Bioequivalence and Bioavailability Forum

Dear d_labes,

❝ I think the discriminatory value of the partial AUC is well enough seen from the examples in the presentation.

❝ Thus the additional metrics only raise the burden.

I agree.
Another practical issue for the sponsor will be power. In a normal BE study with a tablet having just one Cmax, the power is typiclly based on the expected T/R and the expected CV, the latter typically for Cmax as this is oftentimes slightly more variable than AUC.
In simple cases we might write something like P(Success) = P(BE AUC and Cmax) = P(BE AUC) * P(BE Cmax), but in reality Cmax and AUC are correlated and therefore the two components are not independent. It is implicitly assumed anyway when companies just ignore the less variable component and power the study for Cmax and assume that if Cmax is BE then AUC will be as well. This is a proven successful strategy as long as we don't talk inhalation products or other nasty things.
Now for the bi-phasic formulations. Here I am not sure what to do. I imagine a bunch of problems in practice. We will have four parameters that need to be BE, all of them are correlated. One could of course try and ID the most variable of them (C_max,II?) and power on basis of that but I would certainly not feel comfortable to relay the message to management that the chancee of success is 80% if the dimensioning is based on power consideration for just one of the three/four parameters.

Finally, linear kinetics: How to assess it? What if something appears linear based on phase I but not really linear based on phase II?