Pediatric AIDS–Therapeutic Successes Built on a Foundation of Pediatric Clinical Pharmacology with Pharmacokinetic-Pharmacodynamic Modeling

I am very honored to be receiving this award recognition and offer great thanks for those who supported my activities promoting better use of antiretrovirals in pediatrics through application of clinical pharmacology. I am especially grateful given the award's namesake, Sumner Yaffe, MD, who was a trailblazer in pediatric clinical pharmacology, pushing the now-accepted mantra that “children are not miniature adults.” His persistent effort resulted in recognition and understanding at the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) and the FDA that pediatric clinical pharmacology was essential to the safe and effective use of drugs in children. His persistence led to the establishment of the NICHD Pediatric Pharmacology Research Unit (PPRU) Network in 1994, where I was fortunate to learn from him in my journey on conducting pediatric clinical pharmacology trials to better understand drug effects and to help establish rational dosing and inform pediatric drug labeling. I am also very honored to be recognized by PPA, a critical organization in promoting rational therapeutics in children. In a perfect world, we would be able to easily and directly incorporate science-based medicine into the drug therapy for children. However, in reality there are frequently extenuating circumstances and important differences between clinical situations and clinical trial results that limit our ability to directly apply existing literature to drug treatment for infants and children. Therefore, it is critically important that groups such as PPA exist to provide leadership in bridging these gaps and integrating emerging scientific findings with “real-world” patient and logistic factors, including drug formulation, administration methods, co-morbidities, and concomitant medications, to allow for delivery of optimal pediatric therapy.I consider myself extremely fortunate to have had amazing opportunities early in my career, along with supportive leadership that allowed me to participate in clinical pharmacology research focusing on expanding our understanding of HIV treatment in pediatrics. After completing my fellowship training, one of my early pharmacy responsibilities included performing therapeutic drug monitoring at the University of California–San Diego Medical Center. At this academic teaching institution, I quickly encountered James Connor, MD, a pediatric infectious disease specialist who actively directed the pediatric antiviral assay laboratory. This laboratory performed quantitative drug concentration determinations in biologic fluids for pediatric clinical trials. These results were being sent only to study sponsors as raw concentrations but no pharmacaokinetic (PK) modeling or interpretation were performed or shared with participating investigators. Unfortunately, in that era (pre–Best Pharmaceuticals for Children's Act), study sponsors had little incentive to analyze these pediatric data themselves. So the results would sit in their data warehouses, underanalyzed, making only minimal contributions to pediatric therapeutics. In discussions with Dr Connor, we saw this as a missed opportunity and thought about how we could help create more knowledge from the pediatric PK data being generated from our laboratory. We focused on applying population PK (PopPK) methods to the drug concentration data. This approach was well suited for the sparse PK data available from many pediatric clinical trials. Population PK also has the desirable feature of promoting the unifying of analyses of raw data across studies so that data can be leveraged from multiple small pediatric PK studies to produce more robust PK parameter estimates. Our PopPK activities led to a National Institute of Allergy and Infectious Diseases (NIAID) grant award as a pharmacology laboratory to support the Pediatric AIDS Clinical Trials Group. In our proposal not only did we include drug concentration measurements but we also proposed incorporating PopPK analysis methods into subsequent pediatric HIV trials. We were initially funded in 1992 and have been continuously funded for 30 years. Shortly thereafter we broadened our aim in pediatric clinical pharmacology by submitting a competitive grant proposal for the NICHD PPRU Network and were selected as one of the initial sites. As a PPRU site, we joined the PPRU Network, which included a cadre of expert pediatric clinical pharmacologists with complementary expertise and experience to design and conduct the trials needed to promote the safe use of drugs in infants and children across many diseases. I was able to take the lessons learned from the broad PPRU experience into my pediatric clinical pharmacology studies of HIV therapies.Treatment and prevention of HIV infection is one of the recent great medical success stories; however, it is my belief that it is substantially underappreciated, particularly with regard to pediatric HIV infection. Pediatric AIDS is a disease in which children, particularly infants, do not behave like “miniature adults” with respect to drug PK, pharmacodynamics (PD), and disease progression. The natural course of HIV infection in adults presents as a self-limited viral infection that slowly depletes key immune cells, but opportunistic infections and progression to full-blown AIDS takes a decade after HIV infection to develop. In contrast, newborns infected with HIV in utero who do not receive HIV antiretroviral therapy have an approximate 50% mortality rate during their first few years of life because of HIV ravaging the immature immune system. Once we were able to detect the HIV virus in blood and remove HIV-tainted blood from distribution, pre-adolescent pediatric HIV infection occurred almost exclusively via mother-to-child transmission (MTCT). This can occur in utero in the peripartum period or during breastfeeding. Prior to the availability of HIV antiretrovirals (ARVs), HIV-infected mothers transmitted HIV to their babies about 25% to 40% of the time. The initial perinatal ARV study, Pediatric AIDS Clinical Trials Group (ACTG) 076, studied maternal ZDV administration during pregnancy, at delivery, and then post-partum in infants for 6 weeks. It demonstrated1 a two-thirds reduction in infant infection despite having almost no effect on HIV viremia in the mother. Subsequently, NVP, an ARV with a much longer half-life than ZDV, was shown2 to reduce MTCT with only 1 to 2 infant doses. Combination therapy with newer ARVs used to treat the mother during pregnancy frequently suppresses HIV viremia below the limit of detection (HIV RNA <50 copies/mL).3,4 Prenatal care with effective ARV treatment and maternal virologic suppression can limit in utero and peripartum transmission. Furthermore, the use of infant formula instead of breast milk can prevent HIV transmission through breastfeeding. Maternal virologic suppression combined with infant ARV prophylaxis for the first 4 weeks of life have driven infant HIV infection rates down to near zero in developed countries. In low-resource settings, infant formula is unaffordable and formula-fed babies are deprived of the critical immunologic benefits from breast milk, which is urgently needed in low-resource settings. Thus, not only is infant formula unaffordable, the lack of potable water and limited sanitation infrastructure mean that the use of infant formula leads to increased gastrointestinal infection morbidity. Remaining areas of need to prevent infant HIV infection include universal access to pre-natal care, including HIV testing and treatment, and better approaches to preventing transmission during breast-feeding. So while we have the tools and knowledge to prevent infant HIV infections worldwide, there are still more than 150,000 new pediatric HIV infections annually.5 Where the cost, availability, and risks of formula feeding are high, breastfeeding is recommended for HIV-exposed babies, and it continues to represent the major route of HIV transmission in pediatrics. Thus, as I will discuss later, additional efforts are focused on reducing HIV transmission in breastfed infants. This task is hindered by the limited availability of suitable ARV formulations for neonates. While there are more than 20 individual ARVs and many more ARV combinations with FDA approval for use in adults, less than 10 are approved for use in newborns, and most of these represent older, less potent agents first approved in adults more than 20 years ago.I came into pediatric clinical pharmacology with the belief that PopPK analysis with computer data analysis programs such as NONMEM was a powerful approach to help describe ontogeny and age effects on drug disposition, fostering the design of optimal dosing recommendations. In addition, using limited sampling design PopPK can also leverage existing information across studies. However, to develop rational pediatric therapeutics from these data, we desperately need appropriate formulations combined with an expanded understanding of a drug's age-defined PD. Pediatric dosing regimens can be designed to match target plasma trough drug concentrations or AUC values found to be effective in adults, but allometric principles inform us that one cannot simultaneously find a pediatric dose to match both trough drug concentrations and AUCs from adults forin pediatric patients. If one designs a pediatric dosing regimen to match the adult AUC, then one typically generates a higher peak and lower trough in pediatrics than are seen in adults. On he other hand, pediatric dose regimens designed to match adult troughs result in higher peak concentrations and larger AUCs than are seen in adults. Thus, a clear understanding of the drug's PD and the target exposure metric (e.g., peak, trough, AUC) driving the clinical responses will determine the the optimal pediatric ARV dose – a dose designed to match the AUC in adults is different than a dose designed to match the trough in adults. As we are all well aware, pediatric formulation and food effects need to be considered in the design of optimal orally administered pediatric therapies. Differences in formulations and how a drug is administered can have a huge effect on drug PK and therapeutic success in the pediatric patient. While we are slowly moving away from liquid formulations toward chewable and dispersible tablets for infant and young child ARVs, these non-adult formulations rarely have the same absorption properties as the corresponding original solid oral dosage forms designed for adults.Because many antibiotics and antivirals are primarily renally eliminated, I took great interest in understanding the ontogeny of renal function and its variability in neonates. There were detailed descriptions of renal function maturation when I was starting to study ARVs in infants. However, few were derived in a manner that helped describe the individual components driving variability in renal ARV clearance that could be directly integrated into PopPK models. Most infant PK studies were performed during the first 1 to 2 weeks of life, and data were much more limited beyond this period of life and rarely incorporated markers of renal function. Infant dosing guidelines supported by these studies based their dosing on weight, post-natal age (PNA), gestational age (GA) at birth, or post-menstrual age (PMA), but they failed to provide appropriate guidance on dosing for older term infants and infants with elevated serum creatinine (SCR) and renal dysfunction. This led us to study the factors of vancomycin renal clearance during the first 2 years of life.6 In this population, SCR is initially highly correlated with GA at birth, while SCR continues to be highly correlated with PNA and PMA in older infants. We found SCR was a better predictor of vancomycin clearance than was either GA or PNA. We also studied ampicillin and aminoglycoside renal clearance early in life, looking to characterize both early and later changes in renal clearance.7,8 Using a single exponential PNA function or a sigmoid Emax PMA model we described overall maturation of clearance, though both approaches had some bias in renal clearance in the first few days of life.We are exploring a bi-exponential function to describe maturation of renal clearance, which shows promise in characterizing both early and later changes in renal clearance with less bias. This approach is consistent with the shape of changes in SCR during the first year of life.9The first available ARV for treating HIV infection in adults or children was ZDV. As a single agent, ZDV has limited potency, and therefore a maximum tolerated dose approach was used during its development. This resulted in a relatively high adult dose of 200 mg 5 times a day that often resulted in mild, but tolerable, toxicity. Zidovudine is metabolized by uridine glucuronyl transferase, and thus, determining pediatric dosing required extrapolating from the adult dose while incorporating the maturation pattern of this enzyme system in infants and young children. We incorporated a PopPK strategy into the landmark pediatric study, Pediatric ACTG 152, that evaluated ZDV and didanosine (DDI) for treatment of HIV infection in over 800 pediatric patients.10 Our work characterized significant increases in ZDV metabolism over the first 3 years of life.11 Of note, ZDV was associated with increased toxicity and poor growth in this younger age group (<3 years of age), who had higher ZDV concentrations compared with older children. We found that ZDV exposure (AUC) correlated with the early fall in hemoglobin concentrations. The study also demonstrated important efficacy differences between ZDV and DDI that were contrary to results obtained from adult studies. In adults, ZDV was found to be clinically superior to DDI. However in the Pediatric ACTG 152 trial, children did not behave like “miniature adults,” as DDI demonstrated superior efficacy compared with ZDV, in large part because of the higher ZDV toxicity seen in infants.10Preterm birth represents another important challenge to optimal ARV for children. birth is associated with increased of After ZDV used for preventing we found plasma elevated ZDV concentrations in babies higher than had been in term This was because uridine glucuronyl is with develop dosing guidance in infants for the first 6 weeks of life we a study of ZDV that included a therapeutic drug monitoring We of ZDV concentrations to and infant our PopPK we found that ZDV clearance was initially lower in infants than had been in term infants during the first of We also found significant differences in ZDV clearance based on GA at those between and weeks GA had lower and of ZDV clearance over than did those at 30 to weeks GA based on the PopPK model were used to develop and dosing for infants that were incorporated in the of Health and Human HIV treatment has also been to prevent MTCT and has as or 2 infant has been shown2 to prevent peripartum transmission. of to infants can reduce HIV transmission to breastfed infants with less than a of the dose needed to treat infant HIV PK, and dosing information in term and infants is needed to develop We 2 to these knowledge is metabolized by and its This PK from PK studies very and the of rational dosing that are safe and can be easily in We found that infants between and weeks GA required a dose reduction to concentrations to term the lack of ARV PK and data in infants we Pediatric and AIDS Clinical Trials This study was designed to determine the and PK of ARVs and in birth infants. It took an opportunistic approach to ARV and PK and information from infants receiving these drugs as part of their clinical care over the first 6 of Population PK modeling of data from this study and is being performed to help develop optimal dosing in birth infants that include dose increases to for increased metabolism from and are ARVs that do not have infant formulations and cannot be studied by administration to infants. In this we are PK from infants to these mothers to determine the initial infant clearance to help design PK studies infant formulations do We have studied both and PK to characterize PK on the first days of life, when drug metabolism would be considered to be at its of We found a in metabolism after birth, which required 2 dose increases during the first few weeks of life to therapeutic with ARV dosing in infants is how to and consistent absorption for newer ARVs (e.g., during the first 6 of of these agents are making their absorption on the gastrointestinal and by food is metabolized by in poor and a very these PK has been with which and the drug's half-life to allow dosing of the In the first 2 weeks of life has been associated with and renal This in part be to the high and of the liquid pediatric of the for these the FDA a use in the first 2 weeks of life and below weeks older infants, but those still less than 6 of larger than do older infants and other to better understand some of these we a PopPK analysis of across infant studies that included infants and found that absorption in young infants is the of this larger dose we have evaluated use in term and infants after therapy to determine how to dose in this This is a critical age for HIV treatment, as ARV are limited and of viral can have immunologic The was evaluated in infants birth who were between 1 and 2 of with sparse PK sampling for 6 The treatment was safe and well including in the infants who weeks variability in absorption was with a corresponding lower AUC at first PK with absorption with In the first of life absorption of can be a huge to optimal HIV therapy in term infants. In the study using a formulation of the AUC was well below the study target of 20 to and the adult of to The results from these studies that when we are drugs we consider not only the maturation of metabolism but also age effects on because of changes in metabolism and drug because exposure is driven by the clearance not ARV drug therapy a of at 3 different combinations have been developed to reduce and While these have been to in adults using and their drug dose there are additional in pediatrics. In young infants, the rates of maturation are different for different So the drug of the drug based on single drug studies in newborns not be the same as the in older pediatric patients. In addition, the at which dose changes are needed also and of these changes in low-resource the use of some of these I have through the Health Pediatric Group in the of dosing of ARVs in infants and children These dose are designed to help pediatric dosage forms and provide target for to also dose increases at the same for all ARVs than for agent, care to this dose to determine the available to dosing and some differences from the dose for agent, but the of and consistent dosing in my this I have focused on and as they to ARV these factors are not the only of ARV PK variability in pediatrics. was the drug of in adults because of its potency, and ability to be with and into a single While was also used in older children and in a liquid formulation its use in children 3 years of In where ARV are limited there is a need for many ARV has limited with drug therapy and can be used in children. is primarily metabolized by and while significant differences have been in adults and children based on their the drug is FDA approved at a of We performed an study of in HIV-infected children between 3 and of age with and concomitant infection. was administered as and blood were obtained as blood and with the blood used to concentrations in with dose where on limited PK data Pediatric Population of – in to at dosing was developed using with large infant used to in this dose of mg was used for infants, which represents two-thirds of the adult the was for a but concentrations in poor required the study to be to to dosing was with poor receiving 25% of the dose used in and were with all subsequent poor infants concentrations in the target that dosing are essential for poor In contrast, infants receiving the with do not need additional dose as the in is Thus, a dosing strategy using the same of is HIV have been studied as a approach to prevent and treat HIV infection. These have the of and in newborns and of after a single they have no for and work by viral that are not ARV they minimal and no with existing of are not metabolized by are they renally but through the system like other their metabolism can be from by to the the of the of can their to and reduce metabolism by more than We evaluated the first in newborns after of the adult study and the down through older pediatric age We combined the PK results from newborns with PK data from adult studies to a PopPK model of that was then used for this analysis we developed a dosing strategy for infants that would high concentrations when used to treat HIV infection treatment with early during infection limit the of the viral and promote including These effects reduction in ARV that could include a in a of infants in which the immunologic HIV 2 in the can be and a half-life increased of has been developed and studied in infants. a much in serum concentrations than in adults, but this in infants had a on the This is most to the growth in infants having the same effect with both and However, PK still the that a single dose given after birth and another dose at 3 of age would concentrations for 6 could be given at for babies who to HIV via breastfeeding beyond 6 We have evaluated in with another in young children infected with HIV from birth but with The early and PK results of this are multiple can be used to the of antiviral have in our understanding of HIV infection and the clinical pharmacology of ARVs to prevention and treatment in pediatric patients. of the to of MTCT and pediatric HIV to care delivery system and other pediatric clinical pharmacology studies of HIV are to and early antiviral which can result in benefits in HIV-infected infants.