Back to the Top
The following message was posted to: PharmPK
I am having problems with the stability of my LCMS system. I prepared
a standard of several basic drugs and a basic drug labeled/unlabeled
pair at about the 200 ng/ml level.
The instrument response tends to drift upward 15-25% over time. The
labeled/unlabeled response factor for xanax/xanax d5 is fairly
constant, the response shows random error, and relative standard
deviation is always less than 2% over a 5 hour period.
If you use the xanax-d5 standard as the internal standard for the other
4 basic drugs, the response factors are varying systematically, the
relative standard deviation is 5-10%, and the responses show systematic
error.
Do other people have these problems?
Does anyone have a test in their laboratory for instrument stability
when you are having problems or you have used to evaluate an
instrument?
Thanks for the help.
Back to the Top
Dear James Little,
unfortunately I have not direct experience about a drift upward over
time of a LC/MS system, but I think this is an interesting feature.
I can only speculate about some possible problems:
1) If you are using a gradient, be sure your system is really in
steady-state mode. I suspect this is not your problem, since when non-
equilibrium happens the more frequent event is the reduction of
retention time.
2) Be sure you have not a front effect: whenever your compounds
elute near chromatographic front the risk is to interfere with
response efficacy of MS detector by interfering with compounds-
specific ionization. It's possible that for different samples an
unwonted not reproducible noise generates a random response. The
simplest way to check this issue is to increase retention times
moving right the compounds peak: you expect to observe more
reproducible peaks.
3) A possible problem, if you use an isocratic chromatography, is
the injection sample composition effect.
4) Possible other problems could come from ionization source: are
ionisation conditions (temperature, voltage, gas flows, etc.) really
stable?
5) Usually electronics instability are well supervised by software
and you don't expect this kind of problems.
Best regards
Stefano
Dr. Stefano Porzio
Pharmacokinetic and Tox. Dept.
Inpharzam Ricerche SA - ZAMBON-GROUP
Taverne - Switzerland
Back to the Top
The following message was posted to: PharmPK
Dear all:
The mass-spec response goes up over the time
phenomenon, we have seen a lot of this problems, if
you have front set of standard curve and back set of
standard curve included in your batch, you will end up
standard curves divergence, batch will fail.
The most of cases, it is matrix effects. The samples
from biological matrix was not "truly clean" when you
inject into mass-spec for analysis. If you have very
short batch like 40 samples, you may not see the
problem, if you have full 96 well plate batches, also
you have hundreds or even more samples to run, over
time the interference from your matrix will "built up"
and leaks out gradutely and cause either your back
curve have more response ( signal enhancement) or back
curve have less response ( signal suppression), either
way, you will end up standards divergence, batch will
fail.
There are many ways to solve this problem:
You can run your plate certain ways to solve the
problem in short term and get same response to get one
batch to pass. But if you have lots of samples to run,
in long term you should consider change/modify your
extraction method/LC method:
clean up more your samples, go with solid phase
extraction. Get your samples as clean as you could
usually is the best solution for standards divergence
which you see the mass spec response changes over
time.
Lihong Gao, Ph.D.
Back to the Top
The following message was posted to: PharmPK
I agree with all below and have seen this "instrument-based" matrix
effect many times with older systems with in-line sources. I would
also add:
1. If it's feasible, obtain a deuterated label of the test item as
the internal standard, matrix effects should cancel out if this work
is non-regulatory. If it is in support of regulatory work, matrix
effects should be addressed in isolation within the validation (see
CDER 2001 guidelines / 21CFRpart58) and so additional work on the
extraction/analytical method will be required.
2. Again if feasible, try using an system with a source in dual
orthogonal configuration similar to the Z-spray ( Waters/Micromass).
The majority of non-ionisable plasma-based crud does not enter the
source and contains an additional physical filter immediately prior
to entering the optics/Q1. The result of this in our experience has
been very large sample numbers (>200) per batch with no obvious drop-
off or increase in signal response.
3. Use an in-line switching valve immediately after your column to
prevent a large portion of the extracted crud entering the MS in the
first place i.e. divert 0-x min to waste and the rest into the MS.
Hope this helps.
cheers,
Iain
Dr Iain A. Stuart
Preclinical Development
Cyclacel Limited
Back to the Top
The following message was posted to: PharmPK
Dear James,
If your deviation occurs with biological samples you could :
1_ Try to rinse your column with a large step gradient of your organic
mobile phase solvent
2_ Divert to waste the first minutes of your run
3_ In Esi, use a split and add a make up before your source (with your
organic mobile phase)
In APCI, without split, add a make up before your source (with
your
organic mobile phase)
4_ Before the first injection of your samples, you could inject a
serie of
blank samples in order to stabilize your system
5_ Keep your samples in a place with the same temperature as your
autosampler for at least 12 hours (or more ..) before injection
If your deviation occurs with non-biological samples, you should
know if,
after this drift, the signal is stable.
I think about some parameters
6_Have a look at the lab temperature (should be constant and not too
hot)
7_Equilibration of your system (for example, some MS/MS need time for
Cad
gaz stabilisation)
8_Evaporation of your samples in the autosampler, with some differences
between the product and the internal standard
9_Better or worse solubilisation of your products (and internal
stdrd) in
your sample
points 1 to 5 could be try too, to ameliorate your results.
Best regards.
Fabrice Guillet
Bioanalytical Group
Fournier Pharma
PharmPK Discussion List Archive Index page
Copyright 1995-2010 David W. A. Bourne (david@boomer.org)