| Abstract |
Leishmaniases are diseases caused by members of the genus Leishmania, protozoan parasites infecting numerous mammal species including humans, and transmitted by the bite of phlebotomine sandflies. Εven though, they are endemic in 88 countries in which they consist serious public health problem, leishmaniases are among the most neglected diseases (http://www.who.int/tdr/diseases/leish/).
Visceral leishmaniasis (VL), caused by Leishmania infantum (L. infantum) in Europe, is the most severe form which, if left untreated, invariably leads to death in 95% of the cases. Visceral leishmaniasis is endemic in all countries of southern Europe with up to 25% seroprevalence in domestic dogs (the reservoir host) and 700 autochthonous human cases per year (Dujardin et. al., 2008). Currently, it appears that the global incidence of human leishmaniases is higher than before: environmental and human behavioural factors contribute to the changing landscape of these diseases, which show a wider geographical distribution than previously known. Autochthonous Leishmania transmission is being recorded in traditionally non-endemic areas, such as North part of Italy, Hungary and Germany (Koechler et al., 2002; Gradoni et al., 2003; Gramiccia and Gradoni, 2005; Ferroglio et al., 2006, Maroli et al., 2008).
- 93 -
In nature, dogs are considered to be the main reservoir of L. Infantum. Canine leishmaniasis due to L. Infantum is, may be, the most frequent infectious disease in Cyprus and the other Mediterranean countries, as well as in Portugal, whereas the percentage of serological or/ and PCR positive animals may reached the percentage of 80% of the whole animal population (Dereure et al., 1999, Solano-Gallego et al., 2001a).
In Cyprus, leishmaniasis has been considered exclusively a veterinary problem. Before 1945, canine visceral leishmaniasis (CanL) was widespread in Cyprus (Minter and Eitrem, 1989) and Phlebotomus tobbi (P. tobbi) was incriminated as the vector of L. infantum (Adler, 1945; Leger et al., 2000). The Malaria eradication campaign, 1940-1950, greatly reduced the sand fly fauna in Cyprus (Constantinou, 1998) whilst dog numbers fell dramatically (from 46,000 to 6,000) as a consequence of the successful anti echinococcosis campaign of 1970-1975 (Polydorou, 1984). These actions resulted in the almost complete eradication of CanL in the government controlled part of the island, which stayed clear of the disease for over 20 years (Leger et al., 2000; Deplazes et al., 1998; Leger and Depaquit, 2008). However, sand fly populations increased and the number of dogs recovered to an estimated 100,000 (17 dogs/km2) following the end of the two control programmes. As a consequence, CanL re-emerged and dog cases were again recorded in coastal areas in 1996 implicating L. infantum MON-1 as the causative agent (Leger et al., 2000; Deplazes et al., 1998). In contrast to the relatively high infection rates observed in domestic dogs (Minter and Eitrem, 1989; Adler, 1945, Leger et al., 2000) no passive or active surveys of human leishmaniasis have been conducted in the past in Cyprus and only two human cases (infantile VL) are known since 1935 (Minter and Eitrem, 1989; Deplazes et al., 1998).
The aim of the study is the epidemiological investigation of leishmaniases in Cyprus, in dogs and humans, for the control and surveillance of the disease. The survey conducted to resolve this was based on the assumption that if there were to be VL cases, they had to be in the areas with the highest dog seroprevalence. To locate these areas, an epidemiological study was carried out on the dog population. A map of Cyprus was divided into 82 equal squares of which 30 were chosen randomly as the study area. Dog seroprevalence revealed the areas with the highest risk for humans. In such two areas, as well as in one with zero dog seroprevalence, a - 94 -
seroepidemiological study was conducted on the human population and sandfly collections were done to compare species in these three areas.
The Government controlled part of the island, southern Cyprus (Fig.1), covering 5,896 km2 and comprising five prefectures, was divided into 82 equal arbitrary squares on a map. The squares were given numbers and 30 numbers were drawn from a ballot to select the squares to be included in the study. One or more villages falling in each of the 30 squares, according to dog population size, were chosen so as to have at least 200 dogs per square.
A total of 900 dogs (379 males and 521 females; 719 hunting, 87 guard, 94 companion) living in the 30 areas were sampled (30 dogs per area) during 2005 and 2006. Owned dogs, living permanently in the area irrespective of race, age color or health status were included in the study. To select the dogs, the houses in each village in the study area were given numbers and drawn from a ballot; visited with a government veterinarian and the dog owners were informed about leishmaniasis and asked to participate in the study by providing a written consent. The dogs were examined clinically, peripheral blood (900 samples) and lymph node aspirates when enlarged lymph nodes were observed (22 samples) were collected and personal, epidemiological and clinical data were registered in questionnaires for each dog. Severe leishmaniasis symptoms were diagnosed in 18 dogs, the disease was confirmed by serology and PCR, the dogs were euthanized with the consent of the owner and the spleen provided. In addition, 2,056 dog sera were provided by veterinarians throughout the island for routine testing for leishmaniasis.
The two areas (squares) with high dog seropositivity, 25% and 29% (high risk for humans), and one with no seropositive dogs (low risk) were considered for the human survey. The areas were chosen such as to be located near a local hospital. In collaboration with the Ministry of Health, doctors from local hospitals provided samples from a total of 600 people that lived permanently in these three areas (200 people per area), irrespective of nationality, age or health status, visiting the local hospital for any reason. The people were informed of the study and if agreed to participate they provided a written consent and were examined clinically for symptoms of VL and CL and completed a personal questionnaire with epidemiological and clinical data. The ages of the people providing blood samples ranged from 10 to 88 years, with an average age of 52 years (SD). In addition, blood - 95 -
(35), skin tissue (4) and bone marrow (4) samples from 35 patients with suspicious symptoms of leishmaniasis (high persisting fever or skin lesions difficult to cure) were provided from hospitals from different cities of the island irrespective of the sampled areas.
Sampling of sand flies was carried out during May to October, 2006. Live sand flies were collected by CDC light traps (Hausherr’s Machine Works in Tom’s River, New Jersey) from 20 villages with and without human and/or dog cases. The traps were battery operated and placed near animal shelters, at dusk until early morning, near houses with human and/or dog cases (if any in the village) for one or more nights. Species identification was carried out according to morphology based keys.
The 900 dog and the 635 human sera were screened for the presence of Leishmania IgG antibodies by ELISA. Seroprevalence measures were derived solely from the ELISA result in order to have comparable results with previous publications on Cyprus (Deplazes et al., 1998). All sera were further tested by the Indirect Immunofluorescence Test (IFAT) using anti-human or anti-dog anti-IgG antibodies accordingly and Electrosyneresis (ES) which is a very sensitive serological test and can identify the acute stage of the disease. The 2,056 extra dog sera were tested only by ELISA.
All 35 patient samples and all 900 random dog samples were tested by polymerase chain reaction (PCR) on peripheral blood, skin and/or bone marrow from patients and blood, lymph node or spleen from dogs, according to availability.
An attempt to culture the parasite was made from biological samples from all the 900 dogs and the 35 patient samples. Two culture media were used: the NNN medium and the RPMI 1640 (World Health Organization, 1991). The isolates were typed by starch gel electrophoresis using 15 enzymatic systems (Rioux et al., 1990) and the K26 PCR assay(Haralambous et al., 2008).
Dog seroprevalence was estimated for each of the 30 areas and the 5 prefectures and mapped using the geographical information system software (GIS; ArcGIS 9.2). Possible associations between a) dog seropositivity and b) PCR positivity and ten possible risk factors were initially assessed using the chi-squared test and univariate logistic regression models. The risk factors considered were: geographical origin, sampling season, dog use (hunting, companion, guard dog), sex, age, weight, coat colour, length of dog hair, ectoparasite presence and lastly, the - 96 -
presence of at least three of any of the following CanL symptoms (lymph node swelling, alopecia, onychogryphosis, epistaxis, ocular lesions, splenomegaly). Subsequently, a multivariable logistic regression model was fitted including all nine factors and backwards stepwise selection procedures were used to obtain a final predictive model.
Of the 900 dogs, 14.9% were positive by ELISA, 11.8% by IFAT (at the 1/200 cut off titer and 12% at the 1/100), 9.1% by ES, 6.9% by culture and 25% by PCR. In some squares seropositivity by ELISA reached 33.3%. Of the additional 2,056 dogs tested, 402 (19.6%) were seropositive by ELISA. Parasites were isolated from the blood (25 isolates), lymph node (19) and spleen (18) of 62 dogs, 9 of which were healthy looking seronegative animals. All but three dogs with a positive culture gave a positive PCR. Typing showed 61/62 isolates to be L. infantum MON-1 and one L. infantum MON-98 (Rioux et al., 1990; Haralambous et al., 2008). In addition, one dog was positive both for L. infantum MON-1/L. donovani MON-37 (Antoniou et al., 2008).
The statistical analyses showed that geographic location is a high risk factor for a dog to be seropositive and PCR positive with Pafos district to be a high risk area (area with 4/6 human cases). However, the season of sampling is also a high risk factor for a dog to be seropositive and PCR positive with high incidence in spring, summer and autumn. Age, dog use (hunting, guard, companion), gender, weight, coat shade, hair length and presence of ectoparasites were found not to be significantly associated with seropositivity.
All 600 individuals from the three areas under investigation were found seronegative. Of the 35 patients with suspicious symptoms of leishmaniasis, two with high persisting fever developed VL and four, with cutaneous lesions, CL. The VL patients were positive by ELISA and IFAT (both titers 1/200) and only one of the four CL patients had antibodies (by ELISA and IFAT, titer 1/400). PCR was positive in all 6 patients (2 VL from blood sample and 4 CL from skin biopsy) and parasite isolation was made from 5/6 patients. All isolates were typed as Leishmania donovani MON-37 by starch gel electrophoresis using 15 enzymatic systems (Rioux et al., 1990) and the K26 PCR assay (Haralambous et al., 2008).
A total of 1,716 sand flies (649 males and 1067 females), comprising 10 species were collected from the 20 villages; altitude range 37-176m: 1,512 Phlebotomus and 204 Sergentomyia. P. papatasi was found in almost all areas
- 97 -
studied (18/20), P. tobbi in 14/20, P. galilaeus in 7/20, P. sergenti in 4/20, P. alexandri in 1/20, P. mascittii in 1/20 and P. economidesi in 1/20. Since sampling was not done for the same number of nights and in the same month for each village, the results are reported as presence of species found.
Although it is clear that in Cyprus there are competent vectors for the transmission of Leishmania to dogs and humans (Deplazes et al., 1998; Leger et al., 2000) the question why there are no VL or CL cases due to L. infantum, remains. CanL is widespread in the island and seroprevalence had an almost nine fold increase in the last 10 years (overall average from 1.7% (Deplazes et al., 1998) to 14.9%). This demonstrates that the parasite is circulating actively, transmitted by autochthonous sand fly species found in most parts of the Republic (Antoniou et al., 2009) and that conditions favor its geographic spread since it is no longer found restricted in coastal areas (Adler, 1945; Deplazes et al., 1998). The seroprevalence found in the two groups of dogs tested: 14.9% in the 900 randomly selected dogs and 19.6% in the 2,056 extra dog sera sent by veterinarians from all over the country, is comparable.
In the two endemic areas sampled, with 25% and 28,6% dog seropositivity, the 400 persons tested (out of the 21,967 registered permanent residents of this area, according to Census 2001; www.mof.gov.cy/mof/cystat/statistics.nsf) were all permanent residents of the study areas with mean age of 52,7 years, factors which favour them coming into conduct with the parasite (Moral et al., 2002). Yet, the ELISA test performed on the human sera did not produce any positives. Nevertheless the two VL and the one/four CL patients due to L. donovani MON-37 were seropositive with the same ELISA test, which used L. infantum promastigote soluble antigens. Hence, although the planning of the study expected to result a small percentage of seropositives, the negative result needs to be considered with caution and more tests are required before excluding the possibility of missing seropositive individuals.
In the two endemic areas sampled, with 25% and 28,6% dog seropositivity, the 400 persons tested (out of the 21,967 registered permanent residents of this area, according to Census 2001; www.mof.gov.cy/mof/cystat/statistics.nsf) were all permanent residents of the study areas with mean age of 52,7 years, factors which favour them coming into conduct with the parasite (Moral et al., 2002). Yet, the - 98 -
ELISA test performed on the human sera did not produce any positives. Nevertheless the two VL and the one/four CL patients due to L. donovani MON-37 were seropositive with the same ELISA test, which used L. infantum promastigote soluble antigens. Hence, although the planning of the study expected to result a small percentage of seropositives, the negative result needs to be considered with caution and more tests are required before excluding the possibility of missing seropositive individuals.
The absence of VL or CL patients due to L. infantum despite the high dog seroprevalence is intriguing, suggesting a paradox of a minimal risk of L. infantum infections in humans as compared to the high risk observed in dogs in southern Cyprus; a situation not found in the surrounding countries where CanL and VL due to L. infantum coexist. On the contrary, reports indicate that the epidemiology of leishmaniasis in the northern part of Cyprus is different, with CL cases increasing from two cases per year in 1985 to 36 in 1990 in the Turkish Cypriot population (Desjeux, 1991) as well as one VL case (Minter and Eitrem, 1989) implicating L. Infantum (Howard et al., 1992). This could possibly be explained by the presence of P. neglectus, the usual vector of L. infantum in the eastern Mediterranean, in northern but not in southern Cyprus (Leger et al., 2000; Rastgeldi et al., 2005). This species is reported north of the Keryneia mountains. Taking into account that sandflies are weak fliers and do not usually disperse more than a few hundred meters from their breeding places, P. neglectus spreading towards the south is in part prevented by the mountainous terrain of the island with altitudes up to 1024m in the area concerned (Geological Survey Department of Cyprus). Also, its absence in the south is most probably linked to the biogeological/microclimate diversity of Cyprus (Antoniou et al., 2009). The above are further supported by the fact that, although the populations between the two parts of the island mix, especially after 2003 when people were allowed to cross the “green line” freely, no VL cases due to L. infantum have been reported in the southern part.
As L. donovani MON-3 was isolated in Syria from P. tobbi (Rioux et al.,1998) this species is a putative vector of both L. infantum and L. donovani. The detection of both parasite species in one dog (Antoniou et al., 2008) suggests that a sand fly biting both humans and dogs may be involved and whether this is P. tobbi or P.
- 99 -
galilaeus needs to be proved by further studies; although findings so far point to P. tobbi (Antoniou et al., 2009).
The detection of both parasite species in one dog (Antoniou et al., 2008) suggests that a sand fly biting both humans and dogs may be involved and whether this is P. tobbi or P. galilaeus needs to be proved by further studies; although findings so far point to P. tobbi (Antoniou et al., 2009).
Since a man-biting sandfly exists in Cyprus, confirmed also by a number of tourists visiting Cyprus among other endemic countries who developed VL (Wheatley et al., 1983; Valkoun et al., 1985), the absence of human VL or CL cases due to L. infantum may suggest that there is a mechanism protecting the Greek Cypriots from L. infantum but not from L. donovani which caused CL and not VL in healthy individuals, like in Sri Lanka (Nawaratna et al., 2007). It has been shown that host-related factors may determine the evolution of the host-parasite interactions and that, ethnic and familial factors can play an important role in the distribution of VL (Bucheton et al., 2002). The two human VL cases due to L. donovani (a 9 month-old female, gypsy in origin, with Epstein Barr virus co-infection (Koliou et al., 2008) and a 73 year-old British male with no other health complication), both permanent residents of Cyprus (Antoniou et al., 2008) in different geographical areas without any connection, developed severe disease. The other four cases on the other hand, all Greek Cypriots, healthy, 44 to 55 years old (Antoniou et al., 2008) developed CL the less severe form of leishmaniasis. In line with the above, microsatellite analysis showed that the genetic makeup of the strains did not correlate to the clinical manifestation of leishmaniasis.
Thus, the Republic of Cyprus presents a unique situation where two distinct leishmaniasis transmission cycles run in parallel: in dogs with L. infantum MON-1 and in humans with L. donovani MON-37. Recent data using microsatellites (Alam et al., 2009) showed that there are substantial differences between the MON-37 Cypriot strains and the MON-37 strains from India (Moreno, 1989), Israel (Schnur et al., 2001), Sri Lanka (Karunaweera et al., 2003) and Kenya (Moreno et al., 1986). Thus one cannot assume a very recent introduction by immigrants or infected sand fly vectors in Cyprus at least from the above mentioned countries (Moreno, 1989). However, the scenario that Cypriots were infected in Cyprus by parasites imported by infected people from endemic areas not yet studied, cannot be excluded.
- 100 -
The results of the dog seroepidemiological study led the Cyprus government to take measures to prevent the dispersal of leishmaniasis (decree for monitoring the disease in the dog). These measures however will protect the dog but not the human population if indeed L. donovani patients play the role of parasite reservoir. Measures must be taken to protect the population in this emerging disease hot spot otherwise L. donovani may spread fast, as it has done in Sri Lanka (Nawaratna et al., 2007). At a time when VL due to L. infantum is spreading northwards in Europe (Maroli et al., 2008) the emergence of L. donovani MON-37 in a Mediterranean country may result not only in a dramatic change in the epidemiology of leishmaniasis but it could also give the opportunity to the circulating species to generate hybrids (Ravel et al., 2006). To avoid the spread of this species it is crucial that the disease is placed under public health surveillance at the European level since putative vectors occur in many European countries (Myskova et al., 2007).
|
Search
