Alfredo García-Arieta

Interactions between active pharmaceutical ingredients and excipients affecting bioavailability: Impact on bioequivalence ☆

The aim of the present paper is to illustrate the impact that excipients may have on the bioavailability of drugs and to review existing US-FDA, WHO and EMA regulatory guidelines on this topic. The first examples illustrate that small amounts of sorbitol (7, 50 or 60mg) affect the bioavailability of risperidone, a class I drug, oral solution, in contrast to what is stated in the US-FDA guidance. Another example suggests, in contrast to what is stated in the US-FDA BCS biowaivers guideline, that a small amount of sodium lauryl sulphate (SLS) (3.64mg) affects the bioavailability of risperidone tablets, although the reference product also includes SLS in an amount within the normal range for that type of dosage form. These factors are considered sufficient to ensure that excipients do not affect bioavailability according to the WHO guideline. The alternative criterion, defined in the WHO guideline and used in the FIP BCS biowaivers monographs, that asserts that excipients present in generic products of the ICH countries do not affect bioavailability if used in normal amounts, is shown to be incorrect with an example of alendronate (a class III drug) tablets, where 4mg of SLS increases bioavailability more than 5-fold, although a generic product in the USA contains SLS. Finally, another example illustrates that a 2mg difference in SLS may affect bioavailability of a generic product of a class II drug, even if SLS is contained in the comparator product, and in all cases its amount was within the normal range. Therefore, waivers of in vivo bioequivalence studies (e.g., BCS biowaivers, waivers of certain dosage forms in solution at the time of administration and variations in the excipient composition) should be assessed more cautiously.

Computer simulations of bioequivalence trials: selection of design and analyte in BCS drugs with first-pass hepatic metabolism: Part II. Non-linear kinetics.

The objective of this work is to use a computer simulation approach to define the most sensitive analyte and study design of the in vivo bioequivalence study for all types of Biopharmaceutics Classification System (BCS) drugs undergoing first-pass hepatic metabolism under non-linear conditions. A semi-physiological model was developed in NONMEM VI to simulate bioequivalence trials. Eight classes from class I to IV BCS drugs (with high or low intrinsic clearance) in two variability scenarios (high-low) and in six drug products of decreasing quality were simulated in non-linear conditions to complete a total of 96 scenarios that were tested in single dose and steady state studies and compared with the previous results obtained under linear conditions. Parent drug in single dose is the most sensitive analyte and study design for bioequivalence trials in almost all the studied scenarios. However, this general rule has an exception not only in drugs with low permeability (class III and IV) and low intrinsic clearance, for which parent drug in steady state showed differences in the rate of exposure (Cmax) and also in some occasions in the extent of absorption (AUC), that are not reflected with the same sensitivity in the single dose scenario, but it could also be possible for Cmax in class III drugs with high intrinsic clearance. Metabolite data shows less sensitivity detect differences in biopharmaceutics quality in most of the scenarios or it gives the same information as the parent compound.

Computer simulations for bioequivalence trials: selection of analyte in BCS drugs with first-pass metabolism and two metabolic pathways.

The objective of this work is to use a computer simulation approach to define the most sensitive analyte for in vivo bioequivalence studies of all types of Biopharmaceutics Classification System (BCS) drugs undergoing first-pass hepatic metabolism with two metabolic pathways. A semi-physiological model was developed in NONMEM VI to simulate bioequivalence trials. Four BCS classes (from Class I to IV) of drugs, with three possible saturation scenarios (non-saturation, saturation and saturation of only the major route of metabolism), two (high or low) dose schemes, and six types of pharmaceutical quality for the drug products were simulated. The number of investigated scenarios was 144 (4 × 3 × 2 × 6). The parent drug is the most sensitive analyte for bioequivalence trials in all the studied scenarios. Metabolite data does not show sensitivity to detect differences in pharmaceutical quality or it gives the same information as the parent compound. An interesting point to notice is the case of class I drugs administered at a high dose when the principal metabolic route is saturated and the secondary one is not saturated. In this case a substantial reduction in dissolution rate (as it could occur in the case of a prolonged release formulation developed as a line extension of an immediate release formulation) leads to a considerable increase in the AUC of the major metabolite whose formation is saturated supporting the need to require pharmacokinetic and clinical data for new prolonged release medicinal products.

Assessment of the Regulatory Methods for the Comparison of Highly Variable Dissolution Profiles.

The objective is to compare the performance of dissolution-profile comparison methods when f2 is inadequate due to high variability. The 90% confidence region of the Mahalanobis distance and the 90% bootstrap confidence interval (CI) of the f2 similarity factor (f2-bootstrap) were explored. A modification of the Mahalanobis distance (new D-Mahalanobis) in which those points >85% were not taken into account for calculation was also used. A population kinetic approach in NONMEM was used to simulate dissolution profiles with the first-order or Weibull kinetic models. The scenarios were designed to have clearly similar, clearly non-similar or borderline situations. Four different conditions of variability were established: high (CVu2009=u200920%) and low variability (CVu2009=u20095%) for inter-tablet (IIV) and inter-batch variability (IBV) associated to the dissolution parameters (kd or MDT) using an exponential model. Forty-four (44) scenarios were simulated, considering different combinations of IIV, IBV and typical dissolution parameters. The dissolution profiles simulated using a first-order model modified the profile slope. The Weibull model allows profiles with different shapes and asymptotes and crossing each other. The results show that the f2-bootstrap is the most adequate method in cases of high variability. The method based on the 90% confidence region of the Mahalanobis distance (D-Mahalanobis) is not able to detect large differences that can be detected simply with f2 (i.e. low specificity and positive predictive value due to false positives). The new D-Mahalanobis exhibits superior sensitivity to detect differences (i.e. specificity as a diagnostic test), but it is not as good as the f2-bootstrap method.

Investigating the Discriminatory Power of BCS-Biowaiver in Vitro Methodology to Detect Bioavailability Differences between Immediate Release Products Containing a Class I Drug

The purpose of this work is to investigate the discriminatory power of the Biopharmaceutics Classification System (BCS)-biowaiver in vitro methodology, i.e., to investigate if a BCS-biowaiver approach would have detected the Cmax differences observed between two zolpidem tablets and to identify the cause of the in vivo difference. Several dissolution conditions were tested with three zolpidem formulations: the reference (Stilnox), a bioequivalent formulation (BE), and a nonbioequivalent formulation (N-BE). Zolpidem is highly soluble at pH 1.2, 4.5, and 6.8. Its permeability in Caco-2 cells is higher than that of metoprolol and its transport mechanism is passive diffusion. None of the excipients (alone or in combination) showed any effect on permeability. All formulations dissolved more than 85% in 15 min in the paddle apparatus at 50 rpm in all dissolution media. However, at 30 rpm the nonbioequivalent formulation exhibited a slower dissolution rate. A slower gastric emptying rate was also observed in rats for the nonbioequivalent formulation. A slower disintegration and dissolution or a delay in gastric emptying might explain the Cmax infra-bioavailability for a highly permeable drug with short half-life. The BCS-biowaiver approach would have declared bioequivalence, although the in vivo study was not conclusive but detected a 14% mean difference in Cmax that precluded the bioequivalence demonstration. Nonetheless, these findings suggest that a slower dissolution rate is more discriminatory and that rotation speeds higher than 50 rpm should not be used in BCS-biowaivers, even if a coning effect occurs.

Semi-physiologic model validation and bioequivalence trials simulation to select the best analyte for acetylsalicylic acid.

The objective of this paper is to apply a previously developed semi-physiologic pharmacokinetic model implemented in NONMEM to simulate bioequivalence trials (BE) of acetyl salicylic acid (ASA) in order to validate the model performance against ASA human experimental data. ASA is a drug with first-pass hepatic and intestinal metabolism following Michaelis-Menten kinetics that leads to the formation of two main metabolites in two generations (first and second generation metabolites). The first aim was to adapt the semi-physiological model for ASA in NOMMEN using ASA pharmacokinetic parameters from literature, showing its sequential metabolism. The second aim was to validate this model by comparing the results obtained in NONMEM simulations with published experimental data at a dose of 1000 mg. The validated model was used to simulate bioequivalence trials at 3 dose schemes (100, 1000 and 3000 mg) and with 6 test formulations with decreasing in vivo dissolution rate constants versus the reference formulation (kD 8-0.25 h (-1)). Finally, the third aim was to determine which analyte (parent drug, first generation or second generation metabolite) was more sensitive to changes in formulation performance. The validation results showed that the concentration-time curves obtained with the simulations reproduced closely the published experimental data, confirming model performance. The parent drug (ASA) was the analyte that showed to be more sensitive to the decrease in pharmaceutical quality, with the highest decrease in Cmax and AUC ratio between test and reference formulations.

Computer simulations for bioequivalence trials: Selection of analyte in BCS class II and IV drugs with first-pass metabolism, two metabolic pathways and intestinal efflux transporter

&NA; A semi‐physiological two compartment pharmacokinetic model with two active metabolites (primary (PM) and secondary metabolites (SM)) with saturable and non‐saturable pre‐systemic efflux transporter, intestinal and hepatic metabolism has been developed. The aim of this work is to explore in several scenarios which analyte (parent drug or any of the metabolites) is the most sensitive to changes in drug product performance (i.e. differences in in vivo dissolution) and to make recommendations based on the simulations outcome. A total of 128 scenarios (2 Biopharmaceutics Classification System (BCS) drug types, 2 levels of KM Pgp, in 4 metabolic scenarios at 2 dose levels in 4 quality levels of the drug product) were simulated for BCS class II and IV drugs. Monte Carlo simulations of all bioequivalence studies were performed in NONMEM 7.3. Results showed the parent drug (PD) was the most sensitive analyte for bioequivalence trials in all the studied scenarios. PM and SM revealed less or the same sensitivity to detect differences in pharmaceutical quality as the PD. Another relevant result is that mean point estimate of Cmax and AUC methodology from Monte Carlo simulations allows to select more accurately the most sensitive analyte compared to the criterion on the percentage of failed or successful BE studies, even for metabolites which frequently show greater variability than PD. &NA; Graphical abstract Figure. No caption available. Abbreviations: BE: bioequivalence; PD: parent drug; PM: primary metabolite; SM: secondary metabolite.

Global Harmonization of Comparator Products for Bioequivalence Studies

Comparator products should be the products that were shown to be safe and efficacious in pivotal clinical trials to ensure prescribability of generics. The use of a common comparator ensures switchability between generics. The selection of the comparator is a national responsibility and may be different between countries. This paper discusses the current recommendations on selection of comparators, the associated problems, and the possibility of harmonization. Most countries follow the World Health Organization (WHO) recommendations for selecting comparator products and require the comparator product to be obtained from their national markets to ensure switchability between the local comparator and their generics. These recommendations are only feasible in the few countries where the repetition of the bioequivalence study is economically feasible, but they are impracticable in all other countries. Furthermore, the exclusive use of the local comparator to ensure switchability is ethically and scientifically questionable. The innovator product from well-regulated markets should be the global comparator. This harmonization is feasible as the concept already applies in the WHO prequalification program. It is ineffectual to harmonize only the requirements for performing bioequivalence studies, if such a study has to be repeated for every single country simply because of the different comparator products.

Interchangeability between first-line generic antiretroviral products prequalified by WHO using adjusted indirect comparisons.

BACKGROUNDnThe scaling-up of access to antiretroviral therapy, particularly in low- to middle-income countries, was facilitated by the introduction and widespread use of generic antiretroviral medicines and fixed-dose combinations. Generic medicines are approved by regulatory authorities based on the demonstration of bioequivalence with the innovator or reference product, as well as meeting quality standards. In clinical practice, however, it is not unusual for generics to be interchanged between each other. This study investigated the differences in bioavailability between WHO-prequalified first-line antiretroviral generics by means of adjusted indirect comparisons to ensure interchangeability between these generics.nnnMETHODSnData on 34 products containing emtricitabine, tenofovir disoproxil fumarate, lamivudine and efavirenz in single formulations or fixed-dose combinations were included in the analysis. The 90% CI for the adjusted indirect comparisons was calculated using the homoscedastic method that uses the conventional t-test, and assumes homogeneity of variances between the studies and small sample sizes. The combined standard deviation of both bioequivalence studies was calculated from the variability of each individual study.nnnRESULTSnThe adjusted indirect comparisons between generics showed that the differences, expressed as 90% CIs, are less than 30%. Confidence in the interchangeability of two generic products was reduced if the mean difference between the test and reference in the original studies is more than 10%.nnnCONCLUSIONSnFrom a bioequivalence perspective, the generic antiretroviral medicines prequalified by WHO are interchangeable with the reference, as well as between each other without safety or efficacy concerns.

Validation of a semi-physiological model for caffeine in healthy subjects and cirrhotic patients

The objective of this paper was to validate a previously developed semi physiological model to simulate bioequivalence trials of drug products. The aim of the model was to ascertain whether the measurement of the metabolite concentration-time profiles would provide any additional information in bioequivalence studies (Fernandez-Teruel et al., 2009a,b; Navarro-Fontestad et al., 2010). The semi-physiological model implemented in NONMEM VI was used to simulate caffeine and its main metabolite plasma levels using caffeine parameters from bibliography. Data from 3 bioequivalence studies in healthy subjects at 3 different doses (100, 175 and 400mg of caffeine) and one study in cirrhotic patients (200 or 250mg) were used. The first aim was to adapt the previous semi-physiological model for caffeine, showing the hepatic metabolism with one main metabolite, paraxanthine. The second aim was to validate the model by comparison of the simulated plasma levels of parent drug and metabolite to the experimental data. The simulations have shown that the proposed semi-physiological model was able to reproduce adequately the pharmacokinetic behavior of caffeine and paraxanthine in both healthy subjects and cirrhotic patients at all the assayed doses. Therefore, the model could be used to simulate plasma concentrations vs. time of drugs with the same pharmacokinetic scheme as caffeine, as long as their population parameters are known, and it could be useful for bioequivalence trial simulation of drugs that undergo hepatic metabolism with a single main metabolite.