Visceral leishmaniasis (VL), the most serious of the various clinical presentations of Leishmania infection, is fatal if not treated. In East Africa, VL is caused by Leishmania donovani.  The disease is characterized by fever, weight loss, hepatosplenomegaly, lymphadenopathy, anaemia, leucopenia and thrombocytopenia. There are 200-400,000 estimated new cases of VL occurring annually and ninety percent of cases occur in six countries: Bangladesh, Brazil, Ethiopia, India, Sudan and South Sudan. Recent epidemics claimed thousands of lives in South Sudan. 

The WHO recommended treatment for VL in the Eastern Africa region was revised in 2010 from sodium stibogluconate (SSG) monotherapy for 30 days to a combination of SSG and paromomycin (PM) administered for 17 days, following a phase III trial conducted in the region. The efficacy in the intention to treat population using a complete case analysis was 91% for SSG & PM combination in this trial. The new treatment is an improvement to the SSG monotherapy, but a few limitations still preclude its general use. Although the treatment regimen is shorter, it still requires 17 days of painful injections, necessitating patients being hospitalised during the whole treatment period.  In addition, there are life-threatening toxicities associated with the use of antimony-based treatments such as SSG. These include cardiotoxicity and pancreatitis. Therefore there is a need to explore alternatives that are efficacious, safe, short duration, affordable and suitable to be used in remote areas.

To evaluate alternative regimens of VL treatment in Eastern Africa, a phase II clinical trial (LEAP0208) has been conducted in Kenya and Sudan with currently available drugs.  This study assessed the safety, efficacy, pharmacokinetics (PK) and pharmacodynamics (PD) of the combinations Ambisome+SSG, Ambisome+miltefosine and miltefosine monotherapy. Results of trials showed that none of the treatment regimens used reached the satisfactory efficacy level of > 90% to be taken forward in a Phase III trial to compare with the current WHO recommended treatment regimen of SSG+PM.  Ambisome+SSG showed a cure rate of 87% at 6 months follow-up, Ambisome+miltefosine 82%, and miltefosine as monotherapy had 72% efficacy.

Results from the PK and PD studies from LEAP0208 and AMBI0106 indicated that the combination Ambisome+miltefosine regimen could potentially be optimized to reach higher efficacy. However, it would require increased daily dosing as well as longer treatment duration and result in a high cost therapy, which together with other safety and stability limitations would not meet the DNDi VL target product profile (TPP).

One of the key characteristics of the VL TPP is to develop new treatments that can be used in remote areas where VL occurs. In this context, an oral efficacious and safe drug would be more adapted to field conditions, to allow short hospitalization (if needed part of the treatment can be done at home) and not requiring safety monitoring such as laboratory evaluations.

In order to respond to this need, upstream research at DNDi has been focused on identification of new chemical entities (NCEs) that meet these requirements. Currently, a phase II PoC study is under development to assess efficacy and safety of fexinidazole, an oral treatment that is under development for human African trypanosomiasis. Other promising NCEs are expected to transition from pre-clinical to clinical research in the coming years, bringing new opportunities for innovation in VL therapy.  

Therefore, it was decided to focus the development on optimizing the use of miltefosine for potential combinations with another oral drug offering a new treatment option with good efficacy and safety profile, which would reduce hospitalization and be suitable for use in remote areas.

Miltefosine, a phosphorylcholine ester of hexadecanol, is a membrane-active alkylphospholipid, it is the first effective oral agent for VL. Miltefosine interferes with the membrane lipid metabolism (e.g. de novo synthesis of phosphatidylcholine) and acts through numerous interactions with cell membrane components and cell signalling pathways. It also induces apoptosis through the PI3K-Akt pathway and possibly through mitochondrial dysfunction. Numerous immune-modulatory actions of miltefosine have been described, potentially contributing indirectly to its mechanism of action in vivo.

Miltefosine is indeed the only oral drug currently available for VL treatment. Although miltefosine is potentially teratogenic, requiring contraceptive coverage in women of child-bearing potential, an oral combination containing miltefosine could particularly benefit children, the main group at risk of VL in Eastern Africa.

The miltefosine PK data extracted from LEAP0208 indicated that the dose linearly based on weight (mg/kg), calculated as 2.5 mg/kg/day, does not provide similar drug exposure in children as compared to adults. This is in line with previous published miltefosine PK data from India and Nepal (Dorlo JID 2013 and Dorlo AAC 2012). Children are under-dosed with miltefosine, and therefore significantly under-exposed to the drug, compared to adult drug exposure. Miltefosine keeps accumulating over the treatment duration; therefore the end-of-treatment concentration can be compared (maximal accumulation). In the LEAP0208 study, for the 10 day regimen, the mean end-of-treatment concentration was 14.9 μg/mL (IQR 10.4-18.8 μg/mL)  vs 17.6 μg/mL (IQR 15.9-20.9 μg/mL) (difference 18%, p=0.171, ns), respectively for paediatric (7-12y) vs adult (>12y) patients. For 28 days of treatment, this was 22.1 μg/mL (IQR 16.8-29.1 μg/mL) vs 30.2 μg/mL (IQR 24.7-36.3 μg/mL) (difference 37%, p<0.001), respectively for paediatric vs adult patients.

Miltefosine concentration accumulating over time (until end of treatment):

Miltefosine concentration at the end of treatment:

When assessing clinical outcome at D210 (6 months post treatment) by age group, in the Ambisome+miltefosine arm the final cure rate was 74% in patients aged 7-12y vs 90% in patients > 12y (p= 0.25). In the miltefosine monotherapy arm cure was achieved in 59% of patients aged 7-12y vs 86% in patients > 12y (p= 0.061). Although the study was not powered for this sub-group analysis, and the difference is not statistically significant, there is a clear trend of poorer outcome in children as compared to adults. This difference can be at least partially explained by the lower drug exposure in children. This is in line with data from Nepal and India, where the only risk factor found for miltefosine treatment failure was to have an age <12 yrs (Rijal CID 2013, Ostyn PLOS One 2014).

In previous studies in India, similar PK results had been observed, and it has been proposed that a linear miltefosine dose based on body weight (mg/kg/day) does not provide appropriate exposure for children and low weight adults, and a revised allometric dosing should be used, accounting for the observed differences in drug clearance over the whole range of body sizes of patients, to allow similar drug exposure between children and adults. In simplified practical terms this allometric dosing entails administering a relatively higher mg/kg/day dose in patients with lower body weight compared to patients with a higher body weight (Dorlo et al AAC 2012).

Anthropometric data from 973 VL patients previously enrolled in clinical trials in Eastern Africa was used to compare conventional vs allometric miltefosine dosing. There was no significant difference for patients weighting > 30kg, whereas the allometric dose corresponds to an increase in the conventional dose by 20% up to 75% in children with body weights <30kg: lower body weights correspond to a more pronounced difference in terms of mg/kg between conventional and allometric dosing.

Conventional vs allometric miltefosine dose in VL patients in Eastern Africa, with the calculated and rounded mg/kg dose on the y-axis and body weight on the x-axis:

Due to limitations of capsule strength (10mg and 50mg) and limited availability of the 10mg strength, higher doses than the prescribed 2.5mg/kg/day are currently used in routine clinical practice due to practical constraints (e.g.  the Indian national guideline for VL treatment indicates 1 capsule of 50mg for all children 2-11yrs). Therefore safety of allometric dose (especially gastrointestinal AEs) should be closely monitored, but no major safety issues are expected.

The proposed study aims to assess whether drug exposure in children can be increased to equivalent adult drug exposure by using the miltefosine allometric dose given BID for 28 days in paediatric VL patients aged 4-12y and whether this dose is tolerable. The present study is also expected to provide the basis for minimum time to reach sufficient drug exposure for miltefosine activity to guide optimal treatment duration to be used in combination therapy for visceral leishmaniasis. The PK data will be assessed in this trial using a compartmental population PK approach, which means that we can make use of a limited sampling design, significantly reducing the burden of sampling in the pediatric population. Efficacy assessment at end of treatment and 6 months follow-up will be a secondary endpoint for this study.

Since most patients in Eastern Africa are pediatric and the current miltefosine treatment is not optimal in terms of drug exposure in children, this proposed study is pivotal to further optimize miltefosine treatment in children with VL. Results from this are expected to guide how to best adjust miltefosine dose in VL patients (especially children), as well as duration of treatment when used in future combination of oral drugs.

Primary Objectives:

·       To characterize the drug exposure and PK properties of miltefosine using allometric dosing in children with primary VL in Eastern Africa, and assess whether it is equivalent to adult drug exposure

·       To identify PK parameters that are associated with parasite clearance to derive the basis for a definition of what would be sufficient drug exposure for miltefosine activity

·       To assess the safety of miltefosine allometric dose in children with primary VL in Eastern Africa

Secondary Objectives:

·       To assess the efficacy of a 28 days miltefosine allometric dose in children with primary VL in Eastern Africa

·       To identify secondary PK parameters that are associated with clinical outcome up to D210

·       To assess the pharmacodynamics of parasite clearance over time after treatment onset and its correlation with PK parameters and clinical outcome

Primary Endpoints:

Pharmacokinetics (PK):

·       Total drug exposure (e.g. area under the concentration-time curve [AUC], Css/Cmax) will be assessed. Primary PK parameters (CL, Vd etc.) will be estimated using a population PK approach employing non-linear mixed effects modeling (NONMEM) and will be compared to previously obtained miltefosine PK results from the LEAP0208 study.
Combined data from LEAP0208 and the present study will be used to identify secondary PK parameters and PK targets to attain that are associated with parasite clearance and/or clinical response, which will provide a basis to define time to reach adequate drug exposure and guide minimum treatment duration needed for miltefosine in children to be used in combination therapies.

Safety:

·       Frequency of SAEs and AEs requiring treatment discontinuation

·       Frequency and severity of adverse events

Secondary Endpoints:

Efficacy:

·       Cure at the D28 assessment, defined as recovery of clinical signs and symptoms and absence of parasites at D28

·       Cure at D210 follow-up, defined as absence of signs and symptoms of VL at D210 and no requirement for rescue administration during the trial (e.g no relapse or earlier treatment failure)

Pharmacodynamics (PD):

·       Parasite clearance over time (qualitative and quantitative), as measured by qPCR from blood samples, from baseline until D56, and at any suspect of relapse during the trial

·       Correlation of parasite quantification in blood vs tissue samples at baseline, D28 assessment and at any suspect of relapse that requires tissue aspirate for diagnosis confirmation

·       Neopterin levels (marker of macrophage activation) in blood samples measured by ELISA, from baseline until D56, and at any suspect of relapse during the trial.

ngle primary endpoint

This is a multicenter, non-comparative, open-label clinical trial to assess the PK and safety of miltefosine using an allometric dose algorithm in the treatment of children with primary visceral leishmaniasis in eastern Africa. Efficacy and PD will be assessed as secondary outcomes.

Patients who fulfill all inclusion criteria and do not present any of the exclusion criteria will be eligible to be enrolled in this study.

Inclusion criteria:

·       Patients with clinical signs and symptoms of VL and confirmatory parasitological microscopic diagnosis

·       Patients aged > 4 to < 12 years who are able to comply with the study protocol.

·       Patients for whom written informed consent has been signed by parents(s) or legal guardian

·       Weight < 30 kg

Exclusion criteria:

·       Patients who are relapse cases

·       Patients who have received any anti-leishmanial drugs in the last 6 months

·       Patients with severe malnutrition (for children aged <5 years, weight-for-height WHO reference curves by gender, z score <-3; for children 5-12 years,  BMI-for-age WHO reference curves for gender, z score < -3)

·       Patients with positive HIV diagnosis

·       Patients with previous history of hypersensitivity reaction to miltefosine

·       Patients suffering from a concomitant severe infection such as TB or any other serious underlying disease (cardiac, renal, hepatic) which would preclude evaluation of the patient’s response to study medication

·       Patients suffering from other conditions associated with splenomegaly such as schistosomiasis

·       Pregnant or lactating women or female patient in childbearing age (reached menarche)

·       Patients with haemoglobin < 5g/dl

·       Patients with WBC < 1 x 10³/mm³

·       Patients with platelets < 40,000/mm³

·       Patients with abnormal liver function (ALT and AST) tests of more than three times the normal range.

·       Patients with bilirubin more than 1.5 times the upper normal range

·       Patients with serum creatinine above the ULN for age and gender.

·       Patients with clinical signs of severe VL disease such as jaundice and bleeding

·       Patients who cannot comply with the planned scheduled visits and procedures of the study protocol

This study will be implemented taking into account that recruitment period shall last maximum of 4 months + 7 months period for patient participation in the study (28 days treatment + 6 months follow-up post treatment). Therefore, a total of 11 months from first patient in (FPI) to last patient out (LPO). Taking into account the analysis and reporting period, the study shall last 15 months.

In order to meet recruitment target, multiple sites will be prepared to initiate the trial: Kimalel and Kacheliba in Kenya; Amudat in Uganda; Dooka in Sudan. Once approval is obtained in the country, recruitment shall start, and other sites to be added ad hoc as approvals are obtained. Routine active case search and community awareness activities will be boosted to increase recruitment.

All patients in this study will be treated with:

·       Miltefosine administered in a dosage conform the daily allometric dosing algorithm (based on individual weight and height), as per Dorlo et al. (Antimicrob Agents Chemother. 2012;56(7):3864-72), divided in 2 administrations per day for 28 days.

An easy-to-use table with dosing scheme by weight and height will be provided to the investigators to define the exact daily dose to be administered.

Samples size estimation for this trial was based on the power and precision to estimate the PK parameter AUC, as this the main parameter of drug exposure of interest. This innovative approach to determine the minimum sample size needed for PK trials has been presented (Dorlo et al. Sample size estimates for a clinical trial evaluating allometric dosing of miltefosine in children with visceral leishmaniasis in East Africa. PAGE 23 (2014) Abstr 3236 [www.page-meeting.org/?abstract=3236])

In short, anthropometric data from 454 East-African VL patients aged 4-12 years were obtained from previous DNDi sponsored clinical trials. Miltefosine PK data were simulated for all pediatric patients on allometric dosing (for 28d) with an established 2-compartment population PK model using NONMEM 7.2. Secondary PK parameter AUC was calculated and used as a measure of total drug exposure. The Monte Carlo confidence interval approach was used to evaluate achieved power for various sample sizes. Clinical trials (n=1000) were simulated with a random sample of patients (n=10, 15, 20, etc.) drawn from the pediatric population.

The sample size minimally required to achieve the AUC 95%CI within a 15% precision level (85%-118%) and a power of 95% was 25 patients.

In the LEAP 0208, 12/51 (24%) patients treated with miltefosine 2.5mg/kg/day for 28 days presented at least one episode of vomiting (11 Grade 1 and 1 Grade 2). However, compliance was achieved in 100% of patients (90-110% of intended dose).

In order to account for possible non-compliance due to lower tolerability of the miltefosine allometric dose or unforeseen logistics problems with samples, an adjustment of 15% over the sample size will be incorporated, making a total of 30 patients in the trial.

Data management, analysis and reporting will be done at the DNDi Data Center in Nairobi, Kenya.

There is no randomization in this trial, as it is a single allocation arm.

Pharmacokinetics

A population PK model of miltefosine will be developed using the miltefosine PK data from this trial using the gold standard software NONMEM for non-linear mixed effects modeling.  Primary PK parameters (e.g. k_a, CL, V2, Q and V3) will be estimated using this approach, allowing for the disentanglement of between- and within-subject variability. A covariate analysis will be performed to assess whether there are significant (clinical, demographic) factors influencing any of the PK parameters. Secondary PK parameters will be derived from this model for each individual and will include: area under the concentration-time curve, elimination half-lives, Cmax/ss, time and AUC above relevant thresholds (e.g. IC90). The population PK model will be evaluated using golden standard procedures and software (Pirana, Xpose, PsN), such as goodness of fit plots and visual predictive checks (Keizer et al. CPT-PSP, 2013). Miltefosine exposure values in children on allometric dosing will be compared to similar adult results obtained in the LEAP0208 study, to assess if the overall pediatric exposure (in this case AUC_0-∞) in this pediatric population has been increased after administering the allometric dose to a similar level of miltefosine exposure in adults after treatment with miltefosine conventional dose of 2.5 mg/kg/day.

After the PK model has been developed, a population PK-PD model will be developed by integrating PK data with qRT-PCR, which will be used as a marker of parasitological clearance. Clinical outcome and measures (spleen and liver size), appropriate biochemistry values and other selected biomarkers (neopterin) will also be integrated in the PK-PD model, if applicable.

Safety

Adverse events will be reported during treatment and follow-up periods. AE’s will be tabulated by relation to study drug (ADR) and severity (based on the CTCAE v4).

Efficacy

The primary population for efficacy analysis will be the intention to treat population.

For efficacy analysis, cure at D28 or D210 will be defined as the proportion of patients with undetectable parasites and or absence of clinical signs and symptoms of VL without the need for rescue administration. The efficacy estimate will be reported with its 95% confidence interval.