Irina Bondareva (1), Sergey Zyryanov (1,2), Marina Ivzhits (1,2), Mikhail Chenkurov (1,2)
(1) RUDN University, Moscow, Russia, (2) State Budgetary Institution of Healthcare "City Clinical Hospital No. 24 of the Moscow City Health Department", Russia
Introduction: Premature infants have a high risk of developing neonatal seizures [1]. Levetiracetam is a second-generation broad spectrum antiepileptic drug (AED). Routine TDM of LEV is unnecessary due to its favourable pharmacokinetic (PK) profile: linear pharmacokinetics, predictable dose-concentration relationship, wide therapeutic index, minimal side effects, and unlikely clinically significant drug-drug PK interactions [2]. However, significant and rapid changes in pharmacokinetics may occur in neonates, and LEV pharmacokinetics in neonates appears to be highly variable, which makes neonates a special population for TDM use [3, 4].
Objectives:
- Determine the influence of clinically relevant covariates on interindividual variability of PK parameters for intravenously (IV) administered LEV-monotherapy for the treatment of seizures in preterm neonates
- Develop a population model of LEV pharmacokinetics from sparse TDM data of preterm neonates with seizures
Methods: TDM data were routinely collected in the neonatal intensive care unit. LEV daily doses (18.5 – 51, median = 30 mg/kg/24 h) were administered by IV infusion bid. For this PK analysis, demographic and clinical characteristics as well as LEV administration details and measured concentrations were retrieved from patients’ records retrospectively. Blood samples were collected at predose (trough levels) and at the end of 30 min IV infusion (peak levels) at different treatment days, each included subject had 2 – 14 (median = 4) LEV levels. LEV concentrations were measured by high performance liquid chromatography. PK analysis was performed using the Pmetrics software based on the one-compartment model and TDM measurements (peak-trough strategy) of 31 preterm neonates who received IV LEV-monotherapy. Repeated TDM data of 25 preterm neonates were used to estimate the model predictability taking into account change in body weight and maturation process.
Results: All included subjects had a gestational age (GA) 22 – 32, median = 26 wk; at blood probe, post-conceptual age (PCA) was 24.4 – 38.3, median = 34.5 wk; weight 0.62 – 2.4, median = 1.4 kg; the glomerular filtration rate (GFR) estimated using the Schwartz equation (the coefficient was set to 0.33) 8.6 – 38.3, median = 18.6 ml/min/1.73 m2. The geometric means (min – max) of predose and peak LEV concentrations were 8.8 (4 – 19.4) and 16.3 (8.4 – 63.3) μg/ml, respectively. The majority of neonates (more than 90%) had serum trough levels above 6 μg/ml. The median values for total body clearance and elimination serum half-life (T1/2) of LEV were 1.6 ml/min/kg and 14.6 hours, respectively, with interindividual CV > 50%. In average, clearance and GFR increased, and T1/2 decreased with increasing GA. The GFR was strongly correlated with the PCA, and in most patients with repeated measures during the first two months, the GFR increased with post-natal age. For 33 weeks of GA/PCA considered as a cut-off for the capacity of drug renal elimination, the mean elimination serum half-life of LEV was statistically significant higher for less versus more than 33 weeks group (p=0.023). The regression analysis revealed that total body weight at dosing significantly influenced the LEV pharmacokinetics, and GFR significantly influenced LEV elimination.
Conclusion: Despite a small number of studied patients and narrow patient population which included subjects with relatively normal renal function, the results demonstrated that extremely premature and premature newborns, in whom extensive pharmacokinetic changes occur over the first months after birth, require monitoring of their LEV therapy. Bayesian approach for LEV concentration prediction based on minimum sampling steady-state or non-steady-state TDM data appear to be useful. This is especially important, because in preterm neonates, their PK parameters changed significantly during the first months of life and often before steady-state was reached. Bayesian feedback adaptive control and population modeling can improve LEV dosage adjustment for this patient group during the first few months of life.
References:
[1] Scher, M. S. Neonatal seizure classification: A fetal perspective concerning childhood epilepsy. Epilepsy Res., 70, 41–57, 2006.
[2] Patsalos PN. Clinical pharmacokinetics of levetiracetam. Clin Pharmacokinet, 43:707‐724, 2004.
[3] Patsalos PN, Berry DJ, Bourgeois BF, Cloyd JC,Glauser TA, Johannessen SI, et al. Antiepileptic drugs–best practice guidelines for therapeutic drug monitoring: a position paper by the subcommission on therapeutic drug monitoring, ILAE Commission on Therapeutic Strategies. Epilepsia, 49(7):1239-76, 2008.
[4] Jarvie D., Mahmoud SH Therapeutic drug monitoring of levetiracetam in select populations. J Pharm Pharm Sci, 21(1s): 149s-176s, 2018.
Reference: PAGE 28 (2019) Abstr 9016 [www.page-meeting.org/?abstract=9016]
Poster: Drug/Disease Modelling - CNS