- Email: [email protected]
Rift valley fever on the east coast of Madagascar
(~) INSTITUTPASTEUR/ELSEVIER Paris 1991
Res. Virol.
1991, 142, 475-482
Rift Valley fever on the east coast of Madagascar J. Morvan (1), J.-F. Saluzzo (z), D. Fontenille (3), P.E. Rollin (2) and P. Coulanges (l) it) Unitd de Recherche sur les Arbovirus, Institut Pasteur, Antananarivo (Madagascar), ~2) Laboratoire des Fibvres hdmorragiques virales, Institut Pasteur, Paris, and (3) Laboratoire d'Entomologie mddicale, Institut Pasteur, Antananarivo
SUMMARY In March 1990, a Rift Valley fever virus (RVFV) outbreak was suspected in the district of Fenerive on the east coast of Madagascar after an abnormally high incidence of abortions and disease in livest(,ck. Sera from humans and cattle were tested for RVFV antibodies by immunofluorescence assay (IFA) and ELISA-IgM capture. Sma and mosquitoes collected in the same area were tested for virus isolation by tissue culture and suckling mouse intracerebral inoculation, and for antigen detection by an ELIZA antigen capture assay. Among cattle from the area, RVFV antibody prevalence was 58.6 % by IFA and 29.6 % by ELISA-IgM. In contrast, human populations in the same area had a lower RVFV antibody prevalence, with 8 . 0 1 % IFA and 5.4 % IgM-positive sera. No RVFV antigen was detected and virus isolation was unsuc~:essf, I from the sera and mosquito pools tested. Different hypotheses concerning the emergence and diffusion of RVFV in this area and the occurrence of the outbreak are discussed.
Key-words: RVF, Madagascar = Cattle, Man, Mosquito vectors, Antibodies.
INTRODUCTION Rift Valley fever virus (RVFV) was first isolated in 1930 in Kenya during an occurrence of abortions in sheep and cattle, and febrile illness in humans (Daubney et aL, 1931). In the decades which followed, successive outbreaks of zoonotic disease were reported in sub-Saharan Africa. In 1977-1978, an extensive outbreak was reported in Egypt (Meegan, 1979). In Madagascar, RVFV was isolated in 1979 (Clerc et al., 1982) from several pools of mosquitoes collected in the medium altitude, moist tropical forest of Perinet (130 km east of Antananarivo). RVFV
strains isolated in Perinet were found to be biologically and antigenically related to the highly pathogenic. Egyptian strains (Saluzzo et al., 1989). Serological investigations in humans (Mathiot et al., 1989) showed a low prevalence of RVFV antibodies in Madagascar. Neither an epizootic nor an epidemic outbreak had been known in Madagascar and RVFV was not con~idered a public health problem. In March 1990, during a 2-week period, an unusually high rate of abortion in zebu cattle (25/150 pregnant females) was reported by the Veterinary Service of the Ministry of Animal Production in the district of Fenerive-est (100 km
Submitted June 3, 1991, accepted November 12, 1991. Correspondingauthor: JacquesMorvan,Unit6de Recherchesur les Arbovirus,InstitutPasteur, BP 1274,Tananarive(Madagascar).
476
J. AIOR V A N E T AL.
north of Tamatave, on the east coast). Hospital recolds in this area showed that in April several patients were hospitalized with a febrile illness of unknown origin. No haemorrhagic syndrome was reported, but a 28-year-old man died. The epizootic outbreak was identified by serological techniques (Morvan et al., 1991). This report presents detailed results of the virological, serological and entomological investigations.
MATERIALS
AND METHODS
Stud.v area The study area covered several villages located in the Fenerive district (fig. 1) on the eastern coastal plain (at an altitude of less than 50 m). The climate is tropical, ~umid with a high rainfall (2,6003,200 ram/year) and no dry month. The mean annual temperature is 24°C. The vegetation consists of small residual patches of primary rain forest in a vast deforested area of cultivated fields (rice, cloves).
Samples Investigations were conducted by the Pasteur Institute staff during the course of two field trips (mission 1, May 2-6; mission 2, May 26-June 2) in Fenerive area villages where the Fenerive Veterinary Service reported a high incidence of abo~ions in zebu cattle in state farms (Vohilengo 13/50 pregnant females, Namantoana 8/25, Itendro 4/75), Fenerive city, and surroundings villages (fig. 1). A total of 40 abortions were reported in state farms and samples of livestock sera were collected at these farms. Human sera were taken from among an apparently healthy population and from hospital admissions. Human blood was taken by venipuncture using 10-ml vacutainer (Becton Dickinson, France), and animals were bled from the jugular vein. Serum was centrifuged and kept in liquid nitrogen until returned to the laboratory, then stored at - 80°C until tested. Mosquitoes were caught on human bait and with light traps placed in cattle enclosures. After species determination, monospecific pools were stored in liquid nitrogen.
ELISA IFA OD PBS
= = = =
enzyme-linked immunosorbent assay. immunofluorescence assay. optical density. phosphate-buffered saline.
Serological techniques lmmunofluorescence assay (IFA) Sera, diluted 1/16 in phosphate-buffe~'ed saline (PBS) were screened by indirect IFA on slides coated with a mixture of inactivated RVF infected and uninfected Vero E6 cells using fluoresceine-labelled anti-human or anti-bovine IgG (Diagnostic Pasteur, France). Positive sera were titrated using two-fold dilutions. R VFV-specific l g M test Serum were tested for IgM-specific RVFV antibody by an ELISA-IgM capture assay (Digoutte et al., 1989). Microtitre plates (Immulon 2, Dynatech Laboratories, VA, USA) were coated by incubation at 37°C for 2 h with goat anti-human IgM (pt-chainspecific) or goat anti-bovine IgM antiserum (Kirkegaard and Perry Laboratories, MD, USA). One hundred microlitres of each sera diluted 1/400 in PBS pH 7.4 containing 1 070 skimmed milk and 0.5 070 Tween-20 were added to the plate and incubated at 37°C for 1 h. After washing, 100 ~l of RVFV sucrose-acetone extracted mouse liver antigen diluted 1/500 in the same diluent was added and incubated overnight at 4°C. RVF hyperimmune mouse ascitic fluid was diluted l/l,000 in the diluent and 100 ~tl added and incubated at 37°C for 1 h. Following incubation and washing, goat anti-mouse horseradish peroxidase conjugate (Kirkegaard and Perry) and o-tolidine (Sigma) we~'e employed to reveal positive reactions. Results were determined by reading the optical density (OD) at 450 nm on a "Titertek multiskan" spectrophotometer (Flow Laboratories, VA, USA). Sera were coas:.dered to be positive if their OD was greater than the mean background of a panel of animals known to be RVFV-antibody-negative plus 3 standard deviations (SD).
Plaque-reduction neutralization test (PRNT) Confirmatory neutralization tests were 9erformed for IFA- and IgM-positive sera and selected negative sera by the technique previously described (Early et al., 1967). An 80 °70 reduction in plaques was used as the index for virus neutralization.
PRNT RVFV SD
= = =
plaque reduction neutralization test. Rift Valley fever virus. standard deviation.
477
RIFT VALLEY FEVER IN MADAGASCAR
..~r0 ' ~
Tampon'/
FENEBIVE '~
FENERIVE A,NTANANIAR~IVO •
Mananara Tamatave ®
'q
Mahambo I ~
City,Villages
~ /
Commomiaz~_u l_ m~ra!ic~..s
Villageswithabo,'tions incattlezebus Fig. !. Map of the locations where serum was sampled.
Virus isolation and antigen detection Sera and mosquito pools were inoculated onto tissue culture (Vero E6 ceils and A e d e s pseudoscutellaris cells) and intracerebrally into l-2-day-old suckling mice by techniques described by Digoutte et al., 1989. This study was completed by an ELISA antigen detection method in mosquito pools using IgM capture (Digoutte c: a./., 1989; Meegan et ai., 1989). Microtitre plates (Immulon 2) were coated overnight at 4°C with an goat anti-human IgM (Kirkegaard and Perry Lab.). Plates were washed and 100 tzl of human lgM anti-RVF serum diluted 1/200 in P B S / T w e e n ° 2 0 / s k i m m e d milk was added and incubated at 37°C for 1 h. Then mosquito supernatant fluids from sonicated mosquito monospecific pools diluted 1/2 in the same diluent were added and incubated at 37°C for 3 h. RVFV hyperimmune mouse
ascitic fluid diluted 1/1,000 was incubated at 37°C for 1 h. After incubation and washing, goat antimouse horseradish peroxidase conjugate (Kirkegaard and Perry) diluted 1/4,000 was added. Plates were incubated, washed, and then substrate (o-tolidine) was added. Results were determined by reading at 450 nm. Specimens were considered to be positive if the OD was greater than the mean value of a panel of known negative pools plus 3 SD.
RESULTS Serological results In this study, the proportion of h u m a n sera (including 129 subjects bled during the first week o f May, and 345 bled during the second trip)
J. M O R V A N E T AL.
478
Table I. Repartition of human sera by place.
Place
IgM
Zone
No.
IFA +
°70
No.
+
%
1 1 1 2 3 2
18 88 92 91 100 85
5 7 10 9 3 4
27.7 7.9 10.8 9.8 3.0 4.7
14 87 88 91 96 84
4 3 6 8 3 1
28.5 3.4 6.8 8.8 3.1 1.2
Fenerive hospital Fenerive city Itendro Namantoana Tarnpolo Mahambo
Table II. RVFV antibody pzevalence in humans according to sex.
Antibody Males
Females
Place
IFA
lgM
IFA
IgM
Fenedve city Itendro Namantoana ~*) Tampolo Mahambo
3/39 4/51 8/44 2/46 3/37
2/39 3/50 7/44 2/44 1/37
3/47 6/41 1/47 1/54 i/48
1/46 3/38 1/47 1/52 0/47
Results shown as no. positive/total samples. (*) p < (L01 (Fisher's).
Table III. Age-specific prevalence of RVFV
antibodies in humans. Antibody Age < 20 20-29 30-39 40-49 > 50
IFA
IgM
7/93 14/152 5/79 4/55 3/77
3/89 10/150 3/77 2/55 3/75
Results shown as no. positive/total samples.
having RVFV-specific IFA antibody was 38/474 (8.01 070).When tested by IgM capture ELISA, 25/460 (5.4 070)were positive. Among patients hospitalized in Fenerive hospital for febrile illness during the same period, 5/18 (27.7 070)were positive by IFA and 4/14 (28.5 070) were positive by IgM.ELISA. Before his death, one patient was bled; he was positive by IFA (1/400) and by IgM-ELISA. The systemic investigation in Fenerive's surrounding villages (table I)
showed that 33/456 (7.18 °70) human sera were positive by IFA and that 21/446 (4.7 070) had IGM-specific antibodies. No significant difference between sexes was observed (table II), except in Namantoana village where cattle abortions were also more numerous. In this village, RVFV antibodies were more frequep.t in males (19 < 0.01). Repartition of RVFV antibodies according to age group is presented in table ill, in the 20-29-year-old age group, RVFV lgM antibodies were more frequent in males (male 8/65, female 2/85, p < 0.01). In livestock, of the 121 sera tested, RVFVspecific antibodies were detected by IFA in 71 sera (58.6 °70). Among these sera, 35/118 (29.6 070)had lgM antibodies present. Repartition by location is presented in table IV. The rate of positivity was greater (p < 1.10-5) among the cows who had aborted (table IV). RVFV antibodies were present in every age group (table V). PRNT was used to control 82 positive sera and confirmed the specificity of the results.
RIFT VALLEY FEVER IN MADAGASCAR
479
Table IV. RVFV antibody and IgM prevalence in zebu cattle sampled in different villages.
Cows having aborted Pregnant cows Non-pregnant cows Bulls Vohilengo (*) Fenerive Itendro Tampolo Namantoana Mahambo Vavatenina (*) Out of zone (*)
No.
IFA +
%
No.
IgM +
%
40 20 27 34 15 13 13 27 26 17 5 5
36 10 13 12 13 6 7 13 15 8 4 5
90.0 50.0 48.1 35.3 86.6 46.1 53.8 48.1 57.6 47.0 80.0 100.0
40 18 27 33 15 13 13 27 23 17 5 5
23 3 3 6 8 3 3 3 8 3 2 5
57.5 16.6 11.1 18.2 53.3 23.0 23.0 11.1 34.7 17.6 40.0 100.0
(*) Only from cows having aborted.
Table V. Age-specific prevalence of RVFV antibodies in cattle. IFA
IgM
Age
+
%
+
%
< 3 3-4 5-6 7
10/17 24/44 20/32 17/28
58.8 54.5 62.5 60.8
8/16 12/43 8/32 7/27
50.0 27.9 25.0 25.9
Results shown as no. positive/total samples.
Entomological results During the two field trips, adult mosquitoes were collected (11,234 mosquitoes) and 408 pools were constituted (table VI). Potential RVFV vectors (Fontenille, 1989) were present: Anopheles c o u s t a n i (30.4 °70), M a n s o n i a u n i f o r m i s (14.9 °70), and Culex antennatus (4.2 °70). S o m e RVFV vectors in subsaharian Africa (Meegan and Bailey, 1988) were also collected: Eretmapodites quinquevittatus (11 °70) and Aedes circumluteolus (0.9 °70).
Virological results A total of 180 h u m a n a n d zebu sera, a n d 408 mosquito pools were inoculated into two cel-
Table VI. Number of mosquitoes caught and mosquito pools. Species
Total No. pools
Anopheles coustani A. squamosus A. gambiae A. mascarensis A. funestus A. flavicosta A. maculipalpis A. grassei A. brunnipes Anopheles sp. pool
3,424 330 544 1,369 39 661 70 76 22 87
105 14 21 50 3 21 4 2 1 5
Culex antennatus C. tritaeniorhynchus C. bitaeniorhynchus C. decens C. quinquefasciatus Culex sp. pool
478 58 24 141 60 45
21 4 2 8 4 3
A edes circumluteolus A. argenteopunctatus A. albopictus A. phillipi A. albodorsalis Aedes sp. pool
102 370 137 37 36 28
7 15 6 2 2 4 61 34 1 8
Mansonia uniformis Eretmapodites quinquevittctus Mansonia + Coquilletidia Culicidae sp. pool
1,683 1,243 12 158
J. M O R V A N E T AL.
480
lular systems and into suckling mice. No virus was isolated from these samples as all results were negative by the antigen detection ELISA.
DISCUSSION
Although RVFV was not isolated from man, animals or mosquitoes during this outbreak, the reports of disease and the cases of abortion among the cattle and our serological data clearly confirm that RVFV circulation occurred in the east coast of Madagascar during the first half of 1990. Among livestock of the Fenerive district, a high antibody prevalence rate existed, as was observed during the West African epizootic outbreak (Akakpo et al., 1989). High IFA antibody prevalence among of zebu cattle of all ages was noted, as it was during the Egyptian epizootic in 1977-1978 (Meegan, 1979). The high prevalence of specific IgM antibodies observed among all groups sampled reveals recent active virus circulation. The first abortion cases were observed by the veterinary services towards the end of February, confirming IgM serological data. Observations from this outbreak and from a previous published report (Gonzalez et al., 19~9) suggest that the IgM antibodies are likely ~;o ~.uaauIJK;~l,1 I U l ~ ; a a tlli:l.ll 3 l l l O l l t l l ~ .
1111~ l l H l l l d I l
attack was light, but serological investigations have shown that the prevalence was higher than in other regions of Madagascar (Mathiot et al., 1989; Morvan, 1990). The higher proportioo of seropositivity in animals rather than in humans suggests that the cattle may have played an amplifying role. Two epidemiological explanations of RVFV circulation in Africa were proposed. First, RVFV present in the traditional foci, a tropical forest cycle and a cycle involving viraemic animals and mosquitoes, is maintained during the dry season by transovarial transmission of Aedes mosquitoes that breed in temporary ground pools; then Aedes eggs become flooded during the rainy season (Davies et al., 1985; Linthicum et al., 1985). Large rainfall would thus be necessary to hatch a critical mass of transovarially infected Aedes mosquitoes. Vector pullulation transmits RVFV to animals which develop high
viraemia and may act as amplifying hosts. Virus activity expresses itself in a cyclic mode. This mechanism could explain the virus activity and the first isolates encountered in Perinet in 1979. The circumstances of the described epizootic in Fenerive are different, the climate is humid tropical without any dry months. Recent meteorological date are unknown, but rainfall was said to be lower than during preceding years according to information obtained in the field. Entomological data have not led to abundant harvesting of Aedes. On the other hand, other potential vectors (A. coustani, M. uniformis. E. quinquevittatus, C. antennatus) were numerous. -rhooo mosquito species are oh,,,~n~n~ all year round (Fontenille, unpublished data collected in Mananara on the coastal plain, 170 km north of Fenerive). Laboratory transmission experiments (Macintosh and Jupp, 1981 ; Gad et al., 1987; Saluzzo, 1989) to assess the vector competence of different mosquitoe species showed the particular efficiency of C. antennatus, a mosquito with a long lifespan. RVFV was not isolated during this episode, unlike previous reports where Aedes was suspected to have initiated the outbreak (Macintosh et al., 1983 ; Logan et al., 1991). The existence of certain factors retarding transmission, such as an extrinsic incubation period (Macintosh et al., 1983), is possible. Only a longitudinal en~,omological survey could indicate the extent of vectorial involvement in this epizootic outbreak. Another conception of RVFV epidemiology is based on the notion of permanent foci of RVFV circulation from which virus diffuses by successive stages to other areas by herd migrations, as was suggested during the Egyptian outbreak of 1977-1978 (Gad et al., 1986). Introduction and propagation of virus in an area of non-immune livestock would generate animals (sheep or cattle) with high level viraemia which would act as source of infection for even inefficient mosquito vectors, thus sustaining the epizootic (Peters and LeDuc, 1984; Saluzzo et ai., 1987). Such important foci were not observed in Madagascar by serological investigation, but low level RVFV circulation was confirmed in several regions (Fontenille, 1989;
481
RIFT VALLEY FEVER IN MADAGASCAR
Mathiot et aL, 1989). In a recent investigation, 9/156 (5.7 °7o) human sera sampled in November 1989 in Mananara (Morvan, unpublished data) were found to be positive for RVFV antibodies by IFA, but without RVFV IgM. In Madagascar, cattle commerce involves herd migration from western regions to abattoirs. Migration also occurs on the east coast, and the Fenerive area is traversed through the coastal plain to the north (Mananara) or to the west (Vavatenina) through the Central and North Highlands (fig. 1). RVFV could be introduced by cattle from an endemic foci situated in the western breeding regions. No animals had been imported from Africa according to the Ministry o f Animal Production. H u m a n contamination may be explained by direct transmission from sick animals, since the virus persists in animal spleens and may diffuse during slaughter tv,~.~....,o..h..;., . . . . . ,.,,o,,..... ~ et al., 1981), or by vectorial transmission. Understanding the mechanism of transmis= sion is important for prevention strategies. Several methods have been proposed: in the case of cychc viral activity (as in East Africa), rainfall surveys (Davies e ' al., 1985) are used and animal vaccination camlr~figns started if rainfall is higher than n o r m a l ; if tae epidemiological situation is similar to West Africa with important herd migrations, serological and entomological surveys of men and cattle, associated with vaccination of migratory livestock are used (Saluzzo, 1989). E p i d e m i o l o g i c a l c i r c u m s t a n c e s o f this epizootic outbreak in the Fenerive region are still imprecise. A longitudinal serole,,ical survey in humans and cattle, with entomological investigations for viral isolation will be undertaken to try to understand the mechanisms involved in the maintenance and emergence of RVFV in Madagascar.
Fii~vre de la Vall6e du Rift sur la c6te est de Madagascar
En mars 1990, une 6pizootie de fi6vre de la Vallde du Rift fut suspect6e dans le district de F6nerive, sur la c6te est de Madagascar, du fait d'une incidence
anormale, parmi le b6tail domestique, d'avortements et de maladies. Des s6rums humains et bovins ont 6t6 test6s par immunofluorescence indirecte (IFA) et capture immunoenzymatique d'IgM (ELISA-IgM) visb.-vis du virus de la fi~vre de la Vall6e du Rift. Des tentatives d'isolement de virus par inoculation de cultures de cellules et de souriceaux nouveau-n6s ainsi qu'un test de d6tection d'antig~ne viral ont 6t6 pcatiqu6s sur les s6rums et sur des moustiques collect6s dans la zone affect6e. Parmi les s6rums test6s, la pr6valence en anticorps sp6cifiques est respectivement de 58,6 et 29,6 °7o par IFA et ELISA-IgM pour les bovins, et de 8,01 070 (IFA) et 5,4 07o (ELISA-IgM) chez l'homme. Ni virus ni antigone n'ont pu ~tre iso16s ou d6tect6s dans les s6rums ou les pools de moustiques test6s. Plusieurs hypoth6ses sont discut6es sur l'6mergence et la diffusion de la fi~vre de la Vall6e du Rift dans cette r6gion de Madagascar. Mots-cl~s: Fi~vre de la Vall6e du Rift, 1V ~.dagas-
car; Bovins, Anticorps.
Homme, Moustiques
vecteurs,
References
Akakpo, A.J., Some, M.J.R., Bornarel, P., Jouan, A. & Gonzalez, J.P. (1989)• Epid6miologie de la fi6vre de la Vall6edu Rift en Afrique de l'Ouest. -- I. Enqu~te s6rologique chez les ruminants domestiques au Burkina Faso. Bull. Soc. Path. exot., 82, 321-331. C. l. e. r c. .
.
.
Y__• R n d h a i n
. . . . . . . . . . .
•
17 * , ~I 'i ~ i g ur u. ,L tl t~, .~ , i
I D • , / ]All.,; k l l L . . . r l ~ l l d ~.l . ~., ., .
Jem.
R.
~"
Coulanges, P. (1982), Le programme exploratoire arbovirus de l'lnstitut Pasteur de Madagascar" bilan 1976-1980. Arch. inst. Pasteur Madagascar, 49, 65-77. Daubney, R., Hudson, J.R. & Garnham, P.C. (1931), Enzootic hepatitis of Rift Valley fever: an undescribed virus disease of sheep, cattle and man from East Africa. J. Path. Bact., 31,546-579. Davies, F.G., Linthicum, K.J. & James, A.D. (1985), Rainfall and epizootic Rift Valley fever. Bull. Org. mond. Santd, 63, 941-943. Digoutte, J.P., Jouan, A., Le Guenno, B., Riou, O., Philippe, B., Meegan, J.M., Ksiazek, T.G. & Peters, C.J. (1989), Isolation of the Rift Valley fever virus by inoculation into Aedes pseudoscutellans cells: comparison with other diagnostic methods. Res. Virol., 140, 31-41. Early, E., Peralta, P.H. & Johnson, K.M. (1967), A plaque reduction method for arboviruses. Proc. Soc. exp. Biol. (N.Y.), 125, 741-747. Fontenille, D. (1989), Etude des circuits de vecteurs d'arbovirus :~ Madagascar. Arch. inst. Pasteur Madagascar• 50, 11-315. Gad, A.M., Feinsod, F.M., Ailam, I.H., Eisa, M., Hassan, A.N., Soliman, B.A., El Said, S. & Saah, A.J. (1986), A possible route for the introduction of Rift Valley fever virus into Egypt during 1977. J. trop. Med. Hyg., 89, 233-236.
482
J. M O R V A N E T A L .
Gad, A.M., Hassan, M.M., El Said, S., Moussa, M.I. & Wood, O.L. (1987), Rift Valley fever virus transmission by different Egyptian mosquito species. Trans. roy. Soc. trop. Med. Hyg., 81,694-698. Gonzalez, J.P., Guillaud, M. & Wilson, M.L. (1989), Rift Valley fever, in "Rapport Annuel 1988". WHO Collaborating Center for Reference and Research for Arboviruses. Institut Pasteur, Dakar, S6n6gal. Linthicum, K.J., Davies, F.G., Kairo, A. & Bailey, C.H. (1985), Rift Valley fever virus (family Bunyaviridae, genus Phlebovirus). Isolations from diptera collected during an inter-epizootic period in Kenya. J. Hyg. (Camb.), 95, 197-209. Logan, T.M., Linthicum, K.J., Davies, F.G., Binepal, Y.S. & Roberts, C.R. 0991), Isolation of Rift Valley fever virus from mosquitoes (Diptera: Culicidae) collected during an outbreak in domestic animals in Kenya. J. Med. Entomol , 28, 293-295. Macintosh, B.M. & Jupp, P.C. (1981), Epidemiological aspects of Rift Valley fever in South Africa with reference to vectors, in "_Rift Valley fever" (T.A. Swartz, M.A. Klingberg & N. Goldblum) (pp. 92-99). Contr. Epidem. Biostatist., 3. S. Karger, Basel. Maclntosh, B.M., Jupp, P.G., Dos Santos, I. & Rowe, A.C. (1983), Field and laboratory evidence implicating Culex zombaesis and Aedes circumluteolus as vectors of Rift Valley fever virus in coastal South Africa. S. Afric. J. Sci., 79, 61-64. Mathiot, C., Fontenille, D., Georges, A.J. & Coulanges, P. (1989), Antibodies to baemorrhagic fever viruses in Madagascar. Trans. roy. Soc. top. Med. Hyg., 83, 407-409. Meegan, J.M. (1979), The Rift Vailej fever epizootic in Egypt 1977-1978. -- 1. Description of the epizootic and virological studies. Trans. roy. Soc. trop. Aled. Hyg., 73, 618-623. Meegan, J.M. & Bailey, C.H. (1988), Rift Valley fever, in "The arboviruses: epidemiology and ecology"
(T.P. Monath) (pp. 51-76). CRC Boca Raton, Florida. Meegan, J.M., Le Guenno, B., Ksiazek, T.G., Jouan, A., Knauert, F., Digoutte, J.P. & Peters, C.J. (1989), Rapid diagnosis of Rift Valley fever: a comparison of methods for the detection of viral antigen in human sera. Res. Virol., 140, 59-65. Morvan, J. (1990), in "Rapport annuel 1989". WHO Collaborating Center of Reference and Research for Arboviruses. Institut Pasteur, Dakar, S6n~gal. Morvan, J., Saluzzo, J.F., Fontenille, D. & Coulanges, P. (1991), Possible Rift Valley fever outbreak in man and cattle in Madagascar. Trans. roy. Soc. trop. Med. Hyg., 85, 108. Peters, C.J. & LeDuc, J.W. (1984), Bunyaviruses, phleboviruses and related viruses, in "Textbook of human virology" (R.B. Belshe) (pp. 547-598). PSG Pub!. Co. Ira.., Littleton, Massachusetts. Saluzzo, J.F., Chartier, C., Bada, R., Martinez, D. & Digoutte, J.P. (1987), La fi~vre de la valise du Rift en Afrique de l'Ouest. Rev. t~iev. M~d. v~t. Pays trop., 40, 215-223. Saluzzo, J.F. (1989), l~pid6miologie mol~culaire, g~n6tique et pouvoir pathog6ne du virus de la fi~vre de la Vall6e du Rift: application/t l'6valuation d'un vaccin vivant att~nu~ /t usage v6t~rinaire. Th/~se, Universit6 de Clermont-Ferrand. Saluzzo, J.F., Anderson, G.W., Smith, J.F., Fontenille, D. & Coulanges, P. (1989), Biological and antigenic relationship between Rift Valley fever virus strains isolated in Egypt and Madagascar. Trans. roy. Soc. trop. Med. Hyg., 83, 701. Yedloutschnig, R.J., Dardiri, A.H. & Walker, J.S. (1981), Persistence of Rift Valley fever virus in the spleen, liver and brain of sheep after experimental infection, in "Rift Valley fever" (T.A. Swartz, M.A. Klingberg & N. Goldblum) (pp. 72-76). Contr. Epidem. Biostatist., 3. S. Karger, Basel.
Res. Virol.
1991, 142, 475-482
Rift Valley fever on the east coast of Madagascar J. Morvan (1), J.-F. Saluzzo (z), D. Fontenille (3), P.E. Rollin (2) and P. Coulanges (l) it) Unitd de Recherche sur les Arbovirus, Institut Pasteur, Antananarivo (Madagascar), ~2) Laboratoire des Fibvres hdmorragiques virales, Institut Pasteur, Paris, and (3) Laboratoire d'Entomologie mddicale, Institut Pasteur, Antananarivo
SUMMARY In March 1990, a Rift Valley fever virus (RVFV) outbreak was suspected in the district of Fenerive on the east coast of Madagascar after an abnormally high incidence of abortions and disease in livest(,ck. Sera from humans and cattle were tested for RVFV antibodies by immunofluorescence assay (IFA) and ELISA-IgM capture. Sma and mosquitoes collected in the same area were tested for virus isolation by tissue culture and suckling mouse intracerebral inoculation, and for antigen detection by an ELIZA antigen capture assay. Among cattle from the area, RVFV antibody prevalence was 58.6 % by IFA and 29.6 % by ELISA-IgM. In contrast, human populations in the same area had a lower RVFV antibody prevalence, with 8 . 0 1 % IFA and 5.4 % IgM-positive sera. No RVFV antigen was detected and virus isolation was unsuc~:essf, I from the sera and mosquito pools tested. Different hypotheses concerning the emergence and diffusion of RVFV in this area and the occurrence of the outbreak are discussed.
Key-words: RVF, Madagascar = Cattle, Man, Mosquito vectors, Antibodies.
INTRODUCTION Rift Valley fever virus (RVFV) was first isolated in 1930 in Kenya during an occurrence of abortions in sheep and cattle, and febrile illness in humans (Daubney et aL, 1931). In the decades which followed, successive outbreaks of zoonotic disease were reported in sub-Saharan Africa. In 1977-1978, an extensive outbreak was reported in Egypt (Meegan, 1979). In Madagascar, RVFV was isolated in 1979 (Clerc et al., 1982) from several pools of mosquitoes collected in the medium altitude, moist tropical forest of Perinet (130 km east of Antananarivo). RVFV
strains isolated in Perinet were found to be biologically and antigenically related to the highly pathogenic. Egyptian strains (Saluzzo et al., 1989). Serological investigations in humans (Mathiot et al., 1989) showed a low prevalence of RVFV antibodies in Madagascar. Neither an epizootic nor an epidemic outbreak had been known in Madagascar and RVFV was not con~idered a public health problem. In March 1990, during a 2-week period, an unusually high rate of abortion in zebu cattle (25/150 pregnant females) was reported by the Veterinary Service of the Ministry of Animal Production in the district of Fenerive-est (100 km
Submitted June 3, 1991, accepted November 12, 1991. Correspondingauthor: JacquesMorvan,Unit6de Recherchesur les Arbovirus,InstitutPasteur, BP 1274,Tananarive(Madagascar).
476
J. AIOR V A N E T AL.
north of Tamatave, on the east coast). Hospital recolds in this area showed that in April several patients were hospitalized with a febrile illness of unknown origin. No haemorrhagic syndrome was reported, but a 28-year-old man died. The epizootic outbreak was identified by serological techniques (Morvan et al., 1991). This report presents detailed results of the virological, serological and entomological investigations.
MATERIALS
AND METHODS
Stud.v area The study area covered several villages located in the Fenerive district (fig. 1) on the eastern coastal plain (at an altitude of less than 50 m). The climate is tropical, ~umid with a high rainfall (2,6003,200 ram/year) and no dry month. The mean annual temperature is 24°C. The vegetation consists of small residual patches of primary rain forest in a vast deforested area of cultivated fields (rice, cloves).
Samples Investigations were conducted by the Pasteur Institute staff during the course of two field trips (mission 1, May 2-6; mission 2, May 26-June 2) in Fenerive area villages where the Fenerive Veterinary Service reported a high incidence of abo~ions in zebu cattle in state farms (Vohilengo 13/50 pregnant females, Namantoana 8/25, Itendro 4/75), Fenerive city, and surroundings villages (fig. 1). A total of 40 abortions were reported in state farms and samples of livestock sera were collected at these farms. Human sera were taken from among an apparently healthy population and from hospital admissions. Human blood was taken by venipuncture using 10-ml vacutainer (Becton Dickinson, France), and animals were bled from the jugular vein. Serum was centrifuged and kept in liquid nitrogen until returned to the laboratory, then stored at - 80°C until tested. Mosquitoes were caught on human bait and with light traps placed in cattle enclosures. After species determination, monospecific pools were stored in liquid nitrogen.
ELISA IFA OD PBS
= = = =
enzyme-linked immunosorbent assay. immunofluorescence assay. optical density. phosphate-buffered saline.
Serological techniques lmmunofluorescence assay (IFA) Sera, diluted 1/16 in phosphate-buffe~'ed saline (PBS) were screened by indirect IFA on slides coated with a mixture of inactivated RVF infected and uninfected Vero E6 cells using fluoresceine-labelled anti-human or anti-bovine IgG (Diagnostic Pasteur, France). Positive sera were titrated using two-fold dilutions. R VFV-specific l g M test Serum were tested for IgM-specific RVFV antibody by an ELISA-IgM capture assay (Digoutte et al., 1989). Microtitre plates (Immulon 2, Dynatech Laboratories, VA, USA) were coated by incubation at 37°C for 2 h with goat anti-human IgM (pt-chainspecific) or goat anti-bovine IgM antiserum (Kirkegaard and Perry Laboratories, MD, USA). One hundred microlitres of each sera diluted 1/400 in PBS pH 7.4 containing 1 070 skimmed milk and 0.5 070 Tween-20 were added to the plate and incubated at 37°C for 1 h. After washing, 100 ~l of RVFV sucrose-acetone extracted mouse liver antigen diluted 1/500 in the same diluent was added and incubated overnight at 4°C. RVF hyperimmune mouse ascitic fluid was diluted l/l,000 in the diluent and 100 ~tl added and incubated at 37°C for 1 h. Following incubation and washing, goat anti-mouse horseradish peroxidase conjugate (Kirkegaard and Perry) and o-tolidine (Sigma) we~'e employed to reveal positive reactions. Results were determined by reading the optical density (OD) at 450 nm on a "Titertek multiskan" spectrophotometer (Flow Laboratories, VA, USA). Sera were coas:.dered to be positive if their OD was greater than the mean background of a panel of animals known to be RVFV-antibody-negative plus 3 standard deviations (SD).
Plaque-reduction neutralization test (PRNT) Confirmatory neutralization tests were 9erformed for IFA- and IgM-positive sera and selected negative sera by the technique previously described (Early et al., 1967). An 80 °70 reduction in plaques was used as the index for virus neutralization.
PRNT RVFV SD
= = =
plaque reduction neutralization test. Rift Valley fever virus. standard deviation.
477
RIFT VALLEY FEVER IN MADAGASCAR
..~r0 ' ~
Tampon'/
FENEBIVE '~
FENERIVE A,NTANANIAR~IVO •
Mananara Tamatave ®
'q
Mahambo I ~
City,Villages
~ /
Commomiaz~_u l_ m~ra!ic~..s
Villageswithabo,'tions incattlezebus Fig. !. Map of the locations where serum was sampled.
Virus isolation and antigen detection Sera and mosquito pools were inoculated onto tissue culture (Vero E6 ceils and A e d e s pseudoscutellaris cells) and intracerebrally into l-2-day-old suckling mice by techniques described by Digoutte et al., 1989. This study was completed by an ELISA antigen detection method in mosquito pools using IgM capture (Digoutte c: a./., 1989; Meegan et ai., 1989). Microtitre plates (Immulon 2) were coated overnight at 4°C with an goat anti-human IgM (Kirkegaard and Perry Lab.). Plates were washed and 100 tzl of human lgM anti-RVF serum diluted 1/200 in P B S / T w e e n ° 2 0 / s k i m m e d milk was added and incubated at 37°C for 1 h. Then mosquito supernatant fluids from sonicated mosquito monospecific pools diluted 1/2 in the same diluent were added and incubated at 37°C for 3 h. RVFV hyperimmune mouse
ascitic fluid diluted 1/1,000 was incubated at 37°C for 1 h. After incubation and washing, goat antimouse horseradish peroxidase conjugate (Kirkegaard and Perry) diluted 1/4,000 was added. Plates were incubated, washed, and then substrate (o-tolidine) was added. Results were determined by reading at 450 nm. Specimens were considered to be positive if the OD was greater than the mean value of a panel of known negative pools plus 3 SD.
RESULTS Serological results In this study, the proportion of h u m a n sera (including 129 subjects bled during the first week o f May, and 345 bled during the second trip)
J. M O R V A N E T AL.
478
Table I. Repartition of human sera by place.
Place
IgM
Zone
No.
IFA +
°70
No.
+
%
1 1 1 2 3 2
18 88 92 91 100 85
5 7 10 9 3 4
27.7 7.9 10.8 9.8 3.0 4.7
14 87 88 91 96 84
4 3 6 8 3 1
28.5 3.4 6.8 8.8 3.1 1.2
Fenerive hospital Fenerive city Itendro Namantoana Tarnpolo Mahambo
Table II. RVFV antibody pzevalence in humans according to sex.
Antibody Males
Females
Place
IFA
lgM
IFA
IgM
Fenedve city Itendro Namantoana ~*) Tampolo Mahambo
3/39 4/51 8/44 2/46 3/37
2/39 3/50 7/44 2/44 1/37
3/47 6/41 1/47 1/54 i/48
1/46 3/38 1/47 1/52 0/47
Results shown as no. positive/total samples. (*) p < (L01 (Fisher's).
Table III. Age-specific prevalence of RVFV
antibodies in humans. Antibody Age < 20 20-29 30-39 40-49 > 50
IFA
IgM
7/93 14/152 5/79 4/55 3/77
3/89 10/150 3/77 2/55 3/75
Results shown as no. positive/total samples.
having RVFV-specific IFA antibody was 38/474 (8.01 070).When tested by IgM capture ELISA, 25/460 (5.4 070)were positive. Among patients hospitalized in Fenerive hospital for febrile illness during the same period, 5/18 (27.7 070)were positive by IFA and 4/14 (28.5 070) were positive by IgM.ELISA. Before his death, one patient was bled; he was positive by IFA (1/400) and by IgM-ELISA. The systemic investigation in Fenerive's surrounding villages (table I)
showed that 33/456 (7.18 °70) human sera were positive by IFA and that 21/446 (4.7 070) had IGM-specific antibodies. No significant difference between sexes was observed (table II), except in Namantoana village where cattle abortions were also more numerous. In this village, RVFV antibodies were more frequep.t in males (19 < 0.01). Repartition of RVFV antibodies according to age group is presented in table ill, in the 20-29-year-old age group, RVFV lgM antibodies were more frequent in males (male 8/65, female 2/85, p < 0.01). In livestock, of the 121 sera tested, RVFVspecific antibodies were detected by IFA in 71 sera (58.6 °70). Among these sera, 35/118 (29.6 070)had lgM antibodies present. Repartition by location is presented in table IV. The rate of positivity was greater (p < 1.10-5) among the cows who had aborted (table IV). RVFV antibodies were present in every age group (table V). PRNT was used to control 82 positive sera and confirmed the specificity of the results.
RIFT VALLEY FEVER IN MADAGASCAR
479
Table IV. RVFV antibody and IgM prevalence in zebu cattle sampled in different villages.
Cows having aborted Pregnant cows Non-pregnant cows Bulls Vohilengo (*) Fenerive Itendro Tampolo Namantoana Mahambo Vavatenina (*) Out of zone (*)
No.
IFA +
%
No.
IgM +
%
40 20 27 34 15 13 13 27 26 17 5 5
36 10 13 12 13 6 7 13 15 8 4 5
90.0 50.0 48.1 35.3 86.6 46.1 53.8 48.1 57.6 47.0 80.0 100.0
40 18 27 33 15 13 13 27 23 17 5 5
23 3 3 6 8 3 3 3 8 3 2 5
57.5 16.6 11.1 18.2 53.3 23.0 23.0 11.1 34.7 17.6 40.0 100.0
(*) Only from cows having aborted.
Table V. Age-specific prevalence of RVFV antibodies in cattle. IFA
IgM
Age
+
%
+
%
< 3 3-4 5-6 7
10/17 24/44 20/32 17/28
58.8 54.5 62.5 60.8
8/16 12/43 8/32 7/27
50.0 27.9 25.0 25.9
Results shown as no. positive/total samples.
Entomological results During the two field trips, adult mosquitoes were collected (11,234 mosquitoes) and 408 pools were constituted (table VI). Potential RVFV vectors (Fontenille, 1989) were present: Anopheles c o u s t a n i (30.4 °70), M a n s o n i a u n i f o r m i s (14.9 °70), and Culex antennatus (4.2 °70). S o m e RVFV vectors in subsaharian Africa (Meegan and Bailey, 1988) were also collected: Eretmapodites quinquevittatus (11 °70) and Aedes circumluteolus (0.9 °70).
Virological results A total of 180 h u m a n a n d zebu sera, a n d 408 mosquito pools were inoculated into two cel-
Table VI. Number of mosquitoes caught and mosquito pools. Species
Total No. pools
Anopheles coustani A. squamosus A. gambiae A. mascarensis A. funestus A. flavicosta A. maculipalpis A. grassei A. brunnipes Anopheles sp. pool
3,424 330 544 1,369 39 661 70 76 22 87
105 14 21 50 3 21 4 2 1 5
Culex antennatus C. tritaeniorhynchus C. bitaeniorhynchus C. decens C. quinquefasciatus Culex sp. pool
478 58 24 141 60 45
21 4 2 8 4 3
A edes circumluteolus A. argenteopunctatus A. albopictus A. phillipi A. albodorsalis Aedes sp. pool
102 370 137 37 36 28
7 15 6 2 2 4 61 34 1 8
Mansonia uniformis Eretmapodites quinquevittctus Mansonia + Coquilletidia Culicidae sp. pool
1,683 1,243 12 158
J. M O R V A N E T AL.
480
lular systems and into suckling mice. No virus was isolated from these samples as all results were negative by the antigen detection ELISA.
DISCUSSION
Although RVFV was not isolated from man, animals or mosquitoes during this outbreak, the reports of disease and the cases of abortion among the cattle and our serological data clearly confirm that RVFV circulation occurred in the east coast of Madagascar during the first half of 1990. Among livestock of the Fenerive district, a high antibody prevalence rate existed, as was observed during the West African epizootic outbreak (Akakpo et al., 1989). High IFA antibody prevalence among of zebu cattle of all ages was noted, as it was during the Egyptian epizootic in 1977-1978 (Meegan, 1979). The high prevalence of specific IgM antibodies observed among all groups sampled reveals recent active virus circulation. The first abortion cases were observed by the veterinary services towards the end of February, confirming IgM serological data. Observations from this outbreak and from a previous published report (Gonzalez et al., 19~9) suggest that the IgM antibodies are likely ~;o ~.uaauIJK;~l,1 I U l ~ ; a a tlli:l.ll 3 l l l O l l t l l ~ .
1111~ l l H l l l d I l
attack was light, but serological investigations have shown that the prevalence was higher than in other regions of Madagascar (Mathiot et al., 1989; Morvan, 1990). The higher proportioo of seropositivity in animals rather than in humans suggests that the cattle may have played an amplifying role. Two epidemiological explanations of RVFV circulation in Africa were proposed. First, RVFV present in the traditional foci, a tropical forest cycle and a cycle involving viraemic animals and mosquitoes, is maintained during the dry season by transovarial transmission of Aedes mosquitoes that breed in temporary ground pools; then Aedes eggs become flooded during the rainy season (Davies et al., 1985; Linthicum et al., 1985). Large rainfall would thus be necessary to hatch a critical mass of transovarially infected Aedes mosquitoes. Vector pullulation transmits RVFV to animals which develop high
viraemia and may act as amplifying hosts. Virus activity expresses itself in a cyclic mode. This mechanism could explain the virus activity and the first isolates encountered in Perinet in 1979. The circumstances of the described epizootic in Fenerive are different, the climate is humid tropical without any dry months. Recent meteorological date are unknown, but rainfall was said to be lower than during preceding years according to information obtained in the field. Entomological data have not led to abundant harvesting of Aedes. On the other hand, other potential vectors (A. coustani, M. uniformis. E. quinquevittatus, C. antennatus) were numerous. -rhooo mosquito species are oh,,,~n~n~ all year round (Fontenille, unpublished data collected in Mananara on the coastal plain, 170 km north of Fenerive). Laboratory transmission experiments (Macintosh and Jupp, 1981 ; Gad et al., 1987; Saluzzo, 1989) to assess the vector competence of different mosquitoe species showed the particular efficiency of C. antennatus, a mosquito with a long lifespan. RVFV was not isolated during this episode, unlike previous reports where Aedes was suspected to have initiated the outbreak (Macintosh et al., 1983 ; Logan et al., 1991). The existence of certain factors retarding transmission, such as an extrinsic incubation period (Macintosh et al., 1983), is possible. Only a longitudinal en~,omological survey could indicate the extent of vectorial involvement in this epizootic outbreak. Another conception of RVFV epidemiology is based on the notion of permanent foci of RVFV circulation from which virus diffuses by successive stages to other areas by herd migrations, as was suggested during the Egyptian outbreak of 1977-1978 (Gad et al., 1986). Introduction and propagation of virus in an area of non-immune livestock would generate animals (sheep or cattle) with high level viraemia which would act as source of infection for even inefficient mosquito vectors, thus sustaining the epizootic (Peters and LeDuc, 1984; Saluzzo et ai., 1987). Such important foci were not observed in Madagascar by serological investigation, but low level RVFV circulation was confirmed in several regions (Fontenille, 1989;
481
RIFT VALLEY FEVER IN MADAGASCAR
Mathiot et aL, 1989). In a recent investigation, 9/156 (5.7 °7o) human sera sampled in November 1989 in Mananara (Morvan, unpublished data) were found to be positive for RVFV antibodies by IFA, but without RVFV IgM. In Madagascar, cattle commerce involves herd migration from western regions to abattoirs. Migration also occurs on the east coast, and the Fenerive area is traversed through the coastal plain to the north (Mananara) or to the west (Vavatenina) through the Central and North Highlands (fig. 1). RVFV could be introduced by cattle from an endemic foci situated in the western breeding regions. No animals had been imported from Africa according to the Ministry o f Animal Production. H u m a n contamination may be explained by direct transmission from sick animals, since the virus persists in animal spleens and may diffuse during slaughter tv,~.~....,o..h..;., . . . . . ,.,,o,,..... ~ et al., 1981), or by vectorial transmission. Understanding the mechanism of transmis= sion is important for prevention strategies. Several methods have been proposed: in the case of cychc viral activity (as in East Africa), rainfall surveys (Davies e ' al., 1985) are used and animal vaccination camlr~figns started if rainfall is higher than n o r m a l ; if tae epidemiological situation is similar to West Africa with important herd migrations, serological and entomological surveys of men and cattle, associated with vaccination of migratory livestock are used (Saluzzo, 1989). E p i d e m i o l o g i c a l c i r c u m s t a n c e s o f this epizootic outbreak in the Fenerive region are still imprecise. A longitudinal serole,,ical survey in humans and cattle, with entomological investigations for viral isolation will be undertaken to try to understand the mechanisms involved in the maintenance and emergence of RVFV in Madagascar.
Fii~vre de la Vall6e du Rift sur la c6te est de Madagascar
En mars 1990, une 6pizootie de fi6vre de la Vallde du Rift fut suspect6e dans le district de F6nerive, sur la c6te est de Madagascar, du fait d'une incidence
anormale, parmi le b6tail domestique, d'avortements et de maladies. Des s6rums humains et bovins ont 6t6 test6s par immunofluorescence indirecte (IFA) et capture immunoenzymatique d'IgM (ELISA-IgM) visb.-vis du virus de la fi~vre de la Vall6e du Rift. Des tentatives d'isolement de virus par inoculation de cultures de cellules et de souriceaux nouveau-n6s ainsi qu'un test de d6tection d'antig~ne viral ont 6t6 pcatiqu6s sur les s6rums et sur des moustiques collect6s dans la zone affect6e. Parmi les s6rums test6s, la pr6valence en anticorps sp6cifiques est respectivement de 58,6 et 29,6 °7o par IFA et ELISA-IgM pour les bovins, et de 8,01 070 (IFA) et 5,4 07o (ELISA-IgM) chez l'homme. Ni virus ni antigone n'ont pu ~tre iso16s ou d6tect6s dans les s6rums ou les pools de moustiques test6s. Plusieurs hypoth6ses sont discut6es sur l'6mergence et la diffusion de la fi~vre de la Vall6e du Rift dans cette r6gion de Madagascar. Mots-cl~s: Fi~vre de la Vall6e du Rift, 1V ~.dagas-
car; Bovins, Anticorps.
Homme, Moustiques
vecteurs,
References
Akakpo, A.J., Some, M.J.R., Bornarel, P., Jouan, A. & Gonzalez, J.P. (1989)• Epid6miologie de la fi6vre de la Vall6edu Rift en Afrique de l'Ouest. -- I. Enqu~te s6rologique chez les ruminants domestiques au Burkina Faso. Bull. Soc. Path. exot., 82, 321-331. C. l. e. r c. .
.
.
Y__• R n d h a i n
. . . . . . . . . . .
•
17 * , ~I 'i ~ i g ur u. ,L tl t~, .~ , i
I D • , / ]All.,; k l l L . . . r l ~ l l d ~.l . ~., ., .
Jem.
R.
~"
Coulanges, P. (1982), Le programme exploratoire arbovirus de l'lnstitut Pasteur de Madagascar" bilan 1976-1980. Arch. inst. Pasteur Madagascar, 49, 65-77. Daubney, R., Hudson, J.R. & Garnham, P.C. (1931), Enzootic hepatitis of Rift Valley fever: an undescribed virus disease of sheep, cattle and man from East Africa. J. Path. Bact., 31,546-579. Davies, F.G., Linthicum, K.J. & James, A.D. (1985), Rainfall and epizootic Rift Valley fever. Bull. Org. mond. Santd, 63, 941-943. Digoutte, J.P., Jouan, A., Le Guenno, B., Riou, O., Philippe, B., Meegan, J.M., Ksiazek, T.G. & Peters, C.J. (1989), Isolation of the Rift Valley fever virus by inoculation into Aedes pseudoscutellans cells: comparison with other diagnostic methods. Res. Virol., 140, 31-41. Early, E., Peralta, P.H. & Johnson, K.M. (1967), A plaque reduction method for arboviruses. Proc. Soc. exp. Biol. (N.Y.), 125, 741-747. Fontenille, D. (1989), Etude des circuits de vecteurs d'arbovirus :~ Madagascar. Arch. inst. Pasteur Madagascar• 50, 11-315. Gad, A.M., Feinsod, F.M., Ailam, I.H., Eisa, M., Hassan, A.N., Soliman, B.A., El Said, S. & Saah, A.J. (1986), A possible route for the introduction of Rift Valley fever virus into Egypt during 1977. J. trop. Med. Hyg., 89, 233-236.
482
J. M O R V A N E T A L .
Gad, A.M., Hassan, M.M., El Said, S., Moussa, M.I. & Wood, O.L. (1987), Rift Valley fever virus transmission by different Egyptian mosquito species. Trans. roy. Soc. trop. Med. Hyg., 81,694-698. Gonzalez, J.P., Guillaud, M. & Wilson, M.L. (1989), Rift Valley fever, in "Rapport Annuel 1988". WHO Collaborating Center for Reference and Research for Arboviruses. Institut Pasteur, Dakar, S6n6gal. Linthicum, K.J., Davies, F.G., Kairo, A. & Bailey, C.H. (1985), Rift Valley fever virus (family Bunyaviridae, genus Phlebovirus). Isolations from diptera collected during an inter-epizootic period in Kenya. J. Hyg. (Camb.), 95, 197-209. Logan, T.M., Linthicum, K.J., Davies, F.G., Binepal, Y.S. & Roberts, C.R. 0991), Isolation of Rift Valley fever virus from mosquitoes (Diptera: Culicidae) collected during an outbreak in domestic animals in Kenya. J. Med. Entomol , 28, 293-295. Macintosh, B.M. & Jupp, P.C. (1981), Epidemiological aspects of Rift Valley fever in South Africa with reference to vectors, in "_Rift Valley fever" (T.A. Swartz, M.A. Klingberg & N. Goldblum) (pp. 92-99). Contr. Epidem. Biostatist., 3. S. Karger, Basel. Maclntosh, B.M., Jupp, P.G., Dos Santos, I. & Rowe, A.C. (1983), Field and laboratory evidence implicating Culex zombaesis and Aedes circumluteolus as vectors of Rift Valley fever virus in coastal South Africa. S. Afric. J. Sci., 79, 61-64. Mathiot, C., Fontenille, D., Georges, A.J. & Coulanges, P. (1989), Antibodies to baemorrhagic fever viruses in Madagascar. Trans. roy. Soc. top. Med. Hyg., 83, 407-409. Meegan, J.M. (1979), The Rift Vailej fever epizootic in Egypt 1977-1978. -- 1. Description of the epizootic and virological studies. Trans. roy. Soc. trop. Aled. Hyg., 73, 618-623. Meegan, J.M. & Bailey, C.H. (1988), Rift Valley fever, in "The arboviruses: epidemiology and ecology"
(T.P. Monath) (pp. 51-76). CRC Boca Raton, Florida. Meegan, J.M., Le Guenno, B., Ksiazek, T.G., Jouan, A., Knauert, F., Digoutte, J.P. & Peters, C.J. (1989), Rapid diagnosis of Rift Valley fever: a comparison of methods for the detection of viral antigen in human sera. Res. Virol., 140, 59-65. Morvan, J. (1990), in "Rapport annuel 1989". WHO Collaborating Center of Reference and Research for Arboviruses. Institut Pasteur, Dakar, S6n~gal. Morvan, J., Saluzzo, J.F., Fontenille, D. & Coulanges, P. (1991), Possible Rift Valley fever outbreak in man and cattle in Madagascar. Trans. roy. Soc. trop. Med. Hyg., 85, 108. Peters, C.J. & LeDuc, J.W. (1984), Bunyaviruses, phleboviruses and related viruses, in "Textbook of human virology" (R.B. Belshe) (pp. 547-598). PSG Pub!. Co. Ira.., Littleton, Massachusetts. Saluzzo, J.F., Chartier, C., Bada, R., Martinez, D. & Digoutte, J.P. (1987), La fi~vre de la valise du Rift en Afrique de l'Ouest. Rev. t~iev. M~d. v~t. Pays trop., 40, 215-223. Saluzzo, J.F. (1989), l~pid6miologie mol~culaire, g~n6tique et pouvoir pathog6ne du virus de la fi~vre de la Vall6e du Rift: application/t l'6valuation d'un vaccin vivant att~nu~ /t usage v6t~rinaire. Th/~se, Universit6 de Clermont-Ferrand. Saluzzo, J.F., Anderson, G.W., Smith, J.F., Fontenille, D. & Coulanges, P. (1989), Biological and antigenic relationship between Rift Valley fever virus strains isolated in Egypt and Madagascar. Trans. roy. Soc. trop. Med. Hyg., 83, 701. Yedloutschnig, R.J., Dardiri, A.H. & Walker, J.S. (1981), Persistence of Rift Valley fever virus in the spleen, liver and brain of sheep after experimental infection, in "Rift Valley fever" (T.A. Swartz, M.A. Klingberg & N. Goldblum) (pp. 72-76). Contr. Epidem. Biostatist., 3. S. Karger, Basel.