In vitro drug susceptibility of Leishmania donovani from visceral leishmaniasis patients with and without HIV coinfection in Northwestern Ethiopia

Abstract

Leishmania and HIV coinfection is a major public health problem in more than 35 countries worldwide. Leishmania infection is frequently reactivated in VL and HIV coinfected patients. One of the major challenges in the management of VL/HIV coinfection is the lack of an effective drug therapy that not only resolves the first episode of VL but also prevents relapse. This study aimed to assess the in vitro drug susceptibility profile of clinical isolates of Leishmania in HIV positive and negative patients obtained during the first VL episode before treatment and during VL relapses on patients who attended LRTC of University of Gondar, Gondar, northwest Ethiopia, between January 1, 2020, and December 30, 2020. We evaluated the in vitro susceptibility to amphotericin B (AmB), paromomycin (PMM), and miltefosine (MLT) of 30 L. donovani isolates obtained from primary and relapse VL patients with and without HIV co-infection. All 30 strains of L. donovani were tested at the promastigote stage whereas, 10 randomly selected strains of L. donovani were re-tested at the amastigote stage. The mean (± SD) IC₅₀ of AmB, PMM, and MLT tested against promastigotes was 0.221 ± 0.108 µM, 13.49 ± 6.92 µM, 3.77 ± 1.77 µM respectively. Similarly, the mean (± SD) IC₅₀ of AmB, PMM, and MLT against amastigotes was 0.171 + 0.027µM, 10.80 + 4.12 µM, and 3.63 + 1.72 µM, respectively. When the data was disaggregated between primary and relapse cases of VL, the IC₅₀ observed against strains from relapse VL was higher than the IC₅₀ observed against strains from primary VL. The difference in the IC₅₀ against strains from relapse VL with primary VL, was statistically significant (p = 0.03) for AmB against amastigote stages, but not promastigote stage, whereas the difference in IC₅₀ obtained for PMM and MLT against both stages of the parasites was not statistically significant. Likewise, strains from HIV coinfected VL patients with multiple relapses showed an increase in IC₅₀ value for AmB and MLT in both parasite stages. The current study showed that in vitro susceptibility of strains decreased progressively in relapsing patients compared with primary VL patients. Our findings suggest that AmB and MLT resistant parasites may indeed emerge in repeatedly treated relapse VL cases in HIV coinfected patients, warranting the need for continuous monitoring of L. donovani susceptibility to current treatments; and also, the results of such studies reveal the need to re-examine the current therapy policies in HIV coinfected VL patients.

Introduction

Visceral leishmaniasis, also called kala-azar, is caused by species of the Leishmania donovani complex, which mainly targets tissue macrophages of systemic organs, such as the spleen, liver, and bone marrow¹. Reactivation of the infection is possible when T-cell immune responses are compromised, for example, due to post-transplant immunosuppressive therapy, use of immunomodulators, advanced age, or in HIV-infected patients². Rendering to the recent World Health Organization (WHO) report, VL is endemic in 75 countries with an estimated 50,000–90,000 new cases occurring each year. 90% of the global disease burden occurs in just six countries: India, Bangladesh, Sudan, South Sudan, Brazil, and Ethiopia¹. Ethiopia is among the top six high burden countries, with approximately 3.2 million people at risk and 3400–5000 VL cases occurring annually³.

In the past three decades, VL has become one of the most important opportunistic infections in HIV-infected patients. Coinfection with HIV and Leishmania has been reported in more than 35 countries. Initially, most of the cases were from southwestern Europe, but the number of cases is increasing in Sub-Saharan Africa (particularly in Ethiopia), Brazil, and South Asia⁴. Comorbidities are an added risk factor and VL is a common complication for HIV-infected individuals. Coinfection increases susceptibility to VL by 23.2%, whilst VL elevates the progression of HIV to AIDs by 100–2320x⁵, as immune mechanisms required to control either disease are impaired. Therefore, VL may pose a greater health risk in Leishmania-endemic countries where HIV-infected populations are present, although the total burden of this co-infection is underreported, partly due to the remoteness of affected areas⁵.

In areas endemic for VL, many immunocompetent hosts have asymptomatic infection. After primary infection, parasites remain viable in healthy individuals for long periods, creating a population at risk of reactivation if immunosuppression occurs. However, HIV coinfection influences the clinical course of the disease and reactivates latent Leishmania infection, increases severity, and negatively affects treatment outcomes. It substantially increases the risk of progression from asymptomatic Leishmania infection to active disease. In non-endemic areas of the world, VL is mostly an opportunistic infection with up to 70% of adult leishmaniasis cases being related to human immunodeficiency virus (HIV) infection⁶.

In the present situation, chemotherapy is the key strategy for VL control due to the absence of vaccines and the limited impact of vector control, which is based on a very short list of chemotherapeutic agents. Currently available treatments for VL are limited to pentavalent antimonials, amphotericin B, paromomycin, and miltefosine - the only oral treatment available. The therapeutic options for VL depend on different factors, such as the geographical area of the infection, development of resistance to habitual treatments, HIV co-infection, malnutrition, genetic and anthropometric characteristics, social and epidemiological context, and other concomitant infections^4,7. To date, pentavalent antimonials, AmB, PMM, and MLT have demonstrated efficacy against VL in immunocompetent patients. However, there are a few data to guide recommendations for optimal therapy of VL in HIV-infected patients⁸.

Patients with VL/HIV coinfection experience increased drug toxicity and treatment failure rates compared to VL patients, with more frequent VL relapse and death. In the era of VL elimination strategies, HIV coinfection is progressively becoming a key challenge, because HIV-coinfected patients respond poorly to conventional VL treatment and play an important role in parasite transmission¹. The poor therapeutic outcome, the higher rate of relapses, and the polyparasitic nature of VL in HIV-infected persons, as well as the atypical manifestations of the disease that make diagnosis difficult and the impaired access to healthcare resources of co-infected patients, make HIV-infected individuals prone to enlarge the number of human reservoirs in areas where transmission of leishmaniasis is already anthroponotic⁹.

Consequently, incomplete or irregular treatment leads to drug pressure and the rapid development of drug-resistant parasites. Despite increasing reports of Leishmania drug resistance in endemic regions, data on Ethiopian isolates remain limited, particularily in HIV coinfected patients. Hence, it is important to continuously monitor the in vitro susceptibility of antileishmanial drugs to prevent the emergence of drug resistance, against the few drugs which are currently in use. Therefore, this study is useful to assess the antileishmanial drug sensitivity profile of Ethiopian L. donovani, i.e., to determine if strains from relapse VL with HIV coinfection have higher mean IC₅₀ values compared with strains from primary VL and also if strains from relapse VL with HIV coinfection has a shift in IC50 threshold. The study findings can be used as baseline data for further studies.

Methods and materials

Patients and parasites

All patients diagnosed with primary or relapse VL and confirmed by parasitological methods; who were age ≥ 18 years and have given informed consent. Patients who were on antileishmanial treatment for two weeks prior to recruitment, those aged below 18 years of age and those refused to participate were excluded from the study. Primary VL is a patient who is diagnosed with visceral Leishmaniasis for the first time. Relapse: If a person returns with clinical features and positive parasitology consistent with VL, after having been successfully treated for primary VL and discharged improved or with a negative test of cure (TOC), the patient is known as relapse VL case. A relapse episode is defined as a VL episode diagnosed ≥ 4 weeks after the VL cure. A VL episode occurring within 4 weeks after cure is considered treatment failure and treated accordingly.

Clinical isolates of L. donovani were prepared from splenic or bone marrow aspirates of primary or relapse VL patients with or without coinfection reporting to Leishmaniasis research and treatment Center (LRTC) of University of Gondar, Gondar, Ethiopia under the guidelines of the Ethical Committee of the respective Institute.

Parasites were initially inoculated into NNN agar slope with Locke’s antibiotic solution overlay and incubated at 26°c. After achieving an acceptable concentration of active promastigotes in the biphasic NNN blood agar slope medium, the promastigotes were transferred to the sterile RPMI-1640 culture medium, supplemented with 20% (v/v) heat-inactivated fetal bovine serum (FBS), 2 mM glutamine, 100 units/ml penicillin, and 100 mg/ml streptomycin and incubated at 26^oC. Parasites were sub-cultured once weekly and were tested for their in vitro sensitivity to antileishmanial drugs within 10 passages starting from primary isolation to avoid selection of resistant parasites and changes in parasite characteristics due to prolonged cultivation in vitro. Mid logarithmic growth phase promastigotes (from 3–4-day old culture) were used for promastigotes-based assay while late stationary promastigotes harvested on 5–6 days were used to infect macrophages for intracellular amastigotes assay.

Laboratory animals

Female BALB/c mice, weighing 20–25 g and of approximately the same age were used for the study. BALB/c mice were supplied by the animal breeding facility at the College of Health Sciences, Addis Ababa University. Mice were housed in polypropylene cages and fed with a standard diet and water ad libitum.

Preparation of drug stock solutions

Drug stock solutions were prepared as previously described⁹, with slight modification. To obtain a 20 mM stock solution of Amphotericin B, 18 mg Fungizone powder was dissolved in 1 ml 100% dimethyl sulfoxide (DMSO). Further dilutions were made in RPMI 1640 medium supplemented with 20% FBS and immediately used. A 20 mM MLT stock solution was prepared by dissolving 81.5 mg MLT in 10 ml water or PBS by stirring until a clear solution was obtained. The MLT stock solution was kept at 4˚C. Finally, a PMM stock solution (20 mM) was prepared by dissolving 123 mg PMM-sulfate in 10 ml water or PBS. The stock solution was kept at 4˚C till used.

Peritoneal macrophage isolation and Preparation

Stimulation and collection of primary mouse peritoneal macrophages was performed as previously described in with slight modification^10–12. Peritoneal macrophage cells (PMCs) were isolated from BALB/c mice upon peritoneal injection with 2% starch solution in sterile PBS. Macrophages were collected by abdominal lavage with ice-cold PBS supplemented with 3% HIFBS containing 1% penicillin and streptomycin. The collected cells were centrifuged for 10 min at 400 g and 4 °C, washed in RPMI-1640 medium and re-suspended in RPMI-1640 medium containing 20% HIFCS. Cells were counted using a hemacytometer, and the number of macrophages was adjusted to contain 1 × 10⁵cells/ml.

Drug susceptibility assay at promastigote stage

The promastigote assay was performed in 96-well plates in triplicates as previously described with slight modification¹¹. Test drugs were serially diluted 2-fold from highest concentrations of 1.600µM for AMB, 24µM for MLT, and 72 µM for PMM to a six-point dilution. Then 100 µL of 1 × 10⁶ /ml logarithmic stage culture was added to each well. Plates were incubated with the positive control (parasites without test drugs, 0% inhibition) and blank wells (media without parasite, 100% inhibition) at 26 °C for 72 h.

After the 72-hour incubation, 10 µl of resazurin stock was added to each well and the plates were further incubated overnight at 26 °C. The fluorescence intensity was estimated after 72 h in a multilabel microplate reader at 550 nm excitation-590 nm emission. Blank wells containing a complete RPMI 1640 medium were used to monitor the background fluorescence activity of resazurin, and the average value was subtracted from every well. Wells containing freely growing promastigotes in a complete medium without inhibitors were used as a positive control (0% inhibition). The assay was performed in triplicate and results were expressed as the mean percentage reduction of parasite numbers compared with untreated control wells.

Promastigote inhibition (%) was calculated for each concentration using the following formula:

Data were analyzed using non-linear regression analysis (GraphPad Prism version 8.4 software), and the 50% inhibitory concentration (IC₅₀) of drugs were determined.

Drug susceptibility assay at intracellular amastigote stage

In vitro amastigote-based assay was carried out as previously described¹⁰, with slight modification. The assays were set up in Lab-Tek^® (Sigma-Aldrich) 8-well chamber slide™ system. Primary mouse peritoneal macrophages isolated from BALB/c mice were used for this assay. Stationary phase promastigotes were counted and adjusted to 1 × 10⁶ parasite/ml and were prepared in complete RPMI 1640 medium (Sigma Aldrich) supplemented with 20% HIFCS.

Peritoneal macrophage cells were seeded onto sterile tissue culture chamber slides (Labtek Products, Sigma-Aldrich), in RPMI-1640 supplemented with 20% HIFCS at a density of 1 × 10⁵ per well and incubated at 37 °C and 5% CO₂. Then the adhered PMCs were infected with stationary phase promastigotes at a ratio of 10 promastigotes to 1 macrophage and maintained at 37 °C in a 5% CO₂ incubator. After 24 h, the cultures were washed to remove extracellular promastigotes and one slide was fixed with methanol and stained with Giemsa stain to determine the initial level of infection. If a sufficient level of infection was obtained after 24 h, drugs were added over a pre-determined range of concentrations.

Amphotericin B solutions at concentration of 1.6, 0.8, 0.4, 0.2, 0.1, and 0.05 µM, paromomycin at concentrations of 72, 36, 18, 9, 4.5, 2.25, and 1.125 µM and miltefosine solutions at concentrations of 24, 12, 6, 3, 1.5, and 0.75 µM were added in triplicate at each concentration, on the same plate. After 72 h incubation, all slides were methanol-fixed and Giemsa stained. Stained slides were examined under oil immersion (100x) for counting infected macrophages and amastigotes by searching through at least 50 macrophages per well. The data is expressed as the proportion of infected macrophages (IP) and the number of amastigotes per 50 PMCs. An experiment was performed in triplicates.

Determination of IC₅₀ values in amastigote-based assays

The drug activity was evaluated by determining the number of amastigotes per macrophage, as well as by the proportion of infected macrophages. The infection index for each well of the triplicates was determined by multiplying the proportion of infected macrophages (IR) with the average number of amastigotes in the infected macrophages using the formula:

The percentage of inhibition was calculated by comparing the parasite infection index of treated and untreated (control) wells. IC₅₀ values were determined from sigmoidal curves (on semi-log curves/plots) for each of the isolates tested against all 3 standard antileishmanial drugs, and results were expressed in micromolar (µM) concentrations. The IC₅₀ value was defined as the inhibitory concentration of test compounds that reduces amastigotes density (number of amastigotes per infected macrophages) by 50%.

Ethical consideration

Ethical clearance was obtained from the ethical committee of the Department of Microbiology, Immunology and Parasitology, College of Health Sciences, Addis Ababa University, and Institutional Research Ethics Board of the University of Gondar. All patients were informed using written documents and signed an informed consent statement.

The samples were anonymized to ensure that identification of the patient was not possible. The use of laboratory mice was carried out with local ethical approval obtained from the Institutional Research Ethics Board of the University of Gondar and in accordance with the Public Health Service Policy on Humane Care and Use of Laboratory Animals, 2015 of the U.S. Department of Health and Human Services.All study procedures were carried out in accordance with the Helsinki declaration and other local and international research ethics guidelines.

Statistical analysis

All statistical analyses were performed using Graphpad Prism version 8.4 software using the sigmoidal dose-response (variable slope) nonlinear curve fitting function. Both promastigote-based and amastigote based assays were carried in triplicate. IC₅₀ for antileishmanial drugs was expressed as mean values ± standard deviations. The student’s unpaired t-test was used to determine the statistical significance of the values obtained, for comparison of groups, and data values were expressed as the mean ± standard deviation (SD). Tests were considered statistically significant if p < 0.05.

Results

Demographics of study participants, clinic-epidemiological and sample information

A total of 30 L. donovani isolates, 17 isolates from HIV-infected VL patients, and 13 isolates from HIV-negative VL patients were obtained. Out of the 30 isolates, 15 were obtained during the first VL episode (among which two were obtained from HIV-positive patients), while the other 15 were obtained during VL relapses, all recovered from HIV-positive patients. The isolates were recovered from a splenic biopsy (24 cases) and bone marrow aspirates (6 cases). All patients were male, ages range from 19 to 59 years. Amphotericin B was the most frequently used followed by miltefosine, then sodium stibogluconate and pentamidine in the treatment prior to parasite isolation in relapse cases.

In vitro promastigote drug susceptibility

The mean inhibitory concentration (IC₅₀) values were determined from the calculated % inhibition values. The overall mean inhibitory concentration (IC₅₀) values of AmB, PMM and MLT against promastigote stages (n = 30) were 0.221 ± 0.108µM (95% CI 0.181–0.262), 13.49 ± 6.92µM (95% CI 10.90-16.07), and 3.77 ± 1.77µM (95% CI 3.11–4.40), respectively. The mean inhibitory concentration (IC₅₀) values of AmB, PMM and MLT against promastigote stages of strains isolated from primary VL (n = 15) were 0.185 ± 0.051µM (95% CI 0.157–0.213)), 11.32 ± 3.36µM (95% CI 9.46–13.18) and 3.15 ± 0.80µM (95% CI 2.71–3.60)), whereas mean inhibitory concentration (IC₅₀) against strains from relapsed VL (n = 15) group was 0.257 ± 0.136µM (95% CI 0.181–0.332µM)), 15.65 ± 8.82µM (95% CI 10.77–20.54)) and 4.40 ± 2.24 µM (95% CI 3.16–5.64)), respectively. However, the difference in the calculated mean inhibitory concentration (IC₅₀) for the three drugs between strains from primary VL group and strains from relapse VL group was statistically insignificant (Table 1).

Table 1.

Comparison of mean inhibitory concentration (IC₅₀) of AmB, PMM, and MLT against promastigotes isolated from primary VL and relapse VL patients with or without HIV coinfection.

Tested drugs	Category	Mean IC50 ± SD, µM	95% CI	P. value
AmB	Primary VL Isolates (n = 15)	0.185 ± 0.051	0.157–0.213	0.0678
	Relapse VL Isolates (n = 15)	0.257 ± 0.136	0.181–0.332
PMM	Primary VL Isolates (n = 15)	11.32 ± 3.36	9.46–13.18	0.0865
	Relapse VL Isolates (n = 15)	15.65 ± 8.82	10.77–20.54
MLT	Primary VL Isolates (n = 15)	3.15 ± 0.80	2.71–3.60	0.0518
	Relapse VL Isolates (n = 15)	4.40 ± 2.24	3.16–5.64

(−) = Negative, (+) = positive, µM = micromolar, IC₅₀ = mean inhibitory concentration, PVL = primary visceral leishmaniasis, RVL = relapse visceral leishmaniasis.

In vitro intracellular amastigote drug susceptibility

Following the primary antipromastigote assay, 10 isolates (5 strains from primary VL cases and 5 strains from relapses cases) were randomly selected for intracellular amastigote assay. The overall infection rate of PMCs by stationary phase promastigote was ranged from 64% − 82%, with an average parasite burden of 2.560–3.8/macrophage. The infection levels remained static or increased slightly in the untreated macrophages during the subsequent 3 days of incubation.

The mean inhibitory concentration (IC₅₀) value for AmB, PMM, and MLT against the amastigote stages (n = 10) were 0.171 ± 0.027µM (95% CI 0.151–0.190), 10.800 ± 4.120µM (95% CI 7.860– 13.750), 3.630 ± 1.720µM (95% CI 2.400–4.850); respectively.

The mean inhibitory concentration (IC₅₀) value for AmB, PMM, and MLT against amastigotes isolated from primary VL cases (n = 5) was 0.153 ± 0.016µM (95% CI 0.133–0.174), 8.85 ± 2.94µM (95% CI 5.20 -12.51), and 3.46 ± 1.65µM (95% CI 1.41–5.51), whereas in relapse VL cases (n = 5) it was 0.188 ± 0.026µM (95% CI 0.155–0.221), 12.75 ± 4.47µM (95% CI 7.20–18.30), and 3.80 ± 1.98µM (95% CI 1.34–6.25) respectively. A statistically significant variation was observed in the mean inhibitory concentration (IC₅₀) value of AmB obtained for isolates recovered from primary VL (0.153 ± 0.016µM) cases and relapsed VL (0.188 ± 0.026µM) cases (p = 0.03) (Table 2).

Table 2.

Comparison of mean IC₅₀ of AmB, PMM and MLT obtained against amastigote stage of strains isolated from primary VL with that obtained for amastigote stage of strains isolated from relapse VL cases coinfected with HIV between January 1, 2020, and December 30, 2020.

Tested drugs	Category	Mean IC50 ± SD (µM)	95% CI	P. value
AMB	Primary VL Isolates (n = 5)	0.153 ± 0.016	0.133–0.174	0.03
	Relapse VL Isolates (n = 5)	0.188 ± 0.026	0.155–0.221
PMM	Primary VL Isolates(n = 5)	8.85 ± 2.94	5.20-12.51	0.14
	Relapse VL Isolates (n = 5)	12.75 ± 4.47	7.20–18.30
MLT	Primary VL Isolates (n = 5)	3.46 ± 1.65	1.41–5.51	0.77
	Relapse VL Isolates (n = 5)	3.80 ± 1.98	1.34–6.25

Comparison of mean IC₅₀ AmB, PMM & MLT obtained against promastigote and amastigote stage between primary VL & relapse VL isolates with tentative IC50s reported breakpoint in literature

We compared the mean IC50 obtained in our study with previously reported as tentative “Breakpoint” estimates (Table 3) for categorizing drug-susceptibility and drug-resistance for all species of the L. donovani complex by Maes et al.⁹. which was based on results obtained with sensitive reference strains (L.donovani MHOM/ET/67/L82 and L.infantum MHOM/MA/67/ITMAP263).

Table 3.

Mean IC₅₀ of AmB, PMM and MLT obtained against both stages of some strains isolated from multiple relapse VL cases coinfected with HIV between January 1, 2020, and December 30, 2020.

Isolate Id	Episode	AmB(uM)		MLT(uM)		PMM(uM)
		Promastigote	Amastige	Promastigote	Amastigote	Promastigote	Amastigote
RVL-09(+)	3rd	0.452	0.182	6.78	2.36	32.17	10.68
RVL-07(+)	4th	0.385	0.169	5.41	4.03	23.03	12.9
RVL-23(+)	5th	0.521	0.234	9.01	6.71	31.23	19.04
RVL-21(+)	13th	0.468	0.181	8.10	4.19	23.96	13.57

In comparison with this tentative break points, the mean inhibitory concentration (IC₅₀) obtained in our study findings (Table 4) was slightly higher at amastigote stage but comparable at promastigote stage to AmB; lower against amastigote whereas comparable at promastigote stage to PMM and comparable at both promastigote stage and amastigote stage to MLT.

Table 4.

Mean IC₅₀ of AmB, PMM and MLT obtained against both stages of strains isolated from primary VL and relapse VL cases coinfected with HIV between January 1, 2020, and December 30, 2020.

Groups	AmB		PMM		MLT
	Promasigotes (n = 15)	Amastigotes (n = 5)	Promasigotes (n = 15)	Amastigotes (n = 5)	Promasigotes (n = 15)	Promasigotes (n = 5)
Primary VL	0.185 ± 0.051	0.153 ± 0.016	11.32 ± 3.36	8.85 ± 2.94	3.15 ± 0.80	3.46 ± 1.65
Relapse VL	0.257 ± 0.136	0.188 ± 0.026	15.65 ± 8.82	12.75 ± 4.47	4.40 ± 2.24	3.80 ± 1.98
P values	0.067	0.03	0.086	0.14	0.051	0.77

Besides, some strains obtained from multiple relapse VL patients with HIV coinfection (Table 3) ,showed a shift in mean inhibitory concentration, in comparison with the tentative breakpoints reported in the literature (Table 5), which needs further validation.

Table 5.

‘Breakpoint’ estimates^a for categorizing drug-susceptibility and drug resistance against miltefosine (MIL), paromomycin (PMM) and amphotericin B (AmB)⁹.

Drugs	Promasstigotes (Axenic)	Amastigotes (PMCs)	Susceptibility limits(IC50 estimates)
			Sensitive	Resistant
AmB	0.1–0.3	0.01–0.03	< 0.5	> 0.5
MLT	2–5	3–6	< 10	> 25
PMM	15–25	40–50	< 60	> 150

Another observation of our study was that strains isolated from a single VL patient with HIV coinfecton after second and third relapses respectively showed increased IC₅₀ values in the last relapse. A higher IC₅₀ value was found at the 3rd relapse compared to second relapse (Table 6).

Table 6.

Mean IC₅₀ of AmB, PMM and MLT obtained against both stages of strains isolated from consecutive relapse VL cases coinfected with HIV between January 1, 2020, and December 30, 2020.

Isolate ID.	Relapse episode	AmB (µM)		MLT (µM)		PMM(µM)
		Promastigote	Amastigote	Promastigote	Amastigote	Promastigote	Amastigote
RVL-04(+)	2nd	0.282	0.173	6.34	1.65	16.62	7.20
RVL-09(+)	3rd	0.452	0.182	6.78	2.36	32.17	10.68

Lastly, we observed one strain (5th relapse) obtained from VL patient coinfected with HIV having a history of treatment failure to a combination of L-AmB and MLT, exhibited higher mean IC₅₀ at both parasitic stages for AmB and MLT. This finding might suggest an increased tolerance of the parasite to AmB and MLT in response to repeated treatments.

Discussion

Treatment outcome in VL is a complex phenomenon that depends on the interaction between the drug, the parasite, and the host. The current study investigated the in vitro susceptibility profile of L. donovani strains to the commonly used anti-leishmanial drugs amphotericin B, paromomycin, and miltefosine using isolates obtained from primary VL and relapse VL patients, with or without HIV coinfection.

In our study finding we obtained the mean IC₅₀ at promastigotes stage (n = 30) and amastigotes stage (n = 10) to AmB (0.221 ± 0.108µM; 0.170 ± 0.027µM), PMM (13.49 ± 6.92 µM; 10.80 ± 4.12 µM), MLT (3.77 ± 1.77 µM, 3.63 ± 1.72 µM), respectively. This study findings were comparable with the susceptibility test reported against L. donovani reference strain at both promastigote stage and amastigote stage in previous study¹⁰ to AmB (IC50s, 0.6 to 0.7 µM;, 0.1 to 0.4 µM) and MLT(IC50s, 0.4 to 3.8 µM; 0.9 to 4.3 µM ) whereas, susceptibility to PMM was also in range with previous study reported susceptibility against L donovani at promastigote stage (IC50, 6–50 µM ) and at intracellular stage (IC50,8–48.81 µM )^11,14. In comparison with the mean inhibitory concentration (IC₅₀) values reported in literature to be used as tentative “Breakpoint” estimates for categorizing drug-susceptibility and drug-resistance for all species of the L. donovani complex by Maes et al.⁹. ; our strains exhibited higher mean inhibitory concentration (IC₅₀) for AmB and lower mean inhibitory concentration (IC₅₀) for PMM at the amastigote stage. This observation needs further follow-up to determine if the difference may be associated with emerging acquired resistance. Furthermore, variations in the in vitro susceptibility of Leishmania have been associated with distinct plasma membrane composition in different species and lipid content, which seem to influence drug uptake and, consequently, their activity¹².

The calculated IC₅₀ of the three drugs (AmB, PMM, and MLT) against Leishmania strains isolated from relapse VL patients with HIV coinfection was higher at both promastigote and amastigote stages compared with the mean IC₅₀ values obtained against strains from primary VL patients with or without HIV coinfection. The difference in the AmB IC₅₀ values concerning amastigote stages of Leishmania in isolates obtained from relapse and primary VL was found statistically significant (p = 0.03). On the contrary, the difference in the calculated IC₅₀ values for AmB against promastigote stage and IC₅₀ values for PMM and MLT against both stages in primary and relapse VL was not significant. The variation might be attributed to repeated exposure of parasites to tested drugs leading to a gradual shift of the effective IC₅₀, possibly accumulated through selection. In vitro studies have shown that Leishmania parasites can develop resistance when cultured in the presence of MLT¹³, PMM¹⁴, and AmB¹⁵. A previous study¹⁶ conducted on immunocompromised patients with treatment failure observed that the in vitro inhibitory concentration of AmB increased throughout successive treatments, suggesting changes in susceptibility due to exposure to the drug. Another report¹⁷ conducted on L. donovani also confirmed that isolates recovered from relapsed VL patients are more likely to be less susceptible to MLT than isolates recovered from primary VL patients. In another study, the sensitivity of isolates obtained over time from HIV-Leishmania co-infected patients submitted to several courses of treatment with AmB did not seem to change¹⁸. Continuous exposure of parasites to medication is one of the causes for the appearance of a decrease in susceptibility; however, this is unlikely to explain the situation for PMM, as this drug is not the most commonly used medication in VL-HIV coinfected patients.

On the basis of the tentatively proposed ‘threshold’ values of the current antileishmanial reference drugs for all species of the L. donovani complex reported⁹, our strains show: lower susceptibility AmB at amastigote stage while it was comparable at promastigote stage; higher susceptibility to _PMM at amastigote stage while it was comparable at promastigote stage; and comparable IC₅₀ value of MLT at both amastigote and promastigote stages. A few strains obtained from multiple relapse cases of VL coinfected with HIV showed a decreased susceptibility either at amastigote stage or promastigote stage to AmB and/or MLT. One strain collected from a patient with 5th relapse VL coinfected with HIV and a history of treatment failure to a combination of L-AmB and MLT, exhibited a shift in mean IC₅₀ value at both parasite stages for AmB and MLT. A strain isolated from a single HIV coinfected VL patient (see Table 6) after second and third relapses showed increasing IC₅₀ values in the last relapse when tested against AmB and MLT in both promastigote and amastigote stages. A higher mean IC₅₀ value for AmB, PMM and MLT against promastigote and amastigote stage was found at the 3rd relapse ([0.452 ± 0.182 µM], [32.17 ± 10.68 µM], [6.78 ± 2.36 µM]) compared to second relapse [0.282 ± 0.173 µM], [16.62 ± 7.20 µM], [6.34 ± 1.65 µM]), respectively.

These findings might show an increased tolerance of the parasite to AmB and MLT in response to repeated treatments. The relationship between clinical treatment failure and drug resistance in leishmaniasis has been indefinable and debated, although drug pressure is considered as a risk factor for the selection of resistant microbial pathogens¹⁹. Obviously, resistance does not always emerge during treatment, and it could be expected that host, as well as parasite factors, are involved not only in the clinical response but also in determining the degree and impact of the selective pressure imposed by treatment. Treatment failure has previously been linked with age and the amount of treatment^20,21. Treatment failure in visceral leishmaniasis, diffuse cutaneous leishmaniasis, and mucosal leishmaniasis highlight the dependency of the therapeutic response on the host immune response. Numerous studies have shown the importance of T-cell-mediated immunity in the prevention of relapses²². This may explain the high frequency of relapses observed in HIV co-infected VL patients, even after effective treatment.

Previous studies have reported that only 38–87% of HIV co-infected VL patients achieve clinical response and that a parasitologic response is observed in 38–81% of cases²³. The HIV/VL coinfected patients have a high parasite burden and a weak immune response, respond poorly to treatment and have a high relapse rate. The relapsing course is probably due to reactivation caused by the inability of the immune system to eradicate the infection, with the persistence of infection after an apparent cure. In some instances, the development of drug resistance could contribute to the therapeutic failure and relapsing course as observed in HIV-infected VL patients^24,25. Treatment options for patients with VL and HIV coinfections in East Africa are still inadequate. In Ethiopia, a systematic review and meta-analysis on the treatment outcome of VL in HIV positive and HIV negative patients have shown the end of the treatment success rate of 66.8% and 93.2%, respectively, while the corresponding treatment success rate after 6 months follow up were 42.5% and 87.9%. The study also noted greater treatment failures in HIV coinfected VL patients²⁶.

In summary, the overall in vitro susceptibility patterns of our strains to the three antileishmanial drugs, particularly AmB and MLT, revealed a slightly decreased susceptibility of strains from VL relapses with HIV coinfection compared with strains obtained from primary VL patients with or without HIV coinfection. In addition, some strains from multiple VL relapses coinfected with HIV exhibited a shift in mean inhibitory concentration to AmB and MLT at promastigote and or amastigote stages, despite the fact that no isolate with resistant phenotype was found on the basis of the breakpoint reported in the literature.

Author contributions

Conceptualization and design the experiment: A.H and H.F, Consenting process, clinical data completion and sample collection: S.A and Z.M, Culturing samples: R.M and A.Y, Investigation: S.T, M. T, and A.H, Data analysis: S.T and D.K, Funding acquisition and reagents contribution: S.T, A.H, Supervision: A.H, MT, and H. Writing – original draft: S.T, Writing – review & editing: S.T, D. K, B.K, A. H and M.T.

Funding

This study was supported Addis Ababa University.

Data availability

Data is provided within the manuscript.

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval

All methods were carried out in accordance with ARRIVE guidelines (https://arriveguidelines.org).