Liquid Chromatography, In Combination with a Quadrupole Time-of

Page 1 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Liquid Chromatography, in Combination with a Quadrupole Time-of-Flight Instrument (LC QTOF), with Sequential Window Acquisition of All Theoretical Fragment-Ion Spectra (SWATH) Acquisition: Validated quantification of 39 antidepressants in whole blood as part of a simultaneous screening and quantification procedure. Andreas T. Roemmelt, Andrea E. Steuer and Thomas Kraemer*. Department of Forensic Pharmacology & Toxicology, Zurich Institute of Forensic Medicine, University of Zurich, Switzerland. ABSTRACT: SWATH (sequential window acquisition of all theoretical fragment ion spectra) is a data independent acquisition (DIA) method for very fast scanning QTOF instruments. SWATH repeatedly cycles through 28 consecutive 20 Da precursor isolation windows detecting all precursor ions and fragments MSALL like and yet fast enough to generate more than 10 data points over the chromatographic peak. It was already shown in previous publications that SWATH, despite its wide Q1 windows, allows the identification of different substances and that SWATH has a higher identification rate than data dependent acquisition approaches. The aim of this study was a proof of concept study whether these same datasets can also enable validated quantification according to international guidelines, exemplified for 39 antidepressants. The validation included recovery, matrix effects, process efficiency, ion suppression and enhancement of coeluting ions, selectivity, accuracy, precision and stability. The method using SWATH acquisition proved to be selective, sensitive, accurate and precise enough for 33 out of the 39 antidepressants. The applicability of SWATH for screening and validated quantification in the same run was successfully tested with authentic whole blood samples containing different antidepressants and other drugs thus proving the QUAL/QUAN abilities of SWATH. In an additional systematic investigation, it could be shown that calibration curves injected a few days after or before the actual sample can be used for quantification with acceptable accuracy.

INTRODUCTION The simultaneous identification and quantification of all present substances in one analytical run is a major goal in analytical toxicology, especially in clinical and forensic toxicology. Future methods should be based on instruments and methods that enable detection of all ions, their identification, and quantification of relevant substances without the need for reinjection. State of the art instrumentation is liquid chromatography (LC) coupled to tandem mass spectrometers (MS/MS). Almost all current LC MS/MS methods in clinical and forensic toxicology are based on the multiple reaction monitoring (MRM) mode. The MRM mode is ideal for targeted screening of compounds 1-3. However, this detection mode is strictly targeted and cannot be used for actual open screening purposes. For accurate and precise quantification only a limited number of analytes can be included in the method to be quantified, due to the prolongation of the cycle time by adding (too) many analytes 4-6, therefore, a second method and injection are necessary for quantification. With these drawbacks in mind it is time for a paradigm shift from selective reaction monitoring (or multiple reaction monitoring) to high resolution mass spectrometry7. Quadrupole time of flight instruments (QTOF) principally detect all precursor ions, which come off the column and depending on the cycle time, enough data points

should be generated to allow accurate and reproducible quantification, too. Due to isomers in the sample, identification solely based on the precursor can be difficult, so further fragmentation information is indispensable. In order to obtain fragment information, many users apply data (or information) dependent acquisition (DDA; IDA), which more or less randomly selects precursors to be further fragmented to obtain MS2 spectra 8, which can then be searched in databases9. Most recent publications emphasize the advantages of data independent acquisition (DIA) methods like Sequential Window Acquisition of All Theoretical Fragment-Ion Spectra (SWATH) for screening purposes10,11. SWATH technique is already widely applied in proteomics research12-14 , metabolomics15 and biomarker16-18 research. Data recording of SWATH acquisition consists of a recurring cycle of a survey scan followed by a Q1 isolation strategy. In the first step a survey scan with low collision energy covers the user defined mass range (Q1 set to full transmission) to obtain the intact precursor ion. Then the mass range is consecutively scanned by several predefined Q1 windows (e.g. 20 Da) applying a collision energy to produce product ion spectra. In a previous publication the suitability of SWATH for clinical and forensic toxicological screening was systematically investigated. It was shown that SWATH could detect more substances than DDA methods19 and that despite the

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

wide Q1 windows SWATH is still selective enough for unambiguous identification20. In the present proof of concept study, it should be investigated whether the SWATH approach actually allows simultaneous screening and quantification (QUAL/QUAN analysis) in the same analytical run. Therefore, the same LC and MS parameters described in our previous SWATH screening method were applied to a quantitative method that should be fully validated according to international guidelines. Antidepressants representing an important group of drugs relevant in clinical and forensic toxicology were chosen as analytes. It should also be tested whether accurate quantification is possible using calibration curves stored some time before or generated some time after the actual screening/quantification run. This should reflect the real life situation in a clinical or forensic toxicological lab, where quantification of previously unknown drugs is required after a screening. Application was also tested by screening authentic blood samples and quantification of detected antidepressants. Results were compared with those from conventional methods comprising of a screening and a second quantification method.

EXPERIMENTAL SECTION Reagents and Chemicals. The chemicals and solvents for the HPLC system were purchased from the following suppliers: water (LC-MS grade), acetonitrile (LC-MS grade) and acetic acid (99% for analysis) from Merck (Darmstadt, Germany), ammonium acetate from Sigma Aldrich (Buchs, Switzerland). Agomelatine was purchased at Cayman Chemical (Michigan, USA). Amitriptyline, amitriptyline-d3, noramitriptyline, amitriptyline N-oxide, amoxapine, atomoxetine, bupropion, hydroxybupropion, citalopram, norcitalopram , clomipramine, clomipramine-d3, norclomipramine, cyclobenzaprine, desipramine, dosulepine (dothiepin), nordosulepine), dosulepinesulfoxide, doxepin, nordoxepin, duloxetine, fluoxetine, fluoxetine-d6, norfluoxetine, fluvoxamine, imipramine, maprotiline, mianserin, mirtazapine, moclobemide, opipramol, paroxetine, reboxetine, sertraline, tianeptine, tranylcypromine, trazodone, trimipramine, nortrimipramine, venlafaxine and o-desmethylvenlafaxine were purchased from LGC (Wesel, Germany). Norfluoxetine-d6, paroxetine-d6 and trimpramine-d3 came from Cerilliant (Austin, TX, USA). Fluvoxamine-d3, trazodone-d6 and o-desmethyl venlafaxine -d6 were obtained from TRC (Toronto, Ontario, Canada). Venlafaxine- d6 and was purchased at Lipomed (Arlesheim, Switzerland) and citalopram-d6 at Reseachem (Burgdorf, Switzerland). Instrumentation. High Performance Liquid Chromatography separation was performed using an UltiMate 3000 Rapid Separation Liquid Chromatography system (RSLC) (Thermo Fischer Scientific, San Jose, CA), configured in binary high pressure gradient mode and controlled by Chromeleon 6.80 software (Thermo Fischer Scientific, San Jose, CA). Mobile phase A consisted of 25 mM ammonium acetate buffer with 0.1% acetic acid and mobile phase B consisted of acetonitrile with 0.1% acetic acid. The gradient was programmed as follows: 0.00–1.00 min: 5% eluent B, 1.01–5.00 min: gradient increase to 20% eluent B, 5.01–7.00 min: 22% eluent B, 7.01–9.00 min: 30% eluent B, 9.0113.00 min: 40% eluent B; 13.01-14.00 min: 50% eluent B, 14.0116.00 min: hold 50% eluent B; 16.01-17.00 min: 95% eluent B, 17.01-19.00 min: hold 95% eluent B, 19.01-20.00 min: 5% eluent B. For re-equilibration of the HPLC column gradient was set to 5% eluent B for 5 min. The column oven was set at 40 °C and the autosampler was cooled at 7 °C. The flow rate was 0.5 mL min−1.

Page 2 of 9

Ten microliter of the samples were injected onto a Synergi Polar RP column (100 x 2.0 mm I.D.; 2.5 µm particle size, 100A) (Phenomenex, Torrance, CA, USA) guarded with a C18 guard column (2.0 mm I.D. × 4.0 mm; Phenomenex, Torrance, CA, USA). Mass spectrometric detection was performed using a QTOF MS (TripleTOF 5600, AB Sciex, Concord, Ontario, Canada) operated in positive ionization mode with a DuoSpray ion source at a resolving power (full width at half maximum, FWHM, at m/z 400) of 30.000 in MS and 30.000 in SWATH MS2 (high resolution mode). The automated calibration device system (CDS) was set to perform an external calibration every four samples. The source conditions were: temperature 500 °C, curtain gas (CUR) 25, on source gas (GS) 1 and 2 at 45 and 57, respectively, and ion-spray voltage floating (ISVF) at 5.5 kV. One complete cycle of the SWATH method consists of a survey scan and Q1 isolation strategy. The survey scan covered a mass range of m/z 100-700 with low collision energy of 10 eV and an accumulation time of 100 ms. The Q1 isolation strategy covered a mass range of m/z 100-650 with a 20 Da window for Q1 isolation (overlap 1u). In each SWATH window a collision energy of 35 eV with a spread of ± 15 eV and an accumulation time of about 40 ms in high resolution mode was used. The total cycle time was 1.2 s. All MS parameters were controlled by AnalystTF Software 1.6 from AB Sciex (AB Sciex, Concord, Ontario, Canada). The performance of the instrument is tested every day with an injection of a test solution of five different triazine compounds. Peak areas and retention times are plotted and compared to the ones measured days before to exclude fluctuations in the performance. Identification and quantification of the analytes. For the identification of the compounds, MasterView 2.0 (ABSciex), which is integrated in the PeakView 2.0 (ABSciex) software, was used. 1326 compounds and their empirical formula were integrated in an intact accurate mass list (XIC list). A mass error of the precursor of less than 5 ppm was demanded for a positive identification of the measured mass compared to the theoretically calculated mass. Isotope ratio difference was set to 40%. The library consists of 534 compounds, which were recorded with five different collision energies (10, 20, 40, 50, 35 ± 15 eV) with a Q1 window set to 1 Da width and integrated into our in house library. The library search of the generated SWATH MS2 was based on a confirmation search with an intensity threshold of 1%. Library hits were positive when the fit value was above 60 (library score points), still tentative above 25 and negative with less than 25. Quantification in MultiQuant 2.1 (ABSciex) was based on the precursor mass of the antidepressants in the survey scan which was extracted with a window width of ± 20 mDa. Retention time had to be within ± 5% of that of the reference standard. All integrated peaks were manually checked. Furthermore it was also checked, whether quantification via the area of fragments in the respective SWATH windows, could also fulfill validation criteria. Blood samples and sample preparation. Blank blood samples were from volunteers of the Zurich Institute of Forensic Medicine (ZIFM) of the Zurich University, Switzerland. All samples were prescreened using routine MRM based screening procedures. Whole Blood Sample Extraction. The whole blood samples were extracted using protein precipitation (PP) as described elsewhere.21 Briefly, 200 µL of whole blood were mixed with 20 µL of the internal standard (IS) mixture containing 2.0 µg/mL amitriptyline-d3, 0.5 µg/mL citalopram-d6, 1.0 µg/mL clomipramined3, 4.0 µg/mL fluoxetine-d6, 1.5 µg/mL fluoxetine M (nor-)-d6,


Page 3 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1.5 µg/mL fluvoxamine-d3, 1.0 µg/mL paroxetine-d6, 8 µg/mL trazodone-d6, 5 µg/mL trimipramine-d3, 2.0 µg/mL venlafaxined6 and 2.5 µg/mL venlafaxine M (desmethyl-)-d6. The spiked whole blood was shaken for 5 min. Then 600 µL of ice cold acetonitrile were added and the mixture was shaken for 10 minutes and centrifuged at 13 000 rpm for 10 min. An aliquot of 600 µL was transferred into a vial and evaporated to dryness under a gentle nitrogen stream at 50°C. The residue was dissolved in 50 µL of a mixture of eluent A and eluent B (50:50, v/v). Aliquots of 10 µL of this solution were injected into the HPLC-HRMS system. Preparation of stock solutions, calibration standards, and control samples. Methanolic stock solutions (1.0 mg/mL) of each analyte were purchased from different providers or produced by separate weighing of the solid substance and dissolved it in the respective amount of solvent. The stock solution of bupropion M (hydroxyl-) was stored in acetonitrile due to instability in methanol. Working solutions (0.1, 0.01 and 0.001 mg/mL) of each analyte were prepared in methanol or acetonitrile by dilution from each stock solution. The spiking solutions for calibration standards and QC samples were prepared by adding the appropriate amount of each of the analytes of the corresponding stock or working solution to volumetric flasks to obtain the corresponding concentration. All calibration solutions and QC solutions were aliquoted in 500 µL glass vials. The working solution was stored at + 4°C and the rest was frozen at - 20 °C until usage. Calibration standards were freshly prepared every day using 200 µL blank whole blood and 20 µL of the corresponding spiking solution. The concentration of the spiking solution was ten times higher than the blood concentration. QC samples were prepared at three different concentrations (LOW, MED, and HIGH). Each sample was prepared in the same way as the calibration solutions. Selectivity and interferences. Nine blank whole blood samples from different sources were analyzed for interfering peaks with the detection of the analytes or the ISs. Criteria set for an interfering peak or analytes from deuterated standards were as follows: more than 10% signal of calibration sample 1 and matching retention time, isotope pattern and library search. Two zero samples (blank sample + ISs) were analyzed to check for absence in the respective analyte peaks caused by the ISs. Two blank samples spiked with the spiking solutions of calibration standard 6 were analyzed to check for absence of interfering signals of the analytes in the subsequent runs (carry-over experiment). Blank whole blood samples were spiked with different drug classes such as benzodiazepines, neuroleptics and different illicit drugs (resulting in high therapeutic concentrations) and analyzed for interfering peaks with the monitored ADs. Recovery, matrix effects, and process efficiency. Recovery (RE), matrix effect (ME) and process efficiencies (PE) were investigated according to the simplified approach described by Matuszewski et al.22. Briefly, three sets of samples were prepared at QC LOW and QC HIGH: samples set A representing the blank matrix spiked before extraction, samples set B representing blank matrix spiked after extraction, and samples set C consisting of neat spike solution. RE results were obtained by comparison of the absolute peak areas of sample set A with those of the corresponding peaks in sample set B. MEs were estimated by comparing the peak areas of sample set B with those in set C. For PE, peak areas of set A were compared with the corresponding peaks in set C.

Ion suppression and enhancement of co-eluting analytes. The effect of ion suppression or enhancement induced by coeluting analytes and their deuterated analogues was investigated, too. The effect of ion suppression/ enhancement was tested at QC LOW and QC HIGH (n=6) by comparing peak areas of the analytes in presence or absence of the corresponding co-eluting analyte. Calibration model and choice of Internal Standards. Daily calibration curves were prepared with each batch of validation samples. After validation all combinations of analyte and internal standards were used for calculations. The combinations showing the best results were chosen for the final quantification (given in Table S-1). Accuracy and Precision. QC samples (LOW, MED, HIGH) were analyzed according to the procedures described above in duplicate on each of eight days. Accuracy was calculated in terms of bias as the percent deviation of the mean calculated concentration at each concentration level from the corresponding theoretical concentration. Intra-day and inter-day precision were calculated as relative standard deviation (RSD) according to Peters et al.23,24. The acceptance intervals of within-day (repeatability) and intermediate precision were ≤15% RSD (≤20% RSD at LLOQ) and ±15% for bias (±20% at the LLOQ) of the nominal values. Processed sample stability. For estimation of the stability of processed samples under the conditions of HPLC-HRMS analysis, QC LOW and QC HIGH samples (n=6) were extracted as described above. The extracts at each concentration were pooled. Six aliquots of these pooled extracts at each concentration level were transferred to an autosampler vial and injected under the conditions of a regular analytical run at time intervals of 8 hours over a total run time of 48 hours. Stability of the extracted ADs was tested by regression analysis plotting absolute peak areas of each analyte at each concentration versus injection time. Instability of processed samples would be indicated by a negative slope significantly different from zero (p < 0.05). Freeze/ thaw stability and long-term stability. For evaluation of freeze /thaw stability, QC samples (LOW and HIGH) were analyzed before (control samples; each n=6) and after three freeze/thaw cycles (stability samples; n=6).For each freeze and thaw cycle, the samples were frozen at -22°C for 20 hours, thawed, and kept at room temperature for 3 hours. The concentrations of the QC samples were calculated based on the daily calibration curves. The internal standards were added just before extraction. For stability, there were two criteria which had to be met; the ratio of means (stability/control) had to be within 90110%, and the 90%-confidence interval had to be within 80-120% of the control sample. Control samples for the long-term stability were the same as for freeze/ thaw cycles (control samples; each n=6). The stability samples (stability samples; each n=6) were thawed and worked up as described above in an interval of one week over a period of 3 month. The internal standards were added just before extraction. For stability, there were two criteria which had to be met; the ratio of means (stability/ control) had to be within 90-110%, and the 90%-confidence interval had to be within 80-120% of the control sample. Limits of detection. The lowest point of the calibration curve was defined as the limit of quantitation (LOQ) of the method and fulfilled the requirement of LOQ with a signal to noise ratio of 10:1 determined by comparing background signal height after



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

blank sample extraction and extraction of the lowest calibrator. The lowest calibrator was set at a concentration level corresponding to half of the lower therapeutic level. LOD was not systematically investigated. Applicability. Ten authentic blood samples from the routine laboratory of the Zurich Institute of Forensic Medicine were screened and antidepressants detected during the screening procedure were quantified using calibration curves stored some time before or generated some time after the actual screening/quantification run. This should reflect the real life situation in a clinical or forensic toxicological lab, where quantification of previously unknown drugs is required after a screening. Obtained results were compared with those from conventional methods comprising of a screening and a second quantification method.

RESULTS AND DISCUSSION The aim of this study was to proof the concept whether the SWATH approach actually allows simultaneous screening and quantification (QUAL/QUAN analysis) in the same analytical run according to international guidelines. Siegel et al. and Hopfgartner et al. have already shown initial experiments on simultaneous QUAL/QUAN of aldehydes/ketones or bosentan in different matrices15,25. However, blood as typical quantification matrix for clinical and forensic purposes had not been studied and validation according to international guidelines including precision and accuracy testing over several days was not performed. The quantitative assay for 39 antidepressants in the present study was fully validated according to such international guidelines26 as part of a simultaneous screening and quantification method. Specific aspects and/or problems resulting from that QUAL/QUAN concept are also considered and critically discussed in the corresponding sections. Sample preparation. Simple acetonitrile precipitation is unselective and can lead to more prominent matrix effects and matrix peaks than specific work-up procedures such as liquid-liquid or solid-phase extraction27 . If validated quantification is the only goal for a procedure, highly specific sample preparation procedures can improve validation results and avoid e.g. matrix effects or sensitivity issues 28. In the present study, sample preparation had to be unspecific, because it should also be suitable for a broad screening procedure. More specific sample preparation would have led to loss of analytes thus corrupting the concept of simultaneous screening and quantification. It could be shown, that the present method was adequate for precise and accurate quantification of the ADs as can be seen in the corresponding section below. However, the “quick and dirty” sample preparation might also explain the higher variance of repeatability and intermediate precision compared to more specific extraction methods 29-34. The high ion enhancement of bupropion was also observed by Remane et al.31 despite their more specific liquid-liquid extraction. Only Montenarh et al.35 did not detect problems with bupropion also using a liquid-liquid extraction. Since similar issues could also be observed with more specific work-ups regarding matrix effects, recovery and process efficiency, our assay was considered to be suitable for the validation. Further investigations will be necessary to test the recovery and sensitivity of the assay for more substance classes using the QUAL/QUAN method.

Page 4 of 9

Selectivity and Interferences. The validated quantification of the antidepressants was based on the precursor detection in the TOF MS scan. No major interference peaks were detected in nine different blank whole blood samples. All interfering signals with an extraction mass within ± 20 mDa of the analyte could be differentiated by their retention times, e.g. the isomers amitriptyline and maprotiline. The zero samples showed minor peaks for clomipramine, citalopram and trazodone. All signals were less than 5% of the lowest calibration sample (i.e. in the subtherapeutic range). Interferences from other drug classes typically present in forensic samples were tested with benzodiazepines, neuroleptics and other typical illicit drugs, but no interferences were observed except for the isomers tramadol and O-desmethylvenlafaxine, but they could be distinguished by their retention times. Recovery, matrix effects and process efficiency. Matrix effects, recovery and process efficiency for the different analytes are summarized in table S-2. Only in QC LOW, recoveries for amitriptyline, amitriptyline N-oxide, amoxapine, duloxetine and paroxetine were below 50%, but with still acceptable coefficient of variation and the intensity of the substances in the lowest calibration sample was still sufficient for quantification. In QC HIGH, no problems were detected. Matrix effects in QC LOW ranged from 80% to 1257% and from 93% to 1995% in QC HIGH. In QC LOW, matrix effects were observed for amoxapine, bupropion, fluvoxamine and paroxetine. Unacceptable coefficients of variation in QC LOW were calculated only for norfluoxetine. In QC HIGH, bupropion and norfluoxetine showed matrix effects. Matrix effects for atomoxetine were slightly above the acceptance criteria but due to the low coefficient of variation it was still acceptable. The results from process efficiency ranged in QC LOW from 27.8% to 444% and in QC HIGH from 50% to 124% and were overall acceptable. This variability could be compensated using the internal standard. Accuracy and Precision. The results obtained using a daily calibration model are shown in table 1. With the exception of norfluoxetine, amitriptyline- N-oxide and duloxetine, which failed to fulfill the criteria, all analytes met the criteria. The insufficient accuracy and precision for norfluoxetine and amitriptyline-Noxide could be explained by varying matrix effects. Duloxetine could not be validated due to poor ionization under the given conditions. The presented LC-HRMS method was also not suitable for reproducible quantification of tranylcypromine. Thus, the performance in terms of accuracy and precision of the present SWATH method is comparable with MRM based4,21,31 methods but offers additional QUAL capabilities. Ion suppression and enhancement of coeluting analytes. Comparing the corresponding sample sets, no significant influence on the analytes was detected. Stabilities. Data on long term stability for most antidepressants are available. Several publications exist about the stability of antidepressants in plasma and serum and all state that most of the antidepressants are be stable over a long period 36-50 . However, sufficient data on the stability of the analytes in whole blood samples was not available. During freeze and thaw cycles none of the analytes showed significant degradation in whole blood. Over a time of 12 weeks, all analytes were within the required 90-110% interval. Concerning processed sample stability, the analytes were stable in whole blood extracts for a period of 48 h in the autosampler kept at 7°C.


Page 5 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48


Table 1. Summary of the results of the validation for the given analytes.

Agomelatine Amitriptyline Amitriptyline M (Nor-) Amitriptyline M (-N-oxide) Amoxapine Atomoxetine Bupropion Bupropion M (Hydroxy-) Citalopram Citalopram M (Nor-) Clomipramine Clomipramine M (Nor-) Cyclobenzaprine Desipramine Dosulepin (Dothiepin) Dosulepin M (Nor-), Northiaden Dosulepin M (Sulfoxid-) Doxepin Doxepin M (Nor-) Duloxetine Fluoxetine Fluoxetine M (Nor-) Fluvoxamine Imipramine Maprotiline Mianserin Mirtazapine Moclobemide Opipramol Paroxetine Reboxetine Sertraline Tianeptine Tranylcypromine Trazodone Trimipramine Trimipramine M (Nor-) Venlafaxine Venlafaxine M (O- Desmethyl)

nominal conc [ng/mL]

mean cal. conc. [ng/mL]

5.0 30.0 25.0

5.11 29.8 25.1

QC low Accuracy Repeatability bias [%] CV [%]

2.3 -0.6 0.5

7.0 11.3 12.3

intermediate precision CV [%]

nominal conc [ng/mL]

9.4 11.3 12.3

300.0 310.0 310.0

mean cal. conc. [ng/mL ] 325.3 261.2 306.3

QC medium Accuracy Repeatability bias [%] CV [%]

8.4 -14.5 -1.2

intermediate precision CV [%]

nominal conc [µg/mL]

mean cal. conc. [ng/mL]

9.0 11.6 5.3

11.1 11.6 6.1

480.0 480.0 480.0

500.0 435.7 465.1

QC high Accuracy Repeatability bias [%] CV [%]

4.3 -9.2 -3.1

6.7 11.9 7.7

intermediate precision CV [%] 10.0 14.2 7.7

48.0

49.9

4.0

38.0

39.2

920.0

954.8

3.6

28.5

36.7

1440.0

1517.8

5.4

14.1

14.5

8.0 30.0 12.0 24.0 12.0 12.0 24.0 30.0 3.0 60.0 12.0 60.0 24.0 7.0 7.0 18.0 60.0 60.0 30.0 24.0 36.0 7.0 12.0 180.0 30.0 12.0 7.0 18.0 30.0 2.0 300.0 12.0 12.0 7.0

9.1 29.7 12.3 26.2 12.0 12.2 24.7 30.2 2.9 62.4 13.1 62.0 26.5 7.1 7.4 20.5 60.0 56.1 31.3 22.4 40.8 6.5 11.8 189.9 29.5 13.1 6.0 17.5 29.8 n.d. 253.4 11.0 11.6 6.5

13.6 -1.1 2.6 9.1 0.0 1.2 2.9 0.8 -2.5 4.0 9.2 3.3 10.3 1.8 5.3 20.5 0.0 -6.6 4.3 -2.6 13.3 -7.3 -8.2 5.5 -1.5 8.7 -13.7 -3.0 -0.6 n.d. 12.5 -8.2 -3.8 -6.9

9.5 5.6 8.9 4.1 10.8 7.2 7.4 9.0 6.9 5.6 8.3 9.2 9.3 6.5 4.0 25.0 4.0 15.3 3.7 4.1 2.3 5.3 5.8 10.6 12.6 4.4 14.5 3.1 12.3 n.d. 9.4 3.5 9.6 1.9

9.5 8.3 10.7 4.1 10.8 7.2 9.1 10.6 8.1 9.0 8.9 12.2 10.1 8.1 5.3 25.0 10.1 18.3 6.1 6.7 2.5 8.6 7.0 10.8 12.8 4.7 14.5 10.1 12.3 n.d. 13.5 9.1 10.6 6.4

200.0 450.0 250.0 100.0 200.0 200.0 460.0 560.0 40.0 330.0 150.0 220.0 410.0 150.0 350.0 120.0 525.0 525.0 310.0 310.0 265.0 80.0 100.0 4000.0 500.0 150.0 350.0 250.0 510.0 50.0 2600.0 300.0 300.0 750.0

190.3 446.5 238.2 99.1 203.0 203.8 423.9 562.8 37.2 317.0 137.8 213.8 378.8 147.6 338.7 139.6 523.1 565.1 313.1 310.8 292.4 76.7 95.0 3430.0 489.5 136.0 351.8 229.5 558.1 n.d. 2525.9 278.3 299.6 790.3

-4.8 -0.8 -4.7 -0.9 1.5 1.9 -7.9 0.5 -6.9 -3.9 -8.2 -2.8 -7.6 -1.6 -3.2 16.3 -0.4 7.6 1.0 0.3 10.4 -4.1 -5.0 -14.2 -2.1 -9.3 0.5 -8.2 9.4 n.d. -2.9 -7.2 -0.2 5.4

3.2 13.0 8.6 9.0 5.7 3.0 6.7 8.4 4.4 5.2 7.5 5.7 5.2 7.1 7.4 5.14. 7.9 10.2 6.7 5.3 10.1 7.5 4.8 4.3 4.1 9.0 6.8 6.5 4.9 n.d. 4.7 8.3 10.5 10.6

3.3 13.0 8.6 9.0 7.7 7.1 10.0 8.4 7.2 9.5 8.6 6.8 5.2 9.5 8.0 17.5 10.2 15.6 10.7 8.6 10.5 7.5 7.2 8.6 5.5 11.1 6.8 11.5 5.4 n.d. 4.7 9.5 10.5 14.0

320.0 700.0 400.0 160.0 320.0 320.0 720.0 880.0 60.0 480.0 240.0 320.0 640.0 240.0 560.0 200.0 800.0 800.0 480.0 480.0 400.0 120.0 160.0 6400.0 800.0 240.0 550.0 400.0 800.0 80.0 4000.0 480.0 480.0 1200.0

332.8 752.4 418.7 163.4 311.4 309.4 749.1 889.9 56.5 502.8 234.9 317.8 595.5 238.0 560.0 205.8 842.5 936.5 500.2 486.6 427.3 120.9 170.1 5519.5 745.1 243.3 537.6 352.6 899.7 n.d. 3803.8 504.9 501.4 1323.6

4.0 7.5 4.7 2.2 -2.7 -3.3 4.0 1.1 -5.9 4.8 -2.1 -0.1 -7.0 -0.8 0.0 2.9 5.3 17.7 4.2 2.0 6.8 0.8 6.3 -13.8 -6.9 1.4 -2.3 -11.8 12.5 n.d. -4.9 5.2 4.5 10.3

4.7 6.2 10.4 3.6 9.2 7.1 13.5 9.0 9.5 2.9 9.6 3.9 8.3 8.0 8.8 9.8 4.6 6.7 11.0 8.4 8.8 5.7 4.8 4.2 6.7 12.8 5.4 8.1 4.6 n.d. 13.0 6.5 4.4 2.7

4.7 6.5 12.0 4.2 10.2 5.3 14.4 11.2 10.4 6.8 10.5 5.6 9.6 8.3 10.2 12.6 7.5 7.3 12.3 8.7 11.4 5.7 5.9 5.7 7.4 13.7 8.1 14.4 4.6 n.d. 13.0 6.8 4.5 6.0

60.0

66.3

0.5

11.6

14.2

790.0

768.0

-2.8

5.9

6.4

1200.0

1215.9

1.3

6.3

7.7



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 1. Extracted Ion Chromatograms (XIC) of an authentic case, indicating the presence of the given compounds. The detected ADs venlafaxine, its demethyl metabolite and mirtazapine were quantified in the same run using the present method (results given with the compound names). Applicability. Ten authentic whole blood samples from cases of the ZIFM were analyzed and compared to our standard method for antidepressants and neuroleptics using scheduled MRM on a 5500 QTrap instrument21. All samples were routinely screened with our SWATH method employing high resolution MS20 and antidepressants, which were detected in the run, were quantified in the same data file using the present method with calibration curves generated the same day (table 2). In figure 1, corresponding extracted ion chromatograms (XIC) of an authentic case are shown, indicating the presence of phenazone, Odesmethylvenlafaxine, quinine, venlafaxine, mirtazapine, zolpidem, nordiazepam and diazepam. Exemplified SWATH spectra can be found in supporting information. The ADs desmethylvenlafaxine, venlafaxine and mirtazapine were quantified in the same run using the present method. The results from the SWATH method matched very well with the ones from our routine method. In real world forensic laboratory work, samples are screened and detected compounds are often quantified one or more days later, when evaluation of the screening run is completed and the responsible expert has decided on the necessity of quantification. If quantification finally is necessary, the sample has to be workedup and analyzed again using a second method. This can be problematic, especially if only small sample amounts are available, which is often the case in forensic casework. SWATH proved to be superior for that purpose, since quantification of detected analytes can be done in the same data file without need for returning in the laboratory. After validation, this approach was systematically investigated using the generated data files. Quality control samples (LOW, MED, HIGH) of the validation day 1-8 were quantified with the calibration curve from day one. It should be mentioned that the quantification using older calibrations only works, if the instrument performance is constantly monitored over this period. We recommend using freshly made whole blood QCs for this kind of quantification to check the accuracy of the analysis. On three days (day 2, 3 and 8) quantification using an older calibration curve was only possible with analytes with the respective labelled internal standard in the sample. Whereas, on these days the quantification based on a calibration curve from the very same day worked

Page 6 of 9

well with all analytes and results were within the required range. Provided that the instrument is within the required performance compared with the results of the calibration (± 20%) analysis of the rest of the data revealed that the accuracy error of 25 analytes was below 20% for QC LOW and below 15% for QC MED and HIGH. The quantitative error of the rest of the antidepressants were still below 30% and therefore good enough for at least semiquantitative assessment. This proved that stored calibration curves or calibration curves generated after the evaluation of the screening run can be used for quantification. These results show that for most of the validated analytes quantification can be done even using a stored calibration curve. These results are also valid for other LC-MS methods. Even if same day calibration is done, all the laboratory work (extraction, LC-MS/MS analysis) for the already screened sample can be omitted, proving the big advantage of SWATH acquisition. In addition, a good estimation, whether a substance is in the toxic or more or less in the therapeutic concentration range is often sufficient for assessment of a clinical or forensic case. It should also be emphasized that the data file generated using the SWATH acquisition can be re-analyzed over and over again for new analytes, e.g. if information about use of a new psychoactive substance (NPS) has become available10. Table 2. Identified (ID) and quantified substances in authentic cases (id=identified, but not quantified, nd=not identified). Case

Detected Compounds

1

2

3

4

5

Venlafaxine O-desmethylvenlafaxine Methadone EDDP Oxazepam Paroxetine Atropine Flupentixol Melitracene Rocuronium

QUAN [ng/mL] SWATH 55 112 id id id 208 id id id id

QUAN [ng/mL] MRM 47 105 id id id 220 id id id id

Venlafaxine O-desmethylvenlafaxine Mirtazapine Diazepam Nordazepam Phenazone Quinine Zolpidem Citalopram Norcitalopram Trazodone m-CPP

76 82 9 id id id id id 41 33 id id

63 89 14 id id id id id 48 33 id id

Mirtazapine N-Desmethylmirtazapine Amiodarone Melitracene

28 id id id

36 id id nd


Page 7 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Using conventional MRM-based targeted methods, blood samples must be picked out of the deep freezer in a storage unit, worked-up in the laboratory and analyzed again for their presence. This clearly proves superiority of SWATH for those purposes in clinical and forensic toxicology.

(3) Gergov, M.; Ojanpera, I.; Vuori, E. J.Chromatogr.B Analyt.Technol.Biomed.Life Sci. 2003, 795, 41-53.

Other data independent acquisition methods using QTOF instruments are available on the market51-53. Due to the more unspecific acquisition using low and high collision energy fullscan with Q1 in full transmission, overall cycle time is shorter than in SWATH. The windowed acquisition of SWATH leads to higher selectivity. Given that both methods have advantages and disadvantages, a final conclusion on superiority of one over the other is not yet possible until a direct comparison of the different instruments employing the different DIA approaches is made.

(5) Remane, D.; Wetzel, D.; Peters, F. Analytical and Bioanalytical Chemistry 2014, 406, 4411-4424.

(4) Saar, E.; Gerostamoulos, D.; Drummer, O. H.; Beyer, J. Journal of mass spectrometry : JMS 2010, 45, 915-925.

(6) Kirchherr, H.; Kühn-Velten, W. N. Journal of Chromatography B 2006, 843, 100-113. (7) Ramanathan, R.; Jemal, M.; Ramagiri, S.; Xia, Y.-Q.; Humpreys, W. G.; Olah, T.; Korfmacher, W. A. Journal of Mass Spectrometry 2011, 46, 595-601. (8) Andrews, G. L.; Simons, B. L.; Young, J. B.; Hawkridge, A. M.; Muddiman, D. C. Analytical chemistry 2011, 83, 5442-5446.

CONCLUSION In this proof of concept study, it could be shown, that the SWATH approach actually allows simultaneous screening and quantification (QUAL/QUAN analysis) of small molecules in the same analytical run and according to strict international guidelines. Quantitative abilities of SWATH were proven by successful validation and were comparable with MRM based methods. In authentic cases, identified antidepressants were quantified in the same SWATH run without need for returning in the laboratory. SWATH acquisition may therefore be the future of forensic toxicological analysis.

(9) Krauss, M.; Singer, H.; Hollender, J. Analytical and Bioanalytical Chemistry 2010, 397, 943-951. (10) Scheidweiler, K.; Jarvis, M. Y.; Huestis, M. Analytical and Bioanalytical Chemistry 2015, 407, 883-897. (11) Arnhard, K.; Gottschall, A.; Pitterl, F.; Oberacher, H. Analytical and Bioanalytical Chemistry 2015, 407, 405-414. (12) Liu, Y.; Chen, J.; Sethi, A.; Li, Q. K.; Chen, L.; Collins, B.; Gillet, L. C. J.; Wollscheid, B.; Zhang, H.; Aebersold, R. Molecular & Cellular Proteomics 2014, 13, 1753-1768.

ASSOCIATED CONTENT (13) Gillet, L. C.; Navarro, P.; Tate, S.; Röst, H.; Selevsek, N.; Reiter, L.; Bonner, R.; Aebersold, R. Molecular & Cellular Proteomics 2012, 11, 1-17.

Supporting Information A table of concentrations of the calibrants and QCs, one table about the matrix effects, recovery and process efficiency and figures of different runs can be found in the supporting material. This material is available free of charge via the internet http://pubs.acs.org.

(14) Bruderer, R.; Bernhardt, O. M.; Gandhi, T.; Miladinovic, S. M.; Cheng, L.-Y.; Messner, S.; Ehrenberger, T.; Zanotelli, V.; Butscheid, Y.; Escher, C.; Vitek, O.; Rinner, O.; Reiter, L. Molecular & Cellular Proteomics 2015, 14, 1400-1410. (15) Hopfgartner, G.; Tonoli, D.; Varesio, E. Analytical and Bioanalytical Chemistry 2012, 402, 2587-2596.

AUTHOR INFORMATION Corresponding Author *Tel.: 0041 446355641. Fax: 0041 446356852. E-mail address: [email protected].

(16) Sajic, T.; Liu, Y.; Aebersold, R. PROTEOMICS – Clinical Applications 2015, 9, 307-321. (17) Shao, S.; Guo, T.; Aebersold, R. Biochimica et Biophysica Acta (BBA) - Proteins and Proteomics 2015, 1854, 519-527.

Author Contributions All authors contributed equally to the manuscript. Notes The authors declare no competing financial interest.

(18) Di Girolamo, F.; Del Chierico, F.; Caenaro, G.; Lante, I.; Muraca, M.; Putignani, L. Biomarkers in Medicine 2012, 6, 759773. (19) Zhu, X.; Chen, Y.; Subramanian, R. Analytical chemistry 2013, 86, 1202–1209. (20) Roemmelt, A. T.; Steuer, A. E.; Poetzsch, M.; Kraemer, T. Analytical chemistry 2014, 86, 11742-11749.

REFERENCES (1) Mueller, C. A.; Weinmann, W.; Dresen, S.; Schreiber, A.; Gergov, M. Rapid Commun.Mass Spectrom. 2005, 19, 1332-1338. (2) Dresen, S.; Ferreirós, N.; Gnann, H.; Zimmermann, R.; Weinmann, W. Analytical and Bioanalytical Chemistry 2010, 396, 2425-2434.

(21) Steuer, A. E.; Poetzsch, M.; Koenig, M.; Tingelhoff, E.; Staeheli, S. N.; Roemmelt, A. T.; Kraemer, T. Journal of Chromatography A 2015, 1381, 87-100. (22) Matuszewski, B. K.; Constanzer, M. L.; Chavez-Eng, C. M. Anal. Chem. 2003, 75, 3019-3030.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 9

(23) Peters, F. T.; Drummer, O. H.; Musshoff, F. Forensic Sci.Int. 2007, 165, 216-224.

(42) Macek, J.; Ptacek, P.; Klima, J. J Chromatogr.B Biomed.Sci Appl. 2001, 755, 279-285.

(24) Peters, F. T. In Applications of Liquid Chromatography-Mass Spectrometry in Toxicology, Polettini, A., Ed.; Pharmaceutical Press: London, 2006, pp 71-95.

(43) Gschwend, M. H.; Richter, J.; Sennewald, R.; Guserle, R.; Renner, J.; Martin, W. Arzneimittelforschung. 2005, 55, 730-737.

(25) Siegel, D.; Meinema, A. C.; Permentier, H.; Hopfgartner, G.; Bischoff, R. Analytical chemistry 2014, 86, 5089-5100.

(44) Patel, C.; Patel, M.; Rani, S.; Nivsarkar, M.; Padh, H. Journal of Chromatography B 2007, 850, 356-360.

(26) Peters, F. T.; Hartung, M.; Herbold, M.; Schmitt, G.; Daldrup, T.; Musshoff, F. Toxichem. Krimtech. 2009, 76, 185.

(45) Heller, S.; Hiemke, C.; Stroba, G.; Rieger-Gies, A.; DaumKreysch, E.; Sachse, J.; Hartter, S. Ther.Drug Monit 2004, 26, 459-461.

(27) Bonfiglio, R.; King, R. C.; Olah, T. V.; Merkle, K. Rapid communications in mass spectrometry : RCM 1999, 13, 11751185.

(46) Rao, M. L.; Staberock, U.; Baumann, P.; Hiemke, C.; Deister, A.; Cuendet, C.; Amey, M.; Hartter, S.; Kraemer, M. Clin.Chem. 1994, 40, 929-933.

(28) Peters, F. T.; Drvarov, O.; Lottner, S.; Spellmeier, A.; Rieger, K.; Haefeli, W. E.; Maurer, H. H. Anal.Bioanal.Chem. 2009, 393, 735-745.

(47) Yan, J.-h.; Hubbard, J. W.; McKay, G.; Midha, K. K. Journal of Chromatography B: Biomedical Sciences and Applications 1997, 691, 131-138.

(29) Shinozuka, T.; Terada, M.; Tanaka, E. Forensic Sci. Int. 2006, 162, 108-112.

(48) Norman, T. R.; Maguire, K. P. J Chromatogr. 1985, 340, 173-197.

(30) de Castro, A.; Fernandez, M. d. M. R.; Laloup, M.; Samyn, N.; De Boeck, G.; Wood, M.; Maes, V.; López-Rivadulla, M. Journal of Chromatography A 2007, 1160, 3-12.

(49) Peters, F. T. Anal.Bioanal.Chem. 2007, 388, 1505-1519.

(31) Remane, D.; Meyer, M. R.; Wissenbach, D. K.; Maurer, H. H. Anal Bioanal Chem 2011, 400, 2093-2107. (32) Montenarh, D.; Wernet, M.; Hopf, M.; Maurer, H.; Schmidt, P.; Ewald, A. Analytical and Bioanalytical Chemistry 2014, 406, 5939-5953. (33) del Mar Ramírez Fernández, M.; Wille, S. M. R.; Samyn, N. Ther. Drug Monit. 2012, 34, 11-24.

(50) Nyberg, G.; Martensson, E. Ther.Drug Monit 1986, 8, 478482. (51) Roman, M.; Ström, L.; Tell, H.; Josefsson, M. Analytical and Bioanalytical Chemistry 2013, 405, 4107-4125. (52) Dalsgaard PW, R. A., Rasmussen BS, Bjork MK, Petersen DI, Madsen KA, Gammelgaard B, Simonsen KW and Linnet K. Journal of Forensic Science and Criminology 2013, 1, 1-6. (53) Rosano, T. G.; Na, S.; Ihenetu, K.; Swift, T. A.; Wood, M. Journal of Analytical Toxicology 2014, 38, 495-506.

(34) Rani, S.; Kumar, A.; Malik, A.; Singh, B. Chromatographia 2011, 74, 235-242. (35) Montenarh, D.; Wernet, M. P.; Hopf, M.; Maurer, H. H.; Schmidt, P. H.; Ewald, A. H. Anal Bioanal Chem 2014, 406, 5939-5953. (36) Leis, H. J.; Windischhofer, W.; Raspotnig, G.; Fauler, G. Journal of Mass Spectrometry 2001, 36, 923-928. (37) Massaroti, P.; Cassiano, N. M.; Duarte, L. F.; R., C. D.; Marchioretto, M. A. M.; Bernasconi, G.; Calafatti, S.; Barros, F. A. P.; Meurer, E. C.; Pedrazzoli, J. Journal of Pharmaceutical Sciences 2005, 8, 340-347. (38) Geschke, R.; Körner, J.; Eggers, H. Journal of Chromatography B: Biomedical Sciences and Applications 1987, 420, 111-120. (39) Morgan, P. E.; Tapper, J.; Spencer, E. P. J Chromatogr.B Analyt.Technol Biomed.Life Sci 2003, 798, 211-215. (40) Maris, F. A.; Dingler, E.; Niehues, S. J.Chromatogr.B 1999, 721, 309-316. (41) Lantz, R. J.; Farid, K. Z.; Koons, J.; Tenbarge, J. B.; Bopp, R. J. J Chromatogr. 1993, 614, 175-179.


Page 9 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


For TOC only

9


Liquid Chromatography, In Combination with a Quadrupole Time-of

Recommend Documents