Bioequivalence and Bioavailability Forum

We have performed the usual comparison BE study with 2 sequences (AB, BA), 2 Periods, Latin Square Design, comparing two modified release formulations A and B in 16 subjects. Now after completing the cross-over treatments we are bringing back the same 16 subjects and we are going to administer Formulation B with different meals on the same date. {We realize that the best way of doing such a comparison would have been a 3 way cross over balanced study design in the first place}. The formulation comparison (B v A) for the randomized cross-over will of course be done as usual in Phoenix WinNonlin 6.4. We are going to combine the PK data parameters from Formulation B from Phase 1 and Phase 2 and compare them with the PK parameters from the added arm (formulation B) with different meals. I am thinking to treat the data as esentially a parallel group comparison. We believe that such a comparison, although inconsistent with regulatory requirements, will provide useful information at this stage of the programme.

Any thoughts are welcome,

Angus

---

Edit: title changed to lower case [Ohlbe]