Back to the Top
The following message was posted to: PharmPK
Can anyone please tell me, why not the limits set as 100+/-20% in the
similarity (f2) statistics of test and standard. The official limits
are 80 to 125.
Kumaran. V., M. Pharm.,
Back to the Top
Kumarin:
f2 metric and dissolution similarity:
In order to appreciate the utlility of this metric you need to sit
down with a formulation/dissolution type scientist and appreciate the
application of f2 to the dissolution profile. Try a sample
calculation with a dissolution profile you have and follow the
guidance. It is completely different from a BE test applied to
pharmacokinetic profiles.
From Henry Flanner (one of the authors of the original paper).
"f2 is an index from 0-100 with 100 indicating perfect similarity.
An f2>50 is an indicator that the comparator is overall less than 10%
different from the reference. Dissolution curves are not handled
like PK HF".
Hope above helps,
Angus McLean Ph.D,
8125 Langport Terrace,
Suite 100,
Gaithersburg,
MD 20877
tel 301-869-1009
fax 301-869-5737
BioPharm Global Inc.
Back to the Top
Kumarin:
The f2 metric for dissolution is not theoretically based. This is a
controversial statement. I make it because the results of the metric
vary with the number of time points used in the comparison over the
same time period. It has no applicability to PK. I say that with some
trepidation because I am not a pharmacokineticist and I read this
discussion group for different reasons.
Stan Alekman
Stanley L. Alekman Ph.D.
Pharma Consultant
S.L. Alekman Associates Inc.
Back to the Top
The following message was posted to: PharmPK
It sure is 80 -120% provided you are analyzing on the untransformed
scale.
But we do not base our conclusions on the untransformed data because
it is
positively skewed and not normal.
We apply log transform to the data and compare it against
80-125%
for the simple reason that ln(.8) is same as ln(1.25).....(magnitudes)
Hope this will clear your doubt
Sulagna
Associate Scientist
Biostatistics
Back to the Top
The following message was posted to: PharmPK
Dear Kumaran
We can't state anything more than absolute (100 %).
Similary, in case of test and reference, only the
greater one can be treated as 100% (two one sided
hypothesis). Therefore, we can never state 100+/- 20%.
When, Reference is 100% and test is 80% (20%
difference) we say test/ reference (80/100 = 0.80) and
when test is 100% and reference is 80% (20%
difference) we say test/reference (100/80= 1.25)
That is why the official limits are 0.8 to 1.25 (on
converting to percentage it is 80 to 125)
Regards
V.Rajesh
PharmPK Discussion List Archive Index page
Copyright 1995-2010 David W. A. Bourne (david@boomer.org)