Introduction

The leishmaniases represent a group of pathologies caused by parasitic protozoa gathered in 30 species within genus Leishmania (Kinetoplastida, Trypanosomatidae). Distribution of this disease over more than 88 countries is linked to the territory of their vectors, the sandflies (Diptera): it can be encountered in Southern Europe, up to the Cévennes, but the disease is essentially endemic in tropical and sub-tropical regions. Population at risk is approximately 350 millions of persons. Total prevalence is (under-) estimated at 12 million cases and incidence of the most severe form (visceral leishmaniasis) reaches 500.000 new cases each year (TDR leishmaniasis webiste).

Lifecycle of the parasite

A. Leishmania

Leishmania are digenetic parasites: in natural conditions, they need to pass through two hosts (i) invertebrate (sandfly) and (ii) vertebrate (mammals, including humans); however, infection by transfusion is possible. Leishmania parasites can take two morphological appearances:

• The oval amastigote, which resides inside the reticuloendothelial cells of the vertebrate host.
• The slender flagellated promastigote, which replicates in the gut of the sandfly.

These forms correspond to the changes in direct environment and are accompanied by biochemical modifications and changes in gene expression.

Fig.1 Lifecycle of Leishmania

1) Parasites are inoculated into the mammalian host when the sandfly regurgitates infectious promastigotes while the female sandfly is taking a blood meal.

2) Promastigotes are quickly taken up by tissue phagocytes, monocytes and neutrophils, attracted to the biting site due to the damage caused by the sandfly. Within the macrophage, the parasite loses its flagellum and transforms into a non-motile amastigote form.

3 - 4) The amastigote survives and replicates in the very acidic environment of the phagolysosome.

5 - 6a) Infected cells can be lysed by the multiplying amastigotes, which will be freed to infect nearby cells.

6b) When a sandfly takes a bloodmeal from an infected host, it acquires either free amastigotes, or amastigote-infected macrophages.

7 - 8) Parasites ingested during the bloodmeal have to survive, multiply and transform to infectious promastigotes. The swallowed amastigotes are driven into the abdominal midgut of the fly where they are captured together with the bloodmeal in the perithrophic membrane which is secreted by gut cells. As a reaction to the ingestion of a bloodmeal, the vector secretes proteolytic enzymes. In the peritrophic matrix, the gp63 protease of the promastigotes clears the haemoglobin around the parasites, and subsequently the chitinolytic enzymes of the promastigotes lyse the chitin framework of the membrane. This function of gp63 is important because haemoglobin is known to inhibit the secretion of chitinolytic enzymes. Procyclic promastigotes, which are continuously dividing, attach to the microvilli of the thoracic midgut through lipophosphoglycan (LPG) and transform into infective promastigotes with shorter body and longer flagellum. Increasing numbers of non-replicating, rapidly moving promastigotes can be observed in the lumen of the anterior midgut and foregut. These highly infective promastigotes will be introduced into the mammalian host and the lifecycle can rebegin (ref. 1-5).

B. Sandflies (ref. 6)

Out of 700 species of Phlebotomine sandflies (Diptera, Psychodidae), 70 are known vectors of Leishmania (see criteria below). Only female sandflies feed on blood and are host to the parasite life cycle. Two genera contain anthropophagous species: Phlebotomus in the Old World and Lutzomyia in the New World; parasite/vector specificity is higher in the Old World. Besides transmitting Leishmania, some species may transmit bacteria (Bartonella) and virus (sandfly fever).
Breeding and adult resting often takes place in the same microhabitat, such as the soil accumulated in cracks in walls and rock, in animal burrows and shelters, caves, or in damp leaf litter in forests. The main requirements for breeding sites are moisture and the presence of organic detritus on which the larvae can feed. Between 30 and 70 eggs are scattered about the potential breeding site by the ovipositing female and hatching occurs one to two weeks later. The period from oviposition to adult eclosion is generally 20-40 days, but may reach up to several months in diapausing species (like P. ariasi, Mediterranean basin, cool winters).
During the day, adults are not active and seek out cool and relatively humid dark niches. This allows them to survive in very hot and dry climates. Adults become active at night (between 18.00 and 24.00) when the ambient temperature drops and humidity rises. Several abiotic factors affect sandfly numbers, but temperature and rainfall are the most important. Most species in temperate regions have only one generation per year, and consequently a single peak of activity and transmission. Sandflies can be more prevalent in the wet or the dry season, depending on species. Several species can be present in the same area, each with its own annual cycle of activity.
Sandflies occur in a very wide range of habitats from sea level to altitudes of 2800 m (e.g. Andes), and from hot dry deserts, through savannas and open woodlands to dense tropical rain forest. In general each species has fairly specific ecological requirements: f.i. some species of the rain forest are only encountered in the canopy. Feeding behaviour might play an important role in these specificities: both males and females feed on sugars. These can be obtained from aphid honeydew or from plants (ref. 7): some species are restricted to some plants (e.g. Colombian populations of Lutzomyia longipalpis in coffee plantations). Flight capacity of the insects is quite low (maximum of 2 km in open habitats). All these factors are responsible for the high micro-focality of sandflies. It is mostly the mammal host that contributes to the spreading of the parasites.

Several factors contribute to the capacity of a given sandfly to ingest and transmit Leishmania:

• the prevalence of mammal-feeding species in a leishmanial area
• the activities of both vertebrates and invertebrates, that will determine contact between them
• the seasonal distribution
• the feeding habit of the sandfly

Definition of the Leishmania-vector status of a given sandfly species is based on the following criteria:

• overlap between distribution of the sandfly and the infection
• capacity to feed on mammals (anthropophily for vectors dangerous for humans): this can be measured with "sentinel" animals, capture on human baits
• measurement of the host specificity: analysis of the bloodmeals
• presence of Leishmania in their gut (Note: the rate of natural infection is very low, about 1/1000)
• possibility to complete the whole Leishmania life cycle: experimental infections and transmission

C. Hosts (ref. 8)

A diverse range of mammals (more than 100 species) act as reservoirs for Leishmania species: Marsupialia (including opossum), Insectivora (including anteaters), Primata, Edentata (including sloths), Lagomorpha, Rodentia, Carnivora (including dogs and foxes), Hyracoidea (guinea pig), Perissodactyla (horse). Very little information is available on the possibility that some parasites might have a natural host-specificity, even if in experimental conditions, several Leishmania species may grow on the same animal (like the golden hamster). Ecological barriers most probably play a major role in parasite restriction to a given host.

Taxonomy & Geographical Distribution (ref.9)

The taxonomy of Leishmania parasites is quite complex and reflects successive trials to apprehend the biodiversity of the natural parasite populations. The genus Leishmania belongs to the Trypanosomatidae family of the order Kinetoplastida, which consists of a set of organisms characterised by the presence of the kinetoplast. The genus Leishmania is divided into two subgenera, Leishmania and Viannia, on the basis of the localisation of the parasite during its development within the gut of the infected sandfly. Parasites of the subgenus Leishmania are distributed in America, Asia, Europe and Africa; whereas parasites of subgenus Viannia are restricted to the American tropics and sub-tropics. The classification of the Leishmania parasites into species and complexes of related species is based on isoenzymatic genotyping, monoclonal antibody typing, DNA markers, clinical manifestations, reservoirs, vectors and geographical location.

Complex	Species	Distribution	Reservoir	Pathology
donovani	L. (L.) donovani	Asia	human	kala-azar + PKDL
	L. (L.) infantum	Mediterranean	dog	kala-azar
	= L. (L.) chagasi	Latin America	dog	kala-azar
	L.(L.) archibaldi	Africa (East)	dog	kala-azar
tropica	L. (L.) tropica	Middle East, India	human	dry lesion
major	L. (L.) major	Africa (North), Asia	rodents	wet lesion
aethiopica	L. (L.) aethiopica	Africa (East)	rodents	diffuse
mexicana	L. (L.) mexicana	Central America	rodents	chiclero ulcer
	L. (L.) amazonensis	South America	rodents	diffuse
braziliensis	L. (V.) braziliensis	South America	rodents, sloth, opossum	muco-cutaneous
	L. (V.) guyanensis	South America	sloth	muco-cutaneous, lymphargitis
	L. (V.) panamensis	South America	sloth	muco-cutaneous, lymphargitis
	L. (V.) peruviana	Peru	human, dog ?	uta (benign)
		Table: Principal Leishmania species

NB: in the absence of regular sexual recombination (clonal structure of wild populations) the Linnean species criterion is unapplicable: Leishmania species are first of all conventions (ref. 10). Several aspects of this classification are still matter of debate due to different factors, in particular the presence of hybrids between L.(V.) panamensis and L.(V.) braziliensis, L.(V.) braziliensis and L.(V.) guyanensis, L.(L.)major and L.(L.) arabica, and L.(V.) braziliensis and L.(V.) peruviana. In addition, the genetic differentiation between some species initially described on the basis of different geographical origins, for example L.(V.) panamensis and L.(V.) guyanensis, is contested . Finally, the link between clinical manifestation and taxonomic species is not simple, as a single species can cause different clinical outcomes whereas distinct species can sometimes cause similar outcomes.

The Disease

A. Pathology

The disease is essentially due to an immunological disequilibrium induced by the parasite (ref. 11;12).
The major forms are:

• Cutaneous and muco-cutaneous forms: granulome containing infected macrophages, lymphocytes and other immune cells; fixed or disseminated macrophages (metastasis); lesion is the expression of the inflammation; problem = trend to necrosis (ulceration); predominance of Th1 response, exacerbated cell immunity, few circulating antibodies, hence serology is poorly useful.
• Visceral form: in bone marrow, production of precursors of macrophages increases vs. precursors of other lineages decrease (hence anemia, leucopenia, thrombocytopenia); effect on spleen. Predominance of Th2 response, depressed cell immunity, abundance of antibodies thus serology is useful.

B. Treatment (ref. 13;14)

Chemotherapy is critically important in reducing the burden of disease and antimonials (SbV) are the first-line drugs for all clinical forms. Treatment is long, not devoid of adverse side-effects and expensive due to the need for hospitalisation. Drugs can only be delivered after demonstration of the parasite (or the presence of specific antibodies in the case of visceral leishmaniasis). Efficient treatment is therefore linked with the availability of adequate diagnostic tools. Treatment failure is well documented for SbV. Most alarming reports came from Bihar (India), where over 60% of VL-patients are unresponsive to SbV-treatment (ref. 15) leading to the use of conventional Amphotericin B as first-line treatment (ref. 16). In Nepal, antimony unresponsiveness ranged from 5% to 24%, the closer the district to the endemic Indian foci, the higher was the failure rate. The second-line drugs amphotericin B and pentamidine are toxic. Most promising treatments like liposomal amphotericin B and Miltefosine (hexadecylphosphocholine), which can be taken orally, are expensive and not affordable for most of the patients in endemic countries. Miltefosine has already been found to induce resistance in vitro, and the pharmacokinetics suggests that occurrence of clinical resistance to this drug is only a question of time.

C. Vaccination (ref. 17;18)

Vaccination against cutaneous leishmaniasis through the inoculation of culture promastigotes of L. major (leishmanisation) has been used in the past, but could occasionally produce non-healing lesions. Therefore, attempts were done with killed whole Leishmania parasites used with or without BCG (Bacille Camette Guérin) as adjuvant (called first generation vaccines). These were tested in clinical trials in Iran, Sudan, and Latin America, with different results: (i) 0-75% efficacy against CL and (ii) little or no protection against VL. Second-generation vaccines consist in recombinant antigens. Several antigens are potentially available but few protect against more than one species in animal models. This might be due to the genetic diversity presented by some of then. In that context, cocktails of antigens like Leish-111f might constitute a promising alternative. Another problem is the choice of an adjuvant that might induce antigen-specific Th1 responses. Recently, a new adjuvant, monophosphoryl lipid A (MPL&) was developed for that purpose and is likely the first T-cell adjuvant to be approved in human use. In combination with Leish-111f, it has been tested in phase I, and safety and immunogenicity was demonstrated. Current studies are in progress to evaluate the efficacy of this vaccine in combination with standard chemotherapy. Noteworthy, it has been shown that immunity to sandfly saliva protects against Leishmania infection; accordingly, it is likely that salivary gland proteins would be part of future vaccines against leishmaniasis (ref. 19).

Eco-Epidemiology

A. From zoonosis to anthroponosis

In natural conditions, Leishmania encounters very different environments during its life cycle (Fig.2), with consequent implications on the transmission pattern, genetic heterogeneity of the parasites and pathology.

Fig. 2 Major types of Leishmania transmission

In primary environments (e.g. Amazonian forest), parasites essentially circulate between wild mammal hosts and sandflies: this pattern is specific of a zoonotic transmission. Humans are accidental hosts (related to occupation: coca farmers, wood-cutters, gold miners, hunters); very often there is a strong sex and age bias among patients (male adult workers). The disease can be debilitating (e.g. L.(V.) braziliensis and muco-cutaneous leishmaniasis).
In secondary environments (e.g. coffee or hevea plantations), humans are more regularly involved in the life cycle: this pattern is characteristic of anthropo-zoonotic transmission. (e.g. L.(L.) mexicana and Chiclero ulcer)
In domestic environments, humans may be the only reservoir of the parasite: this is an anthroponotic transmission. The parasite may be adapted to the human host, after centuries of contact, hence a benign disease (L.(L.) tropica and Oriental Sore). In other cases, disease is still severe (L.(L.) donovani and Kala-Azar): this could be explained by a recent domestication (adaptation is still not complete). Hypothesis of the 'prudent' parasite: for a parasite it might appear an advantage not to kill its host, and thus less virulent forms should be progressively selected. This is the trend observed for many parasites. However, there are some exceptions where the rapidity and intensity of transmission compensates parasite fitness at population level: parasites are transmitted before the host dies.

B. Control strategies (ref. 20)

Several strategies may theoretically be applied for the control of parasitic diseases:

1) Detection/treatment of human cases
2) Vaccination
3) Intervention on the vector (decrease host/vector contact or eradicating the vector)
4) Intervention on the host

In the case of leishmaniasis, vaccination is currently unavailable. Because of the epidemiological complexity of leishmaniasis, control strategies need to be adapted to each situation. General principles are described here in relationship to anthroponotic or zoonotic transmission.

Zoonotic leishmaniasis
In this case, humans may get infected during a short visit in the endemic areas. The disease as such may not be controlled (how to control sandfly and host populations in the Amazonian forest?). Prevention may occur by the use of protective clothes, repellants and impregnated bednets. Detection/treatment of human cases will only have an impact at individual level. Another typical case is the presence of a permanent human settlement within a focus. Knowing the poor flight capacity of sandflies, deforestation around the village might constitute a good but short-term solution (see below).

Anthropo-zoonotic leishmaniasis
In this situation, action on the vector and the host may be undertaken. With respect to the vector, this will essentially occur by chemical control, and knowledge of the exact ecology of the sandfly is required for a correct spreading of insecticides. Resistance is a classical problem and can be overcome by the rotational use of several insecticides. According to the possibility, hosts may also be controlled. In Tunisia, the rodent constituting the reservoir of L.(L.) major is strictly dependent of a plant (family Chenopodiaceae) for his diet. Eliminating the plant around villages will send the rodents and associated vectors far away, thus decreasing the incidence of the disease in humans. In some countries (Brazil, China), case detection of dogs (reservoir of L.(L.) infantum) is performed, but followed by euthanasy (treatment is expensive and difficult in dogs). This worked well in China, but had a poor epidemiological effect in Brazil: this could be due to differences in terms of wild reservoir.

Anthroponotic leishmaniasis
This form of the disease occurs mostly in or around human settlements in the Indian sub-continent. Breeding and resting sites of sandflies are often in houses or other buildings. From the onset, control of sandflies forms the most important measure to reduce the intensity of parasite transmission as much as possible. This will occur by spreading insecticides or by the use of impregnated bednets: however, sandflies are so small insects that very fine bednets should be used (heat). With the decreasing number of cases, detection/treatment becomes the most important measure. In the case of Indian visceral leishmaniasis (L.(L.)donovani), a particular attention should be paid to detect cases of Post Kala-Azar Dermal Leishmaniasis (PKDL). This form appears after initial healing of the disease and is characterised by abundant cutaneous lesions, full of parasites; sandflies can easily be infected and therefore humans constitute an excellent reservoir for the parasites.

Evolution of Eco-Epidemiological Patterns

Last trends indicate that leishmaniasis is re-emerging worldwide, essentially because of three reasons (ref. 21).

A. Human made and environmental changes

Human made and natural changes to the environment may lead to alterations in the densities of the vectors and reservoirs, hereby increasing human exposure to infected sandflies. For instance, in several southern Mediterranean countries, new villages develop, with inadequate slaughterhouses that are attractive for semi-domestic and/or stray dogs. This is a situation that is likely to favour an increase of the population of these animals and, consequently, the sources of L. infantum for human infection. Likewise, change in climate involving rainfall and temperature, may alter the plant coverage affecting rodent reservoirs and sandfly densities and expanding the range of some sandfly species (ref. 22). In parallel, a high density of susceptible human hosts - populations never exposed to either Leishmania-free sandfly probing/blood meal and/or to Leishmania - is also the consequence of human made change drivers, like widespread migration from rural to urban areas or fast urbanization. In the city of Kabul, Afghanistan, an estimated 270 000 cases of cutaneous leishmaniasis occurred since 2001 among the less than 2 million inhabitants of the city (ref. 23). Other contributors to increased density of susceptible human hosts are population movements for economic reasons such as the development of agro-industrial projects or seeking safe heaven from civil unrest. The latter is best illustrated by the recent epidemics of VL in Sudan, where 100.000 persons (out of 300.000) would have died of VL in Western Upper Nile State (ref. 24). Last but not least, the disruption of public health systems will lead to increasing sandfly density when the vector control is compromised and to an expansion of the disease incidence together with their high severity assessing late diagnosis.

B. Emergence and spreading of drug resistance

Chemotherapy is critically important in reducing the burden of leishmaniasis and pentavalent antimonials (SbV) are the first line drug for all clinical forms. The increasing unresponsiveness to SbV, is a major concern. The most serious documented case of chemotherapy failure is situated in Bihar (India), where more than 60% of visceral leishmaniasis (VL) cases do not respond to SbV (ref. 15). A similar dramatic situation was observed in mucocutaneous leishmaniasis (MCL) where 30-40% SbV unresponsiveness to first round of treatment was observed. Molecular mechanisms leading to the emergence of drug resistance in wild populations are unknown (but have been thoroughly analysed with laboratory-induced resistant strains, ref. 25;26). Spreading of natural drug resistance is also not understood. It may depend on several factors including disease transmission cycle, parasite reproduction mechanisms, stability of the resistant phenotype and virulence of resistant parasites:

• Anthroponosis and zoonosis imply different transmission cycles and the drug selective pressure is different under these two conditions (less pressure in zoonotic conditions). Domestication of the parasite life cycles might change the patterns of spreading.
• Reproduction of Leishmania is essentially clonal (ref. 27) , but sexual exchanges might occur rarely (ref. 28) and lead to horizontal transfers of extrachromosomal genetic elements.
• Stability of the resistance phenotype is dependent on the level of resistance induced, and highly resistant parasites could be transmitted from host to host without alteration of the resistance level. The isolation of naturally resistant parasites from sylvatic vectors might be explained by stability of resistance phenotype, or by the occurrence of primary resistance (without requiring a progressive adaptation of parasites to high SbV concentrations).
• Virulence may be influenced by the resistance genotype; in this case, even in zoonotic conditions, resistant parasites might be fitter and have an increased capacity of spreading.

C. HIV-Leishmania co-infection (ref. 29)

HIV/Leishmania co-infections constitute another major threat and presently concern mostly visceral leishmaniasis. Around the Mediterranean Basin, VL was for long time a disease of children (L.(L.) infantum = Leishmania of infants). In the last few decades, an inversion of epidemiological patterns has been observed with a shift to a higher age category and to people with immuno-depression. Although people are often bitten by sandflies infected with Leishmania protozoa, most do not develop the disease. However, among persons who are immunocompromised (e.g. as a result of advanced HIV infections, immunosuppressive treatment for organ transplants, haematological malignancy, auto-immune diseases (ref. 30), cases quickly evolve to a full clinical presentation of severe leishmaniasis. VL is considered a major contributor to a fatal outcome in co-infected patients.
The Leishmania/HIV co-infection has emerged as a result of the increasing overlap between Leishmania donovani and HIV in both rural and suburban areas. Cases of co-infection have so far been reported from 36 countries around the world, most of the cases have been notified in South-Western Europe: over 1 911 up to early 2001 (WHO fact sheet). The risk of co-infected patients, carriers of Leishmania in many tissues (including blood), to be a source of infection for sand flies or for other humans (through sharing syringes among intravenous drug users), has been recently confirmed (ref. 31). Co-infected patients might become real reservoirs of Leishmania donovani sl, the causal agent of VL, especially in an urban context (ref. 32). Besides HIV, there are other immunosuppressive factors that might contribute to spreading of leishmaniasis (like graft-associated treatments (ref. 30) or concomitant diseases like tuberculosis).
The evolution of Leishmania/HIV co-infection is being closely monitored by extending the geographic coverage of the surveillance network and by improving case reporting. WHO encourages active medical surveillance, especially in South-Western Europe, of intravenous drug users, the main population at risk. Finally, because case notification of leishmaniasis is compulsory in only 40 of the 88 endemic countries, WHO strongly suggests the remaining 48 endemic countries follow suit.