Overview of Emerging Arboviruses

Ann M. Powers


Future Virology. 2009;4(4):391-401. 

In This Article

Geographic Expansion of Chikungunya Virus

In contrast to the limited number of YFV cases in South America, CHIKV has re-emerged in the 21st century in a succession of massive outbreaks, ranging from East Africa to India to numerous islands of the Indian Ocean. Epidemic activity was detected in the Democratic Republic of the Congo between 1999 and 2000,[6] with concurrent outbreaks occurring in Indonesia.[7] These outbreaks were small–moderate in magnitude, similar to the majority of previously documented CHIKV episodes. Then, in 2004, two coastal communities in Kenya experienced significant CHIKV epidemics. First, the resort community of Lamu Island had an outbreak of dengue-like illness that was initially thought to have affected approximately 1300 people. A subsequent serosurvey revealed that the attack rate was actually 75%, suggesting that 13,500 of the island's 18,000 residents had been infected by the virus.[8] A few months later, the virus was identified in Mombasa, a town to the south of Lamu, and genetic evidence implicated the same lineage of virus that was found in Lamu.[9] In early 2005, a suspected dengue outbreak occurred in the islands of Comoros; however, dengue virus-specific tests performed on serum samples were found to be negative. Additional testing for other arboviral pathogens confirmed the presence of CHIKV on Grande Comore Island.[10] Again, genetic characterization of the virus indicated that the origin of the virus was coastal Kenya, indicating that the 2004 epidemics had spread to islands of the Indian Ocean where the virus had never before been documented.[9] Similar to the Kenyan outbreaks, a serosurvey was conducted, demonstrating that 63% of the population had been infected with the virus. Given the population of Grande Comore, almost 215,000 individuals were calculated to have been infected.[10] An outbreak of this magnitude had not been documented since the 1964 outbreak in Madras, India.[11,12] While the outbreak was ongoing in Comoros, transportation of viremic individuals between Comoros and nearby La Reunion Island introduced the virus to another new island.[13,14] Initially, only a small number of cases were identified each week during 2005; however, in early 2006, the epidemic exploded, with approximately 40,000 new cases reported each week. By the time the outbreak subsided in La Reunion, it was estimated that more than 255,000 individuals had been infected.[15] The outbreak did not end, but rather continued its expansion; in 2006, India saw the reappearance of epidemic CHIKV disease after three decades of inactivity.[16–18] The magnitude of the outbreak in India made all the previous activity in the Indian Ocean islands appear almost insignificant. Over 13 states, clustered primarily in the west and southern areas of the country, including Andhra Pradesh, Karnataka, Maharashtra, Madhya Pradesh, Tamil Nadu, Rajasthan, Gujarat and Kerala, were affected, with an official estimate of 1.3 million cases. Sequencing data obtained from Indian samples confirmed the Central/East African genotype of the lineage found in the 2004–2006 outbreak areas was responsible for the Indian outbreak.[19] With the magnitude of the outbreak in India and the significant amount of air travel through this country, it is not surprising that there were coincident cases in Sri Lanka, Malaysia and Indonesia.[20–22] After 4 years of epidemic activity, public health officials of any country with populations of the vector mosquitoes Aedes aegypti and Aedes albopictus were concerned about the risk of establishment of the virus in novel ecological niches. During the summer of 2007, this precise event occurred in Italy when CHIKV was introduced by a viremic individual returning from India, resulting in autochthonous transmission.[23,24] While the Italian outbreak was limited to fewer than 300 cases of febrile illness in two villages and no subsequent local entomological activity was detected during the winter or following spring, this was the first evidence of movement of the virus to a new locale, where establishment of the virus in an enzootic cycle was a true possibility. The virus has continued to cause epidemic activity in India and Southeast Asia into 2009, with no apparent signs of the outbreaks ending.

During the massive outbreaks of CHIKV, not only did the geographic range of the virus change, but the clinical presentation appeared to be evolving as well. Similar to other closely related alphaviruses, CHIKV typically causes fever with a severe and frequently persistent arthralgia of the peripheral small joints.[25] Occasionally, atypical presentations, ranging from mild hemorrhagic manifestations to myocarditis to meningoencephalitis, were reported; however, these were infrequent and isolated cases.[26–29] However, when the virus activity peaked in La Reunion in 2006, there were reports of neurological manifestations, intrauterine infection and neonatal disease, as well as CHIKV-associated mortality.[30,31] Numerous patients found to have CHIKV infections who developed meningitis or encephalitis during their illness were identified; this presentation was unusual for an arthralgic alphavirus and led to the speculation that the virus may have been evolving into a more virulent form. Curiously, no reports of these unusual manifestations were reported in Comoros or Kenya, even though the same number of cases had occurred. The only change in clinical presentation prior to the La Reunion reports was that rash was infrequently identified, even though it was previously considered one of the classic symptoms in the CHIKV fever triad, along with fever and arthralgia. Recently, several reports of clinical cases with atypical presentations from patients in La Reunion and India indicated these patients were commonly older and had underlying conditions that may have predisposed them to more severe or altered patterns of disease.[31,32] These careful, detailed case studies should eventually be useful in establishing whether any causal associations between either neurological or neonatal disease with viral infection are occurring. In addition, the outbreak in La Reunion was the first time a large number of fatalities had been attributed to CHIKV. Previously, there were occasional reports of CHIKV-infected individuals succumbing to their infections; however, the patients were typically found to be coinfected with another pathogen (e.g., a dengue virus) or there may have been other conditions that more directly caused their deaths.[33,34] In La Reunion, crude death rate estimates suggested that over 200 deaths could be attributed to CHIKV.[15] There were some additional reports from India also reporting deaths, but the numbers were small in comparison to the overall outbreak numbers.[35] These crude death rate calculations were, of course, only postulations with regard to the cause of mortality; however, they did motivate clinicians and researchers to carefully investigate the causal relationships between CHIKV and mortality. Importantly, even though there were reports of unusual clinical manifestations, the vast majority of cases presented with classical CHIKV fever.

Finally, perhaps most significantly, the virus also showed indications that it could alter its transmission cycle and more efficiently use an alternate mosquito vector. In the zoonotic cycle in Africa, the virus circulates in an enzootic cycle between nonhuman primates (chimpanzees, monkeys and baboons) and mosquitoes of the wooded savannah, such as Aedes africanus, Aedes furcifer and Aedes luteocephalus.[36,37] While CHIKV most likely evolved from this forest cycle in Central Africa, the virus developed a new urban cycle as it spread through Africa and Asia. Mosquitoes such as Ae. aegypti have been implicated as vectors in Africa in urban settings[38] and, in Asia, transmission of CHIKV has been shown to be almost exclusively between Ae. aegypti mosquitoes and humans.[39,40] Other vertebrates and vectors have occasionally been shown to harbor the virus; the likelihood that they contribute significantly to virus maintenance is low. For example, there have been rare isolations of the virus from Culex species mosquitoes; however, these are likely to be instances of capture of a mosquito that has recently ingested virus after feeding on a viremic human.[41] Furthermore, nonhuman primate serosurveys from Asia have revealed the presence of CHIKV-specific antibody, but, since the levels were often low, it is unclear if these vertebrates contribute to virus persistence.[42,43]

The mosquito vector implicated in the 2004 epidemics in Kenya and in Comoros in 2005 was Ae. aegypti, the traditional urban vector. In Comoros, larval surveys indicated that at least 11 different species were present on the island,[44] three of which (Ae. aegypti, Aedes vittatus and Eretmapodites chrysogaster) had previously been shown to be competent CHIKV vectors in either field or laboratory studies.[45–47] However, virus was isolated from only Ae. aegypti and the minimum infection rate was over 30 for this species. In addition, Ae. aegypti comprised over 60% of the total infected mosquitoes collected, providing further evidence that this was the vector probably causing the outbreak in Comoros.[44] While no entomological studies were performed during the outbreak period in La Reunion, investigations performed after the epidemic indicated that the predominant mosquito was Ae. albopictus, a species that had been shown to be a competent laboratory vector of CHIKV.[48,49] In fact, in some areas of the island, no Ae. aegypti were present at all, implicating alternate vectors not only in the transmission of the virus, but also in the tremendous increase in the rate of cases several months after the virus was first identified on the island. Genomic sequencing was performed on numerous isolates from La Reunion and one mutation in particular was studied in order to examine the shift in vector usage. A change in the envelope gene from an alanine to a valine residue at position 226 was noted and was postulated to affect cholesterol dependence of the virus in the mosquito vector.[14] Laboratory studies of vector competence and viral replication in both urban vectors, Ae. aegypti and Ae. albopictus, with viruses containing either amino acid, do reveal differences in the levels of replication and dissemination of the virus in the secondary vector Ae. albopictus.[48,50] It is important to note that the outbreak had been ongoing for over 1.5 years, affecting over 220,000 people prior to the identification of this mutation; however, the presence of only the secondary mosquito vector in La Reunion may have forced the selection of a variant with altered vector competence.