Accepted Manuscript Title: Method validation and application of a liquid chromatography–tandem mass spectrometry method for drugs of abuse testing in exhaled breath Author: Niclas Stephanson S¨oren Sandqvist Marjan Shafaati Lambert Olof Beck PII: DOI: Reference:
S1570-0232(15)00072-0 http://dx.doi.org/doi:10.1016/j.jchromb.2015.01.032 CHROMB 19298
To appear in:
Journal of Chromatography B
Received date: Revised date: Accepted date:
27-10-2014 20-1-2015 24-1-2015
Please cite this article as: N. Stephanson, S. Sandqvist, M.S. Lambert, O. Beck, Method validation and application of a liquid chromatographyndashtandem mass spectrometry method for drugs of abuse testing in exhaled breath, Journal of Chromatography B (2015), http://dx.doi.org/10.1016/j.jchromb.2015.01.032 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Revised manuscript January 2015
Method validation and application of a liquid chromatography–
us
cr
exhaled breath
ip t
tandem mass spectrometry method for drugs of abuse testing in
M
an
Niclas Stephanson, Sören Sandqvist, Marjan Shafaati Lambert, Olof Beck
te
Stockholm, Sweden.
d
Department of Laboratory Medicine, section of Clinical Pharmacology, Karolinska Institutet,
Ac ce p
Correspondence and requests should be addressed to N. Stephanson ([email protected]), Phone +46858583440 Fax +46858585870
Page 1 of 39
2
Abstract A mass spectrometric method for drugs of abuse testing in exhaled breath employing a sampling device collecting aerosol particles was developed and applied in routine use. Analytes covered
ip t
were amphetamine, methamphetamine, 6-acetylmorphine, morphine, cocaine, benzoylecgonine, diazepam, oxazepam and tetrahydrocannabinol. The method involved eluting drugs from the
cr
collection filter with methanol, quantification using deuterated analogues as internal standards,
us
reversed phase chromatography with gradient elution, positive electrospray ionisation and monitoring of two product ions per analyte in selected reaction monitoring mode. The measuring
an
range was 6.0–1000 pg/filter. The intra- and inter-assay imprecision expressed as the coefficient of variation was less than 7%. Influence from matrix was noted for most compounds but was
M
compensated for the use of co-eluting internal standards. The LLOQ was 6.0 pg/filter with intraassay CV 18MΩ/cm) and prepared in-house. The exhaled breath sampling device was obtained from SensAbues AB (Huddinge, Sweden).
d
2.2 Standards and controls
te
The ampoulled stock solutions containing 100 µg/mL amphetamine, methamphetamine, 6-AM, morphine , cocaine, benzoylecgonine, diazepam, oxazepam and 1000 µg/mL THC were diluted
Ac ce p
with methanol to working solutions containing 0.25, 2.5 and 25 ng/mL of each analyte. The working solutions for preparing quality control (QC) were independently diluted. Internal standards were prepared in methanol at a concentration 25 ng/mL from the ampoulled stock solutions of 100 µg/mL.
The prepared solutions were stored at -20 ºC, with a maximum storage time of 3 months.
Calibration samples were prepared by fortifying blank filters with methanol solutions of analytes. These were prepared by adding 8.0-40 µL of methanol solutions containing 0.25-25 ng/mL of analytes. After drying the devices were prepared for analysis as described below (section 2.5). Calibration graphs were constructed using linear regression analysis, with weighting factor 1/x. The concentrations of analytes in unknown samples were determined from
Page 5 of 39
6 the peak area ratio between analyte and internal standard by reference to the calibration graph. Internal standards were used as follows: amphetamine-d5 for both amphetamine and methamphetamine, 6-AM-d3 for 6-AM, morphine-d3 for morphine, cocaine-d3 for cocaine and
ip t
benzoylecgonine, diazepam-d3 for diazepam, oxazepam-d3 for oxazepam and THC-d3 for THC. A zero standard and three calibration levels (18, 54 and 324 pg/filter) were routinely used with a
cr
lower reporting limit of 18 pg/filter applied for all compounds. Three quality control levels: QC low (20 pg/filter); QC middle (100 pg/filter) and QC high (500 pg/filter) as well as blank filter
us
samples were used in every batch. 2.3 Instrumentation
an
The LC-MS/MS system consisted of a Thermo Fisher Scientific Dionex Ultima 3000
M
UHPLC with a Ultimate 2000 SRD degasser, Ultimate 3000 RS binary solvent pump system, column oven, Ultimate 3000 RS auto sampler connected to a Thermo Fisher Scientific TSQ
d
Vantage triple quadrupole mass spectrometer. The system was operated using Chromeleon
te
Xpress (version 3), Trace Finder Clinical Research (version 2.1) and Thermo TSQ Tune Master (version 2.3.0) softwares. The heated electrospray interface (HESI) was used with the instrument
Ac ce p
operating in the positive ion mode. Nitrogen was used as sheath, auxiliary and ion sweep gas, and argon as collision gas. The chromatography was performed using a 1.7-µm 100 2.1 mm (i. d.) Ethylene Bridged Hybrid (BEH) phenyl column (Waters Co), preceded by a 0.2 µm column filter (Waters Co). The liquid chromatography system was operated in a gradient mode with a flow rate of 650 µL/min (Table 1). Solvent A consisted of 4.0 mmol/L ammonium formate with pH adjusted to 10 with 25% ammonia and solvent B was methanol with 4.0 mmol/L ammonium formate (99.6% methanol) and same amount of ammonia added as solvent A. The injection volume was 2.0 µL and the column oven temperature 65°C. Needle wash was performed before and after injection with 100 µL of a mixture consisting of equal volumes of methanol, acetonitrile, 2-propanol and ultra-pure water containing 0.20% ammonia. The total run time of
Page 6 of 39
7 the method was 4.0 min. The following conditions were used in the mass spectrometer: peak width 0.70 Da for Q1 and Q3, collision gas pressure 1.2 mTorr (Argon), capillary temperature 250 °C, vaporizer temperature 450 °C, sheath gas pressure 55 Arb units, ion sweep gas pressure
ip t
1.0 Arb units, aux gas pressure 18.0 Arb units, spray voltage 3000 V, DCV -4 V. The selected ions, S-lens voltage, collision energy and dwell time used for each compound are presented in
cr
Table 2. Acquisition time was 0.50 – 4.0 min with monitoring of each analyte in a time segment of ±0.35 min of expected retention time.
us
2.4 Sampling of exhaled breath
A sampling procedure using a commercial sampling device was employed (SensAbues AB,
an
Huddinge, Sweden). Micro-particles present in the exhaled breath were separately collected by
M
asking subjects to let the exhaled breath pass through mouth piece constructed for separating saliva and larger particles from micro-particles. Micro-particles passing the mouth-piece were
d
collected on an electrostatic polymer filter inside the device [21]. The sampling procedure was
te
standardised by filling of a plastic bag and collecting about 30 L of exhaled breath which took approximately 1-3 min time. Following sampling, the device was sealed with plugs, transported
Ac ce p
at ambient temperature and stored at -20 ºC (maximal storage time before analysis was 1 week). 2.5 Analysis of exhaled breath samples Following transport and storage the collection devices were prepared for analysis by placement onto a 10 mL disposable glass test-tube after removal of the plugs. Following the addition of 75 µL internal standard working solution (0.25 ng of each) the extraction of analytes from the filter surface was performed by addition of 2.0 mL of methanol, wait 5 min, and add 2.5 mL methanol 2 times, sequentially. The extraction eluate was evaporated with nitrogen gas with gentle heating after addition of 20 µL 10% formic acid. The final dry residue was dissolved in 30 µL of methanol and finally 30 µL of 0.10% ammonia was added (total volume 60 µL).
Page 7 of 39
8 The criteria for identification was a relative ion intensity between qualifier and quantifier ions within ± 20% of the target value and a relative retention time between analyte (both ions) and deuterated internal standard within ± 1% of the target value. Target values for ion ratios and
ip t
relative retention times were taken from the calibration standards and updated for each batch. The acceptance criteria for quantifier and qualifier ions were a signal-to-noise ratio of ≥10 and
cr
≥3, respectively. 2.6 Method validation
us
The influence of matrix was evaluated using two experimental designs. The first was post column continuous infusion of a 1.0 µg/mL solution of each analyte in 0.10% ammonia
an
solution/methanol (50% v/v) with a constant flow rate of 10 µL/min. Simultaneously, blank
M
extracts (in duplicate) without internal standards prepared from blank exhaled breath samples and from reference methanol without filter matrix were injected into the LC-MS system while
d
monitoring both SRM transitions selected for all compounds. The second experiment involved
te
comparison of the response for each analyte in 6 replicates of spiked filter matrix extracts (108 pg/filter) from blank exhaled breath samples with the response from 6 reference solutions
Ac ce p
containing the same amount of each analyte. The linearity of the method was estimated by analysis of six replicates of calibration
standards with concentration levels of: 6.0, 18, 54, 108, 324, 648 and 1000 pg/filter and a zero standard in spiked filter matrix.
For estimation of limit of detection (LOD; signal-to-noise ratio of 3 for both quantifier and
qualifier transitions) and the limit of quantification (LOQ; signal-to-noise ratio of 10 for quantifier and at least 3 for qualifier) diluted calibrator extracts were used at concentrations of 2.0, 6.0, 10 and 18 pg/filter. The lower limit of quantification (LLOQ) was experimentally determined by using ten replicates spiked filter extracts based on the total CV
S1570-0232(15)00072-0 http://dx.doi.org/doi:10.1016/j.jchromb.2015.01.032 CHROMB 19298
To appear in:
Journal of Chromatography B
Received date: Revised date: Accepted date:
27-10-2014 20-1-2015 24-1-2015
Please cite this article as: N. Stephanson, S. Sandqvist, M.S. Lambert, O. Beck, Method validation and application of a liquid chromatographyndashtandem mass spectrometry method for drugs of abuse testing in exhaled breath, Journal of Chromatography B (2015), http://dx.doi.org/10.1016/j.jchromb.2015.01.032 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Revised manuscript January 2015
Method validation and application of a liquid chromatography–
us
cr
exhaled breath
ip t
tandem mass spectrometry method for drugs of abuse testing in
M
an
Niclas Stephanson, Sören Sandqvist, Marjan Shafaati Lambert, Olof Beck
te
Stockholm, Sweden.
d
Department of Laboratory Medicine, section of Clinical Pharmacology, Karolinska Institutet,
Ac ce p
Correspondence and requests should be addressed to N. Stephanson ([email protected]), Phone +46858583440 Fax +46858585870
Page 1 of 39
2
Abstract A mass spectrometric method for drugs of abuse testing in exhaled breath employing a sampling device collecting aerosol particles was developed and applied in routine use. Analytes covered
ip t
were amphetamine, methamphetamine, 6-acetylmorphine, morphine, cocaine, benzoylecgonine, diazepam, oxazepam and tetrahydrocannabinol. The method involved eluting drugs from the
cr
collection filter with methanol, quantification using deuterated analogues as internal standards,
us
reversed phase chromatography with gradient elution, positive electrospray ionisation and monitoring of two product ions per analyte in selected reaction monitoring mode. The measuring
an
range was 6.0–1000 pg/filter. The intra- and inter-assay imprecision expressed as the coefficient of variation was less than 7%. Influence from matrix was noted for most compounds but was
M
compensated for the use of co-eluting internal standards. The LLOQ was 6.0 pg/filter with intraassay CV 18MΩ/cm) and prepared in-house. The exhaled breath sampling device was obtained from SensAbues AB (Huddinge, Sweden).
d
2.2 Standards and controls
te
The ampoulled stock solutions containing 100 µg/mL amphetamine, methamphetamine, 6-AM, morphine , cocaine, benzoylecgonine, diazepam, oxazepam and 1000 µg/mL THC were diluted
Ac ce p
with methanol to working solutions containing 0.25, 2.5 and 25 ng/mL of each analyte. The working solutions for preparing quality control (QC) were independently diluted. Internal standards were prepared in methanol at a concentration 25 ng/mL from the ampoulled stock solutions of 100 µg/mL.
The prepared solutions were stored at -20 ºC, with a maximum storage time of 3 months.
Calibration samples were prepared by fortifying blank filters with methanol solutions of analytes. These were prepared by adding 8.0-40 µL of methanol solutions containing 0.25-25 ng/mL of analytes. After drying the devices were prepared for analysis as described below (section 2.5). Calibration graphs were constructed using linear regression analysis, with weighting factor 1/x. The concentrations of analytes in unknown samples were determined from
Page 5 of 39
6 the peak area ratio between analyte and internal standard by reference to the calibration graph. Internal standards were used as follows: amphetamine-d5 for both amphetamine and methamphetamine, 6-AM-d3 for 6-AM, morphine-d3 for morphine, cocaine-d3 for cocaine and
ip t
benzoylecgonine, diazepam-d3 for diazepam, oxazepam-d3 for oxazepam and THC-d3 for THC. A zero standard and three calibration levels (18, 54 and 324 pg/filter) were routinely used with a
cr
lower reporting limit of 18 pg/filter applied for all compounds. Three quality control levels: QC low (20 pg/filter); QC middle (100 pg/filter) and QC high (500 pg/filter) as well as blank filter
us
samples were used in every batch. 2.3 Instrumentation
an
The LC-MS/MS system consisted of a Thermo Fisher Scientific Dionex Ultima 3000
M
UHPLC with a Ultimate 2000 SRD degasser, Ultimate 3000 RS binary solvent pump system, column oven, Ultimate 3000 RS auto sampler connected to a Thermo Fisher Scientific TSQ
d
Vantage triple quadrupole mass spectrometer. The system was operated using Chromeleon
te
Xpress (version 3), Trace Finder Clinical Research (version 2.1) and Thermo TSQ Tune Master (version 2.3.0) softwares. The heated electrospray interface (HESI) was used with the instrument
Ac ce p
operating in the positive ion mode. Nitrogen was used as sheath, auxiliary and ion sweep gas, and argon as collision gas. The chromatography was performed using a 1.7-µm 100 2.1 mm (i. d.) Ethylene Bridged Hybrid (BEH) phenyl column (Waters Co), preceded by a 0.2 µm column filter (Waters Co). The liquid chromatography system was operated in a gradient mode with a flow rate of 650 µL/min (Table 1). Solvent A consisted of 4.0 mmol/L ammonium formate with pH adjusted to 10 with 25% ammonia and solvent B was methanol with 4.0 mmol/L ammonium formate (99.6% methanol) and same amount of ammonia added as solvent A. The injection volume was 2.0 µL and the column oven temperature 65°C. Needle wash was performed before and after injection with 100 µL of a mixture consisting of equal volumes of methanol, acetonitrile, 2-propanol and ultra-pure water containing 0.20% ammonia. The total run time of
Page 6 of 39
7 the method was 4.0 min. The following conditions were used in the mass spectrometer: peak width 0.70 Da for Q1 and Q3, collision gas pressure 1.2 mTorr (Argon), capillary temperature 250 °C, vaporizer temperature 450 °C, sheath gas pressure 55 Arb units, ion sweep gas pressure
ip t
1.0 Arb units, aux gas pressure 18.0 Arb units, spray voltage 3000 V, DCV -4 V. The selected ions, S-lens voltage, collision energy and dwell time used for each compound are presented in
cr
Table 2. Acquisition time was 0.50 – 4.0 min with monitoring of each analyte in a time segment of ±0.35 min of expected retention time.
us
2.4 Sampling of exhaled breath
A sampling procedure using a commercial sampling device was employed (SensAbues AB,
an
Huddinge, Sweden). Micro-particles present in the exhaled breath were separately collected by
M
asking subjects to let the exhaled breath pass through mouth piece constructed for separating saliva and larger particles from micro-particles. Micro-particles passing the mouth-piece were
d
collected on an electrostatic polymer filter inside the device [21]. The sampling procedure was
te
standardised by filling of a plastic bag and collecting about 30 L of exhaled breath which took approximately 1-3 min time. Following sampling, the device was sealed with plugs, transported
Ac ce p
at ambient temperature and stored at -20 ºC (maximal storage time before analysis was 1 week). 2.5 Analysis of exhaled breath samples Following transport and storage the collection devices were prepared for analysis by placement onto a 10 mL disposable glass test-tube after removal of the plugs. Following the addition of 75 µL internal standard working solution (0.25 ng of each) the extraction of analytes from the filter surface was performed by addition of 2.0 mL of methanol, wait 5 min, and add 2.5 mL methanol 2 times, sequentially. The extraction eluate was evaporated with nitrogen gas with gentle heating after addition of 20 µL 10% formic acid. The final dry residue was dissolved in 30 µL of methanol and finally 30 µL of 0.10% ammonia was added (total volume 60 µL).
Page 7 of 39
8 The criteria for identification was a relative ion intensity between qualifier and quantifier ions within ± 20% of the target value and a relative retention time between analyte (both ions) and deuterated internal standard within ± 1% of the target value. Target values for ion ratios and
ip t
relative retention times were taken from the calibration standards and updated for each batch. The acceptance criteria for quantifier and qualifier ions were a signal-to-noise ratio of ≥10 and
cr
≥3, respectively. 2.6 Method validation
us
The influence of matrix was evaluated using two experimental designs. The first was post column continuous infusion of a 1.0 µg/mL solution of each analyte in 0.10% ammonia
an
solution/methanol (50% v/v) with a constant flow rate of 10 µL/min. Simultaneously, blank
M
extracts (in duplicate) without internal standards prepared from blank exhaled breath samples and from reference methanol without filter matrix were injected into the LC-MS system while
d
monitoring both SRM transitions selected for all compounds. The second experiment involved
te
comparison of the response for each analyte in 6 replicates of spiked filter matrix extracts (108 pg/filter) from blank exhaled breath samples with the response from 6 reference solutions
Ac ce p
containing the same amount of each analyte. The linearity of the method was estimated by analysis of six replicates of calibration
standards with concentration levels of: 6.0, 18, 54, 108, 324, 648 and 1000 pg/filter and a zero standard in spiked filter matrix.
For estimation of limit of detection (LOD; signal-to-noise ratio of 3 for both quantifier and
qualifier transitions) and the limit of quantification (LOQ; signal-to-noise ratio of 10 for quantifier and at least 3 for qualifier) diluted calibrator extracts were used at concentrations of 2.0, 6.0, 10 and 18 pg/filter. The lower limit of quantification (LLOQ) was experimentally determined by using ten replicates spiked filter extracts based on the total CV
