New Paper - Gammaretroviruses - Maureen Hanson, David Bell

[Bob]

Regarding the Lombardi "Cytokine" study that RRM has been discussing:

From the paper:
"Patients and controls. One hundred and eighteen specimens from
patients who tested positive for XMRV and with a confirmed
diagnosis of CFS at the time of collection were obtained from the
Whittemore-Peterson Institutes’ sample repository.
...
"All patients described in this study represent
well-defined cohort of CFS patients who are known to have a ‘viral
or flu-like’ onset and have tested positive for XMRV by PCR and/or
serology."

http://iv.iiarjournals.org/content/25/3/307.full.pdf


To watch the video on youtube, click here.

I originally thought that the wording from the paper (above) was clear that the patients had been selected after they had been diagnosed with XMRV, but actually, on close reading it is ambiguous.

After watching the video, I have to say that I agree with RRM that the XMRV patients were discovered/determined after the study had started.

However I think there is another interpretation to RRM's and the process could have happened as follows:

They originally studied more than 118 patients, and then after they tested patients XMRV positive, they decided to turn it into an XMRV study and dropped the non-XMRV patients from the study to tidy it up.

So, if this is what happened, it would not be the case that they selected 118 random patients, who later all tested XMRV positive. But they would have selected any number of patients to research, and then later focused only on the ones that were XMRV positive in order to publish an XMRV-related study.

 

[Mark]

Regarding the Lombardi "Cytokine" study that RRM has been discussing:

I originally thought that the wording from the paper (above) was clear that the patients had been selected after they had been diagnosed with XMRV, but actually, on close reading it is ambiguous.

After watching the video, I have to say that I agree with RRM that the XMRV patients were discovered/determined after the study had started.

However I think there is another interpretation to RRM's and the process could have happened as follows:

They originally studied more than 118 patients, and then after they tested patients XMRV positive, they decided to turn it into an XMRV study and dropped the non-XMRV patients from the study to tidy it up.

So, if this is what happened, it would not be the case that they selected 118 random patients, who later all tested XMRV positive. But they would have selected any number of patients to research, and then later focused only on the ones that were XMRV positive in order to publish an XMRV-related study.

I agree with that interpretation as a possible explanation that's consistent with the text Bob. It's an interpretation that is consistent with all of the information presented in both the paper and the slides, and does not require that a 118/118 XMRV+ detection took place or was claimed. Since RRM is right that 100% would seem an unrealistic scenario, and this is an interpretation that doesn't imply any confirmation bias or dubious practices on the part of the researchers, this seems to me the most likely interpretation. Though there may be more detail to it than that, one would have to ask the researchers to clarify this to get a clearer picture.

This interpretation is quite similar to what I understood from that sentence, as I described above, with one slight difference due to the information in the video of Lombardi. I said:

Your interpretation, RRM appears to assume that when slide 3 says that the RNA, DNA and plasma were taken in Sep 06 and July 07, that earlier time point was a definition of which 118 patients they would use in this 2011 study. On the contrary, it seems fairly clear to me that they are not saying that at all: they are saying that they looked at which patients from this cohort they had tested positive for XMRV, found 118 of them, then went back to the samples drawn in Sep 06 and July 07, pulled out the relevant subset of the 300 representing the 118 who they tested positive for XMRV, and analysed the cytokine and chemokine profiles of those patients.

The only difference that the Lombardi quote makes to what I said there is that the measuring of the cytokine and chemokine profiles must have taken place earlier, but they pulled out the subset of those earlier results from the patients that tested XMRV positive.

This would seem roughly consistent with what one would expect if the original cytokine and chemokine measurements were taken for the entire cohort of 300: that would mean that, at that point in time, they had tested about 39% of that cohort positive for XMRV, so they decide only to use them for this particular paper.

It's also easy to see how this would seem a very rational course of action at the time, because if they found that those that best fitted this cytokine profile also were those that tested positive for XMRV, they would naturally feel that XMRV was a better identification of the relevant subgroup of ME/CFS, and therefore the profile correlations would be better in that group...and if they found that indeed they were, they now had multiple lines of independent evidence pointing at the same cohort.

It's also quite conceivable that there was time for some of the researchers on this paper to go back to those results, pull out only the subset of data for the XMRV+ patients, and then all they would have had to do during the limited time frame that RRM mentions is perform (or re-do) the mathematical part of the analysis (which I guess amounts to re-running the computer program on a subset of the data) and write that up. That would seem quite feasible for part of the team to do in a short space of time: essentially refine an existing paper to look only at the XMRV+ subset of the results, and publish that with an even stronger correlation with the profile as a result.

So this seems a consistent interpretation of what may have happened, but I must admit it's not easy to reconcile that interpretation with this quote from Lombardi in the relatively informal youtube interview for an audience of patients:

That work was able to accurately distinguish Chronic Fatigue [sic] from healthy controls about 95% of the time, in that group we were loooking at. And it happened to be that when we we were doing the XMRV research they all corresponded to XMRV, so that work was published as a XMRV cohort of Chronic Fatigue Syndrome. Now in actuallity that work was done before I ever knew XMRV existed, so it would really stand on its own, at least based on the cohort of patients that we have."

Lombardi appears to say here that "the group we were looking at" "all corresponded to XMRV", and the implication of his comment that "in actuallity that work was done before I ever knew XMRV existed" seems at first sight to be that when he says "the group we were looking at" he's talking about the entirety of the original group for which the cytokine/chemokine analysis was performed.

But it's also possible to interpret that when he says "it happened to be that when we we were doing the XMRV research they all corresponded to XMRV, so that work was published as a XMRV cohort", by "the group we were looking at" he means "the group we decided to look at for the paper" - the subset of the larger group that corresponded to XMRV positives. When he later says "in actuality that work was done before" he could be talking about the original measurements and analysis on the larger group. You have to read that "and it happened to be..." sentence a few times to realise that it's possible that's what he meant, but it is possible.

I can see two ways in which this could have worked, then (there are perhaps more possibilities). First, they could have selected only the 39% of samples that were testing positive for XMRV, and restricted their cytokine/chemokine profile analysis to that group. Second, they could have originally tested, say, 130 or so of that cohort (and incidentally that subset of the cohort of 300 may have been selected for other criteria which are unspecified in order to study only the most sick patients) and found that they correlated strongly with XMRV, so they picked out the 118 out of 130 that had tested positive.

I'd agree that Lombardi's comment in this informal interview for patients does seem to read as "they all corresponded to XMRV" and when he says "all" that is what suggests "118/118". But that one comment of Lombardi's "they all corresponded" is the only thing suggesting this, and he does say "they all corresponded", which could mean that they, as a group, selected for other reasons, had been found separately (in unpublished work) to correspond very strongly (not 100%) with a positive XMRV test...and then they worked up the data on the subset that tested positive.

It may also be that there's a slight linguistic confusion here too, because I'm not a native American speaker, but I know that at least some Americans do use the expression "you all" or "y'all" to refer to a group, so it may be that "they all corresponded" means "that group corresponded"...something like this, or a slightly loose use of the word "all" in this sentence, may explain why the quote above reads as a more definitive "100%" than is intended. And of course if, when Lombardi says "all", he was speaking somewhat informally and didn't really mean "100%", then RRM's point about how unlikely 118/118 would be doesn't really apply.

So: while I agree that Lombardi's comments in the interview do appear to be saying what RRM claims, that this appeared to be a 118/118 correlation, that interview is the only piece of evidence presented that appears to be saying that, and the paper and slides certainly don't say that. Again, I'd want to hear the full story on all of this from the researchers before concluding that they are really are talking about a 118/118 finding here, and of course they never published it as a 118/118 result even if that is indeed what they (thought they) found.

 

[Mark]

If you want my guess (but I agree it's just speculation):

- It is known that Max Pfost did the first (PCR) testing on 20 samples, and found 2 XMRV positives. After several rounds of tweaking and restesting, all 20 were found to be positive.

Now, my guess is that during all this repeated tweaking, they did not really get better tests (as later blinded retests have shown), but that they just contaminated more and more of their samples. Now (and this is really unknown), but my guess is that they then extended their tests to other patient samples in their repository and not to controls (which were probably purchased at a later time, for the specific purpose of doing the study when the decided to do this). They then contaminated more samples.

After they "knew" they were really onto something (i.e. XMRV infection is correraled to having ME/CFS), they decided to do a "real" study on this, and they then got into contact with the Ruscetti's, purchased control samples, mixed them up with their patient samples (many of which were contaminated by then), ran their "offical study PCR test" on all samples (contaminating some more patient samples and some of their fresh control samples), and so arrived at their 67% vs 4%.

But again I agree this is mere speculation and we'll probably never know.

On this speculation of RRMs, actually I have to agree that this sounds like a rather good and plausible guess as to roughly what might have happened. In the absence of all the raw data I have to also agree that we can't do much better than guess like this. Since I would very much like to see the raw data used to unpick the apparent deception involved in the presentation of the PACE trial results in the UK, and since I feel it's an outrage that the raw data for that publicly-funded study is still not publicly available despite freedom of information requests, I have to also say that I do think that, in general, scientists' raw data should all be publicly available. If it were, it would actually be possible to resolve 'unanswerable' questions like this without any need for speculation and without leaving questions like this as unsolvable mysteries, so I do think this is another outdated aspect of standard scientific practices that needs to change: all such data should be made available, and that should be a requirement for funding and publication.

 

[jace]

RRM, you're not the only one that was not clear enough. In simpler terms, this is my interpretation of what we know.

From 2006 on, the WPI were investigating the Nevada cohort. Before the idea of looking for a retrovirus had even occurred, they ran a cytokine investigation on them. That work was filed and they went on to test for retroviruses. They realised that there seemed to be a cytokine signature common to the samples that were positive for XMRV, so they took the results from those samples forward and, together with control samples, ran the Random Forest Algorithm using what seemed to be a common XMRV positive signature. As we know, the retroviral research was published in 2009, and the cytokine study in 2011, but the majority of the work on the cytokine study predated even the idea of looking for a retrovirus in people with ME aka CFS.

 

[RRM]

this is an interpretation that doesn't imply any confirmation bias or dubious practices on the part of the researchers, this seems to me the most likely interpretation.

​

​

It's an interpretation that is consistent with all of the information presented in both the paper and the slides...

​

I would be nice if this was true but it isn't.​

​

First of all, you are misreading the "cohort of 300" bit in the presentation. The entire Lake Tahoe cohort is about 300 (it's actually "just" 259), not the cohort that they had in their blood repository. These people were diagnosed from 1984-1987. Some people may have died or may have moved. Not all people want to participate. For instance, in a 2001 follow up study by Peterson, 123 of the original cohort of 259 returned questionaires (source: http://www.cfids-cab.org/cfs-inform/Prognosis/strickland.etal01.txt).​

​

Second, although you also mention this, Lombardi's states that "they all corresponded to XMRV" and I think it is very clear what he means here.​

​

Third, there is just no way that they completely did the XMRV testing and statistical analysis in the same time-window as they were preparing the Science study. To quote Judy Mikovits (from her interview in Nature), after the XMRV discovery "we really retooled our entire programme and did nothing but focus on that.”​

​

Finally I want to point out that even in your scenario there are some " dubious practices" going on. After all, the authors would then still have omitted from the paper that their "XMRV cytokine cohort" came from a single outbreak, making their results appear more generelizable than would be warranted by knowing all the underlying data (and, on a side note, it lends credibility to some of the earlier criticism of the Science paper by the Dutch group, who said that Mikovits had stated in an earlier presentation that the XMRV findings were also from the single, Lake Tahoe cohort).​

 

[jace]

What do you think of my suggestion, which chimes with Lombardi's video interview and what we know from the pdf and the studies, RRM? It all seems logical and innocent to me. Nothing odd at all, just investigators following up results, finding correlations and moving on to try something else (with rather exciting results, IMHO).

The 2009 paper had 101 patient samples. The 2011 paper (for which the majority of the work was clearly done before the retroviral research) had 118. There is nothing to say that there weren't samples from other than the Lake Tahoe cohort - indeed, I'd be surprised if the Whittemore's daughter and some of her internet friends hadn't volunteered their blood and/or tissues. I think we are trying to make too much of a youtube video interview and a header slide, reading in things that aren't there.

 

[Mark]

this is an interpretation that doesn't imply any confirmation bias or dubious practices on the part of the researchers, this seems to me the most likely interpretation.

It's an interpretation that is consistent with all of the information presented in both the paper and the slides...

I would be nice if this was true but it isn't.​

I really can't see your logic here, I'll need to see the workings of your argument. I said that mine and Bob's interpretation is consistent with all the information in the paper and slides you presented. If you assert that isn't true, you'll need to present quotes from the paper and slides that are inconsistent with our interpretation of what may have happened. I've read both in full, with particular reference to the sections you quoted, and I don't see anything that contradicts what we've set out.

First of all, you are misreading the "cohort of 300" bit in the presentation. The entire Lake Tahoe cohort is about 300 (it's actually "just" 259), not the cohort that they had in their blood repository. These people were diagnosed from 1984-1987. Some people may have died or may have moved. Not all people want to participate. For instance, in a 2001 follow up study by Peterson, 123 of the original cohort of 259 returned questionaires (source: http://www.cfids-cab.org/cfs-inform/Prognosis/strickland.etal01.txt).

That's not evidence from the paper or slides, so it doesn't have any bearing on whether my statement was false as you claim. But in any case, the info you add about the size of the available cohort does not mean that the cohort they originally tested for levels of cytokines and chemokines can not have had more than 118 people in it, which is your claim. So it's less than 300, it's 259 or less - so what? Did they need permission or questionnaires to run tests on the 259 blood samples they had? I haven't seen evidence that's the case. Even if it was 123 that they studied and they discarded 5 who weren't XMRV+, as Bob and I described, that would be a 96% positive rate and not the 118/118 you are asserting.

​

Second, although you also mention this, Lombardi's states that "they all corresponded to XMRV" and I think it is very clear what he means here.

Again this is a separate source from the paper and slides so does not bear on my statement which you said was false. This evidence you've introduced is a comment by Lombardi in a youtube interview for an audience of patients, and your whole case appears to hinge on assuming that when he said "they all corresponded to XMRV" that he definitely is asserting correctly that 100% of the samples tested positive for XMRV. I really don't think you can assume that. If 95% of them had, might he not ever say, casually in an interview, "they all corresponded to XMRV"? Note "corresponded to", he doesn't even say "they all tested positive for XMRV". This is the only piece of evidence you have produced in support of your claim that they tested 118/118 of those samples as positive, and I don't think it's strong enough to be certain that this is the case. Surely that's why scientists produce papers and scrutinise and peer-review them: to make sure that everything they said is strictly accurate? I don't see how you can rely on that "all" meaning "100% of the Lake Tahoe samples we tested".​

​

Third, there is just no way that they completely did the XMRV testing and statistical analysis in the same time-window as they were preparing the Science study. To quote Judy Mikovits (from her interview in Nature), after the XMRV discovery "we really retooled our entire programme and did nothing but focus on that.”

I really don't accept this argument either. I've set out above how it's perfectly conceivable that they had time. Firstly, in the scenario Bob and I sketched, they had all the data collected, the analysis had been run, the paper was basically ready to go, and then they found XMRV. They tested the samples from that study for XMRV as part of their XMRV work (this testing was unpublished, presumably somewhat exploratory, perhaps this was how they refined their testing in preparation for the full study). They decided to take the samples that tested positive and re-run the analysis on them only. As Silverman mentioned: this was now absorbed into their XMRV work so it's now part of focusing on XMRV. As I mentioned above: there were several names on this paper, and all that was needed was for one or two of those people to re-run the software statistical analysis and use the revised numbers in place of the first set obtained on the entire cohort. Since the software was presumably ready to go, this may have been a matter of editing a file to remove x numbers (where x is between about 5 and 100) and hitting 'run' - it could be done in a day, with no lab testing, by one of the collaborators; they just needed the list of numbers of samples that tested positive. And finally, what this resulted in was the slides, not the paper, which wasn't published until 2011. It's perfectly conceivable that someone in the team had ample time to do this.​

​

The only part of this work that you might say there's a challenge to fit into the time frame is testing the 118 samples for XMRV, when they were also running the testing for the Lombardi 2009 paper as well. I'm not convinced of that, but of course if it were true that they didn't have time, it would nullify your claim that they got 118/118 on these samples and this was evidence of "confirmation bias". That can't be the case if they didn't even test them.​

​

Finally I want to point out that even in your scenario there are some " dubious practices" going on. After all, the authors would then still have omitted from the paper that their "XMRV cytokine cohort" came from a single outbreak, making their results appear more generelizable than would be warranted by knowing all the underlying data

On the first point: yes the paper fails to go into detail about precisely where the cohort came from, except that it says "from the WPI's sample repository", which I think was known to be comprised of samples from Peterson's Lake Tahoe cohort. And yes, it would be ideal to go into just a little more detail and say "from the Lake Tahoe outbreak".

But then, just think about these cohort arguments in context of the history of CFS definitions, and your criticism that this "makes them appear more generalisable than would be warranted" is seen to be completely upside-down. To suggest that this is a "dubious practice" to simply call this 'a CFS cohort' is to turn the world on its head. Just remember how and when the term "CFS" was invented: it was invented specifically as a supposed description and definition of this very disease, when it broke out in Lake Tahoe. So if you are now criticising them by saying "they are calling this CFS, but is it generalisable to all CFS? Is that being misleading?" - that's quite ridiculous when you think about it. What cohort could possibly have a better claim to be "CFS" than a cohort from the outbreak from which the name was coined? This is true "CFS".

And so if results from this cohort are not generalisable to "CFS" in general, and your concern is valid, what then does that mean? Surely it means what we all know to be true: that the invented name and definition "CFS", created to label this outbreak disease, does not describe it accurately and defines a mixed cohort, so much wider that by definition it is impossible to obtain useful scientific results from studying it! So: are you questioning them for studying the actual disease, which was labelled as "CFS", and simply calling it "CFS"? When you talk of "dubious practices", surely you should be talking mostly about the creation of the misleading and obfuscating label "CFS", which effectively rendered study of this particular disease impossible for 25 years?

The same situation applies to a criticism of Lombardi et al that came from voices such as the UK psychiatrists: they said (paraphrasing): "one problem with this paper is it uses a definition of the disease, a set of selection criteria, which are very strange, and not clear, and not internationally recognised". What anybody who knows anything about ME understands full well when they say this is: They are criticising them for studying the actual disease! Using the best and most appropriate criteria possible! Lombardi et al selected people with profound disability, who fulfil the strictest consensus definition available - the Canadian Consensus Criteria - and who fulfil the Fukuda Criteria - and who present with profound disability....and Wessely, White and crew criticised them for not using their own made-up Oxford Criteria which requires only prolonged unexplained fatigue! If you want to talk about "dubious practices" for not mentioning in this paper that the cohort came from Lake Tahoe, then come on: "dubious practices" is not going anywhere near far enough to describe the arguments and research practices of those who use the Oxford Criteria to study "CFS/ME". I'm not going to mince words on this: arguments like their questioning of the Lombardi et al cohort selection criteria are the arguments of scoundrels.

(and, on a side note, it lends credibility to some of the earlier criticism of the Science paper by the Dutch group, who said that Mikovits had stated in an earlier presentation that the XMRV findings were also from the single, Lake Tahoe cohort).

What?! How? The Science paper said the samples it used were collected from about 10 (?) clinics across the US in areas where outbreaks had occurred. The cytokine study was, apparently (based only on slide 3 of the presentation), a study of samples from Lake Tahoe only, and apparently those were also tested for XMRV. Mikovits (is said to have) presented, in an earlier presentation, on the XMRV findings from the Lake Tahoe cohort. The Science paper described a different cohort. Clearly these were different cohorts.

Clearly there's no inconsistency in testing two sets of samples, giving a presentation on one and publishing a paper about another. Unless you have any further evidence you're not revealing, the argument you've presented just suggests that the Dutch group are in the habit of adding two and two to make five. They seem to have leapt to the incorrect conclusion that these two separate cohorts must both relate to the same single study, an assumption which makes a nonsense of the stated facts, and therefore they conclude that the WPI had described their work inaccurately. Since they seem to have repeated the far too familiar pattern of making a series of unjustified and incorrect assumptions about the presented evidence, highlighting inconsistencies between their own incorrect assumptions, and then concluding that the people who presented the evidence must be in error, perhaps they are the ones with the confirmation bias?

 

[RRM]

I said that mine and Bob's interpretation is consistent with all the information in the paper and slides you presented

Bu it isn't.

That's not evidence from the paper or slides

Yes, it is. The slides say that there is the Lake Tahoe cohort of ~300 which was identified from 1984 till 1987, of which they obtained ~100 blood samples in 2006-2007 for these studies.

Did they need permission or questionnaires to run tests on the 259 blood samples they had?

Wait, what? Where did I say that they (or anyone else) ever had 259 blood samples from this group? A total of 259 people were diagnosed by Peterson from 1984-1987. There is no evidence or indication whatsoever that blood was ever taken from this group as a whole.

And that is the whole point. In 2001 Peterson contacted the whole group to do a follow-up study, and 123 cooperated at that time.

your whole case appears to hinge on assuming that when he said "they all corresponded to XMRV"

No, it is not. If you listen to the relevant segment of the interview, you'll find that Lombardi presents the XMRV-cytokine findings as really being ME/CFS-cytokine findings. That is the whole point he is making.

This is the only piece of evidence you have produced in support of your claim that they tested 118/118 of those samples as positive, and I don't think it's strong enough to be certain that this is the case.

It is not the only point. The results from the May 2009 slides, which realistically could not have been done on XMRV in that timeframe, are pretty much conclusive evidence in itself. Besides, even if we (hypothetically) assume that they had the time and resources, they would never have done this "double work" of a cohort within a cohort before they even knew if their original XMRV research would get published.

And yes, I am pretty much certain about this.

It would nullify your claim that they got 118/118 on these samples and this was evidence of "confirmation bias". That can't be the case if they didn't even test them.

What? Of course they tested them for XMRV.

So if you are now criticising them by saying "they are calling this CFS, but is it generalisable to all CFS? Is that being misleading?" - that's quite ridiculous when you think about it.

No, it isn't. It's just entirely possible that some clusters of ME/CFS are caused by X and other clusters by Y. Therefore, whether you are studying the first cluster or the 13th, you need to specify this. It's essential.

To give you an example: if I were to begin studying the first cluster of AIDS in the US, I could conclude that essentially all people with AIDS are gay. It would be silly to suggest that this is an acceptable and scientifically sound practice.

Saying that it is more acceptable because it happened to be the first cluster or that this cluster was reason the disease got defined as it did, is no justification whatsoever.

What?! How?

It confirms that, at least some of the times, WPI's investigators have studied a cluster of patients without saying so in their final paper.

Unless you have any further evidence you're not revealing

Of course I have.


The patient identification numbers of the Science study correspond with patients from the Lake Tahoe cohort that had developped cancer (the CAA checked this out). Mikovits has justified this by saying they renumbered their patients, but I find it illogical to renumber to already existing numbers.

Also, according to a Peterson presentation, there were more cancer patients involved. Incidentally, Frank Ruscetti submitted the Science paper to the NCI for its 2009 advances. He just got the name of the paper wrong, most likely because he just copied it from an earlier draft. Anyway, he submitted it under the name "Detection of a retrovirus, XMRV, in blood products of patients with neurological diseases and cancer". :
http://ccrintra.cancer.gov/news/scientific_advances_2009.asp

All in all, I find it pretty compellng evidence that Peterson's Lake Tahoe cohort was used as a basis for the Science study, or at least to a greater extent as has been asserted since.

 

[jace]

I find it strange that you are reading so much into scant evidence, extracting single slides from presentations, picking up on a few words in a YouTube video. I think my interpretation is equally as valid as yours, but that neither of us can be 100% sure we know the truth. I do wonder why it is so essential to you to convince us of this point of yours, and wonder whether you are underestimating PR members or overestimating your 'evidence'.

The plain fact is that no-one here can know all the fine details of what occurred and when. However, there does seem to be more confirmation of the retroviral hypothesis recently, and we poorly people have our fingers crossed that the research finally finds a way through to dealing with this darned disease.

We could go into a discussion on the nature of reality, and how it is different for different people, modified by experience, throwing in a pinch of confirmation bias with a glance at the doors of perception It would match the discussion on pragmatism, on another thread in the XMRV category.

 

[natasa778]

I find it strange that you are reading so much into scant evidence, extracting single slides from presentations, picking up on a few words in a YouTube video. I think my interpretation is equally as valid as yours, but that neither of us can be 100% sure we know the truth. I do wonder why it is so essential to you to convince us of this point of yours, and wonder whether you are underestimating PR members or overestimating your 'evidence'.

The plain fact is that no-one here can know all the fine details of what occurred and when. However, there does seem to be more confirmation of the retroviral hypothesis recently, and we poorly people have our fingers crossed that the research finally finds a way through to dealing with this darned disease.

We could go into a discussion on the nature of reality, and how it is different for different people, modified by experience, throwing in a pinch of confirmation bias with a glance at the doors of perception It would match the discussion on pragmatism, on another thread in the XMRV category.

LOL jace

I guess if a MuLV or another type retrovirus is found in some unfortunate humans, it won't matter IF those humans are from a cluster. We won't call it a human retrovirus then, but a not-at-all-relevant cluster virus. We shall undiagnose those patients from CFS/ME or prostate cancer or whatever, 'cos we all know that gammaretroviruses ARE shot down by human immune system easily and if we find them in clusters of patients they should be confined to clusters then and not bother the rest of us! And we know that true CFS/ME or ASD cannot be sick with a retrovirus cause they are all really healthy and only having false illness beliefs of Munchausen or were born with their misfortune or belong to genetic clusters. Or whatever

 

[currer]

The reality is that we need more facts, more research and more information, especially from those researchers who say they can find MRVs in human samples. This has always been the crux of the problem, and it is disturbing that precisely those researchers we need to hear from most, are the ones whose work has been supressed and who have been silenced. Until they can defend themselves and explain their research freely, attacks on their integrity should cease.

 

[Sam Carter]

Another curiosity - this was published only two months after the /in vivo/ paper, covers very similar ground but with a smaller cohort, yet fails to reference the much larger study.

Immune correlates of XMRV infection

Vincent Lombardi1, Deborah Goetz2, Max Pfost1, Cassandra Puccinelli1, Judy Mikovits1*

From 15th International Conference on Human Retroviruses: HTLV and Related Viruses

Leuven and Gembloux, Belgium. 5-8 June 2011

Background

CFS patients often display antiviral enzyme RNase L dysfunction underscoring the importance of the innate immune response in CFS. We reported the XMRV detection in the peripheral blood of 67% of a cohort of CFS patients and 3.4% of controls [1]. XMRV infection may play a role in CFS pathogenesis through the dysregulation of the immune response.

Methods

This hypothesis was addressed by multiplex profiling of plasma cytokines and chemokines on a LuminexTM platform and phenotypic analysis of leukocyte subsets by multi-parameter flow cytometry in XMRV infected CFS patients versus uninfected controls. XMRV-infected subject and control samples were assayed blindly. Analysis was performed using the Gene Expression Pattern Analysis Suite and Random Forest tree classification algorithms. For immune profiling, 63 XMRV infected CFS patient samples were analyzed within 6 hours using an LSRII flow cytometer with BD FACSDiva software. Six normal donors and reference values based on a healthy population were used as normal baselines.

Results
16 of the 26 cytokines/chemokines measured were significantly differentially expressed; eleven up-regulated and five down-regulated including: IL-8, IL-6, MIP1alpha, MCP-1, IFNalpha and TNFalpha. XMRV-infected CFS patients showed reduced percentages of CD56+ NK and
CD19+ B cells. The NK phenotype in XMRV-infected CFS patients was altered, with 80% of the patients having a significantly reduced CD56+DIM population. The B cells present in the peripheral blood were CD20+, CD23 + mature B cells.

Conclusion

XMRV infection results in dysregulation of the immune response, either directly by infection of specific leukocyte subsets or indirectly through cytokine modulation.

Author details
1Whittemore Peterson Institute, Reno, NV, 89557, USA. 2Environmental
Sciences,University of Nevada, Reno, NV, 89557, USA.

Published: 6 June 2011

Reference
1. Lombardi V, Ruscetti F, Das Gupta J, Pfost M, Hagen K, Peterson D, Ruscetti S, Bagni R, Petrow-Sadowski C, Gold B, Dean M, Silverman R, Mikovits J: Detection of an Infectious

Retrovirus, XMRV, in Blood Cells of Patients with Chronic Fatigue Syndrome. Science 2009, 326:585-589.

doi:10.1186/1742-4690-8-S1-A221

Cite this article as: Lombardi et al.: Immune correlates of XMRV
infection. Retrovirology 2011 8(Suppl 1):A221.

 

[Sam Carter]

I also can't quite reconcile this with the /in vivo/ paper:

Transcript of Gordon Medical Associates' Jan 17th XMRV event featuring Dr Mikovits and others
by XMRV Global Action on Tuesday, February 15, 2011 at 3:19am

""
We’re looking at diagnosis based on cytokine and chemokine patterns.

SLIDE: DYSREGULATED CYTOKINE/CHEMOKINE PRODUCTION PLASMA FROM ME/CFS PATIENTS

This pattern from an ATL patient looks exactly like the pattern that we derived from 156 patients versus 138 controls, so it’s an inflammatory cytokine/chemokine signature.

SLIDE: RANDOM FOREST ANALYSIS REVEALS CYTOKINE SIGNATURE OF XMRV INFECTION

And you can do this diagnostic test, then, even when you can’t find the virus.
""

Is this another, separate cohort? Or a PowerPoint snafu?

 

[Bob]

RRM, I'm always willing to consider what you say, and I think you've made some helpful and interesting points in this thread. I think you have made a plausible case for some of the issues you have raised. Although, most of our discussion is speculation at this stage.

But, RRM, I think you've also made some unhelpful and misleading points, which really haven't helped your case.
When you announce 'factual' information which then turns out not to be based on evidence, but on speculation, then it means that I'm unlikely to take the rest of what you say seriously.

With the cytokiene study, you are clearly making speculative conclusions, based on partial and informally presented information.
We don't have the full information about where the samples came from. There might be extra samples that were added from other cohorts, which were already in the WPI repository. The paper says that the samples were taken from the WPI's blood bank, and we do not know how many samples they had there, or from which cohorts.

You also seem very eager to denounce the cytokeine study, which, for me, puts a question mark above your reliability (in terms of the information that you present).
I can't see why there is a need to tear it apart so vigorously, when all the evidence you have been presenting is a loose collection of informal, and not directly related, presentations, which do not prove anything.

The cytokine research was a single small interesting exploratory study, that would obviously need further studies to validate its findings.
If it eventually turns out that XMRV was a red-herring, then the study will still be useful, as it was based on a certain cohort of CFS patients. As Lombardi said, the study would still stand without the XMRV element of it.

But whatever your arguments, there are some of us who will remain interested in MLVs, until we have answers to all our questions.
The most obvious question is where all of these sequences originate from, and what is the source of contamination.
The contamination is not just coming from XMRV 22rv1, because there are now a range of fairly diverse sequences.

People have said that we may never know where these sequences originate from, but I would hope that there are quite a number of scientists, such as Denise O'Keefe, who will now be curious enough to investigate further. And a number of them probably are investigating, right now. If everyone had had an incurious attitude, or a disinterested attitude, or a defeatest attitude with the original XMRV sequences, then they would never have discovered XMRV 22rv1 in the cell line, which contaminates so many labs and substrates.

It's not just ME/CFS that many of us are interested in, in terms of ERV research. There have been so many others studies, published under the radar, researching diseases such as breast cancer, that have thrown up various diverse retroviral sequences that many of us are also interested in. We are hoping that the new technologies, and the new interest in ERVs will start to uncover some new findings to answer some of the questions.
If it all turns out to be contamination, then that's fine. At least it will have been properly investigated, and the questions answered, and we will have found the sources of the contamination.

I have no doubt that there will be many scientists who will quietly work away in these fields until they have better answers. It might turn out to be the case that CFS/ME is not associated with MRVs, but many of us are still interested in this field being explored until no stone is left unturned. Personally, I can only imagine that it will thrown up at least some very interesting and unexpected findings, as it already has done.

I acknowledge that not everyone is particularly interested in this field of research, and not everyone expects any interesting discoveries to come from it. But some of us are very interested.

 

[Bob]

The patient identification numbers of the Science study correspond with patients from the Lake Tahoe cohort that had developped cancer (the CAA checked this out). Mikovits has justified this by saying they renumbered their patients, but I find it illogical to renumber to already existing numbers.

Also, according to a Peterson presentation, there were more cancer patients involved. Incidentally, Frank Ruscetti submitted the Science paper to the NCI for its 2009 advances. He just got the name of the paper wrong, most likely because he just copied it from an earlier draft. Anyway, he submitted it under the name "Detection of a retrovirus, XMRV, in blood products of patients with neurological diseases and cancer". :
http://ccrintra.cancer.gov/news/scientific_advances_2009.asp

All in all, I find it pretty compellng evidence that Peterson's Lake Tahoe cohort was used as a basis for the Science study, or at least to a greater extent as has been asserted since.

The WPI used blood samples from their own repository, so I don't see how you can rule out samples being taken from a mixture of cohorts. As pure random speculation, maybe it was predominately the Lake Tahoe cohort that was originally studied, but with 30 extra samples from other cohorts. We just don't know.

 

[Mark]

I also can't quite reconcile this with the /in vivo/ paper:

Transcript of Gordon Medical Associates' Jan 17th XMRV event featuring Dr Mikovits and others
by XMRV Global Action on Tuesday, February 15, 2011 at 3:19am

""
We’re looking at diagnosis based on cytokine and chemokine patterns.

SLIDE: DYSREGULATED CYTOKINE/CHEMOKINE PRODUCTION PLASMA FROM ME/CFS PATIENTS

This pattern from an ATL patient looks exactly like the pattern that we derived from 156 patients versus 138 controls, so it’s an inflammatory cytokine/chemokine signature.

SLIDE: RANDOM FOREST ANALYSIS REVEALS CYTOKINE SIGNATURE OF XMRV INFECTION

And you can do this diagnostic test, then, even when you can’t find the virus.
""

Is this another, separate cohort? Or a PowerPoint snafu?

If you're that bothered about this technicality, ask them.

OK: so you've referenced another paper from the same group analysing cytokines and chemokines in 63 XMRV+ patients, and a presentation on results of the same kind of analysis from 156 patients versus 138 controls. This to add to the paper on 118 patients and 138 controls, and a slide presentation of those results from May 2009.

Well it seems fairly clear that they ran these cytokine and chemokine analyses on a lot of samples, and pulled out the data in a few different ways.

Very likely the 138 controls referred to here are the same as the ones from the 2011 paper. In this presentation there are 156 patients, in the paper focusing only on XMRV+ patients there are 118 patients and 138 controls.

This is just looking like confirmation of mine and Bob's interpretation: Here we are, perhaps, looking at the full data set of 156 patients from the original analysis; in the paper we are looking at the 118/156 of that sample (76%) who they found afterwards tested XMRV+. They published the paper on the results of those 118 only, and here they are presenting the results on the larger original group.

The study on 63 patients can be a completely separate analysis - perhaps an analysis of most of the patients who tested positive in the Science paper - or it can be another subset of the Lake Tahoe cohort, or a combination of samples...the 156 patients, 63 patients, and 118 patients could be completely separate cohorts, or overlapping cohorts...there doesn't seem to be any way to tell.

In any case, the 156 v 138 slide really just backs up the hypothesis of Bob and myself (can we call it a theory now, since Sam's provided supporting evidence?) that the 'in vivo' study was based on the 118 XMRV+ subset of an earlier study of a wider cohort of Lake Tahoe samples.

The paper quoted above mentions that it took 6 hours to run the flow cytometry on 63 patients. Data entry, running the computer analysis, and creating power point slides perhaps takes a couple of days. I don't see anything here that could not possibly be fitted in during a 6-month busy time window by a team of researchers.

This argument is looking more and more ridiculous. Yes, there's incomplete information in the various slides and papers about exactly which cohort the patient samples were from, there are clearly multiple cohorts, there may be overlaps in terms of samples from the same patients being used for multiple studies, some of the results may have been subsetted and computer analysis performed on a subset of the results, these results were presented informally on slightly different subsets of the patients...but there is no inconsistency between the various bits of information presented, no oddness or curiousity about it, just insufficient information to answer these particular questions about the backplot of the detailed relationships between the patients and samples used in these different studies.

So there are just some questions, which are not relevant to the results of the studies, which one might politely ask the researchers if one wanted that supplementary information. There are perfectly reasonable explanations possible. But if these questions are being raised publicly rather than privately, with unfounded suggestions and pure suppositions presented as fact, like "they found 118/118 positive and believed it, that suggests they're bad scientists" and "I think they lied in the Science paper and they really used the Lake Tahoe cohort, that's my guess because they said in a presentation once that they tested Lake Tahoe patients for XMRV"...well, when the speculation is going in that direction you're just questioning the scientists' honesty and methodology based on nothing at all, just some missing information around which you've constructed a conspiracy theory. If these straightforward questions were raised in this disrespectful fashion when they first arose in suspicious minds, perhaps that's why we don't have answers to them already.

I've spent enough time on this: there is nothing here but rumour, speculation, and wild assertions questioning the scientists' integrity, apparently based on nothing but suspicion and false assumptions. I'm not personally interested in the exact details of the relationships between these cohorts and samples in the various papers; it's the results that are interesting to me. Anyone who's curious about it should politely ask the researchers to clarify their confusion. If anyone's got any decent evidence to back up this mud-slinging, let's have it: lay the case out. Otherwise, I'm done with following up all this trail of FUD that leads nowhere, and refuting claims like the 118/118 which seem to have nothing but the most tenuous and speculative basis to them.

 

[Mark]

I said that mine and Bob's interpretation is consistent with all the information in the paper and slides you presented

Bu it isn't.

Then I wish you'd get round to presenting some evidence of inconsistency.

That's not evidence from the paper or slides

Yes, it is. The slides say that there is the Lake Tahoe cohort of ~300 which was identified from 1984 till 1987, of which they obtained ~100 blood samples in 2006-2007 for these studies.

From which you conclude - somehow - that they obtained exactly 118 and certainly no more. How you conclude this is unexplained.

Did they need permission or questionnaires to run tests on the 259 blood samples they had?

Wait, what? Where did I say that they (or anyone else) ever had 259 blood samples from this group? A total of 259 people were diagnosed by Peterson from 1984-1987. There is no evidence or indication whatsoever that blood was ever taken from this group as a whole.

And that is the whole point. In 2001 Peterson contacted the whole group to do a follow-up study, and 123 cooperated at that time.

OK then, so these numbers you gave are irrelevant to the question at issue: fine. Unless you're somehow claiming that the above information proves that they studied exactly 118 patient samples in their original cytokine/chemical analysis, I can't see how it's relevant.

your whole case appears to hinge on assuming that when he said "they all corresponded to XMRV"

No, it is not. If you listen to the relevant segment of the interview, you'll find that Lombardi presents the XMRV-cytokine findings as really being ME/CFS-cytokine findings. That is the whole point he is making.

Yes, and as he explains, they were originally ME/CFS findings and were then presented as an XMRV paper when it was found they tested positive for XMRV too. His use of the word "all" in that interview for patients is your only basis for the 118/118 claim.

This is the only piece of evidence you have produced in support of your claim that they tested 118/118 of those samples as positive, and I don't think it's strong enough to be certain that this is the case.

It is not the only point. The results from the May 2009 slides, which realistically could not have been done on XMRV in that timeframe, are pretty much conclusive evidence in itself. Besides, even if we (hypothetically) assume that they had the time and resources, they would never have done this "double work" of a cohort within a cohort before they even knew if their original XMRV research would get published.

Right: you say they could not have possibly done the work in those slides in that time frame. It's been explained that the cytokine/chemokine analysis could have been done before, and a subset of those results analysed for this paper. So all they had to do was test the patients for XMRV. You assert further down that they did so. So how can you say this work not have been realistically done, when you accept that half of it was done considerably earlier, and the other half you agree was done? And this you consider 'pretty much conclusive evidence'???

And yes, I am pretty much certain about this.

I get that you're confident, but it doesn't seem to me that your confidence is based on any decent evidence, certainly not any that you're sharing with us here.

It would nullify your claim that they got 118/118 on these samples and this was evidence of "confirmation bias". That can't be the case if they didn't even test them.

What? Of course they tested them for XMRV.

OK so they did have time to test them for XMRV. Good. But they didn't have time to also test 101 different samples within those 6 months leading up to May 2009? Is that what you're saying?

So if you are now criticising them by saying "they are calling this CFS, but is it generalisable to all CFS? Is that being misleading?" - that's quite ridiculous when you think about it.

No, it isn't. It's just entirely possible that some clusters of ME/CFS are caused by X and other clusters by Y. Therefore, whether you are studying the first cluster or the 13th, you need to specify this. It's essential.

To give you an example: if I were to begin studying the first cluster of AIDS in the US, I could conclude that essentially all people with AIDS are gay. It would be silly to suggest that this is an acceptable and scientifically sound practice.

Saying that it is more acceptable because it happened to be the first cluster or that this cluster was reason the disease got defined as it did, is no justification whatsoever.

Decent argument. The cluster from which the samples are drawn is relevant and useful information, granted, and should ideally be concluded. But if you want to call this "dubious practice", then please: roll out much stronger language for the infinitely more dubious practices of the people who created the name and definitions of CFS, and the tricksters who research the disease using a wastebasket Oxford diagnosis and a dodgy questionnaire they made up. Get it into perspective when you talk about "dubious practice": if this is "dubious practice" in your eyes, then I guess on that scale of reference you would want to see those using the Oxford Criteria in prison.

What?! How?

It confirms that, at least some of the times, WPI's investigators have studied a cluster of patients without saying so in their final paper.

More precisely, it appears to confirm (assuming the relatively informal slide presentation is accurate) that they have done so at least once. Please don't over-reach with your conclusions, it really winds me up.

The patient identification numbers of the Science study correspond with patients from the Lake Tahoe cohort that had developped cancer (the CAA checked this out). Mikovits has justified this by saying they renumbered their patients, but I find it illogical to renumber to already existing numbers.

Also, according to a Peterson presentation, there were more cancer patients involved.Incidentally, Frank Ruscetti submitted the Science paper to the NCI for its 2009 advances. He just got the name of the paper wrong, most likely because he just copied it from an earlier draft. Anyway, he submitted it under the name "Detection of a retrovirus, XMRV, in blood products of patients with neurological diseases and cancer". :
http://ccrintra.cancer.gov/news/scientific_advances_2009.asp

All in all, I find it pretty compellng evidence that Peterson's Lake Tahoe cohort was used as a basis for the Science study, or at least to a greater extent as has been asserted since.

This is going back a long way, but I never saw a solid case made about these numbers and if it's as clear cut as you say then it shouldn't be difficult to lay it out clearly.

The Science paper, as I recall, said the patients were drawn from about 10 clinics across the US where outbreaks had occurred. I don't recall it being denied that Tahoe was one of them. I also recall that the original submission to Science was returned with comment and they were asked to expand and improve some aspects of their methodology. I don't precisely recall whether one of these comments was that they should expand their study to incorporate more samples from additional outbreak areas, but that sounds like a reasonable explanation.

Supposing that was it, then does this narrative satisfy you as plausible?...

The original analysis they did was all based on samples from Lake Tahoe, which they had previously analysed for cytokines and chemokines and had that data all ready. When they started testing XMRV, they tested on that cohort, and found most of the (156?) positive for XMRV. Their presentation in May 2009 showed the results of the cytokine/chemokine analysis on the XMRV+ subset of those patients. They submitted to Science at that time but were asked to change a number of things, including analyse patients from other areas too. Over the following months, they then analysed other patients from multiple outbreak areas, perhaps including in the final study some patients from Lake Tahoe (perhaps those are the duplicate patient numbers) as one of the outbreak areas. When they presented to Science they were presenting that additional analysis.

Just a guess, really, but it seems consistent with everything you've highlighted, and doesn't require any insinuations of dishonesty or confirmation bias.

 

[Bob]

There are perfectly reasonable explanations possible. But if these questions are being raised publicly rather than privately, with unfounded suggestions and pure suppositions presented as fact, like "they found 118/118 positive and believed it, that suggests they're bad scientists" and "I think they lied in the Science paper and they really used the Lake Tahoe cohort, that's my guess because they said in a presentation once that they tested Lake Tahoe patients for XMRV"...well, when the speculation is going in that direction you're just questioning the scientists' honesty and methodology based on nothing at all, just some missing information around which you've constructed a conspiracy theory.

I've spent enough time on this: there is nothing here but rumour, speculation, and wild assertions questioning the scientists' integrity, apparently based on nothing but suspicion and false assumptions. I'm not personally interested in the exact details of the relationships between these cohorts and samples in the various papers; it's the results that are interesting to me. Anyone who's curious about it should politely ask the researchers to clarify their confusion. If anyone's got any decent evidence to back up this mud-slinging, let's have it: lay the case out. Otherwise, I'm done with following up all this trail of FUD that leads nowhere, and refuting claims like the 118/118 which seem to have nothing but the most tenuous and speculative basis to them.

Mark, I totally understand your frustration.
It's so hard to work out all the facts re MLV research, without the added confusion of people posting innaccurate and misleading information.
So it's especially annoying when people post information as 'facts', when on closer inspection, it turns out to be either baseless or plain wrong.
This has happened twice recently: on this thread, and on another thread re Singh's study.
Presenting misinformation as 'fact' causes other people to do so much work to find out the real facts, that it's simply unreasonable.
I'm particularly pleased that you've managed to establish that RRM's assertions, re the cytokine study, are baseless.
I was quite surprised that RRM would make such baseless allegations, and I'm very annoyed by it.
I'm annoyed that I was misled, and that it's taken so much work of other forum members to work out the truth.

But I really appreciate the time and effort you've taken to explore these issues.
It's been really interesting to read all of the recent discussions, and I've learnt a few things that I didn't know previously.

 

[Sam Carter]

https://listserv.nodak.edu/cgi-bin/wa.exe?A2=CO-CURE;a1ca7759.0812D

Monday, October 13, 2008

Special Issue - Abstracts and Reviews: 7th Joint Conference of the International Cytokine Society and the International Society for Interferon and Cytokine Research, CYTOKINES IN CANCER, INFLAMMATION AND INFECTIOUS DISEASES: TRANSLATING SCIENCE INTO HEALTH

doi:10.1016/j.cyto.2008.07.077

Serum cytokine and chemokine profiles of individuals with myalgic encephalomyelitis (ME) reveal distinct pathogen associated signatures

Vincent C. Lombardi 1,2, Doug Redelman 2, Darren C. White 1, Marc Fremont 3,
Kenny DeMeirleir 4, Daniel Peterson 1, Judy A. Mikovits 1,2,

1 Whittemore Peterson Institute, Reno, NV, USA,
2 Department of Microbiology and Immunology, University of Nevada, Reno, NV, USA,
3 Protea Pharma, Brussels Belgium, Belgium,
4 Free University of Brussels, Academic Hospital Brussels Belgium, Belgium


Myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) is a Heterogeneous disease with unknown etiology. Previous studies have shown that viral specific immune responses and immune abnormalities play critical roles in the pathogenesis of ME/CFS. The central problem in the management of patients with ME/CFS is the lack of biomarkers for patient stratification into subgroups according to distinct immune responses, virus infections and neurological abnormalities. This situation hinders both the diagnostic process and development of specific treatments.

In this study our aim was to subgroup ME/CFS patients based on serum chemokine and cytokine profiles with the ultimate goal of establishing disease parameters on a molecular level that correlate with distinct disease phenotypes. We used suspension antibody microarrays of 25-cytokines and chemokines on a Luminex platform for serum profiling of 168 ME/CFS patients and 140 healthy controls. Our analysis has revealed distinct pathogen associated signatures with significant 5- to 200-fold differences between patients and controls for the inflammatory serum chemokines IL-8, IP-10, MIP-a and MIP-1b, as well as the pro inflammatory cytokines IL-6, TNFa and IL-1b.

Moreover, our data shows for the first time in ME/CFS a cytokine and chemokine profile, which suggests a TH17 shift in subgroups of our cohort. We conclude that cytokine and chemokine patterns in subgroups of ME/CFS can be used diagnostically, as serum biomarkers to stratify patients for appropriate anti-inflammatory, antimicrobial and antiviral therapeutics.

doi:10.1016/j.cyto.2008.07.077

 

[Sam Carter]

...

Otherwise, I'm done with following up all this trail of FUD that leads nowhere, and refuting claims like the 118/118 which seem to have nothing but the most tenuous and speculative basis to them.

Not quite done, it would seem. ;-)