Back to the Top
Do you know if there is a statictical way to calculate the placement
of Standards and qc's?
say your curve should be 10 to 10000ng/mL. where should your
calibration standards be placed?
I would like to use an 8 point standard curve to start. would the
standard practice be a logarithmic scale? evenly distributed?
Back to the Top
The following message was posted to: PharmPK
FINDLAY et al. J. PHARM Biomed. Anal. 21 (2000) 1249-1273 ( page 1260
section 5.2.1 provides a guide. The 2001 FDA guidance for Bioanlytical
methods precisely indicates the LQC as <= 3X the LLOQ. The MQC and HQC
need to be near the middle and high end (generally this is somewhere
around 50% and 80% of theof the dynamic range) Findlay again give more
specifics.
Ed O'Connor, PhD
Technical Director, Immunoanalytical
Tandem Laboratories
115 Silvia Street
West Trenton, New Jersey
609-228-0243
Back to the Top
The following message was posted to: PharmPK
I do think that the better placement for standard is "evenly
distributed" and NOT as a logarithmic scale when the representation
of the curve is linear (y=ax+b). As example for 10 to 10 000 ng/ml
you can choose 6000, 8000 and 10000 for the highest standards and
4000, 2000, 1000 (or 500), 20 (2xLOQ) and 10 (LOQ) for the lowest
part. Your QCs could be 25, 5000 and 9000.
In the other hand, for immunoassays (RIA, EIA) the logarithmic scale
is relevant (i.e.: 10000, 5000, 2500, 1250, 625, 312, 156, 18,
etc...) since the standard curve is logarithmic.
Hope this help
Henri BENECH
CEA, France
Back to the Top
Dear Henri,
I have always been (personnally) troubled when deciding where to
place the QC with respect to the standards of the calibration curve.
Based of your example below; there is a 250 fold difference between
the lowest QC and less than a two fold difference between the middle
and high QC.
If in your study your samples are across the validated range below,
then you are fulfilling the requirements of the FDA. However, if
your samples are between the LOQ and let's say for this remark 4000
ng/mL, then you would not have at least three levels of QC's in the
range of expected concentration.
My following questions is for everyone:
1. Have you had issues similar previously? If affirmative, how did
you handle it?
2. What was your final decision, accept the actual validated range or
respike within the actual range of samples (i.e. between 10 - 4000 ng/
mL) with new QC's concentration to fulfill the FDA requirements?
3. Does anybody uses more than 3 levels of QC's during their
validation assessment?
I hope that this will generate a few responses, as I attended the
conference on Bioequivalence and Dissolution in Budapest last October
and had an interesting discussion with Dr. Vinod Shaw (former Senior
Research Scientist with the FDA and well known for his involvement in
Bioanalytical ) about this issue.
Sylvain Mandeville, Ph.D.
LAB Research Inc.
www.labresearch.com
Back to the Top
Colleagues-
Regarding preparation of QC concentrations across a range of
calibration standards and the questions from Dr. Mandeville, the
bioanalytical groups with which I've been associated prepared QCs at
3 levels across the range of calibration standard concentrations. We
occasionally used one or more extra concentrations if there was a
dilution step for some samples to bring them into the range of
calibration standards. We did not change the QC concentration levels
to span the expected range for a particular sample set (expectations
could be wrong). We never received negative comments from regualtory
authorities or QA on this practice, and received many drug
registrations.
Placement of the mid QC was always a question in our own minds
because it tended to be in the arithmetic mean of the standard curve,
and we sometimes interpreted the guidelines to keep the mid QC
concentration towards the lower part of the mid-range.
Tom
Thomas L. Tarnowski, Ph. D.
Bioanalytical Development
Elan Pharmaceuticals
800 Gateway Blvd.
South San Francisco, CA 94080
thomas.tarnowski.-at-.elan.com
Back to the Top
Dear Sylvain,
"Based of your example below; there is a 250 fold difference between
the lowest QC and less than a two fold difference between the middle
and high QC".
Of course an so? This is not a problem. Do not forget that the
representation of your standard curve is linear and not logarithmic.
In my example, you have a range of about 500 ng/ml between low QC and
medium QC and also 500 ng/ml between medium and high QC.
If in your study your samples are across the validated range below,
then you are fulfilling the requirements of the FDA. However, if
your samples are between the LOQ and let's say for this remark 4000
ng/mL, then you would not have at least three levels of QC's in the
range of expected concentration.
Yes you are right, but this is more an issue of the relevance of the
calibration curve range than an issue of the position of the QCs.
Consider also the samples between 5000 and 10000, if your QCs are,
for example, 20, 200 and 900 ng/ml, you will have not any CQ between
201 and 799 that represents 7/10 of the calibration curve!! Will you
feel comfortable?
Keep in mind, that the FDA states that the medium QC should be "near
the center of the standard curve"
Henri BENECH
CEA, France
PharmPK Discussion List Archive Index page
Copyright 1995-2010 David W. A. Bourne (david@boomer.org)