Current relevance of arbovirus infections in transfusion medicine


  • 3E-S18-01

Anton Andonov, Public Health Agency of Canada

Arboviruses are maintained in nature through biological transmission between susceptible vertebrate hosts by haematophagous (blood feeding) arthropods, such as mosquitoes, ticks, midges and sandflies. Biological transmission can also be transovarian (from an infected female to offspring) and venereal (from a vertically infected male directly to a female vector). Originally, this large group of viruses was called ARBORviruses, a short laboratory name coined in part by William C. Reeves from the words ARthropod BORne viruses; however, the middle ‘R’ was later dropped to avoid potential misidentification with trees (arbor in Latin). The earliest attempts to isolate arboviruses were initiated during the 1930s when Murray Valley encephalitis, Rift Valley fever, St Louis encephalitis, Russian spring-summer encephalitis, Western, Eastern and Venezuelan equine encephalitis viruses were discovered [1,2].

Today, around 520 arboviruses have been catalogued of which over 100 infect humans. This large group of RNA viruses is quite heterogenous both in terms of their genetic features as well as in the variety of disease they cause. Currently, the term arbovirus has long lost its taxonomic significance and is only suggestive of the common arthropodborne transmission strategy. At present, arboviruses are classified among six virus families: Flaviviridae (genus Flavivirus), Togaviridae (genus Alphavirus), Bunyaviridae (genus Orthobunyavirus, Nairovirus, Phlebovirus), Reoviridae (genus Orbivirus, Coltivirus, Seadornavirus-proposed), Rhabdoviridae (genus Vesiculovirus) and Orthomyxoviridae (genus Thogotovirus).

An abbreviated list of medically important arboviruses is presented in Table 1. Other arboviruses for which the available information is insufficient, or which do not seem to cause a disease (although data for pathogenicity may become available in the future) were not considered for this review.

Table 1.   Abbreviated list of medically important arboviruses
VirusGeographic distributionHuman cases*/symptoms
  1. Data presented are summarized from peer-reviewed literature;*indicates average number of cases per year; for the rest, cumulative data based on published reports are presented;**underreported.

Flaviviridae: Genus Flavivirus
A. Mosquitoborne
Dengue virus (DENV) 4 serotypesTropics, subtropics, Southern USA>50 million/fever, rash, DHF/DSS*
Japanese encephalitis virus – JEVAsia>50 000/encephalitis*
West Nile virus – WNVNorth America, but also worldwideThousands/fever, encephalitis
St Louis encephalitis virus – SLEVAmericas (Canada to Argentina)>4000 since 1955/fever, encephalitis
Usutu virusAfrica, Europe2 cases/encephalitis
Rocio virusBrazil>1000 since 1975/fever,encephalitis
Murray Valley encephalitis MVEVAustralia>100 since 1951/fever, encephalitis
Yellow Fever virus – YFVAfrica, South America (sylvatic)>1000/pantropic
Zika virusAfrica, Micronesia-2007>100/fever,rash,arthralgia
B. Tickborne encephalitis viruses
European subtype, Far Eastern subtype, Siberian subtypeEurope, Far East including Siberia and Japan, Mongolia, China, South Korea>150 000 between 1990 and 2007 in Europe only/meningitis, encephalitis
Omsk haemorrhagic feverRussia1488 cases between 1945 and 1949; 165 between 1988 and 1999/haemorrhagic fever
Kyasanur Forest Disease virusIndia100–500/haemorrhagic fever*
Alkhurma virusSaudi Arabia24/haemorrhagic fever
PowassanCanada, US, Russia<100/encephalitis
Louping ill virusEngland, Bulgaria, TurkeyUnknown/encephalitis
Togaviridae: Genus Alphavirus
Chikungunya virusAfrica, Indian Ocean islands, India, Indonesia, Malaysia, ItalyMillions since 2004/severe arthralgia, rash, neurological complications
O’nyong nyong virusAfrica>2 million, Uganda 1959, no recent outbreaks/fever, rash, lymphadenitis
Semliki Forest virusAfrica<100/fever, arthralgia
Mayaro Fever virusSouth America>100/fever, rash, arthralgia
Ross River virusAustralia, Fiji, Samoa, Papua New Guinea, Cook&Solomon islands>5000/fever, rash, arthralgia
Sindbis virus and Sindbis-related Ockelbo, Pogosta and Karelian fever virusesAstraloasia, Africa, Europe>5000/fever, rash, arthralgia*
Western Equine Encephalitis virusAmericas, mostly Western Canada and USThousands/fever, rash, arthralgia
Eastern Equine Encephalitis virusAmericas and the CaribbeanThousands/fever, encephalitis
Venezuelan Equine EncephalitisSouth-Central America, South USThousands/fever, encephalitis
Bunyaviridae; Genus Orthobunyavirus
California encephalitis virus-CALVNorth America1348 (1963–1981, US; 282 (2003–2007, Eastern US)/encephalitis**
La Crosse Virus-LACVNorth America?/encephalitis**
Jamestown Canyon virusNorth America?/encephalitis**
Inkoo virusEurope?/encephalitis**
Tahyna virusEurope, China?/encephalitis**
Bwamba virusAfrica2/encephalitis**
Bynyamwera virusAfrica130 000 (1978–1980, Brazil)/fever, meningitis**
Cache Valley virusThe AmericasThousands/haemorrhagic fever
Oropouche virusSouth America>100 000/meningoencephalitis
Genus Nairovirus
Crimean-Congo haemorrhagic feverEurope, Iran, Turkey?/aseptic meningitis, haemorrhagicfever
Genus Phlebovirus
Rift Valley Fever virusAfrica>100 000/fever
Toscana fever virusSandfly fever Naples virusEurope441 (1985–1989, US)/fever, encephalitis**
Reoviridae: Genus Coltivirus
Colorado tick fever virusNorth America, Europe, Asia?/neurologic symptoms**
Eyach virusEurope, North America?/encephalitis**
Proposed: Genus SeadornavirusKadipiro VirusBanna VirusChina, IndonesiaIndonesia China?/encephalitis**

Typically, the arbovirus transmission cycle in nature involves enzootic amplification (between the arthropod and small rodents or birds) which at times can spill over domestic animals or humans (West Nile, Rift Valley fever, Japanese encephalitis viruses) or human-arthropod-human urban transmission in the case of Dengue and Chikungunya infection (the latter can also be maintained in nature through enzootic (sylvatic) amplification between mosquitoes and non-human primates.

For the most part, various arboviral infections result in mild or no apparent clinical symptoms, with a brief viraemia followed by specific immune response. Whenever there is a clinical presentation, several general syndromes could predominate in individual patients:

  • 1 Self-limited viral fever with or without exanthema and non-specific flu-like symptoms (classical Dengue, West Nile fever, Rift Valley fever, Chikungunya viruses).
  • 2 Febrile illness with skin rash, arthralgia and/or arthritis, myalgia, fatigue, lethargy (Chikungunya, Ross River, Barmah Forest and Sindbis-related viruses, such as Ockelbo and Pogosta viruses).
  • 3 Viral encephalitis, aseptic meningitis, flaccid paralysis (West Nile, St Louis encephalitis, Eastern, Western and Venezuelan equine encephalitis, Murray River Valley, Japanese encephalitis and tick-borne encephalitis Virus (TBEV).
  • 4 Viral haemorrhagic fever, fulminent hepatitis or jaundice (Dengue, Yellow Fever, Rift Valley fever, Crimean-Congo haemorrhagic fever viruses).

Disease manifestation in infected individuals varies accordingly with respect to host factors such as immune status with homologous or heterologous viruses, immunosuppression, age, race, etc.

Secondary Dengue infection caused by a different serotype from the primary infection is considered as a higher risk for Dengue haemorrhagic fever/Dengue shock syndrome (DHF/DSS) especially in young children. On the other hand, certain level of heterologous immunity against one or more previous Dengue episodes may have a net positive effect on the outcome of a secondary/tertiary infection [3–5]. African ancestry and Afro-Brazilian-Cuban-Caribbean ethnicity are protective against DHF/DSS compared to Caucasians [6–8].

Recently, arbovirus incursions have increased on a global scale with a tendency to cause large epidemics; West Nile and Chikungunya virus infections are just few well-publicized examples. Others such as Dengue and Japanese encephalitis viruses have proven not only difficult to contain in their primeval endemic areas, but are invading new territories. West Nile virus (WNV) epidemics in North America posed a major challenge to both public health and the blood banking community; remarkably substantial collaboration between academia, government agencies, industry and the blood establishment proved crucial in addressing and managing the threat from WNV to the blood supply culminating in the rapid implementation of a nationwide screening of blood donations by nucleic acid testing in both United States and Canada. The lessons learned from the WNV experience are valuable and will help us prepare for the next arbovirus calamity.

The recognition that re-emerging arboviruses pose a considerable threat to blood safety is based on the following factors:

  • 1 Worldwide presence of arboviruses; there is hardly a country or geographic region without arboviral infection.
  • 2 Majority of arboviral infections are asymptomatic; therefore, deferral of infected blood donors unlikely.
  • 3 A viraemic phase, although short, is present during most if not all arboviral infections. Published data are available for asymptomatic viraemia for WNV, Dengue, Colorado Tick Fever, European Tickborne encephalitis and Sindbis-related Pogosta virus. Even in cases with overt arboviral infection (Dengue), there is a short prodromal phase of 1–2 days with detectable viraemia preceding the acute phase of disease; although very short during this window period, donor deferral is not possible.
  • 4 The viral load for Dengue, Chikungunya and Tickborne encephalitis is much higher than that of WNV for which transfusion transmission has been documented.
  • 5 Certain arboviruses in endemic areas causing prolonged, large-scale epidemics (Table 1) setting-up a sufficient pool of viraemic donors year round (Dengue, Japanese encephalitis viruses) while others cause explosive outbreaks (Chikungunya, O’nyong nyong, Rift Valley Fever viruses) with a large number of infected people during a very short period of time leading to sharp increase in infected donors.
  • 6 Non-endemic countries to certain re-emerging arboviruses are still vulnerable to potential for outbreaks because of the presence of competent mosquito vectors which could support virus replication in the likely scenario of importation of disease (2007 Chikungunya outbreak in Italy).
  • 7 Vector control has proven inefficient in most developing countries and difficult to successfully implement even in industrialized nations; therefore, once established, an enzootic/epizootic transmission cycle takes time to abolish if at all possible (WNV in North America).

The most prominent among the listed arboviruses is Dengue virus (DENV) which causes an estimated 50–100 million cases of classical Dengue fever annually of which 1–10% may result in severe vasculopathy (DHF/DSS). Despite the colossal number of infections, there are surprisingly few documented reports on Dengue transmission through transfusion [9,10]. Developing countries are bearing the brunt of the Dengue epidemic, and in the midst of a huge outbreak, health care resources are already stretched to the limit making it impossible to properly investigate potential transfusion-associated cases, especially in an environment where health care facilities are often unprotected from mosquitoes. Specific or even heterologous immunity (against a different serotype) may also be a factor in preventing or mitigating the effect of a potential viral transmission through blood products. Further to that the recipients who acquired Dengue through blood seemed to have an overall rather benign course of infection and therefore may not be noticed especially in resource-poor countries. On the other hand, viraemic donors have been identified in Puerto Rico (7·3 per 10 000 units) and Brazil (6/10 000). In 2008, our study in Brazil found much higher prevalence; 1 in 296 viraemic donors in Rio De Janeiro and 1 in 494 in Fortaleza [11]. Although the viral load in asymptomatic Dengue-infected donors may be lower than that of patients with classical Dengue disease and lead to milder clinical symptoms, prevalence rates as mentioned previously are hard to dismiss. A comprehensive assessment of the potential threat of emerging pathogens for blood safety in United States assigned the highest priority among all arboviruses to Dengue virus [12]; coincidently, the same year the report was published continental United States experienced a Dengue outbreak in Key West, Florida [13]. For endemic areas, the threat to blood safety is real although its magnitude remains blurred. A worrisome trend in South-East Asia is the shift of infection from children to adolescents and adults which will affect the donor pool because of deferrals and will eventually increase the risk of transfusion transmission.

Japanese encephalitis is the second most common arbovirus infection in Asia (Table 1). Culex mosquitoes are the most common vector, birds such as herons and egrets are the reservoir but the transmission cycle is widely amplified in pigs. Japanese Encephalitis (JE) incidence is heavily underreported and at least one estimate suggests 125 000 cases annually [14]. The widespread increase in JE is linked to the rapid population growth which doubled for the last 50 years; approximately half of the world population lives in JE endemic areas. Additional factors for its expansion are increased pig farming and irrigated rice production areas. National immunization programs are largely successful in Japan, South Korea and recently in China. With India developing its own vaccination strategy, the long-term trend is for JE incidence to decrease in the region. However, at present, JE is on the rise in low-income countries and expanding in Muslim countries, such as Pakistan and Bangladesh, which were traditionally not affected. There are several circumstances that do not favour Japanese encephalitis virus (JEV) transmission through blood transfusion in endemic areas; (i) Cross-protective flavivirus antibodies mitigate the outcome of infection. People who were exposed to Dengue do not develop severe symptoms when infected with JE [15]. Given the fact that JE-affected areas are also endemic for Dengue, this should be taken into account when assessing the threat of the virus for blood safety. Same seems to be true for WNV antibodies [16,17]. (ii) Viraemia is not only short, but viral loads seem to be much lower compared to DENV, although the data are very scarce and not known how it may affect infectivity.

In typical JE endemic areas, the infection targets mostly children as the herd immunity is very high in adolescents and adults which is beneficial for blood safety. However, adults are equally susceptible as children in new territories claimed by emerging JEV which poses an increased risk for transfusion transmission combined with the fact that JE is often asymptomatic. Despite the perpetual expansion of JEV into the Indian subcontinent, the virus was never introduced to Africa, Europe or through the West Pacific to the Americas. Hypothetically, this could change as the prospect for inadvertent incursion seems to be getting better because of globalization of commerce, tourism, effects of climate change and illegal importation of infected birds or reptiles. Countries such as Canada and United States with large Asian populations may be especially vulnerable as millions are travelling back and forth to their native land. Should that happen, the infection can become easily established as a local zoonosis similar to that with WNV because of the multitude of potential Culex vectors and avian hosts. US Midwest and Canada are also big pork producers although the pig farms are usually in much less-densely populated areas; still let us not forget that pigs serve as a major amplifying host in endemic areas.

Chikungunya and WNV are other major arboviruses known to be a threat to the blood supply; because of the excellent coverage of these agents in the peer-reviewed literature, they will not be a subject of this manuscript.

Rift Valley Fever and O’nyong nyong are also among the arboviruses capable of causing large outbreaks in Africa. Although Rift Valley Fever Virus (RVFV) predominantly affects domestic animals (cattle, sheep, goats) causing major economic losses, it has also resulted in hundreds of thousands of human infections; most are asymptomatic or resemble influenza-like disease. Viraemia is usually high both in animals and humans [18]. In 2000, the virus was introduced to the Arabian peninsula through infected livestock where it caused severe outbreaks (Saudi Arabia, Yemen). A study evaluating the pathways of introduction of RVFV into United States concluded that the most feasible route would be through importation of infected animals, people, mechanical transportation of infected mosquitoes and lastly by bioterrorism [19]. Should this happen, it is important to know the susceptibility of North American mosquitoes to the virus. Two laboratory studies identified several species which can potentially become infected and/or transmit the virus; Aedes canadensis, Culex tarsalis, Aedes vexans, Culex erraticus [20,21]. A similar study identified field-collected Culex pipiens in the Mediterranean (Tunisia, Southern France) as a competent vector and hence the potential for introduction of RVFV [22].

Oinyong nyong virus (ONNC) was detected again in Uganda in 1997 for the first time after the explosive outbreak in 1959–1962 which affected several million people. Anopheline mosquitoes are the vector for the disease, mainly Anopheles gambiae and funestus, which are also found in Asia; other anopheline species found in the Americas and Europe have not been adequately studied for their competence with regard to ONNV; therefore, it poses a hypothetical transfusion risk if inadvertently introduced.

The data provided so far examined briefly the high-incidence arboviruses with regard to proved (DENV) or theoretical transfusion transmission risk. Predicting arboviral threat to the blood supply based on global trends of arbovirus transmission seems somewhat easier compared to projections for local geographic regions, countries, or even continents. Evidently, it is safe to conclude that Dengue, Chikungunya and Rift Valley fever arboviruses are considered a substantial transfusion risk for the tropical/subtropical regions in the future, especially during epidemics. Arbovirus transmission depends on a complex set of local interactions between the vectors, hosts, evolution of local climate change affecting temperature and rainfall patterns; these factors are especially relevant for regions with temperate climate where the effects of global warming according to some experts will be beneficial for further spreading of arboviruses in large. For Europe, two very good examples are the recent incursion of Chikungunya in Northern Italy which had severe repercussion on the regional blood supply and the first two human cases [23,24] of Usutu virus (USUV) infection clinically presented with fever and neurological impairment (encephalitis). As already mentioned, the Chikungunya outbreak was well publicized; however, it is worth to follow the remarkable journey of USUV more closely; isolated 50 years ago in South Africa from Culex mosquitoes, the virus has not been associated with human disease except for a single report describing fever and rash. Around 2000, the virus emerged in Europe and shortly became well established in birds in Austria, Hungary and Italy causing significant avian mortality, and finally in 2009 showed considerable pathogenicity for humans although immunosuppression also may have played a role. Other endemic European arboviruses with potential for transfusion transmission are WNV, TBEV, Sindbis-related Ockelbo, Pogosta and Karelian fever viruses and perhaps to a lesser extent Phleboviruses (sandfly Naples, Sicilian and Toscana viruses). It should be noted that while for the first three agents mentioned previously transmission through transfusion has been documented, the sandfly virus circulation in Mediterranean countries is only a reason for concern for now. WNV outbreaks have been documented in the past (Roumania); since then, mostly isolated cases have been reported. Recently, however, there was increased WNV transmission in Hungary (14 cases in 2008) and Italy (16 cases in 2009). As a result, between August 1 and October 30, 2009 nucleic acid testing has been implemented for blood donation in affected provinces in Italy. Although it is difficult to make predictions, the overall risk from WNV may be mitigated as this infection has been established for a long time in Europe and a possible equilibrium reached within the enzootic cycle between mosquito vectors and avian hosts; it is not clear if the virus may move further North as a result of climate change. Monitoring closely, the situation through entomological, veterinary and human surveillance is imperative.

West Nile virus excluded, in North America Dengue, St Louis encephalitis and Chikungunya viruses have been ranked the highest priority as emerging arboviruses for United States based on a published report from the Transfusion Transmitted Committee of the AABB [12]. In addition, importation of Rift Valley Fever, Japanese encephalitis and O’nyong nyong viruses should also be taken under consideration. Dengue and Chikungunya are not a threat for Canada; unlike the United States, the competent vectors (Aedes aegypti and albopictus) are not present in Canada. Fortunately, the majority of arboviruses and their vectors even if imported will not become established because of the climatic constraints in this country. A notable exception is Rift Valley Fever which if introduced could become established because of the abundance of Culex mosquito species in Canada. Western equine and St. Louis encephalitis viruses as well as Powassan, Colorado tick fever and Jamestown canyon viruses have caused human infections in the past, but are not considered a serious risk.

Dengue remains the major threat to the blood supply in Central and South America, the second ranked is Oropouche virus and to a lesser extent St Louis encephalitis, Mayaro and Rocio viruses.

Ross River Valley is the most common Australian arbovirus causing polyarthritis; viraemia is assumed to exist although its duration and viral load not well known. Murray Valley encephalitis, Japanese encephalitis, Kunjin and Barmah viruses are a potential threat, but DENV is of most concern, causing seasonal small outbreaks on a regular basis in North Australia.

Public health will play a major role in preventing and/or sustaining arboviral incursions when they occur. Key to development of real-time warning system for vectorborne diseases is adequate surveillance. Examples of excellent networks specifically targeting arbovirus ecology are ArboNet in United States, the European Mosquito Control Association (EMCA), the European Network for Diagnostics of Imported Viral Diseases (ENIVD), EpiSouth, a network for the Mediterranean region and the Balkans, the National Arbovirus and Malaria Advisory Committee (NAMAC) in Australia.

Traditional vector control brought spectacular results in the past, eliminating Dengue from Latin America between 1950 and 1980, however, has proven unsustainable in the long run. The current vector control based on insecticide measures is prone to multitude of limitations and more often failures than success stories. Still it should not be dismissed outright; if implemented properly in combination with other prevention programs, it can accomplish a lot; a good example is a low-resource country like Cuba.

A good intervention strategy reducing the risk of arbovirus transfusion is temporary and/or seasonal donor deferral, preparedness to introduce Nucleic Acid Testing (NAT) screening for wide range of arboviruses and technologically advanced pathogen reduction methods capable of inactivation wide range of infectious agents in all blood components.