Bioequivalence and Bioavailability Forum

Dear Najat!

❝ 1. […] integration mode changes automatically within the same run […]

❝ in a haphazard way for both the analyte and the internal standard without any external modification.

That’s quite common.

❝ Would appreciate why this is happening...

Numerical integration depends not only on the parameters you are able to set in the chromatographic data system (e.g., retention time window, RT shift, threshold, peak width, preferred integration method,…), but also on internal parameters of the algorithm itself (how many data points are used in upslope/downslope of the peak, whether polynoms or splines are used, slice bundling, etc. – ask the manufacturer; but don’ be disapointed if they don’t come up with anything usable).
In an actual chromatogram (depending on both external parameters and internal algorithms) the data system may ‘come to the conclusion’ (haha!), that it may be better to change from one integration method to another one. That’s the reason one should always inspect all chromatograms and go with a manual integration if the automatic one obviously failed.

❝ … and would it affect the integrity of the generated data from regulatory point of view (both European and US-FDA)?

For the FDA you must report the reintegrated chromatograms; EMEA’s BE-draft is a little bit strange (see here).

❝ 2. Part of the LC-MS/MS analysis involves performing system suitability at the beginning; the analyst modified the chromatograms to get a better base-line for the internal standard. Calculations of the CV before modification fails the acceptance criteria while those after modification pass the acceptance criteria.

Fine.

❝ Kindly advice how is this looked into from regulatory point of view? Does this invalidate the whole run even if the standard curve and QCs pass?

Oops! The idea of a system suitability test is to check whether the entire equipment is able to perform as intended before actual matrix samples are injected. No calibration and QCs at this time…
I’m not a regulator, but I would be very reluctant to accept such a procedure – personally I would either reject the run, or have a deep look in the system suitability SOP and the training of analytical staff…