Vaccaro et al. Int J Clin Pharmacol Ther 2000;38(12):588-94
Fig. 6.32. Effect of Food
Reprinted with permission from Dustri-Verlag, Rockledge, FL.
would conclude dose proportionality. In the telithromycin example shown in the figure, the statistic test indicated that the AUC was deviated from dose proportionality at the doses tested, possibly due to a saturable first-pass metabolism of the drug. Another commonly used dose-proportionality assessment method is to use a power model, which assumes that log-transformed surrogate is linear to log-transformed dose: Ln(S) = Ln(a) + b • Ln(dose). The surrogate data after log transformation are fit to the power model using a mixed effect model. When b is tested not significantly different from 1, one could conclude dose proportionality. The graph shows a dose normalized AUC versus dose for both single and multiple doses for telithromycin (Fig. 6.31), which indicates some deviation from the dose proportionality .
Normally, early first-time-in-man studies offer a good opportunity to examine the widest dose range, while the targeted crossover study designs with carefully considered washout periods between doses are the best for evaluating within and between subject variability, thus, the most robust way to test dose proportionality. Dose proportionality is an important measure of the predictability of PK when dose adjustment is needed .
Administration of food may change the pharmacokinetics of a drug by possibly delaying gastric emptying, changing the gastric pH, increasing bile flow and splanchnic blood flow, and changing lumenal metabolism (Fig. 6.32). Food can also physically or chemically interact with a dosage form of a drug, and thus food can either increase or decrease the bioavailability or delay the absorption of the drug. The effects of food on BA depend on the physicochemical (solubility) and pharmacokinetic (site, rate, and extent of absorption, first-pass metabolism) properties of the drug and on the dissolution of the drug substance from the drug product. The information derived from the food interaction study can (a) optimize the formulations for early and mid-stage developmental compounds, (b) enable well-designed late-stage clinical trials, (c) provide prescribing options to physician for optimal patient compliance, and (d) avoid excess variability in drug responses or even toxicity. In the figure, the graph demonstrates exemplary impacts of food on drug absorption, with a C over twofold higher and a T of about 1 versus about 7
max ° max hours in the fasted versus fed states; .
Drugs such as ampicillin, aspirin, tetracyclines, and warfarin have reduced drug absorption, and drugs such as acetaminophen, diclofenac, digoxin, and valproate have delayed absorption. Drug absorption of diazepam, propranolol, grise-ofulvin, and carbamazepine are increased with administration of food, while absorption of oxazepam, tolbutamide, telithromycin, and propoxyphene remains unaltered. The bis-phosphonates have a class label that requires patients take the drug first thing in the morning before any food and drinks. The drug in this class can form a complex by chelating calcium or other divalent minerals in the food or drinks, and the absorption of the drug is greatly dampened.
The most common effect of food study is a study evaluating the effect of a high-fat meal. This study is a crossover study where drug is administered to healthy male and female volunteers in fed or fasted state. Plasma samples are collected for 24 hours and the effect on the PK parameters, especially AUC and C , is evaluated using the bioequivalence criteria max
(however, a wider 90% CI, i.e., 70-143%, may be set for C ). Other food-effect studies include time of administra-
max tion of meal with respect to food study and special diet study such as for diabetics.
The drug-drug interaction studies are conducted to evaluate the PK and PD effects but primarily are designed to evaluate for metabolism-based PK drug-drug interaction, because metabolism changes by far are the most common of the mechanisms for DDI (Fig. 6.33). The following factors should be carefully evaluated when design a drug-drug interaction study: study population (healthy vs. patients), study design (crossover vs. parallel group; fixed sequence vs. randomized crossover), dose regimen (single or steady-state studies, dose and duration, timing of co-administration), mechanism of interaction (PK vs. PD; inhibition vs. induction), PK/PD characteristics (e.g., linear vs. nonlinear kinetics, presence of active metabolites, or delayed pharmacological response), wide or narrow therapeutic index, blinded or unblinded if a pharmacodynamic or safety outcome is to be assessed, and how the study will be interpreted in the product label. The bottom line is that the study should be able to maximize the probability for an interaction, yet still ensure the safety of the study subjects.
Several common study designs are used for DDI studies (Fig. 6.34). About 70% DDI studies used a one-way fixed-sequence crossover design by administering multiple doses of both victim drug and perpetrator drug to steady state. Such designs are best in mimicking the clinical therapy or clinical practice. The fixed sequence designs include the following, in which drug A is the theoretical victim drug and drug B is the perpetrator drug:
• Randomized crossover: Drug A (period 1) followed by drug A and drug B (period 2) or drug B (period 1) followed by drug A and drug B (period 2).
• One-sequence crossover: Drug A (period 1) always followed by drug A and drug B (period 2) or the reverse; this can be extended to 3 periods by giving drug B alone in period 2, and then drug A and drug B in period 3.
• Parallel design: Drug A in one group of subjects and drug A and drug B in another group of subjects (period 1).
Sometimes, both drug A and drug B can be victim drugs, or if the victim drug is not clear prior to the study, the design can be more complicated.
The bioequivalence criteria are used for AUC and C to max show if there is a significant interaction. There is also a possibility of conducting a multiple probe or "cocktail" study that would evaluate the effect of the perpetrator drug on various CYP isozymes. Phase III population PK analysis is useful to confirm there is no large and unexpected DDIs. Drug interaction studies are important in drug development, especially r Objectives c Effect of the NCE on the PK/PD of other drugs O Effect of the other drugs on the PK/PD of the NCE r General Features o Healthy subjects, relative homogenous o Single (long t1/2) or multiple dose (short t1/2) c Dose to maximize probability for an interaction yet still ensure safety of subjects C Rationale for selecting interacting drugs:
i Mechanistic understanding for potential interaction based on in vitro, preclinical and human MPK and safety data I Co-prescribe potential o Often conducted in parallel with Phase II or III trials
Was this article helpful?