Dear Sir,Back to the Top
I would like to know the AUC, Cmax and t-half of Diphenhydramine,
Isoproterenol and Vancomycin in mice with different routes of
Please send the results along with the references to the above email id.
Dear colleagues,Back to the Top
Two products are bioequivalent if the 90% confidence interval for the
percentage ratio of the lnAUC and lnCmax of the test product (generic) /
reference product (innovator) fall within the bioeaquivalence interval
May I have your opinion on this - While the Cmax is within the
bioequivalence interval, what if we get a range of, for example,
131.11 for AUC? Can we not say that the test product (e.g. rifampicin)
demonstrated a better bioavailability performance (AUC being the
of the extent of absorption) than the reference product? ANOVA would
show that there is no statistically significant differences in subjects,
treatment sequence, period and formulation means.
Dear JoyBack to the Top
If the test product has higher AUC than reference, you can say that it
has better bioavailability than reference, but in your case they are
not bioequivalent (90% CI > 125).
Considering the 90% CI of 101.51 to 131.11 for AUC, I imagine that the
mean are quite far from 100, which indicates that the test formulation
has to be changed to be bioequivalent to the reference, if this is the
Hope this helps,
Paula Macedo Cerqueira, PhD
Pharmacist- Technical Consultant
National Health Surveillance Agency - ANVISA
Bioequivalence Unity - S\0x201Eo Paulo -SP- Brazil
Hi Joy,Back to the Top
The fact that the 90%CI for log-transformed AUC falls outside (above)
the 80-125% interval indicates the study data have failed to
demonstrate equivalence of the formulations. Since the 90% CI falls
entirely above a ratio of 100% one can have confidence that the test
product demonstrated greater bioavailability (i.e., AUC) than the
reference. For regulatory decisions which require equivalence to be
demonstrated (e.g., generic approval or formulation changes with a new
drug) having greater bioavailability than the reference is not
"better", it is inequivalent.
The assessment of bioequivalence of formulations is performed to use PK
exposure as a surrogate for safety and efficacy analyses on the new
formulation. Since greater bioavailability may be related to increased
safety risks, greater bioavailability is not "better", nor is it
generally approvable (based on bioequivalence only).
The use of the ANOVA analysis you describe is not really an issue for
demonstrating bioequivalence any more. It could be that the ANOVA
analysis lacked significance because the study was under powered.
Hope this helps.
Joel S. Owen, Ph.D.
395 S. Youngs Road
Buffalo, NY 14221
(v) (716) 633-3463 ext. 247
(f) (716) 633-7404
Joy,Back to the Top
In BE studies, the product is either bioequivalent or it is not. There
is no such thing as better bioavailability. Even if the 90% CI for AUC
is 125.1 (or for that matter 79.9), then the product is not
Dear joy,Back to the Top
Unfortunately you cannot. If a generic is bioequivalent to the
means a patient can use the generic instead of the original. When your
higher it means that more of the compound is available in the body.
lead to other safety issues and thus you cannot automatically say that
drug is equal or better than the old one.
If you want to show that your drug is better, you have to do a
study on the effects of the drug. Of course that involves much more
money than a simple bioequivalence trial.
Dear Joy,Back to the Top
Have you considered that you are measuring relative bioequivalence and
the selection of the reference product could affect the outcome of your
study? What criteria were used for this selection, and what tests were
to prove that this product was of a high quality prior to study
We have had similar problems with antitiberculosis drugs, even with
preparations that have been marketed and are available commercially.
Division of Pharmacology
K45-68 Old Main Building
Groote Schuur Hospital
Tel. +27 (0)21 406 6498
Fax. +27 (0)21 448 1989
Dear members,Back to the Top
During two period crossover randomized bioequivalence test of
Loratadine in 40 healthy male subjects, many of the subjects ( about 12
in reference and 14 in test products) do not show any concentration or
very minimal concentration with respect to parent drug as well as
metabolite i.e Desloratadine.There is no clinical sign in the subjects
which can be correlated with this kind of variation.
On observing the PK parameters, they are comparable.
Pharmacokinetic parameters of Loratadine (Mean + SD)
Tmax (hr) 0.980 + 0.599 0.993 + 0.448
Cmax (ng/ml) 3.556 + 2.613 3.225 + 2.656
AUClast (ng*hr/ml) 4.514 + 4.176 4.057 + 3.743
AUCinf (ng*hr/ml) 7.380 + 5.404 6.186 + 3.879
Pharmacokinetic parameters of Desloratadine (Mean + SD)
Tmax (hr) 1.382 + 0.606 1.660 + 1.176
Cmax (ng/ml) 3.039 + 1.727 2.897 + 1.686
AUClast (ng*hr/ml) 18.600 + 15.311 19.418 + 17.471
AUCinf (ng*hr/ml) 30.153 + 17.895 31.565 + 23.905
While, the Ln transformed pharmacokinetic parameters at 90% confidence
interval indicate the non-bioequivalence of the formulations.
All 40 subjects 90% CI Lower 90% CI Upper
Ln(Cmax) 80.46 113.24
Ln(AUClast) 80.99 126.08
Ln(AUCinf_obs) 82.69 116.07
All 40 subjects 90% CI Lower 90% CI Upper
Ln(Cmax) 79.85 104.09
Ln(AUClast) 57.17 103.67
Ln(AUCinf_obs) 67.81 110.35
What could be the reason for this much variability and very less or
nil concentraions in the some of the subjects ( reference & test both).
What would be yours suggestions/comments about this study.
Dear Ruchy,Back to the Top
unfortunatelly no suggestions from my side, just a few comments.
Looking at your data, some points are interesting:
Your results show high mean residual areas (AUClast/AUCinf) of
38.8%/34.4% (test/reference, parent) and 38.3%/38.5% (test/reference,
metabolite). If you primary target in the protocol was AUCinf, your
study may not be accepted by regulators.
Loratadine shows an extensive first pass metabolism and is known for
it's high variability.
You did know that, and therefore included 40 subjects (which would
allow for a variability of 30% at point estimates of \0xB15% from unity
with 80% power).
CVs calculated from your data are for the parent 48.1% (Cmax), 45.9%
(AUClast), and 47.3% (AUCinf); and for the metabolite 36.3% (Cmax),
93.0% (AUClast), and 71.9% (AUCinf). Clearly your study is underpowered
to show BE (or in other words, your expectation of 30% variability was
Since the main metabolic pathway is mediated by CYP3A4, and ranges of
elimination half-lives in adults are reported with 3 to 20 hours for
the parent, and 9 to 92 hours for the metabolite, my first thought was
that the small concentrations of parent are only seen in extensive
(fast) metabolizers, whereas high concentrations are seen in poor
But: this is impossible, since you don't see 'normal' concentrations of
the metabolite in these subjects as well. Very strange.
Generally in BE-studies substances showing polymorphism are not
problematic, since a subject shows either low _or_ high concentrations,
both after test _and_ reference, keeping the intra-subject CV low.
Did you see high/low subject ratios as well (e.g., a subject beeing
'high' on test, but 'low' on reference and vice-versa) ?
What could be the reason for [...] very less or nil concentrations in
the some of the subjects (reference & test both).
If you really have 'nil concentrations' in some of the subjects, how
did you calculate your confidence intervals of ln-transformed data of
Consultancy Services for Bioequivalence and Bioavailability Studies
tel/fax +43 1 9713935
Hi RuchyBack to the Top
My interpretation of your data is that some subjects who do not show
any parent and
metabolite concentrations in blood did not take the drug at all or had
dosing without telling clinical supervisor. That happened once with us
and have increased
data variability a lot, and consequently caused hall study failure!!
PharmPK Discussion List Archive Index page
Copyright 1995-2010 David W. A. Bourne (email@example.com)