Cases and distribution of visceral leishmaniasis in western São Paulo: A neglected disease in this region of Brazil

Visceral leishmaniasis (VL) is one of the most prevalent parasitic diseases worldwide. In 2019, 97% of the total numbers of cases in Latin America were reported in Brazil. In São Paulo state, currently 17.6% of infected individuals live in the western region. To study this neglected disease on a regional scale, we describe the spread of VL in 45 municipalities of the Regional Network for Health Assistance11(RNHA11). Environmental, human VL (HVL), and canine VL (CVL) cases, Human Development Index, and Lutzomyia longipalpis databases were obtained from public agencies. Global Moran’s I index and local indicators of spatial association (LISA) statistics were used to identify spatial autocorrelation and to generate maps for the identification of VL clusters. On a local scale, we determined the spread of VL in the city of Teodoro Sampaio, part of the Pontal of Paranapanema. In Teodoro Sampaio, monthly peri-domicile sand fly collection; ELISA, IFAT and Rapid Test serological CVL; and ELISA HVL serum surveys were carried out. In RNHA11 from 2000 to 2018, Lu. longipalpis was found in 77.8%, CVL in 69%, and HVL in 42.2% of the 45 municipalities, and 537 individuals were notified with HVL. Dispersion occurred from the epicenter in the north to Teodoro Sampaio, in the south, where Lu. longipalpis and CVL were found in 2010, HVL in 2018, and critical hotspots of CVL were found in the periphery. Moran’s Global Index showed a weak but statistically significant spatial autocorrelation related to cases of CVL (I = 0.2572), and 11 municipalities were identified as priority areas for implementing surveillance and control actions. In RNHA11, a complex array of socioeconomic and environmental factors may be fueling the epidemic and sustaining endemic transmission of VL, adding to the study of a neglected disease in a region of São Paulo, Brazil.


Introduction
Visceral leishmaniasis (VL) is one of the most prevalent parasitic diseases worldwide. In 2017, 94% of new cases occurred in seven countries: Brazil, Ethiopia, India, Kenya, Somalia, South Sudan, and Sudan [1]. In 2019, 97% of the total numbers of cases in Latin America were reported by Brazil [2]. Currently, 25 of 27 Brazilian states notified autochthonous cases of canine visceral leishmaniasis (CVL) and 23 notified cases of human visceral leishmaniasis (HVL) [3].
In São Paulo, the richest and most populous state, Lutzomyia longipalpis sand flies were found in 1997, infected dogs in 1998, and infected humans in 1999. In 2020, 106 of 645 (16.4%) municipalities registered cases of HVL [4].
In recent years, from where and how VL reached the western region of São Paulo has been raised. Epidemiological studies demonstrated that the disease crossed the Bolivian border with Brazil and was found in the cities of Corumbá and Ladário, on the western side of the state of Mato Grosso do Sul (MS), expanded to its capital, Campo Grande, crossed the entire state and reached Três Lagoas, located on the western border of São Paulo. From MS to SP, the first diffusion occurred through the cities along with the construction of the Novoeste railway , the (BR-300) highway, completed in 1980, and the construction of a pipeline (1998) [5]. In São Paulo, CVL and HVL were first found in Araçatuba, a city harboring the railway, the highway, and the pipeline [6]. However, following the Candido Rondon highway (BR300), the disease crossed the state from the western region to the east toward the capital [6]. In the 45 municipalities of RNHA11 mesoregion, in the western region of São Paulo, in the border of MS, the first phlebotomine systematic collection was found in Dracena, a city located in the north and considered the epicenter [7]. In 2020, the western region harbored 537 of 3042 (17.6%) individuals infected with VL in São Paulo [4].
It has been hypothesized that in the western region, the presence of socioeconomic and environmental risk factors may favor the biological cycle of the disease, increasing the presence of Leishmania (Leishmania) infantum chagasi parasites, Lu. longipalpis vectors, and infected Canis familiaris, which in turn infect humans. Also, no study has been conducted on the dispersion of VL in the western region of São Paulo using an integrative approach so far. Our study is innovative because it allows the construction of a framework on two geographical scales: the regional scale combines public databases of the 45 municipalities of RNHA11 mesoregion; the local scale combines databases of Teodoro Sampaio, a representative city of RNHA11. To study the spread of this neglected disease in the 45 municipalities of RNHA11 on the regional scale, public databases on vectors, dogs, humans, environment, and socioeconomic risk factors were accessed. On a local scale, an entomological, canine, and human VL survey was conducted in Teodoro Sampaio, considered the capital of Pontal of Paranapanema.

Ethics statement
The project was approved by the ethics committee of Unoeste (number 4030). Canine samples were collected with the written informed consent of owners. Ethical approval was obtained from the Adolfo Lutz/Pasteur Institute Ethics Committee on the Use of Animals (number 02/ 2016). The study is for scientific research purposes and is in compliance with Federal Law no. 11794 of October 8, 2008, Decree no. 6899 of July 15, 2009, and in accordance with regulations outlined by The National Council for Animal Experimentation Control (Conselho Nacional para Controle da Experimentação Animal, CONCEA). and most important rivers in the southeastern region of Brazil, and is a natural border between São Paulo and Paraná states (Fig 1Bc). It is 930 km long and since the 1,980s, it includes 11 large artificial lakes and an extensive flooded area, supporting hydroelectric plants and bridges. These bridges link dozens of small towns and villages connected by a large number of highways ( Fig 1A).

Data on the presence of Lu. longipalpis in the municipalities of RNHA11
In RNHA11, data on phlebotomine sand flies identified in entomological collections were obtained from SUCEN in the period from 2000 to 2019. Entomologic surveys are proposed by the Manual of Surveillance and Control of Visceral Leishmaniasis of São Paulo (VLCPSP) [12], aiming to monitor the distribution of Lu. longipalpis in silent non-receptive, non-vulnerable municipalities, and silent non-receptive vulnerable municipalities as defined by the absence of confirmed autochthonous cases of HVL and CVL, without the presence of the chosen vector or not by estimated values of the distance to characterize the vulnerability. In these municipalities, an entomologic survey is recommended for the detection of the presence of Lu. longipalpis and to provide information on its distribution, allowing identification of the risk areas where control measures should be intensified [12].

Canine and human visceral leishmaniasis and Center of Zoonosis Control in the municipalities of RNHA11mesoregion
According to VLCPSP guidelines [12], in São Paulo state, the municipalities are divided into two groups: silenced for VL, that is, without confirmation of autochthonous human and/or canine cases; and with transmission, that is, with confirmation of autochthonous human and/ or canine cases. The investigation of a case of CVL may occur by a passive or house-to-house survey, conducted by the Center of Zoonosis Control (CZC) or by the Municipal Zoonosis Service (MZS) of each municipality. An active house-to-house search is conducted when Lu. Longipalpis is found in a particular area of the city or every time a dog is diagnosed with CVL; sequentially, all households within a radius of 200 m are visited and the dogs are tested for CVL. A passive survey is conducted when a dog's owner finds the symptoms described by the CZC or MZS as suspicious for CVL, as previously published [10]. Data on CVL, presence, and characteristics in each municipality under the CZC in the RNHA11 mesoregion were obtained from Adolfo Lutz Institute of Presidente Prudente [13]. A suspected case of HVL is confirmed if it meets one of the following criteria: laboratory criteria, that is, identification of Leishmania (Leishmania) infantum chagasi, from a culture and/or inoculation in hamsters and/or by molecular techniques; presence of Leishmania spp. in the direct parasitological examination; presence of anti-Leishmania (Leishmania) infantum chagasi antibodies in serological tests and clinical-epidemiological criteria, that is, a patient from the area of transmission of VL, with suggestive clinical signs and without laboratory confirmation, provided that other diagnostic hypotheses are ruled out, that presents a favorable response to therapeutic treatment. Data on patients infected with HVL in RNHA11 were obtained from the Epidemiological Surveillance Center of São Paulo, which organizes and publishes data on recorded infectious diseases in São Paulo state annually [4].

Mapping the cluster of identification of canine visceral leishmaniasis in RNHA11 mesoregion
In RNHA11 mesoregion, a Moran map was utilized to assess the formation of clusters of municipalities associated with the presence of CVL. Global spatial autocorrelation indicators, such as the Moran Index, present a single value as a measure of the spatial association for the data, which are presented by area, providing an index for the entire study region. However, it is often desirable to examine patterns on a more detailed scale to see if the process stationary hypothesis is verified locally, that is, if there is spatial dependence. Therefore, it is necessary to use spatial association indicators that can be associated with the different locations of a spatially distributed variable. This methodology uses the Local Moran Index (LISA) to find the spatial correlation of these areas. As it is a local indicator, there is a specific correlation value for each area, thus allowing the identification of clusters of areas and outliers [14,15]. Values vary from −1 to 1; values approaching zero indicate the absence of significant spatial autocorrelation of the values with their neighbors; values below 0.50 indicate a weak autocorrelation; values between 0.50 and 0.75 indicate medium autocorrelation; and values above 0.75 indicate strong autocorrelation. Positive values indicate a positive autocorrelation, where the value of what is being evaluated is similar to the values of its neighbors, and negative values indicate a negative autocorrelation [14][15][16].

Mapping the Human Development Index in the RNHA11 mesoregion
The Human Development Index (HDI) was used as an indicator for understanding the socioeconomic profile of the municipalities of São Paulo state. The HDI is a well-known and comparable indicator worldwide, reproducing studies in other states and regions countrywide [17]. The HDI value of the 45 municipalities in the RHN11 was obtained in Fundação Sistema Estadual de Análises de Dados (SEADE), from 1999 to 2016, coded and classified by the denominated quintile as a reference parameter. Thematic maps were created for the HDI using ArcGis 10.7.1 software. The HDI was analyzed with support in a predictive method of interpolation surface generation: local polynomial interpolation (LPI). The Gaussian kernel method was used to generate the surface, classified according to the quintile. The gain in the LPI analysis in relation to the global polynomial interpolation presupposes it overlaps in several concentrations that a spatial representation may have, in our study, through a point in the centroid city of São Paulo state.

Entomological survey in Teodoro Sampaio
In the urban area of Teodoro Sampaio, the phlebotomine sand fly entomological survey began in 2000, monthly, from February to October during the most favourable periods for detecting the presence of the vector. The survey ended in October 2010 when Lu. longipalpis was found in all areas. To standardize collections, the number (ranging from 1 to 40) and frequency (ranging from 1 to 39) of CDC light traps (Horst Ltd, São Paulo, Brazil) installed varied according to the density and infestation of sand flies. The traps were placed approximately 1 m from the ground in places susceptible to the presence of phlebotomies and/or in the shelters of domestic animals, with an average temperature above 20˚C and relative humidity >70%. In each sector of the different areas, four areas with risk factors for the presence of phlebotomies were selected, including a large peri-domicile, and areas with abundant vegetation, accumulation of organic matter in the soil, and the presence of domestic animals from which the female sand fly can potentially obtain an infected blood meal. The traps were operated between 17:00 and 07:00 h. Monthly, four traps were installed for three nights in each sector. After capture, nylon-mesh cages were placed in plastic bags, tagged, and refrigerated (1˚C-7˚C) until identification was performed by the Regional Entomology Laboratories of SUCEN, Coordination of Disease Control, Secretary of State of Health, located in Presidente Prudente, São Paulo state. The sand flies were processed and identified using the taxonomic key of Galati [18].

Canine visceral leishmaniasis survey in Teodoro Sampaio
The canine survey took place in four sectors of the urban area of Teodoro Sampaio, from January 2010 to December 2018, by a passive or house-to-house survey conducted by the Center of Zoonosis Control (CZC). We used recorded data of 3749 domiciled dogs of different ages, submitted to serum survey, using antibody tests to detect Leishmania, according to VLCPSP guidelines [12]. In Teodoro Sampaio, the Center of Regional Laboratory of the Adolfo Lutz Institute of Presidente Prudente (CRL-ALI-PPV) is responsible for the diagnostic tests for CVL. The VLCSP supervises the main actions to reduce morbidity and mortality, aimed at early diagnosis and treatment of human cases, vector control, identification, and euthanasia of seropositive domestic dogs. In 2009, the Ministry of Health began using enzyme-linked immunosorbent assays (ELISAs), produced by Bio-Manguinhos/Fiocruz, Ministry of Health, Rio de Janeiro, Brazil, for screening and confirmatory tests in CVL serological surveys. For ELISApositive samples, a confirmatory test using IFAT, produced by Bio-Manguinhos/Fiocruz, was performed. From April 2012 to the present, the Brazilian Ministry of Health replaced the ELISA/IFAT tests with a new protocol using a dual-path platform CVL rapid test, produced by Bio-Manguinhos/Fiocruz, with ELISA-CVL as a confirmatory test. Animals with positive samples in the screening test (ELISA from January 2009 to March 2012 and the dual-path platform CVL rapid test from April 2012 until the present and a negative result in the confirmatory test (IFAT and ELISA, respectively) were re-sampled. Only concordant samples that were positive in the screening test and positive in the confirmatory test were considered seropositive [10]. Data on the presence and characteristics in each municipality under the CZC in the RNHA11 mesoregion were obtained from Adolfo Lutz Institute of Presidente Prudente [13].

Human visceral leishmaniasis in Teodoro Sampaio
Data on human leishmaniasis was obtained from the Epidemiological Surveillance Municipality of Teodoro Sampaio. The sole case of HVL in Teodoro Sampaio was referred to the Regional Hospital of Presidente Prudente, a tertiary, university, public, and reference center for treatment of HVL in RNHA11 mesoregion. Confidentiality was assured. The case was confirmed in 2018 by clinical/ epidemiologic criteria and laboratory diagnostics according to the VLCPSP guidelines [12].

Human visceral leishmaniasis serological survey and laboratory diagnosis in Teodoro Sampaio
In July 2018, a serological survey was conducted on 159 individuals aged from 18 to 65 years from four different locations: a health care center, a Family Health Strategy Unit; the Social Assistance Reference Center; a State Technical School; and in individuals attending a private clinical laboratory. Blood was collected blood collections tubes under vacuum with and without EDTA as an anticoagulant. Serum was separated by centrifugation and stored at −200˚C until use. The analysis was performed at the CRL-ALI-PPV using a qualitative test for the detection of antibodies against HVL by ELISA, according to the manufacturer's instructions (LEISHMANIA ELISA IgG+IgM; VIRCELL, Santa Fe, Granada, Spain). The test uses recombinant antigens capable of fixing specific antibodies to Leishmania (Leishmania) infantum chagasi with a sensitivity of 97% and specificity of 99%. Positive tests were confirmed by indirect immunofluorescence (Bio-Manguinhos/Fiocruz, Ministry of Health, Rio de Janeiro, Brazil). Hemograms and hepatic enzymes were determined because hepatic function may be affected and pancytopenia may be present in patients infected with VL. Hemograms were performed using a flow cytometer flux counter (Pentra 80, HoribaDiagnostics, Montpellier, France) and the differential leukocyte count was compared with direct microscopic observation of blood smears. Alanine aminotransferase (ALT), aspartate aminotransferase (AST), and albumin were assessed using automated systems according to the manufacturer's instructions; the normal ranges were 7-55 U/L, 8-43 U/L, and 3.5-5 g/dL, respectively. A total of 16,636 individuals were considered to determine the sample size. These samples were processed by Laboratório Bioclínico, Teodoro Sampaio, SP, Brazil.

Kernel maps construction
Dogs that underwent the serological survey and those with positive serology were registered according to the location used by the municipal health surveillance service. The city was divided into four sectors, starting from the northeast of the urban area in a clockwise direction. These sectors, in turn, were divided into 432 blocks, arranged in a polygon shape, used to generate the Geographic Information System and served as a basis for indexing the attributes of the surveyed blocks. With the geometric points generated per block, it was possible to visualize the number of dogs with CVL in each block, the number of dogs in the census for each block, and the number of euthanized dogs per block. The number of positive dogs per block is not consistent with the number of euthanized dogs, because there were some cases in which euthanasia of dogs that tested positive was not performed. For the procedure that identified hot areas of CVL infection, we used the Kernel density maps, using the euthanized dogs as a population field to generate a continuous surface to the urban area of the Teodoro Sampaio. A raster surface was generated using the Gaussian equation, with a search ratio equivalent to a block (100 m) and a pixel size equivalent to a household (10 m). The coordinate system used as reference was WGS1984. Regarding the parameters for reading the data for analysis, the urban area of Teodoro Sampaio measures 4 × 3 km 2 .

Statistical analysis
The results are shown as means standard error of the mean (SEM) (for normally distributed variables). Dichotomous and nominal variables are expressed as frequencies and percentages. Pearson's correlation coefficient was used to describe the functional relationship between the number of cases of HVL per 10,000 inhabitants and the HDI of each municipality. Statistical analysis was performed using GraphPad Software (San Diego, CA, USA) and the Sigma-Stat program (Systat Software, Richmond, CA, USA). ArcGIS 10.7 and ArcGis Pro were used for data analysis and layout design.

Results
The presence of Lu. longipalpis in the municipalities of RNHA11 mesoregion  (Fig 2).

Spatial cluster analysis of CVL cases in RNHA11 mesoregion
The cases of CVL in the 45 municipalities of RNHA11 mesoregion during the period from 1999 to 2019 was totalled by the municipality. The Global Moran Index was applied and presented a weak spatial autocorrelation, but statistically significant (I = 0.2572). Applying the Local Moran Index Cluster Map (LISA) revealed that most municipalities (34 of 45; 75.5%) were non-significant, however, 11 of 45 (24.5%) showed significant spatial autocorrelation, distributed in the following quadrants: high-high with 5 municipalities Q (+/+), Presidente  Venceslau, Ouro Verde, Dracena, Tupi Paulista, and Junqueirópolis; low-low with 3 municipalities Q (−/−), Martinópolis, Taciba, and Presidente Bernardes; low-high with 2 municipalities Q (−/+), Piquerobi and Ribeirão dos Índios; and in the high-low quadrant, only one municipality Q (+/−), Presidente Prudente. In the LISA map, we can see the p value of significant municipalities (Fig 3). In the northern area, all the municipalities' notified cases, most of them at increased rates (17-157 cases), and Dracena (dark red), considered to be the epicenter, showed the highest rates (58-176 cases). On Pontal of Paranapanema, 8 of 32 (25.0%) municipalities registered HVL, and dispersion on the regional scale occurred in three different ways: irradiation, contiguity, and jumps (when municipalities far from the endemic region register cases of CVL or HVL). In municipalities located along the Paraná River, only one did not register cases (Fig 4A).

Paraná and Paranapanema Rivers, dams and artificial lakes in RNHA11 mesoregion and surrounding areas of Teodoro Sampaio
Figs 1A and 4A show that the municipalities in the RNHA11 mesoregion are divided from MS on the western side by the Paraná River, which flows along the border of five cities. On the south side, the region is divided from Paraná state by the Paranapanema River, which flows along the border of ten municipalities. Together, the rivers support seven large lakes and a flooded area of 6175 km 2 and seven dams and bridges (Fig 1A). These bridges link the VL endemic regions of Três Lagoas in MS and Andradina in São Paulo state to the areas of the western region as well as the municipalities of Pontal of Paranapanema to the VL-free areas of Paraná state. The urban area of Teodoro Sampaio (Fig 1Ba) is surrounded by the Devil's Hill State Park (Fig 1Bb), 6.3 km away, and linked by a highway with the administrative offices, museums, laboratories, and environmental educational areas of the park. The municipality of Teodoro Sampaio is served by both the Paraná and Paranapanema Rivers, which surround the urban area of the city at a distance of 1 km (Fig 1Bc). In the region, the average annual precipitation varies between 1200 and 1500 mm, and the average annual temperature remains slightly above 22˚C.

The Human Development Index in the RNHA11mesoregion
In the municipalities in this study, the HDI varied from very low in 4 municipalities and low in 15 municipalities to high only in Presidente Prudente, the most important city in RNHA11 mesoregion. Of the areas with very low and low HDIs, 18 (40%) of the municipalities are located in the Pontal of Paranapanema region. Ribeirão Preto, connecting the Paraíba Valley to the Metropolitan region of São Paulo city, and from Campinas to Piracicaba presents an axis of great development, with the HDI varying from 0.790 to 0.826. The regions with the lowest HDI values are the Ribeira Valley, located in the south of the state, and Pontal of Paranapanema (our study area). In Teodoro Sampaio, rates vary from 0.678 to 0.715. In São Paulo state, the HDI is distributed by regions, and maps are available in Fundação Sistema Estadual de Análise de dados (SEADE), Teodoro Sampaio is surrounded by very poor municipalities, Euclides da Cunha Paulista, Marabá Paulista, and Caiuá, with rates between 0.6770 and 0.704 (Fig 2).
When the functional relationship between the number of cases of HVL per 10,000 inhabitants and the HDI of each municipality was calculated, the correlation (adjusted R value) had a value less than 10% and the P value was not significant, however, when municipalities that did not have cases of VL were removed and the simple linear regression technique was applied, the regression line showed a trend for lower values of VL cases per 10,000 inhabitants as the HDI increased (Fig 5).

Entomological survey in the urban area of Teodoro Sampaio
Teodoro Sampaio covers two areas with regard to operational disease control programs (e.g. dengue fever surveillance and control, CVL and HVL surveillance and control). The two areas are divided into four sectors and the sectors are divided into blocks (Fig 6A and 6B). Lu. longipalpis were initially found in July 2010 in area 1, sectors 2 and 4 in blocks 146 and 298, and in October 2010 in area 1, sector 1, and block 75 (Table 1).

Canine serological survey and spatial distribution of euthanized dogs in Teodoro Sampaio
In 2010, CVL was identified for the first time in the urban area of Teodoro Sampaio. From January 2010 to December 2018, 3749 dogs were selected for serology by the CZC and 298 animals

Cases of human visceral leishmaniasis in Teodoro Sampaio
Teodoro Sampaio, 133 km from the epicenter, was one of the last municipalities to register cases and HVL was detected in 2018. The first autochthonous case of HVL was identified in 2018 in the eastern area, area 1, on the outskirts of Teodoro Sampaio (Fig 6B), where a high number of cases of CVL were found throughout the period. The patient was a 48-year-old man who did not respond to treatment with meglumine antimoniate and was treated with liposomal amphotericin B (5 mg/kg/day) for 5 days with improvement in the symptoms and 50% reduction of splenomegaly.

Human visceral leishmaniasis serological survey, laboratory findings and epidemiologic characteristics of the study population in Teodoro Sampaio
Between January and December 2018, 159 individuals living in Teodoro Sampaio were enrolled in the human visceral serological survey, and 2 of 159 (1.3%) resulted positive; 4 of 159 (2.5%) resulted indeterminate and 153/159 (96.2%) resulted negative. Hematologic abnormalities are among the most common manifestations in patients infected with VL in whom pancytopenia may occur (Table 2)

Discussion
In 2005, HVL reached the municipalities in RNHA11 mesoregion, western São Paulo, and in 2010 the municipalities of Pontal of Paranapanema, a poor and asymmetric region. In Teodoro Sampaio, considered the capital of Pontal of Paranapanema, Lu. longipalpis and CVL were found in 2010 and HVL in 2018. In 2018, 19 of 45 (42.2%) municipalities in RNHA11 recorded cases. A complex array of socioeconomic and environmental factors may be fueling the epidemic and sustaining endemic transmission of VL in this region (Fig 7).
The first Lu. longipalpis outbreak was found in the microregion of Dracena in 2003 [19]. One year later, vectors were found in Tupi Paulista and other adjacent municipalities, which suggested the existence of a possible route of dispersion by irradiation or by contiguity. However, Lu. longipalpis was registered in the region far from the epicenter, suggesting dispersion by jumps and toward the municipalities of the south, and HVL cases reached the municipalities of Pontal of Paranapanema in 2010 (Fig 4A and 4B). An extensive highway network (2382 km), linking endemic regions to small-and middle-sized cities that may function as potential routes of dispersion of vectors and infected dogs was demonstrated in RNHA11 mesoregion. We suggest that the main axis adding to VL dispersion is the Euclides Figueiredo highway (SP562). Starting in Ilha Solteira outside RNHA11, the highway goes down toward the cities of Andradina and Dracena and toward the municipalities of Pontal of Paranapanema, crossing Teodoro Sampaio at the border of Paraná state. Nevertheless, dissemination may also occur in lower rates by a secondary axis, SPV031 and SPV035, starting near Brasilândia, an endemic region in MS, and following the Paraná River through small-sized cities and ending in Teodoro Sampaio must also be considered. All the municipalities crossed by these highways became endemic areas of VL (Fig 4A). The potential role of highways in the spread of VL in the southeastern and southern states of Brazil has been pinpointed in recent years [5,6]. However, a new route has been described and has gained great importance. Connecting Foz do Iguaçu in the west, on the Brazilian side of the triple border and considered endemic for VL, to the port of Paranaguá in the extreme east, a large number of trucks carrying grain, fertilizers, fuel, and goods, circulate daily on BR 277, the most important highway in Paraná state. It was suggested that following BR-277 and transects, coming from Argentina and Paraguay, seropositive dogs and Lu. longipalpis vector spread to the neighboring cities, becoming a region endemic to VL. Furthermore, potentially it is likely an important route of dispersion of infected dogs, vectors, and parasites to other regions of Paraná state [20,21]. In the western region, the proliferation of vectors could be aggravated by the extensive watershed flowing into the Paraná and Paranapanema Rivers, supporting hydroelectric dams, bridges and artificial lakes [22][23][24][25][26]. Porto Primavera dam in Paraná River has an artificial lake with a flooded area of 2.250 km 2 and divides the municipalities of RNHA11mesoregion and MS. Dividing São Paulo and Paraná state, the Paranapanema River flows on the border of ten municipalities with increasing rates of VL to the VL-free areas of Paraná state. Artificial lakes modify the climate of their surrounding areas as well as the environment within which the lake interacts [22][23][24][25][26]. In Presidente Epitacio, a city bordering the lake of Primavera dam, higher yearly temperatures rates of 23.93±3.06˚C (95% CI 22.54-25.32) were found compared with 22˚C in municipalities far from the lakes [23]. The influence of artificial lakes of hydroelectric dams in Paranapanema River on the ecological aspects of the sand fly fauna in 3 different settings of an endemic area of CL in Paraná state was investigated. A wide dispersion of sand flies involved in the transmission of leishmaniasis was found in all the settings [22]. In Panorama, a city located on the border of the lake of the Paraná River, also an endemic area for VL, the high abundance of Lu. longipalpis had a significant correlation with temperature and humidity [27]. A relationship with climate and environmental factors with increased incidence rates of VL and annual precipitation, humidity, enhanced vegetation index, and night temperature values were found in the State of Tocantins, located in the northern region of Brazil [28].
In Teodoro Sampaio, although an entomological survey started in 2000, Lu. longipalpis sand flies were found only in 2010. It is well known that other phlebotomines, such as Lu. migonei, may be permissive vectors of VL in areas where Lu. longipalpis has not been recorded, but human or canine cases do occur [29][30][31]. According to the SUCEN database, in the period from 2000 to 2010, Lu. cruzi a VL vector found in Central-West of Brazil [32], and Lu. migonei were not identified in Teodoro Sampaio urban area. Lu. migonei was found in the wild forest of the Devil's Hill in research carried out for an LT investigation in 1994LT investigation in , 1995LT investigation in , and 1996.
Although a direct association between HDI and the presence of VL in municipalities of RNHA11 mesoregion was not found, linear regression showed a tendency of a lower number of infected individuals in municipalities with higher HDI. It is well known that poverty induces health care inequalities, poor housing conditions, lower income, less education, or lower occupational skill levels, and the population tend to be less healthy than those who experience higher levels of social conditions in other areas [16,32,33]. Supporting the data, demographic, environmental, and behavioral characteristics of inhabitants of rural settlements, a vulnerable population, have been determined. In Teodoro Sampaio, only 7.1% had a high school education, 11.9% lacked elementary instruction, and 77.3% had completed primary school. The income varied from less than 100 US$ (91.6%) to about 170 US$ (8.4%) per month. Similar rates were found in a neighboring rural settlement in Mirante of Paranapanema [34,35]. Globally and countrywide, VL is considered a neglected disease that affects the poorest of the poor [17,33,34].
In Teodoro Sampaio, on a local scale, the periphery of sector 1 had the higher number of infected dogs, demonstrated by critical point hotspots. The first human case of VL in the city was diagnosed in this sector. From 2010 to 2018, 298 (7.9%) dogs were found to be positive for CVL and 110 (36.9%) were euthanized. They were first found in sectors 1, 2, and 3 on the periphery, and in subsequent years, the hot spots spread to all urban areas. In the city, low levels of infected dogs were euthanized (36.9%). Similar to other cities of Pontal of Paranapanema, Teodoro Sampaio is surrounded by a high number of rural settlements [35,36]. Throughout the region, it is anecdotal that owners of dogs infected with VL hide their animals in these settings and may be disseminating the disease to rural areas. Although controversial, euthanasia is one of the most important actions supported by Brazilian Visceral Leishmaniasis Surveillance and Control Program in reducing the prevalence of Leishmania (Leishmania) infantum chagasi infection countrywide [37]. It has been shown that human epidemics of VL are usually preceded by or concomitant with CVL [10]. However, in Teodoro Sampaio, a diagnosis lag of 8 years between finding vectors, CVL, and HVL was found. Up to December 2020, only one individual had been registered. One possible explanation is that the city is located in the neighborhood of the Devil's Hill State Park, an endemic region for cutaneous leishmaniasis (CL) vectors [38,39], and the municipality is considered endemic for CL. There is some evidence that vectors have been present in the urban area for many years and a superposition CL/ VL is occurring and a cross-reaction may be on course, protecting people from VL infection [40]. Supporting these findings, a VL serological survey of 159 individuals living in the urban area was carried out and only 2 (1.26%) individuals were positive by ELISA, with a probable false positive and/or cross-reaction. In addition, the mean residence times in the city was 23.86 ±15.27 years, increasing the possibility of these individuals being immunized against CL. Although VL-infected individuals do not have hallmark signs and symptoms, severe pancytopenia (thrombocytopenia, leukopenia, and anemia) and increased levels of hepatic enzymes may be considered in the differential diagnosis in VL endemic areas [41]. In the population analyzed in this study, no laboratory findings symptomatic of chronic VL were found.
Shedding new light on the complex interactions occurring within the spread of VL in the western region and identifying the existence of spatial clusters of municipalities with increased vulnerability, Moran's I and LISA significance maps were created. Moran's I map identified a cluster of 11 municipalities with significant autocorrelation with CVL, and 5 municipalities were classified as high/high: Dracena, Ouro Verde, Tupi Paulista, Junqueirópolis, and Presidente Venceslau. Overall, 4 of these municipalities are considered the epicenter of the disease in which HVL was found in Dracena and Ouro Verde in 2005; Tupi Paulista in 2006, and Junqueirópolis in 2007 (Fig 4B). Furthermore, Dracena has a higher number of cases of HVL, varying from 58 to 176 infected individuals, followed by Ouro Verde, Tupi Paulista, and Junqueirópolis, varying from 17 to 57 infected individuals (Fig 4A). After all these years, despite the efforts made by the public authorities, Dracena remains the main focus of VL in western São Paulo. Taken together, these areas should be considered the target of local public policies and receive priority in surveillance actions within the scope of VL. As far as we know, this is the first effort to identify priority areas for VL in São Paulo State using Moran's I and LISA spatial analysis. In Brazil and worldwide, these tools have been used to identify vulnerable areas of parasitic, vector-borne neglected tropical diseases [14][15][16]42]. Moran's I and LISA techniques identified 27 priority areas for surveillance and control of VL in Belo Horizonte, an endemic area of Minas Geraes state [13]. In northeast municipalities, a VL Brazilian endemic region, spatial and space-time clusters of VL were identified in sertão and middle-north subregions, overlapping with high social vulnerability areas [14]. In Morocco, significant LISA clustering maps of human leishmaniasis have been identified in different provinces [42].
Different shortcomings should be considered in this study. The mean age of the population screened for VL was 43.50 years, and 72.33% were women. It is well known that VL affects mainly children and old people [40]. Thus, the population analyzed is not well representative of the disease. Municipal surveillance control programs of CVL were discontinued and there were difficulties in obtaining current records with a more accurate surveillance and response system to apply geospatial methods.

Conclusion
In RNHA11 mesoregion, in a regional scale VL is spreading with a progressive extension, from the north, the epicenter, to the south, from endemic to non-endemic areas. A cluster of municipalities in the epicenter showed a significant autocorrelation for CVL and should receive priority in surveillance actions within the scope of VL. Socioeconomic and environmental factors may be fueling the epidemic and sustaining the endemic transmission of the disease (Fig 7). These findings further highlight the asymmetry of the western region compared with other regions of São Paulo state and add new knowledge on the spread of VL, contributing to the study of a neglected disease in a region of São Paulo, Brazil.