NAME: John Hackett, Jr.
POSITION: Volwiler research fellow and manager, emerging pathogens and virus discovery, Abbott Diagnostics
BACKGROUND: PhD, immunology, University of Texas Southwestern Medical Center; Postdoc, UTSMC and University of Chicago
Since its discovery in 2006, xenotropic murine leukemia virus, or XMRV, and related murine leukemia virus-like viruses have been implicated in various human diseases, most notably prostate cancer and chronic fatigue syndrome.
XMRV's link to prostate cancer remains an open question. However, after various independent groups — including researchers from the US Food and Drug Administration, the National Institutes of Health, and the Whittemore Peterson Institute — identified a possible link between the virus and CFS in 2009 and 2010 (PCR Insider, 9/2/2010); several other groups moved quickly to dispel that notion, claiming that the link was likely a result of contaminated commercial PCR reagents or laboratory environments (PCR Insider, 6/9/2011).
In December 2009, the US Department of Health and Human Services formed the Blood XMRV Scientific Research Working Group, or SWRG, in an attempt to get to the bottom of the matter. This group, led by the Blood Systems Research Institute in San Francisco, used 11 nucleic acid, five antibody, and three culture assays developed by nine research groups. The group set out to determine the ability of these assays to detect XMRV and MLV-like viruses and to examine potential contamination issues involving mouse DNA.
The group also aimed to determine whether a recommendation made by nonprofit blood banking association AABB in 2010 discouraging CFS patients from donating blood was necessary, given the hypothesis that the virus might not even be involved in the disease or transmissible through blood donation procedures.
The ultimate findings of the XMRV SWRG — that the virus or murine-like viruses could not be confirmed in the blood of CFS patients — were published in September in Science and seem to have set the record straight on the matter. In addition, just this week, one of the original papers drawing a link between XMRV and CFS — published in October 2009 in Science — was retracted by the journal in light of recent contrary evidence.
Scientists from Abbott, led by John Hackett, participated in the SWRG and developed a real-time PCR assay to screen samples in high throughput using the company's fully automated m2000 molecular testing system. In a paper published in October in the Journal of Virological Methods, Hackett and colleagues described the development and validation of their assay.
In addition, Hackett and scientists from several other organizations, including the American Red Cross and Gen-Probe, recently put the finishing touches on a study that determined that XMRV does not pose a risk to blood recipient safety. The results of that study are currently available online in an early-access version of the journal Transfusion.
PCR Insider caught up with Hackett this week to discuss Abbott's involvement with the XMRV SWRG; the development of the various tools and assays used in these recent studies; and the continuing saga of XMRV's potential role in human disease. Following is an edited transcript of that interview.
How did Abbott get involved with the Blood XMRV Scientific Research Working Group, and what was its motivation for developing this XMRV assay?
The story for us starts fairly early. The very first publication that came out from Urisman et al in 2006 [linking XMRV and prostate cancer] … piqued our interest because of the potential that if there was another retrovirus that was infecting humans, it could be of concern because of potential transmission. We got involved fairly early, because it takes some time to develop assays and tools to track the virus. It was a situation where there was no real evidence of causality, which is one of the key factors when searching for new human viruses. This is a situation [scientists] constantly deal with: We can identify a new virus, but the question is whether it causes disease. A great deal of time and effort goes into trying to establish that.
So we started a collaboration early on with [Robert Silverman and Eric Klein] at Cleveland Clinic, who were involved in the discovery of XMRV [along with scientists from the University of California, San Francisco]. Our initial role was to establish tools that would allow us to track the virus and see whether we could associate it with any human diseases. From our perspective, this was a research exercise, but we were thinking ahead because if it turned out that [XMRV] was involved with human disease, then obviously we would need diagnostic tests and monitoring tools.
When we started these studies, it wasn't clear to us if we had humans that were infected in a way that we could rely on; for example, if they had antibodies to the virus. So we initiated studies in Rhesus macaques where we infected them artificially, and had the opportunity to understand, first, whether or not they [mounted] an immune response. And if there was, what were the characteristics of the response? It also allowed us to have an opportunity to study where the virus goes. There are always questions of where and when to look for the virus.
These studies were fundamental to our ability to create the tools that we did, because in fact the animals did seroconvert, or create antibodies to the virus, and we were able to map out what viral proteins they responded to. And we had a source of bona fide positive control material from seroconverting animals, which allowed us to establish antibody assays and have a sense of how sensitive they were. In parallel, [we had begun work] related to molecular assays.
This is largely why, when the National Heart, Lung and Blood Institute formed the [Blood XMRV Scientific Research Working Group in December 2009], we were invited to be involved in it, because of the early work we had done in the field; and [because] the assays that we had developed … had been characterized very carefully and were on platforms that allowed us to screen for large numbers of samples very quickly. The molecular assays were real-time PCR assays, as opposed to what many in the field were using. All the early reports used nested PCR assays, which carry a risk of contamination from opening tubes after the first round and transferring materials.
Again, we were trying to develop [assays] on platforms that would allow us to get reliable testing results. In essence that was why [NHLBI was] interested in having us participate, along with a number of other laboratories that were developing their own tools to analyze the virus.
Most of the assays that had been used in previous studies of XMRV's role in human disease were not real-time PCR assays?
The original report [in Science in 2009] linking XMRV with chronic fatigue syndrome used nested PCR assays. Ultimately, some of the labs have moved on to the real-time format. A lot of times you're seeking the simplest way to establish an assay with very high sensitivity. But there are risks associated with that, as there are for all assays.
XMRV ended up being particularly challenging because of the fact that it is closely related to murine leukemia viruses, which are harbored within the genome of mice. This raised new opportunities for potential contamination that frankly, in the beginning, investigators weren't even thinking about. This even comes down to some of the reagents that were used for PCR. It is now recognized that one of the reagents used for some studies contained monoclonal antibodies from mice, which means there is the opportunity for trace amounts of mouse nucleic acid to be in that product. The laboratory can be doing everything right, but they unknowingly use a reagent that contains mouse nucleic acid, which could cause problems.
This is the reason we spent a lot of energy studying the components that went into assays, because we ultimately needed to be able to provide tools that work reliably and are as specific as possible.
And the JVM paper showed that this XMRV assay that you developed on the m2000 was both highly specific and highly sensitive?
Correct.
The JVM paper states that you compared the sensitivity and specificity to other tests included in the SRWG phase I study published in Science last month. What kinds of tests did you compare to?
There were a number of different laboratories — nine total — involved in the study, and [many of them] had their own tests. It was set up such that it was up to each laboratory to create their own tools [and] validate their assays as best they could. And this study was set up really well in terms of the blinding aspect; pedigreeing the negative and positive controls that were used; and being very cautious about reagents used, custody of the samples, and where things were done. It was carried out very effectively, which is why it was so important … in determining [that] there really was no clear evidence of XMRV infection.
The scientists from the Whittemore Peterson Institute that originally found a link between XMRV and CFS in the 2009 Science paper were also involved in this latest study from the SWRG, correct?
Exactly right. Also, Frank Ruscetti at National Cancer Institute, who was also involved in that original study, provided some of the culture methods and a flow cytometry assay to detect antibodies. And that was important. In addition, Harvey Alter's group [at the National Institutes of Health], which had previously found evidence … of polytropic murine leukemia virus in CFS patients, participated in the study. It was critical and good that those laboratories were also participating using their own technologies, which they'd used previously, but [this time] in a completely blinded manner, with replicates of samples.
For a long time there was an argument that differences in sensitivity between assays may account for conflicting results in detection of XMRV. We had an opportunity to directly compare sensitivity and specificity of the various assays. Results generated using the blinded panel clearly demonstrated that all of the assays were sensitive. Yet, XMRV/MLV was not reproducibly detected in the specimens tested.
Did Abbott use the m2000 PCR assay described in the JVM paper for the SWRG study published in Science in September?
We used both serologic assays on the Architect platform, to detect antibodies to XMRV p15E and gp70 envelope proteins, which we've published on; and real-time RT-PCR assays on the m2000 platform to detect XMRV pol and env sequences.
Is the presence of XMRV and potential contamination still a vexing problem in laboratory experiments in general? Is there an addressable market for the Abbott real-time PCR assay?
Our focus is actually turning back to the prostate cancer question. That was where XMRV was identified originally, and we're currently re-investigating this issue. We're actually in the midst of a couple of studies, which we'll be finishing up soon, looking at whether there is any association of XMRV with prostate cancer. Again, the use of these tools, which we believe are quite reliable, will ultimately get us to the answer. These are again collaborative studies with other groups, including the Cleveland Clinic and UCSF, with the perspective that combining technologies can be quite advantageous, to see if everyone comes up with the same answer.
In terms of where the field will go, from a CFS standpoint, I think the data is pretty conclusive. We also had a paper in Science earlier in the year [showing no evidence of XMRV infection in samples previously identified as XMRV-positive], and our impression even at that point was that while there could be a potential infectious agent associated with CFS, it would be better to move resources toward looking for something else besides XMRV.
I don't know if there's a commercial market. Essentially we're doing what we set out to do, which is investigate whether [XMRV] has a role in human disease; and if there's not, then there's not. At least we were proactive. And I think the Transfusion paper also is a very large study that, again, shows the benefit of developing these assays and platforms where we can screen large numbers of samples. We screened more than 17,000 blood donors and recipients for evidence of XMRV infection.
But we're attempting to be proactive, because if it did turn out that this was an important virus from a human perspective, obviously we would need to have ways to try to identify infected [individuals] and mitigate its transmission.