Julia M. Penning de Vries (1), Birgit C.P. Koch (2), Enno Wildschut (1), Brenda. C.M. de Winter (2), Matthijs de Hoog (1)
(1) Erasmus MC, Department Intensive Care, Sophia Children’s Hospital; Rotterdam, (2) Erasmus MC, Department of Hospital Pharmacy, Rotterdam
Objectives: Status epilepticus (SE) is a life threatening event which requires immediate medical intervention. The overall mortality in children treated for refractory generalized convulsive SE is 16%-64% [1]. Little consensus consists regarding treatment of refractory SE and there are no randomized, blinded studies in either adults or children. Pentobarbital is the last resource to treat SE. In daily practice dosing is adjusted based on EEG and pentobarbital drug level. However, too low or too high drug levels are often seen. The aim of this study was to design a pharmacokinetic model of pentobarbital, which can be incorporated in therapeutic drug monitoring.
Methods: The design of the study was a retrospective, single-center analysis of medical records of all consecutive pediatric patients who received pentobarbital-coma for refractory generalized convulsive status epilepticus from 2007 through 2012 at the Department of Pediatric Intensive Care, Sophia Children’s Hospital, Rotterdam for the primary dataset and between 2013 and February 2015 for de validation set. Inclusion: age 1 week – 18 years, exclusion all patients receiving pentobarbital for other reasons than SE. For pharmacokinetic analysis NONMEM® version 7.2 was used. Demographic and laboratory parameters were evaluated as covariates. Allometric scaling was used to adjust for differences in bodyweight using a factor of 0.75 for clearance and 1 for volume of distribution. The parameters of the final model were introduced in MW-Pharm to validate for use in daily practice.
Results: 16 patients were included in the primary set (median age 76 [17-1363] days, median weight 5.4 [3-19] kg. Mean loading dose was 8.8 [0-15.65] mg/kg, mean maintenance dose 4 [1-10] mg/kg/h. In the validation set 6 patients were included (median age 120 [63-460] days, median weight 6.1 [6-10] kg). Mean loading dose was 15.4 [15-16.6] mg/kg and mean maintenance dose 4.5 [3-7] mg/kg/h. The data was best described in a two-compartment model (V1 0.89 L/kg (58% BSV ), V2 1.46 L/kg, with clearance of 5.12 L/h (45% BSV). In children below 1 year, clearance was increased. NPDE of the validation dataset using the final model showed accurate results. Finally, prediction of the final data point of the validation set and comparing it with the real data set, resulted in all predicted data being <20% different than the real data.
Conclusion: The PK model was validated and has been incorporated in MWPharm to be used in daily practice.
References:
[1] Kim SJ, Lee DY, Kim JS. Pediatr Neurol 2001;25:217–220.
Reference: PAGE 25 (2016) Abstr 3698 [www.page-meeting.org/?abstract=3698]
Poster: Drug/Disease modeling - Paediatrics