Bioequivalence and Bioavailability Forum

Dear Martin!

❝ what do you think about comparing dose-adjusted AUCs? on assumption of dose-proportionality the ratio of dose-adjusted AUCs should be one.

Yes, you are right. Actually this is the first model in Chow & Liu's procedure. If BE is demonstrated within conventional limits after dose-normalization, dose linearity (a straight line passing through the origin: E(Y|x) = b·x) is proven.

❝ in case of more than two dose-levels, application of an adequate stepwise testing procedure (based on the closed testing approach) on dose-adjusted AUCs can help to identify the dose-range where dose-proportionality holds. this approach keeps the familywise error rate at the pre-specified alpha-level which should be more powerful than the application of the bonferroni correction.

Again yes. Since to my knowledge there are no regulatory specifications given for dose-proportionality testing, any approach (if stated in the protocol and agreed upon by IEC and authority) should be acceptable.