Letter to Harold Varmus, Director of the NCI, re the NCI's YouTube video about XMRV which omits ten key points

Chris Douglas, Patient Advocate, With many thanks to Danny Ze-dog for his significant input to this letter.

Dear Dr Varmus

Expression of concern about NCI YouTube video

I write with concern about the NCI video ‘Investigating XMRV’ which recently has been uploaded to ‘YouTube’ (see url below). www.youtube.com/watch?v=XwG0XlcMgx8&feature=youtube_gdata_player

The video reports on a paper by Paprotka et al (‘Recombinant Origin of the Retrovirus XMRV’) which claims to have found precursors to XMRV in the mice used to create cell line 22Rv1. It concludes that XMRV originated from the passages of xenografts used to make 22Rv1 and, because such a recombination event is unlikely to happen twice, all XMRV found anywhere must be contamination stemming from this one lab accident.

This conclusion is used as the basis for the video’s dismissal of previous research which shows that XMRV, and other MLV-related viruses, have infected humans and are associated with neuroimmune disease (such as chronic fatigue syndrome, or CFS) and prostate cancer.

There are ten key points, however, which the video either omits or fails to explain and my concern is that the absence of these facts render it misleading at best and inaccurate at worst. 1. The video does not make clear that Paprotka et al’s proposed six-step recombination event is only hypothetical, and that they have provided no evidence that it could actually occur. 2. Paprotka et al based the assumption of a ‘one-off’ recombination event on their own (unvalidated) statistical extrapolation and do not provide conclusive evidence that such an event (if it were possible) could occur only once. 3. Paprotka et al do not provide conclusive evidence to support the hypothesis that the resulting recombinant retrovirus has remained within the lab and not infected humans. 4. Paprotka et al made several assumptions which biased their experimental design and interpretations of data. For example, they detected XMRV env in the early xenograft passages but assumed that, since they did not detect gag, this represented detection of pre-XMRV-1, not XMRV itself. However, Lombardi et al (2009) found gag detection extremely inefficient until appropriately optimizing and modifying their assays. For example, they found that single round DNA PCR was largely ineffective and had to use nested RT-PCR to get high detection rates. Further, since the publication of Lombardi et al, a greater diversity of viruses that are variously being called XMRV or MLV-related viruses (MRVs) have been reported to have been found in CFS and prostate cancer patients. By defining XMRV only by the narrow genetic diversity originally found by Urisman et al and Lombardi et al, Paprotka et al may be overstating their conclusions (whether they are correct or not). It is too early in the understanding of human gammaretroviral findings to define terms so narrowly and then to apply the conclusions so broadly. 5. Paprotka et al failed to consider alternative explanations for their findings. For example, they did not consider that their PCR assay (as only one was used for the relevant experiment) might have been incapable of detecting XMRV present in the original grafts at levels below its limit of detection, and that the administration of testosterone during the xenograft process boosted viral copy numbers (via XMRV’s androgen response element) in the late passages to a level they could actually detect it. Also, they did not address the possibility that pre-XMRV-1 and -2 are recombinants themselves and arose as a result of prior infection with XMRV of the mouse strains used for passaging, or of their progenitor strains. 6. The video fails to state that cell line 22Rv1 has never been used in labs which have detected XMRV and/or MRVs in humans, e.g. Lombardi and Mikovits (University of Nevada), Ruscetti (NCI), Lo (FDA) and Alter (NIH). As such, there is no trail to link 22Rv1 ‘contamination’ (or any other cell line alleged to be ‘contaminated’ with XMRV) with these research labs or to indicate that their research findings are false. 7. Paprotka et al based their conclusions on PCR only. The video omits to state that this does not refute findings from separate XMRV detection techniques such as intracellular flow cytometry, Western blot assays and transmission electron microscopy (which other studies have used to demonstrate budding virus). 8. The video fails to state that other researchers have demonstrated an antibody response in human samples to XMRV/MRVs, that there cannot be an antibody response to a lab contaminant, and that Paprotka et al's hypothesis does not explain these findings. 9. The video omits the fact that other researchers are detecting XMRV/MRVs in a significantly higher proportion of patients (e.g. 67+% in CFS patients) than in controls (i.e. ~5%). 10. The video does not discuss the research of Shyh-Ching Lo and Harvey Alter which detected MRVs in CFS patients and identified viral mutation over time, something a lab contaminant would not do. In summary, both Paprotka et al’s experiments (especially the PCR assays) and the computer generated probability calculation need to be validated independently, just as the Lombardi study or any other, before the results can be taken as strong support for their hypothesis.

To allow this single group's hypothesis, based on one study, to be presented as ‘video’ fact is unscientific, especially when there is so much research that contradicts it (including that by the NCI's own internationally renowned retrovirologist, Frank Ruscetti, who continues to find evidence of XMRV infection in humans), and especially when the scientific issue in question is a potentially pathogenic, oncogenic retrovirus that would pose a major public health risk. As such, the video amounts to an advertisement for an unvalidated study which, from a government health department, is particularly lacking in integrity and professionalism.

It would seem prudent for the NCI to either amend the video to include all the facts or withdraw it altogether until it can be edited to include forthcoming research awaiting publication, including that which incorporates immunological evidence for XMRV infection, a finding that renders the contamination theory invalid.

The makers of the video must be aware that the promotion of Paprotka et al’s point of view is likely to encourage a freeze of interest in research into XMRV/MRVs in humans when there is still so much to investigate. This is the antithesis of good science, undermines the principles on which the NCI and NIH are founded, and could help shut the door prematurely on a medical discovery of vital importance. Let the science speak for itself, and let consensus form on its own, or the NCI risks interfering with the natural progress of biomedical discovery.

As a final note, I would suggest that the NCI appoints an XMRV/MRV representative to the Director’s Consumer Liaison Group.

For example, Dr Michael Snyderman is a practicing oncologist who has chronic lymphocytic leukemia and CFS, has tested positive for XMRV and is experiencing health improvement through the use of antiretroviral medication. The inclusion of such a representative should prevent similar research bias occurring in future NCI communications.

Thank you for your attention in this matter and I look forward to your reply.

Yours sincerely, Chris Douglas, Patient Advocate