HPLC Method Adjustment as per USP

[mattmullaney]

My firm is faced with verifying a HPLC method for which the USP specifies a 3.9 mm x 150 mm C18 column containing 4 micrometer particles. I found that I could buy the exact (per PF) stationary phase with the exact column dimension that USP specifies for $800. I also found that I could buy the exact stationary phase in a 4.6 mm x 150 mm column for $590. Is it reasonable for us to perform a partial validation (specificity, intermediate precision, linearity, and robustness) for the 4.6 mm ID column and demonstrate equivalence to the 3.9 mm ID column? I would plan to scale the flow rate proportionately to the 4.6 mm ID column from the 3.9 mm ID column (from 1 to about 1.4 mL/min) and injection volume (20 uL to about 28 uL). This USP method is isocratic...and if I recall correctly USP permits such an adjustment under isocratic conditions. Please, what are your opinions? My thanks, in advance

[DR]

If it is a stability indicating method, you will need LOD/LOQ as well, and these are what will most likely lead to problems, especially w/ small, later eluting peaks.

[mattmullaney]

Hi DR,

My thanks for your response--looks that this method is intended to be stability-indicating...might as well come out and say that the API is ropivacaine HCl. Procedures 1 and 2 for impurities demonstrates if the ropivacaine HCl is contaminated with bupivacaine and dimethylaniline (along with anything else that may turn up upon forced degradation). So it seems my choice is to stick with the $800, narrower column or switch to the $500, wider column...and demonstrate equivalence; properly adding in the LOQ/LOD. I think I'll stick with the $800 column. I expect the eluent will be antagonistic to the HPLC stationary phase as well...its pH is to be 8.0...probably though it hurts me to do so I'll ask for guard columns for this and see how quickly the SP disappears as this lovely method runs.

Once we have our feet on the ground (meaning selling product)...it may be an idea to modernize this method a bit.

[tom jupille]

if I recall correctly USP permits such an adjustment under isocratic conditions.

It does, with the proviso that you must meet system suitability. As DR pointed out, going to a larger column may hurt LOD/LLOQ, but you are also allowed to adjust injection volume, so that may come out in the wash.

That said, "Columns are cheap. Time is valuable!"

[mattmullaney]

Hi Tom,

My thanks for your response! Quite agreed...the IDs are 3.9 and 4.6 mm for the two columns...flow for the larger ID column would be ca. 1.4 mL/min and the injection volume would move up to 28 uL from 20 uL specified for the 3.9 mm ID column. LOD/LOQ could come out in the wash...but...

To your point...and DR's (I think), so we'll pay an extra $300 per column, why do a verification with the 4.6 mm ID column and confirm equivalence to the 3.9 mm ID one? Makes no sense for investment of time and money in the term this method will be in effect at our firm (I'd presume). We'll pay the extra money for these columns now to save money later by updating the method. [Likely also we'll invest in guard columns (I'm usually against these) as the mobile phase is pH 8.0...the stationary phase is Type A silica and will probably not last too long without a bit of help.]

Eventually, this method will have to be updated for modern instrumentation (and HPLC stationary phase material) and validated...for now, verification of the older-style HPLC method will meet requirements.

[Consumer Products Guy]

I'd say "just document everything". If possible for your needs, though: I'd keep injection volume to 20µl.

[HPLCaddict]

...and my 2 cents:
Even if you exactly stick to the compendial procedure (i.e. in this case, use the 3.9 mm column), you are still supposed to perform a verification. Have a look at USP chapter 1226 "verification of compendial procedures".
"Users of compendial analytical procedure are not required to validate these procedures when first used in their laboratories, but documented evidence of suitability should be established under actual conditions of use".
So, no matter if you decide to use the 4.6 or 3.9 column, I'd strongly recommend to verify LOQ.

[mattmullaney]

Hi Folks,

Agreed, plan was to always perform a verification (specificity, intermediate precision, linearity, and robustness) with the column dimensions of choice...and as DR pointed out, we should add in LOQ as well.

(Also LOD...the same separation method will be used as a limit test as well as to determine related substances by area percent.)

[hossam kamal]

Hi ,
I think using a wider column will change peak details ( width and height ) due to molecules diffusion , and with large injection volume it increases and peaks become broad which change system suitability parameters ( resolution) and sensitivity , this is related substances isocratically method , so I expect resolution is critical , and also sensitivity , in my opinion stick with the claimed column in the USP , surely there a purpose for narrow column in USP monograph , and I am afraid that you purchase the wide column and then method optimization can't work or still have deviation in suitability parameters
About pre column I agree with you completely as it may not produce significant improvement in column stability but on contrarily it's degradation may increase background noise and so decrease sensitivity

[mattmullaney]

Hi Hossam,

Quite agreed. Looks like we'll be in for a bit of a process, with luck, the recommended phase and an eye on the column performance over time with this eluent we'll get back on track. With luck, the guard columns will help and not hinder our progress.

My thanks to everyone for their contributions!