More on IS response variations [Regulatives / Guidelines]

posted by Ohlbe – France, 2019-11-12 19:14 (1588 d 16:54 ago) – Posting: # 20794
Views: 3,977

Dear ElMaestro,

❝ Different stocks, different storages, different equipment, pipetting errors, robot programming, different this, different that. And you get some info by also looking at the slope of the cal. curve in those cases.


Sure. This paper seems to have some interesting discussions on slope variations, unfortunately I don't have access to the full paper.

❝ But still, let us discuss how to establish the presence of the phenomenon before addressing causes? Let us say the average IS of cals+QCs in one run is 8x the average IS of cals+QC's in another run and all passing the usual criteria.

❝ Would you be concerned?

❝ Yes? Why?

❝ No? Why not?

❝ Not enough info? Which questions would you ask? (I would e.g. consider matrix sources used for QC's+cals, pipetting, volumes, station, method settings, columns, possibly grid anomalies if I can get such info, but all that will usually leave a blank)


First I would compare the slopes of the two runs. If the slope is comparable: response is seriously decreased in one run, which may make it difficult to still pass the LLOQ.

If you find the same variation in the slope as in the IS response, and there is no IS response difference between CC / QC samples on the one hand and subject samples on the other hand, and assuming they used the same frozen CC samples in both runs: I would check the preparation of the IS working solution and pipetting, mobile phase preparation, HPLC and MS settings, retention times differences between the two runs, peak shape, anything that may result in (or be an indicator of) differences in matrix effects. If the IS is not a stable isotope: I would also check anything which may result in differences in recovery (solvents, SPE column batch...).

I would also look at the ISR data, if available. If there is a high systematic bias in the samples analysed in one of these 2 runs, there is something seriously wrong in your Kingdom, even if the QCs passed.

Regards
Ohlbe

Complete thread:

UA Flag
Activity
 Admin contact
22,940 posts in 4,812 threads, 1,639 registered users;
37 visitors (0 registered, 37 guests [including 4 identified bots]).
Forum time: 12:08 CET (Europe/Vienna)

Those people who think they know everything
are a great annoyance to those of us who do.    Isaac Asimov

The Bioequivalence and Bioavailability Forum is hosted by
BEBAC Ing. Helmut Schütz
HTML5