Update at July 2013

The report of a potential retroviral link to ME/CFS, which is estimated to affect 0.2 to 0.4% of the population in developed countries, including between 400,000 and 900,000 people in the USA according to population-based data, certainly caught the world’s attention on 9th October 2009 — no bad thing for an under-researched and often-overlooked illness! The scientific report, entitled “Detection of infectious retrovirus, XMRV, in the blood cells of CFS patients”, appeared online in Science, one of the most prestigious scientific journals in the world, and described the findings of a consortium of researchers from the Whittemore Peterson Institute (WPI, located at the University of Nevada, Reno), the National Cancer Institute (part of the National Institutes of Health) and the Cleveland Clinic, Ohio.

The initial findings

The headline finding of the research paper was that DNA from a human gammaretrovirus, xenotropic murine leukemia virus-related virus (XMRV), could be detected in the peripheral blood mononuclear cells of 68 out of 101 ME/CFS patients (67%) compared with only 8 out of 218 healthy controls (3.7%). The extent of this difference in proportions is unusual, as it is the norm for scientific researchers to find relatively small yet significant differences between patients and closely matched control groups; in the modern world, novel associations of such magnitude are rarely found between long-standing chronic illnesses and infectious agents. In addition to the headline finding, the researchers claimed that XMRV proteins were being expressed in blood cells from ME/CFS patients at very high levels compared with controls, and through cell culture experiments they showed that patient-derived XMRV was infectious and transmissible. So, as well as being the first to show infection with this novel virus in ME/CFS patients, the researchers appeared to have been the first to be able to isolate XMRV particles from the blood, and to show direct transmission of this virus between blood cells — dramatic observations indeed.

What caught the attention of the scientific world is that these observations seem to fit neatly, at least at a first glance, with what is already known about ME/CFS as a chronic illness. For example, viruses related to XMRV have been reported to be involved in damage to blood vessels and nerves, and natural killer cells (historically low in ME/CFS) are said to be susceptible to infection by XMRV. Also, the fact that retroviruses like XMRV are known to be able to activate some other (latent) viruses might explain why ME/CFS has been associated with a range of different viral triggers, such as herpesviruses like Epstein-Barr, over the years. Again, as Dr Judy Mikovits and colleagues pointed out in their paper, some of the most commonly reported features of ME/CFS include neurological symptoms and immune dysfunction with inflammatory cytokine and chemokine upregulation, and some of these observations could be accounted for by infectious XMRV in lymphocytes. The fact that such pieces seemed to fit so well together was suggestive only, however, and a virologist at Tufts University was surely wise to say in New Scientist that while it’s not impossible that infection with this agent might cause a disease with neurological and immunological consequences, there was a long way to go before we knew for sure.

The background

The scientific journey towards this discovery is an extremely interesting one, and includes several strands: prostate cancer, the RNAse L immune pathway, the discovery of the novel virus XMRV, and ME/CFSXMRV is a human retrovirus similar to HIV, HTLV-1 and a group of endogenous murine leukaemia viruses found in the genomes of wild mice (see the informative presentation on retroviruses by Dr Jones of SAIC-Frederick/NCI-Frederick), and was first identified only in 2006 by Prof. Robert H. Silverman of the Cleveland Clinic, a co-author on the 2009 ME/CFS study. Prof. Silverman initially showed the presence of XMRV in prostate cancer tissue samples (PLoS Pathog, 2006), and subsequent work has confirmed XMRV protein expression in 23% of 334 prostate cancer biopsies (Proc Natl Acad Sci USA, 2009). Importantly, the men with prostate cancer initially studied by Prof. Silverman all had a specific genetic defect in their antiviral defences, the RNase L antiviral pathway which Prof. Silverman had been studying for 30 years, a lifetime’s work of scientific progression described in his fascinating essay, “Journey through the 2-5A/RNase L System”.

RNase L is the terminal enzyme in the 2,5A synthetase/RNase L antiviral pathway, and plays an essential role in the elimination of viral mRNAs. The enzyme has been the focus of research interest in ME/CFS patients for nearly 20 years, and deregulation of this pathway in subsets ofME/CFS patients has been reported extensively in the scientific literature (reviewed by Nijs and Fremont, 2008). In ME/CFS, a wide spectrum of “cleavage” of RNase L can be observed (a phenomenon also seen in multiple sclerosis patients), and such altered RNase L activity profoundly affects cellular physiology, including apoptosis. Overall, an upregulated RNase L pathway in ME/CFS is consistent with an activated immune state and a role for persistent viral infection in the pathogenesis of the disorder — and it is because of these and other findings that many researchers have come to view ME/CFS as primarily a disorder of the innate immune system (see Klimas and Kineru, 2008). It was due to the insight of Dr Judy Mikovits and her team at WPI that the potential connection between RNase L dysfunction in XMRV-infected prostrate cancer and in ME/CFS was recognised, and an exploration undertaken to test for the presence of the virus in the banked blood samples in the WPI tissue repository.

Subsequent investigations

In the immediate aftermath of their publication, the researchers said that they would continue to refine their test for XMRV. Their plan was to continue their in-depth studies of XMRV to clarify its effects on the human immune system, and to validate clinically a blood test for the detection of XMRV in ME/CFS and other human diseases. And they were to begin the work of determining if any currently approved drugs, such as AZT, might be useful for suppressing XMRV.

At the same time, other independent laboratories across the world began the process of confirming the findings in their own local populations of ME/CFS patients. Since the WPI researchers used samples selected from several regions in the US where “outbreaks of CFS” had been documented (using patients diagnosed on CDC-1994 and 2003 Canadian Clinical criteria), blood samples from patients in other countries (possibly diagnosed with less stringent criteria) might throw up very different results.

For this reason, ME Research UK (with the Irish ME Trust) quickly actioned funding to Swedish researchers to test for the presence of XMRV in Swedish ME/CFS patients (see outline). The results of this investigation were published online in October 2011 (Elfaitouri et al, Plos One, October 2011), followed by an expert review on the subject by Professor Blomberg and colleagues in a special issue of Advances in Virology 2011.

Since the initial report in Science, more than fifty distinct studies on XMRV in ME/CFS have been published by other researchers in the UK, USA, China, Germany, the Netherlands, Japan, Norway, Sweden and Canada, all keen to test their own populations of patients. The vast majority of these studies have not been able to find significant levels of the retrovirus in their patient groups, let alone see a dramatic difference in infection between ME/CFS patients and healthy controls. However, one investigation (Lo et al, 2010), instead of finding XMRV itself, detected a more diverse group of closely related murine leukaemia viruses (MLV) in 86.5% of patients compared with 6.8% of controls. Anecdotally, there were suggestions that XMRV positivity can be found in other populations, but (at July 2013) these data have been presented in poster form only (International Workshop on XMRV at Bethesda, Maryland, September 2010) and have yet to be published.

Of the “negative” studies, some were more noteworthy than others. The first (Knox et al) was published in the journal Science in July 2011, and reported finding no evidence of XMRV or MLVs in patients with ME/CFS. It was published alongside an “Editorial Expressing of Concern” that the validity of the original study (Lombardi 2009), which had also been published in Science, was now “seriously in question”. The second investigation was a $500,000 study, published in Science (Simmons et al 2011), from the Blood XMRV Scientific Research Working Group (BWG) in which none of nine labs could reproducibly detect XMRV or its relatives in the samples. And the same issue of the journal Science which carried the online Simmons et al study contained a “partial retraction” of the original paper, after two coauthors, Silverman and Das Gupta, reported finding contamination in their original samples.

As the list of negative published studies grew longer, attention turned to the possible reasons for the inability of independent researchers to confirm the original positive findings, and some of these were first discussed by Dr Robert Silverman, the discoverer of the XMRV virus, in an excellent review (Nature Reviews Urology, July 2010). Most attention, however, centred around the issue of contamination.

In December 2010, four studies appeared in the journal Retrovirology. Collectively, these suggested that XMRV might have originated from the chance recombination of mouse viruses during laboratory experiments, with positive findings reflecting cell-line contamination rather than true infection in humans (see the Wall Street Journal). One of these studies by Hue et al from University College London had compared XMRV gene sequences from the tumour cell line 22Rv1 with sequences found in XMRV-positive patients; virus genetic diversity was found to be greater between cell lines than between patients, suggesting contamination of patient samples by cell line virus in the lab.

Again, in July 2011, a report from Paprotka et al at the National Cancer Institute in Maryland had examined the origins of a previously used human prostate cancer cell line, 22Rv1. The early versions of the cell line, which were still in storage, harboured no XMRV, but DNA matching one half of the virus was found in two different strains of mice that had been studied subsequently. The researchers concluded that XMRV was generated from a unique recombination of pieces of two mouse viruses “that took place around 1993–1996 in a nude mouse”.

The controversy laid to rest

After three tumultuous years, the final act in the drama came on 18th September 2012 with the publication of the results of a multicentre study coordinated by Columbia University’s Prof. Ian Lipkin. This “multicentre blinded analysis” used peripheral blood samples from 147 ME/CFS patients and 146 matched healthy controls recruited using rigorous diagnostic criteria at six sites of excellence across the United States. Each of the samples was divided into aliquots and the blinded sub-samples sent to three different research groups, including those which had described the original association. All the labs used procedures, including polymerase chain reaction (PCR), which had previously been optimised for molecular or serological detection of the viruses.

The Table below encapsulates the stark message – none of the 293 patients or controls tested positive for the viruses using PCR, and the paper concluded that there was no evidence of infection with either virus. At the press conference explaining the results, Dr Judy Mikovits who had published the initial report in Science in 2009 was quoted as saying, “XMRV is simply not there… We are 100% positive in our results… It is time to move forward”. (Read more at ProHealth.)

While this definitive outcome was a blow to us all – since the XMRV hypothesis of ME/CFS was appealing and could have transformed the outlook for patients everywhere – Prof. Lipkin was adamant that research on the causes of ME/CFS must continue; “We’ve tested the XMRV/pMLV hypothesis and found it wanting,” he said, “but we are not abandoning the patients. We are not abandoning the science. The controversy brought a new focus that will drive efforts to understand CFS/ME and lead to improvements in diagnosis, prevention and treatment.

Table. Results of the Lipkin study
Laboratory Sample ME/CFS cases positive for XMRV Controls positive for XMRV
CDC, Atlanta Plasma 0/147 0/146
FDA, Bethesda Plasma 0/121 0/110
FDA, Bethesda Peripheral blood cells 0/121 0/111
Mikovits, Ruscetti and Hanson labs Peripheral blood cells 0/117 0/126

Further information