing an agenda, trying to build a case for something, they will sometimes see what they are looking for, basically cherry-pick their procedures or selection criteria until they get the results they want. Sometimes they inadvertently discover a new way to create false positives. This is a known risk in all forms of research, I was a researcher myself before CFS and have had to address this issue. The only way to determine whether initial findings are accurate is through replication attempts, particularly by outside researchers using their own (different) methods. Not matter how strong an initial finding looks, it can fall apart. I like to mention the 'Cold Fusion' debacle as an example, that research looked very solid, that is, until the replication studies failed. And at one point one of the Cold Fusion research teams did get published in Science... So Science has made mistakes before, they are an interdisciplinary journal. They are not experts in everything. Also, they can not be expected to figure out something that only a replication study will reveal.
This issue will be settled through the scientific process, and not by our reasoning or beliefs, including my own. What will happen may not be what people who are advocating for or against XMRV as the cause of CFS might like to hear.
But by that reasoning any successful replication studies will be questionable too, since a good-faith first replication effort (like we're hoping for) means they'll use the same methods and the same samples. Where does the doubting end?
EXACTLY, that is the way this works. People will also question the replication studies, as they should. The doubting will not end until a consensus finding begins to emerge from many different studies.
The reason I am posting at all is related to what you say, people don't like the 'doubting.' As if doubting XMRV is somehow doubting CFS. Nothing could be farther from reality. We must doubt any bold claims about CFS until they are proven, and one article in Science is not proof. Dr Coffin made that clear at CFSAC, he knows the WPI study was good, but still just the opening play in a very long game. He has been there and done that before...
Kurt, if I am understanding a standard PCR, the risk is that a small amount of contamination will lead to a false positive. Is this the concern you have for the standard PCR? Given the extreme difference in results between healthy cases and cfs cases I think contamination is most unlikely. If you get a positive on a PCR without contamination evidence you can bet the farm the patient is infected.
One of the problems we used to worry about with the mycoplasma incognitus testing was that the bacteria would disappear if the blood was not sent on dry ice and tested asap. I don't know if this would be a concern with a retrovirus.
Paula Carnes
This is a quote on the problem with standard PCR:
Preventing carry-over contamination with uracil-DNA glycosylase (Roche Diagnostics Corporation)
PCR can amplify a single molecule over a billionfold. Thus, even minuscule amounts of a contaminant can be amplified and lead to a false positive result. Such contaminants are often products from previous PCR amplifications (carry-over contamination). One common strategy to avoid such contamination is substituting dUTP for dTTP during PCR amplification, to produce uracil-containing DNA (U-DNA).
http://www.roche-applied-science.com/pcr/applicati...
I addressed that above, contamination is not the only possible cause of false positives. If there are conflicts between studies, this will have to be figured out.
WPI used more than 3 MuLV antibodies. Is it really possible that they got false positives from all three?
Absolutely yes, because they used MuLV antibodies and not XMRV antibodies. MuLV antibodies are a type of general test for XMRV, evidence of some related infection. So WPI is obviously counting on cross-reactivity between MuLV antibodies and XMRV. By definition this type of approach to antibody (And culture) studies can produce false positives. What that would mean is that some other antigen (not XMRV) is present that reacts with MuLV antibodies. That is very possible, in fact there are even endogenous human retroviruses that are similar to MuLV, that are proven to be present in CFS (some HERV classes are considered gamma retroviral). WPI showed that in their May presentation. There could even be some novel unknown antigen causing false positives on the MuLV antibody test. This is the point that has mystified me the most about the Science article. Why they did not address the risk of their antibody test for cross-reactivity thoroughly, that is a big issue. Maybe they did address that off-line and just did not report it. I think we will learn more about WPI's actual testing process when the first replication studies appear and WPI defends their study.
Final thought about testing - until multiple replication studies are reported and some type of consensus emerges about XMRV, any testing is really for research purposes only. I think some people are counting on a positive XMRV finding to help validate their ME/CFS, and to point to a new treatment. I hope they realize they are gambling at this time. Personally I would wait for a validated test before investing my hopes (and money) on XMRV testing. That may take a year, given that some replication studies will not report until Spring and some are just getting started so probably will not report until Summer or Fall of next year.
Addition: Just found an interesting study showing that MLV antibodies are considered cross-reactive with HERV. In this study the researchers studying psoriasis used MLV to illustrate that HERVs were active. So clearly MLV is known to be very cross-reactive... see:
http://www.ncbi.nlm.nih.gov/pubmed/12902038
