Dear all:

We analyse drugs in human plasma samples by SPE-HPLC. Currently we are having a problem. When we analyse freshly prepared spiking samples (STDs and QCs), the data always pass (within 15 or 20%). However, once the samples are frozen overnight, the calibration curve then is no longer good and there are lots of data out of specification. We have no chromatographic problem and it does not seem to be a stability issue. Probably something happens during storage or freezing/thawing. Has anyone experienced such problem? Your input is greatly appreciated.

Without knowing what the drug is its difficult to comment on the chemistry of the problem. It obviously has something to do with the metabolism of the compound in the plasma sample, but without knowing anything else I don't think you'll get much useful feedback.

Initially there was little inclination to enter this chain, as I was already asked this question per phone and already then could not extract relevant informaton for a rational answer. Now I find it worthwhile in order to demonstrate the futility of responding to these "no info" questions: On the phone I was given the impression that the variations included increases as well as decreases, so that an instability seems indeed ruled out. That still leaves sundry other possibilities: The initial mixing of analyte with plasma might not had reached equilibrium, there is an occasional overlap, the SPE is not performed robustly, injection volume and concentration variations are not compensated, etc, etc. No way to know with this info.
Now, the following chain ("Pressure drop") is quite informative as itīs initiator had the courage to admit to a simple mistake.

This is an unsolicited call for slightly more thoughtful contributions which will make these discussions much more useful for all of us.

I am the message originator. All STDs and QCs are treated under the same conditions ( same spiking plasma donor, same freezing/thawing time and temperature). When analysing freshly prepared spiking samples, all data pass and it has been repeated several time, indicating the SPE and HPLC procedure is rugged and mixing is good. As we use ISTD, injection volume is irrelavant (no sensitivity problem). THe variation does include increases and decreases, propably it's not a stability problem. From literature, the drug has no metabolites. The ISTD is added to plasma each time just before SPE (and well mixed).

I guess that it is related to drug/protein binding (dissociation speed). Another possibility is that some drug sticks to denatured protein after thawing. We have seen another similar incident before in the lab (it was a drug metabolite).

Preliminary data show that it is an ISTD stability issue. Further investigation is underway. Many thanks for help.

Thatīs what I mean: more info (especially precise info) and already there is somthing educational there. You may have given us a splendid example for the argument to do without an int. st., or better, as I suggested earlier, using an int. st., but treating it like an ext. st. (identical inj. vol. to get absolute area or hight . . . ).
Still, it would be interesting to know whether you also have a problem of separating analyte (what is it??) from protein. You have to solvate analyte well to get it off proteins.
Also, this chain has shown that intuition is a factor, though highly ambivalent, in chrom., like anonymous 2, I also had the initial thought: something is unstable, but junked the idea due to reason which is, of course, also faulty when based on faulty info.

Hans