In October 2009, a Science report that patients with chronic fatigue syndrome (CFS) were infected with XMRV attracted a great deal of attention, but subsequent studies could not confirm this finding. Now, a new study in the 22 September edition of ScienceExpress reports that the lab tests used to detect the retrovirus called XMRV in patients with chronic fatigue syndrome are unreliable. The study is accompanied by a partial retraction of the original Science paper.
In the new research, scientists at nine different laboratories, including the authors of the original report, analyzed blood samples from 15 individuals previously reported to be infected with XMRV or a related virus, and from 15 healthy controls. Fourteen of the individuals previously categorized as positive for the virus had chronic fatigue syndrome, and the 15th person had been in contact with a CFS patient.
Each of the nine laboratories independently tested the samples—which were blind-coded, meaning the labs didn’t know which samples they were analyzing—to see if XMRV was present. Only the two laboratories associated with the original report detected the virus. In these two labs, the virus was found in healthy controls as often as in the CFS patients. Tests on replicate samples (that is, when a given blood sample was divided into two or three tubes, with each tested independently) produced inconsistent results.
In their study, Graham Simmons of the Blood Systems Research Institute and the University of California, San Francisco and his co-authors conclude that currently available assays cannot reproducibly detect specific antibodies or direct markers of the virus, such as RNA or DNA, in blood samples. Based on this, they further conclude that screening of blood donors for XMRV is not warranted.
In the October 2009 Science report that first reported the link between the virus and the disease, Vincent Lombardi and colleagues used three detection methods to find evidence of XMRV in CFS patients. The methods included polymerase chain reaction, or “PCR,” to look for XMRV DNA sequences; serological methods to look for XMRV-related antibodies; and cell culture assays to try to grow virus isolated from the patients’ samples.
Co-authors Robert Silverman and Jaydip Das Gupta of the Cleveland Clinic Foundation conducted the experiments for the 2009 paper involving the first of these three methods, analyzing DNA from blood cells of CFS patients and controls. Silverman and Das Gupta have now re-examined the DNA samples they used and determined that some of the samples are contaminated with XMRV plasmid DNA.
Figure 1 of the Lombardi et al. paper, plus Table S1 and Figure S2 of the supporting online material are based on the contaminated data, so the authors are retracting those figures and table. The authors feel the Lombardi et al. paper’s conclusions do not stand on these figures and table alone and can be supported by the remaining evidence reported in the study.
Science currently stands by its Editorial Expression of Concern about this paper, published in May 2011. The editors will be discussing next steps with the authors, in light of the new findings.
A News Focus story by Sciencereporters Jon Cohen and Martin Enserink, appearing in the 23 September issue of Science, looks at this research field more broadly and examines the controversy. Science’s news department also hosted a related ScienceLiveChat at 3:00 p.m. US EDT, 22 September, with Michael Busch, a transfusion medicine scientist involved in XMRV research, and retrovirologist Jay Levy.
Read “Failure to Confirm XMRV/MLVs in the Blood of Patients with Chronic Fatigue Syndrome: A Multi-laboratory Study,” by Graham Simmons and colleagues; and the “Partial Retraction” by Robert Silverman and colleagues. (Full papers free with registration at sciencexpress.org)
Read the abstract and related commentary, as well as Science’s Editorial Expression of Concern, for the 2009 Lombardi et al. paper.
Participate in a ScienceLive Chat about chronic fatigue syndrome at 3:00 p.m. US EDT on 22 September.