Lavezzi Silvia M. (1), Iavarone Laura (1), De Rycker Manu (2), Read Kevin (2), Cebrian David (3), Miles Tim (4), and Chen Chao (5).
(1) Clinical Pharmacology, Modelling and Simulation, Parexel International, (2) Drug Discovery Unit, Wellcome Centre for Anti-Infectives Research, University of Dundee, (3) DMPK, In Vitro In Vivo Translation, GlaxoSmithKline (4) Global Health R&D, GlaxoSmithKline, (5) Clinical Pharmacology Modelling and Simulation, GlaxoSmithKline
Objectives: Visceral leishmaniasis (VL) is an infectious disease caused by protozoan parasites of the genus Leishmania. If untreated, VL is typically fatal. None of the available treatment options are ideal for use, due to teratogenicity, resistance, clinical relapse, prolonged treatment regimens and parenteral administration [1,2,3,4]. Thus, there is a need for new oral VL therapies.
To guide anti-infective dose selection, pharmacokinetic/pharmacodynamic (PK/PD) indices are commonly used as in vivo efficacy drivers. Simulations showed that the area under the effect curve (AUEC) could be an attractive alternative to standard PK/PD indices: area under the concentration curve (AUC), maximum concentration (Cmax), and time of in vivo concentration above a cut-off (e.g., T>EC50) [5].
Here, standard PK/PD indices and AUEC were evaluated retrospectively as in vivo efficacy drivers on non-clinical data for 6 oral candidate VL therapies (Compounds A, B, C, D, E, and F).
Methods: Efficacy was evaluated in vitro as percentage (%) reduction of intracellular Leishmania donovani (L.d.) parasites over a range of constant concentrations. A sigmoidal Emax curve fitted the exposure-response data (EC50=in vitro concentration providing 50% parasite reduction; γ=shape parameter).
In vivo, mice were inoculated with L.d. parasites; treatment started 7 days post-infection and continued for 5 or 10 days. On the last dosing day, the % parasite reduction was assessed. Total daily doses ranged from 6 to 100 mg/kg, with twice- (BID) or once-daily dosing (QD). For each compound, a one-compartmental model was fitted (NONMEM 7.3) to total blood PK on last dosing day. With in vivo PK model and in vitro EC50 and γ, the following were derived: AUC over 24 hours, Cmax, T>EC50, T>EC90 (EC90=91/γ × EC50), and AUEC over 24 hours. The latter was calculated as the integral of 100×C(t)γ/(EC50γ + C(t)γ), where C(t) is the in vivo concentration over time. All indices were derived in two scenarios, assuming that in vitro concentrations (EC50, EC90) are comparable to in vivo (1) free plasma or (2) total blood concentrations. Based on % parasite reduction vs PK/PD index data, the best in vivo efficacy driver was assessed via estimation and evaluation (via graphics and R2) of sigmoidal Emax models (drc package, R 3.5.2).
Results: Scenario 1 (in vitro concentrations ≈ in vivo free plasma) appeared preferable to Scenario 2 (in vitro concentrations ≈ in vivo total blood) for Compounds A and F, based on observed data dispersion and higher R2. The opposite was observed for Compounds B, C, and E, while for compound D no scenario was superior to the other.
In vivo data clusters possibly due to dosing interval were observed for Compound A (n=15 for QD, n=12 for BID) and B (n=3 for QD, n=12 for BID). This was not observed for Compound C (n=3 for QD, n=15 for BID), likely because almost all data were approximately at effect plateau. No dose fractionation was performed for Compounds D, E and F (n=9 each for BID); Compounds D and F displayed huge variability in in vivo efficacy at the lowest dose level (6 mg/kg/day).
The best efficacy drivers were AUC and Cmax for Compound A (R2=0.87-0.88); scenario 2 AUEC and T>EC50 for Compound B (R2=0.81); AUC and scenario 2 T>EC90 and AUEC for Compound D (R2~0.7); and AUC, Cmax, and scenario 1 AUEC for Compound F (R2~0.8). For Compound C and E, all scenario 2 indices except T>EC90 had R2~0.9 or higher, due to most data being at effect plateau.
Conclusions: The newly proposed PK/PD index, AUEC, appears promising, being in 5/6 cases among the best in vivo efficacy drivers. However, this retrospective analysis highlighted the importance of appropriate in vivo data. The number of subjects per experiment was low and no proper dose fractionation was performed; hence it was not possible to thoroughly evaluate the benefit of AUEC, in particular its ability to reduce the need of dose fractionation. In addition, R2 is commonly used to establish the best efficacy driver in anti-infectives, but this value, while measuring how close the fitted curve is to the data, does not give any indication on the sensibility of such curve. Furthermore, a sigmoidal Emax model is usually applied to fit in vitro and in vivo data, and, while this appears frequently appropriate, it might not always be the case. Further work (e.g., properly dose-fractionated in vivo experiments) is warranted to adequately explore the potential role of AUEC in VL therapy development.
Acknowledgments: We thank Drugs for Neglected Disease initiative (DNDi) for providing in vitro and in vivo data.
References:
[1] Hendrickx S, Guerin P, Caljon G, Croft SL and Maes L. Evaluating drug resistance in visceral leishmaniasis: The challenges. Parasitology, 2018;145:453-63.
[2] Sundar S, Singh A, Rai M, Prajapati VK, Singh AK, Ostyn B et al. Efficacy of miltefosine in the treatment of visceral leishmaniasis in India after a decade of use. Clinical Infectious Diseases, 2012;55:543-50.
[3] den Boer M, Alvar A, Davidson RN, Ritmeijer K and Balasegaram M. Developments in the treatment of visceral leishmaniasis. Expert Opinion on Emerging Drugs, 2009;14:395-410.
[4] Mueller M, Ritmeijer K, Balasegaram M, Koummuki Y, Santana MR and Davidson R. Unresponsiveness to AmBisome in some Sudanese patients with kala-azar. Transactions of the Royal Society of Tropical Medicine and Hygiene, 2007;101:19-24.
[5] Chen C, Iavarone L, Lavezzi SM. A simulation-based evaluation of the area under the effect curve as an efficacy driver for anti-infectives. ASCPT (virtual conference) 2021.
Reference: PAGE 30 (2022) Abstr 10119 [www.page-meeting.org/?abstract=10119]
Poster: Methodology - Other topics