Back to the Top
Dear group,
We have to carry out a Bioequivalence study with a
drug that has a long terminal half life (almost 7
days). Tmax is 3.5-5 hours; AUC/0-72 h ~ 40 % of the
AUC/0-inf. I have some data about the SD for Cmax,AUC
and T1/2 but no reports on the intrasubject
variability in distribution or clearance.
Considering a cross-over design we should collect
samples for at least 21 days after the oral dose and
wait for 7-10 weeks for wash out between periods. This
would require a lot of time and perhaps, many
dropouts.
I would like to ask if anyone can give me some advice
considering his own experience/references to decide
which is the best deign for this study, considering a
parallel, traditional cross-over or cross-over but
measuring truncated AUC (AUC/0-72 h).
Thanks!
Paula Schaiquevich, PhD
Buenos Aires
Argentina
Back to the Top
Dear Mr. Schaiquevich,
It's very interesting.
Can you give some idea about the apparent half-life of this molecule?
Most of the guidelines suggest that, if the half-life of the drug is
more than 24 hours, you can collect blood samples minimum upto 72
hours. So I think regulatory point of view, you will not have any
problem. As the half-life of your drug is very long, I feel you should
go at least upto 144 hours sampling, so that you will get the proper
curve and AUC i.e. (AUC 0-144), and will reduce your problem of
(AUC/0-72 h ~ 40 % of the AUC/0-inf). Regarding the wash out, it will
be very long and may be more than what you have suggested.
A pilot study with less wash out and with different time points at 96,
144 , ..... etc will definitely help to finalize the time points & wash
out. In pilot study with less wash out you can see, is there any carry
over effect or not and if not, you can reduce the wash out period and
can save some time. For drop out you can always enroll few more
volunteers for this type of study & molecule. I think cross over
truncated (e.g. AUC 0-144) may help for your final design.
Any other suggestions ?
Regards,
Dr. Nirav Gandhi
Back to the Top
Dear Paula,
A cross-over study with 1 month washout period and sampling for adequate
time to be able to determine Cmax, Tmax (for example determining
truncated
AUC/0-72 hours or 0-96 hours) should be ok, providing that your design
is
described, and justified, in the protocol.
Best regards
K. Avgoustakis, PhD
Department of Pharmacy,
University of Patras,
Greece
Back to the Top
Dear Paula,
The best approach will be a parallel design.
You can use the truncated AUC0-72 in the parallel design as well if this
is acceptable to the Regulatory Agency.
Due to the very long washout the crossover design is not appropriate. In
BE studies it is assumed that the individual clearance values do not
change from one dosing event to the other. The longer the washout the
less
probable is this constancy of the clearance.
The only disadvantage is that parallel designs usually requires larger
number of subjects since the inter-subject variability is generally
larger
than intra-subject variability.
I hope this helps.
radu
Back to the Top
Would it make sense to demonstrate pharmacodynamic equivalence rather
than worrying about some pharmacokinetic equivalence? This case might
be the classic one to demonstrate clever use of biomarkers. In fact,
there are examples, Metoprolol IR vs Metoprolol XL (Abrahamsson et al.
J Clin Pharmacol 1990 30 (2 Suppl) S46-54), where biomarkers have been
used demonstrate equivalence.
I could think of the following advantages and disadvantages:
1. More relevant clinically
2. If conc-biomarker relationship is known, selection of good
end-points will allow less number of samples (May be number of
subjects also, but depends on PK &/PD variability) in the study
compared to the PK study.
3. Parallel design
4. Like Metorprolol study, healthy volunteers could be used.
Disadvantages:
1. May not applicable when biomarker-clinical outcome relationship is
not established.
But I think of it as an equivalence trial and no the effectiveness
trial. So as long as we have reasonable mechanistic basis for
selecting a biomarker, we should be okay.
2. May not be applicable under all clinical conditions.
I look forward to see what group has to say.
Paula, I am not aware of the specifics of the indication, so this was
a general reply.
Thanks.
Pravin
--
Pravin Jadhav
Graduate Student
Department of Pharmaceutics
MCV/Virginia Commonwealth University
DPE1/CDER/OCPB/Food and Drug Administration
Phone: (301) 594-2623
Fax: (301) 480-3212
Back to the Top
Though PD equivalence is the desired endpoint, the inherent noise of PD
measurement would necessitate a larger number of N than that required by
PK equivalence.
My 2c worth.
Back to the Top
Dear Radhu,
Your comment: The longer the washout the less probable is this
constancy of the clearance.
Can you explain it some what more clearly and with example, if
possible. Because I feel sufficient wash out may be enough to overcome
this problem in case of crossover design.
Dr. Nirav Gandhi
Back to the Top
Thanks everybody for your comments.
In the FDA guidance, Bioavailability and
Bioequivalence studies for orally administered drug
products, Special Topics, point C, it says " For drugs
that demonstrate low intrasubject variability in
distribution and clearance, an AUC truncated at 72
hours can be used in place of AUC /0-t or AUC /0-inf "
Then, I understand that first I have to carry out a
pilot study to determine the intrasubject variability
in the kinetic parameters.
However, I still have some questions:
1- Do I have to take samples only up to 72 hours after
the oral dose in case I demonstrate low intrasubject
variability?
2- In case we find a high variability in clearance and
distribution, then, I could not use the AUC truncated
but if I carry out a cross-over design that will last
at least 4 months, who knows whether the parameters
will change with time?
I would really apreciate if somebody could give me
references where I can find some information.
Thanks!!
Back to the Top
Dear Paula,
Q- 1- Do I have to take samples only up to 72 hours after the oral
dose in case I demonstrate low intrasubject
variability?
A-1 You can take upto 72 hours and also you can also go upto 96, 144.
... as per your convenience and expected results.
Q- 2- In case we find a high variability in clearance and
distribution, then, I could not use the AUC truncated but if I carry
out a cross-over design that will last at least 4 months, who knows
whether the parameters will change with time?
A-2 I think in this case you can go for parallel study with somewhat
larger sample size and can save time & environmental variability. The
last sampling point may be very prolonged. What about the FDA guideline
regarding sampling ? Here also you can correlate the LOQ of the
analytical method, because it should be sensitive enough to detect the
plasma concentration upto last sampling time point (e.g. the time point
at 21st to 23rd day).
Any comment from group.
Dr. Nirav Gandhi
Back to the Top
>Q- 2- In case we find a high variability in clearance and
>distribution, then, I could not use the AUC truncated but if I carry
>out a cross-over design.......who knows
>whether the parameters will change with time?
I don't understand the point of this question. I read it several times
and looks like you are more worried about the inter-occasion
variability (IOV) in PK than inter-individual variability(IIV).
For such cases, it is well known that there is no advantage in
conducting a cross-over study. In fact, it might obscure equivalence
profile.
.....parameters will change with time.....I think you mean particular
subject's PK parameter, say clearance, in the first period was X and
during the second period it was X+dx (X-dx is also possible,
irrelevant) due to UNKNOWN reasons.
If you are worried about the extent of this variability (IOV) and have
a reason to believe that this variability will be the main contributor
then a cross-over study is not going to help in demonstrating
equivalence. Parallel group design need to be explored.
Better explanation of this topic is provided in the recent review by
Rani et.al. (Indian J Pharmacol Aug 2004 36(4) 209-216) in General
Concepts in Study Design Section. It also has reference to the seminal
articles by Schuirmann et.al., Westlake et.al. and Hauck et.al., if
you have data analysis questions.
Or by change with time... do you mean patient's renal function is
decreasing over time and thus it is responsible for change in
clearance (not strictly IOV but has a mechanistic reason). However,
you are not going to conduct a BE study in patients where this
phenomenon is more likely. So why bother? Am I missing something here.
Pravin
--
Pravin Jadhav
Graduate Student
Department of Pharmaceutics
MCV/Virginia Commonwealth University
DPE1/CDER/OCPB/Food and Drug Administration
Phone: (301) 594-2623
Fax: (301) 480-3212
PharmPK Discussion List Archive Index page
Copyright 1995-2010 David W. A. Bourne (david@boomer.org)