Prevalence of visceral leishmaniasis among people with HIV: a systematic review and meta-analysis

Leishmaniasis is a parasitic infection expressing different clinical phenotypes. Visceral leishmaniasis (VL) is considered an opportunistic infection among people with human immunodeficiency virus (HIV). The objective of this review was to identify published data on the prevalence of Leishmania spp. infection among PWH and to define particular determinants that affect critically the epidemiological characteristics of VL-HIV coinfection and, potentially, its burden on public health. Two independent reviewers conducted a systematic literature search until June 30, 2022. Meta-analyses were conducted using random-effects models to calculate the summary prevalence and respective 95% confidence intervals (CI) of leishmaniasis among PWH. Meta-regression analysis was performed to investigate the impact of putative effect modifiers, such as the mean CD4 cell count, on the major findings. Thirty-four studies were eligible, yielding a summary prevalence of 6% (95%CI, 4–11%) for leishmaniasis (n = 1583) among PWH (n = 85,076). Higher prevalence rates were noted in Asia (17%, 95%CI, 9–30%) and America (9%, 95%CI, 5–17%) than in Europe (4%, 95%CI, 2–8%). Prevalence rates were significantly mediated by the age, sex, and CD4 cell count of participants. Heterogeneity remained significant in all meta-analyses (p < 0.0001). In the majority of included studies, people were coinfected with HIV and Leishmania species associated with VL, as opposed to those associated with cutaneous leishmaniasis. No sign of publication bias was shown (p = 0.06). Our summary of published studies on leishmaniasis among PWH is important to provide prevalence estimates and define potential underlying factors that could guide researchers to generate and further explore specific etiologic hypotheses. Supplementary Information The online version contains supplementary material available at 10.1007/s10096-022-04530-4.


Introduction
Leishmaniasis is a vector-borne parasitic infection with three primary clinical forms: visceral leishmaniasis (VL), cutaneous leishmaniasis, and mucocutaneous leishmaniasis [1]. More than 20 distinct species of the genus Leishmania are responsible for the clinical manifestations of the disease. The global incidence of leishmaniasis is estimated at 700,000 to 1 million new cases per year [2,3]. Though the clinical manifestations of Leishmania infection are traditionally classified into three syndromes, immunocompromised people are more likely to have disseminated disease [1]. Leishmania species that typically cause only cutaneous disease could cause disseminated disease in people with compromised cellular immunity such as people with human immunodeficiency virus (HIV (PWH)).
Over the last decade, advances in healthcare delivery have led to substantial decreasing trends in VL although it remains a fatal disease in endemic areas, such as in some African regions [2,4]. VL is considered an opportunistic infection for PWH in countries/regions, where leishmaniasis is endemic because coinfection rates are much higher in Leishmania-endemic countries/regions [5]. As of 2021, Leishmania-HIV coinfections have been reported from 45 countries [6,7]. High Leishmania-HIV coinfection rates have been noted in Brazil, Ethiopia, and the state of Bihar in India [8][9][10][11]. Leishmania-HIV coinfected people are more likely to develop the disseminated clinical disease, as well as high relapse and mortality rates [6]. Due to the high prevalence of Leishmania-HIV coinfection in these countries and the lack of a single reliable serologic test for leishmaniasis diagnosis, screening for Leishmania coinfection with a combination of laboratory tests would be useful for PWH living in or originating from endemic regions [12]. From a therapeutic perspective, the critical role of highly active antiretroviral therapy (HAART) in the outcome of VL-HIV coinfection has been clearly reported [13]. In particular, once Leishmania-HIV coinfection is diagnosed, early initiation of HAART is recommended as a protective modality against VL relapses. Antiretroviral treatment reduces the development of the disease, delays relapses, and increases the survival of the coinfected patients [14]. Several reviews have synthesized the existing evidence around the prevalence of Leishmania-HIV coinfection, as well as the demographic and diagnostic characteristics of the disease [15][16][17]. For instance, Lindoso et al. reported that VL occurs in 12 countries of Latin America, with 96% of cases reported in Brazil [15]. Moreover, Boelaert et al. showed that specific immunochromatographic test, i.e., rK39 ICT, shows high sensitivity and specificity for the diagnosis of visceral leishmaniasis in patients with febrile splenomegaly and no previous history of the disease, but the sensitivity is notably lower in east Africa than in the Indian subcontinent [16].
Overall, the results of the present study is expected to be useful to public health authorities when implementing better control strategies and guiding future developments in the field. We reviewed published evidence on the prevalence rates of Leishmania spp. among PWH and to quantitatively synthesize the results of published studies.

Search strategy and study selection
A literature search of Medline database was conducted from inception up to June 30th, 2022, following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Supplementary Table 1) [18]. Two reviewers performed the literature search independently using the algorithm (leishmania OR leishmaniasis) AND (immunocompromised OR immunodeficiency OR HIV OR AIDS OR (immune AND deficiency)) AND (incidence OR prevalence OR rate). Both reviewers evaluated the reference lists of all identified eligible studies using the "snowball" applied. Case reports and experimental or animal studies were excluded. Following the literature search, duplicate citations were removed, and the remaining articles were independently screened by two investigators to identify studies that met the pre-determined inclusion criteria. The study selection was conducted in two stages. First, we reviewed article titles and abstracts and removed those that did not meet our inclusion/exclusion criteria. Second, we retrieved and reviewed the full text of the remaining articles. For the remaining studies, the full papers were retrieved for further screening. In case of disagreement in the selection of studies or snowball procedure, the final decision was reached by team consensus. In articles with overlapping populations, the most recent or most complete publication was considered eligible.

Data extraction
For each eligible publication, the following study variables were extracted: publication year, country/region where the study was performed, study design, study period, follow-up period, sample size, age at diagnosis, mean patient's CD4 count, and the proportion of males. In addition, information about the number of patients with Leishmania-HIV coinfection, the type of Leishmania infection, type of Leishmania diagnostic method, the treatment administered, and the outcome of such infection were also extracted. Two reviewers performed the data extraction, and any disagreements were resolved by consensus.

Statistical analysis
A descriptive presentation of the eligible studies was initially performed (Table 1). Thereafter, the prevalence of Leishmania infection among PWH and the respective 95% confidence intervals (CI) were extracted or calculated from the available data using the Wilson's method [19]. Metaanalyses were undertaken using random-effects models [20] to estimate the prevalence of Leishmania-HIV coinfection overall and by study region (Europe, America, and Asia). We performed sub-analyses of prevalence by region where numbers allowed. Between-study heterogeneity was assessed using the Cochran Q and I 2 statistics 20 . The Z-test was applied for the overall effect, and statistical significance was set at p < 0.10. In addition, we performed sensitivity metaanalyses excluding a study per time to identify whether heterogeneity was affected by the inclusion of a specific study. The impact of potential effect modifiers, such as publication year, mean age, percent of males, and mean CD4 cell count on the main results, was explored in meta-regression analysis. Publication bias was calculated using the Egger's test (significance level was set at p < 0.05) [21]. Analyses were performed using the Stata software.
procedure, which entails searching the reference lists of each manuscript for additional eligible articles potentially missed from the initial literature search. We included articles that examined specifically the prevalence rates of leishmaniasis among PWH. No language or other restrictions were  Figure 1 shows the results of the literature search and selection process. Following the exclusion of publications with overlapping populations, a total of 34 eligible studies were finally included in this analysis [12,.

Characteristics of the studies
The descriptive characteristics of the included studies are presented in Table 1. Nineteen of the studies were of cohort design, and the remaining fifteen were found to be of cross-sectional design. The mean follow-up period ranged between 0.5 and 3.8 years. Most studies were conducted in Europe (France, Spain, Italy; n = 18), followed by America (Brazil, USA; n = 9), Asia (Iran, Thailand; n = 6), and Africa (Morocco; n = 1).

Prevalence of leishmaniasis among PWH
Thirty-four eligible studies yielded a total sample of 85,076 PWH, of which leishmaniasis was diagnosed in 1583 people. The average percent of males was 72.4%, while the mean age of participants ranged from 30 to 41 years (median 36.9 years). Diagnosis of leishmaniasis was established through bone marrow biopsy and/or blood tests in the majority of studies (Table 1). The prevalence of leishmaniasis was found to be particularly variable across regions globally, although a summary prevalence of leishmaniasis among PWH was estimated as high as 6% (95% CI, 4-11%; n = 34 studies) with significant between-study heterogeneity (I 2 , 99.3%, p < 0.0001) (Fig. 2). The prevalence was 9% among studies conducted in America (95% CI, 5-17%; n = 9 studies) (Fig. 3a) and even higher among those conducted in Asia (17%, 95% CI, 9-30%; n = 6 studies) (Fig. 3b) compared to the prevalence calculated for European studies (4%, 95% CI, 2-8%; n = 18 studies) (Fig. 3c). Heterogeneity remained significant in the subgroup meta-analyses by study origin.
In the majority of included studies, people were diagnosed with HIV and Leishmania species associated with VL, as opposed to Leishmania species associated with cutaneous leishmaniasis, with the exception of one study where 25 out Full-text articles assessed for eligibility n = 43 Studies included in the review and metaanalysis n = 34 Records excluded (n = 9) 1) No discrete assessment of Leishmania coinfection (combined results from broad category HIVcoinfections including leishmaniasis; n=5) 2) Studies with overlapping populations (n=4) of 27 Leishmania-HIV infected participants had VL, and the remaining two had cutaneous leishmaniasis [41].
In the meta-regression analyses, age (p = 0.05) and CD4 cell count (p = 0.02) were statistically significant positive modifiers of the observed associations, whereas sex (percent of males) had a significant inverse effect modification (p = 0.002). Publication year had a borderline mediating effect on the results of the main meta-analyses (p = 0.06) ( Table 2). Sensitivity meta-analyses excluding a study per time showed similar results with those of the main analyses, without reducing the between-study heterogeneity. Egger's test showed no sign of publication bias (p = 0.06) (Supplementary Fig. 1). Information regarding treatment of Leishmania-HIV individuals was reported in 11 studies only. Liposomal amphotericin B, conventional pentavalent antimonial, pentamidine, allopurinol, itraconazole, and aminosidine were reported as medications; amphotericin B was the most frequently used medication.

Discussion
The present systematic review and meta-analysis of 34 published studies including 1583 individuals with Leishmania-HIV coinfection provides evidence for 6% prevalence of leishmaniasis among PWH living in or from populations of Leishmania-endemic regions. It should not be overlooked, however, that the abovementioned prevalence practically reflects the overall frequency in VL-PWH in the specific countries identified and included in the present study. As it was expected, we observed higher prevalence in studies from Asia and the Americas compared to studies from Europe; data on leishmaniasis among PWH were unavailable for other regions where Leishmania is endemic including Africa and the Middle East. In addition, the present results are consistent with the already well-described caveat that the prevalence of species causing VL is higher in regions (i.e., the "Old World") where such species are endemic and regions with migration from areas where such species are endemic. Similarly, we found the prevalence of cutaneous and mucocutaneous leishmaniasis to be the highest in regions where the species causing these disease manifestations are endemic (e.g., the "New World"). Prevalence rates varied significantly by age, sex, and CD4 cell count of participants. Publication year also had a borderline significant mediating effect, which could be potentially explained by treatment advances, i.e., the increased availability of ART as time progressed. Overall, the results were based on highly heterogeneous studies mainly conducted in Europe. Indeed, the variation of findings by the study setting  Leishmaniasis is considered an opportunistic infection among PWH, especially among those living in VL-endemic sub-tropical and tropical regions around the world, including the Mediterranean [55]. Globally, the occurrence of Leishmania-HIV coinfection has been enhanced by the spread of HIV into rural areas and the concurrent spread of leishmaniasis to suburban/urban areas [56,57]. Some experts expect the number of cases of Leishmania-HIV coinfection will increase as the distributions of the two infections further overlap [58]. In several Southwestern European countries, such as France, Italy, Spain, and Portugal, increasing incidence rates of Leishmania-HIV coinfection have been noted, with most cases concerning VL [57].
Studies on the immunological background of Leishmania-HIV coinfection report that both microorganisms infect and multiply within myeloid or lymphoid cells, enhancing thus the reciprocal modulation of Leishmania and HIV pathogenesis [57]. Additionally, given that recovery from leishmaniasis is associated with long-term persistence of parasites at the primary sites of infection and their draining lymph nodes, it is likely that HIV-mediated immunosuppression due to CD4( +) T cell depletion could lead to reactivation leishmaniasis, particularly in immunocompromised patients [59][60][61]. Indeed, the present results suggest a potential prevalence effect modification due to the PWH population control of HIV infection, which CD4 count is a surrogate marker for. Of note, the mean CD4 count for nearly all cohorts was less than 200, a finding clearly carrying increased risk of leishmaniasis clinical manifestations.
The lethality of VL increases by 4.6% to 16.6% when associated with HIV infection [62]. The presence of HIV can lead to severe forms of VL, which are difficult to control and manage. Furthermore, PWH are more likely to have severe and/or prolonged VL with a more difficulty treatment course, which further may increase the HIV replication and the clinical evolution of AIDS [62]. Therefore, this coinfection warrants concern and must be recognized and treated in a timely manner. Recent reports from the World Health Organization recommend that patients with Leishmania-HIV coinfection should be treated with liposomal amphotericin B, whereas early administration of HAART acts as a protective modality against VL relapse [6,55].
The sound methodological approach has been a strength of the current study given that meta-analyses, sub-analyses by study origin, sensitivity meta-analyses, and meta-regression analyses were run. However, the present results should be cautiously interpreted in view of limitations inherent to the study design and data availability of eligible studies including variable criteria for selection of the comparison groups, statistical analysis methods, and variable follow-up periods.
The highly heterogeneous results across the included studies could either indicate that different factors may be implemented in the prevalence of Leishmania-HIV coinfection in each country or could point to the heterogeneity of the methodological approaches, e.g., in statistical analysis, implemented by each study. Moreover, information on other variables of interest, such as ART at time of leishmaniasis diagnosis, mortality of people with Leishmania-HIV coinfection, and use of secondary chemoprophylaxis post-leishmaniasis treatment, were available by only 1-2 studies, thus not allowing generalization of the results.
Although the articles included in the present study are biased towards specific regions (Europe and Brazil) and cases visceral leishmaniasis only, it becomes evident that the field of VL in HIV patients remains of particular scientific interest. Further studies, both retrospective and prospective, should be specially designed and conducted in order the role of critical individual determinants of the disease to be defined and the between-country comparisons to be feasible.

Conclusions
In view of the still rising rates of the HIV/AIDS pandemic in some parts of the world, we noted a high prevalence of Leishmania infection among PWH. Heterogeneity issues and diverse study settings should be taken into account when interpreting the results of the present study. However, and beyond the above factors, this attempt to summarize current findings is of significant importance in providing effect estimates, as well as to point towards potential underlying factors. Further research based on well-designed studies is needed to generate and explore specific etiological hypotheses, as well as to compare prevalence and determinants between countries reliably and effectively.

3
Data availability The datasets generated during and/or analyzed during the current study are not publicly available due to potential compromise of individual privacy but are available from the corresponding author on reasonable request.
Code availability Not applicable for this submission.

Declarations
Ethics approval Not applicable for this submission.

Competing interests
The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.