Bioequivalence and Bioavailability Forum

❝ At our facility we faced an odd case where some pre-dose samples had an interfering peak of >5% of the Cmax within it's respective analytical run.

Dear Shireen,

If this was a multi-period study (i.e. not a parallel) then it would be important to know if these positive pre-dose levels occured in period 2 (or later, if applicable). If so then it would indicate a lack of washout time. This might not per se be a problem, as regulators seem to be softening a little bit in this regard. Check your period effect. Find out: If the phenomenon was observed in period 1, are you then absolutely sure the subjects did not ingest the active substance before the study? If the phenomenon occured in period 2 (3/4/5...) could it just be a lack of washout time? Are you sure your assay was as good as it shoudl be? Were some samples swapped in error (or labeled wrongly, or ...) during the analysis?

EM.

Dear Shireen,

❝ Question # 2

❝ Doesn't the above mentioned case indicate a problem in the method, or in the analysis of that specific drug itself?

If the interference is more than 5 % of Cmax, and less than LLOQ, it means that your LLOQ is more than 5 % of Cmax. Isn't this a first problem ? Will you be able to follow your concentrations long enough ?

You will be in trouble in future if this sentence in the draft revision of the EU guideline remains as such in the final version:
If there are any subjects for whom the pre-dose concentration is greater than 5 percent of the Cmax value for the subject in that period, the statistical analysis should be repeated with those subjects excluded. Results from both analyses should be presented, but the analysis with the subjects excluded should be considered as primary.

The text uses the word "concentration", not "signal", or "interference". Which means that you need to calculate a concentration. Meaning that it should be higher than LLOQ ! Which means in turn that your LLOQ should not be higher than 5 % of Cmax.

❝ "Any interference less that the LLOQ detected in the zero standard sample of the volunteers should not be repeated even if is more than the 5% of the Cmax of the phase, since the value obtained is not accurate and precise":

You may still wish to repeat these samples to check whether you still have a signal, and that it is not due simply to a contamination or carry-over. This may be part of your investigations to try and understand why you have these pre-dose concentrations higher than 5 % of Cmax (especially if you have them in the first period !).

Regards
Ohlbe

Dear Shireen

please note that a pre-dose level (prior first dosing) may indicate an intrinsic level of the substance to be investigated. The assay to asses the concentration in the plasma may not be able to distinguish between the endogenous substance and the administered substance. In this case, not adjusting for an intrinsic level may lead to a biased assessment of the terminal half-life/terminal elimination rate. In fact, terminal half-life increases as your last sampling time point increases.

hope this helps

Martin

PS.: A very simple adjustment (from a personal point of view) is to subtract the pre-infusion value from all contractions assessment post study drug administration. If a concentration level at a specific time point post study drug administration is lower or equal than the corresponding pre-infusion level, this concentration level and all subsequent concentration levels should not be used for calculation of at least terminal half-life / terminal elimination rate.

dear shireen !

you may find this article of interest:

C. Dansirikul, H. E. Silber and M. O. Karlsson (2008). Approaches to handling pharmacodynamic baseline responsess. Journal of Pharmacokinetics and Pharmacodynamics, 35(3): 269-283.

hope this helps

martin