Advertisement
Research Article

No Evidence for XMRV in German CFS and MS Patients with Fatigue Despite the Ability of the Virus to Infect Human Blood Cells In Vitro

  • Oliver Hohn,

    Affiliations: Centre for Biological Security 4, Robert Koch-Institute, Berlin, Germany, Centre for Retrovirology, Robert Koch-Institute, Berlin, Germany

    X
  • Kristin Strohschein,

    Affiliation: Institute for Medical Immunology, Charité - Universitätsmedizin Berlin, Berlin, Germany

    X
  • Alexander U. Brandt,

    Affiliation: NeuroCure Clinical Research Center (NCRC), Charité Universitätsmedizin Berlin, Berlin, Germany

    X
  • Sandra Seeher,

    Affiliation: Centre for Biological Security 4, Robert Koch-Institute, Berlin, Germany

    X
  • Sandra Klein,

    Affiliation: Centre for Biological Security 4, Robert Koch-Institute, Berlin, Germany

    X
  • Reinhard Kurth,

    Affiliation: Robert Koch-Institute, Berlin, Germany

    X
  • Friedemann Paul,

    Affiliation: NeuroCure Clinical Research Center (NCRC), Charité Universitätsmedizin Berlin, Berlin, Germany

    X
  • Christian Meisel,

    Affiliation: Institute for Medical Immunology, Charité - Universitätsmedizin Berlin, Berlin, Germany

    X
  • Carmen Scheibenbogen equal contributor,

    equal contributor Contributed equally to this work with: Carmen Scheibenbogen, Norbert Bannert

    Affiliation: Institute for Medical Immunology, Charité - Universitätsmedizin Berlin, Berlin, Germany

    X
  • Norbert Bannert equal contributor mail

    equal contributor Contributed equally to this work with: Carmen Scheibenbogen, Norbert Bannert

    bannertn@rki.de

    Affiliations: Centre for Biological Security 4, Robert Koch-Institute, Berlin, Germany, Centre for Retrovirology, Robert Koch-Institute, Berlin, Germany

    X
  • Published: December 22, 2010
  • DOI: 10.1371/journal.pone.0015632

Reader Comments (1)

Post a new comment on this article

A critique of the methodoloy of Hohn et al 2010

Posted by gerwynm on 04 Jan 2011 at 16:42 GMT

No information is given regarding the diagnosis of the participants in regard of who made it. No mention is made of their symptomology. This is contrary to the Fukuda guidelines (1994). A different patient cohort to the one investigated in the Lombardi et al study (2009) would be a clear confounding variable

There is also no information given regarding blood collection and handling proceedures used. Such proceedures are critical as far as the survival of viral DNA or RNA are concerned. Problems in this area alone would explain a negative finding. Also the use of heparin tubes is considered suboptimal for PCR purposes and indeed many virology laboratories refuse to accept heparinised blood for PCR

Elisa based assays have either persistently failed to detect wild type XMRV (Hohn et al 2008, Switzer et al 2010) or have produced detection rates several orders of magnitude lower than that detected by other serological assays (Arnold et al 2010, Schalberg et al 2009). The latest Elisa assay produced by Hohn et al in this study has demonstrated sensitivity and specificity towards mulv proteins but once again, sadly, not towards wild type XMRV

The amplification assay used has the real potential to induce LAK activity which could selectively lyse cells containing xmrv totally destroying the viral DNA. Hence the amplification method used could well have had the opposite of the desired effect

Hohn et al used the nested PCR approach used by Urisman and others (2006) and Switzer et al 2010. Unfortunately they did not use the same primers as Urisman et al which have proven ability to detect the virus but chose the primers of Switzer et al which are without any suchrecord of success

They did not use primers with any proven ability to detect the virus not even the ones from a recent study in which Oliver Hohn was involved

Oliver Hohn was involved in a recent study which did detect XMRV in 10% of immunocompromised patients and 3% of people who were not (Fischer, 2010). This successful method was not used in this study

Thus a study failing even to detect even population levels of XMRV is indicative of inadequate methodology

Primers based on the GAG sequence of the VP62 clone as is the case in this study have never been able to detect wild type virus. Indeed Danielson et al (2010) have clearly demonstrated that primers so constructed are incapable of detecting XMRV in known clinically positive samples.Why researchers persist with such an approach is somewhat of a mystery

PCR is comfortably the least sensitive method for detecting XMRV (Schalberg et al 2009) (Arnold et al 2010) (Mikovits et al 2010). Therefore any conclusion based on PCR and or unproven assays regarding the absence of XMRV are fundamentally unsafe


References

K Fukuda; SE Straus; I Hickie; MC Sharpe; JG Dobbins; A Komaroff; and International Chronic Fatigue Syndrome Study Group; The Chronic Fatigue Syndrome: A Comprehensive Approach to Its Definition and Study Annals of Internal Medicine Dec 15, 1994;vol 121:12: pp953-959

VC Lombardi, FW Ruscetti, J Das Gupta, MA Pfost, KS Hagen, DL Peterson, S K. Ruscetti, RK. Bagni, C Petrow-Sadowski, B Gold, M Dean, RH Silverman and J A Mikovits Detection of an Infectious Retrovirus, XMRV, in Blood Cells of Patients with Chronic Fatigue Syndrome Science 23 October 2009: Vol. 326 no. 5952 pp. 585-589 DOI: 10.1126/science.1179052

O Hohn, H Krause, P Barbarotto, L Niederstadt, N Beimforde, J Denner, K Miller , R Kurth and N Bannert; Lack of evidence for xenotropic murine leukemia virus-related virus (XMRV) in German prostate cancer patients; Retrovirology 2009, 6:92doi:10.1186/1742-4690-6-92

Switzer WM, et al. (2010) Absence of evidence of xenotropic murine leukemia virus-related virus infection in persons with chronic fatigue syndrome and healthy controls in the United States. Retrovirology 7:57.

RS Arnold, NV Makarova, AO Osunkoya, S Suppiah, TA Scott, NA Johnson, SM Bhosle, D Liotta, E Hunter, FF Marshall, H Ly, RJ Molinaro, JL Blackwell, JA Petros; XMRV Infection in Patients With Prostate Cancer: Novel Serologic Assay and Correlation With PCR and FISH; Urology - April 2010 (Vol. 75, Issue 4, Pages 755-761, DOI: 10.1016/j.urology.2010.01.038)

Schlaberg R, Choe DJ, Brown KR, Thaker HM, Singh IR (2009) XMRV is present in malignant prostatic epithelium and is associated with prostate cancer, especially high-grade tumors. Proc Natl Acad Sci USA 106:16351–16356

Urisman A, Molinaro RJ, Fischer N, Plummer SJ, Casey G, et al. 2006 Identification of a Novel Gammaretrovirus in Prostate Tumors of Patients Homozygous for R462Q RNASEL Variant. PLoS Pathog 2(3): e25. doi:10.1371/journal.ppat.0020025

BP Danielson, GE Ayala and JT Kimata; Detection of Xenotropic Murine Leukemia Virus-Related Virus in Normal and Tumor Tissue of Patients from the Southern United States with Prostate Cancer Is Dependent on Specific Polymerase Chain Reaction Conditions; J Infect Dis. (2010) 202 (10): 1470-1477. doi: 10.1086/656146


N Fischer, C Schulz, K Stieler, O Hohn, C Lange, C Drosten, and M Aepfelbacher; Virus-related Gammaretrovirus in Respiratory Tract; Emerging Infectious Diseases, Vol. 16, No. 6

JA Mikovits, Y Huang, MA Pfost, VC Lombardi, DC Bertolette, KS Hagen and FW Ruscetti; Distribution of Xenotropic Murine Leukemia Virus-Related Virus (XMRV) Infection in Chronic Fatigue Syndrome and Prostate Cancer; AIDS Rev 2010; 12: 149-52

No competing interests declared.

RE: A critique of the methodoloy of Hohn et al 2010

RRM_ replied to gerwynm on 05 Jan 2011 at 10:47 GMT

A few remarks to the comment by Gerwynm:

- The original study (1) did actually use heparin tubes themselves, (2) while the study by Switzer et al (3) received criticism from patient organisations for not using the same heparin tubes. (4) The criticism aimed at this study does not seem to be consistent with these observations.

- The patients are pretty well described in the methods section, actually much better than was the case in the original study. (1)(2). On their web site, the principal investigators of the original study (1) state that "there is nothing unique about these patients" (5) from the original study (1).

- 'Wild type XMRV' has no scientific merit in this discussion. Whether or not Lombardi et al (1) discovered 'wild type XMRV' in the blood of CFS patients is exactly what studies from independent laboratories are trying to determine. Assuming a priori that they have in fact found 'wild type XMRV' is thus a critical logical error that leads to circular reasoning. If true, every investigator that would erronously report a finding in the future, may then conclude that subsequent studies are just failing at finding what he or she has "found".

- The original study by Lombardi et al. (1) did not supply other researchers with an exact blood collection and processing protocol. It is the task of the original researchers to provide the scientific community with methods that make it possible to replicate the reported results, not the other way around.

References

1. Lombardi VC, Ruscetti FW, Das Gupta J, et al. (October 2009). "Detection of an infectious retrovirus, XMRV, in blood cells of patients with chronic fatigue syndrome". Science 326 (5952): 585–9. doi:10.1126/science.1179052. PMID 19815723.

2. Lombardi VC, Ruscetti FW, Das Gupta J, et al. (October 2009). "Supporting Online Material for Detection of an Infectious Retrovirus, XMRV, in Blood Cells of Patients with Chronic Fatigue Syndrome". Science 326 (5952): 585–9. doi:10.1126/science.1179052. PMID 19815723

3. Switzer WM, et al. (2010) Absence of evidence of xenotropic murine leukemia virus-related virus infection in persons with chronic fatigue syndrome and healthy controls in the United States. Retrovirology 7:57.


4. Suzanne Vernon. "Blood from a stone". http://www.cfids.org/xmrv...


5. http://www.wpinstitute.or...

No competing interests declared.

RE: RE: A critique of the methodoloy of Hohn et al 2010

omerbasket replied to RRM_ on 09 Jan 2011 at 19:16 GMT

Actually, the original study supplied a long, detailed desctription of the exact blood collection and processing protocol:
www.sciencemag.org/cgi/co...

No competing interests declared.

RE: RE: RE: A critique of the methodoloy of Hohn et al 2010

omerbasket replied to omerbasket on 09 Jan 2011 at 19:28 GMT

The direct link is the following one:
http://www.sciencemag.org...

No competing interests declared.

RE: A critique of the methodoloy of Hohn et al 2010

gerwynm replied to gerwynm on 08 Jan 2011 at 22:15 GMT


Thank you for your opinion that the term wild type is not appropiate.you do not however explain why you hold that opinion.


Ironically satisfying the consensus criteria while being a comment means that the patient must have post exertional malaise and neuroendocrine symptoms otherwise they would not fulfill the selection criteria

Thus your comment that this study contained more patient information than lombardi et al is entirely erroneous

As I,m sure that you are aware the study by lombardi et al consisted of three seperate experiments.

The experiment that hohn et al have attemped to partly replicate did not involve heparin .Thus you are once again mistaken

One of the experiments did involve drawing a fresh blood sample into heparinised tubes which were then handled according to recognised methodology with the various being seperated and optimally stored within six hours of the draw, sadly we have no information about the blood collection and storage proceedures in the hohn study. prolonged exposure to heparin(again as I am sure you are aware) is not appropiate if a PCR assay is to be carried out on that sample.

Inadequate blood collection alone would easily explain a false negative result

A false negative is ,almost cetainly, due to study design flaws as the methodology could not even detect the expected levels of XMRV in the general population detected by a study involving Oliver Hohn himself

your apparent objection to the use of the term wild type is somewhat suprising as it is it is a standard term which differentiates a real virus with, an unknown level of sequence variability, from a synthetic clone used by Hohn et al. Using a PCR assay which has had its sensitivity demonstrated by its ability to detect this clone in a spiked blood or water sample has always failed to detect wild type virus.I am quite happy to use the term real virus if that would help clarify my comments

I note that you are resident in Afghanistan. Please accept my best wishes for the future of your country and that of your fellow citizens

references as per the original article

No competing interests declared.

RE: RE: A critique of the methodoloy of Hohn et al 2010

RRM_ replied to gerwynm on 09 Jan 2011 at 10:57 GMT

First, I note that you are from the UK. I would strongly advise you to seek some help in English or, if you are a dyslectic, in getting a friend or relative to help you write these kinds of responses. I do simply not understand the actual meaning of some of the sentences you have written and thus are unable to respond to some of the arguments that you are trying to articulate. However, as I am not from an English speaking country myself, I am sure some of the confusion arises from my own lack of knowledge of the English language, so please accept my apologies in not understanding much of what you have written.

That said, I think the most important issue you seem to have with my response is the comment regarding 'wild type XMRV', so I will focus my attention on that subject.

First, it is worthy to note that I did explain why I hold my opinion regarding 'wild type XMRV'. For your reference, I did so right after my assertion that the construct has no scientific merit in this discussion. But because I was apperently unable to covey the intented argument to you, I will rephrase using a (hopefully) simple example:

Suppose I am a retrovirologist, and a lousy one at that. I have no idea what I am doing, and one day I decide to check for Some Magical New Retrovirus (SMNR) in patients that have broken their legs. Because I am a lousy retrovirogist, I somehow manage to (erronously) find SMNR in people with broken legs and not (or less so) in healthy controls. I publish my 'SMNR is correlated with broken legs' in a journal.

All the good scientists in the world will try to duplicate my findings. Some will use my primers and some of my other methods, while others will search it using their own methodology. Suppose they will not be able to find anything that confirms my initial findings, simply because my SMNR findings are incorrect (which is what those other scientists are independently trying to determine).

Now I, the lousy scientist, am easily able to set aside all criticism. How? I will just say that all those other scientists have never shown that their assays would work on 'wild type SMNR' in patients with broken legs. In fact, no investigation could ever prove me or my methods wrong, because I would be the "brilliant" scientist, the only one in the world that can detect the bug in "wild" samples.

Don't you agree that it would be unscientific for me to use this argument when other scientists are trying to determine if SMNR is "real"? It has no scientific merit because we would be arguing in circles. When you are trying to independently determine whether a previous investigation is correct, it is unwise to a priori assume this previous investigation is correct. Any scientist that would calibrate his assays against my "positive wild samples", would not be actually detecting SMNR, but would at best be detecting the same junk (e.g. high levels of paracetamol) as I had been actually finding (thinking it was some magical new retrovirus).

Therefore the use of "wild samples" to calibrate assays has no scientific merit in the discussion regarding whether XMRV is a real human pathogen that is connected to ME/CFS or not.

Even beyond the above, your argument is very flawed scientifically (that is, if I am understanding you correctly). Your argument, that the "real virus" is of an "unknown level of sequence variability", does apply to your proposed 'wild type' method equally: even if a lab tested their assays against a WPI sample (or 2, or 10) of 'wild type XMRV', as you propose, it would give no more guarantees in any way that the XMRV in the 'wild' samples that they would be actually testing for their study after that, were also detectable by that assay.

No competing interests declared.

RE: RE: RE: A critique of the methodoloy of Hohn et al 2010

LC12 replied to RRM_ on 09 Jan 2011 at 14:42 GMT

The existence of wild-type XMRV is not in question. Several studies have merely suggested that it may be a mouse virus, but have not provided evidence to support this conclusion.

No competing interests declared.

RE: RE: RE: RE: A critique of the methodoloy of Hohn et al 2010

RRM_ replied to LC12 on 09 Jan 2011 at 16:11 GMT

Two corrections to your short response:

- The existence of 'wild type XMRV' is very much in question. It is exactly what are scientists are currently trying to determine. It would therefore be unscientific to use samples that are positive for XMRV according to some scientists, using unvalidated assays.

- Several studies have provided evidence for the hypothesis that XMRV is a laboratory artifact instead of a retrovirus that is actually infecting humans (1). You probably mean that these studies have not provided physical evidence that the original studies had actually contaminated their samples. That would be true (for the most part*), but science isn't 'Murder, she wrote": in science you test a hypothesis through experimental evidence. When all of your experiments support a given hypothesis, this would be considered evidence for that hypothesis. When the experiments from others support that hypothesis, the evidence could become considered to be (temporarily) conclusive by the scientific community. And the fact of the matter is that all experiments done in the Retrovirology paper support the hypothesis that XMRV is a contaminant. In other words, you wouldn't expect these results a priori if XMRV weren't a laboratory contaminant (ceteris paribus). Of course, it is always possible to later explain away evidence within any given hypothesis, which (understandably) is precisely what patients are doing on internet forums.

I would like to add that at this moment, the case 'for' XMRV nor the case 'against' it, is considered to be conclusive by the scientific community.

References

1. Robert A Smith. "Contamination of clinical specimens with MLV-encoding nucleic acids: implications for XMRV and other candidate human retroviruses". Retrovirology 2010, 7:112doi:10.1186/1742-4690-7-112. (See also the references therein)

*The fact that Sato et al (see the reference above) found that the brand of reagents that both Lo et al and Lombardi et al had used in their respective studies, is contaminated by mouse DNA, is of course some evidence for actual contamination of those studies but, in itself, not enough evidence.

No competing interests declared.

RE: RE: RE: RE: RE: A critique of the methodoloy of Hohn et al 2010

LC12 replied to RRM_ on 09 Jan 2011 at 17:48 GMT

I think you misunderstand. No one has claimed that XMRV does not exist, merely that it is either a mouse or human retrovirus.

No competing interests declared.

RE: RE: RE: RE: RE: A critique of the methodoloy of Hohn et al 2010

LC12 replied to RRM_ on 09 Jan 2011 at 18:17 GMT

XMRV is the name given to a wild-type virus. The naming convention used will be incorrect, IF evidence demonstrates XMRV to be a mouse retrovirus, not a human one. This does not in anyway impact on whether the virus is "real".

You state that it would be unscientific to use samples that are positive. Well if those samples did not exist, according to your reasoning, then how could they be used and what possible impact could they have? But more importantly what should they not be used for? You do not say.

Several studies hypothesised that XMRV was not a human retrovirus, but they did not state at any time that it was a mouse retrovirus.

You also state that if "all your experiments support a given hypothesis, this would be considered evidence for that hypothesis." Yet how can you conclude that all experiments have been conducted? I think you will find that experiments have been quiet focus, and there are a multitude of further experiments which could be and should be undertaken before assuming that all basses have been covered.

Your next statement is also a bit of a puzzle. I'm guessing here because again you are unclear, and do not indicate which paper you are referring to, but I will make a guess that it is Hue et al. The authors of this paper did not support the hypothesis of XMRV being a contamination. However, a press release did, and this may be the source of your confusion. They in fact stated that: "We suggest that XMRV as a human virus does not conform to this criterion." The word 'suggest' is the important one, they do not say prove. And they also state that: "We conclude that future screens for MLV-related sequences use more rigorous PCR containment procedures" Not MLV sequences, as in mouse, but 'related' ones.

Nothing in science is ever conclusive, but you knew that right.

No competing interests declared.

RE: RE: RE: RE: RE: RE: A critique of the methodoloy of Hohn et al 2010

RRM_ replied to LC12 on 09 Jan 2011 at 19:02 GMT

- I am not saying that it is never prudent to use confirmed positive samples. However, at this moment in time, there are no confirmed positive samples as there is no validated assay that scientists agree on. Some "proponents" of XMRV are arguing that assays should be validated using a "confirmed" XMRV positive sample but, as I explained, that would not be a wise thing to do: we would then not be independently confirming results but might be making the very same mistakes the original scientists might have made - that is not the way to go for independent validation.

- The rest of your arguments and questions regarding using known positives are based on misconceptions. I was merely responding to a previous comment. If you do not understand the value (or rather the lack thereof) of using 'wild type XMRV' in independent validation studies I suggest you ask that commenter about the proper context.

- I was not specifically referring to Hue et al. Like I explained, all the studies that were recently published in Retrovirology, supported the hypothesis that XMRV is a laboratory contaminant through evidence gathered by experiments. I also explained how that evidence does not conclusively prove that Lombardi et al and Lo et al actually contaminated their samples.

- Nothing in science is conclusive, which is why I explicitly used the word "temporarily" the first time I explained how science got its conclusive answers. You did read that, right?

- How would one ever prove the original studies to be incorrect if they happened to be incorrect? Would you propose a method, please? Or are the original studies infalsifiable according to you?

No competing interests declared.

RE: RE: RE: RE: RE: RE: RE: A critique of the methodoloy of Hohn et al 2010

LC12 replied to RRM_ on 09 Jan 2011 at 20:11 GMT

So now you are admitting that wild-type XMRV does exist, and are now saying that the clinical positives of those able to find XMRV should not be used to calibrate assays. Are you sticking with this conclusion now? However, yet again, you do not explain why the use of these positives would be a problem?

Again, you have erroneously attributed conclusions to not one but several papers. None of the four papers published in Retrovirology on the 20th December 2010, stated that XMRV is a laboratory contaminant.

Robinson et al. actually stated that, "It is true that we cannot formally rule out the possibility that the samples in question are infected with XMRV and simultaneously contaminated by mouse DNA"

Oakes et al. stated that, "..our data are compatible with the conclusion that the detection of MLV-related sequences in human genomic DNA samples could be due to contamination with minute and variable quantities of mouse DNA, most likely contained in various laboratory reagents." Key words here are 'could' and 'likely', not proven. Contaminating your own lab with MLVs is not proof that XMRV is a contaminant, it merely shows to anyone, and I don't know who they would be, that it is possible to do so.

Finally Sato et al. states that "Therefore, in the investigation of XMRV infection, researchers should prudently evaluate test kits for the presence of genomes of endogenous MLV as well as XMRV." Again they do not prove contamination, it is again a study that highlights the dangers of mouse contamination. But who doesn't know that.

In conclusion, no study has demonstrated that XMRV is a contaminant, they have merely suggested the possibility, by either finding mouse contamination in their own labs or kits, and by trying to fit some evidence to a hypothesis of contamination being the cause of negative studies.

Only further studies and investigations will settle the issue of why some are unable to detect the retrovirus. Your final comment is quiet revealing, thank you.

No competing interests declared.

RE: RE: RE: RE: RE: RE: RE: RE: A critique of the methodoloy of Hohn et al 2010

RRM_ replied to LC12 on 10 Jan 2011 at 07:49 GMT

You do not seem capable of understanding some of my arguments. For instance, where have I stated, explicitly or implicitly, that I 'have admitted wild type XMRV'does exist?

Also, the fact that Sato et al found that the brand of reagents which were used by Lombardi et al and Lo et al can be contaminated IS in itself evidence . I repeat that this finding in itself is not conclusive evidence.

Finally, you quote selectively. For instance, Hue et al explicitly state that:

“XMRV detected by sensitive PCR methods in patient samples is the likely result of PCR contamination with mouse DNA.”

This already makes your statemant regarding the four papers incorrect.

No competing interests declared.

RE: RE: RE: RE: RE: RE: RE: RE: RE: A critique of the methodoloy of Hohn et al 2010

LC12 replied to RRM_ on 10 Jan 2011 at 13:22 GMT

This is what you previously said "The existence of 'wild type XMRV' is very much in question." Now you are saying it is not.

Sato et al. found a one-step RT-PCR kit from Invitrogen were contaminated with RNA derived from PmERV. They did not find that reagents were contaminated. Lombardi et al. never used that testing kit.

Again, the quote you provide uses the word 'likely' not proven.



No competing interests declared.

RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: A critique of the methodoloy of Hohn et al 2010

RRM_ replied to LC12 on 10 Jan 2011 at 14:17 GMT

"This is what you previously said "The existence of 'wild type XMRV' is very much in question." Now you are saying it is not."

Where exactly have I said "it is not" in question?

"Sato et al. found a one-step RT-PCR kit from Invitrogen were contaminated with RNA derived from PmERV. They did not find that reagents were contaminated. Lombardi et al. never used that testing kit."

Sato et al did not test the kit that Lombardi et al or Lo et all used. sato et al, however, did test a kit from the same brand that was used by Lombardi et al. They found that kit (I believe out of a total of 4 kits) to be contaminated.

No competing interests declared.

RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: A critique of the methodoloy of Hohn et al 2010

LC12 replied to RRM_ on 10 Jan 2011 at 15:20 GMT

Top of your third post you said "The existence of 'wild type XMRV' is very much in question.", now you are saying that wild-type XMRV does exist.

You originally said that Sato et al. tested a reagent, now you are saying testing kit. At least you are learning. I think the rest of your comment again speaks for itself.

No competing interests declared.

RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: A critique of the methodoloy of Hohn et al 2010

RRM_ replied to LC12 on 10 Jan 2011 at 15:54 GMT

I will explain this one more time very clearly. If one asserts that an other person has made contradictory statements, both statements are essential to provide evidence for this assertion.

I have repeatedly agreed to have said that the existence of 'wild type XMRV' is still very much in question. I have however never stated that there is at this moment scientific consensus on the assertion that 'wild type XMRV' does exist.

But even if it did, it wouldn't matter much to the argument, as I have explained: there wouldn't be positive samples that all scientists would agree on, because of the lack of a proper, reliable, reproducable, standardized, assay. If there were samples we would agree on, scientists wouldn't be having these discussions in the first place. So using samples to somehow solve the puzzle, is circular reasoning.

The root of science is reproducability. One thus cannot rely on the findings of the original researchers. This should not pose a problem because the original researchers also didn't have to rely on the findings of the original researchers (which would be themselves). This logically proves that any scientific experiment that is true, can be independently reproduced by others.

Instead of cherry picking semantic points, can you please cover this argument I have repeated several times? Otherwise, enjoy your perceived victories in online discussions. Simon Wesseley would find confirmation of his findings if he were reading this.

No competing interests declared.

RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: A critique of the methodoloy of Hohn et al 2010

LC12 replied to RRM_ on 10 Jan 2011 at 22:19 GMT

"As yet, we can really only refer to one virus, and that's XMRV. That's the only one of these sequences that we know is in a replicating virus."

Dr John Coffin, '1st International Workshop on XMRV: Q&A Wrap-up Session' (NIH, 8 September 2010, 3mins 32secs in) http://videocast.nih.gov/...

Coffin may still be unsure if it's a mouse or human retrovirus, but one thing is certain, that XMRV is a virus. A wild-type virus.

[Part of this comment has been deleted because it does not comply with PLoS policies.]

No competing interests declared.

RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: A critique of the methodoloy of Hohn et al 2010

RRM_ replied to LC12 on 11 Jan 2011 at 07:46 GMT

In no way, shap or form does that quote establish that XMRV is a 'wild type' virus. Furthermore, you should really use good references, for instance from peer reviewed papers. A quote from a scientist, no matter how wel resepected, can never substantiate your argument (even if it supported your argument, which it doesn't).

[Part of this comment has been deleted because it does not comply with PLoS policies.]

No competing interests declared.

RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: A critique of the methodoloy of Hohn et al 2010

LC12 replied to RRM_ on 12 Jan 2011 at 23:45 GMT

Amusing, but no. There are no researchers saying the retrovirus does not exist. Only that it is either in humans or in mice. Is that simple enough?

No competing interests declared.

RE: RE: RE: A critique of the methodoloy of Hohn et al 2010

gerwynm replied to RRM_ on 09 Jan 2011 at 20:12 GMT


I am sorry that you have difficulties with English

I will endevour not to use any technical terms

Danielson and others have now demonstrated that while no specific PCR conditions are required to detect a synthetic (artificial) copy of a virus in vitro(in a test tube) this is not true when trying to detect a real human gammaretrovirus in the blood of a person known to be infected. Danielson and others demonstrated that their PCR assay which could easily detect very few artificial virus copies in a test tune could not detect a real virus in a person known to be infected by other detection methods

So far PCR assays which can easily detect xmrv in 22rv1 cells or the vp62 clone in a spiked sample have a 100% failure rate in detecting viruses in people known to be infected

The people who have used the same PCR conditions when viewed as a whole have a 100% of success

[Part of this comment has been deleted because it does not comply with PLoS policies.]

No competing interests declared.

RE: RE: RE: RE: A critique of the methodoloy of Hohn et al 2010

RRM_ replied to gerwynm on 10 Jan 2011 at 07:55 GMT

Thank you for your response. The use of technical terms is not a concern to me. I was merely confused by the lack of proper grammar an spelling in your reply to me. For instance, I am not confused by the use of a "test tube" as technical term, but I wouldn't understand what a "test tune" is.

- Danielson et al have not conclusivey demonstrated what you think they have demonstrated, at least not to my knowledge. Can you please supply convincing evidence for your assertion, instead of mere conjecture?

- Like I have argued twice, you are reasoning in circles when you are arguing that some PCR assays "have a 100% failure rate in detecting viruses in people known to be infected" while others (presumably the studies you "support") "have a 100% of success". Please see my reply and example above to understand why your reasoning is circular. For further help you could also check:
http://en.wikipedia.org/w...

No competing interests declared.

RE: RE: RE: RE: RE: A critique of the methodoloy of Hohn et al 2010

LC12 replied to RRM_ on 12 Jan 2011 at 23:54 GMT

Danielson et al. Did indeed come to the conclusion Gerwyn highlights. Only further research could potentially dismiss this finding.

You fail to deal with the current research in it's entirety and are singling out information to support your beliefs. There is a distinct difference between the positive and negative studies.

No competing interests declared.

RE: RE: RE: A critique of the methodoloy of Hohn et al 2010

asleep replied to RRM_ on 10 Jan 2011 at 17:46 GMT

I like stories too, so let me tell another equally plausible version of your tale:

A group of scientists find evidence of Very Real Human Retrovirus (VRHR) in a heretofore highly maligned disease. From a divine, omnipotent perspective it is clear that this virus is a true human pathogen. But we lowly humans must conduct scientific inquiry to determine this.

Humans have previously discovered VRHR and have created a cloned reference strain. By some cosmic misfortune, it just so happens that this clone differs significantly in certain regions from "wild type" strains infecting humans. The scientists who found VRHR used probes that (by luck or by insight) target reasonably well-preserved regions of "wild type" strains. This is how they were able to detect the virus in human samples.

After publication, others attempt to find the virus by probing for clone-based regions that are not well-preserved in "wild type" strains. Naturally, they fail to find it. However, in a move that it surprising from both an omnipotent as well as scientific perspective, they declare that VRHR is not a human pathogen.

These subsequent negative studies are then put forward as proof that the original study is false, though they have not actually proven this.

Though this is a story, it should help you understand the flaw in your reasoning: the inability to detect something by a *different* method is not conclusive proof that the original finding is false. It is proof that the method employed cannot detect virus in the samples tested, nothing more.

You ask how the Lombardi et al finding could be falsified. Presumably any serious attempt at falsification would begin with a replication of their precise methods to see if one can find what they are finding. Then, from this common ground, independent inquiry could attempt to prove that what is found is not an exogenous virus (but instead an artifact or junk). To not proceed along this route (esp. after others have shown that a clone-based probe is inadequate) is scientifically lazy if not outright disingenuous.

No competing interests declared.

RE: RE: RE: RE: A critique of the methodoloy of Hohn et al 2010

RRM_ replied to asleep on 10 Jan 2011 at 21:56 GMT

Thank you very much for a response that actually addresses the topic at hand.

However, I have neverv stated that 'the inability to detect something by a *different* method is conclusive proof that the original finding is false'. I and every other scientifically trained mind in the world would agree that such a statement would be incorrect. Your argument fails because it relies on the assumption above.

Like I have explained above, reproducability lies at the root of finding objective truth through the scientific method. One cannot rely on the findings of the original researchers in the way that the original commenter is proposing. The following is important to understand: this should not pose a problem at all because the original researchers also didn't have to rely on the findings of the original researchers (which would be themselves). This logically proves that any scientific experiment that is true (from that divine, omnipotent perspective you speak of), can be independently reproduced by others.

The fact that you consider your story to be equally plausible to my story (although I wouldn't agree, it may very well be), does not solve the circular problem of your story. Your methodology would only get the correct result if the original result were true, and even then perhaps because of the wrong reasons. It thus would add little if any value to the original result, just like the findings of RedLabs or VipDx add little to the argument of the original researchers.

But even beyond the above, the argument for using 'wild type XMRV' is very flawed scientifically (as I have already explained). The argument, that the "real virus" is of an "unknown level of sequence variability", does apply to the proposed 'wild type' method equally: even if a lab tested their assays against a WPI sample (or 2, or 10) of 'wild type XMRV', as you propose, it would give no more guarantees in any way that the XMRV in the 'wild' samples that they would be actually testing for their study after that, were also detectable by that assay. For what it's worth, the as of yet unpublished Blomberg study did test samples that were supplied by the WPI - I assume Blomberg was sent 'wild type' samples. Still, it didn't satisfy any proponents of a connection between XMRV and CFS: more sequence diversity was assumed and this negative study was just as quickly dismissed as the other studies.

To call the dozens of scientists that have tried to solve this puzzle to the best of their efforts "lazy"is pretty ironic considering that they have done more than you have, while asserting it may be "outright disingenuous" is against basic scientific principles.

Finally, don't forget that your accusation also applies to Alter and Lo, which also haven't replicated the original researchers' methods: surely they are also "scientifically lazy if not outright disingenuous"? I guess you would answer that they are not, merely because you happen to like their results?

No competing interests declared.

Faulty logic

peterw replied to RRM_ on 12 Jan 2011 at 21:30 GMT

RRM, with respect this reasoning is completely fallacious and reminds me of an old joke:- Goofy tells Mickey he has lost a coin outside. Mickey later finds him looking for it in his room. When questioned "why" he says "because the light is better in here".

Or, to put it in your terms:- suppose you are a brilliant retrovirologist. You do indeed discover a retrovirus that correlates with broken legs. Others attempt to replicate your work. But alas, they fail to take note of what you have (brilliantly) learned of SMNR. And they fail to detect it. Therefore, they are like Goofy looking for his coin - choosing to look where it is convenient, not where it is realistic.

Further, no test that you design could ever prove them wrong. They will simply say "we looked with the best methods and we saw no SMNR".

Using your example with you as the brilliant retrovirologist, the options in logic available to those looking to confirm or contradict your work are:-

A. They assume a priori that your original assertion is incorrect. Use there own methods:
A1. They find SMNR --> confirmation of your work
A2. They do not find SNMR --> suggestive that your finding is incorrect but not definitive. Further work required

B. They assume a priori that your original assertion is correct and follow your methods:
B1. They find SMNR --> suggestive that your finding is correct but not definitive. Further work required
B2. They do not find SNMR --> definitive refute of your proposition

Most XMRV studies are sitting at A2. They raise valid questions. But failing to prove a positive does not imply proof of the negative.

No competing interests declared.

RE: A critique of the methodoloy of Hohn et al 2010

PLoS_ONE_Group replied to gerwynm on 13 Jan 2011 at 21:50 GMT

We have deleted or edited some of the comments on this post because they did not adhere to the ONE commenting policies: http://www.plosone.org/st....

Please maintain a professional and civil tone in future comments and refrain from making any personally insulting statements. If we notice any further personal insults, we will be forced to remove the comments and moderate the discussion.

I’d also like to remind the participants that all users must unambiguously identify themselves in their profiles with first and last names and geographical locations (http://www.plosone.org/st...). Failure to do so may result in the suspension of user accounts.

No competing interests declared.