Bioequivalence and Bioavailability Forum

Dear MGR!

❝ Can anybody tell me about Meta Analysis method used in BA/BE Studies (For example: one passed pilot Be study and pivotal BE study, one failed pivotal BE study).

Please see this post for the proper terminology.
A pilot study is always exploratory in nature, i.e., it's purpose is to get estimates of CV_intra and the point estimate, and to check the suitability of the dosage regimen, the sampling schedule, and the analytical method. If a pilot study gives a confidence interval within the accteptance range, that's nice, but no proof of BE because such a study is

• generally just underpowered to do so (pragmatic view),
  
• not planned to do so (hardcore statistical view).

There are anecdotal reports that drugs got approval based on the results of a pilot study only (if properly planned, executed, and sized - whatever that means), but I would not bet on it.
The only references I know are http://www.fda.gov/cder/guidance/5356fnl.pdf FDA's BE-guidance (Section III.A.2., last sentence) and a personal conversation with a Swedish regulator I had in May 2006…
BTW, if you have a passed pilot and a pivotal BE, why did you perform yet another pivotal BE study?

Canada and Japan allow for add-on designs – but that's a different cup of tea…

❝ If there is any sites or references regarding this kindly provide.

A rather questionable reference is

E Zintzaras and P Bouka
Bioequivalence studies: biometrical concepts of alternative designs and pooled analysis
Eur J Drug Metabol Pharmacokinet 24, 225-232 (1999)

Chapter 13 of Chow & Liu (2001) also deals with Meta-Analysis of average BE.

❝ Is it acceptable by the regulatory bodies?

Very doubtful...

---

Edit: Link corrected to latest archived copy. [Helmut]