Introduction
Leishmaniases are zoonotic and anthroponotic diseases caused by several protozoan species in the genus
Leishmania that are transmitted by the bites of phlebotomine sand flies. They represent a major public health problem in endemic countries, with regular increases reported in the past decade [
1‐
3]. Different species present with diverse clinical symptoms, with a certain degree of specificity in the clinical presentation depending on the species [
4]. Leishmaniases are thus classified into cutaneous (CL), mucocutaneous (MC), and visceral leishmaniases (VL) [
5,
6].
Whereas VL in non-endemic areas is usually seen in immunocompromised adult patients living or having lived in endemic areas, imported CLs are easily linked to international travels. In recent years, these imported CLs have been on the rise due to international tourism, military operations, and the influx of immigrants from endemic countries [
7‐
9]. The French National Reference Centre reported a stable number of cases until 2012 [
10], but now reports an annual increase in CL declarations, from 130 in 2013–2017 to 214 in 2018, mostly (90%) from North Africa (
https://cnr-leish.edu.umontpellier.fr/files/2019/05/Rapport_CNRLeishmanioses_Act-2018.pdf). We were therefore interested in analyzing the data collected in our hospital located in northeast Paris.
We were also interested in validating our two-step diagnostic strategy, which includes a real-time quantitative PCR (qPCR) assay for the positive diagnosis targeting the consensus sequence of the highly repeated kinetoplast DNA, followed by the amplification and sequencing of a cytochrome
b (
cytb) gene fragment for species identification [
11]. Indeed, the species identification step is necessary for making the best therapeutic decisions according to species [
5,
6]. If VL is mainly due to
Leishmania infantum, different species can be responsible for CL in both Latin America and the Mediterranean basin, and they have different progressions and treatment options [
3,
12].
Discussion
We report on a comprehensive collection of 89 leishmaniasis cases seen over more than 7 consecutive years in a university hospital located in Paris, France. These cases consisted of 77 Old World and 12 New World Leishmania cases. The cohort primarily included CLs from North Africa (39 cases) or West Sub-Saharan Africa (21 cases).
Diagnostic qPCR targeting the 18S rRNA gene [
11] was more sensitive than microscopy for routine diagnosis, as has already been shown using other PCR gene targets and methods for CL [
14] and New World
Leishmania spp. [
15]. Besides the closed tube format to avoid contamination with amplicons and thus false positive results, the main interest in qPCR is to exclude false negative results by accurately quantifying the parasitic load and determining if PCR inhibitors are present [
16]. We confirmed that a CL diagnosis could be corrected when microscopy was negative, even when the lesions appeared old and were healing either spontaneously or due to previous treatment. For VL diagnoses, a more sensitive tool than microscopy is crucial for proper diagnosis and follow up.
A different target was chosen for the identification step than for the diagnostic step, namely the
cytb gene, as previously proposed for
Leishmania sp. identification [
17]. This precaution was designed to limit false positives caused by routine laboratory contamination. Indeed, opening amplicon-containing tubes for secondary analyses such as sequencing leads to a risk of contamination from the laboratory environment. Moreover, in contrast with non-sequence-based methods such as ITS-RFLP (internal transcribed spacer region-restriction fragment length polymorphisms), sequencing identifies SNPs that can be useful for comparison with databases [
18]. This target also allows for the correct identification of
L. killicki within the heterogeneous
L. tropica complex [
19]. However, the number of copies of the
cytb gene (~ 50 [
17]) is less than that of the 18S rRNA gene (~ 50–200 [
20]), explaining why species identification failed for five samples with low parasite loads.
For VL occurring in non-endemic areas, the issue is to increase the suspicion index and confirm the diagnosis using qPCR [
21]. Indeed, the clinical presentation is not specific and can be seen in different situations of immunosuppression such as HIV infection [
22], anti-TNF treatment [
23,
24], after solid organ transplantation [
25], or after hematopoietic stem cell transplantation [
26]. Excepted for HIV co-infection in endemic areas [
22], the prevalence of VL is often very low [
23,
24]. The accepted physiopathology is the reactivation of a persistent parasite from a previous primo-infection subsequent to an acquired immunodeficiency [
27]. Thus, the disappearance of circulating DNA after treatment does not mean the parasite has been cleared from the organism, hence the possible reoccurrence of the disease [
22]. Conversely, detection of leishmania DNA is not synonymous of an ongoing clinical active disease [
28]. Visceral leishmaniasis is also a differential diagnosis of leukemia-like syndromes in infants living or traveling to endemic countries, as observed here for a pediatric case [
29,
30].
For CL, the travel history easily differentiates patients returning from Latin America from those returning from Africa and the Middle East. In the former, the clinical lesions are often exudative, large, and prone to secondary bacterial infections. The main goal then is species identification, and we indeed identified different species, in accordance with the wealth of
Leishmania spp. in this region [
31,
32]. Importantly, none of the cases presented with mucosal lesions, confirming that patients seek medical advice soon enough after the appearance of such lesions. The treatments were diverse, according to patient and species (Table
1), but effective for all patients.
The largest contingent of patients with CL in our series was composed of migrants or French citizens born in France visiting relatives in North Africa or West Sub-Saharan countries. The delay between the appearance of lesions and the microbiological diagnosis was more than 3 months, which suggests these patients only consult when no spontaneous healing occurs. Consequently, the actual burden of CL is probably grossly underestimated. Three species responsible for the Old World CLs were identified:
L. infantum,
L. tropica/killicki, and
L. major. Our clinical observations support previous studies reporting that lesions caused by
L. major are more often multiple and located on limbs, whereas lesions due to
L. tropica are usually single and face-localized [
33]. These differences can be explained by the different vector behaviors [
33]. Clinically, the
L. major lesions were more exudative and wet than the
L. tropica/killicki lesions, while the
L. infantum lesions were more nodular infiltrative. Despite these clinical differences, however, there were numerous overlaps, and it is not possible to exclude a given species based on its clinical aspect. Moreover, the distribution areas of the species also frequently overlap [
3]. This highlights the importance of molecular identification for both epidemiological purposes and to avoid misdiagnosing
L. infantum CL, which can lead to secondary VL [
34], or alternatively, to reinsure patients in cases of
L. major lesions, which do not disseminate in the case of HIV infection [
11].
The age distribution of the Old World CL cases showed a double peak, corresponding to children and adults, including the elderly. While the occurrence of CL is not surprising in naïve children born in a non-endemic area, the occurrence of CL in older people suggests either they were naïve because (i) the parasite was not contracted during childhood, (ii) the patients lost their immune status after long periods spent in France, or (iii) that previous exposure to the parasite is not in any way protective. Mandall et al. suggest that despite the induction of a protective immune response, secondary
L. major infections can effectively establish themselves in a previously infected host, supporting the non-protective hypothesis [
35]. Moreover, Bousslimi et al
. have reported cases of CL not just in Tunisian children, but also in Tunisian adults (57.1%), which also supports the non-protective scenario [
33]. In this series, one can underline the high number of CL cases after travel to south-eastern Tunisia (governorate of Tataouine), where both
L. major and
L. tropica/killicki are endemic [
33]. This continual transmission of CL relates to environmental changes, which impact mammal reservoirs and sand fly populations, as well as demographic and human behavioral factors [
36], factors specifically present in Tataouine [
3].
Data on the molecular identification of
Leishmania species from West Sub-Saharan Africa are scanty although numerous outbreaks have been reported, mainly based on serology surveys or clinical diagnoses [
37]. In a recent study, 8 cases of
L. major infections were reported from Mali using an end-point PCR assay [
38]. Here, we added 20 confirmed
L. major infections from West Sub-Saharan Africa. In accordance with the absence of
L. tropica and
L. infantum CL in this region, only
L. major was reported [
37]. Interestingly, we observed two synonymous mutations in the
cytb gene (A108G and C624T) that distinguished the
L. major cases from West Sub-Saharan Africa from those from North Africa and the Middle East. Using an isoenzyme analysis, a particular zymodeme (MON26) reported in Mauritania, Senegal, and Mali has previously been shown to differ from the common zymodeme (MON 25) from North Africa [
39]. If our results are confirmed, this observation could serve as an epidemiological marker that is easily available using PCR.
We readily acknowledge the limits of our observational study to draw any epidemiological conclusions based on the frequency and intensity of transmission in the visited countries. We have no data on healthy returning travelers or those with self-limited and/or spontaneously healing lesions who did not seek medical advice. In addition, we cannot exclude that word-of-mouth information led to patients coming to consult in our hospital specifically, leading to a false impression of an increase in cases from a specific region.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.