Mem Inst Oswaldo Cruz, Rio de Janeiro, Vol. 89(3):435-443, jul./sep. 1994 435

The Dermal leishmaniases of Brazil, with SpecialReference to lhe Eco-epidemiology of lhe Disease in

Amazonia

R Lainson, JJ Shaw, FI Silveira, AAA de Souza, RR 8raga, EAY Ishikawa

Seção de Parasitologia, Instituto Evandro Chagas, Caixa Postal 691, 660 17-970 Belém, P A, Brasil

Six species of Leishmania are at present known to cause cutaneous andjor mucocutaneousleishmaniasis in Brazil, and they are alI to be found in lhe Amazon region of this country. 11reeco-epidemiology of each is discussed, with lhe observation that lhe Amazonian leishmaniases are alIzoonoses, with their source in silvatic mammals and phlebotomine sandfly vectors. With man'sdestruction of lhe natural forest in southern Brazil, some sandfly species have survived by adapting toa peridomestic or domiciliary habitar in rural areas. Some domestic animais, such as dogs and equinesare seemingly now involved in lhe epidemiology of lhe disease. No such process has yet been reportedin lhe Amazon region, but may well taie place with lhe continuing devastation of its forest.

Key words: Leishmania (Vwnnia) braziliensis -Leishmania (Vwnnia) guyanensis -Leishmania (Vwnnia)lainsoni -Leishmania (Vwnnia) shawi -Leishmania (Vwnnia) naiffi -Leishmania (Leishmania) amazonensis-

eco-epidemiology -Amazonia -Brazil

Until the Iate 1960's, alI fonos of dennalleishmani~is in Brazil were attributed to thesingle species Leishmania braziliensis. At thattime, however, epidemiological studies were in-itiated at the Instituto Evandro Chagas, in Belém,State of Pará, entailing the examination of mànypatients and a wide variety of wild mammals andphlebotomine sandflies (Psychodidae:Phlebotominae). They soon indicated thepresence of another species of Leishmania infect-ing man in the Amazon region of Brazil whic~by virtue of its biological and biochemicalresemblance to Leishmania mexicana, the causalagent of "chiclero's ulcer" in Central America,was given the subspecific name of Leishmaniamexicana amazonensis Lai~n & Shaw, 1972.

Over the ensuing years, and largely as a resultof these studies, it became clear that there ex-isted, in fact, a multiplicity of different neotropi-cal Leishmania species. M~t of them have beenshown to cause human dermal leishmaIÚasis inone fonn or another, with each showing its ownpeculiar ecological and epidemiological charac-teristics.

reservoir -sandfly vector -mano More detailedaccounts have previously been published el-sewhere (Lainson 1982, Anon. 1984).

Hurnan infection: detection and isolation of theparasite

Microscopic diagnosis -Clinical aspects canbe deceiving, and differential diagnosis is re-quired for dennallesions due to a variety of otheraetiological agents. There is no more sure diag-nosis than seeing and isolating the parasite, but ininexperienced hands, or in poor laboratory condi-tions, the detection of amastigotes in stainedsmears mar be sadly inefficient. Badly preparedsmears, containing a host of contarninative bac-teria, fimgal spores, dust particles and stainingdeposit, mar result in the failure to flnd Leish-mania ar, even worse, the registration of a falsepositive. Points to remember are the use of ultra-clean and dry slides, avoidance of the necrosedcentral part of ulcers, adequate cleansing of theskin surrare before making a biopsy or scrapingfrorn the rnargin of the lesion, avoidance of ex-cessive blood in the srnear, and rapid drying ofthin srnears for immediate fixation in absolutemethanol which has not been allowed to absorbatmospheric moisture. Giernsa stain is that rnostwidely used, but the quality of staining mar varywith the mnd: buffered distilled water should beat 7.0 -7.6 pH, and free of algal contaminants.Placing the slides face down in flat stainingdishes, rather than uprlght in Coplin jars, willavoid staining deposit if the surface scum ispushed to one side.

MATEIUALS AND METHODS

Before discussing the eco-epiderniology ofArnazonian denl1alleislunaniasis it is appropriateto f~t outline the methods used in elucidatingthe "epiderniological triangle" of wild animal

Receiv~ 31 January 1994Aocepted 2 May 1994

436 Le;shman;a sp. in Amazonian Brazil .R lainson et ai.

In vitro culture -Parasites may be so scanty,especially in old lesions, that they defy detectionin stained srnears. Diagnosis will now depend onthe culture of tissue juice ar biopsied skin frag-ments from the border of the lesion in a suitableblood-agar medium, and the inoculation of suchmaterial into laboratory animaIs. Isolation of theparasite for subsequent identification is of utmostimportance in eco-epidemiological studies: theuse of monoclonal antibodies and DNA analysisfor immediate identity of Leishmania species onglass slides or other substrates may hold prornisefor the future, but these sophisticated techniquesare still not fully adapted to most field conditionsand beyond the means of rnany laboratories indeveloping countries.

Growth of different Leishmania species mayvary greatly in different culture media, andworkers must ascertain those which are mostsuitable for the parasites likely to be encounteredin their study areas.

Isolation in laboratory animais -The Syrianharnster remains the animal of choice: the mouseis much less susceptible to some leishrnalÚas,particularly to those of the braziliensis complexoAspirated tissue juice ar a saline triturare of biop-sied tissue is usually inoculated intraderrnallyinto the skin of the Dose or feet, and the animaIsexamined periodically for the appearance of skinlesions. Some Leishmania species will producean inapparent skin infection, or one which is of atransient nature. For this reason at least twoharnsters should be inoculated, so that cultoresmay be made from the inoculation-site of oneafter a few weeks, while the other(s) are retained.A year or more may elapse before some parasitesproduce a visible skin lesion.

11Ie intradermal ("Montenegro") skin-test -This widely used skinítest is highly specific andof great use in quantitative epidemiology. Indiag~is, however, the following points must beborne in mind: (a) a positive reaction may be dueto a past infection and have nothing to do withthe present lesion ofthe patient; (b) positive reac-tions may be registered in persons with nopresent ar past evidence of infection, and (c) asmall percentage of parasitologically positivepatients rnay produce no reaction. The test isregularly negative in cases of diffuse, anergicleishrnalÚasis.

into the descriptive labelling of that parasite,using the coding rneth<xl recommended by theWorld Health Organization (Anon. 1984).Strains should be p-eserved in liquid N2 as soonas possible after isolation, and deposited in one ormore intemational reference centres.

Pin-pointing the sandfly vector

This mar be relatively simple in regionswhere there are few sandfly species, but is clearlymore difficult in areas where there are large num-bers of anthropophilic species -as in the Amazonrain-forest. In our study area on the Serra dosCarajás, State of Pará, for example, we haveregistered 25 species of sandflies known to biteman! Logical steps in determining which of theseare vectors in the transmission of the variousLeishmania species to man are as follows: (a) ac-curately pin-pointing the exact area of transmis-sion by carefully questioning patients, (b) es-tablishing the time of year when transmission isat its peak, and (c) indicating all theanthropophilic species of sandflies in the area.Initially this will necessitate human bait, but onceman-biters are recognized alternative methodsshould be used, including such well-tried devicesas the Shatmon and C.D.C. light-traps. When thehabits of the suspected vectors are known theymar be collected from their resting-sites, such astree-tnmks and animal burrows.

The final process of incriminating the vectorof each Leishmania species in the area willdepend on detecting, isolating and identiíying theparasites of infected sandflies during mass-dis-sections of the captured insects. When the studyarea is lar from borDe-base this, and the inocula-tion of the flagellates into hamsters and culturerores, must be dane in the field.

With luck, infected flies mar be encounteredduring the initial dissections made to establish thecomposition of the local sandfly population (it is,incidentally, easier to recognise the species of afresh1y dissected female sandfly than it is to iden-tiíy specimens flXed in alcobol). If, bowever, thisis done when the population is at its peak thechances of fmding infected flies are slim, becausea high proportion of the sandfly catch will benewly hatched, nulliparous females. Muchhigher infection-rates are found when the popula-tion is declining, at the onset of dryer weather.

One swallow does not make a swnmer, andone infected fly does not conclusively indicare avector. Dissections and isolations must contin~,to find a significant number of infections in agiven sandfly species and to demonstrare heavy.promastigote infection in the anterior part of thefly's intestine. Concluding evidence will be ob-tained by indicating an intimate reservoir-host/sandfly/man contact, and by experirnentally

Identification, description and preservation ofLeishmania isolates

Correct identification of the parasite byrecognized biological and biochemical criteria(Lain.;on & Shaw 1987) is imperative Ü1 eco-epidemiological studies, especially where a nurn-ber of different leislunanias are circulating in na-ture. ~ original code of the laboratory in whichan isolare is rnade should always be incorporated

Mem Inst Oswaldo Cruz, Rio de janeiro, Vai. 89(3), jul./sep. 1994 437

transmitting lhe parasite by lhe bite of lhesuspected vector species.

Incrirnination of the wild mamJnalian reservoir-host(s)

Entornological data on the biting habits andpreferential hosts of the sandfly vector will giveclues as to the most likely wild mamJnalian hosts:canopy dwellers and sandflies resting on treetrunks are likely to derive their infection from ar-boreal animaIs, and a vector found only at groundleveI clearly indicates a terrestrial reservoir.Analysis of fresh bloodmeaIs from sandflies ispotentially useful, but must be interpreted withcare because many of these insects have verycatholic feeding habits. Finally, the screening ofwild animais should always include the examina-tion of both skin and viscera, and it should beremembered that most Leishmania speciesproduce a benign, inapparent infection in theirnatural mamJnalian hosts. Skin-snips frorn theears, Dose, tail and feet are triturated in saline andinoculated intradermally into one pair ofhamsters, and a similar suspension of liver andspleen intradermally into two others. Culture ofskin and viscera fragments follows the sameprocedure as that used for the isolation of Leish-mania frorn mano

A seventh species, Leishmania (Leishmania)deanei Lainson and Shaw, 1977, of the porcupineCoendou, has yet to be found infecting mano

Leishmania (Y:) braziliensis

Geographical distribution: what appears to bethe same parasite has been reported from theStates of Bahia, Ceará, Espírito Santo, MinasGerais, Pará, Paraná, Rio de Janeiro, Rondônia,São Paulo and southem Amazonas. The exactdistribution of the parasite remains doubtful,however, due to past inadequate methods of iden-tification.

Known mammalian hosts: man -among wildanimais, L (Y:) braziliensis sensu lato has beenrecorded from the rodents Proechimys,Rhipidomys, Oryzomys. Akodon and Ranus, andthe marsupial Didelphis. In domestic animais in-fections have been registered in dogs and equi-fies, principally in the extensively deforestedareas of southeast and northeast Brazil.

Recorded sandfly hosts: in the primary forestof Serra dos Carajás, Pará, the principal vector isundoubtedly Psychodopygus wellcomei. This is ahighly anthropophilic fly, biting man avidly atnight and, in overcast weather, during the day. Itis active only during the rainy season (ap-proximately November to April), and enters intodia pause during the dry season. In southeastBrazil, in the States of Rio de Janeiro and SãoPaulo, the vector is most probably Lutzomyia in-termedia, with possible involvement of Lu.migonei and Lu. pessoai in São Paulo. In MinasGerais and some northeastem States (Bahia andCeará), Lu. whitmani has been incriminated.Contrary to Ps. wellcomei, the latter species haveadapted to a peridomestic or even intra-domiciliary habitat, doubtless as a result ofman'sdestruction oftheir natural silvatic habitat.

Defining the eco-epidemiology of L. (Y:)braziliensis is at present rendered most difficultby the occurrence of what appears to be the sameparasite in areas of very different ecology, andwhere the sandfly fauna is quite different. Ps.wellcomei, for example, is absent in manylowland areas of Pará, yet many isolates of Leish-mania from cases of cutaneous and/ormucocutaneous leislunaniasis in these regions ap-pear to be L (Y:) braziliensis (or very similarvariants or serodemes, as indicated by isoenzymeand monoclonal antibody identification). Further-more, parasites recorded as H L braziliensisw, wL.braziliensis braziliensis W or H L braziliensis sensu

latoH are reported in geographic regions as farapart as Belize in Central America, and the Stateof São Paulo, Brazil! Clearly, the sandfly andmammalian CauDas and the ecology of suchregions are very different, and we are faced withthe choice of one of two explanations: either L(Y:) braziliensis is exceptional in having a wide

Quantitative studies

These shou1d accompany the qua1itativestudies, to provide infonnation on the preva1enceand incidence of the human disease due to thedifferent Leishmania species of a given loca1ity,and the eco10gy and popu1ation fluctuations ofboth the sandfly vectors and the reservoir-hosts -alI of which are a basic necessity before any con-tro1 measures can even be contemp1ated.

~ULTSTo date we have indicated the presence of six

species of Leishmania causing human cutaneous1eishmaniasis in the Amazon region of Brazil. Inrecent pub1ications (Lainson & Shaw 1987, 1992)we have allocated tive of these to the subgenusVlannia Lainson and Shaw 1987, and one to thesubgenus Leishmania Saf'janova 1982, as fo1-10ws:

Leishmania (Vlannia) braziliensis Vianna, 1911Leishmania (Vlannia) guyanensis Floch, 1954Leishmania (Viannia) lainsoni Si1veira et aI.1987Leishmania (Vlannia) shawi Lainson et aI. 1989Leishmania (Viannia) naiffi Lainson and Shaw,1989Leishmania (Leishmania) amazonensis Lair1sonand Shaw, 1972

438 Leishmania sp. in Amazonian Brazil .R lainson el ai.

range of totally unrelated sandfly vector species,or we are dealing with an extensive complex ofrelated species and subspecies, for which ourpresent metlK>ds of separation are inadequate.

Leishmania (J-:) guyanensis

Geographical distribution: the Amazon basin,north of the river Amazonas, in Amapá, northPará and Amazonas, possibly extending intoRoraima: in neighbouring countries, Guyana,Surinam, French Guyana and Colombia. Rarecases are registered from Pará, south of the riverAmazonas.

Known mammalian hosts: man -the principalwild mammalian hosts are the two-toed sloth,Choloepus didactylus and lhe lesser anteater,Tamandua tetradactyla. Secondary hosts arefound in marsupiais (Didelphis) and, more rarely,rodents (Proechimys).

Recorded sandJly hosts: the major vector isLu. umbratilis, a forest canopy and tree-tnmkdwelling sandfly. This insect is found, often iRenormous numbers, resting on the tronks of thelarger trees, in primary forest, from where itreadily attacks man when disturbed. Transmis-sion of L (J-:) guyanensis to man is, therefore,principally during the daylight hours, especiallyin the early morning, and men engaged in cuttingand clearing primary forest are at particularlyhigh risk. As sloths are rather sedentary animais,tending to stay in a given spot for some time, theinfection-rate of Lu. umbratilis on a tree harbour-ing an infected animal mar reach very highproportions. The simultaneous bite of numerousinfected flies, disturbed from their resting-site, isthe most likely explanation for the frequency ofmultiple lesions in cases of leislunaniasis due toL. (J-:) guyanensis.

Occasional infections have been recorded inLu. anduzei, which is also a canopy and tree-tnmk dweller and perhaps has some role as aseCOndary vector. Infections reported in Lu. whit-mani sensu lato (Lainson et ai. 1981) wereprobably those of L (J-:) shawi. (There is nowstrong evidence that wLu. whitmaniw from Pará isdifferent frorn the type material of this speciesfrom Bahia).

The enzootic of L (J-:) guyanensis, as seen inprimary, forest is unlikely to survive in secondaryforest or man-made plantations, where there arefew or no large trees. The microhabitat on treesof small girth is unfavourable for the sandfly vec-tor, doubtless due to low humidity and lhesmooth surface of the small tree-tronks. In addi-tion, such trees are not a favourable environrnentfor a large and relatively heavy animal such asthe sloth. In plantations of pine, gmelina andeucalyptus, this animal is deprived of its normaldiet of foliage and fruits of native trees.

A higb prevalence of cutaneous leishmaniasisdue to L (~) guyanensis rnay be associated withhuman habitations in or very near to primaryforest, leading to the impression that the sandflyvector has adapted to a peridomestic or domi-ciliary habitat (Arias & Naiff 1981). There is noevidence as yet, however, that Lu. umbratilis isable to dispense with its arboreal tire, and suchcases of leishmaniasis are ahnost certainly theresult of infected sandflies being attracted tohurnan dwelling pIares from the neighbouringforest, presumably by the lights of houses atnight. Clearing the forest to a distance of about400 m arolmd a village surro\mded by primaryforest, in French Guyana, was found to entirelyeliminate peridomestic transmission to man (LePont & Pajot 1981).

Finally, although infection of the marsupialDidelphis has been found to be very rare or ab-sent in prirnary forest where there is intensivetransmission of L (~) guyanensis (Arias et aI.1981, Lainson et aI. 1981), a high prevalence hasbeen folmd in those captured in the peridomestichabitat associated with nearby primary forest(Arias & Naiff 1981). Reasons for this are uncer-tain: neither is it clear as to whether or not theseinfected opossums serve as an effective source ofparasites for sandflies, or if they merely representdead-end, accidental hosts.

Leishmania (~) lainsoni

Geographical distribution: until now thisparasite has been recorded only from the State ofPará, Brazil, but it doubtless exists in otherregions where the sandfly vector and reservoir-host occur together.

Recorded mammalian hosts: man -amongwild animaIs, it has so far been found only in the"paca", Agouti paca (Rodentia: Dasyproctidae).

Known sandfly host: Lutzomyia ubiquitalis.Steps in unravelling the eco-epidemiology of

L (~) lainsoni have made an interesting story,commencing in 1981 when we isolated a Leish-mania from a case of cutaneous leishmaniasis ac-quired in forest near the Pirelli rubber plantation,about 20 krn from Belém, and referred to it simp-Iy as an "unnamed parasite of the subgenus Vlan-nia" (Lainson & ShaW 1987).

In 1983 we isolated another Leishmania froma single specimen of the sandfly Lu. ubiquitalis,captured in forest in the foothills of the Serra dosCarajás, Pará, and placed it in the same category.

One year later, working on the doomed Islandof Tocantins, destined to be submerged beneaththe Tucuruí Reservoir formed by the newhydroelectric dam, we made isolates fromnurnerous specimens of A. paca. Once again, theparasite was considered merely as an unidentifiedmember of the subgenus: Vlannia (unpublishedobservations).

Mem Inst Oswaldo Cruz, Rio de janeiro, Vai. 89(3), jul./sep. 1994 439

likely that other arboreal animais mar be in-volved.

Recorded sandfly hosts: Lutzomyia whitmanis.l. (There is now strong evidence that "Lu. whit-mani"

from Pará is different from the typematerial of this species from Bahia).

In 1948 Leorúdas Deane noted the presenceof amastigotes in spleen and liver smears from atwo-toed sloth, C. didactylus captured in Abae-tetuba, Pará. He remained Wlcertain, however, asto whether they belonged to a Leishmania speciesar represented a stage in the life-cycle of En-dotrypanum (an intra-erythrocytic flagellate ofsloths).

What was probably the same parasite waslater isolated from the viscera of C. didactylus,from Acará, Pará, by Lainson and Shaw (1972).This time, however, there was no doubt regardingthe organism's inclusioll in the genus Leish-mania, as it was studied in both hamsters andblood-agar cultures. It was considered to "...differfrom L b. braziliensis..." and to be "...closest toL. b. guyanensis..." (Lainson & Shaw 1979,1987).

In 1988 we recorded similar infections in theskin of C. apella and Ch. satanas (Cebidae), theviscera of a three-toed sloth, B. tridactylus, andthe skin of a coatimWldi, N. nasua, ali fromprimary forest in the Serra dos Carajás, Pará. Acomparison with L. (V:) braziliensis, L. (V:)guyanensis and L (V:) panamensis, usingmonoclonal antibodies and isoenzyme profiles,led us to the conclusion that the organism was"...a hitherto undescribed para si te of thebraziliensis complex", and it was given the nameof L (V:) shawi (Lainson et ai. 1989).

Incrimination of the sandfly Lu. whitmani s.l.followed after the biological and biochemicalcharacterization of 14 heavily infected flies. Twoof these were caught in a C.D.C light-trap baitedwith a caged C. apella, and two in a Shannontrap, ali in the same study area in which the in-fected animais were obtained. The remaining teninfections were in specimens of Lu. whitmani s.l.caught in light-traps ':>1 with human bait, inanother of our work areas in Tucuroí, Pará.

It was clearly now important to complete thethird point of our epidemiological triangle -didL (V:) shawi infect man? Accordingly, we tookfrom our cryobank some isolates of Leishmaniamade from rnan. and which we had previouslyfound to be similar to, but not identical with, L(V:) guyanensis. The isoenzyme profiles ofeighteen of these showed twelve to be indistin-guishable from the type strain of L (V:) shawi:very slight differences in the PEP profiles of theother six led us to regard these merely as isoen-zyme variants of the same parasite (Shaw et ai.1991).

Later on. nmnerous isolates of a Leishmaniamade from patients coming to the InstitutoEvandro Chagas had aroused our interest by thepeculiar morphology of their amastigotes andpromastigotes and, fo11owing the observation thatthe parasite's isoenzyme profiles distinguished itfrom a11 Other known species of Leishmania, theorganisms was named L (V:) lainsoni (Silveira etaI. 1987).

A major break-through carne in 1990, whenwe investigated an area of forest where one ofour patients had acquired bis infection with L(V:) lainsoni: among our sandfly catches wefound eight specimens of Lu. ubiquitalis heavilyinfected with this parasite (Silveira et aI. 1991b).The jigsaw puzzle was fina11y to become adefinite picture, however, when we compared theisoenzyme profiles of a11 our old and new isolatesof Leishmania from Lu. ubiquitalis and A. pacawith our type species of L (V:) lainsoni fromman, and found them to be indistinguishable (Sil-veira et aI. 1991a,b).

Seemingly, then, we had completed our eco-epidemiological study of this parasite -but forone puzzling feature that remained to be ex-plained. We had Dever recorded Lu. ubiquitalisbiting man in the forest and, we were forced toadmit, therefore lacked the most crucial evidencein incriminating any art.hropod vector of a humandisease! Consequently, we began a study of thebehaviour of Lu. ubiquitalis, both in thelaboratory and in the forest.

In the ftrst place, we found that this sandflywould feed on man, quite avidly, when broughtfrom the forest and maintained in the laboratoryfor some hOUlS. This suggested, then, that this flymust also bite man in the forest, under certainconditions -which we had yet to define. Fina11y,any doubts of this were removed when a fu11y en-gorged Lu. ubiquitalis was eventua11y taken whilebiting the arm of a man who was, appropriatelyenough, standing in the same spot in the forestwhere we caught our first infected Lu. ubi-quitalis, nearly ten years previously (Lainson etaI. 1992).

Clearly, however, this sandfly is not par-ticularly fond of hmnan blood, which explainsthe relatively low rate of infection with L (V:)lainsoni in man, compared with that of L (V:)braziliensis and L (V:) guyanensis, both of whichhave highly anthropophilic sandfly vectors.

Leishmania (V:) shawi

Geographic distribution: at present, onlyknown from Pará, Brazil.

Recorded mammalian hosts: man -reservoir-hosts among the forest animais include monkeys(Cebus apella and Chiropotes satanas), sloths(Choloepus didactylus and Bradypus tridactylus)and the coatimundi (Nasua nasua). It is quite

440 Le;shman;a sp. in Amazonian Brazil .R lainson et ai.

The steps leading up to the elucidation of theeco-epidemiology of L (Y:) lainsoni and L (Y:)shawi clearly emphasise the enonnous impor-tance of religiously cryopreserving alI isolates ofLeishmania from sandflies, wild or domesticanimais and mano

Leishmania (V:) naiffi

Geographic distribution: tlús has been poorlystudied, but till now isolates appear to be lirnitedto the States of Pará and Amazonas.

Recorded mammalian hosts: man -amongwild animaIs the parasite is known only from thenine-banded armadillo, Dasypus novemcinctus.

Known sandjly hosts: Psychodopygus paraen-sis, Ps. ayrozai and Ps. squamiventris have alIoccasionally been found infected. Wlúch of theseis the principal vector, however, remains doubt-fuI.

Amazonas, and what appeared to be the sameparasite from a human skin lesion, based onbiological characters of the organismo

Fina1ly, Lainson et aI. (1990) confmned thatthe parasite recorded from man by Naiff et aI.(1989) was indeed L (V:) naiffi, and registeredanother human infection with this parasite aftercomparison of the isolates with the type strainfrom D. novemcinctus, by way of monoclonal an-tibodies and isoenzyme profiles.

L (V:) naiffi rarely produces a discernableskin lesion in the hamster, and we feel that forthis reason human infection with this organismmar frequently have been missed in the past,when inoculation of this animal with lesion-biop-sy material has been the sole method of isolationattempted.

Nonetheless, judging from our isolates ofLeishmania from man in Pará during the past 25years, L (V:) naiffi does not appear to be a veryimportant pathogen of mano Either because themajor sandfly vector is not very anthropophilic(Ps. ayrozai'l), or because the parasite commonlyproduces only an occult, benign infection in theskin of man, as it does in the hamster.

Leishmania (L) amazonensis

Geographic distribution: typically the Ama-zon basin, but rare records exist from the Statesof Bahia and Paraná.

Known mammalian hosts: man -the most im-portant wild animal hosts are rodents (particular-Iy species of Proechimys and Oryzomys), but theparasite is also commonly found in marsupiais(Didelphis, Philander, Marmosa, Caluromys andMetachirus). We have recorded a single isolatefrom the fox, Cerdocyon thous.

Recorded sandfly hosts: the principal vector isLu. jlaviscutellata. Infections have less frequentlybeen found in Lu. olmeca nociva, a very closelyrelated fly of the olmeca complex, to which Lu.jlaviscutellata belongs: it mar piar a secondaryrole within the enzootic.

L (L) amazonensis is extremely common inAmazonia, in a wide range of terrestrial animais.Furthermore, it occurs in ali manner of woodland,including primary high forest, varzea, swampforest, and any low, dense secondary growthresulting from deforestation. The enzootic haseven spread into man-made forests such as plan-tations of pine and gmelina, due to the readyadaptation of rodents, marsupiais and Lu. jlavis-cutellata to such habitats.

It is fortunate, indeed, thatthe sandfly vectoris not greatly attracted to mano If it were, theproblem of human cutaneous leislunaniasis inAmazonia would be very much greater than it al-ready is: not only in terms of sheer numbers ofcases, but due to the fact that L (L) amazonensisis the cause of "diffuse, anergic cutaneous leish-

The nine-banded arnladillo for long enjoyed afolkIoric reputation as the fundamental source ofcutaneous leislunaniasis in the Amazon region.The reasons for this are obscure, but probablyhave something to do with the frequent presenceof phlebotomine sandflies in their burrows. TheTupi indian name for the annadillo is "tatu", andthat for the sandfly, "tatuquira" -literally, the "ar-madillo-fly". Unfortunately for this hypothesisthe sandfly species living in annadillo holes arenot those which commonly bite man, and the caseagainst D. novemcinctus remained somewhatthin.

In 1979, however, we added some substanceto the myth when we isolated a Leishmania inblood-agar cultures made from the liver andspleen of an annadillo obtained from primaryforest in Monte Dourado, north Pará (Lainson etai. 1979).

In spite of the uni que biological andbiochemical nature of the parasite it did notreceive its name until ten years later (Lainson &Shaw 1989), and there still remained no finnevidence that it was the cause of humancutaneous leislunaniasis. In that year, however,Menezes et aI. (1989) did record the isolation of aparasite from the skin lesion of a man corningfrom the Vale do Rio Trombetas region in Pará,which had an isoenzyme profile "...similar to thegroup of seven zymodemes of isolates from man,armadillos and sandflies from the Amazonregion...". They gave no details of these,however, and described the organism as havinground, vacuolated amastigotes which were"...larger than any recognized species of the sub-genus Vlannia". This is not at ali suggestive of L(11:) naiffi, which has amastigotes which are con-siderably smaller than most members of the sub-genus.

Naiff et ai. (1989) registered the presence ofL (11:) naiffi in D. novemcinctus from the State of

Mem Inst Oswaldo Cruz, Rio de janeiro, Vai. 89(3), jul.jsep. 1994 441

TABLEIIAmazonian phlebotonú11e sandflies fowKl with

Leishmania or promastigotes of doubtful taxonomy

Psychodopygus amazonensis Lutzomyia anduleiPs. chagasi Lu. anlunesiPs. paraensis Lu. dendrophilaPs. s. squamivenlris Lu.furcataPs. s. maripaensis Lu. gomeziPs. wellcomei Lu. h. hirsuta

Lu. shawiLu. tuberculata

Lu.vuilli

Today, in Bahia, ~ Gerais, Rio de Janeiro,São Paulo and other affected regiDos, somesandfly species survived by adapting to aperidomestic or even intradomiciliary habitat inrural areas, and their reservoir of infection marnow aIso include some domestic animaIs such asdogs and equines, which are frequently foundwith leishmanial skin lesions. It is very likely,however, that the primitive source of lhe parasitestilllurks in lhe degraded forest remnants, whichstill harbour a variety of rodents and marsupiaIs.The predominant surviving leishmanial parasiteis L (JI:) braziliensis s.l. Many other Leishmaniaspecies were doubtless doomed to extinctiontogether with their sandfly vectors, and it remainsto be seen if this same process will ultimately berepeated in Amazonia, following man's con-tinued devastation of its great forest (Fig.).

ACKNOWLEDGEMENTSTo the Wellcome Trust, London, for their unstinted

support during 27 years of research on leishmaniasis inAmazonia by the Wellcome Parasitology Unit (1965-1992). Also to the Instituto Evandro Olagas, FwtdaçãoNacional de SalXie, whose collaboration made thesestudies possible.

maniasis" in individuaIs with a faulty cell-mediated immunological system. This disease isoot only highly mutilating, but also incurable byour present methods of treatment.

DISCUSSIONFrom this brief account of the Leishmania

species known to cause dennal leishmaniasis inPará and other parts of Amazonian Brazil, we seethat these diseases are essentially zoonoses, withreservoirs in a wide variety of silvatic mamrnals,among which theY are transmitted by differentspecies of phlebotomine sandflies. Transmissionis by no means an haphazard, indiscrirninate af-Cair, however: there exist distinct associationsbetween certain mamrnalian and sandfly species,in specific ecological niches, and governed bycertain environrnental factors.

The number of Leishmania species or sub-species circulating in the Arnazonian fauna isanybody's guess. Certainly we have but scratchedthe surface, and rnany more than we have dis-cussed here will be discovered in the future.Tables I and n list marnmals and sandflies inwhich we have recorded unidentified leishmanialparasites, or promastigotes which are mostprobably those of Leishmania. We either failed toisolate these organisms in blood-agar culture andhamsters, or the isolates were lost before theycould be characterised.

In Amazonia, the cutaneous leishrnaniasescan be regarded as occupational diseases, con-tracted principally by forest-workers engaged insuch activities as the constroction of highways,hydroelectric dams, or other projects requiringthe penetration and cutting of native forest. Atrisk, on a smaller scale, are topographers, geo-logists, botanists, zoologists, hunters, -and theoccasional tourist!

This predominantly silvatic ecology of thedifferent Leishmania species was doubtless thesituation throughout Brazil and most other partsof Latin America before the lberian colonization,and the relentless deforestation that followed.

T ABLE I

Arnazonian manunals fowtd wilh Leishmania ofdoubtful taxonomy

Rodentia

REFERENCESAnon. 1984. The leishma"iases: report of a WHO e;x-

pert committee. Tech Rpt Series 701: WHO,Geneva, 1984, 140pp.

Arias IR, Naiff RD 1981. ~ principal reservoir hostof cutaneo~ leishmaniasis in the Urban areas ofMana~, Central Amazon of Brazil. Mem /nst 0$-waldo Cruz 76: 279-286.

Arias IR, Naiff RD, Miles MA, Souza AA 1981. ~opossum, Didelphis marsupialis (Marsupialia:Didelphidae), as a reservoir host of uishmaniabraziliensis guyanensis in the Amazon Basin ofBrazil. Trans R Soc Trop Med Hyg 75: 537-541.

Deane LM 1948. Abstract of discussion on CAHoare's paper The relationship of the haemoflagel-Iates. Proc 4th Int Congr Trop Med Malar,Washington OC, 10-18 May.

Aoch H 1954. uishmania tropica guyanensis n.ssp.,agent de Ia leishmani~ tégumentaire des Guyanas

Marsupialia

Rhipidomys leucodacrylusProechimys spp.Rattus rattus

Didelphis marsupialisPhilander opossumTamandua tetradacrylaHomo sapiensAn unidentified bat

Xenarthra

Primates

Chiroptera

442

\{

-Cl.

Cl.

C/)

~-z~~~C/)-W-J c::of-UL.LJ>~L.L-~Z<t:

C/) zof-~f-C/)

LJ.Jc::oU

-LJ.J1:::1

u.o(/).--J=

>(/)LJ.Jc::

>1

~~I U>

31

Leishmania sp. in A

mazonian

Brazil .R

lainson

et ai.

o-I-' c::o

>'--1-'

--I-'u)aJU

aJ~

:JL--L-O--I-"!-c::

U)

:JaJ>'

"C

L-U

co

-aJE-1-'>

-O

-L-O

>

c.

N

L-C

:::J'!-W

U)O>1~-.JI

~I ~~31

-Q)

s:::cuQ

)"O~-JI'-"~

I ('"-

CIJ

CIJ

UL-~li-

31

<n

L ('".

L- o

L-Ic.

o ~

c.

<nc:-

<1)

<n

"""<1)

CO

E

""'<n>

~

O

L- C

OC

: U

<

1) N

<

1) E

::]

-.cOC

OC

O-~

L-L-::] <

1) O

>

- C

O cr

.o

3: C

O

c. <

n ::]

"C.c:

<nl

CO

<

nI<

n1

<nI

::]1

a Q

..a..a..-J

<n

<n

c:<1)

c:O'I-

<n

N

'I- c:

CO

L-

CO

C

O

.o c:

-...

~I

~I

~I

...-.JI

-.J I -JI

>-

L-<

1) (íJ

>"0<

1)-c

E>

O

O

L- U

U

~

<1)

<1)

(/) (/)

.c

-I-' L-

-I-'..c

O

.c C

'.01

O1U

-"O--

E

<1) ~

-I-'

"O

(/)U

(íJ

<1)

-L- -E

-1-'01-1-'00<

1)(/)"00"0<

1)-N

L-L-

CC

O<

1)W

-II-Q.

-::::~'

~~ -~

(tj+

-'(tjQ

)+

-'::JU(/)

>(tj

4-

31 I:: I

Q)

""'1::3:C

U"O

""'E

V"O

::)

3:1:: O

ucuu

UJ

..+-'

Q)

+-'

UJ

UJ

>U

JI::Q)

Q)

o E

>

L O

L O

+-,"O

::) cu

UJ

+-,L

>-I::Q

)U

LCU

o.""'C

U

+-'"O

o.Q)

01:: E

OO

Q)O

NU

"OU

I:: Q

) C

U

Q)

LJ.JUJE

.o

-JI---

~I

j.

"-~.

...

I

31 ~(/)

(/)cuc=C

UE

)-C

(/)

(/)

cu Q

)

CU

-CU

C

U

~

L--Q

)-JI

u~

(/)

->-J I '-'

I-I.J(/)C

l)~o'I-

coC

l)"C

O-I.J-I.J

co--I.Jcot::-

-I.J0.co

-co-I.J-I.J-Co."C

CO

UC

l)"C-

>

co -I.J

-(/)>

>.

Cl)

~

.Q

E::)

O(/)

t:: "C

o-(/)-~C

O

-I.J C

l):I:

CO

o.

Ê.

'\\

~o]..~§=.g~.B~"Oo~'t~ao=]ao~~.g~1~~~>>

.C1..~1~.~'".~~~~~.~~ã.22!~ci)tt

~~

o

-Jo'(,)

w-J'<

tW~o

m~

~,<

t ~

1 -,

~.

~.

." ~

~~

~,. "o

.:

W.J~~~

~

:-.; '

>

_"-.~W

>

-=--

.J'<-,.

-;:;

O~

20=>

UO

w~"

442Leishm

ania sp. in Am

azonian B

razil .R

lainson et ai.

442

Mem Inst Oswaldo Cruz, Rio de janeiro, Vol. 89(3), jul./sep. 1994 443

et de r Ameriq~ Centrale. Arch Inst Pasteur GuyFranc 15: 1-4.

Lainson R 1982. Leislunanial parasites of manm1als inrelation to human disease, p. 137-179. In MA Ed-wards & U McDonnell (eds), Animal disease inrelation to animal conservation: Symposia of theZoological Society of 1.ondon, 50. Academic Press,London.

Lainson R, Shaw JJ 1972. Leislunaniasis in the NewWorld: taxonomic problems. Brit Med Buli 28: 44-48.

Lainson R, Shaw JJ 1977. Leislunanias of neotropicalporcupines: Leishmania hertigi deanei nov subsp.ActaAmazônica 7: 51-57.

Lainson R, Shaw JJ 1979. The role of aJÚmals in theepidemiology of South American leishmaniasis, 1-116 In WHR Lurnsden & R Killick-Kendrick (eds)Biology of lhe Kinetoplastida Vol 2, AcademicPress, 1.ondon.

Lainson R, Shaw JJ 1987. Evolution, classification andgeographic distribution, 1-120 In W Peters & RKillick-Kendrick 11ze Leishmaniases in Biolagy andMedicine Vol 1 Biology and Epidemiology,Academic ~, 1.ondon.

Lainson R, Shaw JJ 1989. Leishmania (Vlannia) naiffisp.n. a parasite of the annadillo, DasypUS novem-cinctus (L.) in Amazonian Brazil. Ann Parasit HumComp 64: 3-9.

Lainson R, Shaw JJ 1992. A brief history of the genusLeishrnania (Protozoa: Kinetoplastida) in theAmericas, with particular reference to AmazonianBrazil. Ciência e Cultura 44: 94-106.

Lainson oR, Braga RR, de Souza AAA, Póvoa MM,Ishikawa EA Y, Silveira FT 1989. Leishmania(Viannia) shawi sp.n. a parasite of monkeys, slothsand procyonids in Amazonian Brazil. Ann ParasitHum Comp 64: 200-207.

Lainson R, Shaw JJ, Braga RR, Ishikawa EA Y, deSouza AAA, Silveira Fr 1988. Isolation of Leish-mania from monkeys in the Amazon Region ofBrazil. Trans R Soc Trop Med Hyg 82: 231.

Lainson R, Shaw JJ, Ready PD, Miles MA, Póvoa MM1981. Leishmaniasis in Brazil XVI. Isolation andidentification of Leishmania species from sandflies,wild manunals and man in north Pará State, withparticular reference to Leishmania braziliensisguyanensis, causative agent of Hpian-boisH. Trans RSoc Trop Med Hyg 75: 530-536.

Lainson R, Shaw JJ, Silveira Fr, Braga RR, IshikawaEA Y 1990. Cutaneous leishmaniasis of man due toLeishmania (Vlannia) naiffi Lainson & Shaw, 1989.Ann Parasitol Hum Comp 65: 282-284.

Lainson R, Shaw 11, de Souza AAA, Silveira Ff,Falqueto A 1992. Further observations on Lut-zomyía ubiquitalis (Psychodidae: Phlebotominae),the sandfly vector of Leishmania (Vwnnía) lain-som Mem Inst Oswaldo Cruz 87: 437-439.

Lainson R, Shaw JJ, Ward RD, Ready PD, Naiffi RD1979. Leishmaniasis in Brazil XIU. Isolation ofLeishmanía from armadillos (Dasypus novem-cinctus), and observations on the epidemiology ofcutaneous leishmaniasis in north Pará State. TransR Soc Trop Med Hyg 73: 239-242.

Le Pont F, Pajot F-X 1981. La leishmaniose en Guyanefrançaise 2. Modalités de Ia transmission dans unvillage forestier: Cacao. Cahiers ORSTOM Sér EntMéd Parasitol18: 359-382.

Menezes JA, Brazil RP, Momen H, Grimaldi G 1989.Atypical human leishmaniasis d~ to a parasite ofthe Leishmania braziliensis complex in northBrazil. Trans R Soc Trop Med Hyg 83: 429.

Naiff RD, Freitas RA, Arias JR, Barrett TV 1989.Aspectos epidemiológicos de uma Leishmanía detatus (Dasypus novemcinctus) Prog Resum XICongr Bras Parasitol, p. 24.

Safjanova VM 1982. ~ problem of taxonomy withLeishmanía, p. 5-109. In The Leishmanias, SerProtozool Soviet Acad Sci. Science (Leningrad) 7:(in Russian).

Shaw JJ, Ishikawa EA Y, Lainson R, Braga RR, Sil-veira Ff 1991. Cutaneous leishmaniasis of man d~to Leishmanía (Vwnnía) shawi Lainson, de Souza,Póvoa, Ishikawa & Silveira, in Pará State, Brazil.Ann Parasitol Hum Comp 66: 243-246.

Silveira Ff, Lainson R, Shaw JJ, Braga RR, IshikawaEA Y 1991a. Leishmaniose cutânea na Amazônia.Isolamento de Leishmanía (Vwnnía) lainsoni doroedor Agouti paca (Rodentia: Dasyproctida), noEstado do Pará, Brasil. Rev Inst Med Trop S Paulo33: 18-22.

Silveira Ff, Shaw JJ, Braga RR, Ishikawa E 1987.Dermal leishmaniasis in the Amazon Region ofBrazil: Leishmanía (Vwnnía) lainsoni sp.n. a newparasite from the State of Pará. Mem Inst OswaldoCruz 82: 289-292.

Silveira Ff, de Souza AAA, Lainson R, Shaw JJ,Braga RR, Ishikawa EA Y 1991 b. Cutaneous leish-maniasis in the Amazon region: natural infection ofthe sandfly Lutzomyía ubiquitalis (Psychodidae:Phlebotominae) by Leishmanía (Vwnnía) lainsoniin Pará State, Brazil. Mem lnst Oswaldo Cruz 86:127-130.

Vianna G 1911. Sôbre uma nova espécie de Leish-mania (Nota preliminar). Bras-Méd 25: 411.