****************************************************************************** From: "Daniel L." Date: Fri, 16 Apr 1999 21:51:14 +0200 Subject: TSQ7000 - Drugs in plasma Organization: Club-Internet (France) I've been measuring drugs in plasma for one year by LC/MS with TSQ7000, APCI probe. The technique was based on a precipitation with acetonitril. I was used to have series of about 100 runs without problems. Unfortunately, the drug I must quantify now has no response with APCI. So, I use ESI probe. When I make analysis of pure solutions, no problems. But when I try to inject plasma precipitates, the intensity of signal decrease very quickly: about 50% lost with 10 runs on the internal standard. Does somebody know this kind of problem and could explain it? Does somebody know what I can do to about it? Maybe there are web sites where I could find some pieces of advice... Welcome to all suggestions!! Thank you very much for your help. Daniel Labarre ****************************************************************************** From: "M Sweeney - MSMS Consulting" Date: Fri, 16 Apr 1999 22:22:38 -0700 Subject: Re: TSQ7000 - Drugs in plasma Organization: EarthLink Network, Inc. info you gave: 1)ESI assay with acn plasma crash 2)pure stds "work" 3)plasmas with istd show signal drop off ideas in response The column is accumulating materials that interfere with ionization and these materials slowly elute. You don't mention the HPLC method. If you are using a plasma crash (acn ppt.) then you are not removing the more lipophilic materials and they are sticking on the column and then slowly eluting. I would bet you are trying a fast gradient or isocratic method. If this is the problem then putting on the pure standard solutions after the failure state would yield the low intensities that slowly improve to close to 'normal'. This would be because the column is now washing as the pure set runs. Eventually these components are rinsed to a large degree. Putting on the plasma derived samples again would then produce accumulation of late eluting analyte/ISTD suppressing compounds. The sensitivity would then go down again. ESI seems to be more subject to this type of behavior because of the ionization process. It can be helped by working on the sample prep. A sep-pak type method might be nice if developed to produce recovery of the ISTD and analyte and keep the lipophilic material on the sep-pak. So if the compound comes off in 50:50 acn:h20 on the analytic column then a 70:30 acn:h2O (or so) might just get all the compound and keep the fatty stuff on the sep-pak. You may even improve things using 100% acn (min. volumes) giving good recovery of analytes and still keeping the fats on the column. You might try mixed mode or a different stationary phase and solvent than your analytical technique uses. There is an outside chance that you have a stability issue of the istd in plasma. But I bet not. Another possibility is the source is getting dirty or clogged. If the sensitivity comes back after a long HPLC rinse then you know it probably is not dirty source. The log files also show source pressure and you can look at this value to see if the pressure is going down with sensitivity. This would mean that the hole is clogged and you are sampling less of the spray. But I doubt it. The following set would confirm the ion supression theory. Run it as one big run of about 40 or more samples. Feel free to shorten up the first set of stds. Another way would be to take the HPLC out of line and do FIA loop injections. It may take a few minutes to get any residue out of the plumbing (try looking full scan q1ms) but if the sensitivity comes back you know that it is HPLC 0)Possibly do FIA with out column to get base line value on ISTD pure 1) pure std curve with istd (4-10 samples) 2)10 plasmas with istd/analyte (all at one conc maybe) 3)re-inject same pure std curve with istd (4-10 samples) 00)do an FIA with out the column if you want here if sensitivity has not returned. 4)inject pure ISTDS (you don't really need analyte) until area counts return to about normal (you can also try - 5 blanks and then a pure ISTD, 5 blank and then pure ISTD pattern. 5)re-inject same pure std curve with istd (4-10 samples) 6)re-inject 10 plasmas with istd/analyte (all at one conc maybe) Plot the ISTD areas vs time/injection # Look at the graph. I bet you will see group 1) flat equal areas group 2) roll off from area cts in group 1) group 3) area counts no higher than the last from group 2) - may gradually increase or decrease as big wide slowly eluting glops come off the column. This is not predictable as it depends on the chemistry/column etc. I have seen peaks come off hours later from dirty samples (even at maximum organic). group4) this is a rinse step and it depends on chromatography and sample chemistry. Hopefully you can get it rinsed and eventually recover area counts. group5) should have flat area counts like group 1) group 6) rolls off from group 5) counts (you got the column dirty again) Fix it with sample prep and HPLC col. rinse steps. It is common in quan work to not see this as one often starts working exclusively in SRM mode and you don't see this stuff at other masses coming off the column. You can check it out by going back to Q1MS mode and looking out to even fairly high mass (over 1000). Finally many times a 50% drop in ISTD is acceptable to the lab and normal to boot! You are the driver. Feel free to write with further questions or to tell me I am all wrong. Matt Sweeney mattsweeney@earthlink.net Mass Spec Consulting Available for any of the following please contact me. Training/Operations/Consulting/Method Development LC/MS Pharmacokinetics, Peptides, Proteins, Metabolism, Maintenance Classes, Specialist in Finnigan Equipment and Software ****************************************************************************** From: "Tata, Prasad" Date: Wed, 21 Apr 1999 09:01:01 -0400 Subject: Re: TSQ7000 - Drugs in plasma Organization: * Mr. Labarre: This is a common problem, protein precipitation is a good tool for quick sample clean up. But it leaves other proteins in acetonitrile, which builds up over the time. My suggestion would be to use extraction with other organic solvents, add a solid phase clean up after protein precipitation, or evaporate the acetonitrile and reconstitute in acetonitrile-ammonium acetate mixture. Ammonium acetate aids ionization and also aids volatilization of the extraction. In general, protein precipitation is not good for assays that are intended for use in long -term routine studies. Hope this helps. Prasad NV Tata, Ph.D. ****************************************************************************** From: "Stuart" Date: Thu, 13 May 1999 23:25:18 +1000 Subject: Re: TSQ7000 - Drugs in plasma Organization: Monash Uni I have been developing a method based on the Supelco Hisep column which precludes proteins and other high MW compounds from the C18, which means they wash off early in the chromatography. The addition of a bypass valve between the column and the MS then allows you to divert the nasty stuff away from the MS (and other detectors). Although I have only had the column a couple of weeks it seems to work perfectly. The only downside is the fact that the column has a limited lifespan and is expensive ($A800ish). Cheers Stuart Thomson Manager - Mass Spectrometry Facility Victorian College of Pharmacy Monash University