Bioequivalence and Bioavailability Forum

Dear all,

I ran into a BE data set in which 40% of the subjects have individual extrapolated AUC >20% in at least test or reference and in at least one period. However, the 'average' extrapolated AUC is <20%. The PK data from this study shows the elimination half-life of 3 hrs when other studies showed 7-9 hrs.

Although USFDA and EMEA guidelines mention about this issue but I would like to know what we should do in this situation:

1. Should we consider 'individual' or 'average' extrapolated AUC? In EMEA 1998 guideline, it states clearly that we should look at individual extrapolated AUC but in the later guideline (2001 and the 2008 draft) the word 'individual' is discarded.
   
2. If you are a (FDA) reviewer, should this study be accepted? What will you suggest the investigator? Or was there any study ever rejected with the same issue?
   
3. As an investigator, if only 'few' subjects show the extrapolated AUC >20% should the data from these subjects be excluded before doing statistical analysis?

Best Regards,
OL