Back to the Top
Hi All,
What is the most accepted way to determine outliers in sample analysis for a PK curve ? Does FDA have a guideline for this?
Thanks!
Hao Pan, Ph.D.
PK Scientist
Regeneron Pharmaceuticals, Inc.
Back to the Top
The following message was posted to: PharmPK
Hi Hao,
The regulatory agencies do not allow identification of outliers based on PK. In fact, the other way is to look for outliers or for repeat analysis is from bioanalytical perspective (bad chromatography, IS variation, injection needle failure etc.). For this you should have in-house SOP's. If the concentration in question is fine analytically and still it does not fit into the PK curve, i don't think it can be excluded. Other option is to submit the data to regulatory with and without this anomalous concentration.
Comments from the experts are invited.
Thanks
Tausif Ahmed, Ph.D.
Principal Scientist,
DMPK, Sai Advantium Pharma Ltd., Pune, India
Back to the Top
That is not entirely true. It is more relevant to id outliers for PK parameters in PK analysis rather than in bioanalytical. Bioanalytical should be constrained as to whether or not an assay or specific sample result met acceptance criteria.
Pharmacokinetic analysis should present summaries both inclusive and exclusive of the point(s) in question and the reason for exclusion (re-analysis may be requested with justification). As with most other components, the re-analysis should be done twice (a total of three times including the initial analysis).
Back to the Top
The following message was posted to: PharmPK
Dear all,
I disagree with Tausif Ahmed - and agree with Edward O'Connor.
To quote FDA's Guidance for Industry on Bioanalytical Method Validation (2010, page 15,
http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/ucm070107.pdf
):
It is important to establish an SOP or guideline for repeat analysis and acceptance criteria. This SOP or guideline should explain the reasons for repeating sample analysis. Reasons for repeat analyses could include repeat analysis of clinical or preclinical samples for regulatory purposes, inconsistent replicate analysis, samples outside of the assay range, sample processing errors, equipment failure, poor chromatography, and *inconsistent pharmacokinetic data*. Reassays should be done in triplicate if sample volume allows. The rationale for the repeat analysis and the reporting of the repeat analysis should be clearly documented.
A similar formulation is given Health Canada's - and many other countries' guidelines.
Imagine a situation where something (you could not control and did not notice at all) happened prior to bioanalysis preventing a reliable result. Examples: a stabilizer was forgotten to be added to the whole blood sample, the sample was not put on ice, or even the vacutainer was contaminated. And by contaminated I don't mean by the drug itself, but anything which may result in a matrix effect in LC/MS-MS... In such a case you will repeat the analysis forever and get the same erroneous result.
I strongly suggest to perform a blinded plausibility review of bioanalytical data as early as possible by an experienced pharmcokineticist. My procedure call for at least 1/3 of subjects in the study and set up a ruleset based on the variability of the profile. It's up to you to come up with a clever set of rules. Suggestions: >+/-2SD in log-scale, >1.5times interquartile range, etc. Now calculate an estimate of the suspected value based on its neighbours. Linear interpolation before tmax, loglinear afterwards. Apply your ruleset. If the suspected value is outside, initiate reanalysis. If the value is confirmed according to your SOP, the value may be dropped from the profile.
Remark: FDA definitely don't like that, but there's even a form to submit such data (Table 9 in "Formatting of Bioequivalence Summary Tables").
http://www.fda.gov/downloads/Drugs/DevelopmentApprovalProcess/HowDrugsareDevelopedandApproved/ApprovalApplications/AbbreviatedNewDrugApplicationANDAGenerics/UCM120957.pdf
As far as I know, there is only one region, where PK-based reanalysis is not acceptable, namely the European Union, which states in their recent guideline (January 2010, page 13,
http://www.ema.europa.eu/pdfs/human/qwp/140198enrev1fin.pdf
):
Reanalysis of study samples should be predefined in the study protocol (and/or SOP) before the actual start of the analysis of the samples. Normally *reanalysis of subject samples because of a pharmacokinetic reason is not acceptable*. This is especially important for bioequivalence studies, as this may bias the outcome of such a study.
A similar formulation is given in EMA's Drafted Guideline on Validation of Bionalytical Methods (November 2009, page 12,
http://www.ema.europa.eu/pdfs/human/ewp/19221709en.pdf
):
[Same formulation as above, followed by:] However reanalysis might be considered as part of laboratory investigations, to identify possible reasons for results considered as abnormal and to prevent the recurrence of similar problems in the future.
Best regards,
Helmut
-
Ing. Helmut Schuetz
BEBAC - Consultancy Services for
Bioequivalence and Bioavailability Studies
Neubaugasse 36/11
1070 Vienna, Austria
Back to the Top
Dear all-
Regarding reassay for inconsistent PK profiles, My experience is that regulatory authorities are very skeptical of any process that could be perceived of reanalysis in order to produce a desired result. Reassay for technical purposes are allowed and should be defined by SOPs. A case of PK reason that may also be acceptable is apparent sample switching: a timepoint in a placebo profile is positive for drug presence and the corresponding timepoint in an actively dosed subject (blood drawn at the same time) is negative (when the surrounding timepoints for that subject are positive. Even in this case the re is some concern that other samples may have been switched, but the switch would not be apparent from the results (switch between 2 dosed subjects.) And I agree that it would be advisable to perform the PK analysis both with and without the switch or reassay.
Tom
Thomas L. Tarnowski, Ph.D.
Bioanalytical Development
Elan Pharmaceuticals, Inc.
800 Gateway Boulevard
South San Francisco, CA 94080
thomas.tarnowski.-at-.elan.com
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 "How to determine outliers on a PK curve" as the subject | Support PharmPK by using the |
Copyright 1995-2011 David W. A. Bourne (david@boomer.org)