It's not easy being small

I t is well known that physiologic differences between children and adults can affect drug disposition. Developmental differences in drug absorption, plasma protein or tissue binding, functional maturation of excretory organs, metabolic enzyme expression, and distribution of drug in the various tissues of the body can result in differences in systemic drug exposure for children compared with adults treated with the same dose [1–3]. The most dramatic changes in excretory organ function and body composition occur during the first few days to months of life (Fig. 1). Thus, from a pharmacotherapeutic perspective, infants present a multi-dimensional challenge to treating clinicians. If asked, pediatric oncologists would detail the meticulous manner in which we prescribe anti-cancer medications, systematically measuring the child’s height and weight, calculating the corresponding body surface area (BSA), and individualizing the dose of drug for each child. It should thus not be surprising that as oncologists we would be rightfully concerned about undertreatment if a 12-month-old infant received a daunomycin dose of 10 mg/m in place of a 30 mg/m dose. Moreover, we would prepare for life-threatening toxicity if that same infant inadvertently received 150 mg/m of anthracycline. Although such dosing errors are thankfully rare, the actual drug exposures realized by children with cancer, and especially infants, often span a remarkably wide range. In this issue of Pediatric Blood & Cancer, Hempel et al. report the results of a pharmacokinetic study of 21 infants treated with daunomycin for their leukemia. Using a population based approach, the investigators found that daunomycin drug clearance varied over an almost 15-fold range. For example, following a 20 mg/m dose of daunomycin, one 3-month-old infant’s drug exposure, as measured by the area under the concentration–time curve (AUC), was approximately 2,020mg hr/L, whereas following a higher 30 mg/m dose, a 10-month-old infant’s AUC was only 190mg hr/L. Overall, however, the authors did not find that infants cleared daunomycin slower than older children, raising the question as to whether the empiric dose reduction routinely applied for infants is warranted. Important caveats to this finding include (i) that even though a population based approach was taken, the infant population studied was small; (ii) there was an unexplained high residual error in the population model; and (iii) clinically, infants less than 6 months were more likely to experience an infection than older infants. How then can pharmacokinetic data lead us to improved therapeutics for infants with cancer? The ideal situation is one in which there is a well-defined relationship between drug exposure (pharmacokinetics) and drug effect (pharmacodynamics). Over the past several decades such relationships have been defined for a number of therapeutic classes of drugs, including cardiovascular agents (e.g., digoxin monitoring was developed more than 30 years ago [4]), anti-epileptic drugs, anti-infective agents, and psychotropic drugs. Yet relative to other subspecialties, and despite utilizing the most toxic drugs administered to patients today, pediatric oncology is essentially still in the Dark Ages when it comes to understanding PK–PD relationships. We can count on one hand the number of drugs that have adequately defined a toxic exposure threshold. Moreover, with rare exception, we simply do not know the therapeutic concentrations for anti-cancer drugs. Dosing guidance for most anti-cancer drugs have been empirically derived based on toxicity. Individualization of dose, should it occur, is almost invariably based on subsequent tolerability following initial empiric dosing. To account for the changes that occur during the early months and years of life, broad, empiric modifications based on body weight, body size, or age are applied. Selection of infant doses varies widely, as can be evidenced by current dosing guidance for doxorubicin (Fig. 2A) and daunorubicin (Fig. 2B). Depending on a child’s age or weight, the anthracycline dosing recommendations vary by as much as fourfold. The paucity of pharmacokinetic data in the infant population underlies these highly variable, empiric recommendations. We have relatively limited knowledge on how such reductions affect toxicity, and virtually no understanding of their impact on survival. A multi-step approach should be undertaken to begin to address the problems surrounding dosing guidance for infants and young children. The first is to move away from the ‘‘quantum’’ dosing guidance currently used in which an infant’s drug dose can double from one day to the next based on an arbitrary age or weight cut-off (Fig. 2). One potential approach would be to use an allometric scaling [5–7] of drug dose (e.g., scaled to body weight, kg), an approach that can, in part, be derived from existing data. Further, additional well-designed, prospective population PK–PD studies that can help us understand, and ultimately reduce, the wide degree of interpatient variation in drug exposure that occurs in infants should be conducted. It’s not easy being small, but successfully addressing the pharmacotherapeutic challenges posed by infants is well within our reach.