Ian Lipkin on XMRV

XMRVLate last year virologist Ian Lipkin was asked by National Institute of Allergy and Infectious Diseases head Anthony Fauci to coordinate a multi-center study of CFS patients. Newly drawn blood samples from 100 CFS patients and 100 healthy controls from around the US will be blinded and sent to three groups – FDA, CDC and the Whittemore Peterson Institute – and assayed for the presence of XMRV. After the recent publication by Ila Singh on XMRV in CFS patients, Dr. Lipkin sent me the following note:

Dear Vince-

We have a plethora of explanations for how CFS/XMRV/MLV studies could go awry. However, we don’t have evidence that they have. Absent an appropriately powered study representing blinded analyses by Mikovitz and Lo/Alter of samples from well characterized subjects using their reagents, protocols and people, all we have is more confusion.

I remain agnostic. We won’t have answers until the end of 2011.

The NIH will post something on our study today.

Ian

240 thoughts on “Ian Lipkin on XMRV”

  1. There are really two uses of the word agnostic:

    – Assertion (x) is unknown (to me)
    – Assertion (x) is unknowable

    I am an agnostic on XMRV in the sense that it can never be proven that XMRV (or any other virus) is not in humans. Even when (say) archived hidden video camera footage would “prove” that the Mikovits and Lo studies were contamination, there would still be a chance that XMRV is in humans and has “just” not been detected by doing the right tests. Perhaps it’s in other tissues, or just in my little toe. Just like any other virus, I might repeat. You can’t prove a negative isn’t just a soundbite. In this sense, I am and always remain agnostic.

    However, I would agree that he appears to use the term “agnostic” in the context of him not actually being convinced by the data.

  2. I bet that it will be summer 2012, as Mikovits will change her culturing protocol to 83 days…

    It took her 60 days to partially report back on a couple of samples doing just PCR for hase IIb of the Blood Working Group.

    I think it would actually be best if Lipkin lowered the number of patients/controls somewhat, as it seems clear that not-detecting XMRV has not been a cohort issue and lowering the amount of samples would speed up the process.

  3. Dr Racaniello, do you have a link to the “something” on his study that Dr Lipkin said the NIH would be posting? I’ve hunted for it and couldn’t find it anywhere. If you come across it, would you please post it? Thanks.

  4. Hahaha!

    Ddrosha, the proud scientist, is presented clear, sourced evidence that reveals the absurdity of her claim that Dr. R has behaved without prejudice and remains objective. All she can muster in response is “And?”. What a joke.

    I’m beginning to wonder what it takes to get a PhD these days besides the suffocating self-righteousness on display here. Sturdy knee-pads?

  5. @profvrr:disqus First Lipkin says the Singh study (and the others) just recreated more confusion, then you, Prof. R., claim Lipkin said there is no ‘confusion’ for scientists, implying that the Singh study shows Mikovits in incorrect to scientists. But that non-scientists will never understand that until an appropriately powered study is done, which will probably never happen, then he supposedly says his own study is exactly such a study, results expected end of 2011.

    I submit (1) there is in fact confusion, not only among the confused non-scientists like me, but also among some scientists and (2) some scientists are not as agnostic as they claim in writing. : )

  6. @font-face {
    font-family: “Cambria”;
    }p.MsoNormal, li.MsoNormal, div.MsoNormal { margin: 0cm 0cm 10pt; font-size: 12pt; font-family: “Times New Roman”; }div.Section1 { page: Section1; }

    A “known positive” sample is precisely what’s under debate
    here, and is discussed extensively already. I won’t repeat that discussion.

    For the sake of many people that may be confused I would
    like to be clear about how this could be addressed.

    What experiments would you ask the Singh lab to do? Would it be
    something like this?

    1) Repeat the PCRs with half the concentration of Platinum Taq &
    dTTP (i.e. *exactly* the previously used conditions)

    2) Test a sample of human blood spiked with XMRV, to check that their
    nucleic acid extraction & PCR will definitely amplify from human blood

    3) Test a sample(s) that was actually PCR positive in the Mikovits lab
    (whether from infection or contamination)

    If they did 2 & 3 and they worked (i.e.
    yielded positive results), would they have shown that their assay works well
    enough to amplify XMRV genome if it really were there? And therefore not need to do 1?

  7. L J, maybe you should just speak for yourself and not even a small minority of the CFS community. It is not right to discount another persons believe. Sometimes truth is stranger than fiction.

  8. It’s not part of his methodology that validated assays are being used. Note that he in not demanding that the WPI use their “Science” methodology or Alter/Lo theirs. He couldn’t care less if they use a newly developed assay that was never calibrated to anything. He is just saying: ‘throw at these new samples whatever you wanna throw at it, and let’s settle the debate’.

    Thus, it is not part of his methodology that participant use validated assays. If it were, he WOULD be using “known positives” in this study, which he isn’t.It’s a subtle, but simple and essential difference, really.

  9. ALERT:

    Some person from the other side of the debate is reaching a new low as this is not a post from me (RRM).

    Please refrain from doing this again.

  10. The point is that it the Lombardi emthodology not part of Lipkin’s methodology. Everybody is free to choose their methodology in the Lipkin study, whether it is good or bad. WPI could ask advice from a voodoo man and Lipking wouldn’t care.

    The same applies to the “known positives” part. Lipkin is not using “known positives” in his study, but because labs are free to do whetever they want to do, they can calibrate thier assays to their own supposed positives (or go to the nearest voodoo guy for advice).

    Although it is almost certain that WPI will calibrate their samples against their own supposed ‘wild type’ positives (though they will most definetely not use the exact Lombardi methodology), they are perfectly free not to, and therefore it is not part of Lipkin’s methodology.

  11. It is not a fallacious appeal to history, as I have already explained multiple times. If someone asserts that certain events have occured historically, it is not a fallacious appeal to history to ask for an example of such an event. Only when you are conceeding that this has never happened historically but are arguing that this should be the “new”‘ way to go, asking for an historical example would be a logical fallacy.

    Thus, are you either stating that:

    a) this “exact replication” is the normal way to go in retrovirology, or
    b) this has actually never happened but are saying how it should in the future?

    Please do answer, Steve.

    Furthermore, concerning your points:

    1) Again, this is circular reasoning. You assume that YOUR defintion of “replication” is true and that therefore I am damning en entire field of virology. However, my point is that your definition of what is considered to be a “replication” in this field is wrong and that the way that this finding is being (in)validated is the proper and only way in which retrovirology findings are (in)validated. Rather ironically, it is you who is damning an entire field of science by implying, based on your flawed defintion of replication within the field, that all of XMRV research, of course except for the positive findings, is unscientific at its core. Which it is not.

    2) Because this point hinges on the last, the same really applies here. You are reasoning from a flawed assumption. Rather, you implicitly accuse Mikovits of having a strong aversion to experimental rigor just as well, as Mikovits has never in her life performed a more rigorous replication study than the one done by Singh.

    Again, and this is not a fallacious appeal to history as I have explained, if your aasumption would be true, you should easily point me to a few examples. As I pointed out not from Mikovits, as she has apparently been to lazy to do an exact replication study herself just once.

  12. It is not a fallacious appeal to history, as I have already explained multiple times. If someone asserts that certain events have occured historically, it is not a fallacious appeal to history to ask for an example of such an event. Only when you are conceeding that this has never happened historically but are arguing that this should be the “new”‘ way to go, asking for an historical example would be a logical fallacy.

    Thus, are you either stating that:

    a) this “exact replication” is the normal way to go in retrovirology, or
    b) this has actually never happened but are saying how it should in the future?

    Please do answer, Steve.

    Furthermore, concerning your points:

    1) Again, this is circular reasoning. You assume that YOUR defintion of “replication” is true and that therefore I am damning en entire field of virology. However, my point is that your definition of what is considered to be a “replication” in this field is wrong and that the way that this finding is being (in)validated is the proper and only way in which retrovirology findings are (in)validated. Rather ironically, it is you who is damning an entire field of science by implying, based on your flawed defintion of replication within the field, that all of XMRV research, of course except for the positive findings, is unscientific at its core. Which it is not.

    2) Because this point hinges on the last, the same really applies here. You are reasoning from a flawed assumption. Rather, you implicitly accuse Mikovits of having a strong aversion to experimental rigor just as well, as Mikovits has never in her life performed a more rigorous replication study than the one done by Singh.

    Again, and this is not a fallacious appeal to history as I have explained, if your aasumption would be true, you should easily point me to a few examples. As I pointed out not from Mikovits, as she has apparently been to lazy to do an exact replication study herself just once.

  13. RRM, it seems to be clear that you’re not really a fan of Dr. Mikovits. But how could you explain that the Drs. Alter, Lo and Ruscetti also are authors of positive XMRV/MRV studies? And have not distanced themselves from those studies in any way, so far.

    I wish someone would reply to my posts here.

  14. Once again, the ME/CFS community is most of all concerned about being abandoned yet again. It matters more that good research is done than that XMRV is validated. If research on XMRV is stopped after one thorough negative study, that would be too soon (science results are expected to be conflicting in the first few years while things are being sorted out), but a few high-quality negative studies from appropriate patient populations would convince most.

    However, the important thing to note here is that the usual yearly spending for ME/CFS research is $4 million. The current figure has gone up to all of $6 million. Real adequate for a “Priority 1 Emerging Infectious Disease” which is at least as debilitating, if not more so, than MS, Lupus, heart disease, and ESRD, but has no approved treatments and no approved diagnostic markers, no?

    Complicate that by understanding that intramural research (CDC) is spent on studying random tired people (not selecting for ME/CFS particularly) and extramural research is generally allocated by a committee with few or no experts qualified to examine the needed biomedical research.

    If you need confirmation of anything I’ve said, I can email (or post) references, or you can ask Lenny Jason of DePaul U.

    The concern from the community is more about what happens next, if XMRV doesn’t pan out.

    Can you offer any reassurance?

  15. Good scientists are wrong all the time. Even brilliant scientists. There is nothing wrong with being wrong. In every field, scientists that are much smarter than me (or you) will disagree with other scientists that are much smarter than me (or you). It’s (IMO) not smart though to focus on one side of the discussion (“Behe believes in intelligent design and is much smarter than you so who are you to believe in evolution”) and thus to ignore where the consensus really lies (“evolution rulezz”).

    Right now, the cumulative weight of the scientific community (and no, this isn’t an appeal to authority in the light of your question) is heavily leaning towards the idea that XMRV and CFS are not (meaningfully) connected. No, it doesn’t conclusively prove they are right, but all of these scientists have taken into account all of the data, including the fact that world class scientists like Ruscetti and Alter did report finding XMRV/MLV-like sequences and that many other world class scientists could not.

    There is nothing wrong with a few good scientists disagreeing with the consensus opinion (as that will lead to new discoveries or to firmer establish the consensus opinion), but that doesn’t mean I have to explain how these few accomplished people have produced these results and have not yet distanced themselves from their studies. It simply happens.

    As for Mikovits, she seems to be constantly changing her position on what is necessary for proper validation and that is what I don’t like. She is also constantly attacking the motives of scientitst that do not agree with her, as are other people from her institute (see the latest tripe from Deckoff-Jones for instance). Mikovits even once said that fellow scientists had “skewed their experimental design in order to not find XMRV in the blood”.

    Even if Mikovits turns out to be right and XMRV is truly in blood, it is clear that it is at least not as easy and reproducible to find as was first thought. Therefore, the above comment way out of line. Still,I’ll readily admit that she deserves every award she may get if she turns out to be right. : )

  16. Ok, a good answer, thanks. It didn’t change my mind and as long as i don’t see Ruscetti, Alter and Lo distance themselves from the positive studies, it’s by far not over in my opinion, but thanks anyway.

    I agree that a study like the one by Ian Lipkin or the BWG is the best way to resolve the question. But i’ll honestly tell you that as long as not every step in the process is controlled by both sides of the argumet, i will have trouble believing a negative outcome.
    I’m not paranoid (not that i know), but this is such an important question and there has been so much conflict between the different parties and also XMRV seems to be very difficult to study, so that i would like to see any chance for a wrong result ruled out, as far as this is possible.
    Like in a political vote, for example, when members of both parties check the ballot boxes, count the votes etc. I would really want to have every step of the process monitored and checked by both sides.

    And even in case we would ever arrive at the sure conclusion that the positive studies were wrong, i think they should try to find what went wrong. In this case there would probably be something to be learnt there too.

    I have donated to the WPI and in this moment i still don’t change my view regarding the positive studies, but if their studies turn out to have been wrong, it might even be the right thing to start a criminal investigation.
    Same thing if there are signs XMRV might have escaped from a lab and is harmful.

    Eric

  17. If you are suggesting that Shin et al used Nested RT-PCR, then you have not read the paper. There is no mention of Nested RT-PCR isolating RNA.

    Don’t you think you are in the wrong job if you are claiming differently?

  18. Neither is Singh, Coffin, or Racaniello
    . Facts are not an attack and it is plain that you wish to suggest to others that this is the case in order that they do not become aware of them.
    qPCR and Nested PCR are not Nested RT-PCR.

  19. The WPI and FDA/NIH labs will be using validated assays. Lipkin will insists that the CDC validates their assays too. This is your failure to understand how science progresses.

  20. If they wish to refute the results of Lombardi et al, they should first replicate the methodology of their paper. That is the Nested RT-PCR on gag.

    Then they can attempt to create new assays, but they too would need validating with clinically positive samples.

  21. Then you have not understood the premise for the study. Lipkin will ask for assays to be validated.

  22. There is no personal definition of replication, it means you do everything the same. I’m certain you will now demonstrate your foolishness again and claim that they can’t use the same lab or techs. Of course not, but they can use the same methodology.

  23. Thanks for your reply.

    I think the problem is that you trust the scientists based on their results. One week ago, you could have included Singh as a reliable scientist on your side of the argument, but now that she produced negative results, it’s easy to argue with hindsight that she didn’t do everything as she should have.

    The problem is, if that thought were true, the very same Lo and Alter you mention are even far worse scientists than Singh, as Singh came much closer to a replication of Lombardi et al. (and even consulted the original authors on multiple occasions) than Alter/Lo did. They didn’t do serology, they didn’t do culture, and if they had reported negative results, it would have probably been the easiest study to critique by the forums.

    The problem with your view of the Lipkin study, is that there will always be a possibility for conspiracy. Even though the WPI is controlling and approving every step of the process, nobody can guarantee that the phlebomotist(s) that will collect blood from patients/controls, is not an evil conspirator hired by the still alive Elvis to “bury” this important finding.

    However, in reality, someone independent from WPI will have to collect and/or deidentify and/or store and/or distribute all those samples. This will always leave some opening for conspiracy nuts.

    For the record: in my view, collecting/storing/sending is not a great problem/risk, simply for the fact that hundreds of untrained patients have collected, stored and sent their blood to WPI/VipDx and gotten positive results.

  24. You are saying that Lipkin will be using “known positives” from WPI and that he insists that CDC will validate their assay against these “known positives”?

    If so, where do you get all of this misinformation from? I can guarantee that this will not be the case.

  25. Thus you conceed that you cannot calibrate to the “known positives” of the original authors in areplication study, as “calibrating to known positives” was not part of the original methodology?
    [Note that I am actually addressing your post, while your post doesn’t address my post at all. I’m happy to restate: the point of my post was that, if someone asserts that “X is normal in science”, he should be able to give me an example of “X in science”. I note that you are still unable to do so.

  26. I don’t think it’s so easy. Yes, i have ME/CFS and so the outcome of those studies matter to me and so probably it’s not possible to see things enitrely unbiased. The same is of course then true for the scientists involved themselves and many others, as they have a stake in it as well.

    I’m not saying i don’t trust Ila Singh. I don’t think i’ve ever said that. I don’t know those people and am too far away to be able to judge wheter i can trust them or not.
    What i don’t trust are premature conclusions. I don’t trust the results of many of those studies. Which one is correct is impossible to tell with certainty at the moment, for me.

    In the time since Lombardi et al. i have seen too many questionable things happen to trust the conclusions that are made by investigators, journalists and others after almost every study. So what i don’t necessarily trust is the process and the result, who is honest or not and for what reason is impossible to tell for me, so i will leave it open.
    I do trust the scientific process in that in the end it will get to the truth, but there might be a lot of strange turns before that.
    I think it’s a fact that this is potentially a very big deal, a new human retrovirus that might be in the blood supply, explains previously unexplained illnesses etc. There is a lot of money potentially to be made or lost, carreers are on the line and it’s also a political question. So i think it’s right to be cautious.

    I’m not judging the scientific quality of the studies, i am not qualified or able to do that. I never said a study is good or bad because it’s a replication or not. Also those were Dr. Singh’s first results regarding ME/CFS, i think, so i don’t know why i should have seen her as a reliable scientist before, if i was judging scientsists based on wheter i like the outcomes of their studies or not. But since i’m not aware i’m doing this, i don’t think it matters.

    Regarding the Lipkin study, please note that i never talked about a conspiracy. But as in any critical operation, one should design it in such a way that any possible sort of error or negative influence is ruled out, as far as this is possible. Such errors or negative influences that you can think of, but also other sorts, even things that have never happened before.

    Nothing personal, but please don’t talk about the “still alive Elvis” to discredit my point, that i think is valid.
    I’m pretty sure i would be able to design this study in a way that such influences can be ruled out, so for people in this business it should be even easier. For example, something could be wrong with the tubes. This possibility could be more or less ruled out, if the labs were allowed to provide their own tubes. If the tubes were then randomly assigned to cases or controls, i think it would be safe. In my view, the samles should then never be out of sight of representatives of all the participating labs. What’s at stake here is not smaller than in cases where similar measures are taken, i think. I’m also pretty sure they will think of a way how to store the codes, so that they are safe. Is Ian Lipkin a conspiracy nut because of that? These things should be in everybody’s interest, because it would also more or less prevent that a party will be able to find arguments to attack an outcome they don’t like.

    I don’t think it’s necessary to use the term “conspiracy nut”. I’d prefer to keep it factual. Unless you’d like to call me a conspiracy nut.

  27. Some more questions, if you have the time and don’t mind.

    Do you think the association between XMRV and prostate cancer is true?

    What is your background? Are you a virologist or something of that kind? Just curious. Me, i’m not.

  28. No big deal, but i just realized i didn’t express myself clearly, probably. I meant “i don’t think it’s so simple” in my first sentence of the post above.

  29. I don’t think the association between XMRV and prostate cancer will hold up given the current information I have read.

    I am not a virologist at all, so I will readily admit to not knowing much about the exact specifics.

    Like I said on the other blogpost with now over 500 replies, a person within my family has ME/CFS. While the XMRV link sounded like good news, for me the common sense (although many here will not consider it that) alarm bells started ringing pretty soon (as fast as I heard about the 99% percent positive rate after retesting) As I have always been a follower of (pseudo)science, it was not hard to get interested in this debate after the first negative findings were reported.

    I really regret to say that it is exactly turning out to be as any (probably) false finding, only much more intense than usual (I think for several reasons). That doesn’t mean that this finding cannot be an extraordinary excpetion, but I think it is not very realistic at this point to think that the data supports an XMRV-CFS connection.

  30. Thanks again for the reply. I don’t have time to answer extensively, but I don’t think your a conspiracy nut based on what you wrote. The Elvis example was not for ridicule, but was meant as an extreme example to show that (almost) all conspiracy theories could be true in theory and can at least never be disproven beyond all doubt.

  31. So to take this a step further, and using the current assay, what about these outcomes? We can imagine all the possible results and what they could mean.

    I have shortened this a bit for clarity (e.g. left out healthy controls). Results (made up) are all PCRs from from the Singh lab:

    OUTCOME 1

    (A) Mikovits Lab (WPI) sample positive in the past: POSITIVE

    (B) Spiked human blood sample (with XMRV): POSITIVE

    (C) Fresh blood sample from someone positive in the Mikovits study: NEGATIVE

    (D) Assay negative control: NEGATIVE

    (E) Assay positive control (XMRV clone): POSITIVE

    Means: Singh lab assay works from patient samples known to have XMRV. No XMRV is found in fresh samples. Calls into question the link between XMRV with CFS.

    OUCOME 2

    (A) Mikovits Lab (WPI) sample positive in the past: NEGATIVE

    (B) Spiked human blood sample: POSITIVE

    (C) Fresh blood sample from someone positive in the Mikovits study: NEGATIVE

    (D) Assay negative control: NEGATIVE

    (E) Assay positive control (clone): POSITIVE

    Means: Singh lab assay works with lab derived clone but does not work from human blood, so can’t answer the question. Casts no light on the link between XMRV and CFS either way.

    (NB – B could also be negative and lead to this conclusion. If A & B are not the same it casts doubt on the assay.)

    OUTCOME 3

    (A) Mikovits Lab (WPI) sample positive in the past: POSITIVE

    (B) Spiked human blood sample: POSITIVE

    (C) Fresh blood sample from someone positive in the Mikovits study: POSITIVE

    (D) Assay negative control: NEGATIVE

    (E) Assay positive control (clone): POSITIVE

    Means: Singh lab has the same findings as the Mikovits lab and suggests a link between XMRV and CFS.

    (Note that (C) has been done and is negative. This means the only way this could happen is after outcome 2, above, followed by changing the assay.)

    OUTCOME 4

    (A, B C): all POSITIVE

    (D) Assay negative control: POSITIVE

    (E) Assay positive control (clone): POSITIVE

    Means: PCR is messed up or contaminated – repeat the PCR.

    OUTCOME 5

    (A, B C): all NEGATIVE

    (D) Assay negative control: NEGATIVE

    (E) Assay positive control (clone): NEGATIVE

    Means: Taq enzyme has gone off/PCR machine doesn’t work/power cut/other reagent not working etc. Repeat the PCR.

    This all hypothetical, of course. But alot seems to hang on whether the Singh study is replication on this blog (replication not to be confused with vindication). I propose that his would test the degree to which the Singh study is or isn’t a true replication.

    Bascially, Singh takes XMRV positive samples from the Mikovits lab. If she finds XMRV, it shows her assay works.

  32. Ok, no problem. For me, the high percentage of XMRV being reported, actually was something that convinced me. Because i never believed in that hypothesis that ME/CFs is some sort of “collection” of many different illnesses. Sure, it’s not exactly the same in every patient, but i guess most illnesses are like that. For me it was always very puzzling that they were not able to find something that’s wrong in all of us, the same thing. So when they corrected the number up, that felt like now it fits better with what i would expect.

    What i agree with is that i think if the link to ME/CFS were to be disproven i would expect the same for prostate cancer.
    I don’t really see why they think that association is more likely to be true.

  33. 1. I’m not a “her”

    2. I don’t have a PhD (at least not yet).

    3. That copy pasted stuff is hardly evidence of anything.

    4. At least I’m not pretending to be someone else.

    5. You’re coward if by hiding behind someone else’s screen name.

  34. Dear Dr. Racaniello

    There’s something i’d like to know. Assuming the studies that found XMRV in prostate tissue were correct and XMRV is in the population, would it even be possible that the virus is not present in the blood even at low levels in all of the tested subjects? Can a virus be in the body, but absolutely absent from the blood?

    Also there was a study from Germany last year (Fischer et al.) that found XMRV in respiratory secretions. Why wasn’t this pursued any further? Dr. Singh, as long as she believes in her prostate cancer work, probably also believes in a background rate of XMRV in the population, of a couple of percent, as a number of studies have found XMRV in healthy controls. If she couldn’t find XMRV in the blood, why not try respiratory secretions to try to verify or falsify the positive studies?

    Thanks

  35. That is not a direct quote from Coffin but a conclusion by the journalist.
    However, since you ask, I’ll quote Coffin from the most recent ScienceMag article about XMRV:”Singh bent over backward to try to use the same assays as published, allowing her to knock down what I consider to be a real straw man but that nonetheless was out there,”http://news.sciencemag.org/scienceinsider/2011/05/more-bad-news-for-chronic-fatigue.html?ref=hp

  36. The WPI and NIH/FDA will use validated assays, proven to detect the virus. Lipkin being a scientist will ensure the CDC do the same.

  37. Using unvalidated assays renders the results of studies worthless. However, it is not the only parameter that can influence the results and at this time much is unknown about this retrovirus. Shin et al was not a replication study as the methodology was different, if it was a replication study, the assay would have already have been validated and there would be no need for clinically positive samples.

  38. Coffin: CROI 16

    “To say something is successul because you can’t detect it (…) doesn’t mean it is not there. (…) it’s telling you in a sense more about the inability of your assay to detect something than about what is going on. “

  39. If you don’t replicate the only other way to validate your assays and prove they can detect the virus is to use clinically positive samples. That is the only way to dispute the findings of the two positive papers Lombardi and Lo et al. If you want to claim those patients do not have the virus.

Comments are closed.

Scroll to Top