| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In January 2005, public health officials noted increased reports of dengue-like febrile illness with severe debilitating joint pain in Grande Comore (Ngazidja) of the Union of the Comoros. Initially, screening for dengue virus infection was performed on a subset of the samples but the percentage of dengue-confirmed patients was too low to account for the increase in febrile illness. In February 2005, the World Health Organization (WHO) African Regional Office (AFRO) was informed of the outbreak and 25 human serum samples were delivered to the Kenya Medical Research Institute (KEMRI) in Nairobi where preliminary results indicated that an alphavirus was responsible for the illnesses. IgM antibody testing was performed on the samples against both CHIKV and ONNV. CHIKV-specific IgM antibodies were detected in 9 samples and antibodies against ONNV were present in 1 specimen. Additionally, sequencing analysis on RT-PCR amplicon products indicated that the infecting virus was indeed CHIKV.
In early March 2005, a team including members from CDC, KEMRI, WHO, and public health officials from Comoros initiated an investigation of the outbreak in which over 1,100 cases had already been reported.10 This report describes the results of field work focusing on entomological investigations conducted from March 11 through March 31; during this period, approximately 2,500 additional cases were documented. Fortuitously, this time frame corresponded to the peak of the outbreak as determined by the number of clinical cases reported suggesting that timing was optimal for carrying out our objectives of identifying the primary vector(s) involved in transmission of the virus and noting entomological parameters that contributed to the epidemic.
|
MATERIALS AND METHODS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Grande Comore is covered with lava exhibiting a porous surface that is unable to maintain ground water. The island has no rivers or wells that can provide reliable water sources. Additionally, there are no piped water or sanitation systems in place; therefore, the inhabitants of this island harvest rain-water that is stored in large concrete cisterns. Most of the cisterns are either only partially covered or are completely uncovered. Predatory fish to control mosquito larvae were found in only some cisterns. Apart from large cisterns for more extended storage, water is also commonly stored in smaller containers including plastic and metal reservoirs at individual households for cooking and drinking purposes. Additionally, similar small containers that serve as ornamental plant holders and decorative vessels are a common feature around homes and business premises in the city. In contrast, the other 3 islands of the Comoros have rivers and piped water systems resulting in a reduced need for cisterns and small water-storage vessels. However, the effect of these differences on virus transmission has not been confirmed.
Garbage collection facilities do not widely exist in the cities of the island. For this reason, discarded plastic and metal water-holding containers, used tires, aluminum soda cans, and empty food cans were found widespread throughout the city. This was particularly pronounced near the urban center and dock areas of Moroni.
Adult mosquito collection. An adult mosquito survey was carried out using two main methods. Host-seeking mosquitoes were collected using human landing collections while resting mosquitoes were collected using vacuum aspiration; both methods have been found to be highly productive for collecting adult Ae. aegypti mosquitoes.11,12 In the landing collection method, mosquitoes were captured in glass vials when they arrived to feed on exposed skin of public health and/or mosquito control team members. All volunteers were only involved in collections if they had previously recovered from CHIKV-like symptoms and collections were performed with approval by WHO and Comorian Health Office personnel. All investigations reported here were approved by the Comoros Ministry of Health as part of the emergency public health response to the outbreak and determined by this body to not represent research requiring review by an ethical review group. Collections were performed during peak daytime biting hours from 12:00 to 18:00 each day for 3 weeks starting March 11, 2005 at the peak of the outbreak. A 3-week time period allows the investigation of nearly 2 life cycles of Ae. aegypti. This amount of time is more than sufficient to obtain information regarding density and infection rates particularly when the timing of the collections is found to correspond to the peak of the epidemiologic curve. Collections were done near the homes of clinically ill patients primarily within the transmission areas of Moroni where the majority of the clinical cases were identified; some collections were also made in Mitsamiouli and Foumbouni on Grande Comore. For the vacuum aspiration method, a battery-operated backpack aspirator was used to aspirate mosquitoes resting in bedrooms and living rooms of houses in the neighborhoods of affected individuals. Landing collections and aspiration collections were performed in the same localities to collect the complete range of mosquitoes present in outbreak areas. The use of alternate collection techniques was attempted but unsuccessful for several reasons. For example, using CDC light traps was frustrated by lack of access to dry ice as bait and the overwhelming abundance of containers serving as oviposition sites precluded the use of gravid traps.
All collected mosquitoes were sorted by sex, species, and collection method into cryovials at a temporary facility set up in Moroni. Samples were initially frozen at –20°C then later transferred to liquid nitrogen charged dry shipping tanks and transported to the CDC at Fort Collins, Colorado, where they were stored at –70°C until processing.
Larval surveys. All indoor and outdoor water-containing receptacles at randomly selected domestic and business centers within 5 distinct zones in the city of Moroni were inspected for mosquito larvae from March 27 through March 31. The 5 zones were selected to ensure no bias existed in the collections due to differences in habitat. The zones included the old/central part of Moroni with dense housing, coastal Moroni adjacent to the ocean and ports with little vegetation, "suburban" housing areas with more ornamental vegetation, northern Moroni where houses are more dispersed, and eastern Moroni where the city transitions to a mountain habitat. A total of 100 houses were sampled in the city with 20 sampled per zone. Live larvae observed in positive receptacles were sampled and examined for identification. Containers positive for Aedes aegypti larvae were recorded. The larval indices calculated from the collected data include container (percentage of water-holding containers examined that contained Ae. aegypti larvae), household (percentage of houses examined that have Ae. aegypti larvae in at least one container), and Breteau (total number of containers with Ae. aegypti larvae per 100 houses) indices.
Mosquito processing. Mosquitoes were retrieved from storage at –70° and held on a chill table while being sorted and identified to species using appropriate taxonomic keys and references.13–17 Specimens were assigned to pool by species, sex, location, trap method, and date of collection. Each pool (not exceeding 40 mosquitoes) was homogenized in 2 mL DMEM diluent.18 The homogenate was clarified by centrifugation and used for virus isolation and nucleic acid detection. Minimum infection rate (MIR), or the [number of positive pools / total specimens tested] x 1,000, was calculated for all species cohorts that were found to be positive for virus or viral nucleic acid.
Virus isolation. For virus isolation, 1 mL of undiluted homogenized mosquito suspension was inoculated onto 25-cm2 flasks of Vero cells. Cells were incubated for 1 hour at 37°C to allow attachment of virus; then, 4 mL of DMEM diluent was added to inoculated cells. Cells were incubated at 37°C in a 5% carbon dioxide incubator and observed daily for 10 days. Supernatant was harvested when cytopathic effects (cpe) were observed. All harvested supernatants were titrated using a standard plaque assay to confirm the presence of virus and to determine titer.19
Virus detection by RT-PCR. Reverse transcription-polymerase chain reaction (RT-PCR) was used to detect and identify CHIKV-specific nucleic acid. RNA was isolated from mosquito homogenates by using QIAamp viral RNA minikit (QIAGEN, Valencia, CA), following the manufacturer’s protocol. RT-PCR assays were performed using Titan One tube RT-PCR kit (Roche Molecular Biochemicals, Indianapolis, IN). Each reaction contained 10 µL of RNA and 20 pmol of each CHIKV-specific primer, 7028 forward (5'TGCGCGGCCTTCATCGGCGACTAC 3') and 8288 reverse (5'CCAGGTCACCACCGAGAGGG 3'), in a 50 µL total reaction. Amplified product was analyzed by gel electrophoresis and all positive samples were repeated for confirmation using the above primers or a CHIKV-specific real-time RT-PCR assay. CHIKV specificity was confirmed by sequencing the amplicons.
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
.
|
Pools were further tested for the presence of CHIKV by attempting virus isolation in Vero cells. All cultures that showed any cytopathology were then harvested and reexamined for CHIKV in a plaque assay. One culture, from a pool of 4 Ae. aegypti females, contained viable virus and an isolate was obtained. This pool corresponded to one of the positive pools from the RT-PCR analysis. The inability to isolate virus from other pools that were RT-PCR positive may have been due to the different sensitivities of the assays or possibly the loss of virus viability due to suboptimal temperature maintenance conditions of the mosquitoes during collection and shipment. Sequence analysis (data not shown) confirmed CHIKV that was found to be most closely related to strains from the Central/East African genotype.8,20
Larval survey. Because a large proportion of the mosquitoes were suspected to be Ae. aegypti and it was unknown if Ae. albopictus were present, a larval index study was conducted to provide a measure of risk to the population due to exposure to mosquitoes in household areas in Moroni by randomly sampling in and around homes in 5 distinct zones exhibiting different ecological patterns. The 5 zones examined ranged from urban settings in central Moroni, neighborhoods adjacent to the coast and port areas, sparsely populated regions approaching the mountains, and rural or suburban habitats with more dense vegetation. Selected homes were examined for the presence of containers with standing water and mosquito larvae in those collection vessels. There were 9 general groups of water-containing vessels identified including natural sites (e.g., puddles, split coconuts) as well as artificial containers (e.g., cisterns, tires, cans, jars, cooking pots). All of these, with the exception of waste pits/septic tanks, had larvae that were collected from them in at least one home.
Several indices were calculated to estimate the Ae. aegypti population density including the house index, container index, and Breteau index.12 The overall percentage of homes examined that had Ae. aegypti larvae in containers (house index) was 68% with a container index of 31%. Additionally, the Breteau index was 126; a Breteau index above 50 historically has been used to indicate a high risk of urban yellow fever virus transmission while more recent estimates indicate YFV epidemic risk with Breteau indices of 5–5012,21 and dengue virus outbreak risk has been noted with a Breteau index of 
1.22
In addition to the overall high density figures, several types of containers were found to have extremely high infestation rates. The most significant of these were all artificial containers present at virtually every home examined. Discarded tires were the containers that had the highest prevalence of Ae. aegypti larvae with almost 60% of those containing water positive for larvae. Only slightly less important were temporary water storage jars with 52% of these positive (Table 2). Natural water collection receptacles (e.g., split coconuts) were found to have Ae. aegypti larvae far less frequently than the extremely abundant artificial containers. A complete listing of containers and infestation rates is presented in Table 2.
|
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
By performing entomological collections throughout this outbreak, we had an opportunity to evaluate which mosquitoes may be vectors of recent epidemic CHIKV in East Africa. During the first documented CHIKV outbreak in 1952–1953, Ae. aegypti was considered the principal vector,28 particularly in urban settings. In Comoros, the outbreak occurred largely in urban areas suggesting the possible involvement of Ae. aegypti. Our investigations therefore focused upon optimizing collections of these and other domestic or peridomestic mosquitoes. Both immature and adult collections contained significant numbers of Ae. aegypti with high larval indices and a high percentage of total adults collected; abundance of a particular species is certainly one criteria for determining vector status during an outbreak. Recent literature has questioned the predictive value of these larval indices for risk of Ae. aegypti-borne infections22,29 and numerous attempts to identify or develop more appropriate pupal and other indices has been undertaken.30–33 However, to date, there is no solid consensus as to which immature indices best provide this risk estimate and the container and Breteau indices in particular are still frequently used as population density indicators. Furthermore, because Ae. aegypti are day-biters, anthropophilic, and display interrupted feeding patterns, even a moderate mosquito population (assuming there are infected mosquitoes present) could contribute significantly to the spread of disease. Finally, the fact that infectious virus and nucleic acid were obtained virtually exclusively from Ae. aegypti further supports the hypothesis that this species was the primary vector during the outbreak in 2005.
Because limited options for mosquito collections precluded the examination of mosquitoes not closely associated with human habitation, the involvement of other species in the transmission of CHIKV is still a possibility. Previous reports have implicated members of other genera including Anopheles rufipes, An. coustani, and Culex ethiopicus in West Africa2 as well as Coquillettidia fuscopennata3 and Mansonia africana.34 Additionally, although virus has not been isolated from wild collections, Eretmapodites chrysogaster has been shown to have a higher vector potential than Ae. aegypti through experimental infection studies in the laboratory.35,36 Our collections did contain some Eretmapodites spp. and Ae. simpsoni complex mosquitoes but unfortunately, the number of specimens was too small to provide meaningful information on infection potential. Further, our collections did not extensively consider species that preferentially feed on non-human vertebrates. Previous seroprevalence studies have frequently detected CHIKV-specific antibodies in monkeys and on occasion in rodents and birds suggesting that enzootic transmission cycles may also be present.2 This possibility could have implications for long-term or repeated outbreaks of CHIKV in the human populations. However, Grande Comore is not known to have any monkeys on the island so further studies to characterize any enzootic maintenance of the virus are warranted.
One finding of these entomological investigations was the identification of prime Ae. aegypti breeding sites. Considering the extent to which large water cisterns were used to provide water for the residents of Moroni, it was expected that these were the key premises for Aedes aegypti larvae. Because collections from the cisterns were not practical, the role of these storage vessels in mosquito production is still unknown. However, the smaller, common household containers and discarded tires yielded significant numbers of larvae. Although further characterization of the large vessels is still warranted, the information concerning abundance of larvae in "disposable" breeding sites provided some concrete information on sources of vectors that allowed the development of both long-term and short-term control strategies for public health officials in Comoros.
Received October 28, 2006. Accepted for publication August 8, 2007.
Acknowledgments: We thank the people of Comoros for their support during the sampling operations in their homes. We also thank the officials of Ministry of Health of Comoros for inviting us and facilitating our operations, the staff of the WHO country office for logistic and administrative support, and John T. Roehrig, Roy L. Campbell, Ned Hayes, and the Division of Vector Borne Infectious Disease (DVBID) travel group for facilitating the participation and travel of the CDC team members.
Disclaimer: This article is published with permission of the director, Kenya Medical Research Institute and the Comoros Health Ministry office. The findings and conclusions in this report are those of the author(s) and do not necessarily represent the views of the Centers for Disease Control and Prevention.
* Address correspondence to Ann M. Powers, CDC, 3150 Rampart Road, Fort Collins, CO 80521. E-mail: APowers{at}cdc.gov 
Authors’ addresses: Rosemary C. Sang, Kenya Medical Research Institute, Nairobi, Kenya. Ouledi Ahmed, Ali Ahmed Yahaya, Ibrahim Mmadi, and Ali Toilibou, Ministry of Health, Moroni, Union de les Comores. Ousmane Faye, Naftali Agata, Allarangar Yakouide, and Mamadou D. Ball, World Health Organization, African Regional Office (AFRO), Brazzaville, Republic of Congo and Moroni, Union de les Comores. Kibet Sergon, Field Epidemiology and Laboratory Training Program, Nairobi, Kenya. Robert F. Breiman, U.S. Centers for Disease Control and Prevention—Kenya, Nairobi, Kenya. Cindy L. H. Kelly, Jennifer Brown, Barry R. Miller, and Ann M. Powers, Division of Vector Borne Infectious Diseases, Centers for Disease Control and Prevention, Fort Collins, CO, USA.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
