Crossing barriers: Pharmacotherapy and the pediatric injured brain

Severe traumatic brain injury (sTBI) remains one of the leading causes of morbidity and mortality worldwide. Children are an especially vulnerable group given the significant developmental changes the brain undergoes throughout childhood. Analgesia and sedation are a cornerstone of supportive management in sTBI. Nonetheless only low-level evidence exists for analgosedation therapies and overall choices on type of medication and subsequent dosing are left to the discretion of the treating physician or institution. This practice is undesirable given serious concerns about safe and effective pharmacotherapy. Also, this practice is potentially unnecessary. Despite the challenges of deriving evidence-based pharmacotherapy of central nervous system (CNS) drugs in the pediatric population there are techniques and methods to overcome these.

This thesis describes the current field of analgosedation practice in pediatric sTBI and investigates different methods by which we can reach a better understanding of pharmacokinetic (PK) properties of commonly used drugs. The ultimate aim of improving pharmacotherapeutic strategies in sTBI is to minimize or prevent secondary cerebral injury and thereby improve outcome. Outcome is explored in terms of the specific patient in relation to pharmacotherapeutic interventions and in general regarding end-of-life decisions.

The data presented in this thesis are the result of international collaborative efforts from preclinical and clinical institutions (Erasmus MC, Rotterdam, The Netherlands; Red Cross War Memorial Children’s Hospital, University of Cape Town, Cape Town, South Africa; Leiden Academic Center for Drug Research, Leiden, The Netherlands, Amsterdam University medical Center, Amsterdam, The Netherlands).


Part 1 gives an overview of pharmacotherapeutic practice in pediatric sTBI.
Chapter 1 gives a general introduction to the three factors that influence pharmacotherapy: the patient, the disease and the medication. The disease investigated in this thesis is sTBI requiring invasive neuromonitoring and admission to a pediatric intensive care unit (PICU). The patient is a child which means significant developmental changes that need to be accounted for. The medication investigated is analgosedation, a cornerstone of sTBI management, and focusses on the aspect of PK (‘what the body does to the drug’). Chapter 2 provides a narrative review of international analgosedation practices in pediatric sTBI and summarizes key PK and pharmacodynamic (PD) features. Potential short and long-term adverse effects are addressed which need to be considered when choosing medication type and dose. Finally, a framework of how to advance our understanding of (pediatric) CNS pharmacotherapy is provided and includes preclinical initiatives such as physiologically-based PK models (PBPK), mathematical methods such as population-based PK models and clinical technical advances like cerebral microdialysis (MD) which enable sampling from CNS target sites. Part 1 is concluded by chapter 3 regarding the application of cerebral MD in sTBI drug therapy as detailed in a neuroscience book chapter. The spectrum of steps involved with CNS drug investigation are covered ranging from the technical aspects of MD to mathematical and statistical techniques in PK modelling as well as suggestions how to link PK data to PD markers.

Part 2 describes the results of two PK studies of commonly used drugs in the management of pediatric sTBI.
A pilot study is presented in chapter 4 whereby a physiologically (animal-)based PK (PBPK) model for morphine was translated for pediatric human use to predict morphine concentrations in the extracellular fluid of the brain (brainECF). The aim of this prospective study was to evaluate the predictive value of this PBPK morphine model for concentration time profiles of morphine in brainECF by comparing predicted values to samples from pediatric sTBI patients as retrieved by cerebral MD (n = 6). Overall the PBPK morphine model showed good predictive value both for morphine concentrations in blood plasma and brainECF. It was of interest to observe that the predictive power was more accurate if patient samples were sampled from MC catheters located in relatively ‘un-injured’ brain compared to samples from ‘injured’ brain. This alludes to the role of the blood-brain barrier in medication disposition. The successful development of this PBPK morphine model demonstrates the feasibility of such an approach in designing predictive PK models. The next step is to validate this PBPK morphine model to enable its use in further pharmacological studies, such as linking target site concentrations to PD end points and thereby improve patient specific dosing strategies. Chapter 5 describes the development and validation (both internal and external) of a population-based PK model for pentobarbital. In addition, pentobarbital dosing simulations were performed with this population-based PK model to improve our PK understanding of different dosing regimens. Retrospective data were retrieved from 36 pediatric patients who received pentobarbital for either status epilepticus (SE) or sTBI. A one-compartment pentobarbital PK model with allometrically scaled weight on clearance and volume of distribution best described the observed pentobarbital concentrations. Serum creatinine was the only significant covariate whereby it was undetermined whether this reflected actual renal failure or overall degree of critical illness. Dosing simulations showed that patients with an elevated serum creatinine failed to achieve a steady state of pentobarbital concentration within 72 hours but progressed to toxic levels. This high PK variability underlines the importance of therapeutic drug monitoring and PK model development to facilitate and adjust individual dosing.

Part 3 focusses on different aspects of outcome in sTBI as the ultimate goal of all management and research efforts is to improve outcome of this potentially devastating disease.
Oxygen is the most commonly used drug in the PICU and hyperoxia has been associated with worse outcome. The study presented in chapter 6 compares hyperoxia classification of pediatric sTBI patients during the first 24 hours in the PICU by using conventional PaO2 cutoff analysis and area-under-the-curve (AUC) PaO2 cumulative analysis. The rationale being that patient classification is crucial to how we subsequently associate hyperoxia to outcome. Seventy-one patients were included in this retrospective analysis and showed a high variability in hyperoxia classification: 52% in the PaO2 cutoff group were classified in the highest hyperoxia category versus 26% in the AUC group. Classification variability was reflected by a Pearson correlation coefficient of 0.40 (p 0.001). We consider classification by cumulative oxygen exposure better approximates (patho-)physiological circumstances due to its time and dose dependent approach. Further prospective studies are necessary to explore the association between cumulative oxygen exposure, physiological parameters and outcome.
End-of-life decisions and subsequent withdrawal of life sustaining therapy occur frequently in children with sTBI. However there are currently no formal guidelines on this topic to assist the medical team and families. The process by which such decisions are reached and the clinical factors taken into account are described in chapter 7 for a retrospective cohort of 78 patients at our PICU. Challenges in this specific patient group include meeting clinical brain death definitions in conjunction with standardized supportive management strategies (such as administering sedation) that hamper neurological examination. Prospective one-year follow-up data was also evaluated to further understand the trajectory of survivors from this cohort.
Chapter 8 outlines the potential of biomarkers for neuroprognostication in pediatric TBI in an editorial regarding a study that investigated serum levels of total tau (t-tau), a marker of axonal injury, in both healthy children and pediatric TBI patients. The study generated age-related normative data to assist in biomarker level interpretation. Awareness that there are age-related differences in biomarker thresholds is crucial and emphasizes the differences between children and adults. The multitude of other items that need to be taken into account before a transition from preclinical to clinical application can be made are detailed against the background of current international initiatives.

In part 4 the general discussion (chapter 9) reflects on the findings of the studies presented in this thesis. It places them in the broader scope of patient, disease and pharmacotherapy to discuss barriers in current pediatric CNS pharmacology and provides suggestions on how these barriers can be crossed. Chapter 10 gives a summary in Dutch and English of the thesis content.

The conclusion of this thesis is that efforts towards crossing current barriers in the development of evidence-based neuropharmacology recommendations in (pediatric) sTBI incorporate collaboration between centres and disciplines as well as innovative approaches towards studies. The latter involves alternative study designs and statistical approaches as well as technical advances such as cerebral MD to obtain target site samples. Ultimately, the heterogenic character of the underlying disease against the backdrop of a developing individual, suggest the way forward is tailored (neuro-)pharmacology by means of individualized drug PKPD templates. Such patient-specific approaches to neuropharmacology will enable more safe and effective drug therapy which should ultimately contribute to improved outcome.