Method Equivalency

[wanda50]

Preface: I am in the API pharmaceutical industry and am working on improving an LC method that is used for release testing.

I am substituting one L1 column for another L1, changing manufacturer as well as particle size from 5 to 3.5 um. No other method parameters are changing.

What validation elements would you include in your protocol? We have no SOP that covers this, although it sounds like a good one for me to write. Personally, I think all elements should be done (other than stability related elements), but management gets testy with instrument and Wanda time.

I will do the obvious - accuracy as injection reproducibility, linearity, specificity. should I determine LOD/LOQ? If so, then I obviously will do accuracy as spike recovery. What about robustness? I imagine interlaboratory precision needs to be evaluated since this method is used by Quality group.

I thought I would ask the LC gurus of the world their opinion so I can substantiate the required Wanda/instrument time.

thanks!

[Richard987x]

Wanda- It seems you have folks from two different schools:

1. Those that would focus on the accuracy and precision parameters.
2. Those that would focus on the system suitability test- i.e., same selectivity and sensitivity
3. I forgot to mention, a third group of folks like me that would do (1) and (2) plus a column evaluation study, i.e., (3) column lifetime to ensure robustness of the method and (4) lot to lot reproducibility- using columns manufactured from 3 different lots of silica, same chemistry to verify consistency.

Hope it helps

[wanda50]

I have never done a column lifetime study. I have noticed that our customers are asking for this more often. Can you elucidate?

Thanks

[Richard987x]

Wanda-

Here is an example; goal is to increase your confidence on column selected- operate the HPLC continuously for 1 week, 2 weeks, etc while evaluating chromatographic parameters are consistent. It is difficult to forecast the long term performance of a column (I dont have a crystal ball) but this type of study can mitigate to some extend these uncertainties. The study becomes more critical when operating methods that utilize highly aqueous mobile phases, alkaline pH and complex samples. I dont think there is a guideline out there, but it is justifiable. If column looks good then oK to proceed to validate column per SOP.

Column performance evaluation
Procedure:
• Using DP reference standard, prepare Resolution solution per the method.
• Perform repeated injections of the Resolution test.
Analysis:
• Integrate chromatograms
• For injection No. 1, determine system suitability parameters per the method.
• For repeated injections- evaluate capacity factor, resolution, tailing factor, plate number, and impurity profile (available from Chromatographic database)
Acceptance criteria:
• For initial injection: Must meet method system suitability acceptance criteria as stated in the method (same as above).
• For repeated injections, consistency in chromatography: capacity factor, resolution, tailing factor, plate number, and impurity profile.
Data Reporting
• Report overlay chromatograms, number of injections, and consistency of chromatography.

[danko]

am working on improving an LC method

Your key validation tests should be focused on the performance parameters which prompted the change/optimization that you made. What was unsatisfactory? Resolution? Then you should validate the specificity! Unstable column? So put the new one on trial? And yes, several of them (3 different batches?).
Finally, LOD and LOQ should always be validated when altering the stationary phase, even if the change is “onlyâ€

[wanda50]

I forgot to add column equivalency to my protocol thoughts. I assumed you could read my mind! My plan is sitting right here on my desk, and even plan to throw in quaternary pumps with a PDA vs binary and VWD.

The reason to improve the method was to separate semi-coleuting isomers. When the method was developed, the isomers did not appear in our product. since the product is derivied from plant material which is subject to growing seasons and country of origin, we have been seeing this isomer grow over time, indicating it is a related substance rather than a process impurity. It is time to separate the isomers for better quantitation so that we can track the amounts.

I never meant to imply that I wanted to short cut an equivalency. I am a firm believer in doing everything applicable to avoid customer and regulatory body questions later. I just wanted other opinions to justify the process to my superiors.

thanks for the comments and keep them coming while I write the protocol.

[AdrianF]

Much of this is covered in a previous topic

http://www.sepsci.com/chromforum/viewtopic.php?t=6890

One of the contributors in that thread suggested that an SOP should be written to explain what you are doing when you update a method. If it contains sound scientific reasoning the regulatory authorities should be satisfied!

[wanda50]

thanks for the reference. I was actually part of that discussion which seemed to be more about what to substitute. My question deals with what to validate after the best possible substitutions were found. but, I could have missed something. I have a pounding validation headaches today.