The XMRV Files: The Mystery Continues

1632-QuestionMark..jpgThe inability of anyone other than the WPI (and the NCI and Cleveland Clinic) to find XMRV is mysterious. In the face of three failed attempts to find any trace of XMRV in several hundred chronic fatigue syndrome patients questions are, of course, growing about the WPI’s find.

Yes, questions have been raised about the efficacy of each of the three studies – all of them may have had their flaws and depending who you talk to, some of them have major flaws – yet it’s alarming that no one has yet found even one CFS patient with an XMRV infection.

Support organizations respond

The inability of the two more XMRV studies- from a ‘friendly’ research group and an ‘unfriendly’ Dutch group – tto find any XMRV in a very large sample of patients was rough news for sure

The ME Association believes the third study considerably darkened XMRV’s potential for being a major factor in chronic fatigue syndrome stating “these three negative studies now place a very serious question mark over the proposition that XMRV is present in a significant proportion of the ME/CFS population and that the infection plays a significant role in most cases of sporadic ME/CFS.”

With regards the different procedures used by the different studies they stated “the various laboratory investigations for finding XMRV have been carried out in very reputable microbiological research centres and involved retrovirologists of international repute.”

They believe the problem lies not in different cohorts or different geographical spread of the virus but in the laboratories doing the testing and urged that the same samples be sent to and tested by the variety of different lands. They also stated that “the MEA Ramsay Research Fund is very willing to consider funding high quality research proposals”.

CFIDS Association’s response

Dr. Vernon stated that with three negative studies under our belts it’s was getting much harder to pin the problem on any one methodological error. Instead Dr. Vernon turned to the question of the patients in the original cohort. Were they from an infection cluster (or clusters) as was originally reported?

Did they have the immune abnormalities reported or, as was later reported, not? Dr. Vernon is suggesting that the reason these studies are turning up negative is that none of the patients in them were similar to the patients in the original study; i.e. she believes the original cohort contained a very unusual type of patient.

Since requests for more information on that original group have not been fulfilled Dr. Vernon looked at various bits of evidence from the CFSAC meeting and elsewhere and concluded that least five of the 32 patients she could get information on had had cancer.

Dr. Vernon’s worry is that if XMRV is only found in the blood of a very specific type of patient that unless researchers search for XMRV in this particular set of patients, the search for XMRV will soon be over. She and others have implied that research enthusiasm for continuing the hunt may be waning in the face of these negative results.

The fact that it now appears that XMRV is more readily found in the organs than in the blood suggests less ill patients could carry it without showing it in the blood. If that’s true finding it in them will take more digging but for the research community to commit to doing that they’ll have to find it in the blood first.


Dr. Mikovits has said that the patients weren’t particular disabled or particularly sick; they simply met either the Fukuda definition or the Canadian Consensus Criteria and there were no lymphoma patients in the study.

Annette Whittemore reported that the patient samples were taken randomly from samples sent to the WPI from 2006 to 2008 from physicians across the country. Dr. Peterson reported focus on, for the want of a better phrase, viral ridden patients has sparked concern that the patients in the initial study were quite different.

But the WPI has also been gathering samples for its repository from physicians across the country and this is apparently where the samples from the study came from. This, of course, would make it unlikely that they were any one type of patient.

The patients were not confined to Incline Village or Nevada or even the United States; instead they came from all over the US as well as the UK, Ireland, Germany and Australia. As for information on illness duration, illness severity and treatments if the WPI hasn’t provided it yet it doesn’t look like this information will be forthcoming.

Friendly or not?

The Groom group has been described as being ‘friendly’ to CFS because it contained several longtime ME/CFS researchers(Dr. Kerr, Dr. Gow) committed to a physiological interpretation of this disease. It’s probably safe to say that while they may be friendly to CFS they’re probably longer with the WPI.

The WPI reportedly freely offered their reagents and materials to researchers yet this first ‘friendly’ group to look for XMRV didn’t avail themselves of them. Nor did they apparently contact the WPI as it became clearer and clearer they were not going to find any XMRV infection in their CFS patients.

For their part they apparently fully expected to find the virus using the techniques they had. Still, since inertia does count for something in science another negative study – particularly coming from this group – couldn’t have helped XMRV’s chances of getting a full examination.

The WPI’s Guide to Finding XMRV – In their response to the Groom UK paper (the ‘friendly’ group) the WPI defined the pitfalls they believe researchers face in validating their work – thus giving them a specific roadmap to follow.

  • The WPI way or no way? The WPI took the UK study to task somewhat for not using their procedures etc. stating that there is only one way to look for XMRV in the blood that’s been validated and that’s their approach and they have a point. Outside of a unpublished Japanese study that looked for antibodies, the WPI was the first group to ever look for XMRV in the blood using PCR and they validated their results using the Cleveland Clinic and National Cancer Institute labs. Indeed we know that European investigators have had problems finding XMRV in prostate tissues; given that it might not be surprising that investigators might be having trouble finding it in the blood.
  • An unfortunate necessity – In the best of worlds a true replication study isn’t necessary and is, in fact, irrelevant. How one found the virus, after all, is not particularly important; pathogens can be uncovered by several techniques and researchers typically use different techniques to validate the presence of a pathogen. In fact a positive validation study using a different technique is considerably more valuable than a replication study because it definitively demonstrates the pathogen is there. It’s only when the validation studies are unable to validate a finding that the issue of a true replication study becomes necessary- as it appears it now has.
  • To be fair the Groom study researchers, some of whom have long track records in CFS research, didn’t have any reason to think their procedures wouldn’t work since apparently they do work for most viruses and the Dutch group used some of the same primers as the WPI. Neither, however, attempted to replicate the first study.
  • Different blood harvesting and storage procedures wiped out XMRV – The idea that different blood storage and harvesting procedures in the Groom study could’ve altered the results seems possible. Indeed one microbiologist at the Phoenix Rising Forums pointed out that the Groom used few of the guidelines deemed necessary in the US to find latent (low level) viruses. We know, though, that XMRV is robust enough for Dr. Peterson to be able to pull it out of a 20 year frozen sample. Still the possibility does remain that all three studies used blood storage techniques unsuitable for XMRV, while unlikely, is possible.
  • Different patients – The idea that the WPI had one set of patients and everyone else had a very different group has come up again and again. The very large size of the British study with patients from several different groups appears to make this scenario an unlikely one (and a decidedly unattractive one since then XMRV would apply only to very special patient groups.) Laymen’s conclusion – not ‘it’ either but it again it could lower prevalence rates. On our way to zero?
  • Very, very low levels of a very difficult-to-detect virus – While the other issues could reduce an investigators ability to find the virus with the exception of the blood storage issue it’s hard to believe they could result in being unable to find any XMRV in a large group of patients. The low viral level issue seems to be more significant, however. The WPI did two things the other groups didn’t do to find the virus.

They usually grew the virus in white blood cells first in order to increase its numbers

This involves hitting the white blood cells with a substance designed to enhance viral replication. Dr. Mikovits reported earlier that while XMRV appears to be able to readily infect immune cells it simply doesn’t have the tools to replicate readily in them – hence its viral loads are expected to be low. Not culturing white blood cells first could, then, reduce prevalence rates markedly – but not completely – as the WPI said they usually, but not always, had to use this technique.

WPI researchers had to search multiple times both in time and space to find the virus

Not only did they look at a sample multiple times they had to look at samples taken at different times from the same patient in order to find the virus. One wonders if the WPI shot themselves in the foot a bit by not announcing this earlier. The Science paper did report that the WPI ALWAYS cultured the cells first – a clear indication of very low viral loads.

In an email Dr. Mikovits stated that viral loads are so low that they would need DNA from a million infected cells to find the virus in unstimulated cells. The UK studies presumably looked at the same samples once or at most twice. The UK and Dutch researchers rebuttal to this is that they believe they used sensitive enough techniques to find the virus in very low copy numbers. After the fact the lead Dutch researcher stated they looked multiple times or the virus and ‘challenged themselves’ to find it.

All the way to zero?

Put the last two issues together plus perhaps a bit different cohort, perhaps reduced rates of XMRV infection in Europe, maybe some blood storage issues and perhaps you have a scenario where it’s ‘highly unlikely’, as the WPI stated, that these studies would find the virus.

1634-VIPDx..jpgThe other ‘study’

We also know that despite the fact that no studies have been able to find XMRV that VIP Dx Labs is finding it every day. Thus far about half of the people completing the Phoenix Rising Forums poll tested positive using either PCR or culture from VIP Dx Labs using the WPI’s techniques.

How is little VIP Dx labs able to find XMRV when these larger research labs are not? It seems that there can be only two answers to this question; their culturing technique is in fact critical to finding the virus or the lab has somehow become contaminated with the virus.

A layman’s conclusion

The fact that zero of several 100 people with CFS have tested positive for the virus is alarming given the reported sensitivity of their techniques. Even if there are problems with the cohorts or some methodological issues it’s difficult to account for zero results.

Dr. Mikovits believes the fact none of the groups cultured the virus first is critical given how low viral levels are. The fact that, even after culturing the virus, the WPI often had to search several times in different samples to find indicates how low the viral levels are. The other researchers believe they still should have been able to find it. Somebody is obviously wrong!

We’re at a real conundrum. Aside from an unpublished Japanese study that found XMRV antibodies in blood, the WPI is the first group, to my understanding, to ever find XMRV in the blood of humans. Their internal validation studies were done by their partners at the Cleveland Clinic and at the National Cancer Institute. These were all small ‘studies’ but they were done at highly reputable labs. They seemed to be doing everything right and yet here we are.

As we wait for more PCR studies we should note that the PCR test – which was probably the most important aspect of the Science study – wasn’t the only evidence the WPI, Cleveland Clinic and the NCI had of XMRV in CFS.

Other XMRV tests

The WPI also found the virus using did antibody tests and grew the virus in specialized cells. The antibody tests are intriguing because they indicated the patients bodies were responding to the virus; something that would not be happening simply if their samples had become contaminated.

The scientific community appears to be focusing on PCR because it provides more or less definitive evidence but the antibody tests are another pillar of the WPI’s argument. A negative PCR result has no effect on the efficacy of a positive antibody result – it just makes it harder to explain.

It’s worth noting as well that no study has come close to matching the comprehensiveness of the WPI’s original study; the WPI isolated the virus, cloned and sequenced it, compared it to similar viruses and to the human genome, demonstrated an immune response to it, showed it was infectious and snapped a picture of it and used two of the top laboratories in the US to do all that.

All that will mean nothing, of course, until labs start validating or replicating their results.

The questions

  • Testing procedures at the WPI – Dr. Mikovits revealed at the Santa Barbara conference that some patient’s samples by PCR were examined three or four times before the virus was found. Because the control samples were from a different source it was impossible to give them the same kind of exacting treatment. Because WPI researchers looked harder for XMRV in the patient samples than they did in the control samples this suggests that infection rate of XMRV in healthy controls could have been greatly understated. This, in turn, could mean the virus is not very specific to CFS and therefore not likely a causal factor.
  • The fact that the antibodies testing at the WPI found about the same frequency of XMRV in controls as did the PCR test suggested this was not a problem. Still THAT result was unexpected if researchers did indeed look three or four times harder for the virus in the patient’s blood than in the controls. Thus far none of the studies suggest substantially higher rates of infection are present in the healthy controls; if anything the opposite has been true. Still, this an area of concern that will have to be resolved by future studies.
  • Contamination – Few people are talking publicly about possibility of contamination but this is a concern for some CFS professionals. The concern appears to be not contamination with an endogenous retrovirus but contamination with XMRV itself; it proposes that XMRV somehow ‘got loose’ at the Cleveland Clinic or the WPI and started showing up in more and more samples. Indeed, the virus’s remarkable genetic homogeneity in the WPI’s samples could be the result of the samples coming from a single source; ie a contaminant.
  • On the other hand Dr. Mikovits reported that as they sequence more strains of the virus they’re finding more and more genetic variation. The genetic homogeneity could also suggest that the virus is a recent entry into the human population. Dr. Coffin has suggested it’s been in humans about 40 years. It’s also unclear why XMRV didn’t show up in the healthy control samples as well. Still contamination could explain the disparity in the results to date.
  • Validation study problems – Gerwyn, a microbiologist with CFS on the Phoenix Rising Forums, has taken issue with the techniques used in the validation studies. He has not been backed up by critiques by the ME Association and the CAA and not all agree with his assertions but he is knowledgeable. Could he be correct?


The WPI now believes the focus on the PCR test by the scientific community and testing labs is a mistake because of the very low viral loads. About a month ago VIP Dx Labs abruptly ditched the PCR test altogether and closed its website altogether and website and the WPI released a rather confusing press release alluding to a new test on the horizon.

Dr. Klimas, Dr. Bateman and Dr. Bell all warned about the dangers and the possible emotional distress that might come from using an unvalidated test and they were right.

What about those patients with negative results? Will they have to pay for the new, more definitive test? A Phoenix Rising Forum participant reported receiving an email stating VIP Dx will run the new culture test on their banked sample for free and Dr. Mikovits confirmed that the Lab is in the process of doing that with their banked samples.

Better test on the horizon

Meanwhile, the XMRV group headed by Dr. Singh in Utah developed an antibody test for the XMRV in prostate cancer patients. Most importantly it showed that the immune systems of patients who were not XMRV positive according to the PCR were nevertheless still producing cells (antibodies) designed to track down and kill the virus ; i.e. they were infected but the PCR can had failed to pick up the virus.

Dr. Mikovits showed a slide of 10 prostate cancer patients and three chronic fatigue syndrome patients all of whom of whom were negative for the virus using PCR but positive using the antibody test. Interestingly, Dr. Bannert of the German XMRV prostate cancer study that failed to find evidence of XMRV is using the new antibody test developed by Dr. Singh at the University of Utah to retest his samples.

Dr. Mikovits stated that the best test for XMRV will be an antibody test. Even then there will be problems because she also noted that some ME/CFS patients simply do not produce many antibodies. (It seems clear that there will be several tests; if the serology test is negative perhaps a culture test or a better PCR test will be in order).

Share this!