BE assessment regarding first time point Cmax [Study As­sess­ment]

posted by Ohlbe – France, 2018-12-11 00:05  – Posting: # 19675
Views: 1,261

Dear Mann,

» Could you please share your views and suggestions from US-FDA and ICH perspective?

There is no ICH guideline on bioequivalence, and I'm not a specialist of FDA regulations and guidelines. I won't comment on the outliers part as I'm not too sure what FDA's current position is on this topic. But the issue remains the same whatever the guideline: you have a shorter Tmax with your product, resulting in a higher Cmax, hence a failure to demonstrate BE. Having an earlier time point would not change this, and may actually make things worse (even higher Cmax with your product). I doubt that any regulator anywhere would approve the exclusion of 23 of 38 subjects on this basis.

If you look at it from a regulator's perspective: they're supposed to have a conservative attitude in order to protect the patients. If your study shows BE but some subjects have a first point Cmax, regulators may challenge it because your estimate of Cmax may not be reliable. But if you fail to show BE, they're not going to challenge that and allow you to exclude the data that make your study fail. Especially if this is obviously due to a different behaviour of your product.

By the way, there is one thing you forgot to say in your first message: did you have anything in your protocol regarding the exclusion of subjects with first point Cmax ? And did the protocol plan for an outlier test to be performed at all ?

Regards
Ohlbe

Complete thread:

Activity
 Mix view
Bioequivalence and Bioavailability Forum |  Admin contact
19,302 posts in 4,102 threads, 1,317 registered users;
online 6 (1 registered, 5 guests [including 5 identified bots]).

Restlessness is discontent –
and discontent is the first necessity of progress.
Show me a thoroughly satisfied man –
and I will show you a failure.    Thomas Alva Edison

The BIOEQUIVALENCE / BIOAVAILABILITY FORUM is hosted by
BEBAC Ing. Helmut Schütz
HTML5