RP-HPLC method validation for biopharmaceuticals

[miro2009]

Hello chromatographers
I have some questions about HPLC method validation generally:

1- 1st a general question, are there specific grades of salts to be used for preparing mobile phases for HPLC methods? eg. NaCl, sod. or pot. phosphate. We generally use ACS reagant grade, are there "HPLC" grades as the case with organic solvents such as methanol?
2- if I change the brand of components used for preparing the mobile phase (for eg. sod. phosphate from Merck instead of Sigma) of a validated method, would I need to revalidate any parameters?
3- for the multiple analyst and multiple day tests, how exactly should they be performed? just preparing the 80, 100, 120% levels by different analysts/days and checking % recovery on a common calibration curve, or also preparing and making the calibration curve by the several analysts/days? and would n=3 be enough in either case?
is there any reference or literature that specifically describes this? I find ICH guidelines very generalized!
4- acceptance critreria for each test are never specifically described in ICH guidelines, does anyone know where to get them? and would they be any different in case of biopharmaceutical analytes?

These are the questions I have for now and hope to share your experiences, looking forward to a fruirful discussion thread.
Thanks!

[lperelman]

1. There are generally very high purity grades of almost all reagents. If price isn't an object, then I recommend using the best grade you can afford.

2. You would want to do at least some quick testing to ensure that pH of mobile phase, retention time, and signal response are not affected by the change.

3. ICH guidelines are purposely vague, as different analytical tests require different levels of rigor to be considered robust (compare, for instance, measurement of solution pH vs. ELSD-based chromatography). You need to have a plan in place to address these issues. As with most ICH tests, if you can demonstrate the concept you are trying to demonstrate, then your test protocol is very likely to be acceptable.

4. Acceptance criteria for limit tests (e.g., residual solvent) are definitely clearly stated. You can find additional information in the USP. For other tests, it is going to depend on your analyte and your detection modality. For instance, you may put error bars of 1% RSD on a UV-based measurement, but 3% on an ELSD-based measurement. In early phase development, the FDA will give you fairly broad latitude in this area, and you should absolutely take advantage of this. Even if a given method has tight specifications, you will want to give yourself wiggle room for future development. As you progress through development towards NDA, the FDA will want to see improvements in your analytical methodology, and you want to leave space for this.

[Consumer Products Guy]

3- for the multiple analyst and multiple day tests, how exactly should they be performed? just preparing the 80, 100, 120% levels by different analysts/days and checking % recovery on a common calibration curve, or also preparing and making the calibration curve by the several analysts/days? and would n=3 be enough in either case?
is there any reference or literature that specifically describes this? I find ICH guidelines very generalized!

For the second analyst/second day/second instrument assays, the second analyst just assays finished product 5 times, and his results must be with 2% RSD, and must match first operator's numbers within 3% RSD. That's our intermediate precision. 80, 100, 120% levels are used for Accuracy testing, and later on for test method transfer. Anyway, that's what our QA department says, and what we do.

[miro2009]

Thanks all for the helpful input.

I have another general question for validation of a quantitative HPLC method for a certain finished lyophilized product; this product has a labelled amount of let's say 10 ug per vial, but actually could have a higher amount as the process is designed that the final stage during formulation and before filling has an overage of 10% to compensate losses later in the process, in addition to losses on patient reconstitution, in other words, the concentration of the liquid just before filteration, filling (1ml/vial) and lyophilization is 11 ug/ml and not 10ug/ml.

During method validation, what concentration should be considered my 100%? 10 or 11 ug/ml?
Are there any guidelines described anywhere regarding this specific issue? please send link.

Thank you

[lperelman]

I am a little confused about your process here. Do you fill 11 ug/mL into the vial based on in-process testing of the product before fill/finish? Or is the 11 ug/mL a theoretical value based on assumed loss through the process before fill/finish? Are you processes validated with a known tolerance for material loss? Is the loss due to residual material in your final fill/finish vials?

I would recommend that batches get tested prior to fill/finish into the vial so that you are accurately filling with a final dosage form. But in any case, if you are knowingly providing a 10% overage, then your targeted dose is 11 ug/mL, and It think this would be 100%. I am hesitant to say that with confidence, as I need to understand your upstream process to provide a meaningful answer. Bottom-line: how much do you know, and when in your process do you know/measure it?

[miro2009]

I am a little confused about your process here. Do you fill 11 ug/mL into the vial based on in-process testing of the product before fill/finish? Or is the 11 ug/mL a theoretical value based on assumed loss through the process before fill/finish? Are you processes validated with a known tolerance for material loss? Is the loss due to residual material in your final fill/finish vials?

I would recommend that batches get tested prior to fill/finish into the vial so that you are accurately filling with a final dosage form. But in any case, if you are knowingly providing a 10% overage, then your targeted dose is 11 ug/mL, and It think this would be 100%. I am hesitant to say that with confidence, as I need to understand your upstream process to provide a meaningful answer. Bottom-line: how much do you know, and when in your process do you know/measure it?

I will try to elaborate more:
In the final stage of the formulation phase, we reach a concentrated "liquid final bulk" of around 40 ug/ml. This concentrated liquid is then further diluted with the form. buffer to a final concentration of 11 ug/ml (not 10ug/ml), all based on HPLC in process analysis.
Then this diluted liquid undergoes a final stage of sterile filteration, then filling (1 ml/vial), then lyophilization and capping. By this stage, some loss of the API is expected. Finally, further loss of the API is expected by the patient during reconstitution of the vial. So the overage is to cover all these losses and the patient finally receives a 10ug/ml dose, which is what is put on the label. I hope this explains it for you.

For me, I'm personally inclined to consider the 10ug/ml as the 100%, as this is the "actual" or expected dose of our API. Also, it's a further challenge to the method to use the lower number, if the method is valid for 10ug/ml, then it should be also valid for a higher concentration. Again, any guidelines about this would be helpful.

So, can you let me know your opinion now?

Thanks again!