Leishmaniasis is a vector-borne disease affecting humans and animals caused by more than 20 species of the protozoan Leishmania (1). They are endemic in 92 countries, where more than one billion people are at risk of infection (2). Depending on the region, Leishmania is transmitted by different phlebotomine sand fly vectors (3). There is no human vaccine (4, 5) and the prevalence of leishmaniasis is closely linked to the human development index and environmental degradation (6, 7). This zoonosis remains a serious public health problem and is recognized as one of the 20 major neglected tropical diseases (8). The World Organization for Animal Health also lists leishmaniasis among the diseases that require worldwide notification (9). Visceral leishmaniasis (VL) is the most fatal form of leishmaniasis in humans caused by Leishmania donovani/Leishmania infantum (Old World) and Leishmania chagasi (same as L. infantum) in the New World (10).

In southern Europe, Africa, Asia, and America, L. infantum causes zoonotic leishmaniasis in humans (VL) (11–13) and in domestic dogs (CanL) (14, 15). Leishmania infantum infection has also been detected in wild animals (16, 17) and other domestic species including cats, horses, cattle, sheep, and goats (18–20). Natural infection and disease have been described in domestic cats to a lesser extent than in dogs (21–23). A recent meta-analysis including 36 years of cross-sectional studies from 12,635 cats in Leishmania endemic areas worldwide found an estimated seroprevalence of 11% and an overall PCR positivity of 10%. At least 63% of the infections were caused by L. infantum (24). Other studies detected 3–4% of cats with L. infantum in Italy, Greece, and German by serology (25, 26). Those recent studies have demonstrated that Leishmania is also circulating and cats and this should be taken as an alert in the regions where this parasite occurs.

In South America, the epidemiological cycle of L. infantum is complex and involves wild and domestic vertebrate and invertebrate hosts (27–30) as a result of social/economic inequality and environmental changes (31). CanL is a severe zoonosis because the dog is highly susceptible exhibiting elevated parasite loads and disease. Besides, it is very close to humans acting as a competent domestic reservoir (32, 33). In Brazil, seroprevalence in dogs and cats ranges from 2.4 to 32% (34) and 0.7 to 30.48%, respectively (35–37). However, the role of cats in L. infantum epidemiology is still in its infancy and should be more explored.

Recent investigations have focused on domestic cats' potential in L. infantum transmission (22, 38). Standardized serological tests for cats and a more detailed knowledge on their immune responses against L. infantum infection are needed (39) since coinfections may predispose cats to develop feline leishmaniasis (FeL) and other diseases (18, 40). These relevant aspects would improve the pivotal role of cats as L. infantum hosts in the context of One Health. As part of a wider study on leishmaniasis epidemiology, natural feline L. infantum infection from VL and CanL endemic areas is reported. Also, the clinical presentation and natural evolution of an infected cat is provided in detail.

This is the first study to report the infection by L. infantum in cats in the state of Bahia, Brazil, where 7.8% of cats were infected by L. infantum. The finding of L. infantum DNA by PCR in the bone marrow of the tested cats indicating parasite circulation simultaneously with CanL and VL that have been notified in the last 12 years (44, 65, 66). Comparatively, in other endemic Brazilian regions for both VL and CanL with similar socioeconomic and environmental characteristics, the frequency of L. infantum infection in cats ranged from 4.0 to 31.9% suggesting that it my vary depending on the region (21, 37, 67, 68).

Brazilian health authorities declared one of the studied municipalities (Feira de Santana) as an area of intense VL transmission, and an average of 20.2 human cases have been reported from 2014 to 2018 (42). Endemicity of zoonotic leishmaniasis in that municipality had been assessed in a former study reporting 0.6 to 2.1% of “a global prevalence of canine cases” between 1995 and 2000 (69). A recent study on CanL serology reported a seroprevalence of 34.4% in the same area (46). Although this municipality has favourable characteristics for harbouring FeL, a low adherence from the households to participate in the study was noticed. Only one pet-owner agreed to participate, and his cat was negative. This might be a result of the better knowledge of the control measures used for positive dogs. Since this is an intense VL/CanL transmission area, probably the households were aware of Brazilian health authorities' control measures. This would favour underreporting and hinder conclusions in a particular area.

There were positive cats in municipalities considered sporadic VL transmission including São Félix, Muritiba, and Camaçari (Table 1). In Muritiba, a recent study reported on a seroprevalence of 15.7% of dogs with CanL (44) whereas in Camaçari, where 11 VL cases were notified, CanL seroprevalence was 19.8% (70). In São Félix and Simões Filho, where <3 VL cases between 2014 and 2018 were notified, no available data on CanL was found. Despite gaps in notification by the health authorities, there is strong evidence of endemic zoonotic circulation of the parasite in these municipalities where local veterinarians have been increasingly reporting CanL. Consistent with these observations, the highest frequency of L. infantum infection found in the present study was in one of these localities (São Félix). Perhaps the lack of positivity in two of the municipalities investigated in this study was because of the low number of animals and households evaluated.

PCR was chosen as a confirmatory method of infection due to its high sensitivity and ability to confirm parasite species (71, 72). Among the 90 examined cats from a highly endemic area, the frequency found of L. infantum positivity was 7.9%. Moreover, only one out of seven infected cats developed FeL, suggesting that they may not be as susceptible to infection as dogs. In this context, other studies suggested that felines appear to be resistant to infection by L. infantum (73, 74). A lower tissue parasitism in cats compared to dogs might hinder cytological examination less accurate during diagnosis. It has been demonstrated in previous studies on dogs that the spleen is the best choice for a more sensitive parasitological or molecular detection of L. infantum (14, 71, 75). Moreover, studies have showed that a higher parasitic load in the canine host is associated with more intense disease, and that these more susceptible dogs are more infectious to the phlebotomine vectors (33, 76). Factors such as the parasite's tissue tropism and intrinsic pathophysiological mechanisms make organs such as the spleen more intensely parasitised than other ones in infected dogs (77–79). Thus, in dogs, the use of samples from such organs is likely to improve diagnostic sensitivity of methods based on the detection of parasites and/or its DNA (71, 75, 80). All these aspects have crucial importance under the view of One Health and remain to be more investigated in cats in the diverse endemic regions for L. infantum infection not only in Brazil, but also in other countries. Here, we advocate that feline bone marrow could be one of the target organs for FeL since we detected a higher parasite load in infected macrophages (Figure 3C). Bone marrow is commonly a chosen target for sampling tissue for PCR diagnosis because it is known to be consistently parasitised in dogs (71) and humans (81) with Leishmania active infection. Bone marrow has been shown to be a suitable sample for the diagnosis of L. infantum infection in cats by PCR (73, 82). Bone marrow collection under proper sedation also causes a minor impact on the welfare of cats (49). Consistent with those observations, the present study reports the identification of L. infantum in different cats by using several PCR techniques and sequencing.

Five cats tested positive in TR DPP®, but no parasite DNA was detected in the bone marrow in four of them. These data may suggest a natural lesser susceptibility in cats whose residual antibody response might only indicate a transient parasite exposure. Besides, cross-reactions with other pathogens should be considered, as already reported for dogs (83, 84). Susceptible canine hosts develop high antibody levels as the infection progresses, together with the development of clinical alterations and a higher chance to infect vectors (66, 79, 85, 86). Based on our data, this does not seem to happen in cats. Furthermore, the TR-DPP® used in the present study might not have detected feline immunoglobulins against L. infantum which would certainly exist as a consequence of the constant exposure to infected vectors (87). Indeed, among the seven cats that were positive for L. infantum DNA detection in the PCR test, only one was seropositive in the TR DPP®. Moreover, the only cat with active infection described herein had a negative result in the TR DPP® assay at the very moment when it was exhibiting overt signs of severe FeL. We emphasise that we have evaluated cats that were highly exposed to infection because of their close coexistence with infected dogs in areas of high transmission. Comparatively, a study on L. infantum naturally infected cats reported a considerable 75%-seropositivity for anti-Leishmania immunoglobulins as detected by an in-house indirect ELISA test (22). Thus, the lateral flow assay used in this study does not seem to be accurate for detecting feline antibodies against Leishmania antigens. This reinforces the need of more studies on immune responses during FeL and the development of more specific tests for L. infantum detection in cats.

Typically, because of the PCR's high sensitivity to detect L. infantum DNA, a positive result can be obtained with minimum amounts of parasites from a host sample (88). In this sense, a positive PCR result may indicate an initial infection or even an infection under the host immune system's resolution process. To characterize an active infection that may progress to the clinical disease, cytological examinations were performed on different tissue samples from the seven PCR positive cats, and of the remaining four cats that were positive only in the lateral flow assay. The molecular approaches confirmed the isolated strain to be L. infantum. This reinforces that several diagnostic methods should be employed for a correct diagnosis not only in dogs but especially in cats when a specific method warrants further development.

Three L. infantum-infected (PCR-positive) cats showed clinical signs ranging from cutaneous to systemic alterations suggestive of FeL as previously described (18). However, there was no laboratory evidence of active infection as established in our diagnosis criteria. One of these cats exhibited gum lesions indicative of a gingivitis-stomatitis complex. This cat lived in a shelter in close contact with other cats and dogs, some with diverse diseases. Living in an overpopulated environment is a well-described stress factor for cats. This favours immunosuppression and susceptibility to pathogens and other non-infectious morbidities (89). Likewise, clinical signs observed in the other two infected cats might have been associated with other pathogens since they had no Leishmania in their cytological samples and their condition improved upon further examinations and symptomatic treatment.

Among the substantial majority (92.2%) of PCR negative cats for L. infantum infection, a high proportion (73.5%) exhibited clinical signs of disease that could be attributed to FeL in a presumptive approach (18). We emphasise that the region is endemic for many other infectious diseases, and that these cats are subjected to the same conditions as their owners, namely undernourishment, high demographic densities, and precarious sanitation infrastructure. Therefore, the clinical presentation of unhealthy cats living in endemic areas for L. infantum does not necessarily reflect an active infection by this parasite. Accordingly, a careful assessment linking PCR positivity to clinical signs must be pursued during the clinical evaluation of a presumptive feline case. Indeed, in a different context of coinfections, a recent study performed in an endemic region for CanL in Spain, on a population of clinically healthy stray cats, still found a 5.6% positivity in blood samples using qPCR (90).

The clinical course of the cat with an active infection in this study suggests a natural pathophysiological evolution in a susceptible mammal host (91, 92). In the first examination, the cat had no clinical alteration, but its bone marrow was already heavily infected. No proper clinical follow-up and treatment could be performed because of the owner's non-compliance. Therefore, after the cat develop FeL, the owner looked for assistance again. Then, we could describe the natural infection progression. Some studies relate FeL onset to immunosuppressive factors such as retroviral infections, such as FeLV or FIV (18, 40). There was no evidence of a pre-existing infectious immunosuppressive condition in the cat with FeL. In the fourth month after the first PCR diagnosis, clinical signs consistent with FeL were also apparent in different bodily systems, confirming the disease's systemic character. Lesions in tissues such as ocular conjunctiva, oral mucosa, in addition to enlargement of lymph nodes, as observed in the present cat, have also been previously described by other authors (18, 93, 94). Hyperproteinaemia associated with hyperglobulinemia was the most evident clinical pathological finding when the cat exhibited overt clinical FeL. This finding agrees with other authors that considered hyperglobulinemia as the most relevant alteration found in FeL (95). Importantly, the proportions of gamma globulins in cats with FeL have been described as variable and often not detectable in serological tests, probably due to a predominantly cellular immune response (18, 96, 97).

The cat identified with active FeL herein lived in close contact with a dog previously diagnosed with CanL, and this dog could have been a parasite source for vectors. Indeed, a recent study on sheltered dogs and cats showed that dogs could be more infectious to sandflies due to higher parasite loads than cats with the same PCR-positivity rates. Moreover, the study argued that cats were more frequently seropositive than dogs (22). Despite xenodiagnoses techniques already proving the transmission of the parasite in sand fly feeding on infected cats, no consistent data has demonstrated that L. infantum-infected cats can maintain the parasite circulation in the environment (48, 98, 99). Besides, there is no robust data estimating the transmission risk for humans who cohabit with infected felines (38).

Brazilian municipalities with the highest number of VL reports low human development indexes (7, 100). In this context, as expected, cats of low-income individuals do not receive proper health care. They may be more susceptible to developing active infections with high parasite loads and therefore possibly contributing to the parasite transmission. As we mentioned before, the current study has some limitations especially due to underreporting in some areas. The studied feline population sample that could be assessed did not allow a precise prevalence estimation of the parasite infection. In the locality with the highest notification numbers of VL, a low number of cats could be evaluated, and these animals resided in the same household, so there is a possibility that the negative results found might not reflect reality. Yet, we can assume that TR-DPP® was not a reliable method for assessing humoral responses L. infantum-infected cats. Nevertheless, despite all limitations, the present study confirmed that L. infantum is circulating among cats using parasitological and molecular methods.

Since the evolution of the infection in cats is often subclinical and with a low parasitic burden, more detailed studies with representative samples are needed to ascertain L. infantum prevalence in endemic areas (73, 94, 101). Different from dogs, cats exhibit a more nocturnal behaviour and may walk longer distances and environments. This may increase the chances to encounter a sand fly in nature and acquire and/or transmit the parasite. More importantly, co-infections with other feline endemic pathogens including viruses (FIV, FeLV) (102), bacteria (Mycoplasma haemofelis, Ehrlichia spp) (103, 104) other protozoans (Cytauxzoon spp.) (105), and, helminths (enteric or lung worms) (106) may lead to immunosuppression of the cats. They would be more susceptible to L. infantum infection thus favouring higher parasite burdens. Indeed, in a recent study in Italy and Greece, where L. infantum may also occurs in cats and dogs, Aelurostrongylus abstrusus and Troglostrongylus brevior were found in cats and in a fatal case in a dog (107–109). Those worms have been increasingly reported as important in feline medicine and co-infections would impact animal management. Thus, further investigations should attempt to characterize hosts and co-infections in the epidemiological cycle of zoonotic leishmaniasis. This would promote better control and prevention measures to achieve the principles of One Health.

Molecular identification and sequencing of L. infantum from naturally infected cats indicates that the parasite is circulating in areas where CanL and VL are endemic. These data demonstrate the feline's exposure to the parasite and reinforce the need for further epidemiological investigations to understand how cats participate in the cycle of this zoonotic leishmaniasis. Additionally, the clinical evolution of a subclinical infection in a cat toward the development of patent feline leishmaniasis (FeL) illustrates a picture of active infection. A specific immunodiagnostic assay to assess feline humoral response to L. infantum is needed.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

The animal study was reviewed and approved by Committee for Ethics on the Use of Animals of the School of Veterinary Medicine and Zootechny of the Federal University of Bahia (Permission number CEUA-13/2017). Written informed consent was obtained from the owners for the participation of their animals in this study.

SB-M and FP conceived and designed the study. NS, BV, TN, JR, and FR performed the investigation and drafted the manuscript. NH, NA, and RP provided laboratory material resources and validation of experiments. NS, FP, NH, NA, RP, MS, SB-M, and RS analyzed the data. FP and MS conducted statistical analyses, data curation, and visualization. SB-M, FP, and RS provided funding acquisition, project management, experiments supervision, and manuscript editing. All authors contributed to the writing of the manuscript and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.