Dengue and Chikungunya viruses in blood donations: risks to the blood supply?

Abstract

Arbovirus epidemics are raging in tropical areas. Dengue virus (DENV), dengue shock syndrome (DSS), and dengue hemorrhagic fever (DHF) affect millions of individuals every year and cause significant mortality in Latin America, Africa, and Asia. Chikungunya virus (CHIKV) has caused recurrent epidemics in the Indian subcontinent and recent epidemics in Reunion and other islands in the Indian Ocean, with recent detection in areas of Europe. This issue of TRANFUSION includes two articles on the detection of DENV RNA in blood donors from epidemic areas, one article on the clearance of spiked DENV in the course of the manufacture of plasma derivatives, and one article on statistical modeling of transfusion risk during an epidemic of mosquito-borne CHIKV. The surprising seriousness of recurring epidemics of West Nile virus (WNV) in North America has heightened concerns about the potential for introduction and similar epidemic spread of other arbovirus infections in the United States. Dengue has received particular attention since cases have been recognized in the United States at the border between Texas and Mexico.1 The mosquito species that transmit DENV, Aedes egypti and Aedes albopictus, are present in the Southern and Southeastern parts of the United States, raising the specter of significant spread within the United States if the virus is introduced and efficiently spreads. Does dengue represent a risk to the safety of the US blood supply? Transmission by transfusion (TT) is often difficult to evaluate in the midst of an epidemic because the infection could have been acquired through a mosquito bite, through a transfusion, or even through a needle stick. Doubts about TT of WNV were eliminated when the virus was transmitted to multiple recipients of organs and, subsequently, multiple recipients of blood components obtained from the blood donors determined to be viremic.2 Despite the recognition of millions of cases of DENV infection and disease every year, there are very few published reports of transfusion transmission. One report was submitted to the public health agency of Hong Kong and, although accessible online, has not been published in a peer-reviewed journal. The other report describes a recent cluster of TT-DENV in Singapore, which is currently in press (P.A. Tambyah, National University of Singapore, personal communication, 2008). There are also reports of transmission by needle sticks and one case associated with a marrow transplant in Puerto Rico (references are included in the articles being published). Considering the available information, should precautionary measures be considered to prevent TT-DENV in the United States or even more important in countries in which DENV epidemics are occurring at expanding rates? What are the data required for decisions to implement measures that mitigate this potential risk? The articles published in the current issue of TRANSFUSION contribute data that will be needed for policy decisions that may be required in the not so distant future. Linnen and coworkers3 used a transcription-mediated amplification (TMA) assay for detection of DENV RNA to screen 13,372 specimens collected from blood donors in Honduras, Brazil, and Australia, countries with ongoing seasonal dengue epidemics. They identified 9 donors in Honduras who were repeat reactive for DENV RNA by TMA. Eight were confirmed by a polymerase chain reaction (PCR) assay that identified three different DENV serotypes, DENV-1, -2, and -4. Infectious virus could be recovered from four of these donors. Three samples from Brazil were repeatedly reactive on TMA. One was typed as DENV-1 and the other as DENV-3. None of the Australian specimens was repeatedly reactive. Samples were also tested by enzyme-linked immunosorbent assay for immunoglobulin M and immunoglobulin G antibodies to DENV and by plaque reduction neutralization assay to determine the existence and serotype of prior DENV infections in the viremic donors. Five of the donors in Honduras and one of the donors in Brazil appeared to have secondary DENV infections. The authors discuss the implications of reinfection with heterotypic subtypes because of the increased risk of DHF in these individuals. The study thus documented the presence of asymptomatic viremic donors in Honduras and Brazil who could theoretically transmit the virus to blood recipients. Mohammed and colleagues4 screened specimens from 16,521 blood donations made during an 11-week period to the American Red Cross in Puerto Rico, starting 2 weeks after the peak of dengue activity at the end of 2005. Twelve were DENV RNA repeatedly reactive using the same TMA assay used by Linnen and coworkers; 5 of the 12 were reactive in a pool of 16 (the operational pool size for WNV using TMA), and 4 of the 5 were positive on PCR. The four PCR-positive samples had viral loads ranging from 2000 to 80 million copies per mL, of which 3 were typed as DENV-2 and 1 as DENV-3. Infectious virus was recovered from all three of these donors by either mosquito inoculation or cell culture. Interestingly, the authors indicated that some of the 77,000 units of whole blood collected in Puerto Rico are exported to the continental United States each year, implying that these imported units could potentially transmit DENV to continental US recipients. A total of 722 cases of dengue have been reported in Puerto Rico in the first 6 weeks of 2008, so such transmissions could be occurring on an ongoing basis. With regard to growing concerns over potential TT-CHIKV, Brouard and colleagues5 used statistical sampling methods to estimate the risk of viremic blood donations during the CHIKV epidemic that raged through Reunion Island in the Indian Ocean from 2005 to 2007. They placed the peak risk at 1500 viremic donors per 100,000 donations between January and March 2006. Similar to DENV, there were no reports of TT despite estimates that more than 300,000 people were infected by CHIKV during this period, some of whom undoubtedly donated blood during the presymptomatic viremic phase of infection. The French government promptly suspended whole-blood collections in Reunion Island during this epidemic and provided the blood needs from the mainland. Platelet collections by apheresis continued locally, but the collected products were subjected to a process of pathogen inactivation. The statistical approach used by Brouard and coworkers5 is identical to that used by Biggerstaff and Petersen6 when they estimated the risk of transmission of WNV by transfusion during the 1999 epidemic in Queens, New York City, in a prescient article published in TRANSFUSION in August 2002. There had been no reports of TT of WNV at that time. However, 2002 was the year in which WNV TT was demonstrated and the year with the highest number of documented TT cases in North America. Blood donor screening by nucleic acid testing (NAT) for WNV was implemented in July 2003 and is viewed as a great success with interdiction of thousands of viremic blood components. Finally, Xie and colleagues7 address concerns over risk of arbovirus transmission by plasma derivatives by evaluating clearance of spiked DENV serotype 2 as a result of fractionation and processing of therapeutic plasma proteins. They concluded that the multiplicity of steps, including treatment by solvent detergent, cumulatively reduced the risk of viral transmission by albumin by more than 10 log and for immunoglobulins by more than 14 log. The article speaks to the robustness of viral inactivation processes currently utilized by the plasma industry and lessens concerns about the transmission of these viruses by plasma derivatives. These results are quite similar to those that have been observed with WNV that led the FDA to the conclusion that plasma destined for further manufacture does not need to be screened for WNV. What is the reason why the number of reports of TT of DENV and CHIKV are so few? There are many differences between these viruses and WNV, but they do not clearly explain the lack of TT reports for DENV and CHIKV. For instance: WNV infects a large number of birds and mammals. Birds are highly efficient amplification vectors, presenting very high levels of viremia. Many species of mosquitoes that transmit WNV bite both animals and humans. DENV amplification occurs in the salivary glands of A. egypti and A. albopictus. The mosquito transmits DENV from human to human in densely populated areas. There may be differences between WNV, DENV, and CHIKV in the number of days with asymptomatic viremia during which an infectious donation could be given and the titers of infectious virus during this period; for WNV, the period of preseroconversion viremia is 7 days but the levels of viremia are higher for DENV or CHIKV. DENV and CHIKV epidemics currently occur primarily in developing countries. Large numbers of individuals are affected simultaneously, overwhelming hospital emergency rooms, making impossible accurate anamnesis, physical examination, and appropriate reporting. The environment is not conducive to clinical studies, even observational, that could adequately document case reports, let alone estimate rates of TT. This is likely the main reason why TT infections are not recognized. In epidemic regions, blood is diverted to the many cases with DHF and DSS. Thus, many of the patients that receive blood transfusions during the height of the epidemic are already infected with dengue. Postponement of other hospital activities like elective surgeries reduces the opportunity of transmission of infection to naïve patients by transfusions. Lookback is rarely performed in developing countries because of limited resources. If TT of DENV or CHIKV is documented to occur at “significant” rates, it will be important to establish whether disease penetrance is more or less serious than disease transmission from other vector-borne viruses. This issue is of particular interest with respect to DENV, since the more serious hemorrhagic syndromes are thought to be related to high-level anamnestic immune responses. It is possible that TT dengue in immunosuppressed patients will result in no symptoms or mild disease. There are reports suggesting that dengue infections in recipients of kidney transplants are milder than dengue infections of healthy individuals.8, 9 The authors of these reports speculate that immunosuppression interferes with the patient's immune response to the virus and reduces the opportunity for development of DHF. This clinical course is the opposite of what is observed with WNV infection. Immunosuppressed individuals like the elderly and recipients of organ transplants appear to be more susceptible to meningoencephalitis than young patients. It will be important to determine the risk of serious disease associated with TT of dengue and CHIKV because TT of these viruses almost certainly represents a very small proportion of the total infections during an epidemic (as is true of WNV). The value of implementation of donor screening or other high-cost prevention measures to protect blood safety would require careful consideration taking into account prevalence of viremia in donors, transmission rates, and disease penetrance in infected recipients. The availability of a potential donor screening assay for DENV using the same platform as the assays for WNV RNA is welcome and reassuring. It will also be wise to address issues related to potential testing algorithms in future studies, including the definition of a positive specimen, the need for minipool or individual-donation NAT, the evaluation and implications of antibody status of donors, the development of confirmatory assays, and the value of determining DENV serotype. The authors of the studies performed in Honduras and in Puerto Rico have indicated that further studies will be performed in an attempt to better document TT of dengue and estimate the actual risk. The need for a DENV assay and the benefits of introduction of blood donor screening are dependent on the outcomes of these studies. Finally, we hope that public health authorities, regulatory agencies, blood banking organizations, transfusing clinicians, and patient groups continue to support development of technologies that may be useful for pathogen reduction of cellular components. Pathogen reduction is a more generic and proactive approach to the risks associated with arboviruses, precluding the need for implementation of donor screening assays. The example of clearance by plasma fractionation and inactivation procedures is remarkable and should encourage further pursuit of methods applicable to cellular components. It would address DENV, CHIKV, and other reemerging agents like yellow fever virus that is reappearing in South America both in wild monkeys and in humans, with several reported human deaths, as well as other viral agents yet to be identified.10