Laura Iavarone (1), Silvia Maria Lavezzi (1), Chao Chen (2)
(1) Parexel International; (2) GlaxoSmithKline R&D
Objectives: There are broadly two types of approaches for predicting the therapeutic dose of anti-infectives from animal experiments. Highly mechanistic approaches can reflect various aspects of the infection in human [1-3]. These approaches require extensive in vitro experiments to correct the animal models’ inadequate reflection of the disease pathology in humans. The other type is rooted in the practice of clinical pharmacy of antibiotics, where a drug’s efficacy is deemed to be either exposure- or time-dependent [4-6]. These methods rely on the so-called PK-PD indices such as Cmax, AUC and the duration that in vivo concentration remains above an effective concentration, e.g., Time > MIC, Time > EC50, etc. While these latter methods are simpler, more time-efficient and more cost-effective, they lack adequate reflection of pharmacology, which drives efficacy. This report describes a simulation-based assessment of a new potential PK-PD index, which is pharmacology-based yet relatively simple: the area under the effect curve, or AUEC.
Methods: In vitro, the pharmacology of a hypothetical anti-infective was described by an Emax PK-PD model. In vivo efficacy was described by a saturable pathogen dynamic model where the drug acted on the net growth rate, also according to an Emax model. For simplicity, the PK of the drug was assumed to be mono-exponential disposition following first-order absorption; and the model parameters were allometrically scaled from values similar to those typically seen in human. The AUEC was calculated as the area under the in vivo effect curve, which was constructed by replacing each concentration of the in vivo PK curve with its corresponding pharmacology derived according to the in vitro pharmacology PK-PD function. Dose fractionation efficacy experiments in rodents were simulated for a dose range of 0.1 – 3.0 mg/kg/day, with several dosing intervals of 3 to 24 hours. Ten animals per dose per regimen were simulated. A broad range of pharmacology properties of the drug and the experimental conditions were simulated, as described below in the Results section. In each situation, the correlation between the efficacy readout and the PK-PD indices – Cmax, AUC, T>EC50, T>EC90 and AUEC – were calculated.
Results: Assessed over a range of treatment durations, the strong correlates of efficacy were AUC (R2 0.88-0.93) and AUEC (R2 0.96-0.98). When the drug efficacy (Emax) was assumed to vary in relation to the in vivo net growth rate of the pathogen, the strong correlates were again AUC (R2 0.71-0.88), T>EC50 (R2 up to 0.84) and AUEC (R2 0.78-0.96). Across different simulated infection stages, the strong correlates were AUC (R2 0.85-0.89), T>EC50 (R2 up to 0.83) and AUEC (R2 0.95-0.96). Under the conditions of a chronic treatment by a highly efficacious drug for a late-stage, more detailed assessments were conducted. Scenarios of the drug potency (relative to dose range) and its in vitro–in vivo miss-match, and the steepness of the pharmacology/efficacy curves and its in vitro–in vivo miss-match, revealed the R2 values to be: AUC (0.36-0.89), Cmax (up to 0.66), T>EC50 (up to 0.79) and AUEC (0.76-0.97). Furthermore, dosing regimen – by extension, PK profile shape – showed clear impact on the predicted efficacy dose for Cmax, T>EC50 and T>EC90, moderate impact for AUC, and negligible impact for AUEC.
Conclusions: The AUEC approach represents a middle ground between the more sophisticated mechanistic modelling and the largely PK-based traditional PK-PD index approach. It is simpler than the former yet connects more closely with pharmacology than the latter. Under the conditions tested, the AUEC almost always showed a strong correlation with in vivo efficacy. Its correlation with in vivo efficacy was generally independent of the in vivo PK profile. These observations suggested AUEC could be a more robust efficacy predictor than the conventional PK-PD indices in certain situations. If these findings from simulation are confirmed in real-life drug development programs, using AUEC as a tool to translate in vivo efficacy from animal to human would make dose fractionation experiments less relevant. This would drastically reduce animal use in preclinical efficacy experiments, leading to resource savings. Although these simulations were for anti-infectives, the AUEC concept is underpinned by the principle that pharmacology directly drives efficacy. As such, this approach should be explored for indications beyond infectious diseases.
References:
[1] Zaloumis S, Humberstone A, Charman SA, Price RN, Moehrle J, Gamo-Benito J, McCaw J, Jamsen KM, Smith K, Simpson JA. Assessing the utility of an anti-malarial pharmacokinetic-pharmacodynamic model for aiding drug clinical development. Malar J. 2012 Aug 30;11:303.
[2] Nielsen EI, Friberg LE. Pharmacokinetic-pharmacodynamic modeling of antibacterial drugs. Pharmacol Rev. 2013 Jun 26;65(3):1053-90.
[3] Best K, Guedj J, Madelain V, de Lamballerie X, Lim SY, Osuna CE, Whitney JB, Perelson AS. Zika plasma viral dynamics in nonhuman primates provides insights into early infection and antiviral strategies. Proc Natl Acad Sci U S A. 2017 Aug 15;114(33):8847-8852.
[4] Ambrose PG, Bhavnani SM, Rubino CM, Louie A, Gumbo T, Forrest A, Drusano GL. Pharmacokinetics-pharmacodynamics of antimicrobial therapy: it’s not just for mice anymore. Clin Infect Dis. 2007 Jan 1;44(1):79-86.
[5] Turnidge J, Paterson DL. Setting and revising antibacterial susceptibility breakpoints. Clin Microbiol Rev. 2007 Jul;20(3):391-408.
[6] European Medicines Agency. Guideline on the use of pharmacokinetics and pharmacodynamics in the development of antimicrobial medicinal products. https://www.ema.europa.eu/en/documents/scientific-guideline/guideline-use-pharmacokinetics-pharmacodynamics-development-antimicrobial-medicinal-products_en.pdf (2016). Accessed 09 March 2021.
Reference: PAGE 30 (2022) Abstr 10228 [www.page-meeting.org/?abstract=10228]
Poster: Methodology - Other topics