Analyte in spiked depleted matrix vs patient sample

[JulieRay]

When I spike my analyte into a double charcoal stripped matrix and analyse by LC MSMS I can see the peaks but when I spike analyte into real patient sample and analyse by LC MSMS I see no peaks.
Sample prep for both both depleted as well as patient matrix: Protein crash, SPE, derivatization and then LC MSMS.
I have performed ion suppression studies and there is no suppression at the retention time of the analyte. I have analysed washes from SPE and there are no peaks for the analytes.
Overall recovery is poor during SPE (which I am working on) but why should spiked patient samples show no peaks at all? Effect of sample prep should be felt similarly by both matrices.
Please help!

[Don_Hilton]

Any chance that there is something lost in the charcoal stripping that binds the analyte in samples that have not been charcoal stripped?

[michael.williams]

Yes tricky one this !! charcoal stripping can remove a lot of compounds other than what you're looking for which I guess is endogenous hence the stripped matrix.

Some pitfalls when using stripped matrix to be aware of include
Enzymes which degrade analyte of interest may be removed/deactivated by charcoal stripping. If you know what they are you can check it out using the appropriate enzyme inhibitors or post spiking the PPT extract.
Charcoal stripping tends to pull out similar compounds including ones that are likely to interfere so you'll need to look at selectivity with care try diluting sample matrix with endogenous levels. The same applies for stability.
Compounds removed in stripping may have a significant effect on recovery/stability.

Great so how can you tell well...

best thing is to prespike with analyte in stripped/nonstripped matrix then post spike into a separate blank of each and derivatise and detect/quantify, derivatise in a relevant control solvent as well then you can assess matrix effects and recovery. Do this at each stage of extraction to pinpoint where the problem is and you can go from there.

Good luck

[JulieRay]

Thanks for your answers!
Julie

[carls]

This is a very common problem in bioanalytical, i.e. using a different matrix to develop and validate the assay. As has been pointed out, using charcoal stripped matrix when the patient samples will not be stripped is a bad idea. The problem is bad enough when the matrices are matched as closely as possible. Many use pooled patient samples from a previous study to reduce the likelihood of differences between the matrix used for validation and the patient samples but problems are seen even with this approach. There are many approaches for dealing with enzymatic degradation of analytes documented in the literature.

Bottom line: Use matrix as similar as possible to the patient samples to develop your assay. If the patient samples wont be stripped then dont use stripped matrix for development.