XMRV, Q fever and other emerging infections: impact on management of blood safety


  • 3C-S5

Roger Y. Dodd, Research and Development, American Red Cross Holland Laboratory, 15601 Crabbs Branch Way, Rockville, MD 20855, USA
E-mail: dodd@usa.redcross.org


As a result of continuous improvement efforts, the blood supply is now extremely safe. However, emerging infections offer a real or potential challenge to such safety. Since 2003, two new tests have been implemented in the United States in order to reduce the risk of transfusion transmission of West Nile virus and Trypanosoma cruzi. In addition, Babesia microti and dengue viruses are considered to be high-priority threats. Q fever outbreaks in the Netherlands illustrate a localized threat that has been assessed and managed through human and agricultural public health measures and by selective testing of blood donors. The recently recognized gammaretrovirus XMRV has generated scientific and social controversy and illustrates a number of problems in managing blood safety policy.


The microbiological safety of the blood supply is continually improving, albeit as a result of considerable effort. The major focus has been HIV and hepatitis viruses and bacterial contamination, with somewhat lesser attention paid to HTLV and syphilis. In the developed world, the residual risk of adverse outcomes from these agents has been driven to almost negligible levels and creditable progress has been made in the developing world. At the same time, the recognition that other infections continue to emerge or re-emerge has expanded the necessary focus of blood safety. Indeed, in the USA since 2003, two new tests have been introduced for the emerging blood-borne pathogens Trypanosoma cruzi and West Nile virus (WNV) and active research is in place to evaluate means to reduce the risk of transfusion-transmitted Dengue virus and Babesia microti. In other locations, there has been concern about Chikungunya virus, Coxiella burnetii (the agent of Q fever) and simian foamy virus.

In fact, there are many infectious agents that are known, or that have properties that would predict the ability, to be transmissible by blood transfusion. A recent publication listed some 68 infectious agents with this characteristic, but for which interventions have not been implemented [1]. Since that publication, additional agents offering transfusion safety risk have been identified. This paper will discuss the issue of emerging infections and some of the measures that have been taken recently, and will discuss C. burnetii and XMRV and other mouse leukaemia virus related viruses in more detail. Each illustrates key concepts that must be considered in the context of blood safety.

Emerging infections

Infectious agents are said to be emerging if their incidence has increased over the past 20 years, or if it appears that their incidence will increase in the near term. There are numerous factors that underlie emergence, whether the agents be ones that were previously unrecognized or are well-known pathogens that are expanding into new geographic regions or host populations. Some 60–70% of emerging infections are zoonoses: that is, agents that normally infect animals, but which have for a number of reasons, entered the human population. There are many factors associated with emergence; it is however, important to note that in most cases several of these factors work in concert. A number of conditions must be met before an emerging agent can be considered to be a risk to blood safety. The agent must have an asymptomatic blood-borne phase, and the agent must be able to survive in the conditions used for the collection, preparation and storage of blood for transfusion. In addition, the agent must be transmissible by the intravenous route and must cause disease in at least some of the recipients. Until relatively recently, it was anticipated that any new transfusion transmissible disease would share the epidemiologic properties of HIV or HBV – namely a lengthy symptomatic carrier state and a predominantly parenteral/sexual transmission route. We now know that this is not the case, as exemplified by, for example, West Nile virus, a mosquito-transmitted agent with a short incubation period, causing acute infection and disease. Indeed, it is now clear that it is not possible to predict the emergence of infectious agents, neither is it possible to define what will next affect blood safety.

At the same time it is reasonable to consider how best to prioritize among the many agents that are judged to offer some risk to blood safety, provided there is adequate information. This is illustrated in the analysis of agents offering potential risk [1]. However, it must be recognized that this analysis was specific to North America and that, in a different region, there would likely be an entirely different priority grouping. A key issue in undertaking the exercise of priority setting is that two dimensions must be considered. First, the axis relating to the potential frequency and severity of the transmitted disease must be considered, along with the likely pattern of emergence. Second, however consideration has to be given to public and political concern about the infection. This factor is not necessarily proportional to the potential for harm, but relates more to issues of fear and dread. Unfortunately, this is a factor that is not easily quantitated.

Illustrative examples

West Nile virus, USA

Although WNV was originally identified in 1937 and is endemic in parts of Africa, the Middle East and Southern Europe, it was not present in the Western hemisphere until 1999, when a small outbreak occurred in New York City. The virus rapidly spread across the United States and into Canada over the next few years and was the subject of considerable public attention and concern [2]. Although the infection is acute in nature, it was recognized that transfusion transmission was possible and a preliminary risk estimate was published in 2002 [3]. Shortly afterwards, some 22 cases of transfusion transmitted West Nile virus infection were reported in the US [4]. As a result of a cooperative and coordinated effort involving transfusion medicine, regulators, public health agencies and industry, tests for WNV RNA were developed and implemented by blood centres by July of 2003. Subsequently, it was recognized that pooled nucleic acid testing was not sufficiently sensitive to detect all potentially infectious blood donations and mechanisms were put in place to implement single donation testing in locations and times of high incidence for WNV infection. This approach, when properly implemented, has been shown to be effective in eliminating the risk of transfusion transmitted WNV. The management of WNV in the context of blood safety in the US has been recognized as a model of effective management and cooperation [5]. However, as will be shown below, circumstances are not always as clear cut.

Chikungunya virus, Indian Ocean and Caribbean

Chikungunya virus is another mosquito-borne arbovirus that has epidemiologic properties similar to WNV (a flavivirus), even though it is an alphavirus. This virus has been responsible for a number of explosive epidemics involving very high proportions of affected populations. To date, no case of transfusion transmission has been reported, but in some locations, there have been significant efforts to manage this threat. Perhaps best known were the actions taken in la Réunion, an island in the Indian Ocean which is an overseas Department of France. In the face of an overwhelming epidemic in 2005–2007 [6], French authorities implemented three measures. First, the blood system stopped the collection of red cells, providing them from the French mainland. Second, nucleic acid testing for viral RNA was implemented within the island, and third, pathogen reduction methods were introduced for treating locally prepared apheresis platelet concentrates.

Trypanosoma cruzi in the USA and Europe

Trypanosoma cruzi is a flagellated protozoan parasite that is the causative agent of Chagas disease. It is endemic in human populations in South and Central America and parts of Mexico. It is transmitted via an insect vector (reduviid bugs) and the main reservoir is likely a variety of wild and domestic mammals: the insect may live in the walls and roofs of substandard housing. Because the infection is most often lifelong, it may be introduced into non-endemic countries as a result of population movements. Ongoing research studies clearly demonstrated that there was a measurable prevalence of T. cruzi infection in the US blood donor population and that most infected donors had likely been exposed in endemic countries. In addition, a number of cases of transfusion transmission had been recognized. In 2007, blood donor testing was implemented in the USA. Initially the adoption of such testing was not uniform although the majority of blood establishments did adopt a policy of testing every donation (universal testing). However, subsequent to the initiation of testing, routine lookback studies showed that the frequency of T. cruzi infection among recipients of prior donations from donors who tested positive, was very low (2 of 253). This suggested that the donor testing program likely had a low benefit, prompting interest in selective testing strategies to reduce the overall cost and resource usage of testing. The leading strategy, which was approved by regulatory authorities, was to establish procedures whereby every donor was tested once and, if found negative future donation would be accepted without a further test. This approach has led to a 70% reduction in the amount of testing.

Babesia microti in the USA

Babesia spp. are malaria-like protozoan parasites normally transmitted by ticks. At least in the US, the predominant species affecting humans, B. microti appears to be geographically restricted to the coastal Northeast and the upper Midwest, but its range is increasing. It is readily transmitted by transfusion and around 100 clinically recognized cases have been documented since 1977. Although the disease is treatable, the mortality of transfusion transmitted disease appears to be substantial, particularly among those most at risk. Research has shown that the prevalence of infection may be high among blood donors in endemic areas and that there may be a lengthy period of asymptomatic infection (six months or more) [7]. Despite these data and continuing research, only recently has significant attention been paid to any organized intervention to reduce or prevent transfusion transmission of this agent. No licensed test is available although recently, there have been efforts to establish new tests and to implement clinical trials of at least one testing service. It has also become clear that it would be most realistic to initiate blood donor testing on a geographically selective basis; an approach that is not of great interest to test kit manufacturers.

Dengue virus in the USA

Dengue virus is another mosquito-borne flavivirus that causes large epidemics, largely in the tropics. Like WNV, it is transmissible by transfusion, with three reported clusters to date. Studies on blood donors have shown appreciable prevalence rates for viraemia during epidemic periods, using tests for dengue virus RNA or even a soluble NS1 antigen [8,9]. At the end of 2010, an advisory committee to the FDA recommended that interventions to reduce the risk of transfusion transmission of dengue virus should be implemented in areas of the USA that are endemic for dengue. In particular, a great deal of work has been performed in Puerto Rico, a US territory in the Caribbean. However, it is unclear whether appropriate tests for this purpose will be made available in the USA, as the potential market is limited.

New challenges

Two recent situations will be discussed in greater detail. Each illustrates particular aspects of management of an emerging infection. The first situation, a continuing, but seasonal outbreak of Q fever in the Netherlands, shows how human behaviour may promote unexpected disease outbreaks and how concerted local action led to appropriate interventions. The second case is that of xenotropic murine leukaemia virus-related virus (XMRV) and related mouse-derived gamma retroviruses. The case illustrates the difficulties of making effective decisions in the face of limited or inconsistent data and the impact of public concern on the management of such a situation.

Q fever in the Netherlands

Q fever is caused by the small, rickettsia-like bacterium, Coxiella burnetii. The infection is quite widespread among domestic mammals in which it may be transmitted by ticks, and is an occasional cause of human disease. Humans are readily infected by inhalation, as infection may result from a single organism. Additionally exposure may occur through consumption of unpasteurized milk and via exposure to a number of fomites. Disease is characterized by high fever and headache lasting 7–14 days, with pneumonia or hepatitis in 30–50% of infections. About 1% of infections become chronic. The mortality rate is about 1–2% in acute infections and approximately 65% in untreated chronic infections.

Over recent years, there were major clusters of human cases in the Netherlands; these clusters were geographically associated with a number of goat farms. Crowded animal housing along with environmental contamination with infected amniotic fluids during birthing resulted in airborne infection downwind of the farms. Subsequently, public health actions, animal vaccination and changes in agricultural practice seem to have brought the outbreaks under control. The potential threat to blood safety was also recognized and investigated, as it was known that bacteraemia occurred during both acute and chronic infection in humans. Thus, PCR tests were developed and used to test blood donations from areas of high risk [10]. A small number of donors were found to exhibit bacteraemia and there was suggestive evidence that in one or two cases, the infection might have been transmitted to blood recipients. Consequently, in 2010, blood donations from areas considered to offer risk of human infection were tested for C. burnetii by PCR. Fortunately, at the same time, the broader measures to reduce the risk of infection had apparently been effective.


In 2006, a gammaretrovirus was identified among a number of prostate cancer (PC) patients. The virus was related to certain xenotropic murine endogenous retroviruses, hence it was named xenotropic murine leukaemia virus-related reterovirus (XMRV) [11]. Initially, the virus was thought to be associated with a particular mutation of the RNASEL gene (a component of the interferon effector pathway), but subsequent studies by others failed to confirm this relationship. In fact, not all subsequent studies were able to find the virus in association with PC. In 2009, XMRV was reported in 67% of a cohort of patients with chronic fatigue syndrome and 3·7% of 218 healthy controls [12]. In that study, multiple methods were used and infectious virus was isolated from the blood cells of some of the patients. It was noted in the paper and an accompanying commentary that these findings could imply a potential for transmission of XMRV by blood transfusion. Two workgroups were established to evaluate this concern and to advise the transfusion medicine community on appropriate responses. These groups will be described below.

Somewhat analogously to the situation with PC, the observations on CFS also became controversial, as the majority of subsequent studies failed to replicate the initial findings. Perhaps most confusing has been a paper by Lo, Alter and others which reported on a finding of related, murine-leukaemia virus (MLV) sequences from 86·5% of 37 CFS patients and 6·8% of 43 blood donor controls [13]. The sequences found in this study were more closely related to polytropic MLVs than to XMRV and demonstrated quite broad variations in sequences, in contrast to the homogeneity of sequences previously reported for XMRV. Also confusing the issue, a number of papers clearly showed that sample or reagent contamination could generate results comparable to those that had been previously published. As pointed out by Weiss and his colleagues, there have been many reports of putative new human retroviruses, but the vast majority of these have not been confirmed [14]. It is reasonable to suppose that the orderly progress of science will resolve the conflicting evidence. However, in the meantime, there is considerable public debate.

More specifically, some CFS patients have drawn attention to XMRV and MLVs and have suggested that they represent a threat analogous to HIV and AIDS. Further, there has been pressure from these same groups to require that individuals with CFS or a history of CFS be deferred from blood donation. It appears that this is designed to attract attention to the disease and to promote the need for more effective funding. A great deal of attention is also given to promoting the philosophy that XMRV and MLVs are indeed etiologically related to CFS and in denigrating research that suggests otherwise. Furthermore, the US FDA has considered the possible impact of CFS and XMRV/MLV on blood safety. At the end of 2010, an advisory committee to the FDA did, in fact recommend that individuals with CFS or a history of CFS should be permanently deferred from donation, despite the existence of voluntary actions taken on the recommendation of the AABB. The FDA has publicly expressed concern about the possible threats of retroviruses to the blood supply, even in the absence of evidence of disease causation, although they have not as yet (March 2011) issued any guidance on the management of XMRV/MLV in this context.

As noted above, two working groups have been established to consider the issues relating to XMRV/MLV and blood safety. The NIH scientific working group has been charged with providing data on three issues: defining the prevalence of XMRV/MLV infection in the US blood donor population; determining whether these viruses are transmissible by transfusion; and if such transmission occurs, defining its clinical significance. To date, the group has focused on defining the performance characteristics of available tests in order to establish a standardized approach. Initial studies on sample panels constructed by spiking blood and plasma with virus or infected cells containing an original isolate of XMRV from PC patients has shown that those nucleic acid test systems evaluated to date appear to be sensitive and broadly comparable [15]. Subsequent data, reported but not published and using samples from CFS patients and pedigreed negative controls have been less easy to interpret. They have also provided a brief literature review of epidemiologic data that fail to support a relationship between blood transfusion and PC or CFS.

An AABB working group has provided information for transfusion medicine specialists in the USA. In brief, they suggest that there is insufficient scientific information to make recommendations relating to XMRV/MLV. However, they have recommended that blood collecting organizations should actively discourage blood donation by those with a current or past medical diagnosis of CFS [16]. Materials were provided to educate potential donors about the issues. This recommendation is consistent with existing recommendations that CFS patients should refrain from giving blood, largely for their own protection. However the recommendation also recognizes the fact that many chronic infections have potentially been linked to CFS.


The author declares that there are no potential conflicts of interest.