Bioequivalence and Bioavailability Forum

Dear everyone,

I have questions about the FDA PSG of conjugated estrogen tablet. The screenshot is too large to upload, so I will quote the original texts here, and a link to original file. It's said:

"for the BE study under fasting condition, A replicate study design (TRTR and RTRT) is recommended to distinguish between unequal sequence and carryover effects in the single-dose fasting studies.The use of these two sequences (TRTR and RTRT) is the best design for balancing and minimizing bias should there be an unequal carryover effect in the study. A non-replicate (two-way crossover) study design may be conducted in lieu of the replicate study design, but the appearance of a statistically significant sequence effect may cast doubt on the conclusion of bioequivalence for the product."

My questions are:
1. Why FDA proposed this recommendation on using replicate design, is it to tell if the unequal sequence or carryover effects exist?
2. I check one of the previously conducted project which also use estrone (endogenous) as the analytes, and it does show significant sequence effect in ANOVA (P=0.01). This is unusual to other non-endogenous analytes as subjects are randomised to each arms. Is there any possible explanation?
3. Why FDA mentioned the appearance of a significant sequence effect will cast doubt on the conclusion of bioequivalence?

Sincerely
Darborn