Back to the Top
The following message was posted to: PharmPK
In a Ph I trial, one subject (Sub A) showed a very low
exposure profile out 6 who received active drug. Cmax=7 and AUC=300
while for
the other 5 subjects Cmax and AUC ranged from 130-210 and 700-950,
respectively. The Cmax for Sub A was similar to the minimum Cmax
observed at
the lowest dose cohort (10-X lower dose) and its AUC was comparable to
the max
AUC observed at the lowest dose cohort. Review of dosing, drug
accountability,
blood sampling, processing, labelling, storage, etc and BioA data did
not
suggest any apparent problems. In fact, 3 samples from Sub A were
included in
the incurred sample analysis and the original results were confirmed
(%CV=1-4%),
and therefore, no further re-assays were done. Reported AEs suggest the
exposure in Sub A should have been at least 100-fold higher than what
was
reported based on the relationship between exposure and AEs observed
from all
cohorts in the trial. We are planning to present and discuss the PK
data both
including and excluding the Sub A. However, I would like to know that
based on
the exposure profile of this subject mimicking the lowest dose cohort
and
expected conc to cause AEs can we call this subject an outlier? Or
this is no
point of doing this.
Rostam
Back to the Top
Dear Rostam:
What will you do if you call that subject an outlier? So
what? You still have shown a rather significant incidence (1/6) of a
very low bioavailability or absorption. Will you similarly plan to
discard the others you will probably find? How will you plan to
develop a dosage regimen for such a varied population?
It might be a good move to use a nonparametric (NP)
approach to pop modeling here. You may well find similar subjects in
the future, and that should be a major concern. The advantage of the
NP pop modeling procedure is that there are NO outliers. There is NO
assumed shape of the parameter distribution. In addition, as you
proceed to your next stage of study, development, or of clinical
application, you will always be able to develop, even in a population
with unusual subpopulations, the dosage regimen (multiple model [MM]
dosage design) which hits your desired target goal with maximum
precision (minimum weighted squared error. You might look at the
recent PharmPK discussion on uncertainties in Bayesian posterior
models, which overlaps a good deal of this one.
The main point here is that with the NP approach there are
NO outliers, and with its natural link to MM dosage design, the NP
approach to pop modeling and Bayesian adaptive control provides, at
each step of the study or clinical application, the most precise way
to plan the next move.
Also, you might look at
1. Jelliffe R: Goal-Oriented, Model-Based Drug Regimens: Setting
Individualized Goals for each Patient. Therap. Drug Monit. 22:
325-320, 2000.
2. Jelliffe R, Bayard D, Milman M, Van Guilder M, and Schumitzky
A: Achieving Target Goals most Precisely using Nonparametric
Compartmental Models and "Multiple Model" Design of Dosage Regimens.
Therap. Drug Monit. 22: 346-353, 2000.
3. Jelliffe R, Schumitzky A, and Van Guilder M: Population
Pharmacokinetic / Pharmacodynamic Modeling: Parametric and
Nonparametric Methods. Therap. Drug Monit. 22: 354-365, 2000.
4. Bustad A, Terziivanov D, Leary R, Port R, Schumitzky A, and
Jelliffe R: Parametric and Nonparametric Population Methods: Their
Comparative Performance in Analysing a Clinical Data Set and Two Monte
Carlo Simulation Studies. Clin. Pharmacokinet., 45: 365-383, 2006.
5. Bayard D, and Jelliffe R: A Bayesian Approach to Tracking
Patients having Changing Pharmacokinetic Parameters. J. Pharmacokin.
Pharmacodyn. 31 (1): 75-107, 2004.
6. Macdonald I, Staatz C, Jelliffe R, and Thomson A: Evaluation
and Comparison of Simple Multiple Model, Richer Data Multiple Model, and
Sequential Interacting Multiple Model (IMM) Bayesian Analyses of
Gentamicin and Vancomycin Data Collected From Patients Undergoing
Cardiothoracic Surgery. Ther. Drug Monit. 30:67-74, 2008.
Very best regards,
Roger Jelliffe
Back to the Top
The following message was posted to: PharmPK
Hi Rostam,
I wouldn't worry about one subject at this early point of development
stage of your compound. This, however, does not mean that you should
forget about him/her. There could be valuable information to learn
from Subject A. An interesting observation is that this subject had
AEs in line with those receiving much higher doses. Could this be due
to extensive metabolism during the first pass, resulting in large
metabolite concentration, which causes the AEs? Have you identified
the metabolite(s) of your compound and what enzymes are involved in
its metabolism? Is this subject an extensive metabolizer, e.g.
genotyped to have a lot of enzyme that metabolize your drug?
It is also noticeable that the Cmax is so much lower (20-20x) compared
to the AUC (2x). How does the concentration-time curve of this subject
compares with others?
I am also somewhat confused/curious about your study design. You
mention that this is 1 out of 6 subjects. It seems, however, that
there were several dose levels involved in the study (both higher and
lower). Does it mean that you have (possibly) many more subjects who
received the drug (at different dose levels)? How does the drug PK
looks across the dose levels? Any trends that other subjects might
show similar low concentrations, although maybe not as pronounced as
this subject?
Again, it seems to be a little too early to conclude anything based on
a first study and one subject but the observation can certainly lead
to ask new questions, which is a good thing early on in your
development program.
Toufigh
Toufigh Gordi, PhD
Clinical Pharmacology, PK/PD analysis consultant
Back to the Top
The following message was posted to: PharmPK
Hello,
Was this high dose given as multiple tablets?
Unfortunately, hand and mouth checks are not always successful in
subjects avoiding dosing.
Maybe they just took 1 and palmed the rest.
Susan
Back to the Top
The following message was posted to: PharmPK
Thank you Roger for your comments and the reference list and
also Walt, Toufigh and Susan. I was not keen to call this subject an
outlier
and wanted to see if there are any reasons to do so. As mentioned in my
original posting, we have not found any factors that could readily
explain this
observation including age, gender, diet, ethnicity or metabolism (less
certain
as limited data available). The site insisted that dosing was accurate
and hand
and mouth check is reliable. Previous records indicate that this
subject is a
trialist. Rostam
Back to the Top
You have already identified the subject as an outlier, that is
objective and should be supported by a statistical evaluation. The
underlying reason for values being outside what you use to classify
data as non outliers is an entirely other matter and necessitates some
investigation. The site's insisting on the dosing accuracy is fine
but how is that claim supported? What is a trialist?
--
I was curious too
trialist - noun, a person who participates in a trial, in particular
* a person who takes part in a sports trial or motorcycle trial.
* a person who takes part in a clinical or market test of a
newproduct. - Apple Dictionary - db]
Back to the Top
The following message was posted to: PharmPK
Hi, Rostam I have a couple of thoughts for you.
If I remember correctly you said that the kind of adverse events
observed in this subject were in line with an exposure that was much
higher than the one you measured.
If this is the case the subject is likely to have been exposed and I
would question the bioanalytical determinations.
You said that the plasma concentrations have been confirmed. That is
fine but this does not rule out a matrix effect in this subject.
What kind of bioanalytical method have you used? If LC-MS a matrix
effect could be an issue.
Try to get the baseline pre-dose plasma sample from this subject and
use this sample and prepare QCs. See if the measured and theoretical
concentrations are in line. If yes matrix effect could be ruled out.
If not this is the explanation and some endogenous compound present in
this subject produced a lover ionization in LC-MS.
Have you collected urine from this subject?
If yes try to see if Ae is in line with that observed (matrix effect
in urine might not be the same).
What about the safety and Ae produced by the drug metabolites; are
they the same as those observed with the parent drug? Could be this
subject an extensive metaboliser so that plasma concentrations of the
parent are lower but adverse events the same as generated by the
metabolite/s?
I hope this helps
Stefano
--
STEFANO PERSIANI, Ph.D.
Director
Translational Sciences and Pharmacokinetics Department
ROTTAPHARM | MADAUS
R&D Division
Rottapharm Spa
Via Valosa di Sopra, 9
20052 Monza - ITALY
stefano.persiani.-a-.rottapharm.com
Back to the Top
The following message was posted to: PharmPK
Metabolites are inactive. The fact that this subject reported AEs
similar to other subjects with much higher conc in the same cohort
deserves some attentions but subjects are talking to each other and a
trailist can pretend to blend with the rest. Thoughts on subject-
specific matrix effect is intersting but not that practical. Many
thanks for your comments.
Rostam
Back to the Top
The following message was posted to: PharmPK
Rostam: If this is a phase I, then you are following some regulatory
guidance, using SOP. If you have declared this an outlier using those
guidances and SOPs, you will need to explain 1) why you are changing
your
statistical approach and define your new approach. 2) now think of the
your
QA and regulatory examining the data finding first that you identified
the
subject as an outlier but are now reclassifying the subject using a new
evaluation.
The suggestion to examine that subjects serum/plasma or whole blood is a
good approach. You can spike into the material and determine recovery,
if
the patient has something-antibody or soluble receptor for your drug-
the
recovery will be decreased explaining the apparent decrease in
circulating
levels. If you have pre-dose samples you can run the evaluation.
Again what is a trialist
Back to the Top
The following message was posted to: PharmPK
Trialist: A subject who frequently participate in clinical trails
e.g., max his participation per year.
Want to post a follow-up message on this topic?
If this link does not work with your browser send a follow-up message to PharmPK@boomer.org with "Outlier Subjects" as the subject | Support PharmPK by using the |
Copyright 1995-2011 David W. A. Bourne (david@boomer.org)