Interferance from plasma at RT of analyte

[prashant.pqa]

Dear all
Seeking your valuable guidance for the development of bioanalytical method.
I am developing the bioanalytical method for glimepiride. As the drug Rt is 10.5min with developed mobile phase.
I tried the solvent for liquid-liquid extraction for extraction of drug from plasma. I tried with TBME, ethyl acetate, Combination of TBME and DCM(7:3) also with ethyl acetate and TBME(7:3).
Out of that i selected TBME:DCM(7:3). In that i observed one late eluting peak is coming to next injection and interfering with the analyte also one more co-eluting peak is also coming on the same RT of drug.
Published methods are by using diethyl ether. Is TBME and diethyl ether will make that much difference?
Also i tried with protein precipitation by perchloric acid, centrifugation followed by evaporation and reconstitution with diluent (ACN:Water 1:1). But here i am getting so much white residue after evaporation. It is not possible to inject. Also i tried protein precipitation using ethanol:ACN(1:1) followed by centrifugation and evaporation and reconstitution but iam getting one interference on analyte RT.
Glimepiride pKa 6.3 . I tried with by acidifying plasma with perchloric acid followed by liquid-liquid extraction but it is also not working the interference is there.
I optimized the rinsing solvent also so it is not carry over from syringe also.
Kindly help me out

Thanking you

[JGK]

Have you tried deproteination with ZnCL2 solution.

ZnCl2 Solution:
Dissolve 30 g of Zinc Chloride and mix for approximately 5 minutes (some crystals will remain). Vortex vigorously to dissolve the salt. The solution will become very hot. Store at room temperature and discard after 1 month. Mix well prior to every use to ensure that the reagent is homogeneous. 10 µL of solution added to a 75 - 100 µL plasma.

How bad is the interference (guidelines allow for < 20% of LLOQ peak response)?

for the late eluting peak can you alter the injection cycle to "move" the late eluting peak away from the analyte in subsequent injections.

[Victor]

I do not want to hijack this thread, but I would like to extend this discussion. I am sure it is possible for some drugs to become adsorbed on proteins in the sample. If the protein is precipitated by some of the methods proposed above, what happens to the adsorbed drug? Is it also precipitated and therefore lost from the analysis? I am not suggesting this happens in the particular problem discussed here, but what methods can be used to overcome such a problem? Is it possible to use proteolytic enzymes, or could the drug then just adsorb to the peptide fragments?

[Kostas Petritis]

Once you cleave your protein to peptides they loose the tetriary structure and thus do not bind with small molecules any more (with some very few exception I guess). Most peptides cleaved with trypsin for example will probably have an a-helix configuration only... Denaturation of the proteins through other means (i.e. organic solvents or urea etc) can also result in unbinding of your molecules but for sure cleaving them should work even better (with the downside that you can not get rid of it any more and you will have the peptides all over the place in your chromatogram...)

[Bryan Evans]

You could use a direct inject column (such as Cadenza HS-C18)
to study drug plasma protein binding.

[Uwe Neue]

The best approach to get rid of interferences is solid-phase extraction. You can finetune your method to the analyte.

As a first approach, I would acidify the plasma sample (perchloric acid is fine). Add internal standard. Load this prepared sample onto an Oasis HLB or MCX cartridge (from Waters), after the cartridge has been activated with methanol and water. Then wash with acidified water with 5% methanol. Then elute the analyte with a higher concentration of methanol. As a first cut, I would elute with 100 % methanol to make sure that I recovered everything.

It is not clear to me, which packing will work better, the HLB or the MCX packing. I think that you got a better shot at removing interferences with the MCX, but the HLB may give better retention for your analyte.

More details on this procedure or similar procedures can be found on the Waters webside.

[prashant.pqa]

Thank you for your valuable suggestion . Today i will try out this and let you the status. The interference area is 10 times more than LLOQ. I think ZnCl2 conc. will be 30g in 100ml of water.
As i am using C18(250 X 4.6 mm, 5u ) column. As i am adding ZnCl2 is it make any problem to column?

Have you tried deproteination with ZnCL2 solution.

ZnCl2 Solution:
Dissolve 30 g of Zinc Chloride and mix for approximately 5 minutes (some crystals will remain). Vortex vigorously to dissolve the salt. The solution will become very hot. Store at room temperature and discard after 1 month. Mix well prior to every use to ensure that the reagent is homogeneous. 10 µL of solution added to a 75 - 100 µL plasma.

How bad is the interference (guidelines allow for < 20% of LLOQ peak response)?

for the late eluting peak can you alter the injection cycle to "move" the late eluting peak away from the analyte in subsequent injections.

[HW Mueller]

Another one of these recurring discussions. If you decide on protein precipittion than you better make sure that the analytes are more soluble in the liquid than in the protein. This holds for ultrafiltration as well. In the case of solid extraction, as suggested by Uwe, everybody does not seem to have a similar problem in recognizing that the analyte better stick more strongly to the stationary phase than to the protein.
There is another aspect of such anayses: Often, literature versions have been worked out for reasonably high analyte concentrations. If one has to go far below that, especially if one has to put in a concentration step, a 3rd chromatography step (or other separation technique) might have to be implemented.
Getting rid of the proteins via enzymatic reaction is an interesting idea, albeit I think it will create a horrendous mess.

[AdrianF]

There are several references out there to the assay of glimperide.eg
http://cat.inist.fr/?aModele=afficheN&cpsidt=16467109

It is generally much better to use /modify an established method rather than reinventing the wheel.

[Victor]

There seems to be some disagreement here. Kostas Petritis seems to suggest that denaturing proteins and precipitating them with the usual reagents can result in unbinding of drugs adsorbed to proteins, but HW Mueller seems to suggest that this might not work.

Surely this problem is a common one in drug analysis (ie the adsorption of small molecules to proteins), and there is a generally accepted procedure to deal with it?

[AdrianF]

It is essential to use the same medium ie serum/plasma for your standards, so any problem with recovery is compensated for.

As part of the validation it is important to calculate percentage recovery by comparing with standards made up in solvent.

If it is less than about 80% the method is unlikely to be robust

[HW Mueller]

Victor,
Kostas is talking about highly specific interactions, these will likely be terminated by denaturation. I am talking about nonspecific interactions: If you have a substance which is not soluble in your solvent it will cling to anything else around, if there is nothing in solution it will cling to container walls. Example: Cortisol (this substance is said to have specific interaction, probably with albumin, but if you precipitate the proteins you loose a large portion of cortisol unless you add some MeOH prior to the precipitation).

[prashant.pqa]

There are several references out there to the assay of glimperide.eg
http://cat.inist.fr/?aModele=afficheN&cpsidt=16467109

It is generally much better to use /modify an established method rather than reinventing the wheel.

Thank you for providing the wonderful reference.
Now discussion is really good one. My aim si to develop simple method for estimation of Glimepiride using liquid-liquid extraction or protein precipitation which can be used for common labs analysis.
Thanks lot

[prashant.pqa]

Dear All
Thank you for your valuable suggestion and all.
I got some what new findings. If i will process 5 different lots of plasma and inject the interference will start from second injection onwards. There is no any inference in first injection chromatogram. I was suspecting it may be problem with injector so i changed washing from 70%methanol to 80%acetonitrile but in both times i am getting same pattern.
What will be your personal experience for this?
Is it late eluting peak but i tried for that also i gave run time up to 35 min but there was no late eluting peak also.


Kindly provide you guidance.

Thanking you
PRASHANT