Bioequivalence and Bioavailability Forum

Dear Nilima!

❝ Dear Sir

... and what about the madams in the forum...

❝ QC samples in duplicate at 3 concentrations ( LQC, MQC and HQC ) should be

❝ arranged in each assay run. What is meant by run, all samples

❝ analysed in a day using same Calibration curve,

Almost. A synonym for 'run' in this context is 'batch' - which must not be performed within one day, but your guess of 'using same CC' is right.

❝ secondly what is meant by duplicate ( 2XLQC, 2XMQC and 2XHQC - duplicate

❝ or replicate...?)

Replicate analysis means separate sample preparation up to the final detection - as opposed to e.g., replicate injections of the same extract into an LC-system.
Every sample which is prepared more than once (singlet), is called a replicate (2x - duplicate, 3x - triplicate,...)

Dear Nilima,

Let's start with the easiest part. "In duplicate" indeed means 2xLQC, 2xMQC and 2xHQC. For each QC level you should process two individual samples in the same way you treat your subject samples. I mean this is not just a duplicate injection of a single QC sample, extracted once. Remember also that you may need to have more than 2 replicates per level depending on the number of samples you analyse in your run, duplicate is the minimum value set up by the guideline.

Now the more difficult question: what is a run ? There is no definition in the guideline. One interpretation is indeed one set of samples analysed under the same conditions, with the same reagents, mobile phase etc., and using the same calibration curve. The problem starts when all samples are not prepared at the same time. Let's say for instance that you have 2 subjects in one run, in the morning you extract the calibration samples, the samples of one subject and some QCs, and later in the day (possibly by a second person) the samples from a second subject and some QCs. All are analysed on the same instrument without any interruption. On a chromatographic point of view this would be a single run. But it includes two "batches" (to use a GMP wording) of samples, prepared separately and possibly under different conditions (not the same analyst, possibly different room temp, even worse different batches of buffers or solvent, and possibly with mistakes made or problems happening for one batch and not for the other). I would consider it as a must to have some QC samples in each of these "batches". But then how many QCs should be analysed in each "batch", and how should they be interpreted ? This could lead to serious discussions.

To avoid such problems I always prefer to see "runs" comprised of a single "batch" of samples, if allowed by stability considerations, and even if this means having shorter runs. Which is not feasible anyway with manual SPE on a manifold that will host a maximum of 20 columns at a time

But this is just my own opinion and I would really be interrested in the group's.

Regards
Ohlbe