Effect of time, injury, age and ethanol on interpatient variability in valproic acid pharmacokinetics after traumatic brain injury

Traumatic brain injury (TBI) results in an increase in hepatic metabolism. The increased metabolism is in significant contrast to a large body of in vitro and in vivo data demonstrating that activation of the host-defence response downregulates hepatic metabolism. Theoretically, this occurs because of activation of the pro-inflammatory cytokines tumour necrosis factor-alpha, interferon-gamma, interleukin (IL)-1 and IL-6. As part of a large double-blind, placebo-controlled clinical trial evaluating the use of valproic acid for prophylaxis of post-traumatic seizures, we obtained extensive valproic acid concentration-time data. Valproic acid is a hepatically metabolised, low extraction-ratio drug. Therefore, unbound clearance (CL(u)) is equal to intrinsic or metabolic clearance. The objective of this study was to evaluate the time-dependent effects of TBI on the pharmacokinetics of total and unbound valproic acid with the goal of identifying patient factors that may predict changes in total clearance (CL) and CL(u). In addition, by determining the factors that influence the magnitude and time course of induction of hepatic metabolism and understanding their interaction with the host-defence mediators, we can further our insight into the mechanism(s) responsible for the changes in CL and CL(u). Valproic acid plasma concentration data were obtained from 158 TBI patients. Unbound valproic acid plasma concentrations were estimated using total valproic acid plasma and albumin concentrations following a Scatchard equation binding model previously developed in a subset of TBI patients. The effect of 13 patient factors on CL and CL(u) was evaluated initially in a univariate analysis. The significant factors were then included in a multiple linear regression analysis by use of step-wise selection and forward selection procedures. CL and CL(u) were significantly increased after TBI in a time-dependent manner. The average increase was >75% by weeks 2 and 3 post-injury. The magnitude of the induction of CL was increased with decreased albumin concentrations, in addition to the presence of ethanol on admission, increased severity of head injury, tube feeding and total parenteral nutrition (TPN). The magnitude of induction of CL(u) was increased by older age, presence of ethanol on admission, increased severity of head injury, tube feeding, TPN, and if the patient had a post-injury neurosurgical procedure. The time to normalisation of CL(u) was significantly longer in pati