UPLC Transfer

[shaun78]

I would like to bring this up again. I did a quick forum search that returned a few threads from 2006 and earlier. That's a long time ago and am curious to know people's thoughts now; especially in light of recent USP stimuli, revisions, and acceptance of a mobile phase linearity velocity scaling (though I am not pleased with the current accepted method).

Basically, this really only applies to everyone in a regulated industry.

What I am after here is what do we feel needs to be done to transfer a method from a HPLC to a UPLC, given the current draft revision of USP <621>, where linearity velocity takes into account column particle size. I'll defend that any day to an FDA auditor, no matter if the USP ends up accepting it or not. We must remember that the USP is a guideline to be followed, and that we do not have to follow the USP to the letter as long as we can scientifically justify our actions.

Having said that, what is necessary for demonstrating a transfer? I would say linearity, LOQ/LOD, and specificity due to different detector types and the general acceptance that not all C18 columns, for example, are created equal.

That amount of work would be only if you wanted to play it safe, which I would assume most QA departments would request.

But really, what are your thoughts? I'm really after plain and simple benchmarking here...

[shaun78]

Yikes...

The lack of responces here make this even more scary than I previously imagined.

Well, for other people looking for the same information I am. I've gotten QA and RA agrement to proceed as I discussed below. Linearity, verification of LOD/LOQ, and specificity to give us a warm fuzzy feeling inside.

We will see how it goes...

[krickos]

Hi

Well essentially/identically this is very similar to more narrow bore columns than <621> currently accepts. Meaning that you need some additional validation, your approach is sound, but in the end it is all about convincing regulators that your proposed procedure at minimum is equivalent and preferbly better (specificity, LOQ perhaps).

From a global perspective US is only one market which has it ups and downs, but if you sit for instance with an older EU product, you may face a serious amount of country specific market approvals (at least until legislation is changed). Which basiclly from a company perspective means that not only do you need to show that is better/equivalent but also that you save money in long run (ie payback for regulatory activity).

A downside with going outside variations allowed in <621> for a monograph is that you have to maintain the addtional validation over time compared to a monograph procedure.

Btw should add that I have not read the latest stimuli article/s.