Population PK/PD modeling of blood- and liver-stage activities of cabamiquine in malaria
Perrine Courlet, Claude Oeuvray, Wei Gao, Akash Khandelwal
Merck KGaA, Darmstadt, Germany
Objectives:
Cabamiquine is a novel antimalarial under development for the treatment and prevention of malaria, active against multiple life-cycle stages of Plasmodium. It is important to have a good understanding of cabamiquine activity against all life-cycle stages. The blood stage activity of cabamiquine is well characterized (1). Modeling liver stage activity of antimalarial treatment is challenging due to the lack of direct measurement of parasites in hepatocytes. The objective of this analysis was to characterize cabamiquine effects on the liver and blood stages of malaria infection.
Methods:
Pharmacokinetics (PK) and longitudinal parasitemia data from two clinical studies conducted in healthy volunteers were used: (a) MS201618-0013 consisting in a first part investigating the safety, tolerability and PK of cabamiquine following single ascending dose (n=49, 50-2100 mg) and a second part assessing antimalarial activity using an induced-blood stage malaria (IBSM) human challenge model (n=22); and (b) MS201618-0003 is a sporozoite challenge model (SpzCh) study where participants received different doses of cabamiquine (n=30, 30-200 mg) or placebo (n=9) in early and late liver-stages.
Individual PK parameters were derived using a PK model formerly established (2) for subsequent pharmacokinetic/pharmacodynamic (PK/PD) models.
In step 1, the blood-stage model was developed using data from the IBSM challenge study MS201618-0013. In step 2, liver stage model was incorporated into blood stage model, using data from the SpzCh study. System parameters were fixed based on biological knowledge (i.e. transfer function between liver and blood, parasite growth rate constant in liver, fraction of inoculated sporozoites invading hepatocytes (3)). Additionally, some parameters estimated during step 1 were fixed (i.e. baseline parasite count with IIV, parasite growth rate constant in blood with IIV, parasite clearance rate constant with IIV, delay rate constant, and Hill coefficient). Drug effect at blood and liver stages (EC50,blood and EC50,liver, respectively) were estimated.
Data below the limit of quantification (BLOQ) were handled using the M3 method. The model was evaluated by comparing the success rates obtained with model-based simulations based on the conditions of the trials against the observed data. Success rates were estimated as the proportion of participants without recrudescence (defined as > 5,000 parasites/mL and ≥ 2-fold increase within 48 hours) or with parasitemia < 100/mL within 28 days of sporozoite challenge, in IBSM and SpzCh studies, respectively. The Lixoft suite (version 2020R1) was used for model development and evaluation.
Results:
The previously developed PK model adequately described cabamiquine concentrations. Overall, 551 parasitemia measures were evaluable and 60% (837/1388) were BLOQ. Since the effect of cabamiquine at the liver stage could reduce the density of parasites reaching blood circulation, estimated EC50,blood value in SpzCh population was lower than that in IBSM population. The minimum inhibitory concentrations (MIC) at blood stage were derived from model parameters and calculated to be 6.57 ng/mL (95% CI 6.15-6.91 ng/mL) and 1.38 ng/mL (95% 1.29-1.47 ng/mL) for IBSM and SpzCh populations, respectively. The MIC in the liver was estimated to a lower value of 0.75 ng/mL (95% CI 0.67-0.83), indicating a more potent activity of the drug in the liver than in the blood.
Despite a large variability likely due to the small number of participants, model-based simulations highlighted the adequate predictive performances of the model, with simulated confidence intervals (CI) of success rate overlapping the observed ones. For instance, in the SpzCh population receiving 80 mg of cabamiquine at early liver-stage, the observed and predicted median success rates were 83% [CI95% 36-100%] and 67% [CI95% 33-100%], respectively.
Conclusions:
A PK/PD model was successfully developed to characterize the activity of cabamiquine on both blood and liver stages of malaria. The model suggested higher potency of cabamiquine at liver stage compared to blood stage which is consistent with the observed clinical data in healthy volunteers requiring lower doses for chemoprevention than cure. This framework provides a quantitative understanding of the antimalarial activity of cabamiquine and can be used to support dose selections and study designs for future cure and chemoprevention trials.
References:
- (1) Wilkins et al, PK/PD modelling of M5717 in malaria. PAGE meeting 2021. Abstr 9588.
- (2) Wilkins et al, Population pharmacokinetics analysis of M5717, a novel antimalarial agent. PAGE meeting 2019. Abstr 9044.
- (3) Cherkaoui-Rbati et al, A Pharmacokinetic-pharmacodynamic model for chemoprotective agents against malaria. CPT-PSP. 2022.