Research Article: Complexities of Assessing the Disease Burden Attributable to Leishmaniasis

Date Published: October 29, 2008

Publisher: Public Library of Science

Author(s): Caryn Bern, James H. Maguire, Jorge Alvar, Simon Brooker

Abstract: Among parasitic diseases, morbidity and mortality caused by leishmaniasis are surpassed only by malaria and lymphatic filariasis. However, estimation of the leishmaniasis disease burden is challenging, due to clinical and epidemiological diversity, marked geographic clustering, and lack of reliable data on incidence, duration, and impact of the various disease syndromes. Non-health effects such as impoverishment, disfigurement, and stigma add to the burden, and introduce further complexities. Leishmaniasis occurs globally, but has disproportionate impact in the Horn of Africa, South Asia and Brazil (for visceral leishmaniasis), and Latin America, Central Asia, and southwestern Asia (for cutaneous leishmaniasis). Disease characteristics and challenges for control are reviewed for each of these foci. We recommend review of reliable secondary data sources and collection of baseline active survey data to improve current disease burden estimates, plus the improvement or establishment of effective surveillance systems to monitor the impact of control efforts.

Partial Text: Leishmaniasis comprises a complex of vector-borne diseases, caused by more than 20 species of the protozoan genus Leishmania, and ranging from localized skin ulcers to lethal systemic disease [1],[2]. Leishmaniasis is classified as one of the “most neglected diseases” [3], based on the limited resources invested in diagnosis, treatment, and control, and its strong association with poverty [4]. Published disease burden estimates place leishmaniasis second in mortality and fourth in morbidity among all tropical diseases [5].

The most common syndrome is localized cutaneous leishmaniasis (CL), most frequently caused by Leishmania major and L. tropica in the Old World, and L. braziliensis, L. mexicana, and related species in the New World [1],[15]. Spontaneous healing is the rule, but requires months to years, and varies by species [15]. Mucosal leishmaniasis (ML) usually occurs months or years after healing of primary CL, most commonly due to L. braziliensis, and can cause destruction of the nasal septum, palate, and other mucosal structures, leading to devastating facial mutilation and, rarely, death from airway involvement [16]. Other complicated forms include disseminated cutaneous leishmaniasis (DCL), diffuse nodular non-ulcerating disease, and leishmaniasis recidivans, localized slowly progressive non-healing lesions. Both are rare, difficult to treat, and can be severe.

The leishmaniases are transmitted to humans in sylvatic, domestic, and peridomestic cycles ranging from cities to deserts and rain forests on every continent except Australia and Antarctica (Table 1). Nevertheless, the human disease burden is relatively concentrated; 90% of VL cases occur in India, Bangladesh, Nepal, Sudan, Ethiopia, and Brazil, while 90% of CL occurs in Afghanistan, Algeria, Iran, Saudi Arabia, Syria, Brazil, Colombia, Peru, and Bolivia [26],[27]. The distribution is dynamic: Colombia and Ethiopia have recently joined this list, and Pakistan currently faces a large epidemic of CL in Baluchistan and Sindh (World Health Organization [WHO], unpublished data). Climate change and other environmental changes have the potential to expand the geographic range of the vectors and leishmaniasis transmission in the future [28].

The most objective measures of disease burden are incidence, prevalence, and mortality. Several derived measures incorporate indicators of disease severity, disability, and/or quality of life into composite outcomes that can be compared across diseases [33]. Currently, the most widely used measure is “disability-adjusted life years lost” (DALY) [5],[33]. Leishmaniasis DALY estimates are based on (1) figures assumed for regional incidence and prevalence, (2) assumed case-fatality rates, and (3) assigned disability weights for CL and VL [33]. For leishmaniasis, there are major uncertainties and sparse documentation for the assumptions underlying all three of these components. The empirical basis and derivation of global and regional leishmaniasis incidence and prevalence figures have not been documented since 1991 [34],[35]. Passive surveillance is generally given as their basis, but leishmaniasis is notifiable in only 33 of 88 endemic countries [26]. Substantial underreporting is widely acknowledged [26], but its magnitude has rarely been measured, and in studies, has varied from 2-fold to 40-fold [36]–[38]. In many countries, the majority of leishmaniasis cases are treated by non-governmental organizations or in the private sector, but these cases are not usually included in surveillance data, exacerbating underreporting. Underreporting is likely to vary greatly, not only among countries and depending on the clinical syndrome, but even between localities in the same district, based on distance to health care, availability of private providers and of antileishmanial drugs, the presence of research groups, and local awareness of the disease. Differential underreporting is labor-intensive to document, and precludes valid generalization of incidence and reporting rates presented in research studies. Underreporting of deaths is even more pronounced. One study from Sudan estimates that 91% of all kala-azar deaths went unrecognized [39], while data from a village-based study in India suggest that as many as 20% of VL patients, disproportionately poor and female, died before their disease was recognized [40]. The origin and derivation of the disability weights assigned for VL and CL are not documented [33]. In the absence of data to address these shortcomings, this article will describe the characteristics of leishmaniasis that contribute to the disease burden in the most important foci, and aspects that should be taken into account in future attempts to quantify its impact and monitor control programs.

In global estimates, the Horn of Africa (Sudan, Ethiopia, Kenya, Somalia) accounts for the second largest number of annual VL cases, after South Asia [41]. Transmission dynamics are complex, involving parasites identified by standard laboratory techniques as both L. donovani and L. infantum[42],[43], and two distinct ecological settings, semi-arid regions in the north where Phlebotomus orientalis is the major vector, and the savanna and forest areas in the south where P. martini and P. celiae are found in association with Macrotermes termite mounds [44],[45]. Investigators have suggested that VL originated in the Sudan, based on the ancestral position of the circulating parasites in genetic analyses [46],[47]. While sporadic sylvatic VL transmission is well recognized [48], sustained peridomestic and domestic cycles in villages, and explosive epidemics affecting populations displaced in recent wars, account for the bulk of human cases [42],[43],[49]. In zoonotic foci, both sylvatic rodents [50] and domestic dogs [51] may act as infection reservoirs, but large outbreaks are usually thought to involve anthroponotic transmission.

If Sudan is the original home of VL, South Asia is its domestic heartland. In the 19th century, devastating outbreaks of a chronic progressive febrile illness with cachexia, hepatosplenomegaly, and high fatality rates were reported in Bengal and Assam, and retrospectively thought to be the first recorded VL epidemics [19]. In 1903, Leishman and Donovan first described the organism that now bears their names in patients infected in India [67],[68]. Today, South Asia is estimated to account for 60% of the global VL disease burden [41], with a sustained endemic focus stretching from Bihar and Bengal in northeastern India, across the border into southeastern Nepal, and to the east into central and western Bangladesh. The parasite in South Asia is transmitted by Phlebotomus argentipes, an endophilic vector that rests in human and animal dwellings in densely populated agricultural villages. Kala-azar incidence fell substantially during the indoor residual insecticide spray campaigns of the malaria eradication effort of the 1950s and 1960s, but the disease returned in the 1970s and transmission has been sustained since then [69],[70].

From Mexico to Argentina, L. infantum (synonym L. chagasi) is transmitted from dogs to humans primarily by Lutzomyia longipalpis, a vector well adapted to the domestic and peridomestic environment. Brazil accounts for 90% of reported VL cases in the Americas, and is the third most important VL focus globally [82]; unlike other major VL foci, case reporting is mandatory in Brazil, and surveillance data are more complete. However, American VL is in the midst of dramatic changes in transmission patterns and a marked geographic expansion, superseding previous estimates of disease burden.

At least 12 different Leishmania species cause American CL, and the disease occurs in every country from the United States to Argentina, except Uruguay and Chile. Until recently, Brazil and Peru reported the first and second highest incidence in the Americas [15],[26]. However, with more than 15,000 reported cases in 2005 and 2006, Colombia now ranks second after Brazil (>30,000 per year); thousands of cases occur in Peru (∼6,500 per year) and elsewhere in Latin America as well [15],[89],[94],[95]. The epidemiology of CL in the Americas is complex, with intra- and inter-specific variation in transmission cycles, reservoir hosts, sand fly vectors, clinical manifestations, and response to therapy [96]. Studies often demonstrate five or more species causing lesions in the same area [97]–[99].

Of the major forms of leishmaniasis, the only historically urbanized form is anthroponotic CL due to L. tropica, as illustrated by its vernacular names, “Baghdad boil”, “Aleppo boil”, “Balkh sore”, and others [46]. Although infections in dogs and other animals have been documented, the disease is characterized by large outbreaks in densely populated cities, especially in the setting of war and large-scale population migration. In Syria, especially the traditional focus in the city of Aleppo, a marked increase to more than 15,000 cases per year was documented during the 1990s, with only a temporary decline when insecticide spray programs were instituted in 1991 [109]. A huge CL epidemic has occurred in Afghanistan since 1992, with estimates of 200,000 cases in Kabul alone [110]. The annual CL incidence in Kabul peaked at 12% in 1996, and averaged 3% per year from 1992 to 2002 [29],[62]. The association between migration and CL transmission may be more complex than originally postulated: while migrants within Kabul were at the highest risk of CL, possibly because of economic disadvantage, immigrants from outside Kabul were at no higher risk, but appeared to fuel local transmission by adding to the pool of susceptible residents [62]. Transmission occurred within the household, even up to second floor apartments [111], and often resulted in facial lesions, especially in women and children [62]. Women with lesions were considered unfit to marry, have children, or breastfeed, and children with lesions were sometimes ostracized by playmates [112].

Current methods of assessing disease burden fail to take into account the clinical and epidemiological diversity of leishmaniasis, and the intense medical, social, and economic impact within highly affected foci. Furthermore, existing passive surveillance data are grossly inadequate to be used to make reliable estimates. Active, rigorous assessments of the true incidence, morbidity, mortality, current transmission patterns, and non-health effects of leishmaniasis are urgently needed. The following steps are recommended to achieve better estimates of current disease burden, and establish systems to monitor the impact of control measures:



Leave a Reply

Your email address will not be published.