Leishmaniasis in an Era of Conflict in the Middle East

Estimated disease burden

Only some of the countries in the Middle East have accurate data as to the number of cases of cutaneous and visceral leishmaniasis, and none have separate information on the causative organism. The rates quoted throughout are a synthesis of various publications, and are only a guide to the overall problem. In Jordan, for example, a recent survey found that, in the endemic area of the Jordan River Valley, the average incidence rate was 75/100,000 for CL, which is as high as for Syria 73.8/100,000, one of the countries named by World Health Organization as in the highest category (Mosleh et al. 2008). In Israel, the rate for CL in the general population is 3/100,000, but it is as high as 196/100,000 in the military, with the cases appearing in December, 3 to 4 months after the peak for civilian cases (Mimouni et al. 2009). This high number among inductees is due to their tendency to use raised animal burrows to rest on during night time maneuvers, exposing themselves to infected sand flies, emerging from the holes in the ground. Iran too has high incidence rates of CL in the endemic areas, where 38.1/100,000 was reported from the Jajarm district in 2005 (Alavinia et al. 2009). Saudi Arabia and Iraq have incidence rates of 10/100,000 and 5/100,000, respectively, which have decreased over time, but may be due to under reporting. A recent report from the Alhaweja district in Iraq estimates the incidence rate at 450/100,000, where young men were the majority of cases (Al-Samarai and Al-Obaidi 2009). U.S. military personnel returning from Middle East have been notably affected; by November 2004, 1178 cases (Lay 2004) had been reported, whereas by January 2009 (Coleman et al. 2009), the actual number of cases were between 3000 and 5000.

Data on incidence for VL in the region are scanty, but in Iraq in 2001, it was reported as high as 10.9/100,000, and in the Thi Qar governate, it was 55.5/100,000 (Jassim et al. 2006). In northwest Iran average VL incidence rates were reported as 2.8% per year with all ages equally at risk (Davies and Mazloumi-Gavgani 1999). In Turkey, VL is seen in some areas, notably Manisha/Alasehir in the Aegean region, the Marmara, Aegean, Mediterranean, and eastern Black Sea coastal areas (Ozensoy et al. 1998), with annual estimated incidence of 4.25/100,000 (Dujardin et al. 2008). Although there are very few human cases reported from Jordan, Lebanon, and Israel, on the West Bank the incidence rate is 2.79/100,000 (Amro et al. 2009). Although VL is widespread in Yemen, especially in the Taez region, no reliable incidence data have been published lately.

Diagnosis and identification

The advent of enzyme electrophoretic and molecular techniques and their application to the study of leishmaniasis has radically changed the overall understanding of the species that infect humans and the vectors that transmit the disease. In the past, researchers had relied on seeing the microscopic amastigotes from skin scrapings or bone marrow biopsies on stained slides, or culturing them in Novy-MacNeal-Nicolle (NNN) or defined media, where motile promastigotes could be observed (Schnur and Jacobson 1987). Similarly, entomologists laboriously dissected sand flies for speciation and cultured any promastigotes found. However, by using various polymerase chain reaction (PCR) technologies, diagnosis and characterization has become much easier, but has complicated the overall picture. It is now possible to identify the causative organism directly from stained smears by PCR techniques that amplify part of the small subunit ribosomal RNA gene and the ribosomal internal transcribed spacer (Schonian et al. 2003). Comparisons of microscopy, rK39 antigen dipstick, and PCR showed that the latter was more sensitive for diagnosing VL (Ozerdem et al. 2009). More recently, high-resolution melt analysis has been used to identify four Leishmania species from mammalian hosts and sand flies, allowing for rapid diagnosis, classification, and quantification, which should be very useful in both epidemiological and clinical studies (Talmi-Frank et al. 2010). In a study of parasites and reservoirs from the Middle East, Central Asia, and East Africa, it was concluded that L. major was genetically diverse, consisting of subpopulations, and that they preferentially infected their known reservoirs (Elfari et al. 2005).

In one of the few Phlebotomus/Leishmania gene homology studies, oligonucleotide primers that were originally designed to amplify salivary gland cDNA sequence (kindly provided by Dr. J.M.C. Ribeiro), cloned from female P. papatasi, had a high homology with other α-amylase genes in the gene bank. One primer pair showed significant variation between the Leishmania species, with a predominant single product size for all L. major isolates regardless of origin, and with specific polymorphic PCR product patterns for all other Leishmania species (Fig. 2) (Jacobson et al. 2001). These results indicate that the α-amylase gene phenotype is highly conserved between the sand fly and these parasites while suggesting a genotypic variation between the species of Leishmania. This study will be of particular interest to those who are researching the various aspects of parasite–vector co-evolution.

OPEN IN VIEWER

FIG. 2.

Characterization of different Leishmania species by primer combinations designed from the Phlebotomus papatasi α-amylase sequence, used in polymerase chain reaction reactions of genomic DNA isolated from Leishmania. Country of origin: IL, Israel; TM, Turkmenistan; IQ, Iraq; JO, Jordan; ET, Ethiopia; SD, Sudan; IN, India; TU, Tunisia. Source of strain: H, human; SF, sand fly; P, sand rat; D, dog. The products were compared by gel electrophoresis on 3% agarose gels. (Adapted from Jacobson et al. 2001.)

L. tropica was originally thought to cause the anthroponotic cutaneous epidemics in cities like Kabul in Afghanistan and Aleppo in Syria (reviewed by Jacobson 2003). However, micro-satellite analysis has shown that it is a much more complex organism, and while most Middle Eastern isolates cluster together, there is a wide geographical heterogeneity (Schwenkenbecher et al. 2006). L. infantum and L. donovani are the causative agents of infantile visceral leishmaniasis and Kala-Azar, respectively, but in recent years it has been reported that they have been found in CL (del Giudice et al. 1998).

Epidemiological surveys still tend to rely on the serological methods, such as fluorescent antibody and direct agglutination tests, leishmanin skin tests, and in-house ELISA and PCR methods to estimate prevalence of the disease in endemic areas for both humans and reservoirs (Doğan et al. 2008, Moshfe et al. 2009). Noninvasive techniques have been developed whereby urine samples are tested for excreted parasite DNA (Motazedian et al. 2008), which will be particularly useful for HIV/AIDS and Leishmania co-infections, a phenomenon that is on the rise (Pourahmad et al. 2009).

Cutaneous leishmaniasis due to L. major (ZCL)

The list of proven and suspected vectors and reservoir hosts of L. major in the Middle East is shown in Table 1. L. major usually causes a simple lesion that resolves in 3 to 6 months and is found in rural areas where the reservoirs and vectors abound. Afghanistan was known as one of the hyperendemic areas for ACL, but it has been recently revealed that ZCL is also very common (Faulde et al. 2008a). Reports from soldiers returning from Afghanistan have alerted the medical military teams' perception that both their soldiers and the local public were at risk. Faulde and his colleagues (2008a, 2008b) carried out extensive surveys to reveal the extent of the problem. They showed that ZCL had a different seasonality than ACL with the maximum number of cases of ZCL seen in September and October and that P. papatasi was the vector with R. opimus the reservoir, as expected. Some of the returning soldiers' lesions did not resolve so easily, and needed extensive treatment with hexadecylphosphocholine (Miltefosine) (van Thiel et al. 2010).

In Iran, ZCL is widespread, with a wide range of possible reservoirs and vectors. In central Iran, besides R. opimus, both Meriones lybicus and Meriones persicus have been found harboring L. major parasites and are thought to be secondary reservoirs (Parvizi et al. 2008b). Tatera indica and some Gerbillus sp. were trapped in the southern Iranian endemic region of Larestan, and were found positive for L. major by PCR (Mehrabani et al. 2007). An additional reservoir, Nesokia indica (short tailed bandicoot), was found positive in relatively large numbers in the northern district of Damghan in conjunction with a small number of Libyan jirds (Pourmohammadi et al. 2008). While P. papatasi is the proven vector of L. major in Iran (Azizi et al. 2010), it has recently been confirmed that Phlebotomus (Paraphlebotomus) caucasicus, Marzinowsky 1917, also harbors these parasites in the Shahrood district of central Iran (Rassi et al. 2008). This is an unusual finding as this subgenus is more associated with L. tropica infections.

In the other countries, P. papatasi is the proven vector and Ps. obesus is the main reservoir: Israel (Schlein et al. 1984), Jordan (Saliba et al. 1994, Janini et al. 1995), Saudi Arabia (Killick-Kendrick et al. 1985, Peters et al. 1985), and Syria (Rioux et al. 1992, Depaquit et al. 2008). This sand fly is ubiquitous and can be found from Portugal to the Indian subcontinent, manifesting molecular homogeneity (Depaquit et al. 2008), but is a nonpermissive vector, only transmitting L. major. The fat sand rat has a more restrictive desert habitat from Syria, Jordan, Israel and the West Bank, Saudi Arabia, and south and west to North Africa. In Iraq, while P. papatasi is the proven vector, the reservoir is only described as a gerbil (Coleman et al. 2007).

In Israel, both Meriones crassus and Gerbillus dasyurus have been found with the parasite (Wasserberg et al. 2002). Outbreaks are common in the country as the disease spread from the Jordan Valley area into the Negev and Arava regions and has now been reported from Sde Eliyahu (Beit She'an region) where, for the first time, voles (Microtus guentheri) have been found infected by L. major (Warburg et al. 2008).

Cutaneous leishmaniasis due to L. tropica, L. infantum, and L. donovani

The list of proven and suspected vectors and reservoir hosts of cutaneous leishmaniasis in the Middle East caused by species other than L. major is shown in Table 2. The L. tropica lesion, which can persist as an erythematous papule, unchanged for more than a year, and in hyperendemic regions, there can be multiple lesions on the exposed parts of the body. The lesions often resemble flattened volcanoes, becoming firmer as they heal and can last for >3 years. One unusual form of this disease in known as leishmaniasis recidivans, some times called lupoid leishmaniasis, due to its similar appearance to cutaneous tuberculosis. The lesions of these patients never heal spontaneously; often spread across the face and, even with treatment, may last many years. Another nonhealing variety has been described where the lesion continues as erythematous boggy nodules and plaques, which can be either active or quiescent, lasting decades and seen mainly in elderly patients.

ACL due to L. tropica has raged almost unchecked in Aleppo, Syria, for >20 years (Tayeh et al. 1997) and the extraordinary incidence rate of 29/1000 per year was reported from Kabul in Afghanistan (Reithinger et al. 2003). Another large outbreak occurred at Sanliurfa in eastern Turkey with over 11,000 reported cases (Ok et al. 2002). However, the vector was never found, but experimentally it was shown that Rattus rattus and P. sergenti from the area could maintain the cycle (Svobodová et al. 2003). In the Cukurova region of southern Turkey, hundreds of cases of cutaneous leishmaniasis occur every year, and it was discovered that the causative agent was L. infantum, not L. tropica, and the vector was Phlebotomus (Larroussius) tobbi, Adler, Theodor and Lourie 1930, which harbored identical parasites as those isolated from patients (Svobodová et al. 2009). In Hatay province, many of the human CL cases and some canine VL cases were characterized as L. infantum as analyzed by internal transcribed spacer 1 region—PCR followed by restriction fragment length polymorphism restriction enzyme analysis (Töz et al. 2009a).

In Israel, recent outbreaks of cutaneous leishmaniasis outside of the traditional desert habitats prompted a multitask force to investigate. The first major outbreak was 15 km east of Jerusalem in a new village that spread to a neighboring town with a population of 33,000. It was then established that P. sergenti was the definitive vector transmitting L. tropica, though the reservoir was unknown (Schnur et al. 2004). Subsequent studies revealed that the incidence rate in the Jerusalem district rose dramatically to 11/100,000 in 2005 and that hyraxes were suspected to be the reservoir (Singer et al. 2008). In fact, the reservoir of L. tropica in northern Israel, where an outbreak occurred in and around the city of Tiberias, had already been shown to be the hyrax Procavia capensis (Jacobson et al. 2003). This and subsequent studies showed that there were two distinctive transmission cycles occurring: one with Phlebotomus (Adlerius) arabicus, Theodor 1953, as the vector in the northern focus and the other with P. sergenti in the southern focus. The parasites from the two foci, though both L. tropica, were antigenically distinct by monoclonal assays and that the P. sergenti were refractory to the northern focus isolates (Svobodová et al. 2006). Sand flies of the subgenus Adlerius, though widespread in the Middle East, are notoriously difficult to identify, which may be the reason that this was probably the first time a sand fly of this subgenus had been definitively incriminated. In the ancient city of Jericho on the West Bank, a cluster of L. tropica cases has been reported in an area endemic for L. major (Al-Jawabreh et al. 2004). Although in the absence of a known vector (P. papatasi dominates the area and is refractory to L. tropica), it is difficult to know whether the patients were mainly immigrants who had brought the disease with them or a new unrevealed cycle has emerged.

In a recent prospective study of 155 cases of CL in the northwestern region of Yemen, PCR–restriction fragment length polymorphism analysis revealed that 85% of the cases were due to L. tropica, 11% were L. infantum, and 3.2% were L. donovani, confirming previous reports of predominantly viscerotropic species causing CL (Khatri et al. 2009). In Lebanon, where the prevalence of CL is as low as 0.18%–0.41%, one of the causative organisms appears to be from the L. donovani complex and yet putatively L. archibaldi (Guerbouj et al. 2001). However, the validity of this species has been challenged and it is more likely that L. donovani sensu lato is the parasite in question (Jamjoom et al. 2004).

In Jordan, Iran, Iraq, and Saudi Arabia, L. tropica is found with P. sergenti as the proven or suspected vector (reviewed by Jacobson 2003). There are indications that zoonotic transmission of the disease occurs in Bani Kinana and Salt districts in Jordan (Kamhawi et al. 1995, Saliba et al. 1997). Large numbers of ACL cases are reported from both population centers such as Bam and Shiraz and rural areas of Iran, and in Mosul in Iraq. In the Talil Air Base in South East Iraq, where thousands of American military personnel were at risk, Coleman and his colleagues conducted a comprehensive survey of sand flies and found that L. infantum and L. tropica were present in a few pools of sand flies tested by real-time PCR assays (Coleman et al. 2009). The vector species trapped were P. papatasi, P. sergenti, and Phlebotomus (Paraphlebotomus) alexandri, Sinton 1928, the latter a known vector of L. infantum. Most of the human cases were reported as CL and it is unclear whether any armed forces personnel had dermotropic lesions due to anything but L. major.

Visceral leishmaniasis

The list of proven and suspected vectors and reservoir hosts of visceral leishmaniasis in the Middle East is shown in Table 3. Infantile or Mediterranean VL in the Middle East is a disease of the young and poor and is often associated with dogs that die from the disease. The incubation for the disease can be 2 to 6 months with acute febrile periods with rigors, weight loss, and sometimes vomiting. This is followed by enlargement of the spleen, anemia, and leucopenia, and then abdominal pain, jaundice, and hepatomegaly, which can be fatal if untreated. In a region rampant with infectious diseases, it is known as a silent killer, as it can be easily confused with malaria, tuberculosis, typhoid fever, brucellosis, and other tropical diseases.

The causative agent of VL is nearly always L. infantum, although very rarely L. tropica has been implicated (Jacobson 2003). In a recent study from southern Iran, 63 patients were found to have VL caused by L. infantum, while in one child PCR analysis identified the parasite as L. tropica (Alborzi et al. 2006). In other regions of Iran, it is particularly common in Khuzestan (southwest), Ardabil (northwest), and Fars province (southeast), where 367 pediatric cases were recently reported (Ashkan and Rahim 2008). The scientists in that country have been very prolific and have found infected vectors such as Phlebotomus (Larroussius) perfiliewi transcaucasicus, Perfil'ev 1937 (Oshaghi et al. 2009b), and have isolated L. donovani from the M. persicus and L. infantum from Mesocricetus auratus, the golden or Syrian hamster (Mohebali et al. 2004). This species of hamster, though seldom seen in the wild, is often used as an experimental model for leishmaniasis research, and finding an infected one completes the circle started by Adler's experiments, >70 years ago (Adler 1938).

In Turkey, where VL is endemic, a new focus of both human and canine leishmaniasis has been reported from Hatay province, and another at Kağizman, eastern Turkey (Buyukavci et al. 2005, Töz et al. 2009b). While there is still no definite proof of the vector, P. transcaucasicus [sic] predominates in the endemic area of north central Anatolia, as previously reported in Iran (Ertabaklar et al. 2005).

In Israel, there are very few cases of human cases of VL (Ya'ari et al. 2004), but canine leishmaniasis is now a major problem, due to the rapid re-emergence of the golden jackal (Canis aureus) (Baneth et al. 1998). In the West Bank, infantile VL has become a serious problem, where in some villages the sero-positive rate for young children ranged from 8% to 21%. P. tobbi, Phlebotomus (Larroussius) syriacus, Adler and Theodor 1931, and Phlebotomus (Larroussius) neglectus, Tonnoir 1921, are the suspected vectors in both territories. Of the 56,000 Ethiopian immigrants who arrived in Israel since the mid-1980s, only 4 have been found to have VL due to L. donovani and are not therefore considered a public health risk.

The annual incidence of 3000 cases/year (mean 12.1/100,000, 95% CI 7.9–16.5) of VL was reported from the greater Baghdad area, central Iraq, after the First Gulf War in 1990–1991 (Anonymous 2003). Since then, the disease has spread southward to Thi-Qar, Basra, and Missan, increasing the risk to the population due to military operations, development of new agriculture and industrial projects, and unplanned and fast-growing urban development. The transmission season is May to October and P. alexandri is the suspected vector (Colacicco-Mayhugh et al. 2010).

Control

Attempts to control phlebotomine sand flies and reservoir animals in the Middle East have ranged from insecticide-treated bed nets (ITN) and curtains, residual spraying and fogging, and dog collars and environmental modifications. In Iraq, both public health officials and coalition forces made serious efforts to reduce the sand fly populations and protect the at-risk populations. Both indoor residual spraying and bed net distribution combined with a health education program were tried at Thi-Qar and resulted on a 55% reduction in cases of VL between 2003 and 2004 (Jassim et al. 2006). In the Tallil Airbase, U.S. military teams offered personal protection measures such as insect repellants, ITNs, and impregnated battle dress uniforms, but there was little compliance due to the severe heat, gusting sand, and lack of user's knowledge (Coleman et al. 2006). As part of the leishmaniasis control plan, ultra-low-volume insecticides and residual sprays were used in and around the tents in conjunction with rodent and canine control, which had little or no impact on the sand fly population (Coleman et al. 2006). However, it is thought that the high incidence rates of leishmaniasis from 2003 to 2004 (∼200/1000 on the Iran/Iraq border) have declined significantly, due to better housing and health education (Aronson 2007).

In northern Afghanistan integrated preventive measures were implemented in a German Army camp that included building a 3-m stone wall around the military base, stone paving to a distance of 100 m outside the camp wall, issuing impregnated battle dress uniforms, bed nets, and curtains. Rodenticides were employed using poison bait boxes, and both residual and nonresidual fogging was used for direct vector control (Faulde et al. 2009). This program was very successful that CL attack rates in the camp were nonexistent within a 2-year period.

In an urban setting in the Iranian city of Mashad, where ACL is endemic, deltamethrin-impregnated bed nets and curtains were distributed and combined with a health education program (Moosa-Kazemi et al. 2007). These nets provided good protection against sand fly bites, but there was no significant reduction in sand fly density. Similar trials of ITNs were tried in rural villages in the Aleppo Governate in Syria, which led to a reduction in ACL incidence in intervention villages compared with those where insecticide residual spraying was the only method of control (Jalouk et al. 2007). Successful use of ITNs and other impregnated household linens against sand flies was reported from Kabul, Afghanistan, with a 65% efficacy rate (Reyburn et al. 2000) and a decrease in incidence in Sanliurfa, Turkey (Alten et al. 2003). However, even with these reported positive outcomes, sand flies do tend to bite early in the evening, before people retire to bed, and the Middle East custom of sitting outside in the cooler evening climate may influence the exposure to the disease.

In the absence of a vaccine, the control of VL has been carried out either by the fruitless exercise of culling infected dogs or by the effective use of insecticide-impregnated dog collars. In a trial in Iran, nine villages where all domestic dogs received deltamethrin-impregnated collars were compared with nine control villages, where no collars were issued (Gavgani et al. 2002). These authors reported that sero-conversion rate for children and dogs in the intervention villages was lower than in the control villages, though they stress the human data should be interpreted with caution. It is interesting that the new canine vaccine Leishmune^®, which was so successfully used in Brazil (Palatnik-De-Sousa et al. 2009), has yet to be tried in the Middle East, where VL is still a fatal disease.

In Israel, research into control measures is ongoing and three methods, environmental modification, indoor repellants, and fine netting fencing, have distinct possibilities. Sand flies require sugar meals for energy, which they obtain from plants. This propensity was used to test whether certain plants, which contain natural toxins and attractive to phytophagous sand flies, would actually harm them. In a series of experiments it was found that in sites where mature ornamental Bougainvillea glabra plants grew, P. papatasi were eight times less numerous than in control sites without this plant (Schlein et al. 2001).

A deltamethrin-impregnated vertical net (Permanet) was tested in a section of a town where an outbreak of ACL had occurred (Faiman et al. 2009). The net was 60 m long and 2 m high and draped over a perimeter fence, and the lower fringe was anchored into the ground with soil. Although sand flies are able to pass through PermaNet (225 holes/in²), they must come in contact with the insecticide and this was the reason the authors reported 50%–60% reduction in sand fly numbers. In Israel, sleeping under bed nets is uncommon and therefore space repellants are more popular against insect pests. A comparison was made between a space repellant geraniol, a natural plant-derived product, and a diffusible insecticide, the synthetic pyrethroid prallethrin (Sirak-Wizeman et al. 2008). Prallethrin was highly effective in reducing the number of P. papatasi in bedrooms by 95% and P. sergenti by 59% in tents, whereas geraniol was found to be unsatisfactory as a space repellant for sand flies.