Tamara van Donge

General discussion and future perspectives 237 11 concentration is known which will result in the desirable clinical outcome, then PKPD models can be used to evaluate different dosing strategies. For instance, despite the development of various standardized scoring scales, such as the Finnegan Neonatal Abstinence Severity Score or the Neonatal Narcotic Withdrawal Index, and the careful training of staff utilizing these scoring tools, the observation of response to treatment of neonates with neonatal abstinence syndrome (NAS) remains to a large extent subjective. 32,33 In addition, these scoring systems are generally developed for term neonates, whereas preterm neonates exhibit different signs of withdrawal, and for instance, less severe symptoms of withdrawal may be observed due to neurologic immaturity. As a result, these scoring systems might under- or overpredict the severity of NAS. 20,34 This reflects the urgent need for the development for new and prospectively validated scoring systems, not only related to NAS but to other neonatal diseases as well, especially for pretermneonates who often present different symptoms as opposed to neonates born at term. 35 Antibiotic treatment in neonates is often initiated empirically immediately after the suspicion of sepsis (pathogen not identified yet) and discontinued after approximately three days if the clinical sepsis was not confirmed with a positive bacterial culture. Therefore, new and quick diagnostic tools to identify the pathogen and infection would prevent unnecessary exposure with antibiotics in this population. Thus, rather than specifying dosing regimens based on individual patient characteristics such as weight, PNA or GA, it might be more appropriate to determine dosing decisions on the exact indication and its microbiological epidemiology: the assumptions (PKPD targets, MIC concentrations) informed by the target infection contribute to a large extent to target attainment compared to the individual patient characteristics. Therefore, better in vivo characterization of PKPD parameters, preferably using novel randomized trial designs combined with pharmacometric modelling and simulation, could considerably improve beta-lactam use with respect to efficacy and toxicity in the neonatal population. 36 It is surprising that although modeling and simulation of PK and PD data is frequently reported in literature and also is an integral part of the health authority’s assessment, the results and adjusted dosing recommendations are not frequently implemented in daily clinical practice. 37 However, moving forward we are confident that pharmacometrics will continue to play a key role in pediatric drug development by contributing towards personalized dosing strategies, constructing feasible trial designs, and finally supporting the labeling of drugs in pediatrics. Personalized pharmacotherapy for the pediatric population could be obtained by the initiation of treatment with the optimal dose, obtained from simulations from a (recently developed or literature) PKPD model. If needed, a posterior dosage adaptation could be realized based on the estimation of the individual PK parameters depending on TDM results. In order to determine the right dose for every