On Internal Standard

[ghie malig]

Hi all!

Is there a rule when to spike the internal standard in a sample in a bioanalytical method? Is it always before extraction of the biological matrix or can we add it after extraction? Do we always have to account the recovery of the internal standard? Please enlighten me. Thanks!

-ghie malig-

[DR]

Generally, an internal standard is added before any extraction steps. It is not a bad idea to periodically check your internal standard recovery, just to see how consistent and efficient your extraction procedures are, but it plays no part in the calculated assay results.

[Mark Tracy]

This subject has been discussed before on this forum. You can find many illuminating comments by searching the forum for that topic.

In the terms used by environmental analysts, "internal standard" is added to the sample immediately prior to the determination (e.g. HPLC) to correct for unstable instrument calibration. A more general definition is that it is added prior the the stage of the method where the uncontrolled variability is introduced. The term "surrogate" refers to a standard added at the beginning of the sample preparation to check for adequate recovery.

Calibration using internal standards tends to result in higher %RSD and higher MDL, and is rarely beneficial for well executed HPLC/UV methods. (One such case would be when you are doing manual, partial-loop injections. The better fix is to use full-loop injection, or buy a good autosampler.)

Using surrogate recovery to correct the results should be justified by statistical analysis. The assumption that the surrogate is recovered the same as the analyte must be tested.

[ghie malig]

This is actually my real problem. I'm using Glibenclamide as internal standard/ external standard for determination of gabapentin in plasma after precolumn derivatization with OPA-MPA. When I spike the Glibenclamide in plasma, I'm getting poor recovery of gliben because of protein binding with the plasma. When I add gliben after protein precipitation I'm getting a bigger peak (ofcourse). I'm not quite so sure if external standard quantitation is suitable for this method. Gabapentin is not protein bound and gliben is protein bound. I'm not so sore either if this is a good practice for bioanalytical methods. I've found some journals that use external standard instead of internal standard though. But they have justifications of using external standard instead of internal standard.Pls. enlighten me. Thanks

-ghie-

[HW Mueller]

How do you know that gabapentin is not protein bound?
Also, I can´t think of anything "protein bound" that can not be unbound.

[Uwe Neue]

I agree with Hans. Your sample preparation protocol should be able eliminate protein binding of your internal standard.

[Mark Tracy]

Is there some good reason for needing an internal standard? If you use external calibration with gabapentin, do you have excess variability in your standard curve? Variable spike recoveries? Low recovery?

[Uwe Neue]

Ladies and gentlemen,

May I recommend to read Matuchevski, Constanzer, and Chavez-Eng in Anal. Chem. 75 (2003), 3019-3030. It covers in great detail what one should do to uncover problems with a bioanalytical method.

As far as I am concerned, this is the best paper describing the problems and the solutions to the problems.

Uwe

[ghie malig]

Dear all,

Gabapentin is not totally unbound from protein but a lot of it is unbound. We use direct precipitation using 20% Perchloric acid. We tried spiking Glibenclamide before precipitation as internal standard and after precipitation as external standard. We spiked the same concentration of gliben for both preparations. But we are getting different results. Higher for external standard. All because we cannot recover a lot of gliben from its being protein bound. We ought to use it as external standard instead. We tried to make a calibration curve and we are getting 0.996 to 0.999 correlation value by using it as external standard. But our manager does not want us to use external standard method because of doubt that it is not a usual practice in bioanalytical procedures. So we have to start again from scratch and look for another internal standard if ever. I'm asking for opinions if this practice is valid for bioanalytical procedures. Thanks.

-ghie-

[Uwe Neue]

OK. Let us summarize what you just told us.

Gababentin, your analyte, is protein bound and the recovery is not perfect. You tried to eliminate the problem by using glibenclamide as the internal standard. This did not work well, because the glibenclamide is even more protein bound than your gabapentin. Now you ask, since this did not work, why can't you ignore the fact that gabapentin in protein bound by just using the glibenclamide as the internal standard AFTER the protein precipitation.

I am wondering where the logic is in this. The purpose of the analysis is to accurately determine the amount of gabapentin in the plasma sample, and not after protein precipitation. Am I right?

You need to find a method that elimates the loss of analyte (and internal standard) in the sample preparation method. The solution to this is to improve the sample preparation method. There are many choices available to do that.

To improve the recovery of your analyte and your standard without more sophisticated tools, I would first attempt to combine the perchloric acid precipitation with the addition of acetonitrile (3 units of MeCN per 1 unit of plasma) for an acetonitrile precipitation. If this does not get you there, your next approach should be solid-phase extraction method. This is a bit more work, but you get first class results if you execute it properly.

[ghie malig]

We are not having problems with the recovery of gabapentin. we are having problems with the recovery of the internal standard. since we are having problems with the recovery of the internal standard glibenclamide we decided to use it as external standard. i'm asking if this would be appropriate for bioanalytical methods. thanks.

[HW Mueller]

Again, there is, most likely, a relatively easy method to get a good recovery of both compounds, Uwe suggested some possibilities.
Also, recovering a spike of an unknown doesn´t necessarily mean that you are recovering your unknown. Some certainty would be imparted if the unknown does not increase upon experimenting with stronger methods to separate protein and analytes. In other words, if you recover the gibendamide you should have quite a robust recovery of the gaba...

Other present chains have given some opinions on IS.

[Mark Tracy]

Another thing I found confusing is your use of the term "external standard" appears different that most of us would use that phrase. The usual meaning of external standardization is your series of gabapentin standards, not ratio'd to anything else. Internal standardization is the addition of glibenclamide at any stage of the sample treatment, and using that to correct the response of gabapentin. When to add the IS is a matter of experimental design.

There are several artificial amino acids that have been used for IS in amino acid analysis of biological samples: 2-aminoadipic acid, norvaline, norleucine, and 2-amino-3-guanidinopropionic acid. Perhaps one of these would be better behaved in your system.

I agree with Uwe, you need to solve the recovery problem with good chemistry.

[Uwe Neue]

Please obtain the paper that I mentioned! It has all the answers that you need. It covers LC bioanalytical methods in general and its applicability is not specific to LC/MS, although this is where the focus is.

[ghie malig]

Mr. Tracy, some journals would refer to the method of addition of another compound after sample treatment as external standard. there were papers i read that after going through sample prep of their major analyte, they added the IS after it and used the term "as external standard." They analyze the compounds together and still use peak area ratio. In my case, i still believe that it is right to use "internal standard," though there are also some confusions happening here in our lab about the "term." And reading through those journals even make me believe that "addition of IS after sample preparation" would refer to "external standard." You said the internal standard is added at any point of sample preparation but do we have writings to show this statement? I will be able to convince our superiors if we have papers to show that at any point of sample preparation, internal standard can be added. The journals i got demonstrated that addition of compound after sample preparation would call that compound "as external standard."

To quote:
From Journal of Chrom. B
Author: Amita Joshi
Title: Simultaneous Quantitation of an anti inflammatory compound and a potential metabolite in human plasma and urine by hplc

"After centrifugation at 1800g... and dried under N2 at room temperature, the residue was reconstituted with 250 uL of acetonitrile containing 12 ng/mL of anthracene as the external standard."


Author: Reeuwijk et al.
Development and validation of bioanalytical assay for e-5-(2-bromovinyl0-2-deoxyuridine
After SPE...
"After drying the column under vacum, the analyte was eluted with 0.6ml methanol. the methanol fraction was evaporated to dryness at 30 degrees and the residue was dissolved in 0.1 mL water which contained the external standard 5-fluorouracil.

From Journal of Chrom.
Author: T.K. Yeh et al.
HPLC determination of pentamidine in plasma

From discussion part:

"However, melphalan was readily eluted from the SPE-C8 extraction column by the preconditioning methanol washes and could not be used as internal standard. On the other hand, the high and reproducible extraction yield of pentamidine suggested that an external standard, could be used in the analysis. Merphalan was used as and external standard and was added between the clean up and the elution steps."

Thanks.
-ghie-