Reader Comments

Post a new comment on this article

Thank you, Erlwein et al.

Posted by Ed_D on 07 Jan 2010 at 21:15 GMT

Thanks to Erlwein and colleagues for this prompt investigation of XMRV/CFS and also for their courteous responses to questions. As an uninvolved researcher, I would like to comment on three of these issues.

First, the WPI cohort is said to be vastly different from the cohort in this paper because Dan Peterson's patients satisfy both the Fukuda and the Canadian criteria, while McClure's samples are from patients who satisfied the Fukuda criteria. Compared with the widely used Fukuda or CDC criteria, the Canadian criteria have been relatively little-used and cited other than by their own authors, including Peterson. However it is highly unlikely that none of the 186 patients in the Erlwein et al study would satisfy both the Fukuda and Canadian criteria. Leonard Jason has written that the Fukuda and Canadian criteria overlap substantially and that both distinguish CFS from other, presumably "psychological," cases of fatigue.

Second, while a geographical distribution is possible, it is also unlikely to be so drastic as to explain zero prevalence in a British CFS population. The North American samples were not from one rural hamlet, but from all around the United States. A greater than 95% prevalence of XMRV in CFS, as WPI now claims, and a 4% prevalence in the general US population are inconsistent with confinement within US borders. Erlwein et al’s use of additional primer sets targeting a highly conserved XMRV/MLV region shows that geographically restricted variants are also unlikely.

Third, that Simon Wessely provided the samples for this study has no more bearing on primer specificity than does the fact that Dan Peterson provided the samples for the WPI study, or that another Science author was developing a diagnostic, or that the discoverer of the virus was a co-author. Scientific debate is never furthered by personal attacks.

No competing interests declared.

RE: Thank you, Erlwein et al.

mmcclure replied to Ed_D on 07 Jan 2010 at 22:19 GMT

This balanced and intelligent response is appreciated. Thank you.

Competing interests declared: Author

RE: Thank you, Erlwein et al.

JohnM replied to Ed_D on 07 Jan 2010 at 23:44 GMT

Patient selection has been and continues to be an extremely important yet often overlooked issue in CFS research, due to certain investigators holding forth that CFS is a behavorial disorder characterized by 'abnormal illness beliefs' combined with a phobic avoidance of exercise, while others hold that it is an organic disease process with an as-yet undetermined cause.

This results in a widely recognised and problematic discrepancy regarding patient cohorts, with psychiatric investigators including people with psychogenic fatigue states such as anxiety disorders, depression and PTSD in their CFS studies which biomedical CFS researchers would not include, and with biomedical researchers including people with signs and symptoms that psychiatric researchers would weed out of their studies due to the different concepts of what constitute 'exclusionary criteria'.

The idea of exclusionary criteria were developed to keep patients who suffered from known diseases such as simple hypothyroidism, hepatitis, etc. from being included in CFS studies. However the exclusionary criteria were not meant to exclude every single abnormality, and with some psychiatric researchers excluding people who have such abnormalities as Postural Orthostatic Tachycardia Syndrome(POTS) and abnormal Romberg tests from their CFS research studies, while other researchers view these as actually strengthening their CFS diagnosis, where patients come from actually do make a world of difference and a psychiatric referral service which specializes in behavorial interventions is hardly the first place I would look for 'well characterized' patients to include in a study of such potential import.

To try and minimize the discrepancy between patient cohorts, the Canadian Criteria were developed in an effort to emphasize cardinal features of CFS which are optional in earlier definitions such as post-exertional malaise and neurocognitive impairment, as well as to try and minimize the chances that patients suffering from psychogenic fatigue states would be diagnosed as having CFS.

So while it may be uncouth to single out a single individual, it is by no means uncalled for to bring up one of the most important subjects in the field since the authors do not feel the need to do it themselves.

Competing interests declared: CFS patient

RE: RE: Thank you, Erlwein et al.

AngelaKennedy replied to JohnM on 08 Jan 2010 at 20:10 GMT

I concur with JohnM's response to Ed D. The issue of discrepant patient samples and inappropriate extrapolation of results dogs 'CFS' research, specifically because of the problems documented in JohnM's posts.

It is therefore entirely appropriate to bring up one of the most important aspects, and problems around 'CFS' research, and yes, I would have expected this to have been addressed carefully by the authors at paper submission stage, and by the reviewers during the peer review process. Indeed, I would be interested to find out if this was the case.

Competing interests declared: Social scientist critically evaluating psychogenic explanations for somatic illnesses. Parent of adult (was child) initially given a 'CFS' diagnosis.

RE: Thank you, Erlwein et al.

ClaireC replied to Ed_D on 10 Jan 2010 at 13:31 GMT

The description of patient selection in the WPI study included the following quote from the supporting online materials and methods:

"their diagnosis of CFS is based upon prolonged disabling fatigue and the presence of cognitive deficits and reproducible immunological abnormalities. These included but were not limited to perturbations of the 2-5A synthetase/RNase L antiviral pathway, low natural killer cell cytotoxicity (as measured by standard diagnostic assays), and elevated cytokines particularly interleukin-6 and interleukin-8. In addition to these immunological abnormalities, the patients characteristically demonstrated impaired exercise performance with extremely low VO2 max measured on stress testing."

Were these markers present in the UK patient cohort?

As for the suggestion of a 4% prevalence in the general U.S. population as being inconsistent with confinement within U.S. borders, the WHO data on prevalence rates of HIV in different countries varies from around 0.1% to 25%, a 250-fold difference between countries.

No competing interests declared.

RE: RE: Thank you, Erlwein et al.

npatel replied to ClaireC on 10 Jan 2010 at 22:34 GMT

Requiring the use of such markers in the inclusion criteria can possibly further complicate the issue by introducing a channeling bias. Since the immunological markers that you mention are not exclusive to patients who present with CFS symptoms, it is possible that Lombardi et al. inadvertently selected patients who were more likely to be infected with XMRV (i.e. individuals with immunological abnormalities are more likely to be infected with XMRV, regardless of whether or not they have CFS) . It may well be the case that XMRV has an etiological role in a subset of CFS patients, but this is not clear from the Lombardi et al. study for this reason. Future case-control studies should therefore include appropriately matched controls who also have immunological disturbances but do not have CFS. In addition to this, other lines of evidence are necessary to infer a causal role of XMRV. Without such evidence, both the Lombardi et al. and Erlwein et al. should be considered preliminary.

No competing interests declared.