Population pharmacokinetic modelling and simulation of tranexamic acid in adult trauma patients

Abstract

Aims: The aim of this study is to describe the disposition of tranexamic acid (TXA) in adult trauma patients and derive a dosing regimen that optimizes exposure based on a predefined exposure target.

Methods: We performed a population pharmacokinetic (popPK) analysis of participants enrolled in the Tranexamic Acid Mechanisms and Pharmacokinetics in Traumatic Injury (TAMPITI) trial (≥18 years with traumatic injury, given ≥1 blood product and/or requiring immediate transfer to the operating room) who were randomized to a single dose of either 2 or 4 g of TXA ≤2 h from time of injury. PopPK analysis was conducted using nonlinear mixed-effects modelling (NONMEM). Simulations were then performed using the final model to generate estimated plasma TXA concentrations in 1000 simulated participants. Dosing schemes were evaluated to determine maintenance of TXA plasma concentrations >10 mg/L for ≥8 h after administration of the initial dose.

Results: TXA PK was best described by a two-compartment model with proportional residual error and allometric scaling on all parameters. Platelet count, skeletal muscle oxygen saturation measured by near-infrared spectroscopy and interleukin-8 concentration were significant covariates on TXA clearance. Based on simulations, a 2 g IV bolus dose, repeated 3 h later, best achieved the target exposure.

Conclusions: According to simulations from a popPK model of TXA, a 2 g IV bolus with a repeated dose 3 h later would be most likely to maintain concentrations >10 mg/L for 8 h in >95% of adult trauma patients and should be considered for patients with ongoing haemorrhage.

Keywords: clinical pharmacology; emergency medicine; pharmacokinetics.

Figure 1.. Concentration-time profiles for recipients of 2 g and 4 g doses of tranexamic acid.

Figure 2.. Observed versus population predicted and individual predicted plots.

2A. Observed versus population predicted plot. 2B. Observed versus individual predicted plot.

Figure 2.. Observed versus population predicted and individual predicted plots.

2A. Observed versus population predicted plot. 2B. Observed versus individual predicted plot.

Figure 3.. Monte Carlo simulations of multiple-dose regimens of tranexamic acid.

All simulations were performed using median population values for informative covariates. The target for simulations was maintenance of concentrations above 10 mg/L (solid black lines) for 8 hours. The median (solid blue lines), 2.5^th and 97.5^th percentile (red dotted lines) of simulated concentrations for each regimen are shown. A: 2 g IV bolus followed by a 1 g IV bolus 3 hours later. B: 2 g IV bolus followed by a 2 g IV bolus 3 hours later. This regimen was identified as most reliably achieving the target concentration of 10 mg/L at 8 hours for 95% of simulated subjects. C: 3 g IV bolus followed by a 0.5 g IV bolus 3 hours later. D: 3 g IV bolus followed by a 1 g IV bolus 3 hours later. E: 1 g IV bolus (over 10 minutes) followed by 1 g as an 8-hour infusion.