proteinLC method for impurities_range and acuracy determintn

[alicee!]

hi all..
how to validate range and accuracy for protein LC methods.. the method is say an RPHPLC method for impurity quantitation.. gradient LC.. estimation at 210 or 215 nm wavelength.
my method quantitates multiple impurities in a single drug product.
these are impurities/related forms like the oxidized, reduced, demidated and similar other forms
are these forms supposedly not very differently absorbing and behaving than the native form?
I donot have values of these impurity % from another validated procedure as the ICH suggests, n i think currently getting this for each form is difficult for me.
let me know of other ways in which range and accuracy for impurities for my method can be qualified.
thanks..

[danko]

Regarding range, you could do the following exercise:
Inject/load decreasing amounts of your degraded control/standard, until you see changes in the values (i.e. increasing purity of the main component or decreasing values for the impurity peaks). This would be the lowest point in the valid range.
Then inject/load increasing amounts of your control/standard, until you see changes in the values in the other direction [i.e. decreasing purity of the main compound or increasing relative (%) values for the impurities]. The latter is due to the overload of the main peak. This would be the highest point in the valid range.

Finally, if you don’t have real values for the degradation products and the impurities, you’ll need to show that another (orthogonal) method of analysis confirms the obtained results. This could be another chromatographic technique, like ion exchange for instance.

By the way, you don’t need the monitor the degradation products at 210 nm. 215 nm is adequate and as you point out yourself, the absorption coefficient at this wavelength doesn’t change just because the protein has been oxidized or deamidated. It’s the peptide bonds that cause the UV absorption at 215 nm and they are assumed intact.

Best Regards

[alicee!]

thanks for the response!

for range and accuracy.. for impurities quantitation.. can i not construct a linear calibration curve using the native form of protein that i have.
i mean i would run a 0.2% to say 4% of native protein.. and then quantitate impurities based on the response obtained when impurity elutes as part of the sample in values near 1%.

i donot have validated orthogonal methods for addressing all impurities.
let me know how else this can be done.

what does ICH mean.. when it says check linearity and range in the region of 80-120% of the specification?

thanks..

[Pradeep Iyer]

hi alicee!!!
according to me you can construct the calibration curve with the native (pure form of protein) after establishing that there is no major difference (in terms of precision of the area values) between the impurity and the pure protein for which you might have to run pure impurity and check. Why I suggest this is because an impurity (say deamidated or oxidised or reduced) will have a changed surface hydrophobicity in an RP method and might show a different elution patters (interms of area along with the retension times). So it becomes important to establ;ish a correlation between the impurity and the pure or native protein to use the calibration curve of the native protein to quantitate impurities...
hope it helped..
cheers!!!

[HW Mueller]

How that?:

"and might show a different elution patters (interms of area along with the retension times)."

alicee, we have been through this many times regarding assumption and knowing.

[Pradeep Iyer]

hi aliccee
it seems the previois post owner knows or thinks that all the patterns and the information of behaviour of impurities should be known beforehand. in tat case we all are jobless!!! lol lol... i guess he s not understood the essense of the topic and the reply in turn..

cheers!!

[alicee!]

hey thanks pradeep..
i think u understood my problem really well!

i m dealing with these protein forms which like u said.. are expected to behave different from the native form.. in terms of elution patterns.
there are multiple forms/peaks that i see on my chromatograms i've got at my method development stage.. n i have no idea of how these different forms may behave!

the only way i see is like u said.. get an isolated impurity.. compare the behavior in terms of response (absorbance may b same at 215 nm but elution pattern would be differnt) and establish conversion factors.

thanks again, it definitely helped!!!

[Pradeep Iyer]

ya you are welcome and i hope that HW Mueller also understood the problem. Its easy pointing out errors.. that too when they do not exist!!!!

[grzesiek]

alicee! could you tell us more about the problem

how many of these different forms are expected?
will the method be used for stability studies?
do you have these forms standards or do you get them from degradation study?
what are the expectations of the client?

[HW Mueller]

Pradeep Iyer, you didn´t tell us how a different elution pattern changes the area.

[Pradeep Iyer]

Hello Mr. (Dr.?) Muller

protein impurities like oxidized, deamidated and specifically reduced forms will be structurally different and have different surface hydrophobicity due to the different exposure of the hydrophobic aminoacid residues. this is also obvious as these impurities elute as separate peaks in a single chromatogram. hence, the different retention times.
the change in surface hydrophobicity would also lead to different elution recovery i.e. loss on column (no complete elution). that is differnt forms are expected to have different %recovery and hence same amount of different forms would give different area values. this different elution pattern is expected and needs to established individually for each different impurity. this may be different for each one of them!
also, for impurities with sharp peaks, the integrable area would be high as compared to those with a broad peak shape. for a broad/wide peak, the edges of the peak may be merging with the baseline and the overall integrable area will be less.
lastly, differences could arise from absorption extinction coefficient of these forms at a particular wavelength. for eg: at 280 nm structural differences would lead to different exposure of aromatic amino acids like tryptophan, phenylalanine etc and hence different absorbance and area. but since we are talking of absorption at 215 nm which is peptide bond absorbance differences resulting from absorption will not be observed in alicee's case

hope that explains!!!

[HW Mueller]

Am I supposed to understand (I am referring to: "and i hope that HW Mueller also understood the problem") that bad recovery is good chromatography?
Or that accepting broad peaks, which can no longer be correctly integrated, is good chromatography?

Also, this is the reason for my pointing out the problem of knowing or assuming: "lastly, differences could arise from absorption extinction coefficient of these forms".

[Pradeep Iyer]

the topic is not what is good chromatography. Please refer!!!
I really think you should read the entire querry and responses as a topic and not as what you think it is!!!
Let us end this unhealthy discussion which is leading us nowhere. I atleat dont have much time to waste!!!
The author seems satisfied and then we should not be wandering around the topic in question but answering it!!!

[HW Mueller]

Ã

[Pradeep Iyer]

there are guidelines which are followed and the drug is out after much introspection so you need not worry about what you swallow unless you get what you read!!!
All the best!!!