|Home | About | Journals | Submit | Contact Us | Français|
Development of optimized pediatric formulations for oral administration can be challenging, time consuming, and financially intensive process. Since its inception, the biopharmaceutical classification system (BCS) has facilitated the development of oral drug formulations destined for adults. At least theoretically, the BCS principles are applied also to pediatrics. A comprehensive age-appropriate BCS has not been fully developed. The objective of this work was to provisionally classify oral drugs listed on the latest World Health Organization’s Essential Medicines List for Children into an age-appropriate BCS. A total of 38 orally administered drugs were included in this classification. Dose numbers were calculated using age-appropriate initial gastric volume for neonates, 6-month-old infants, and children aging 1 year through adulthood. Using age-appropriate initial gastric volume and British National Formulary age-specific dosing recommendations in the calculation of dose numbers, the solubility classes shifted from low to high in pediatric subpopulations of 12 years and older for amoxicillin, 5 years, 12 years and older for cephalexin, 9 years and older for chloramphenicol, 3–4 years, 9–11 and 15 years and older for diazepam, 18 years and older (adult) for doxycycline and erythromycin, 8 years and older for phenobarbital, 10 years and older for prednisolone, and 15 years and older for trimethoprim. Pediatric biopharmaceutics are not fully understood where several knowledge gaps have been recently emphasized. The current biowaiver criteria are not suitable for safe application in all pediatric populations.
The online version of this article (doi:10.1208/s12248-016-9885-2) contains supplementary material, which is available to authorized users.
Recently, there has been a growing emphasis on enhancing pediatric accessibility to quality essential drugs which subsequently has raised the number of drug formulations tested in and labeled for use in pediatric populations (1). Accordingly, health regulatory authorities in the USA and Europe have incentivized the pharmaceutical industry to devote more resources and research toward developing optimized pediatric formulations. Such incentives include 6 months of added exclusivity and additional support from public funding agencies for pediatric drug discovery and clinical testing (2–4). Unfortunately and despite these incentives, children have remained largely “therapeutic orphans” (2, 5).
Development of optimized pediatric formulations for oral administration can be challenging, time consuming, and financially intensive process (1). Furthermore, generation of detailed biopharmaceutic data in pediatric population is often hampered by ethical, technical, and logistic constraints that limit recruitment and inclusion of children in clinical trials (2, 6, 7). Due to these hurdles, and despite incentives, pediatric biopharmaceutics has remained a recognized under-researched area with several accepted knowledge gaps where additional research is highly required (2, 8, 9). In practice, pediatric formulations are often derived from adult formulations after necessary extrapolation and modeling (10–12). Following the guidance documents published by the Food and Drug Administration (FDA) and the International Conference on Harmonization (ICH), extrapolation from adult data into pediatric population can be justified when the course of disease and response to therapeutic intervention are sufficiently similar in adults and children (10–12). However, while extrapolation can be done with respect to efficacy, there remain challenges with regard to predicting safety or the need for dose adjustments (1).
Using principles founded in the biopharmaceutical classification system (BCS), initial predictions of the intestinal absorption characteristics of an active pharmaceutical ingredient (API) are frequently based upon its solubility in the gastrointestinal (GI) milieu and its permeability across the enterocyte membrane (13–16). The BCS categorizes drug molecules into four classes based on their aqueous solubility and intestinal permeability. Today, the BCS is widely used in support of waivers (biowaivers) of in vivo bioequivalence (BE) studies for immediate release (IR) solid oral formulations containing BCS class I (high solubility and high permeability) and, more recently, class III (high solubility low permeability) drugs (17, 18). Health regulatory authorities allow the use of in vitro dissolution testing as a surrogate for conducting in vivo BE trials for highly soluble compounds. Accordingly, biowaivers reduce regulatory burden, processing, and approval time, thus providing economic benefits while maintaining high drug quality standards for therapeutic equivalence (19).
Although the BCS is implicitly based upon the GI physiology of adults, it is uncertain how such classifications may translate to the same drugs when administered in pediatric patients. Theoretically, the basic concepts should be applicable irrespective of patient age (2). However, fundamental differences exist between adult and pediatric populations in terms of developmental changes from birth to adolescence which contribute to age-associated variation in drug absorption, distribution, metabolism, and elimination of drugs (9). The initial gastric volume (V 0) is one of these clear differences. According to regulatory guidelines, a drug is classified as highly soluble when the maximal dose strength is soluble in 250 mL of aqueous media whose pH values range from 1.0 to 6.8 (17) or 1.0–7.5 (20). The volume of 250 mL was derived from a typical BE study protocol in which oral drugs are administered to adult volunteers with a glass of water (about 8 oz). However, BE standards in pediatric populations have not been established and recommendations are still needed (2).
Solubility of drugs in adult GI milieu is expected to be different from those in pediatrics due to differences in GI fluid composition and volume occurring over time with age and development. Hence, children are more of a heterogeneous population and solubilities of drugs are expected to differ with age and development status of the child. Therefore, age-appropriate pediatric BCS (PBCS) criteria are needed to properly classify traditional and novel therapeutic agents in pediatric populations (2). However, classification of traditional drugs into BCS classes using age-appropriate gastric volumes across pediatric age populations was not previously performed.
Since the introduction of BCS, there has been growing interest in provisional BCS classifications of drug lists and databases like the top 200 drugs in different countries and the World Health Organization’s (WHO) Essential Medicines List (EML) for adults (21–25). These investigations used molecular properties like the n-octanol/water partition coefficient (LogP), the pH-dependent partition coefficient (LogD) at certain pH points, and the polar surface area (PSA) as indicators for intestinal permeability to judge if molecules can be assigned a high- or a low-permeability BCS class. The PBCS Working Group of the National Institute of Child Health and Human Development provisionally classified 56 commonly used pediatric drugs based on their aqueous solubility and predicted n-octanol/water partition coefficient (cLogP) values as indicators of their intestinal permeability (9). Gandhi et al extended the work of the PBCS Working Group and attempted to classify five drugs commonly used in pediatric populations into PBCS applying pediatric biopharmaceutical considerations (8). Recently, Charoo et al assessed the risks of extending the BCS-based biowaiver decision on IR oral dosage forms of fluconazole in adults to pediatric population and concluded that it can be safely granted a biowaiver provided using safe excipients (26). Previous studies neither included the drugs listed on the WHO’s EML for children nor classified drug molecules into age-appropriate PBCS for all children groups capable of swallowing solid oral medications (8, 9, 26).
The present study emphasizes the care that must be taken with assigning BCS solubility classes in pediatric populations. Considerable heterogeneity is obvious among pediatric populations as the V 0 varies with age and body size. Because the solubility class is assigned based on the solubility of the drug in a volume equivalent to the V 0, misjudgments regarding the solubility class can arise when considering pediatric populations as a homogenous group and not using age-appropriate V 0 values.
This study aims to investigate the effects of variable age based V 0 on the BCS solubility class of oral drugs particularly available on the WHO’s EML for children. Oral drugs available on the WHO’s EML for children are also provisionally classified into PBCS on the basis of their dose number (D 0) values in pediatric populations. The results of this study extend those started by the PBCS Working Group of the National Institute of Child Health and Human Development.
The latest EML for Children (fourth list, 2013) was used in this study. The list is available at the WHO’s website (http://www.who.int). Only drugs available in solid oral dosage forms were selected for this provisional classification.
Although a compound’s permeability classification can be influenced by paracellular transport or by transporter activity, for the purpose of this evaluation, drug permeabiity solely from the perspective of transmembrane transport was considered. Accordingly, permeability was determined on the basis of the n-octanol/water partition coefficient as calculated (cLogP) using ALOGPS (The Virtual Computational Chemistry Laboratory, VCCLAB, Germany) software package. The experimentally determined n-octanol/water partition coefficient (LogP) values were obtained from (27). Correlation of human jejunal permeability with partition coefficients and the use of benchmark reference drug molecule were carried out as previously described (9, 21, 23). Detailed correlations are shown in the supplementary materials.
According to the FDA, EMA, and WHO guidelines, drugs are administered to adult healthy volunteers with a glass of water that has a volume of 250 mL (17, 18, 28). Therefore, a drug would be considered a high-solubility class if the highest dose strength was soluble in 250 mL or less of water over the pH range specified by the regulatory guidelines at 37±1°C. The FDA, EMA, and WHO used to have different pH ranges, but in the recent guidance draft, the FDA revised the pH range and proposed a range of 1–6.8 which is similar to the range used in the EMA and the WHO guidelines (17, 18, 28). A V 0 of 250 mL was used in calculating the D 0 of drugs to account for their aqueous solubility in adults.
In the US, children as young as 4 years old could be taught to ingest oral tablets (29). However and although this might vary depending on the child, generally, the age at which children can safely take tablet and capsule dosage forms is around 6 years (30, 31). Therefore, children of 6 years old and above might be prescribed oral solid dosage forms and younger children might be prescribed oral suspensions. In this study, age-based V 0 calculations of D 0 values were made for neonates, 6-month-old infants, and infants and children 1 year old and above.
Two fundamental parameters are key factors in determining the solubility of a compound in a medium; these two parameters are composition and volume. In general, it is widely accepted that the gastric pH in children above 2 years of age is similar to that in adults (7, 32). The age-related gastric volumes in children were not intensively investigated. The fasted gastric volume in children was reported to range from 0.40 to 0.56 mL/kg (33–35). This volume is equivalent to 28.0–37.1 mL/kg in a 70-kg fasted male (9). The fasted gastric volume was typically reported as 40 mL in a 70-kg fasted male (36).
To extrapolate to children, median weights of neonates, infants 6 months old, infants and children aging from 1 to 17 years were obtained from the Centers for Disease Control and Prevention growth charts (37). Extrapolating the 250-mL V 0 value used in adults to pediatric subpopulation based on the 0.56 mL/kg value was done according to Eq. 1 (9, 33–36). The median weight in different age groups, gastric volumes, and normalized V 0 are shown in Table TableII
A drug would be considered a high-solubility class if the highest dose strength was soluble in a volume of water equivalent to or less than the normalized V 0 in children over the specified pH range at 37±1°C.
The maximum dose strengths (in milligrams) for all drugs available in solid oral dosage forms were obtained from the WHO’s EML for Children. The maximal per oral dose for each pediatric subpopulation was taken from the British National Formulary for Children (38). The British National Formulary for Children expressed dosing recommendations for many drugs in terms of mg/kg and for some drugs in μg/kg. However, in many cases, a maximal dose in milligrams or micrograms was specified for different pediatric subpopulations. When specified, the maximal dose for each subpopulation was used; otherwise, dosing recommendations in terms of mg/kg or μg/kg and median weights of neonates, infants 6 months old, infants and children aging from 1 to 17 years obtained from the Centers for Disease Control and Prevention growth charts (37) were used to derive age-specific doses. The age-specific maximal dose strength (M 0) was used to calculate D 0 as shown in Eq. 2.
Experimental aqueous solubility data in (mg/mL) were obtained from Yalkowsky & He (39) and DrugBank databases (27); data were originally from (40) in the former database. The most conservative data measures were used. Data were preferred when the pH and temperature at which the aqueous solubility was measured were mentioned.
where D 0 is the dose number, M 0 is the maximum dose strength, C s is the saturated solubility, and V 0 is the initial gastric volume. If a medication was available in different strengths, then the highest strength was used to calculate the D 0.
When the experimentally determined LogP values were plotted against software calculated (cLogP) values, the plot showed a high linear correlation (R 2=0.92) for 35 drugs. The linear correlation plot and the correlation of human jejunal permeability values with partition coefficients are shown in the supplementary materials. Permeability classification based on cLogP of metoprolol as a benchmark for high/low permeability showed that 20 (52.6%) of the 38 drugs were assigned to high permeability class, whereas the rest of 18 (47.4%) of the 38 drugs were assigned to low permeability class. Classification of the 38 drugs into high and low permeability classes is shown in the supplementary materials.
Drugs with D 0≤1 were classified as high-solubility-class drugs and, conversely, drugs with D 0>1 were classified as low-solubility-class drugs.
Calculated D 0 values showed that 29 (76.3%) of the 38 drugs would be assigned to either high- or low- solubility class and did not change the assigned solubility class when calculated using age-appropriate D 0. These drugs are presented in Table TableIIII.
The following compounds are shown to be affected by age-associated changes in M 0 and gastric volume: amoxicillin, cephalexin, chloramphenicol, diazepam, doxycycline, erythromycin, phenobarbital, prednisolone, and trimethoprim. For each of these compounds, their age-specific solubility classes, based on their calculated D 0 values, are provided in Table TableII.II. In some cases, as age (and therefore V 0) increased, the drug went from being classified as low solubility (youngest group) to high solubility as the child matures (Table (TableIIII).
The orally administered drugs listed on the WHO’s EML for children were provisionally classified into BCS and PBCS classes on the basis of their D 0 values which were calculated based on various V 0 and their cLogP values. Table III displays provisional classifications of the 38 drugs based on V 0 in pediatrics and adults.
In this study, solubility data and a commonly used simple molecular descriptor (cLogP) were used to provisionally classify drugs available in IR oral dosage forms on the WHO’s EML for children. The majority of the drugs classified in this provisional classification were anti-infective and anthelmintic agents (Table (TableII).II). This might reflect their urgent need in maintaining healthcare delivery and might reflect treatment priorities in primary healthcare settings.
Data pertaining to pediatric biopharmaceutics are still scarce. Despite the lack of complete understanding, it is generally accepted that intestinal permeability of drugs in children ages 2 years or older are equivalent to that observed in adults (7, 9). Therefore, in this work, adult standards of intestinal permeability were used to designate the drugs as being high or low permeability. However, it is important to note that greater paracellular permeability is expected in younger pediatric subpopulations. Although this could be a limitation of this study, similar standards were used in the provisional classification of the PBCS Working Group of the National Institute of Child Health and Human Development (9). Software-predicted cLogP values were used to assign a high or low permeability class to each drug with metoprolol serving as the benchmark for differentiating between high/low intestinal permeability. This reference drug was consistently used in previous studies (9, 21, 23). Partition coefficients are easily accessible molecular descriptors that can be calculated using validated software packages (9, 21, 23–25). The use of this benchmark was validated by correlating human Peff data of a set of 28 drugs including 14 drugs which were recommended by the FDA as reference compounds for permeability (20, 21, 23). In this study, correlation between the software calculated cLogP to the experimentally determined LogP values are provided in the supplementary materials. The high level of correlation observed between observed and predicted values suggest that software-predicted cLogP values can be applied with some confidence in the prediction of intestinal permeability.
Drug solubility within the GI milieu is a key factor determining the bioavailability of a drug molecule. In the present study, drugs were considered highly soluble when the calculated D 0 was ≤1 using various V 0 adopted for different pediatric subpopulations as well as for adults. Although the findings of this study showed that this shift did not affect the solubility classification for the majority of drugs, interestingly, the solubility class showed a shift for amoxicillin, cefalexin, chloramphenicol, doxycycline, diazepam, erythromycin, phenobarbital, prednisolone, and trimethoprim (Table (TableII).II). The effect of the V 0 used in the calculation of D 0 on the assignment of the solubility class was highlighted previously (9, 42). It was suggested that in case the administered dose is also reduced in certain populations, the overall effect of the reduced V 0 on the D 0 and subsequently the BCS solubility class might remain minimal (9, 42). However, in this study, the maximal dose strength given to each pediatric subpopulation along with the corresponding V 0 were used for calculation of age-appropriate D 0. It is important to note that in this classification, experimental solubility data recorded at ambient temperature were used and, as the solubility of a compound is expected to increase at 37°C, the solubility classification used in this study could be conservative.
In the GI milieu, solubility of a drug molecule depends on the volume as well as the composition of the GI luminal fluid. The composition of the GI luminal fluid shows marked compositional changes with development (43). In a recent study, Maharaj et al assessed the solubility of seven BCS II drugs using biorelevant media reflective of age-specific pediatric populations (43). Using the ratio of solubility in pediatrics to solubility in adults as an indicator, six of the seven drugs studies fell outside the BE criterion of 0.8–1.25 in at least one of the developed age-specific pediatric media. Extrapolating what was seen in Maharaj et al’s study, it is highly probable that the solubility of some drugs would show discrepancies in some pediatric subpopulations. Taking a conservative approach in using the BCS-based biowaiver principles, more risk is considered when the solubility class of a drug shifts from high to low than vice versa. To account for possible discrepancies in solubility of drugs in pediatric subpopulations, percentages by which experimental solubility used in this study might change to shift the solubility class from high to low were calculated. For some drugs, percentages were as low as 1% in some pediatric subpopulations, especially neonates, and as high as 99.8% in older pediatric subpopulations. These percentages are shown in the supplementary materials.
Using adult standards for solubility, 18 (47%) drugs used in this study belonged to the high-solubility class (BCS I and III) in at least one population (Table III). As expected, more drugs belonged to the low-solubility BCS classes (BCS II and IV) when the V 0 decreased. This could have implications on the performance of the formulations in the GI environment.
Globally, age-appropriate formulations are urgently needed as, according to some estimates, off-label use of drugs in children ranges from 60 to 90% (1, 44, 45). These estimates suggest that in practice, drug therapy in children is still largely empirical (1). Probably, there is no single formulation that is ideal for pediatric population of all ages. Considering all age groups from birth through adulthood, the magnitude of doses may vary 100-folds, and theoretically, a range of different strengths of a dosage form should be available to permit simple, accurate and safe administration of drugs in pediatric populations (46). Development of age-appropriate formulations is highly desirable to include all age groups (47). It was suggested that a PBCS should at least consider six age groups which are (a) neonates (≤40 weeks post-conception), (b) infants (0–6 months), (c) infants (6–12 months), (d) toddlers (1–3 years), (e) children (4–6 years), (f) children (7–12 years), and (g) adolescents (13–18 years) (2, 48). However, more categories were said to be needed which could be physiologically meaningful for a more comprehensive evaluation (48). For more comprehensive classification, drugs in this study were classified in neonates, 6-month-old infants, and age–year-wise from the age of 1 year to adulthood (Tables II and III).
Focus on orally administered solid oral dosage forms is important as recent findings suggested that there was no significant difference in refusal rates when children were to receive liquid or solid formulations (49). Nahirya-Ntege et al showed that children as well as their providers preferred oral tablets over syrups when administering anti-retroviral drugs (50). Furthermore, palatability of oral liquid dosage forms could be an issue affecting refusal rate and pediatric patient compliance (1). In practice, oral drugs are often diluted in beverages to enhance palatability and patient acceptance in pediatrics. This practice could compromise their efficacy. Moreover, long-term storage, storage, and transport under extreme temperatures can be issues compromising the efficacy of liquid oral dosage forms, especially in developing countries (51). Therefore, flexible solid oral dosage forms provide advantages over liquid oral dosage forms in these circumstances. Development of drugs for pediatric populations is obligatory in the USA and Europe. The US Pediatric Research Equity Act and the European Pediatric Investigation Plan are required by the FDA and EMA, respectively (52–54). BE of the pediatric formulation to that in adults is recommended but not required.
Findings of this study could be important in managing formulation bridging. In practice, it is common to undertake a relative bioavailability study, or it might be surrogated by an in vitro dissolution testing in healthy adult volunteers to ensure that the pharmacokinetic profile is equivalent for two formulations (7). Subsequent extrapolation to pediatrics is followed by a dose determination/confirmation study (7). In BCS-based biowaiver, bridging can be performed without conducting in vivo pharmacokinetic studies, especially for IR formulations containing BCS I and more recently BCS III drugs (7, 17, 18). The US Pediatric Formulation Initiative (PFI) workshop has emphasized the use of BCS in children (55). The PFI has identified several knowledge gaps that need research including (a) drug metabolism and transport in the GI tract, (b) age-relevant changes in the physiology of the GI tract, (c) changes in permeability and fraction absorbed in children, and (d) developing and validating pediatric physiologic and population-based pharmacokinetic databases (2). In addition to the difference in the V 0 between adults and pediatrics, key factors like gastric and intestinal pH, gastric emptying, intestinal residence time, volume of gastric fluids, immaturity of secretion and activity of bile and pancreatic fluid, expression of influx and efflux transporters, metabolizing enzymes, and membrane permeability differ in pediatric compared to adult populations (56–58). Therefore, the pharmacokinetics and drug disposition in pediatrics might be fundamentally different from that in adults. Despite these differences, in general, studies are often conducted in adults assuming absence of drug absorption differences between children aging 2 years and above and adult population because accurate measures of intestinal absorption in children compared to adults are lacking (1, 7, 9).
Development of a definitive PBCS requires deeper understanding of the GI physiology and intestinal permeability in pediatric populations. Age-based changes in the GI fluid composition, volume, and pH are essentially critical for the prediction of age-based solubilities.
Taken together, the results of this study show that considerations for adult BCS do not apply directly to pediatric subpopulations and the regulations that allow waiver of BE studies based on in vitro dissolution studies instead of in vivo BE studies might not be appropriately applied on pediatric formulations. Standardization in conducting BE studies for pediatric formulations in pediatrics and adults might assist in understanding how drugs perform across age groups (9). Deeper understanding of the GI environment in children is needed before developing predictive dissolution testing in pediatric populations. More investigations are needed to establish age-specific PBCS that would serve as a framework in the development of formulations for the target age population.
Below is the link to the electronic supplementary material.
(PDF 213 kb)