Serological studies of the epidemiology of sandfly fever in the Old World

Serological studies on the epidemiology of sandfly fever in the Old World R. B. TESH,' S. SAIDI,2 S. JA. GAJDAMOVPC,s F. RODHAIN,4 & J. VESENJAK-HIRJAN 5 Selected human sera from 59 different localities in Africa, the Mediterranean littoral, eastern Europe and Asia were examined by plaque reduction neutralization test against eight sandfly (Phlebotomus) fever virus serotypes (Sicilian, Naples, Arumowot, SudAn 754-61, Karimabad, Salehabad, Gordil and Saint Floris) known to occur in the Old World. Results of these studies provide new information on the geographic distribution and preva- lence of human infection with each of the viruses. Specific neutralizing antibodies were detected against all of the agents except Salehabad. Naples and Sicilian antibodies were encountered most frequently and had the widest geographic range; moreover they were found only in areas where Phlebotomus papatasi occurs. Age-specific antibody rates for several of the viruses are presented. These data and the epidemiology of sandfly fever are discussed. Although sandfly (or Phlebotomus) fever has long been recognized as a disease of some public health importance in countries bordering on the Mediterranean and in central Asia, little information is available on the extent of its geographical distribu- tion, the virus serotypes responsible for the illness, or the prevalence of infection among indigenous human populations. To date, eight antigenically distinct Phlebotomus fever virus serotypes (Naples, Sicilian, Arumowot, SudAn 754-61, Karimabad, Salehabad, Gordil, and Saint Floris) have been isolated in the Old World (1-4). While the human pathogenicity of the Naples and Sicilian viruses has been well documented (5, 6), the other six serotypes have been recovered only from insects or animals and their ability to infect man is unknown. In order to answer some of these questions, we recently examined human sera from 59 different localities in Africa, the Mediterranean region, 1 Pacific Research Section, Laboratory of Parasitic Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, P.O. Box 1680 Honolulu, Hawaii, 96806, USA. ' Department of Epidemiology and Pathobiology School of Public Health and Institute of Public Health Research, University of Tehran, Iran. ' Arbovirus Department, Ivanovskij Institute of Virology,(WHO Virus Collaborating Centre), USSR Academy of Medical Sciences, Moscow, USSR. Unite d'Ecologie Virale, Institut Pasteur, Paris, France. Andrija Stampar School of Public Health, Medical Faculty, University of Zagreb, Yugoslavia. eastern Europe, and Asia for neutralizing antibodies against the eight Old World representatives of the Phlebotomus fever group. The present paper reports results of this study and presents new data on the epidemiology of sandfly fever. MATERIALS AND METHODS Source of human sera and populations sampled Table 1 lists the 59 localities studied, the type of populations sampled, and the year in which the sera were obtained and Fig. 1 shows the approximate geographical locations. The human sera utilized in this study came from many different sources. For this reason and because of the difficulties in transporting sera from different countries, some of the specimens were received frozen, others were lyophilized, and a number were obtained dried on filter paper discs. Most of the sera from a given locality were from populations living within a single geographical region (i.e., city, town or province); in a few cases, however, the exact origin of the donors was unknown or the specimens were collected from several areas within a country. In the latter cases, only the country of origin is listed in Table 1. The sources of the sera are given in Annex 1. Viruses The sera were tested against eight different Phlebotomus fever virus serotypes known to occur 3563 -663 - BULL. WORLD HEALTH ORGAN., Vol. S4, 1976 664 R. B. TESH ET AL. Table 1. Prevalence of neutralizing antibodies against 6 Phlebotomus fever virus serotypes among selected human populations in the Old World Locality a Population ~Total no. Percentage positive ? sampled and date sera Arumo- SudAn Kari- Sale-tested Sicilian Naples wot 754-61 mabad habad 1 Senegal Children -1966 2 Zigida, Liberia Mixed -1972 3 Ghana Adults-1963 4 Nigeria Mixed -1970-74 5 Northern Nigeria Adults -1966 6 Northern Kenya Mixed -1972 7 A Giohar District, Somalia Adults -1966 7 B Giohar District, Somalia Mixed -1960-63 8 Territory of the Afars and Issas c Mixed -1974 9 A Omo Valley, Ethiopia c Mixed -1972 9 B Omo Valley, Ethiopia C Mixed -1974 10 Khartoum Province, Sudan Adults -1975 11 El Gezira Province, Sudan Adults -1975 12 Equatoria Province, Sudan Adults -1975 13 Northem Province, Sudan Adults -1975 14 White Nile Province, Sudan Adults -1975 15 Upper Nile Province, Sudan Mixed -1960-63 16 Qalyub, Egypt Mixed -1952-54 17 Cairo, Egypt Children -1975 18 Giza, Egypt Adults-1960-63 19 Alexandria, Egypt Adults-1960-63 20 Baltim-Borg Burulus, Egypt Mixed -1960-63 21 Luxor, Egypt Adults-1960-63 22 Siwa, Egypt Mixed -1960-63 23 Sidi Barrfni, Egypt Mixed -1960-63 24 Bahig, Egypt Mixed -1960-63 25 El Daba, Egypt Adults-1 960-63 26 Aswan, Egypt Adults-1 960-63 27 Tunisia Mixed-1975 28 Tamanrasset, Algeria Mixed -1975 29 Djanet, Algeria Mixed -1975 30 Marrakoch Province, Morocco Mixed -1976 31 Beni-Mellal Province, Morocco Mixed -1976 32 d'Oujada Province, Morocco Mixed -1976 33 Midelt-ltzer, Morocco Mixed -1976 34 Crete, Greece Adults -1976 35 Athens, Greece Mixed -1973-75 36 Arachova, Greece Children -1973 37 Brad, Dalmatia Province, Yugoslavia Adults -1975 38 Kosovo Province, Yugoslavia Mixed -1975 39 Antalya, Turkey Adults -1955 40 Massayeb-AI-Kabir, Iraq C Mixed -1972-73 41 Saudi Arabia Adults -1967 42 Tabriz, East Azerbaijan Province, Iran Adults-1971 43 Rasht, Gilan Province, Iran Adults- 1971 44 Khorasan Province, Iran c Mixed-1971-75 45 Tehran Province, Iran C Mixed-1971-75 46 Kermanshah Province, Iran Adults -1975 47 Isfahan Province, Iran c Mixed -1974-75 48 Khuzestan Province, Iran c Mixed -1973-75 49 Moldavian SSR, USSR Adults -1976 50 Azerbajd±an SSR, USSR Adults -1976 51 Uzbek SSR, USSR Adults -1976 52 Tadfik SSR, USSR Adults -1976 53 Turkmen SSR, USSR Adults -1976 54 Karachi, Pakistan Mixed -1965 55 Dacca, Bangladesh Mixed -1965 56 Rangoon, Burma Mixed- 57 Southern Viet Nam Mixed -1972 58 Northern Peninsular Malaysia Adults -1973 59 Northern China Adults-i 974-75 50 99 34 94 115 24 50 48 127 60 89 207 80 5 6 7 91 252 181 16 37 45 64 51 21 64 36 55 155 46 37 20 30 27 35 38 632 96 212 104 50 40 34 100 93 336 257 32 620 486 163 98 197 158 100 75 75 51 95 103 379 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 12.0 0.0 0.0 2.1 0.0 2.1 0.0 3.1 0.0 0.0 3.3 0.0 0.0 0.0 0.0 13.0 14.0 1.4 13.8 16.3 0.0 20.0 0.0 80.0 16.7 33.3 16.7 14.3 14.3 0.0 6.6 0.0 72.5 22.6 31.0 0.0 5.0 6.1 0.0 43.8 43.8 0.0 27.0 21.6 0.0 8.9 6.7 0.0 59.4 56.3 1.6 2.0 3.9 0.0 4.8 9.5 0.0 6.3 25.0 1.6 8.3 13.9 8.3 43.6 47.3 0.0 1.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 5.7 2.9 0.0 0.0 13.1 - 8.5 24.7 - 0.0 0.0 0.0 15.6 57.6 - 9.6 27.9 - 22.0 62.0 - 2.5 7.5 - 20.6 5.9 0.0 12.0 26.0 - 12.9 21.5 - 9.5 17.9 - 21.4 30.4 - 9.4 28.1 - 21.8 13.5 - 20.4 13.2 - 3.7 2.5 - 3.1 3.1 - 7.6 4.1 - 6.3 2.5 - 2.0 6.0 - 2.7 9.3 - 2.7 12.0 - 0.0 0.0 - 0.0 0.0 - 0.0 0.0 - 0.0 0.0 - 0.0 0.0 0.0 0.0 0.0 0.0 - - 2.6 0.0 0.0 0.0 0.0 - 0.0 - - 0.0 0.0 0.0 0.0 0.0 - 0.0 0.0 - 0.0 - - 27.5 4.3 - 26.3 1.3 - 60.0 0.0 - 16.7 0.0 - 0.0 0.0 - 34.1 11.0 0.0 0.0 0.0 0.0 0.0 0.0 - 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.1 0.0 0.0 0.0 0.0 0.0 9.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 7.3 1.8 0.0 0.0 0.0 - 0.0 0.0 - 0.0 0.0 - 0.0 0.0 - - 0.0 - - 8.6 0.0 0.0 - - 0.0 - - 0.0 - 0.0 0.0 - - 1.0 0.0 - 0.0 0.0 - 11.0 0.0 - 8.2 0.0 - 0.0 0.0 62.1 0.0 - 0.2 0.0 - 0.0 - 1.0 - 1.0 - - 10.1 - - 53.0 - - 0.0 0.0 - 0.0 0.0 - 0.0 - - 0.0 0.0 - 0.0 0.0 a Number refers to locality identification shown in Fig. 1. b No. positive sera/total no. sera tested. Sera producing >80 % plaque inhibition were recorded as positive. Dash = not tested. c Sera tested at 1: 20 dilution: the remainder were tested at 1: 10 serum dilution. SANDFLY FEVER IN THE OLD WORLD Fig. 1. Map showing the approximate location of the 59 populations sampled. The numbers on the map are identi- fied in Table 1. in the Old World (1-4). The virus strains used were -as follows: the prototype Sicilian and Naples strains; Arumowot, strain Ar 1284-64; an undescribed serotype, strain SudAn 754-61; Karimabad, strain 1-58; Salehabad, strain 1-81; Gordil, strain DakAnB 496d; and Saint Floris, strain DakAnB 512. The Sicilian, Naples, Arumowot, Karimabad, and Salehabad strains were obtained from the American Type Culture Collection, Rockville, MD. SudAn 754-61 virus was received from the Center for Disease Control, Atlanta, GA. Gordil and Saint Floris viruses were provided by Dr R. E. Shope, Arbovirus Research Unit, Yale University, School of Medicine, New Haven, CT. Pools of each of the viruses were prepared from infected Vero cell cultures for use in neutralization tests. Neutralization tests All human serum specimens were screened at a single dilution (1:10 or 1: 20) as noted in Table 1. Serum or blood specimens received on filter paper discs (Schleicher and Schuell, No. 740-E) were eluted overnight at 5°C in phosphate-buffered saline (PBS-G), pH 7.2, containing 0.5% gelatin, penicillin (400 IU/ml) and streptomycin (400 ,ug/ml) as described in an earlier publication (7). Hyperimmune rodent sera were tested in serial twofold dilutions from 1:10 to 1:20 480. All serum dilutions were made in PBS-G. Prior to testing, the sera were inactivated at 56°C for 30 min. All specimens were tested by the plaque reduction neutralization method (2, 6) in 24-well microtitration plate cultures of Vero cells against a fixed virus dose of 40-100 plaque forming units. A single well was used for each human serum; duplicate wells were used for titrations of rodent antisera. The serum- virus mixture was incubated overnight at 5°C prior to inoculation. A double overlay system, employing 1.6% gum tragacanth in the initial overlay, was used and has been described previously (2, 6). Magnesium chloride (25 mmol/litre) was added to both overlays as this salt was found to increase the size and improve readability of the virus plaques (8). Cultures were read 6-10 days after inoculation. Those sera producing 80% plaque inhibition or more were recorded as positive, indicating specific neutralizing antibodies. All serological tests were performed at the Pacific Research Section, Honolulu. RESULTS The results of cross-neutralization tests between the eight Old World Phlebotomus fever virus serotypes and their specific immune sera are sum- marized in Table 2. With the exception of Saint Floris, each of the virus serotypes was distinct by 665 R. B. TESH ET AL. Table 2. Results of cross-neutralization tests with 8 Phlebotomus fever virus serotypes a Virus Immune serum SudAn Saint Kari- Sale-Naples Sicilian Arumowot 754-61 Gordil Floris mabad habad Naples 2560 <10 <10 <10 <10 40 <10 <10 Sicilian <10 10240 <10 <10 <10 <10 <10 <10 Arumowot 10 <10 _>5120 <10 <10 40 <10 <10 SudAn 754-61 <10 <10 <10 2560 <10 <10 20 <10 Gordil 20 <10 <10 <10 2560 1280 <10 <10 Saint Floris <10 <10 <10 <10 <10 10240 <10 <10 Karimabad <10 <10 <10 40 <10 <10 1280 <10 Salehabad <10 <10 <10 <10 <10 10 10 5120 a The figures given indicate the highest serum dilution producing >80 % plaque inhibition. the plaque reduction neutralization test and little heterologous crossing occurred. A one-way cross between Saint Floris virus and Gordil antiserum was noted, however, indicating that these two virus serotypes are antigenically more closely related. Table 1 lists the neutralization test results obtained with Sicilian, Naples, Arumowot, SudAn 754-61, Karimabad, and Salehabad viruses on the human sera from each of the 59 localities sampled. Fig. 1 shows the approximate location of the populations sampled: Fig. 2 the known distribution of Phlebo- tomus papatasi. Sicilian and Naples viruses Sicilian and Naples virus neutralizing antibodies usually occurred together, thus their geographical distribution was very similar (Table 1 and Fig. 1). Antibodies against Sicilian virus were observed Fig. 2. A partial map of the Old World. The hatched areas represent the known distribution of Phlebotomus papatasi. (This map was prepared by Dr D. J. Lewis, Natural History Section, British Museum, London.) 666 SANDFLY FEVER IN THE OLD WORLD Table 3. Prevalence of Sicilian and Naples virus neutralizing antibodies in 3 representative populations (percentage positive) Qalyub district, Egypt Isfahan Province, Iran Khuzestan Province, Iran Age group (years) Total no. Sicilian Naples Total no. Sicilian Naples Total no. Sicilian Naples tested tested Ttested 1-5 52 13.5 a 17.3 a 98 0.0 2.0 110 13.6 0.9 6-10 51 21.6 31.4 161 3.1 2.5 169 19.5 7.1 11-20 89 20.2 38.2 76 10.5 15.8 121 23.1 24.8 21-30 14 42.9 21.4 48 41.7 22.9 43 25.6 23.3 31-40 14 42.9 28.6 41 36.6 34.1 17 17.6 23.5 41-50 11 36.4 27.3 34 32.4 44.1 13 15.4 30.8 51-60 1 1 9.1 33.3 20 40.0 50.0 4 50.0 0.0 > 60 10 40.0 50.0 8 12.5 50.0 9 55.5 33.3 a Percentage positive. among human populations living in Bangladesh, Egypt, Greece, Iraq, Iran, Morocco, Pakistan, Saudi Arabia, Somalia, Sudan, Tunisia, Turkey, the USSR, and Yugoslavia. Naples antibodies were also found in all of the above countries except Somalia and Tunisia. In addition, Naples antibodies alone were present in sera from Ethiopia and the adjoining Territory of the Afars and Issas. No antibodies against Sicilian or Naples virus were found among populations sampled in West Africa, South-East Asia, and China. Interestingly, antibodies to these 2 agents were detected only in areas where Phlebotomus papatasi occurs (Fig. 1 and 2). Table 3 shows the prevalence of Sicilian and Naples virus neutralizing antibodies by age group in 3 different communities in Egypt and Iran. In general, the prevalence of antibodies to both agents appears to increase with age. Results obtained in Qalyub with Naples and Sicilian, and in Khuzestan with Sicilian indicate that these viruses are endemic in both areas and that many of the inhabitants are infected during childhood. Unfortunately, in most of the communities tested, the sample size and age distribution of the population were inadequate to determine the age-specific antibody rates. Arumowot and SudAn 754-61 viruses Since Arumowot and SudAn 754-61 viruses have been isolated only in Africa (1, 2), neutralization studies with these two agents were restricted mainly to human sera from that region. Arumowot anti- bodies were found among populations in Egypt, Somalia, and Sudan (Table 1). Arumowot infection rates in most communities were low, except for two provinces in southern Sudan (Equatoria and West Nile) where the antibody prevalences were 80.0 and 72.5 %, respectively. Neutralizing antibodies against SudAn 754-61 virus were observed among persons living in Egypt, Morocco, Nigeria, and Sudan. In Sudan the preva- lence of SudAn 754-61 antibodies varied from 0 to 60.0%; however, human infection rates in the other three countries were generally low. Karimabad and Salehabad viruses Karimabad virus neutralizing antibodies were recorded in 2 geographically isolated regions of Central Asia and Africa (Table 1 and Fig. 1). The highest prevalence was recorded among residents of central and north-eastern Iran (Isfahan, Teheran and Khorasan Provinces) and neighbouring regions of the USSR (Tadlik and Turkmen Soviet Socialist Republics). A second focus of Karimabad antibodies was observed in southern Egypt (Aswan) and Sudan. Table 4 compares the prevalence of Karimabad virus neutralizing antibodies among residents of Isfahan and Khorasan Provinces in Iran. In both populations antibody rates increase with age. Results from Isfahan Province indicate that most of the residents are infected during childhood. No antibodies against Salehabad virus were detected in any of the human population tested (Table 1). 667 R. B. TESH ET AL. Table 4. Prevalence of Karimabad virus neutralizing antibodies among residents of Isfahan and Khorassan Provinces, Iran Isfahan Province Khorasan Province a Agegro)uP No. positive/ Percentage No. positive/ Percentae(er) total no. positive total no. goiie tested tested pstv 1-5 38/98 38.8 0/17 0.0 6-10 101/161 62.7 0/31 0.0 11-15 36/51 70.6 9/46 19.6 16-20 20/25 80.0 2/8 25.0 21-30 33/48 68.8 7/16 43.8 31-40 37/41 90.2 7/11 63.6 41-50 26/34 76.5 7/11 63.6 51-60 16/20 80.0 0/1 - >60 7/8 87.5 3/5 60.0 a Ali-Abad and Esmaiil-Abad villages only. Gordil and Saint Floris viruses Since these two agents were received near the end of our study, only a subsample of the African sera were tested against them. The results of these studies are summarized in Table 5. Neutralizing Table 5. Prevalence of Gordil and Saint Floris virus neutralizing antibodies in selected African populations Percentage Total no. positive Locality a sera tested Gordil Saint Floris 5 Northern Nigeria 51 0.0 0.0 6 Northern Kenya 24 0.0 0.0 7A Giohar District, Somalia 24 4.2 - 7 B Giohar District, Somalia 47 0.0 2.1 9 A Omo Valley, Ethiopia 57 0.0 0.0 15 Upper Nile Province, Sudan 91 1.1 13.2 26 Aswan, Egypt 55 0.0 1.8 28 Tamanrasset, Algeria 21 0.0 - 29 Djanet, Algeria 37 0.0 a Number refers to locality identification in Table 2 and Fig. 1. b No. positive sera/total no. sera tested. Sera producing > 80 % plaque inhibition were recorded as positive. antibodies against Gordil virus were detected in one serum each from Somalia and Sudan. Saint Floris antibodies were found in residents of southern Egypt, Somalia, and Sudan. Despite the close antigenic relationship between Saint Floris and Gordil viruses noted previously with rodent hyperim- mune sera (Table 2), none of the human sera neutralizing Saint Floris virus were positive with Gordil or vice versa. DISCUSSION The results of this study provide new information on the geographical distribution and prevalence of human infection with the eight known Phlebo- tomus fever virus serotypes of the Old World. Previous knowledge of their occurrence has been obtained mainly from field isolations. Although important, these recoveries often reflect the location of virus laboratories and do not necessarily indicate the complete geographical distribution of the viruses. Furthermore, most of the field isolations have been made from animals or insects, conse- quently little is known about the ability of many of these viruses to infect man. Sicilian and Naples viruses The prototype strain of Sicilian virus was isolated by Sabin & Sweet (9) from pooled sera collected from two sick soldiers in Italy in 1943. Subsequent recoveries of this agent been made from humans and/or sandflies in Egypt, India, Iran, and Pakistan (1, 10, 11). Results of the present study (Table 1) indicate that Sicilian virus is also present in Bangla- desh, Greece, Iraq, Morocco, Saudi Arabia, Somalia, Sudan, Tunisia, Turkey, the southern European and central Asian republics of the USSR, and Yugoslavia. Previous serological studies have found a low prevalence of Sicilian haemagglutinating antibodies among human residents of southern France (C. Hannoun, personal communication, 1976) and Portugal (12), suggesting that the virus may also be present in these two countries. Naples virus was first isolated from a febrile patient in Italy in 1944 (9): additional recoveries have been made in Egypt, India, Iran, Pakistan and the Soviet Union (1, 11, 13). Our serological results (Table 1) extend the known distribution of this agent to include Bangladesh, Ethiopia, Greece, Iraq, Morocco, Saudi Arabia, Sudan, Territory of the Afars and Issas, Turkey, and Yugoslavia. The absence of Naples and Sicilian virus neutralizing 668 SANDFLY FEVER IN THE OLD WORLD Table 6. Expected and observed frequencies of persons with virus neutralizing antibodies in 3 different communities Naples and Sicilian GalyubEgypt Isfahan Province, Iran Khuzestan Province,Qalyub, (person >10 years) Iran Naples positive 0.310 (78/252 a) 0.300 (68/227) 0.132 (64/486) Sicilian positive 0.226 (57/252) 0.278 (63/227) 0.204 (99/486) Expected frequency of double reactors 0.070 (0.310x0.226) 0.083 (0.300x0.278) 0.027 (0.132x0.204) Observed frequency of double reactors 0.087 (22/252) 0.106 (24/227) 0.047 (23/486) a No. positive/Total no. tested (see Table 3). antibodies among populations in central Africa and eastern Asia suggests that these agents do not occur in these two regions. The presence of Naples and Sicilian virus neutra- lizing antibodies among residents of Dalmatia, Yugoslavia (Table 1) is of some historical interest, since this is the region where the original studies on the epidemiology of sandfly fever were carried out by Doerr et al. (14). Although the virus serotype studied by Doerr and his coworkers is unknown, the predominance of Naples antibodies among the present adult residents of Dalmatia (ages 33 to 80 years) suggests that it may have been the Naples serotype. Several characteristics of the human antibody patterns observed with Naples and Sicilian viruses imply that both agents have a similar ecology. First, their geographical distribution is quite similar and seems to parallel that of P. papatasi. Reported isolations (1, 10, 11, 13) as well as the results of our serological studies indicate that Naples and Sicilian viruses are present in the Mediterranean coastal regions of Europe and North Africa, the Nile valley, most of South-West Asia, areas adjacent to the Black and Caspian Seas, and central Asia as far east as Bangladesh (Table 1 and Fig. 1). As shown in Fig. 2, this closely approximates the known distribution of P. papatasi. Available evidence indicates that this sandfly species is the principal vector of both viruses (10, 15, 16). Inter- estingly, both cutaneous and visceral leishmaniasis are also endemic in most of the regions where Naples and Sicilian viruses occur (17). Secondly, in many of the populations sampled, the prevalence of Naples and Sicilian neutralizing antibodies was similar (Tables 1 and 3). In order to confirm this observation and to check the speci- ficity of our neutralization test results, we compared the expected and observed frequencies of persons with antibodies to both Naples and Sicilian viruses in three different communities. The results of these calculations are shown in Table 6. In each com- munity the observed frequency of double positives was slightly higher than the expected, perhaps indicating some cross-reactivity. However, the higher observed frequency is also compatible with the hypothesis that persons with high sandfly exposure and infected with one Phlebotomus fever virus serotype are at greater risk of being infected with a second. The failure to demonstrate heterolo- gous neutralizing antibodies in man (6) or in animals (Table 2) experimentally infected with Naples and Sicilian viruses also support the latter hypothesis and the specificity of our neutralization test results. A third pattern that is noted with the Naples and Sicilian (Table 3) as well as with the Karimabad (Table 4) test results is that the respective antibody rates increase with age during the first 20 to 30 years of life, but do not continue to rise in the older age groups. There appear to be two possible explana- tions for this phenomenon. The most obvious is that neutralizing antibodies to Phlebotomus fever group viruses may have a relatively short life, their titres falling to nondetectable levels 20-30 years after infection. If the neutralization test detects only those persons with infections of relatively recent origin (i.e., within the past 20 years), then this might account for the plateau effect observed among the older age groups. However, this hypo- thesis is not supported by results of serological studies among Athens residents (18), which indicate that detectable levels of Naples and Sicilian neutra- 669 670 R. B. TESH ET AL. lizing antibodies persist for at least 30 years after infection. Another explanation for the static antibody rates observed among the older population groups (Tables 3 and 4) might be that not all persons in a community have equal exposure to sandflies and that some people apparently escape infection. This latter hypothesis is compatible with field observations that the distribution of sandflies within a community is usually patchy (7, 19). These insects often show a -marked preference for certain quarters of a city or even for certain houses within a village. In view of the short flight range of these insects (20-22), the above data may indicate that some individuals or families within a community are at greater risk of virus infection than others. Such focal patterns of human infection and clustering of cases within a community have been described previously with cutaneous (23, 24) and visceral leishmaniasis (kala azar) (25), two other sandfly-borne diseases. Arumowot virus Arumowot virus was first isolated in 1963 from mosquitos (Culex antennatus) collected in Sudan (1). Subsequent recoveries of this agent have been made in the Central African Republic,a Ethiopia, Nigeria, South Africa, and Southern Rhodesia all from animals or insects (1, 4, 33). Our results (Table 1) indicate that the virus is also present in Egypt and Somalia. Collectively, these data suggest that Arumowot virus is widely distributed in Africa. While the disease potential of Arumowot virus for man is unknown, the serological data show that man is infected with this agent. Furthermore, the high prevalence of Arumowot neutralizing antibodies in man in the Upper Nile and neighbouring Equatoria Provinces in Sudan suggest that the virus is endemic in this region. Interestingly, the six arthropod isolations of Arumowot virus have been from mosquitos (1, 33). Experimental studies also have demonstrated that Arumowot virus multiplies in Aedes albopictus and Culex fatigans following intrathoracic inoculation (26), suggesting that this agent may actually be mosquito-borne. SudAn 754-61 virus SudAn 754-61 virus, an undescribed member of the Phlebotomus fever group, was originally isolated in 1961 from animals in Sudan (4). Additional a Renamed Central African Empire in 1976. recoveries of this agent have been made in Benin, Nigeria, and Senegal (1, 4, 27). Our neutralization test results (Table 1) add Egypt and Morocco to the known geographical distribution of the virus. Although these serological data indicate that human infection with SudAn 754-61 virus is fairly common in several African countries, its disease potential for man is unknown. During arbovirus studies in West Africa between 1966 and 1970 (4), a total of 53 isolations of SudAn 754-61 was made from wild animals. Most of the recoveries were from rodents, suggesting that these animals are involved in the ecology of the virus. Interestingly, this agent has never been recovered from an arthropod. As noted previously with Naples and Sicilian, the geographical distribution of Arumowot and SudAn 754-61 viruses also appears to overlap. Because of the high prevalence of antibodies to the latter two agents observed among residents of Upper Nile and Equatoria Provinces in Sudan (Table 1), we considered the posibility that some of these positives might be due to nonspecific cross-neutralization. In Upper Nile Province, neutralizing antibody rates for Arumowot and SudAn 754-61 viruses were 72.5 (66/91) and 34.1 % (31/91), respectively. Using these values, the expected percentage of double positives would be 24.7% (72.5 x 34.1/100). The observed percentage of double reactors to Arumowot and SudAn 754-61 among this population was 26.4% (24/91), suggesting that the neutralization test results were specific. This hypothesis is further supported by the results from Khartoum and El Gezira Provinces (Table 1) where 27.2% (78/287) of the combined population samples had neutralizing antibodies to SudAn 754-61 virus but only 1.0% (3/287) were positive to Arumowot. Karimabad virus The prototype strain of Karimabad virus was isolated from an unidentified pool of Iranian sandflies in 1959 (1). A total of 11 subsequent isolations of this agent have been made in Iran from P. papatasi, the presumed vector (10). Although the disease potential of Karimabad virus for man is unknown, our results (Table 1) indicate that there is a significant amount of human infection. The high Karimabad infection rates observed in Iran among residents of Isfahan and Khorasan Provinces (Table 4) suggest that the virus is highly endemic in this region. Interestingly, while Kari- mabad virus is present in some of the areas where SANDFLY FEVER IN THE OLD WORLD 671 the Naples and Sicilian serotypes occur (Table 1), it appears to have a much more limited geographical distribution than the latter two agents. Our sero- logical data indicate that Karimabad virus is present in central and north-east Iran as well as in several of the neighbouring republics of the Soviet Union (Fig. 1). The presence of Karimabad neutralizing antibodies in residents of southern Egypt and Sudan was unexpected, suggesting that this virus or an antigenically related agent also occurs in the Nile Valley. Salehabad virus Salehabad virus was originally isolated from a single pool of Iranian sandflies in 1959 (1). No subsequent recoveries of this agent have been made. The absence of Salehabad virus neutralizing anti- bodies in any of the sera tested (Table 1) suggests that human infection with this agent is rare or does not occur. Gordil and Saint Floris viruses The two remaining Phlebotomus fever virus serotypes, Gordil and Saint Floris, were both isolated from rodents by French workers in the Central African Republic a in 1971 (1, 3). No other recoveries of these viruses have been reported nor have they been associated with human illness. Since these two agents were received midway through our study, only a subsample of sera from selected African populations was tested against them. The results (Table 5) indicate that Gordil virus also occurs in Somalia and Sudan and that Saint Floris is present in these two countries as well as in Egypt. With the exception of Upper Nile Province, Sudan, where 13.2% (12/91) of the sample population had antibodies against Saint Floris virus, the prevalence of antibodies to these two agents was generally low, suggesting that they infrequently cause human infection. The absence of Sicilian, Naples, Karimabad or Salehabad neutralizing antibodies in sera of residents of Burma, Viet Nam and Peninsular Malaysia (Table 1) suggests that these viruses do not occur in South-East Asia. Both P. papatasi (Fig. 2) and leishmaniasis (17) are also absent from this region. The absence of Phlebotomus fever virus antibodies among former residents of North China is rather surprising, since an earlier report in the literature a Renamed Central African Empire in 1976. suggested that sandfly fever was present in this region. In 1915, R. A. Bolt, an American physician working in Peking, described a disease suggestive of classical sandfly fever that occurred each summer among foreign missionary personnel and students from South China (28). The disease affected only newcomers to the region and followed closely upon exposure to sandfly bites. Although P. papatasi does not occur in China (Fig. 2). P. chinensis, an important man-biting species and vector of visceral leishmaniasis, is widely distributed in the central and northern provinces of the country (29). In view of this, our data suggest that if sandfly fever occurs in China, it is probably due to an as yet unknown Phlebotomus fever virus serotype(s). This would not be surprising, since there are already at least 25 different serotypes recognized from various regions of the world (1-3). Unfortunately, our serological studies do not answer the essential question-what is the real public health importance of Phlebotomus fever group viruses? The demonstration of specific neutralizing antibodies in human serum indicates previous virus infection, but it does not necessarily mean that a clinical illness occurred. While the pathogenicity of Naples and Sicilian viruses for man is well documented (5, 6), the human disease potential of the other six Old World Phlebotomus fever virus serotypes is unknown. Additional studies on their pathogenicity are necessary before their public health importance can be fully evaluated. Furthermore, there is some evidence that the clinical manifestations of sandfly fever may vary with age. In susceptible adults, inoculation of the Naples or Sicilian viruses produces classical sandfly fever, an acute self-limited illness of 2-4 days duration, characterized by fever, headache, anorexia, myalgia, photophobia, low-back, and retro-orbital pain (5, 6). Several authors have suggested, however, that the disease in children is somewhat milder (30-32). This implies that the age at which infection occurs may influence the severity of the resulting illness. In areas where sandfly fever is endemic, most of the population is probably infected during childhood (7, and Tables 3 and 4) and suffers only a relatively mild illness. In contrast, when nonimmune adults (tourists, settlers, soldiers, etc.) enter an endemic area, classical sandfly fever usually results. Thus the Phlebotomus fever group viruses may be of potentially greater importance to outsiders than they are to persons residing within the endemic areas. 672 R. B. TESH ET AL. ACKNOWLEDGEMENTS The authors are indebted to all of the persons (listed in the Annex) who provided sera for this study. Without their interest and cooperation, the study would not have been possible. We also thank Dr P. Bres, Virus Diseases, WHO, Geneva and Dr H. Hoogstraal, US Naval Medical Research Unit No. 3, Cairo, for their help in obtaining some of these specimens. We are especially grateful to Dr D. J. Lewis, Natural History Section, British Museum, London, for preparing the map (Fig. 2) showing the known geographical distribution of P. papatasi. RtSUMt ETUDES SEROLOGIQUES SUR L EPIDEMIOLOGIE DE LA FIEVRE A PHLEBOTOMES DANS L'ANCIEN MONDE On a etudie des serums humains provenant de 59 loca- lites differentes de I'Ancien Monde; on y a recherche, par la neutralisation au moyen de la methode de reduction des plages, des anticorps contre 8 s6rotypes de virus responsables de fievre A phlebotomes, A savoir les sero- types Sicile, Naples, Arumowot, SudAn 754-61, Karima- bad, Salehabad, Gordil et Saint Floris). La distribution geographique et la pr6valence de l'infection humaine due A ces divers virus sont exposees, d'apres les resultats de ces etudes serologiques. Les anticorps neutralisants contre les virus Naples et Sicile ont e deceles dans des populations humaines vivant sur le littoral m6diterraneen, dans la Vallee du Nil, les regions voisines de la Mer Noire et de la Mer Caspienne ainsi qu'en Asie centrale et, vers l'est, aussi loin que le Bangladesh. En general, les anticorps A l'gard de ces deux virus etaient trouves ensemble, et exclusivement dans les localites ou Phlebo- tomus papatasi etait present. Au contraire, les anticorps contre les virus Arumowot et SudAn 754-61 ont ete observes surtout en Afrique centrale. Ainsi, des anticorps contre l'un de ces agents, ou contre l'un et l'autre, ont e signales en Somalie, au Soudan, en Egypte, au Nigdria et au Maroc. Des anticorps neutralisants contre le virus Karimabad ont ete decouverts en Iran et dans les regions limitrophes de l'URSS, aussi bien qu'au Soudan et dans le sud de l'Egypte. II n'a ete trouv6 aucun indice d'infection humaine par le virus Salehabad. Les taux d'infection par les virus Gordil et Saint-Floris etaient generalement faibles; des anticorps contre l'un de ces agents ou contre les deux ont ete observes dans des populations en Somalie, au Soudan et en Egypte. On trouvera representes les resultats des epreuves de neutralisation; les consequences de ces resultats, en matiere d'epidemiologie de la fievre A phlebotomes dans l'Ancien Monde, sont examinees. REFERENCES 1. BERGE, T. O., ed., International catalogue of arbo- viruses, 2nd Edition. Atlanta, GA, Department of Health, Education, and Welfare, US Public Health Service, 1975 (Publication No. (CDC) 75-8301). 2. TESH, R. B., ET AL. Antigenic relationships among phlebotomus fever group arboviruses and their implications for the epidemiology of sandfly fever. The American journal of tropical medicine and hygiene, 24: 135-144 (1975). 3. Rapport annuel, Institut Pasteur, Bangui, 1971, p. 44. 4. KEMP, G. E., ET AL. Isolation of viruses from wild mammals in West Africa, 1966-1970. Journal of wildlife diseases, 10: 279-293 (1974). 5. SABIN, A. B. Experimental studies on Phlebotomus (Pappataci, sandfly) fever during World War II. Archiv far die gesamte Virusforschung, 4: 367-410 (1951). 6. BARTELLONI, P. J. & TESH, R. B. Clinical and serologic responses of volunteers infected with Phlebotomus fever virus (Sicilian type) The American journal of tropical medicine and hygiene, 25: 456462 (1976). 7. SAIDI, S., ET AL. Studies on the epidemiology of sandfly fever in Iran. 2. The prevalence of human and animal infection with 5 Phlebotomus fever virus serotypes in Isfahan Province. The American journal of tropical medicine and hygiene, 26: 288-292 (1977). 8. WALLIS, C. & MELNICK, J. L. Magnesium chloride enhancement of cell susceptibility to poliovirus. Virology, 16: 122-132 (1962). 9. SABIN, A. B. Recent advances in our knowledge of dengue and sandfly fever. The American journal of tropical medicine and hygiene, 4: 198-207 (1955). SANDFLY FEVER IN THE OLD WORLD 673 10. TESH, R., ET AL. Studies on the epidemiology of sandfly fever in Iran. 1. Virus isolates obtained from Phlebotomus. The American journal of tropical medicine and hygiene, 26: 282-287 (1977). 11. GOVERDHAN, M. K., ET AL. Isolation of Phlebotomus (sandfly) fever virus from sandflies and humans during the same season in Aurangabad District, Maharashtra State, India. Indian journal of medical research, 64: 57-63 (1976). 12. FILIPE, A. R. Serological survey for antibodies to arboviruses in the human population of Portugal. Transactions ofthe Royal Society ofTropical Medicine and Hygiene, 68: 311-315 (1974). 13. GAIDAMOVICH, S. YA., ET AL. Aetiology of Phleboto- mus fever in Ashkhabad studied in retrospect. Acta virologica, 18: 508-511 (1974). 14. DoERR, R., ET AL. Das Pappatacifieber. Leipzig, Franz Deuticke, 1909. 15. SCHMIDT, J. R., ET AL. Phlebotomus fever in Egypt. The American journal of tropical medicine and hygiene, 20: 483-490 (1971). 16. GEORGE, J. E. Isolation of Phlebotomus fever virus from Phelebotomus papatasi and determination of the host ranges of sandflies (Diptera: Psychodidae) in West Pakistan. Journal of medical entomology, 7: 670-676 (1970). 17. LYSENKO, A. JA. Distribution of leishmaniasis in the Old World. Bulletin of the World Health Organiza- tion, 44: 515-520 (1971). 18. TESH, R. B. & PAPAEVANGELOU, G. The effect of insecticide spraying for malaria control on the incidence of sandfly fever in Athens, Greece. The American journal of tropical medicine and hygiene, 26: 163-166 (1977). 19. HERTIG, M. Phlebotomus and Carri6n's disease. III. Field studies on Phlebotomus. The American journal of tropical medicine, 22: 23-59 (1942). 20. QUATE, L. W. Phlebotomus sandflies of the Paloich area in the Sudan (Diptera, Psychodidae). Journal of medical entomology, 1: 213-268 (1964). 21. FOSTER, W. A. Studies on leishmaniasis in Ethiopia. III. Resting and breeding sites, flight behaviour, and seasonal abundance of Phlebotomus longipes (Diptera: Psychodidae). Annals of tropical medicine and parasitology, 66: 313-328 (1972). 22. CHAIOTIs, B. N., ET AL. Horizontal and vertical movements of phlebotomine sandflies in a Pana- manian rain forest. Journal of medical entomology, 11: 369-375 (1974). 23. NADIM, A. & ROsTAMI, GH. S. Epidemiology of cutaneous leishmaniasis in Kabul, Afghanistan. Bulletin of the World Health Organization, 51: 45-49 (1974). 24. SHARMA, M. I. D., ET AL. Epidemiological and entomological features of an outbreak of cutaneous leishmaniasis in Bikaner, Rajasthan, during 1971. Journal of communicable diseases, 5: 54-72 (1973). 25. ADLER, S., ET AL. Investigations on Mediterranean kala azar. XI. A study of leishmaniasis in Canea (Crete). Proceedings of the Royal Society of London, Series B, 125: 491-516 (1938). 26. TESH, R. B. Multiplication of Phlebotomus fever group arboviruses in mosquitoes after intrathoracic inoculation. Journal of medical entomology, 12: 1-4 (1975). 27. KEMP, G. E. Viruses other than arenaviruses from West African wild mammals. Bulletin of the World Health Organization 52: 615-620 (1975). 28. BOLT, R. A. Sandflies (Phlebotomus) in China and their relation to disease. The China medical journal, 29: 78-86 (1915). 29. Kala-azar Prevention Conference. Control of kala- azar in China. Chinese medical journal, 67: 24-46 (1949). 30. PESCHLE, B. Osservazioni cliniche su un epidemia di febbre da papataci. Pediatria, 44: 41-51 (1936). 31. TAYLOR, R. M. Phlebotomus (sandfly) fever in the Middle East. In: Proceedings of the 6th International Congress of Tropical Medicine and Malaria, 5: 149- 158 (1959). 32. GUELMINO, D. J. & JEVTIC, M. An epidemiological and hematological study of sandfly fever in Serbia. Acta tropica, 12: 179-182 (1955). 33. McINTosH, B. M., ET AL. Culex (Eumelanomyia) rubinotus Theobald as vector of Banzi, Germiston and Witwatersrand viruses. 1. Isolation of virus from wild population of C. rubinotus. Journal of medical entomology, 12: 637-640 (1976). Annex 1 SOURCES OF SERA Sera from Ghana, Liberia, Senegal, Somalia (1966), and Saudi Arabia were obtained from Dr K. M. Johnson and Mr C. F. Peters, Center for Disease Control, US Public Health Service, Atlanta, GA, USA. Specimens from Nigeria (1970-74) were collected during arbovirus studies by the Virus Research Laboratory, University College Hospital, Ibadan, and were received through the courtesy of Dr A. Fabiyi, Dr 0. Tomori, Dr A. Fagbami, and Dr T. Monath. Human sera from R. B. TESH ET AL. northern Nigeria (1966) were collected during a World Health Organization survey for yaws and were provided by Dr R. E. Shope, Arbovirus Research Unit, Yale University, School of Medicine, New Haven, CT, USA. Specimens from Kenya, Territory of the Afars and Issas, Ethiopia (1972), and Algeria were obtained during arbovirus studies by the Institut Pasteur, Paris, France. Additional blood specimens from Ethiopia (1974) were collected by Dr G. Fuller and were kindly supplied by Dr 0. Wood, US Naval Medical Research Unit No. 5, Addis Ababa. Sudanese sera (1975) were provided by Dr 0. Marcus-Jones, Virology Department, National Health Laboratories, Khartoum. Specimens from Somalia (1960-63), Sudan (1960-63) and Egypt (1952-54 and 1960-63) were obtained from the WHO Reference Serum Bank, Yale University, through the courtesy of Dr A. S. Evans. Sera from Cairo were sent by Dr I. Z. Iman, Egyptian Orga- nization for Biological and Vaccine Production, Agouza, Egypt. Tunisian sera were kindly provided by Dr B. Nabli, Centre Ophtalmologique, Labora- toire de Recherches Virologiques, Tunis. Sera from Morocco were furnished by Dr M. A. Chabaud, Institut Pasteur du Maroc, Casablanca. Turkish serum specimens were supplied by Dr D. C. Gadju- sek, National Institute of Neurologic Diseases and Stroke, National Institutes of Health, Bethesda, MD, USA. Greek sera were provided by Dr G. Papaevangelou, Department of Hygiene and Epi- demiology, University of Athens Medical School and by Dr Leon Rosen, Pacific Research Section, Honolulu. Specimens from Yugoslavia were obtained by the Andrija gtampar School of Public Health, Medical Faculty, University of Zagreb. Blood specimens from Iraq were collected during investiga- tions of an outbreak of mercury poisoning and were provided by Dr T. W. Clarkson, Department of Radiation Biology and Biophysics, University of Rochester, School of Medicine and Dentistry, Rochester, NY, USA. Specimens from Iran were obtained during arbovirus studies by the School of Public Health and Institute of Public Health Research, University of Teheran. Sera from the Soviet Union were provided by the Arbovirus Department, Ivanovsky Institute of Virology, USSR Academy of Medical Sciences, Moscow. Human sera from Pakistan and Bangladesh were supplied by the WHO Serum Reference Bank, Institute of Hygiene and Epidemiology, Center of Epidemiology and Microbiology, Prague, Czecho- slovakia through the kindness of Dr L. Syrucek. Burmese sera were collected from patients hospi- talized in Rangoon during a chikungunya virus outbreak and were obtained from Dr S. B. Halstead, Department of Tropical Medicine and Medical Microbiology, School of Medicine, University of Hawaii, Honolulu. The samples from southern Viet Nam were received from Dr N. T. Thanh, Department of Microbiology, Faculty of Medicine, Saigon. These specimens originated from many areas of the country and were submitted for sero- logical confirmation of suspected cases of dengue. Malaysian sera were obtained from persons living in several different states in the north-east portion of the country by Dr R. Donaldson, US Army Medical Research Unit, Institute for Medical Research, Kuala Lumpur, during a survey for melioidosis. Sera from former residents of northern China were provided by Dr J. H. Cross, US Naval Medical Research Unit No. 2, Taipei, Province of Taiwan. These specimens were collected from retired Chinese servicemen born in North China but now living in the Province of Taiwan. 674