Los Angeles Times - Article (The Weiss Paper)

[VillageLife]

http://www.latimes.com/health/boostershots/la-heb-chronic-fatigue-09292010,0,3081476.story?track=rss

Chronic fatigue and XMRV -- what one researcher (who's been there) has to say

It's been a year since the journal Science published a paper linking a retrovirus called XMRV to chronic fatigue syndrome -- an illness nobody has been able to explain or treat very effectively, to the enormous frustration of people diagnosed with it.

The paper was met with expressions of hope and joy from many in the chronic fatigue syndrome community, who saw it as potentially leading to diagnostic tests, treatments and even, maybe, a vaccine and a cure.

But studies, even those published in prestigious journals such as Science, turn out to be wrong all the time, and their results need to be replicated before anybody can be relatively sure the findings were correct.

To that end, other research teams have attempted to find evidence of XMRV in chronic fatigue syndrome patients. Five different teams, including one led by the U.S. Centers for Disease Control and Prevention and another from the National Institutes of Health and the U.S. Food and Drug Administration, have published papers failing to find such evidence.

One of the teams, from the NIH and FDA, found evidence of a different retrovirus, one associated with mice, in the patients, but nobody knows what that could mean. It does not confirm the original paper.

So what is going on?

One esteemed retrovirologist, Dr. Robin Weiss of University College London, has just published his thoughts on the situation. They are worth a look.

Weiss' main point is that the history of retrovirology is littered with the debris of papers finding a link between a virus and a disease that later turned out to be false results caused by contamination. "There has been a long succession of ‘rumor' viruses posing as tumor viruses and promulgated as the cause of chronic human diseases," Weiss wrote.

Researchers at the Whittemore Peterson Institute for Neuro-Immune Disease, which led the team that published the original paper, have repeatedly denied they could have a contamination problem.

Weiss knows something about the issue. In 1997, his own team reported finding a retrovirus genome in patients with rheumatoid arthritis. Two teams -- one in Sweden, one in the U.S. -- "confirmed" his results. It seemed to be a breakthrough in rheumatoid arthritis.

But four years after Weiss reported his findings, he discovered he was actually detecting contamination from a newly discovered rabbit retrovirus.

"I raise an eyebrow when investigators declare that contamination is ‘out of the question'; once bitten, twice shy," Weiss said.

"My own skepticism," he wrote, "derives from a strong feeling of* dj vu."

Read "Push and pull over chronic fatigue syndrome."

-- Trine Tsouderos /*Chicago Tribune

 

[VillageLife]

So these labs are only contaminated for cfs samples and not prostate cancer?

 

[anciendaze]

Here's the original paper. Yet another PCR-only detection. He made a blunder in publishing before he had confirmation via other means, and now ignores the variety of tests used in the Lombardi paper. Had he looked for immune response he could have distinguished contamination from infection in patients. It will take more time to discover how contamination produced the substantial skew in his data, (did he throw out valid results along with contamination?)

 

[CBS]

Here's the original paper. Yet another PCR-only detection. He made a blunder in publishing before he had confirmation via other means, and now ignores the variety of tests used in the Lombardi paper. Had he looked for immune response he could have distinguished contamination from infection in patients. It will take more time to discover how contamination produced the substantial skew in his data, (did he throw out valid results along with contamination?)

I love the whole "watch out for fools" angle from the guy who made a fool out of himself a decade ago.

Do you really have to go to the fool for the story? Everyone else saw his mistake and they are well aware of the pitfalls.

Do you ask the 'A' student or the 'D' student for their take on the potential for errors?

 

[Otis]

Do you ask the 'A' student or the 'D' student for their take on the potential for errors?

Depends if you're an 'A' teacher or a 'D' teacher, apparently.

 

[anciendaze]

We have two threads derived from the same opinion piece. Check out the other, here. The actual article is worth reading, to avoid misinformation added by reporters.

I'm not sure where he is going with some statements, except to say that the problem doesn't fit his preconceptions. Anyone who has followed the subject for a while could tell you to expect surprises. One statement I agree with, though not as he intended.

For instance, the proposition that murine mammary tumor virus (a betaretrovirus) is present in human breast cancer has been in circulation since 1972; but to my mind, the evidence is no stronger for simple reiteration.

Obviously, simple reiteration will not solve this problem. Nor will reiteration of studies which fail to detect virus in any human. If what you have been doing does not produce results, I suggest you do something different.

Under these circumstances, tentatively accepting results for prostate cancer while firmly rejecting those for CFS/ME is pure bias. Those studies demonstrated absence of ability to detect infected human beings irrespective of their diagnosis. (He himself ignores lack of validation for diagnostic criteria used by the CDC, to say nothing of Jason's study invalidating those same criteria by showing the ability to distinguish CFS from primary depression was little better than flipping a coin. Truthfully, diagnosis is irrelevant to those published negative results.)

I also note a common error concerning DeFreitas' claims. She explicitly said the virus she found was not HTLV-II, yet negative tests for that particular virus were claimed to discredit her.

Weiss is sufficiently honest to admit that he has no good explanation for the dramatic difference in contamination rates for his patient and control samples. (And, he also still seems to think using water as a negative control is good laboratory practice!) This comes down to not actually tracking the contamination in his results. What we have here is another example showing that elimination of all possible results eliminated all contamination. You could deduce this without wasting research money.

The dates on this piece show it was rushed out in response to the XMRV workshop. It appears the opposition is reduced to sniping from the sidelines with opinions and anecdotal reports of laboratory problems never really understood.

 

[Mark]

Obviously, simple reiteration will not solve this problem. Nor will reiteration of studies which fail to detect virus in any human. If what you have been doing does not produce results, I suggest you do something different.

Crucial point of course, and it exposes the dishonesty on the sceptical side, because still, one year later, nobody has even attempted to replicate the WPI study methodology. It's pretty basic science that you can't disprove somebody's findings by running a different test. Run the same test if you want to prove the results were due to contamination! If you don't do that, then expect to be distrusted.

And the fact that they don't dig deeper, look further, try to understand the results, but instead just dismiss them, is a pure and plain demonstration that they didn't ever believe the findings to begin with. We didn't even need them to do so: they actually explicitly stated it in the press before undertaking the study. "My expectation is that we will not..." - Reeves said this, but I thought I remembered Wessely using the same words; maybe I remember wrong, but he did say he expected his study to fail.

"Expectation" is an interesting word as well, by the way. A manager may "expect" things of his staff...

But the value of the article to me was a salutory warning from history: this sort of finding keeps coming up again and again, and every time, it is never disproved, never explained, just doubted, dismissed, and forgotten. That's the take-home message of Weiss' article for me.

Under these circumstances, tentatively accepting results for prostate cancer while firmly rejecting those for CFS/ME is pure bias. Those studies demonstrated absence of ability to detect infected human beings irrespective of their diagnosis.

Again an excellent point. There seems to be no basis for taking a different attitude to the prostate cancer and CFS/ME studies, unless that attitude is based on a preconception. And that preconception came out in the press from day one, straight after the Lombardi study was published. "I doubt this is true" was the basic message. Why? Because we have looked before and not found. It gives the game away, as far as Wessely is concerned at least. Nowadays he will always publicly say - because he must - that there may be a viral cause, at least as one of many factors. Yet when one is found, straight away he says he doubts it's true. As you rightly say, pure bias, and very obviously so. So much for scientific objectivity and keeping an open mind.

I also note a common error concerning DeFreitas' claims. She explicitly said the virus she found was not HTLV-II, yet negative tests for that particular virus were claimed to discredit her.

Thanks for mentioning that, the "HTLV-II" comment jumped out at me too, and I thought "hold on, surely DeFreitas never claimed it was HTLV-II, it was always unclear exactly what it was?". But I have barely researched that bit of history at all. Yet even I knew from a cursory knowledge of the subject that this was inaccurate and misleading. How can somebody like Weiss excuse inaccuracies like that? Such incorrect information does seem to say a lot about the writer - when somebody so senior and responsible for being accurate, fair and unbiased throws in the occasional error, you have to wonder why that's happening. I would just answer: bias, and lack of objectivity and fairness. Which comes out time and again in small and apparently irrelevant - but actually crucial - details. Since I'm not a retrovirologist, I have to form my judgment on the controversy primarily based on clues like this. And I'm afraid these sort of 'convenient errors' are there all over the place, for anyone who cares to dig into the matter.