can I use a different particle size col for a validated meth

[Yue Ma]

We are doing method transfer for a validated HPLC method. In the tech transfer document, a Nucleosil SA 10um column from Macherey-Nagel was used. But I have difficulty to purchase the 10um column, can I use 5um particle size but same dimension column? What kind of validation work and regulation paper work I have to do?

[Uwe Neue]

If you sned me an e-mail or give me your e-mail, I can send you a paper that we submitted last year to the USP. It outlines an approach to use columns with different particle sizes and how to scale a separation.

[krickos]

Hi

Well first, USP/EP/JP are recognized standards however they are not the ones who audit or regulatory approve.

If the changes you intend to do is whithin allowed in the pretty much harmonised allowed changes in USP/EP with regard to chromatographic procedures you might be able to avoid revaildation and more focus on equivalence, if outside allowed changes you may face troubles if you do not revalidate.

Some examples of changes:

-company specific method exchanges for a pharmacopiea method. Typically only requires equivalance data, tend to happen when a new monograph appear and company is forced to comply with monograph.

-company specific for new company specific. Full validation data and equivalance reuired on top on justifaction to why change is needed.

-Change outside outside whats permited in pharmacopiea method. Typically requires justification, full validation data and equivalence data.

Equivalance data typically requires data from three different production batches/lots from both procedures including equivalent or better discussion around specifity, sensitivity, precision or accuracy.

[Consumer Products Guy]

USP <621> was updated about a year ago and allows more leeway now on changes from validated procedures. For example, going to a smaller particle size of same packing is fine (going larger is not). They realized that their +/- 25% for column diameter for HPLC didn't get one to the next commercially availablediameter, so that was updated to making the linear velocity the same.

In other words, I think you can change to 5u or 3u without doing any more than updating your procedure, and stating why. You could write a short memo to your cGMP file as well. If your QA department insists, assay a few sample using columns of both dimensions, and document that %RSD is less than 3% (same as for typical Intermediate Precision).

USP also has a separate document somewhere on its web page about what modifications can be made, actually is not as strict as some folks seem to think. Think about it, a single supplier might sell a dozen or more packings all designated as "L1", and obviously they won't all behave the same, but all meet the USP designation...

[unmgvar]

these days,

if it is the same vendor with the same silica chemistry you can do as follow
use the same Dp with a longer column
use a smaller Dp as long as you have the same or better resolution and have checked linear velocity in order to compensate in case of shorter column or smaller ID lenght when suitable

if you move to another vendor you will need to do some validation by USP,
How much is an interesting question
if you do move to another L1 column check, check surface coverage, check pore size, check carbon loading, and also the capping solution if possible. these days most column are type II silica, but moving from a type I to a II also holds differences of selectivity even for the same vendor.
it will all make a difference in the Retention time your compounds will elute

[Rob Burgess]

Here is a scenario to consider here:

You move to a smaller particle column i.e. 5 to 3.5 µm and you start getting better resolution and notice a new peak on the tail (or front). What then?

In my mind it always relates back to the resolution you see (unless it is a simple one peak assay of course) - if it changes (then data reporting results may change) and thus re-valiadtion may be required.

[Uwe Neue]

With a properly scaled method - as outlined in the stimuli article in the Pharmacopeial Forum - you get exactly the same resolution using different column length and particle sizes, but at a constant ratio of column length to particle size and properly adjusted flow rate. Same resolution = same results. Same results = no extra peaks. No extra peaks = nothing to worry about...

One will of course need to explain to regulators that a change results in no change

[pph31]

If you sned me an e-mail or give me your e-mail, I can send you a paper that we submitted last year to the USP. It outlines an approach to use columns with different particle sizes and how to scale a separation.

can you share the paper with us?

pph31@yahoo.com

[Uwe Neue]

pph31: done
I have aslo drafted a summary for the on-line lab magazine Separation Science, which will probably be published in the November issue.

Also, my colleagues from the USP have recent written about the same in one of the lab magazines for the pharmaceutical industry, but I did not keep a copy and I don't recall in which one.