Human herpesvirus‐8: what (not) to do?

Human herpesviruses (HHVs) are a varied bunch, running the gamut from benign to villainous. Some are ubiquitous, and some are relatively uncommon. Some are transmitted sexually, some by inhalation, some by saliva, some perinatally, and some by blood transfusion. All are persistent, and although they generally establish benign latent infections in healthy hosts, some may cause recurrent disease and or aggressive tumors, particularly in immunocompromised hosts. HHV-5 (cytomegalovirus) is a known cause of transfusion-transmitted disease, and specific measures have been taken to reduce (but perhaps not eliminate) its transmission to patients at particular risk for the development of disease. HHV-4 (Epstein-Barr virus), which is ubiquitous and probably transmitted by transfusions, may rarely be responsible for lymphoproliferative disease among blood recipients, but there are no specific measures available to prevent transmission, although leukoreduction may reduce or eliminate risk. On the other side of the coin, specific immune globulin preparations exist for HHV-3 (herpes zoster virus) and for cytomegalovirus and are of clear prophylactic value in some circumstances. HHV-8 is the most recently described virus in this group.1 It is known to be the causative agent of Kaposi's sarcoma, both classical and human immunodeficiency virus–associated, and is the likely causative agent of several other rare disorders, including primary effusion lymphoma and multicentric Castleman's disease.2 It is naturally transmitted through saliva or sexual contact and has been transmitted through organ transplantation. In some cases, recipients of infected organ transplants have subsequently developed Kaposi's sarcoma.3-5 Because of the pathogenic potential of this virus and because of the evident potential for blood-borne transmission, there has been a continuing undercurrent of concern about HHV-8 among those involved with blood safety. It is important to note, however, that HHV-8 is not a new agent: rather, it is newly described. Presumably, blood from HHV-8-infected individuals has been transfused for many years, yet there is no extant evidence of an association between transfusion and Kaposi's sarcoma or other HHV-8-associated disease states. Evidence for transmissibility of HHV-8 by transfusion has been equivocal at best and there continues to be a somewhat anxious wait for the definitive answer to this key question. In 1997, Blackbourn and colleagues6 reported on the detection of HHV-8 DNA in the blood of a seropositive blood donor; based on evidence of in vitro passage of the virus to allogeneic cells, the authors expressed concern about the potential for transmission by transfusion. In the same year, however, Operskalski and colleagues7 used the TSS repository to examine samples from 10 recipients of blood from 14 donors who were subsequently shown to be HHV-8-seropositive. None of the recipients showed any evidence of infection. Similarly, negative data on recipients of seropositive blood units have been reported from Jamaica.8 In contrast, Cannon and associates9 demonstrated an increased prevalence of HHV-8 antibodies among women injection drug users—a result that was interpreted as evidence favoring blood-borne transmission. Other studies generated similar results. Finally, there have been reports of an increased prevalence of HHV-8 antibodies among some, but not all, groups of blood recipients.10, 11 It is perhaps worth pointing out that there is, as yet, no clear gold standard for HHV-8 antibodies and that the performance characteristics of available tests differ considerably and do not seem to correlate well with the presence of viral DNA.12, 13 False-positive reactions have also been noted. In this issue of TRANSFUSION, Dollard and colleagues14 report on a study of patient samples from the FACT study, which was a major effort designed to determine the frequency of transfusion-associated infections in a large cohort of cardiac surgery patients from 1986 to 1990. Among the 284 patients who were initially seronegative, 2 had unequivocal increases in HHV-8 antibody titers 6 months postoperatively. This translated to a risk of 0.082 percent per unit transfused. Although there were no infections among a control group of 75 nontransfused individuals, the difference was not significant. The overall frequency of existing positive test results in the study population was quite high, at 11.3 percent, although, in a different study, the donor prevalence was found to be 2.4 percent. As with other studies, there is no clear donor-recipient linkage or real proof that transfusion-associated transmission had occurred. Linked donor samples were not available. The study adds to the weight of circumstantial evidence, however, that HHV-8 may be transmissible via transfusion. The authors rightly discuss factors that suggest that their study may overestimate the risk of infection as it is today and are careful to point out that the relationship between infection and the subsequent development of disease is unclear. Nevertheless, the article, along with other data, will inevitably raise the question of whether to invoke the precautionary principle and implement some procedure(s) to interdict presumed transfusion transmission of this virus. My own opinion is that, even without considering the fact that there is little in the way of an available intervention, such a suggestion is premature. By way of illustrating the continuing concern about viral transmission by transfusion, however, the FDA recently asked its Blood Products Advisory Committee (BPAC) whether it was appropriate to defer donors because of a remote risk of exposure to the apparently nonpathogenic simian foamy virus, even in the absence of any evidence of transmission of the agent through transfusion of human blood.15 As it happens, BPAC did not support the proposal to defer donors. There are insufficient data to make a rational decision on the need for an intervention relating to the potential transmission of HHV-8 by transfusion, although it is clear that further work in this area should be encouraged. Despite a great deal of effort, there are still no definitive data to establish, or disprove, the transmissibility of HHV-8 by this route. As far as interventions go, testing is not an immediate option (although the potential for development of appropriate tests does exist). Leukoreduction is likely to reduce the risk of transmission of this cell-associated virus, although it has not been validated in this context. The major risk factors for HHV-8 infection (men who have had sex with men and birth in Africa) are already cause for deferral but there appear to be no other risk factors that could be ascertained by donor questioning. Where does this leave us in the face of the precautionary principle? Its strict interpretation does not allow the option of calling for the development of more data: rather, it calls for action pending the availability of such data. The principle is, however, not absolute and it is recognized that certain conditions should be met before it is invoked. Specifically, authoritative commentary says that the measures to be employed should not be disproportionate to the desired level of protection and must not aim at zero risk. Further, comparable situations should not be treated differently and measures taken should be similar in scope to those already taken in comparable areas in which all the scientific data are available. It has also been suggested that a cost-benefit analysis of action or lack of action may be appropriate.16 Although it is clear that cost-benefit assessment has had little impact on blood safety decisions, it is apparent that there is willingness to consider the potential impact of interventions upon the availability of blood for transfusion. It may be worth considering whether there are situations with which the management of the potential for HHV-8 transmissibility may be compared, although none seems to fit exactly. The TTV-SEN-V complex of circoviruses provoked considerable discussion, were clearly transmissible, and for a time, were thought to be associated with hepatitis—perhaps even fulminant hepatitis. No intervention was implemented. In contrast, testing for antibodies to human T-lymphotropic virus was implemented with relatively little evidence of harm to recipients and limited evidence of transmissibility. Although no final decision has been made, BPAC publicly recommended against an intervention relating to simian foamy virus. Profound measures were taken to reduce the risk of transmission of variant Creutzfeldt-Jakob disease before any evidence of its transmissibility. A decision has yet to be made on testing for antibodies to Trypanosoma cruzi or for Babesia spp. No pattern emerges, so there seems to be little to help us to interpret the precautionary principle as we think about HHV-8. On balance, however, a policy of continuing active surveillance and data development may be most appropriate and would not be unprecedented, but the warning gun has gone off.