Bioequivalence and Bioavailability Forum

Hi ElMaestro & Compliance!

❝ […] in my opinion the LLOQ in itself is not always king. Often it is, and often companies strive to get it as low as possible but other factors weigh heavily in when it comes to judging if an assay is good enough.

Well said! Remember that one? No need to go ‘as low as possible’ (as ýou stated in your #1).
Methods used for quantitative measurement [of analytes in any given biological matrix] must be

1. reliable and reproducible
   
2. for the intended use.

#1 is the usual stuff (accuracy, precision, selectivity/sensitivity, reproducibility, stability, ), but very important is #2 (working range: LLOQ ≤5% C_max - carry-over, ULOQ ≈C_max^*, AUC_t/AUC_∞ ≥80%).

❝ 3. […] plasma or serum. Traditionally both are generally acceptable, as far as I can tell.

Right. But if ever possible analysts would opt for plasma. There’s a manifold of problems to be expected with serum:

• Requires sufficient time for clotting – no cooling allowed. Analyte must be stable / stabilized.
  
• Potential problems after thawing. Turbidity may require centrifugation – does the lab have a cooled centrifuge?
  
• Sometimes problems become evident only in multiple freeze-thaw-cycles. This is part of validation, but not necessarily of method development. I have seen surprises here.

If I see a method in serum I ask myself (well, and the analyst) why this matrix was chosen. The only excuse would be that none of the available anticoagulants ‘worked’.

Ad #4: Sometimes new columns show bad performance. They need some run-in / conditioning (adsorption of matrix components) to work within specs (resolution, peak shape, k’, R_s, …). Not so few analysts keep wonder-columns in a drawer labeled ‘Only for method X – don’t touch!’.

---

• Never use the mean C_max from literature. Try to set the ULOQ in such a why that you ‘catch’ all subjects’ C_max. Validate dilution with matrix; needed if a value is outside the working range.