If two medicines are bioequivalent there is no clinically significant difference in their bioavailability. Although bioequivalence is most commonly discussed in relation to generic medicines, it is important to note that bioequivalence studies are also performed for innovator medicines in some situations such as:
- Between early and late clinical trial formulations or between the formulations used in clinical trials and the product to be marketed for new medicines.
- When changes in formulation have occurred after an innovator product has been approved, for example a change in one or more excipients (inactive ingredients).
Bioequivalence is determined based on the relative bioavailability of the innovator medicine versus the generic medicine. It is measured by comparing the ratio of the pharmacokinetic variables for the innovator versus the generic medicine.
The acceptance criteria are such that to be classified as bioequivalent, plasma concentrations of the generic medicine will not differ significantly compared with the innovator medicine. Studies have demonstrated that actual differences between observed mean plasma concentrations of generic and innovator medicines were no greater than 5%. In order to determine that two medicines are bioequivalent there must be no more than a 20% difference between the AUC and C max. This is based on international consensus that differences less than this are not clinically significant. In order to establish this, the AUC and C max for the generic medicine are compared to that for the innovator medicine.
Bioequivalence is based on a comparison of ratios where the ratio of generic to innovator for each pharmacokinetic variable does not differ by more than 8:10; this is how the range for the confidence intervals is defined:
8/10 = 0.80 gives the lower limit
10/8 = 1.25 gives the upper limit
The 90% confidence intervals for the ratios of both C max and AUC should be contained within the limits 0.80–1.25. Thus bioequivalence is based on ratios where the nominal equality is 1. It is not based on differences in absolute values. In practice, the generic product should have a ratio of mean values (AUC and C max generic: innovator) close to 1, indicating equality. If the observed ratio is closer to 0.8 or 1.25, then the data would have to contain little or no variation from the mean for the 90% confidence intervals of the ratio to lie in the 0.8 to 1.25 range that is necessary to demonstrate bioequivalence.