Paediatric extrapolation: the panacea for paediatric drug development?

Evidence‐based medicine is required for adults and nothing less so for the dynamic, growing, and developmentally changing paediatric and neonatal populations. However, most drugs receive regulatory approval for adolescents, children, infants, and neonates many years after these drugs are already available for use in adults. This has not only resulted in a large amount of off‐label use of these drugs in neonatal and paediatric patients, but it also has significantly slowed down clinical investigations of medicines that are necessary to obtain paediatric and neonatal data of sufficient quality 1.

In their review for BJCP 2, Ollivier et al. postulate that “the use of paediatric extrapolation is not only a strategy to increase the efficiency of medicines development; it is part of a new paradigm in paediatric medicine development and an ethical imperative.” Clearly, this is a very strong and very general statement, and the question arises if there is enough evidence‐based information to support this statement at this very moment.

Especially the “ethical imperative” part of this statement is difficult to defend at this stage and as such too premature to be part of this overall postulation. Furthermore, it needs to be emphasized that it is extrapolation of efficacy and the review does not sufficiently address other very relevant issues such as determination of drug exposure as well as assessment of paediatric safety data. All these areas need to be taken into consideration.

There are also differences between various therapeutic areas and paediatric age groups. Evidently, extrapolation of efficacy from adults to adolescents is easier as compared to extrapolating data from adults to younger children or even from adults to infants and neonates. Adolescents and children might be small adults, but young infants and neonates are for sure not little children! This is also true for dosing, where the prediction of neonatal dosing is poor without information from infants 3.

Extrapolation of efficacy plays a major role in paediatric drug development, subsequent marketing approval and labelling to reduce off‐label medicine use in children. Modelling and simulation allows one to utilize data from other sources, including adults and other paediatric populations, to model/simulate data responses in a specific paediatric population. These approaches have proven useful in dose finding and defining the degree of the similarity in exposure‐response relationship between adult and paediatric patients.

Furthermore, extrapolations of population pharmacokinetic (PopPK) covariate models between drugs sharing an elimination pathway have enabled accelerated development of paediatric models and dosing recommendations 4.

Calvier et al. tried to identify the conditions for which this kind of approach will consistently result in accurate, pathway specific, plasma clearance scaling from adults to children for drugs undergoing hepatic metabolism. A physiologically based pharmacokinetic (PBPK) simulation workflow utilizing mechanistic equations defining hepatic metabolism was developed. Based on this research, it became clear that drugs eliminated via the same pathway require similar paediatric dose adjustments only in specific cases, depending on drugs extraction ratio, unbound fraction, type of binding plasma protein and the fraction metabolized by the isoenzyme pathway for which drug plasma clearance is scaled. In conclusion, between‐drug extrapolation of paediatric covariate functions for drug plasma clearance is mostly applicable to low and intermediate extraction ratio drugs eliminated by one isoenzyme and binding to human serum albumin in children older than 1 month 5.

However, there are several issues that deserve additional attention. Paediatric extrapolation carries inherent risks if any of the assumptions used during extrapolation turn out to be incorrect. Given the variability in individual patient responses, passive adverse event reporting and other factors, erroneous assumptions may not be detected and reported in the clinical setting for years, if at all. The potential impact and the severity of the consequences on paediatric and neonatal patients would clearly depend upon the nature of the erroneous assumptions and the intended use of the drug 6.

It is also important to realize that paediatric extrapolation likely has a publication bias—negative studies where extrapolation has failed are not reported. Therefore, journals need to encourage publication of failed paediatric extrapolations so that pharmacometricians can learn from them and will not make the same mistakes.

Furthermore, a very important assumption made in paediatric extrapolation is constant oral bioavailability from adults to children. The allometric scaling equation and maturation function were designed to work for scaling total systemic clearance assuming absorption is a constant in adults and children. This may or may not be the case for every drug. The Biopharmaceutics Classification System Class can change from adults to children due to differences in gastric volume, and it seems likely that bioavailability may change as well. This factor should be at least considered in any paediatric extrapolation.

Study‐ and population‐related differences seem the most likely cause to explain the difference between the observed and simulated data. Two of these differences, patients versus healthy volunteers and differences in meal composition, were of sufficient magnitude to suggest that these differences seen in published accounts could explain the discrepancy.

So it will be important to 1) careful examination of differences in pharmacokinetics between adults and older children and infants/neonates; 2) differences in drug administration and possible changes in BCS class between adults and children and 3) examination that oral bioavailability is a constant between adults and children 7.

There are also clear differences between various therapeutic areas as illustrated here. The Pediatric Epilepsy Academic Consortium for Extrapolation (PEACE) addressed the following two assumptions in focal‐onset seizures, the most common seizure type in both adults and children: 1) a similar disease progression and response to interventions in adults and children and 2) similar exposure response in adults and children. Based on their extensive review, the consortium concluded that extrapolation of efficacy data in adults to paediatrics in focal‐onset seizures is supported by strong scientific and clinical evidence. However, safety and pharmacokinetic data cannot be extrapolated from adults to children. When done in conjunction with pharmacokinetic and safety investigations in children, extrapolation of adult efficacy data from adults to children can reduce the time delay between approval of effective and safe antiepileptic drugs in adults and approval in children 8.

Finally, to add one more aspect of relevance to this important area of extrapolation, the selection of a primary efficacy endpoint is vital to a well‐designed study and serves as the basis for robust evaluation of the clinical impact of the therapeutic intervention. The use of appropriate endpoints that are well defined, reliable, interpretable and directly applicable to the disease or condition being studied is critical. Given its influence on trial outcomes, the selection of efficacy endpoints in paediatric trials should involve a rigorous process that is dependent on scientific and practical considerations.

Green et al. did a systematic review and found that two attributes of endpoints were identified as being different between those trials that met their primary endpoint and were successful in obtaining paediatric labelling for the indication pursued compared to those that did not. Those successful attributes included 1) using an endpoint in a paediatric trial that was the same as the endpoint measure in the corresponding adult trial and 2) having paediatric and adult patients enrolled in the same trial. Of particular interest was the observation that incorporating unique endpoints into paediatric trials that differed from those measured in adult trials represented a risk factor for trial failure.

Understanding the paediatric disease process and understanding the endpoint, as well as careful consideration of trial designs, are all integral components of obtaining success in paediatric drug development 9.

In summary, modelling and simulation is a very useful and important tool that helps inform decision making, but it is “new science” or new understanding of the pathophysiology or expression of the disease in children that will drive change in extrapolation activities. Currently, studies are already separating paediatric drug development programmes into full extrapolation of efficacy and no extrapolation of efficacy based on recent data 1. As a consequence, we are starting to understand the paediatric disease process in comparison to the adult process. As a “new science,” we should let this play out and not rush the process.

Most importantly, the accurate determination of a safe and effective dose of a drug being prescribed to a neonate, infant, child or adolescent is dependent on understanding the pharmacokinetics and pharmacodynamics of that particular drug, as well as the clinical characteristics of the unique neonatal or paediatric patient being treated with this drug. The discipline of developmental pharmacology aims at understanding the impact of maturation on drug disposition and action in the neonatal and paediatric population. It encompasses 1) developmental pharmacokinetics, 2) developmental pharmacodynamics and 3) the ways by which human growth and development will change the pharmacokinetic/pharmacodynamic relationship of a drug in each unique neonatal or paediatric patient. Despite the many advances in technology, study design and pharmacometric analysis, our understanding of the exact influence of age and disease on drug disposition and effect remains challenging 10.

Competing Interests

There are no competing interests to declare.

van den Anker J. (2019) Paediatric extrapolation: the panacea for paediatric drug development?, Br J Clin Pharmacol. 85, 672–674, 10.1111/bcp.13836.