Back to the Top
Dear all,
I am interested in your opinion on the following issue with regard to
bioequivalence:
Let's say that for a particular study, that one of the primary
variables is AUCinf, but out of a sample size of 20, 10 of the AUCinf's
were not included in the analysis since they were extrapolated by more
than the recommended 20%. The CI from the remaining 10 AUCinf's was
within the acceptable range of 80 to 125%. What exactly can we conclude
from this. Are we able to conclude that bioequivalence was demonstrated
with respect to AUCinf, or does the fact that half of the subjects'
data were not included in the analysis invalidate the results obtained?
Your opinion will be greatly appreciated.
Hanna-Liza Jooste
Biostatistician
QdotPharma, South Africa.
Back to the Top
The following message was posted to: PharmPK
Dear Hanna-Liza,
The guidelines written for BE studies refer to the actual data
available and not the number of subjects. By that, I mean that if you
calculate that you need to include 100 subjects in a study, this
calculation refers to the number of subjects you need to get data from
and not the number of subjects who participated. If you do a study
including 20 subjects but can only use data from 10, you are obviously
10 subjects short. The conservatives would say that you must do all
the study over again, including 20 new subjects and making sure you
have long enough sampling so data from all subjects can be used. Some
people might argue that it would be a waste not including the data
from those fully studied 10 subjects. So just include 10 new subjects
and pool these new 10 with the old 10 and you have your 20 subjects.
As long as the formulation used is the same batch (and there is no
evidence of the compound being degraded within the time period from
the first to the complementary study), it might sound OK. One might
argue, however, that in order to make sure that the data from these
new subjects are useable, one might collect samples for a longer
period of time. A longer sampling period, however, makes the two
groups different and non-comparable.
Toufigh Gordi, PhD
Back to the Top
The following message was posted to: PharmPK
In addition if you ae mixing data from two different or subsequent
studies (as is described below) then an element of bias gets added
to statistical evaluation.
You already know results of 10 subjects and you have not achieved
the desired results. And therfore you afer adding more volunteers
to get the desired result. to counter the effect of bias the alpha
level needs correction.
With best regards
Dr. Prashant
Back to the Top
The following message was posted to: PharmPK
Dear Hanna-Liza,
some comments:
Hanna-Liza Jooste wrote:
> Dear all,
> I am interested in your opinion on the following issue with regard to
> bioequivalence:
> Let's say that for a particular study, that one of the primary
> variables is AUCinf,
>
I think there could be only one primary variable in a BE study (either
AUCinf or AUClast, otherwise you may run into problems with
multiplicity, i.e., inflation of the type I error). But you are free to
test anything else you like (virtually any PK parameter) descriptively
in an exploratory manner.
> but out of a sample size of 20, 10 of the AUCinf's were not included
> in the analysis since they were extrapolated by more than the
> recommended 20%.
>
If you have stated it that way in the protocol, fine.
> The CI from the remaining 10 AUCinf's was within the acceptable range
> of 80 to 125%. What exactly can we conclude from this. Are we able to
> conclude that bioequivalence was demonstrated with respect to AUCinf,
> or does the fact that half of the subjects' data were not included in
> the analysis invalidate the results obtained?
>
> Your opinion will be greatly appreciated.
>
I think there is a rather high chance of getting questions form
regulators concerning the appropriateness of your sampling schedule
and/or the sensitivity of the analytical method. Unfortunatelly you
cannot change anything in these parts of the study post-hoc....
Since AUClast is the primary variable in the European Union I would
suggest looking at the BE of AUClast of the whole data set. Doubling
the samle size improves your statistical power, and the (exploratory)
assessment of AUClast should support your claim of bioequivalence.
> Hanna-Liza Jooste
> Biostatistician
> QdotPharma, South Africa.
>
Best regards
Helmut
Helmut Schütz Biokinet GmbH / Dept Biostatistics
Neubaugasse 36/11 Nattergasse 4
A-1070 Vienna/Austria A-1170 Vienna/Austria
tel/fax +43 1 9713935 tel +43 1 4856969 62
no cell phone ;-) fax +43 1 4856969 90
http://www.goldmark.org/netrants/no-word/attach.html
Back to the Top
The following message was posted to: PharmPK
For BE study, you need Cmax, AUC last, and AUC inf. All three criteria
need to meet the guideline set by the FDA. If the design of the BE
experiment yielded AUCinf outside the extrapolate range, you will have
to repeat the BE to correct for that fact. Without AUC inf, you can't
claim BE.
Back to the Top
Dear Vuong,
I would suggest not to answer questions by referring to a guideline
only and expressing something like "you have to...".
By the way, the question came from South Africa, maybe FDA guidelines
simply don't apply there.
If you write
[...] without AUC inf, you can't claim BE.
this must read "without AUCinf, you can't claim BE in the US and some
other countries".
According to EMEA guidelines (EU: 380 Mio people) main target parameter
is AUClast, and for long half life drugs even for the FDA (US: 280 Mio)
truncated areas are ok.
I do respect the FDA, but please remember this a global (rest of the
world: 5600 Mio) list, so questions should be seen according to this
perspective (eg., Brazil adopted FDA rules [and went even further!],
whereas Turkey and Malaysia EU rules...) and answered accordingly.
Best regards
Helmut
Helmut Schütz Biokinet GmbH / Dept Biostatistics
Neubaugasse 36/11 Nattergasse 4
A-1070 Vienna/Austria A-1170 Vienna/Austria
tel/fax +43 1 9713935 tel +43 1 4856969 62
no cell phone ;-) fax +43 1 4856969 90
http://www.goldmark.org/netrants/no-word/attach.html
PharmPK Discussion List Archive Index page
Copyright 1995-2010 David W. A. Bourne (david@boomer.org)