- On 10 Jun 2014 at 20:52:57, shiym7 (shiym7.at.163.COM) sent the message
Dear all,
I ran a PK study and LLOQ = 0.05 ng/mL, most of the plasma cocentrations were in the range of
0.05~30 ng/mL, I am wondering for the detected plasma concenrations near LLOQ, eg. plasma
concentration was 0.0495456 ng/mL, should I round it to 0.05 ng/mL then include this concentrtion
into PK parameters calculations? And for LLOQ=0.05, how many number of significant digits should be
selected to report? Should all the values before rounding lower than LLOQ be reported as BLQ?
Any help would be much appreciated.
Best regards,
Yuming
Back to the Top
- On 10 Jun 2014 at 12:46:00, Edward O'Connor (efoconnor.aaa.gmail.com) sent the message
What does your bioanalytical method direct? What is the greatest precision known? This is usually
on the reference standard for the method. How many sig figs were used on the reference standard COA
or label? You can match but should not exceed that number of sig figs. Your method claims a
certain level of accuracy/precision, you need to stay with that so yes round before reporting and if
they are you report, 0.0495456 rounds to 0.0495 with 3 sig fig, this is it. And now your are in the enviable position of altering your acceptance criteria a posteriori
Back to the Top
- On 10 Jun 2014 at 15:37:13, ttarnowski1.aaa.aol.com sent the message
Hi Yuming-
Regarding your question of how to report values that are numerically less than the LLOQ (e.g., LLOQ
= 0.05 ng/mL, a value is 0.0495456 ng/mL) and the number of significant figures to report, I have
some comments.
1. A standard followed by many in the field is 3 significant figures. You should have an SOP to
describe the way numbers are to be reported (e.g., both rounding and also number of significant
figures or otherwise) and follow it.
2. The reported LLOQ values should also follow the SOP.
3. In the specific case you cite, if the LLOQ is stated as 0.0500 (3 significant figures) and the
result rounded to 3 significant figures would be 0.0495, the result value would be reported as BLQ.
Tom
Thomas L. Tarnowski, Ph.D.
Director, Bioanalytical Chemistry
Gilead Sciences, Inc.
thomas.tarnowski.at.gilead.com
ttarnowski1.-a-.aol.com
Back to the Top
- On 10 Jun 2014 at 22:16:47, Prof. Dr. Zafar Iqbal (zafar_iqbal.aaa.upesh.edu.pk) sent the message
Dear Yuming
Look at the peak shape and noise level, if these are good then I think you can round these figures
Dr Zafar
Back to the Top
- On 10 Jun 2014 at 23:35:35, chirag khatri (cakhatri.aaa.gmail.com) sent the message
Dear All,
Rounding off should be done according to in-house SOP (SOP is must). If SOP is not available, a
proper procedure to be followed for rounding should be documented on a memo, approved by QA and then
apply rounding off. This document should be part of the study raw data and also information should
be captured in the Bioanalytical report.
However since the results are likely not to be affected there should not be any regulatory concern
Regards
Chirag
Back to the Top
- On 11 Jun 2014 at 09:44:00, shiym7 (shiym7.aaa.163.com) sent the message
Dear all,
Thank you very much for all the helps. We will keep 3 significant figures for all the PK results and
method qualifications results.
Thanks,
Yuming
Back to the Top
- On 11 Jun 2014 at 09:44:46, Roger Jelliffe (jelliffe.-a-.usc.edu) sent the message
Dear All:
I would strongly suggest that you all look at
Jelliffe RW, Schumitzky A, Van Guilder M, Liu M, Hu L, Maire P, Gomis P, Barbaut X, and Tahani B:
Individualizing Drug Dosage Regimens: Roles of Population Pharmacokinetic and Dynamic Models,
Bayesian Fitting, and Adaptive Control. Therapeutic Drug Monitoring, 15: 380-393, 1993.
You should see that using CV% is incorrect, and that the reciprocal of the assay variance at a
measured concentration is the correct measure of precision. Why the lab community keeps on with CV%
is beyond me. Please look at this article. It clearly shows that if you use the correct 1/var, you
can track the SD of an assay all the way down to and including zero, and that the concept of LLOQ is
an illusion brought about by the use of CV%. The lab community and the FDA and the College of
American Pathologists should all look at this and re-examine the obsolete policy of CV%.
Very best regards,
Roger Jelliffe
Back to the Top