Comment on “Detection of an Infectious Retrovirus, XMRV, in Blood Cells of Patients with Chronic Fatigue Syndrome” Andrew Lloyd,1 Peter White,2 Simon Wessely,3 Michael Sharpe,4 Dedra Buchwald5
Lombardi et al. (Reports, 23 October 2009, p. 585) reported a significant association between the human retrovirus XMRV and chronic fatigue syndrome (CFS). However, the cases with CFS and the control subjects in their study are poorly described and unlikely to be representative. Independent replication is a critical first step before accepting the validity of this finding. The finding of a significant association between xenotropic murine leukemia virus– related virus (XMRV) and the enigmatic clinical illness, chronic fatigue syndrome (CFS), has the potential to revise our understanding of the pathogenesis of this condition and raise serious public health concerns (1). It is unusual to find such a strong association between an infectious agent and a well-defined chronic disease, much less an illness like CFS. As such, critical evaluation of the results is paramount. Experienced CFS researchers will remember the 1991 “discovery” of an HTLV-2–like retrovirus in CFS (2), which subsequent studies failed to replicate (3–6).
For reliable results, clearly defined cases should be compared to a control group similar in all aspects other than the disease. The 101 patients studied in (1) were “patients fulfilling the 1994 CDC Fukuda criteria for chronic fatigue syndrome and the 2003 Canadian Consensus Criteria for chronic fatigue syndrome/myalgic encephalitis (CFS/ME) and presenting with severe disability” [supporting online material for (1)], but the latter requires physical signs precluded in the former. The patients were selected, in part, with immunologic perturbations such as in RNase L pathways, yet there are no biomarkers that have demonstrated reliability in affirming the diagnosis of CFS (7). Lombardi et al. (1) provided no description of demographics, illness duration, pattern of onset, or evaluation for exclusionary medical and psychiatric conditions. Likewise, the characteristics of the controls, and details of collection, handling, and storage of specimens, are not described. It is therefore impossible to critically evaluate the findings.
CFS is likely to arise from complex genes-xenvironment risk factors, making a simple causative link between XMRVand CFS unlikely. Even if confirmed, further research will be needed to demonstrate causality. In relation to prostate cancer, with which a comparable association with XMRV was reported (8), negative replication studies are mounting (9). Similarly, three negative replication studies in well-characterized cases of CFS have now been published (10–12). This outcome serves as a cogent reminder of the need for independent replication before findings such as this can be accepted.
References and Notes
1. V. C. Lombardi et al., Science 326, 585 (2009).
2. E. DeFreitas et al., Proc. Natl. Acad. Sci. U.S.A. 88, 2922 (1991).
3. J. W. Gow et al., J. Clin. Pathol. 45, 1058 (1992).
4. M. Honda et al., Microbiol. Immunol. 37, 779 (1993).
5. A. S. Khan et al., Ann. Intern. Med. 118, 241 (1993).
6. W. Heneine et al., Clin. Infect. Dis. 18 (suppl. 1), S121 (1994).
7. M. Lyall, M. Peakman, S. Wessely, J. Psychosom. Res. 55, 79 (2003).
8. R. Schlaberg, D. J. Choe, K. R. Brown, H. M. Thaker, I. R. Singh, Proc. Natl. Acad. Sci. U.S.A. 106, 16351 (2009).
9. O. Hohn et al., Retrovirology 6, 92 (2009).
10. O. Erlwein et al., PLoS ONE 5, e8519 (2010).
11. H. C. Groom et al., Retrovirology 7, 10 (2010).
12. F. J. M. van Kuppeveld et al., BMJ 340, c1018 (2010).
13. M.S. has been paid by insurers for independent advice on medical claims that might include CFS. He has also been paid by lawyers for independent advice on litigation that might include CFS. 22 October 2009; accepted 18 April 2010 10.1126/science.1183706
