Bioequivalence and Bioavailability Forum

Dear nvs!

❝ Cases:

❝ (For example if we take MQC)

❝ 1. Both QCs are not meeting the acceptance criteria.

❝ 2. Both QCs (MQC-1, MQC-2) has unacceptable internal standard response.

❝ 3. One QC has not meeting the acceptance criteria & remaining one has unacceptable internal standard response.

❝ In case: 1 we considered that batch failure so we are doing repeat analysis for that batch.

❝ In case of case: 2 & 3 whether we have to repeat the batch or not is there any scientific justification for that.

Cases 2 & 3 are not different from case 1. You will have to repeat the batch, because it is not valid.
See also Ohlbe’s post.

If you are using an internal standard (IS), your calibration is based on the ratio of detector responses of analyte(s) and internal standard (peak areas or heights).
The reason for using an IS is to be protected against errors in volume transfers in the sample preparation steps.
Just an example: we have a concentration of 100 ng/mL, both of the analyte and the IS in our sample (ratio 1). The method calls for the transfer of 500 µL. The detector response for this sample is e.g., 10000 arbitrary units for the analyte and 9000 for the IS.
If we transfer 500 µL responses are 10000 and 9000 (ratio 1.1111).
If we transfer 350 µL only (by chance, erroneously, whatsoever) responses will be 7000 and 6300 (ratio 1.1111).
So we can make any error in volumes getting still the same ratio of analyte and IS. Fine.

Let’s have a look at your case 2. How do you define ‘unacceptable internal standard response’?
We must assume, that the preparation of the calibrators and the QCs were done correctly; only in such a case they will serve their job.
Getting (to stay with the previous example) an analyte’s response of 10000 and IS response of 7000 (ratio 1.429). Calculated concentration of the analyte will be 129 ng/mL and we are dead. Now we get the impression that something went wrong in adding the IS. This might be true, because the area of the analyte is as expected. But our method employs an IS and is not validated for constant volume transfer (ES).
We cannot judge incorrect IS addition (causing the low response) by the comparison with a correct analyte’s response (this is a circular argument).
So there’s no way to get out of this dilemma – unless you videotaped the IS standard addition documenting an mistake by the analyst.

Case 2 ~ Case 3!

❝ I would like to know various references/websites regarding this.

You already know the commonly applied guidelines.
I attended the ‘Arlington Conference’ in 1990, where these rules have their origin in.*
And I must confess, although nothing is written about your particular examples, my answer might give you an idea of the spirit of validation.

---

• Shah VP, Midha KK, Dighe S, McGilveray IJ, Skelly JP, Yacobi A, Layloff T, Viswanathan CT, Cook, CE, McDowall RD, Pittman KA, and S Spector
  Analytical methods validation: Bioavailability, bioequivalence and pharmacokinetic studies
  Int J Pharm 82, 1–7 (1992)

Dear nvs

did you consider to do Re-injection for those samples?
did you had the same problem?
some times silly things can do major issues, such as the sample wasnt vortexed mixed well.
do you prepare one sample from each QC concentration levels for the same batch?
is there a hurm in preparing two samples from each QC?

---

Edit: Full quote removed. Please see this post! [Jaime]