alex3619
Senior Member
- Messages
- 13,810
- Location
- Logan, Queensland, Australia
One Phoenix Rising member has suggested I update my comments on XMRV since the following post:
http://niceguidelines.blogspot.com/2010/12/my-take-on-cfs-xmrv-as-of-today.html
I still agree with these points. I would like to add the following comments:
1. This commentary includes in the heading "as of today" for a reason. The science can change very rapidly. It only takes one really good new scientific paper to do this.
2. People have commented that the epidemic ME outbreaks do not match our understanding of XMRV lifecycle. On the contrary they are a good match using the two hit hypothesis which has been repeatedly discussed on Phoenic Rising. Indeed I have discussed the two hit hypothesis on another forum since before we even knew about XMRV.
3. All of the contamination papers do one thing: they raise a possibility of contamination. None of them prove contamination in all the studies that could find XMRV. Indeed, the fact that some studies showed mouse DNA contamination demonstrates that careful screening can detect such problems. Contamination with live XMRV is another issue, which I have also discussed before. See additional commentary below.
4. None of the contamination papers explain why the control groups in any of the studies are not massively contaminated. Again, see additional commentary below.
Contamination Findings
All the contamination studies do is show that contamination could be a problem. This is already known to researchers. The mouse contamination issue has been so well dealt with in various places that I am not going to comment further. The idea that samples could be contaminated by ... XMRV, deserves further comment.
The two CROI papers by Coffin et. al. that discuss ancestral viruses and recombination events show that XMRV might have been from recombination of viruses from nude mice used to develop the 22Rv1 cell line:
http://www.retroconference.org/2011/Abstracts/42569.htm
http://www.retroconference.org/2011/Abstracts/42508.htm
XMRV was previously shown to match the sequence of a combination of two endogenous MLVs, but showing that these viral precursors exist in some lab mice that are used for xenografts is new.
The key point from the CROI conference is additional confirmation that XMRV is a real infective virus, not that it might be a lab artifact. The claim that XMRV might have arisen in these cell lines due to recombination is possible. It is also possible that XMRV may be present due to contamination from people (or other sources) who carry the virus.
The argument that XMRV could have arisen only once in the 22Rv1 cell line is probabilistic. It might arise once, but it is highly improbable it could have arisen several times. For the probability to be a viable tool, this has to be purely random. I very much doubt that is a purely random event, given that it is to MLV's evolutionary advantage that such crossovers occur.
This paper has also not shown that XMRV did not arise from recombination in the early 1900s (or any other time) either - the arguments against them are essentially the same as the arguments against the Lombardi et. al. findings, except Lombardi et. al. have way more evidence and XMRV has been found in many labs in controls and patients.
Controls Groups
The purpose of control groups is either to perform a check on the experiment to highlight problems, or creating a benchmark to compare the results to. If XMRV were a contaminant you would expect that it would be present in the same percentage in patients and controls. While it might be argued that the patients are more likely to be contaminated due to extra handling, could this be true for lab after lab, sample after sample?
The overwhelming probability is that the low percentage in the control groups, plus much higher percentage in patients, completely rules out all contamination hypotheses. Unless the contamination hypotheses can explain this, they are just hypotheses.
What this study does have is a possible origin for XMRV, a possible mechanism for its creation, and the presence of XMRV. It is enough to make us stop and think, it is not proof of contamination.
What about the zero zero studies - studies in which zero or close to zero XMRV is found in both patients or controls? These tell us several things too.
First of all, if contamination is such a problem, how come these studies are not routinely contaminated? The obvious argument is of course that only those labs that have an XMRV contamination problem will produce positive XMRV findings - but this is both not proven and many of the labs never used the 22Rv1 cell line.
If this is a real virus, which the evidence strongly supports, and it is present in the population, which is likely, then why cannot these studies discover a background rate? There seem to be only three likely reasons. First, the background rate might be very low and with limited geographical distribution, especially if the cell line origin hypothesis is correct. Second, while this cell line might be pumping out virus particles they might have been contained. Third, these zero zero studies might not be finding XMRV because assumptions made during design and testing may not be valid in a human population infected with XMRV. Which answer is correct is still debatable.
Other MLVs including polytropic MLVs may also be an issue. None of the contamination papers have really examined these. While antibodies can cross react with other antigens, the presence of viral proteins indicates a replicating virus.
In my view the balance of evidence is still that XMRV is a real virus, and that it is present in the general population at a low rate (maybe up to 10%), a minority of prostate cancer patients and many ME/CFS patients. Whether it is primarily causal for CFS is still to be determined, there is a distinct possibility that ME/CFS is due to an immune response to fight the retroviruses.
Thank you for reading this far, even if you disagree with me, Alex Young
Recommended Reading:
http://www.mecfs-vic.org.au/sites/www.mecfs-vic.org.au/files/Article-2010Mikovits-Detection.pdf
While you are reading you might like to read my post on writing scientific articles, which is hidden away in the projects section (okay, okay, a shameless plug):
http://forums.aboutmecfs.org/showth...ely-we-can-have-medical-and-scientific-impact.
http://niceguidelines.blogspot.com/2010/12/my-take-on-cfs-xmrv-as-of-today.html
I still agree with these points. I would like to add the following comments:
1. This commentary includes in the heading "as of today" for a reason. The science can change very rapidly. It only takes one really good new scientific paper to do this.
2. People have commented that the epidemic ME outbreaks do not match our understanding of XMRV lifecycle. On the contrary they are a good match using the two hit hypothesis which has been repeatedly discussed on Phoenic Rising. Indeed I have discussed the two hit hypothesis on another forum since before we even knew about XMRV.
3. All of the contamination papers do one thing: they raise a possibility of contamination. None of them prove contamination in all the studies that could find XMRV. Indeed, the fact that some studies showed mouse DNA contamination demonstrates that careful screening can detect such problems. Contamination with live XMRV is another issue, which I have also discussed before. See additional commentary below.
4. None of the contamination papers explain why the control groups in any of the studies are not massively contaminated. Again, see additional commentary below.
Contamination Findings
All the contamination studies do is show that contamination could be a problem. This is already known to researchers. The mouse contamination issue has been so well dealt with in various places that I am not going to comment further. The idea that samples could be contaminated by ... XMRV, deserves further comment.
The two CROI papers by Coffin et. al. that discuss ancestral viruses and recombination events show that XMRV might have been from recombination of viruses from nude mice used to develop the 22Rv1 cell line:
http://www.retroconference.org/2011/Abstracts/42569.htm
http://www.retroconference.org/2011/Abstracts/42508.htm
XMRV was previously shown to match the sequence of a combination of two endogenous MLVs, but showing that these viral precursors exist in some lab mice that are used for xenografts is new.
The key point from the CROI conference is additional confirmation that XMRV is a real infective virus, not that it might be a lab artifact. The claim that XMRV might have arisen in these cell lines due to recombination is possible. It is also possible that XMRV may be present due to contamination from people (or other sources) who carry the virus.
The argument that XMRV could have arisen only once in the 22Rv1 cell line is probabilistic. It might arise once, but it is highly improbable it could have arisen several times. For the probability to be a viable tool, this has to be purely random. I very much doubt that is a purely random event, given that it is to MLV's evolutionary advantage that such crossovers occur.
This paper has also not shown that XMRV did not arise from recombination in the early 1900s (or any other time) either - the arguments against them are essentially the same as the arguments against the Lombardi et. al. findings, except Lombardi et. al. have way more evidence and XMRV has been found in many labs in controls and patients.
Controls Groups
The purpose of control groups is either to perform a check on the experiment to highlight problems, or creating a benchmark to compare the results to. If XMRV were a contaminant you would expect that it would be present in the same percentage in patients and controls. While it might be argued that the patients are more likely to be contaminated due to extra handling, could this be true for lab after lab, sample after sample?
The overwhelming probability is that the low percentage in the control groups, plus much higher percentage in patients, completely rules out all contamination hypotheses. Unless the contamination hypotheses can explain this, they are just hypotheses.
What this study does have is a possible origin for XMRV, a possible mechanism for its creation, and the presence of XMRV. It is enough to make us stop and think, it is not proof of contamination.
What about the zero zero studies - studies in which zero or close to zero XMRV is found in both patients or controls? These tell us several things too.
First of all, if contamination is such a problem, how come these studies are not routinely contaminated? The obvious argument is of course that only those labs that have an XMRV contamination problem will produce positive XMRV findings - but this is both not proven and many of the labs never used the 22Rv1 cell line.
If this is a real virus, which the evidence strongly supports, and it is present in the population, which is likely, then why cannot these studies discover a background rate? There seem to be only three likely reasons. First, the background rate might be very low and with limited geographical distribution, especially if the cell line origin hypothesis is correct. Second, while this cell line might be pumping out virus particles they might have been contained. Third, these zero zero studies might not be finding XMRV because assumptions made during design and testing may not be valid in a human population infected with XMRV. Which answer is correct is still debatable.
Other MLVs including polytropic MLVs may also be an issue. None of the contamination papers have really examined these. While antibodies can cross react with other antigens, the presence of viral proteins indicates a replicating virus.
In my view the balance of evidence is still that XMRV is a real virus, and that it is present in the general population at a low rate (maybe up to 10%), a minority of prostate cancer patients and many ME/CFS patients. Whether it is primarily causal for CFS is still to be determined, there is a distinct possibility that ME/CFS is due to an immune response to fight the retroviruses.
Thank you for reading this far, even if you disagree with me, Alex Young
Recommended Reading:
http://www.mecfs-vic.org.au/sites/www.mecfs-vic.org.au/files/Article-2010Mikovits-Detection.pdf
While you are reading you might like to read my post on writing scientific articles, which is hidden away in the projects section (okay, okay, a shameless plug):
http://forums.aboutmecfs.org/showth...ely-we-can-have-medical-and-scientific-impact.