What is the experimental unit in premix bioequivalence ? June 2010 Didier Concordet

2 Bioequivalence concepts Exposure Close Concentration profiles Close effects Concentrations Effects Time pionneer generic

3 Bioequivalence concepts Time Close PK parameters AUC, Cmax Cmax Tmax AUC Close Concentration profiles Close effects

4 Current definition of bioequivalence Average Average Generic Time Concentration Reference

5 Current bioequivalence definition The two formulations are considered as bioequivalent if their concentrations profiles are close in a virtual animal : the average animal. Close population mean of AUC, Cmax. Does implicitly assume that the population is homogeneous with no specific strata. Sampling randomly animals is enough to insure absence of bias

6 Current bioequivalence definition Time Concentration The average individual is not "well defined"

7 What is the individual ? Individual : That cannot be divided into several (independent) pieces without altering what is observed. Can the animal be the individual when the drug is given in food? Concentration Time Animal fed alone The same animal fed with others Food intakes are not independent

8 What is the individual ? The individual could be the pen : - size of the pen, - social behaviour on food intake,… the breeder : - type of food, - modalities of meal preparation, - modalities of premix distribution,… the pig company ? If the animals' food intakes are not independent, the animal cannot be the individual.

9 What you are thinking about right now … The pen (size of pen, social behaviour) effect and the breeder effect (type of food, modalities of meal preparation, modalities of premix distribution) impacts both the pioneer and the test premix in the same way. So, why should we choose the pen or the breeder as the individual ? It is probably true, but…maybe not important. It depends on what we mean exactly by bioequivalence.

10 How to define the exposure for a pen, a breeder, a population ? A pen A breeder = an individual A population A sample of individuals

11 Exposures Exposure Time Concentration For a single animal For a two animals For a pen

12 Exposures Exposure For a pen Exposure For two pens

13 Exposures For a sample of pens Exposure For a breeder

14 Exposures Exposure For a breeder For a sample of breeders For a population

15 Three main types of bioequivalence Average bioequivalence : The pioneer and generic premix formulations should give close "average" exposures. Population bioequivalence : The statistical distribution of the drug exposures of the two formulations among breeders/pen should be close Individual bioequivalence : For most breeders/pen, the formulations should give close exposures. Average bioequivalence : The pioneer and generic premix formulations should give close "average" exposures.

16 Average bioequivalence Exposure for the pioneer formulation Exposure for the generic formulation Identity line Population mean Equivalence range Population mean The two formulations are "average bioequivalent"

17 Average bioequivalence Exposure for the pioneer formulation Exposure for the generic formulation Identity line Population mean Equivalence range Population mean The two formulations are NOT "average bioequivalent"

18 Average bioequivalence Only the population means are involved in the ABE definition - number of individuals = number of breeders (pens) - need of a representative sample of individuals - factors influencing PK profiles can be considered as noise that render imprecise estimation of population means - the study can be organised so that the between individual variability is minimized The argument that such or such factor impacts (or not) differentially the pioneer and the generic PK profiles is not meaningful in this case.

19 Three main types of bioequivalence Average bioequivalence : The pioneer and generic premix formulations should give close "average" exposures. Population bioequivalence : The statistical distribution of the drug exposures of the two formulations among breeders/pen should be close Individual bioequivalence : For most breeders/pen, the formulations should give close exposures.

20 2.5% Population bioequivalence Exposure for the pioneer formulation Exposure for the generic formulation Identity line Equivalence ranges The two formulations are "population bioequivalent"

21 Population bioequivalence It refers to prescriptability : An ‘individual’ who take a product, for the first time, is expecting to obtain the same therapeutic efficacy whatever the selected formulation. Two formulations would not be “population bioequivalent” if a breeder (pen) that intents to use the new formulation cannot expect the same effect as another breeder (pen) that uses the pioneer one - number of individuals = number of breeders (pens) - need of a representative sample of individuals - the between individual variability to be precisely estimated (the study should not be designed so that the between individual variability is minimized) -The possible existence of a formulation*breeder or formulation*pen interaction has no relevance when discussing prescriptability

22 Three main types of bioequivalence Average bioequivalence : The pioneer and generic premix formulations should give close "average" exposures. Population bioequivalence : The statistical distribution of the drug exposures of the two formulations among breeders/pen should be close Individual bioequivalence : For most breeders/pen, the formulations should give close exposures.

23 2.5% Individual bioequivalence Exposure for the pioneer formulation Exposure for the generic formulation Identity line Equivalence ranges The two formulations are "individual bioequivalent"

24 2.5% Individual bioequivalence Exposure for the pioneer formulation Exposure for the generic formulation Identity line Equivalence ranges The two formulations are NOT "individual bioequivalent"

25 Individual bioequivalence It refers to switchability : Two formulations of the same drug are ‘switchable’ if an individual (breeder/pen) using one formulation can expect the same therapeutic effect after being switched to the other formulation. Two formulations would not be “individual bioequivalent” if a there exists a formulation*breeder or formulation*pen interaction. - number of individuals = number of breeders (pens) - need of a representative sample of individuals - the interaction individual*formulation variability to be precisely estimated (the study should not be designed so that the between individual variability is minimized) -The possible existence of a formulation*breeder or formulation*pen interaction is relevant when discussing switchability.

26 Conclusion Need to choose the individual - breeder - pen - pig company ? Need to choose a definition of bioequivalence - average - population - individual. Whatever the chosen individual and definition of bioequivalence, a population data analysis will necessarily be used.