Overview of Emerging Arboviruses

Ann M. Powers


Future Virology. 2009;4(4):391-401. 

In This Article

Geographic Expansion Combined with Significant Epidemiological Changes

While CHIKV has expanded its geographic range, causing epidemics of unprecedented magnitude, the clinical presentation has remained predominantly consistent since the virus was first described. By contrast, WNV has been evolving epidemiologically with geographic range expansion.

Since its identification in 1937[51] until the mid-1990s, WNV was considered to be a minor human and equine pathogen, causing only occasional outbreaks of mild, febrile illness. During this period, the virus was found to be widely dispersed throughout Africa, the Middle East, Eastern Europe and parts of Asia.[52,53] However, disease was mild, outbreaks were infrequent and were typically restricted to rural localities. Beginning in 1994, WNV was responsible for a number of human and equine outbreaks of increased severity in the Eastern hemisphere.[54–60] This trend began with a small outbreak in Algeria, with 50 human infections, resulting in two deaths.[61] A larger outbreak occurred in Romania 2 years later, with 17 deaths among 393 cases.[62] This outbreak was significant in that it was the first documented urban WNV epidemic. Outbreaks of human, equine and avian disease have continued to this day throughout the Mediterranean basin and Eastern Europe.[63–66]

While WNV outbreaks were increasing in frequency and severity in the Eastern hemisphere, the virus was simultaneously moving to the novel ecology of the Western hemisphere. In 1999, the virus was first identified in New York (NY, USA) during an outbreak of human encephalitis.[67,68] In total, 62 human cases, including seven fatalities, were detected that year in a limited geographic region.[202] One major feature found to be associated with this outbreak was the massive number of bird deaths, particularly among corvids. While occasional avian mortality had been noted in the previous outbreaks in the Eastern hemisphere,[69] the large magnitude of bird deaths in North America was unprecedented.

Since the 1999 outbreak in New York appeared to be relatively small in scope, it was postulated that the virus may have become extinct in the USA over the winter. Unfortunately, the virus was found again in 2000. The total number of cases identified in 2000 remained small (21 human cases); however, the virus had expanded its range to include 12 states in the Northeast USA. Further geographic expansion, with no increase in human cases occurred in 2001, with 28 states reporting detections of WNV. The number of annual human cases remained low until 2002, when over 4100 cases were identified. By that time, the virus had moved across the USA and was reported in 44 states.[202] In addition, WNV was detected for the first time in Canada, where approximately 400 human cases were reported in the province of Ontario.[70] Through 2008, the virus continued to be detected in virtually all states in the USA, with 2539–9862 human cases per year.[202] The virus also tremendously expanded its range in Canada, with seven provinces affected in 2007, resulting in 2215 human infections.[203] After 10 years of activity in North America, WNV is considered endemic and shows seasonal activity, with focal human outbreaks occurring each year. Furthermore, WNV is now the most common cause of arboviral neuroinvasive disease in the USA. Interestingly, the clinical presentation is not only encephalitis, but also includes a large number of acute flaccid paralysis cases,[71–73] a finding not found before the mid-1990s.

The expansion of WNV also progressed southward in the Western hemisphere. Evidence of WNV presence was widespread in the Caribbean, Mexico and Central America by 2003,[74–78] finally reaching South America (Colombia, Venezuela and Argentina) in 2004. However, while antibody against the virus could be found or the virus could be isolated from mosquitoes, there was little indication that WNV was causing human or equine disease in Latin America. Ongoing studies attempt to determine why severe disease or mortality is not associated with WNV in Latin America; to date, no conclusions have been found.

Whether in the Western or Eastern hemisphere, one major question associated with WNV emergence is why the virus changed from a minor human pathogen, causing little severe disease, to one responsible for a significant number of neuroinvasive infections. Numerous genetic studies have been performed to elucidate the role of the viral genome in this transition. Virtually all strains responsible for the recent epidemics in humans, equines and avians belong to lineage 1,[79–81] suggesting that perhaps this group is more neurovirulent. However, lineage 1 also includes several earlier isolates from Northern Africa and the Middle East (where the strain introduced to the Western hemisphere was presumed to have originated). Interestingly, lineage 2 strains (isolated primarily from South Africa and Madagascar), while not historically associated with severe human disease, have also recently been associated with neurovirulence.[82] This finding may simply reflect the likelihood that there is more active surveillance in areas with populations never before exposed to WNV.

In the Western hemisphere, widespread severe disease could easily result from the rapid dissemination of the highly virulent strain introduced into an area with no alternative (i.e., less virulent) strains. Phylogenetic studies have indeed shown that all the North American isolates are derived from the original 1999 New York progenitor strain.[79,81] There has been evolution of the virus as it moved across the USA, with a genotype possessing a point mutation in the envelope gene virtually displacing the original variant.[83] The role of these minor changes in virulence of WNV in the Western hemisphere is the subject of much research.[80] Similarly, in South African lineage 2 strains, mutations that may be associated with increased virulence are being identified. Changes in the nonstructural proteins, particularly NS5 and the 3′ noncoding region, are possible determinants of pathogenicity.[82] Further studies are needed in both lineage 1 and 2 strains to completely evaluate the viral virulence factors affecting humans.