In Part II of the XMRV section of the conference we look at the Coffin/Mikovits presentations, Hanson’s update on her XMRV work and look at the recent XMRV research findings.
The pMLVs Strike Out – Hanson Disavows Former Results
“The sensitivity of the PCR assays used requires extreme caution in interpreting results
Dr. Maureen Hanson”
Background – Dr. Hanson’s was one of two labs (Lo/Alter) to find not XMRV but closely related murine leukemia viruses called pMLV’s, in CFS blood samples. Dr. Mikovits heralded the findings as proof that a family of murine leukemia viruses were present in chronic fatigue syndrome but the their similarity to endogenous retroviruses raised a red flag.
Neither Hanson’s lab (a plant molecular biologist) nor the Lo/Alter labs specialized in retroviruses and several retrovirologists expressed surprise that the WPI allied themselves with work they suspected would not turn in the long run.
At the IACFS/ME Conference, Dr. Hanson presented the latest results from her work.
First she sought and again failed to find a sequence from XMRV from the env gene in her patient samples from Dr. Bell. Then Hanson switched to looking for the gag sequence – a more conserved region of the virus. She appeared to have found them, but further analysis determined that they did not contain the classic 24 bp deletion found in XMRV – she had not found XMRV after all.
She then tried to culture the samples using the WPI’s methods and found the gag sequence for XMRV, but no other sequences – which was troubling given that she should have been able to find the other sequences as well. Even more troubling was the fact that she found as many gag sequences in the healthy controls as in the ME/CFS patients – suggesting that contamination had occurred. (The WPI and NCI Ruscetti labs will later have similar results in the BWG study. )
Culturing has been described as a doubled-sword; it may be more effective in picking up some viruses but it is more susceptible to contamination. The long culturing periods Dr. Mikovits stated were necessary to find XMRV – far longer than most culture tests – increased the risk of contamination still more. Dr. Hanson decided that culturing for 20/30 days left too much opportunity for contamination and abandoned this test.
More Rigor Suggests Hansen’s Original Results Were Due to Contamination – Worried that contamination had somehow gotten into her samples, Hanson then became more rigorous. She went from using the original method of PCR in the Science paper (nested PCR, which is also more susceptible to contamination), to single PCR, reduced her processing of samples, implemented the IAP test for contamination, changed her blood collection procedures and her positive GAG results disappeared altogether. She concluded her original results were due to contamination.
Antibody Test Results are Positive – Then she did antibody tests developed by the WPI – which do not have the contamination problems of PCR – and she found that antibody tests for XMRV were positive in half her CFS patients and in about 15% of healthy controls. She emphasized that this did not mean they had been exposed to XMRV (she never found XMRV). She suggested that people with ME/CFS could have a retrovirus other than XMRV and urged more work in this area.
The Mikovits-Coffin ‘Debate’
Dr. Mikovits Presentation
Mikovits noted that 6/29 samples sent to Silverman showed evidence of contamination with the VP62 plasmid. The WPI has a master set of samples from the original study. They went back and sent samples from them to an independent lab which found no evidence of mouse contamination (IAP) and no evidence of the VP62 plasmid. (This suggests, of course, that the plasmid was introduced at Silverman’s lab. Reportedly, Silverman vigorously rejects this.)
Mikovits then showed that a contaminated sample was able to infect other cells – something a plasmid should be unable to do. Another contaminated sample with a viral protein suggested that another form of XMRV was present.
She stated that she felt the WPI made a critical mistake in focusing on the samples they did and stated there is far more variability in XMRV than has been revealed. Finding more variability is critical to the WPI’s claim that they have found a virus that was not inadvertently produced in a lab.
(About two years into the XMRV controversy, the WPI finally did release more gene sequences to GenBank. The CFIDS Association of America reported an independent analysis indicated that they were similar to the prior sequences deposited. The WPI has stated that they will do more full length sequences of the virus. Dr. Coffin’s response to Dr. Mikovits’ claim that more variability is present was, “Show me the sequences.”)
Dr. Mikovits also referred to the macaque study in which the virus became undetectable by PCR and antibody tests. She also noted a factor called 5-azacytidine which stimulates XMRV in the PBMCs, which should make the virus easier to find. She asserted that the WPI’s antibody assay picks up more types of XMRV than the Abbot test and noted 4 patients who had tested positive to it but not to the Abbot test. (Unfortunately, the WPI’s antibody test picked up more evidence of XMRV in the healthy controls than in the CFS patients in the Blood Working Group study suggesting that it is cross-reacting with something that is not disease producing.)
The WPI is also using a next generation technology that they feel is more effective at picking up XMRV. Dr. Mikovits has generally stayed away from reporting on the treatment side in the past, but this time she reported that a patient on anti-retroviral drugs did better, relapsed and then improved with Tenofovir.
Controversy – Dr. Mikovits’ presentation at the conference became overshadowed by a report that she doctored a slide at the conference to fit her theory. At issue is a slide in which Dr. Mikovits asserted that using a product called 5-azacytidine had revealed proteins from gammaretroviruses other than XMRV were present in CFS patients.
It was later revealed, however, that the very same slide was used in the Oct 2009 Science paper to provide visual proof that XMRV was present. Using the same slide as evidence for two conflicting theories at a conference isn’t exactly kosher, but it’s hardly a deal breaker.
Doctored Photo in the Original Science Paper – A more serious concern surrounds Dr. Ruscetti’s statement that a) the 2011 Ottawa photo was the original one, suggesting that the photo in the 2009 Science paper was altered to fit the findings of the paper, and b) the chemical agent 5-azacytidine was used to flush out retroviruses in the original Science study in at least some ME/CFS patients but not in the healthy controls. That fact was not reported in the Science paper.
Specifically, the parts of the photo that purportedly showed no evidence of XMRV in healthy controls in the Science paper actually showed no evidence of XMRV in CFS patients. Doctoring a photo in a scientific journal article to fit the paper findings would not alter the paper findings but it would be a serious breach of etiquette and a major embarrassment. Dr. Ruscetti apologized for the mistake, stating that in a rush to get the data out mistakes were made.
5-azacytidine – Subtle differences in sample storage and processes can significantly affect how easy or difficult it is to detect viruses. Now the 5-azacytidine saga suggests some patient samples but not healthy control samples may have treated with an agent that allowed viral proteins to be more easily identified.
It remains unclear how many patient samples were treated with 5-azacytidine and Ruscetti and Mikovits both denied that their use of it materially affected their findings, stating the using the factor was not ‘germane to the results’.
However, a retrovirologist at the NCI, Vinay Pathak, disagreed, stating in an article in Science magazine that “If [5-azacytidine] was used in the original experiment, it’s an egregious error to leave it out of the Science paper. . . It makes a difference how I would interpret the results.”
A PCR specialist I talked with agreed with Dr. Ruscetti and Dr. Mikovits’ statement that using the factor should not have affected the results of the paper. Dr. Mikovits did not argue that using 5-azacytidine made it easier to find XMRV; she argued that using the factor made it easier to find other gammaretroviruses.
Dr. Coffin’s Presentation
Dr. Coffin began by noting how excited retrovirologists were at the prospect of an infectious human retrovirus. It wasn’t just that XMRV was possibly only the third infectious retrovirus found in humans; the fact that it was associated with an already well-studied group of viruses, the gamma-retroviruses, made it doubly exciting. This was a virus the research community was clearly ready, willing and able to study.
(With 183 papers published on XMRV since the Oct 2009, Science paper, history has borne Dr. Coffin’s assertion out. Far from not wanting to find the virus (as has been suggested) researchers in many different fields went on something of a tear as they tried, fruitlessly, to find XMRV in every disorder from CFS to HIV to autism to auto-immune disorders to transplant patients to fibromyalgia. Since the Oct 2009 Science paper, the rather dismal toll for and against XMRV is one study for the virus in CFS (the original paper) and a few for it in prostate cancer and over 40 studies unable to find it at all.
The idea of contamination did not begin to arise until a variety of studies both inside and outside of CFS failed to find XMRV. In the last year or so six papers have reported different ways contamination problems have caused researchers to think they had found the virus and then, upon further checking, concluded they had not.
Coffin then explained why he believed XMRV was a contaminant rather than a virus that is infecting humans. First he noted that XMRV had to come from somewhere, and that somewhere must have been mice, yet it had not turned up in extensive surveys of wild mice populations around the world. (A similar search for the origin of HIV, of course, found its progenitor in primate populations in Africa.) That left laboratory mice – and XMRV’s progenitors did finally turn up in some strains of laboratory mice. This suggested, of course, that if XMRV leapt to humans at all, it did so from a lab.
XMRV’s remarkably homogeneous genetic presentation, however, worried many researchers. Viruses that infect humans almost invariably become genetically altered as they as they fight off and adapt to attacks from the human immune system. XMRV, however, looked the same whether it was found in samples from people living in New York or Nevada. A paper by Dusty Miller, which indicated that a prostate cancer cell line called 22RV1 was spewing out very large amounts of a virus identical to XMRV, notched the alarm level one step higher.
Genetic analysis indicated that the XMRV found in patient samples probably came from the 22RV1 prostate cancer cell line, and was not infecting humans but contaminating lab materials and samples. The final blow came when a study found that XMRV had passed from mouse tissues used to grow the prostate cancer cell line into the cell line itself at some point in the early to mid-1990s. At that point, it was clear that XMRV was a lab creation which had escaped: into lab materials but not into humans.
During the panel discussion, Dr. Mikovits peppered Dr. Coffin with a series of questions.
Dr. Mikovits asked Dr. Coffin how he explained the different sequences found, which suggested that more than one virus was present. Coffin replied that he felt it was a big mistake to say that there was more than one kind of XMRV and, in turn, asked for evidence (i.e. gene sequences deposited in GenBank) that different kinds of these viruses (human gamma retroviruses) existed.
What about the infectious nature of XMRV? The WPI, after all, had shown that XMRV could infect human cells in the lab. Dr. Coffin replied that it was easy to make endogenous proviruses that make beautiful virus particles which are infectious. (VP62 is a laboratory engineered clone of XMRV that produces abundant levels of infectious virus.)
What about the electron microscope pictures of XMRV emerging from human cells? Didn’t that indicate the virus was infecting humans? Coffin replied that it was difficult to pinpoint what kind of virus was present in electron microscopy photos. He suggested that instead of a gamma retrovirus, the virus in the photos could have been an alpha retrovirus. In any case, he did not believe these photos ever constituted hard evidence.
Coffin then compared the history of two retroviruses: HIV and XMRV.
The first HIV paper was, as he put it, “terrible”, but it was rapidly reproduced and labs were quickly able, even 25 years ago, to produce their own PCR and antibody assays that found the virus. XMRV, on the other hand, started out well but no labs, even after two years, have been able to reproduce the WPI’s original findings.
Where are We Now?
Much of the WPI’s case for XMRV collapsed with the findings that their tests for it are inaccurate, and with Dr. Silverman showing that contamination was present in at least some of the original samples. The Silverman findings led to major portions of the original Science paper being retracted, and Dr. Simmons strongly hinted that future papers will indicate that the XMRV found in the prostate cancer studies was a contaminant as well. With the retrovirology community now apparently viewing XMRV as a closed issue, it appears that the WPI, Dr. Mikovits and Dr. Ruscetti are on their own.
Dr. Mikovits’ sudden termination and her legal problems involving the theft of documents from the WPI has left the XMRV/HGRV finding in an even more tenuous situation. It’s hard to imagine given the NCI’s public stance several months ago that XMRV is a contaminant, that Dr. Ruscetti will be getting more funding from his employers there. The WPI does have some grants that may provide some money to work from, but at the present time its hard to imagine Dr. Mikovits finding a funding source to continue her work.
Last Chance for XMRV: Lipkin Study – the biggest question, of course, involves the large Lipkin study and how it will proceed now that the major investigator at the WPI is gone.
It’s not clear how critical Dr. Mikovits’ presence is to the progress of the Lipkin study. In one interview, Dr. Mikovits stated that Pfost, a graduate student, was doing the PCR testing for the WPI and he is still apparently employed there. Dr. Lombardi, of course, was the lead author of the Science XMRV paper, which usually suggests that he did much of the lab work. Dr. Mikovits, however, specialized in the culture testing, and she is a far more experienced researcher than Dr. Lombardi and has the Ruscetti connection on her side. On the face of it, the WPI would appear to be ready to move forward on some parts of the Lipkin study but not all.
The WPI’s ability to complete the study is another question. The WPI was late and had problems with contamination on both stages of the BWG study, and was unable to deliver a relatively small set of samples for the second. The theft of research notebooks dating back five years is still unresolved. How they will be able to complete a much larger study with this issues facing them is unclear.
Dr. Mikovits has also questioned the validity of tests done by VIP Dx/UNERV labs, which Dr. Lombardi created, as well as Dr. Lombardi’s ability to carry out some tasks. It seems unlikely, given her mistrust of other researchers and colleagues in the past, that she would countenance any negative results from a Lipkin study she was not involved in.
That would leave a major figure outside the box, so to speak, leaving the ME/CFS community back in the situation when Dr. DeFreitas declared that her retrovirus was real, despite two double-blinded studies that had the same result as the BWG study. The DeFreitas retrovirus finding was never nearly as well studied as the XMRV finding came to be, but multiple labs did fail to validate her results (and her earlier retrovirus findings in multiple sclerosis were eventually debunked).
(Almost 20 years later, pathogen assays by the WPI failed to find evidence of an HTLV-like virus as well). Nevertheless, a good portion of the ME/CFS community believed that the DeFreitas retrovirus was probably present in CFS, and that belief engendered several decades of anger toward researchers, particularly at the CDC.
Clearly, for the peace of mind of the CFS community, it would be best if Dr. Mikovits was either included in the Lipkin study or was given the opportunity to bring her work to a conclusion. After the accusations that she stole material from the WPI, however, that may not be possible.
Genetic Sequencing – Finding more genetic variability in XMRV is critical at this point. Genetic sequencing should not require Dr. Mikovits’ presence as the WPI will have an outside lab do that. If the WPI has the money to do that, that project should move forward.
More Blanks for XMRV
While the drama at the conference and afterward continued, the research world marched on with a slew of negative XMRV papers. Joliceur worked with Dr. Mikovits (at least for a time) in the design of his study but he found no evidence of XMRV in Canadian CFS patients and neither did another Canadian study. DERSE has been touted by Dr. Ruscetti and Dr. Mikovits but a study employing DERSE and other techniques found no evidence of XMRV in CFS patients either. A US study found no evidence of the pathogen in people with CFS, rheumatoid arthritis, Bechet’s disease, and systemic lupus erythematosus patients. Two more CFS studies, one from the US and one from Sweden, drew blanks.
An Abbot Labs paper examining over 2,000 patients including HIV and HTLV patients found no evidence of XMRV either. (A 5% positive rate in HTLV patients was due to cross-reactivity; i.e. HTLV carried gene sequences that are very similar to those found in XMRV.) A Spanish study found no evidence of XMRV in HIV patients and an Oct paper found no XMRV in prostate cancer samples using culture techniques. (The last six prostrate cancer studies dating back a year have not found XMRV.) One study indicated that false positive results could result from reagent contamination. Finally, a Chinese study found no evidence of XMRV in blood donors.