Quantitation of Benzodiazepines in Urine, Serum, Plasma, and

0 downloads 0 Views 187KB Size Report
(10 to 2500 ng/mL) and within ± 20% for meconium (10 to 5000 ng/g). In all, 205 patient specimens were analyzed, and the results compared to a previous ...
Journal of Analytical Toxicology, Vol. 32, September 2008

Quantitation of Benzodiazepines in Urine, Serum, Plasma, and Meconium by LC–MS–MS Stephanie J. Marin1,*, Rebecka Coles2, Miles Merrell2, and Gwendolyn A. McMillin2,3 1ARUP

Institute for Clinical and Experimental Pathology, Salt Lake City, Utah; 2ARUP Laboratories, Inc., Salt Lake City, Utah; of Pathology, University of Utah School of Medicine, Salt Lake City, Utah

3Department

Abstract A single method for confirmation and quantitation of a panel of commonly prescribed benzodiazepines and metabolites, α-hydroxyalprazolam, α-hydroxyethylflurazepam, α-hydroxytriazolam, alprazolam, desalkylflurazepam, diazepam, lorazepam, midazolam, nordiazepam, oxazepam, temazepam, clonazepam, and 7-aminoclonazepam, was developed for three specimen types, urine, serum/plasma, and meconium. Quantitation was by liquid chromatography tandem–mass spectrometry (LC–MS–MS) using a Waters Alliance-Quattro Micro system. The instrument was operated in multiple reaction monitoring mode with an electrospray ionization source in positive ionization mode. The method was evaluated for recovery, imprecision, linearity, analytical measurement range, specificity, and carryover. Average recovery and imprecision (within-run, between-run, and total % CV) were within ± 15% of the target concentrations for urine (10 to 5000 ng/mL) and serum/plasma (10 to 2500 ng/mL) and within ± 20% for meconium (10 to 5000 ng/g). In all, 205 patient specimens were analyzed, and the results compared to a previous in-house gas chromatography–MS method or LC–MS–MS results from an outside laboratory. Oxazepam glucuronide was evaluated as a hydrolysis control for the urine and meconium specimens.

Introduction Benzodiazepines are a schedule IV class of psychotropic drugs prescribed for their sedative, anxiolytic, and anticonvulsant properties. Indeed, of the top 200 generic drugs ranked by number of prescriptions written in 2006 (1), five were benzodiazepines: alprazolam was ranked number 7, lorazepam 17, clonazepam 23, diazepam 41, and temazepam 60. Physiological and psychological dependence can lead to misuse and abuse. The analytical method described here was designed to support medical (not forensic) purposes. The detection of benzo* Author to whom correspondence should be addressed. Stephanie J. Marin, ARUP Institute for Clinical and Experimental Pathology, ARUP Laboratories, Inc., 500 Chipeta Way, Salt Lake City, UT 84108-1221. E-mail: [email protected].

diazepines in urine is clinically important for compliance monitoring and identification of abuse. Quantitation in serum or plasma may help optimize chronic dosing, verify compliance, and identify changes in pharmacokinetics. Detection in meconium can identify neonates exposed to drugs during the prenatal period to guide treatment and improve outcomes for children exposed to drugs in utero. To operate at maximum efficiency and increase throughput, it was desirable to have a single method in our laboratory for the preparation and analysis of commonly used benzodiazepines and their metabolites of interest in clinical applications. Urine is composed of over 95% water, plus sodium, ammonia, phosphate, sulfate, urea, creatinine, proteins, and products processed by the kidney and liver, including drugs and metabolites. Serum and plasma are over 90% water and also contain ions, dissolved gases, proteins, hormones, nutrients, and tissue products like creatinine, urea, and lactate, in addition to parent drugs and their metabolites. Meconium is a very complex sample matrix consisting of water, epithelial cells, lanugo, mucus, amniotic fluid, bile acids and salts, fatty material from the vernix caseosa, cholesterol and sterol precursors, blood group substances, enzymes, mucopolysaccharides, sugars, lipids, proteins, trace metals, various pancreatic and intestinal secretions, and drugs or other materials ingested by the mother (2). These three specimen types have unique compositions that can lead to different analytical interferences. Benzodiazepines are extensively metabolized. Many have common metabolites, and many of the metabolites are also prescribed drugs (3). Most are eliminated to a considerable extent as a glucuronide conjugate. Enzyme hydrolysis of urine and meconium specimens to convert the analytes to their free form improves detection. Numerous methods have been reported in the literature and consolidated in several review articles that describe the quantitation of benzodiazepines in urine or blood using liquid chromatography–tandem mass spectrometry (LC–MS–MS) (4–14), but no work on the analysis of benzodiazepines in meconium has been published. We have developed and validated a single method for the identification and quantitation of 13 benzodiazepines and metabolites in urine, serum/plasma, and meconium using LC–MS–MS: α-hydroxyalprazolam,

Reproduction (photocopying) of editorial content of this journal is prohibited without publisher’s permission.

491

Journal of Analytical Toxicology, Vol. 32, September 2008

α-hydroxyethylflurazepam, α-hydroxytriazolam, alprazolam, desalkylflurazepam, diazepam, lorazepam, midazolam, nordiazepam, oxazepam, temazepam, clonazepam, and 7-aminoclonazepam. This replaces a gas chromatography (GC)–MS method that quantitated only seven benzodiazepines in urine. Alprazolam, desalkylflurazepam, diazepam, midazolam, clonazepam, 7-aminoclonazepam, and two additional specimen types were added to this new LC–MS–MS assay. We have also included the use of oxazepam glucuronide as a control to monitor the effectiveness of enzyme hydrolysis on urine and meconium specimens.

Experimental Reagents and standards

All calibration standards and internal standards were of 99% purity and purchased from Cerilliant (Austin, TX). All solvents were reagent or HPLC grade and purchased from ThermoFisher Scientific (Waltham, MA) or VWR (West Chester, PA). Nanopure water was generated using a Barnstead Nanopure Infinity ultra-pure water system (Thermo-Fisher Scientific). Trace-B columns, 35 mg/3 mL, were supplied by SPEWare (San Pedro, CA). Plasma was obtained with potassium oxalate or sodium fluoride preservative. Serum was obtained without preservatives. Drug-free urine, serum, plasma, and meconium pools were prepared from excess patient specimens that tested negative by EMIT and LC–MS–MS and were used to prepare calibrators, controls, and fortified samples. Benzodiazepines were quantitated in residual patient specimens by GC–MS or another LC–MS–MS method, and the specimens were de-identified to protect personal health information. Apparatus

Homogenation of the meconium specimens was performed with an Omni Tissuemiser Homogenizer (Thermo-Fisher Scientific). Solvent evaporation after initial sample preparation was achieved with a Jouan centrifugal vacuum evaporator (CVE) system, model RC10.10 (Thermo-Fisher Scientific). Extraction was performed on a 48-place Cerex positive pressure manifold (SPEWare). A CEREX 48-place sample concentrator station (SPEWare) was used for final solvent evaporation. Samples were eluted into max recovery autosampler vials (VWR). Instrumental analysis

LC–MS–MS analysis was performed on a Waters/Micromass Quattro Micro LC–MS–MS system equipped with a Waters Alliance® HT HPLC system (Milford, MA). The HPLC system included a solvent delivery/separation module, sample management system (autosampler), and column oven. The instrument was operated in multiple reaction monitoring (MRM) mode with an electrospray ionization (ESI) probe in positive ESI mode. The LC–MS–MS instrument was controlled from a desktop computer using Micromass MassLynx software. The MS source temperature was maintained at 120°C, and the desolvation temperature was 400°C with a nitrogen desolvation gas flow of 800 L/h. The resolution of both

492

quadrupoles was maintained at unit mass resolution with a peak width at half height of 0.7 amu.

Methods Sample preparation

Meconium (1.00 ± 0.02 g) was weighed into a 16 × 50-mm polypropylene tube, followed by the addition of 3 mL of methanol. Each meconium specimen was homogenized until uniform, then centrifuged at 14,000 rpm and 0°C for 15 min. The supernatant was transferred to 16 × 100-mm culture tubes and the solvent was evaporated to < 1 mL under vacuum in the CVE at 60°C. Urine specimens were prepared by measuring 1 mL of each specimen into a labeled culture tube. Serum and plasma specimens were prepared by measuring 1 mL of each specimen into a culture tube and adding 2 mL of 0.1 M sodium acetate buffered to pH 5.0. Calibrators (20, 50, and 200 ng/mL or ng/g) prepared in urine and meconium were prepared and extracted with each batch of patient specimens. Serum/plasma specimens were quantitated using the urine calibrators. Deuterated analogues were included as internal standards at 100 ng/mL or ng/g for all 13 analytes: α-hydroxyalprazolam-d5, α-hydroxyethylflurazepam-d4, α-hydroxytriazolam-d4, alprazolam-d5, desalkylflurazepam-d4, diazepam-d5, lorazepam-d4, midazolam-d 4 maleate, nordiazepam-d 5 , oxazepam-d 5 , temazepam-d5, clonazepam-d4, and 7-aminoclonazepam-d4. Matrix-matched positive controls containing all 13 analytes at 50 ng/mL or ng/g and negative controls containing only internal standard at 100 ng/mL or ng/g were prepared and extracted for all specimen types. Hydrolysis controls containing oxazepam glucuronide were included for the urine and meconium specimens. The fortified concentration for the oxazepam glucuronide control was 81 ng, which would yield 50 ng of oxazepam after hydrolysis. The extraction and analysis of the meconium concentrate continued with the appropriate solidphase extraction method that follows in the same manner as the urine, serum, and plasma specimens. Hydrolysis and solid-phase extraction

Two milliliters of β-glucuronidase enzyme from bovine liver [5000 units/mL in 0.1 M sodium acetate buffer (pH 5.0)] was added to each urine and meconium sample. The samples were briefly vortex mixed, and then urine and meconium samples were incubated at 60°C for 2 h. All samples were then centrifuged at 3500 rpm and 0°C for 5 min. The samples were loaded onto the Trace-B columns at 4 drops per second. The samples were washed at 1 drop per second with 1 mL each of pH 9.0 sodium bicarbonate buffer, then water. The samples were dried on the columns using nitrogen at 25 psi for 10 min. The samples were then eluted into 2-mL max recovery autosampler vials with 1 mL 98:2 ethyl acetate/ammonium hydroxide. Samples were dried in the autosampler vials using the CEREX 48-place sample concentrator with nitrogen at 40°C for approximately 15 min. The residue was reconstituted in 200 µL 1:1 acetonitrile/water (150 µL for meconium samples), capped, vortex mixed, and analyzed by LC–MS–MS.

Journal of Analytical Toxicology, Vol. 32, September 2008

LC–MS–MS analysis

A Waters XTerra® MS C18 analytical column (3.5-µm particle size, 2.1 × 150 mm) was heated to a constant 30°C in the column heater. The LC flow was isocratic at 0.125 mL/min with 55% acetonitrile, 40% nanopure water, and 5% 100 mM ammonium formate buffer at pH 3.0. The ions monitored with the respective cone energy and collision voltage for all analytes are listed in Table I. Method validation

The method was validated for recovery, imprecision, linearity, analytical measurement range, specificity, carryover, and ion suppression. Accuracy (recovery) was determined by analyzing samples fortified at 15 concentrations from the limit of determination (LOD) to the upper limit of quantitation (ULOQ). Two of each concentration were analyzed for urine and serum/plasma samples (n = 30 per analyte), one fortified sample was analyzed for the meconium specimens (n = 15 per analyte). The concentrations for urine and meconium samples were 10, 20, 25, 50, 75, 100, 150, 200, 300, 500, 1000, 1250, 1500, 2500, and 5000 ng/mL or ng/g. The concentrations for the serum/plasma samples were 10, 20,

50, 75, 100, 150, 200, 300, 400, 500, 750, 1000, 1250, 1500, and 2500 ng/mL. Positive patient specimen results from this method were compared with the previous validated GC–MS method or an LC–MS–MS method at another certified laboratory. Correlation of results was determined using 205 residual patient specimens (which were de-identified prior to use to protect personal health information), many of which contained more than one benzodiazepine. Thus, the total number of data points for the patient specimen correlation was 337. Linearity was determined from the same fortified sample data used for the accuracy determination. Average recovery, total imprecision, within-run imprecision, and between-run imprecision [% coefficients of variation (CV)] were determined by analyzing samples fortified with all 13 analytes and internal standards at the lower limit of quantitation (LLOQ) (20 ng/mL or ng/g), the high calibrator (200 ng/mL or ng/g), and the ULOQ (5000 ng/mL for urine, 2500 ng/mL for serum/plasma, and 5000 ng/g for meconium) in triplicate on three different days (nine samples per concentration for each specimen type). Fortified samples at 100,000 ng/mL or ng/g were used to determine carryover.

Table I. Retention Times (min), MRM Transitions Monitored (m/z), Cone and Collision Voltages for Benzodiazepines Analyte

RT

MRM

Cone

Collision

Nordiazepam

6.02

271.00 > 140.00 271.00 > 208.10

35 35

25 25

Oxazepam

4.93

286.95 > 241.05 286.95 > 104.00

25 25

Temazepam

5.95

300.95 > 255.05 300.95 > 193.15

Diazepam

7.75

Lorazepam

Internal Standard

RT

MRM

Nordiazepam-d5

5.98

276.00 > 140.05 276.00 > 213.15

35 35

25 30

25 35

Oxazepam-d5

4.93

295.00 > 246.10 295.00 > 109.05

25 25

25 35

20 20

20 35

Temazepam-d5

5.95

306.00 > 260.10 306.00 > 198.15

20 20

20 35

285.00 > 154.05 285.00 > 193.10

35 35

25 30

Diazepam-d5

7.59

290.05 > 154.05 290.05 > 198.15

35 35

25 30

5.03

320.90 > 229.10 320.90 > 194.10

25 25

30 40

Lorazepam-d4

5.03

324.95 > 233.10 324.95 > 198.15

25 25

30 40

Alprazolam

4.93

309.00 > 205.10 309.00 > 274.25

35 35

40 25

Alprazolam-d5

4.93

314.00 > 210.15 314.00 > 279.25

35 35

40 25

α-OH-Alprazolam

4.48

325.00 > 297.00 325.00 > 205.10

35 35

25 45

α-OH-Alprazolam-d5

4.48

330.00 > 302.05 330.00 > 210.15

35 35

25 45

α-OH-Triazolam

4.41

358.95 > 176.05 358.95 > 277.00

35 35

25 35

α-OH-Triazolam-d4

4.41

362.95 > 176.05 362.95 > 281.00

35 35

25 35

α-OH-Ethylflurazepam

5.12

333.00 > 109.00 333.00 > 211.10

30 30

25 35

337.00 > 113.05 337.00 > 215.15

30 30

25 35

Desalkylflurazepam

5.64

289.00 > 140.05 289.00 > 226.10

35 35

30 25

Desalkylflurazepam-d4

5.64

293.00 > 140.05 293.00 > 230.15

35 35

30 30

Midazolam

4.03

326.00 > 291.05 326.00 > 249.15

35 35

25 35

Midazolam-d4

4.03

330.00 > 295.20 330.10 > 227.10

35 35

25 35

Clonazepam

5.23

316.20 > 214.20 316.20 > 241.20

35 35

35 35

Clonazepam-d4

5.23

320.20 > 274.10 320.00 > 245.10

35 35

25 35

7-Aminoclonazepam

3.93

286.40 > 222.10 286.40 > 250.10

35 35

25 20

7-Aminoclonazepam-d4

3.93

290.40 > 226.10 290.40 > 254.10

35 35

25 20

α-OH-Ethylflurazepam-d4 5.12

Cone Collision

493

Journal of Analytical Toxicology, Vol. 32, September 2008

LOD was the lowest concentration for which the analyte met all qualitative criteria for correct identification (retention time ± 2%, all monitored ions present with acceptable ion ratios) and the peak of interest had a signal-to-noise ratio that exceeded 3:1. LLOQ was defined as the lowest concentration which could be identified and quantitated, all qualitative criteria were met, and the peak of interest had a signal-to-noise ratio greater than 5:1. ULOQ was the highest concentration which could be identified and quantitated, all qualitative criteria were met, and the accuracy and imprecision were ± 15% for urine and serum/plasma and ± 20% for meconium. Analytical measurement range, linearity, recovery, and accuracy of the fortified samples required that all concentrations were within ± 15% of target concentration for urine and serum/plasma, and ± 20% for meconium, and all qualitative

criteria were met. Linear regression data for the fortified samples had slopes that were between 0.85 and 1.15, the y-intercepts were less than the LLOQ, and the R2 were greater than 0.99. Average recovery was the mean of all samples at a specific target concentration for each specimen type (nine results per concentration per analyte for each specimen type). Average within-run CV were determined from the average of the standard deviations for each day at each target concentration. Between-run CV were reported as the standard deviation of the average concentrations for each day at each target concentration. Total CV was the square root of the sum of the squares of the CV for within-run and between-run imprecision at each target concentration. Imprecision also required CV within ± 15% (± 20% for meconium), and concentrations within ± 15% (± 20% for meconium) of the target concentration. Ion mass

Table II. Linear Correlation Data for Fortified Samples* Urine Equation y = 0.938x + 12.310 y = 1.004x + 11.050 y = 0.922x + 16.090 y = 0.888x + 8.727 y = 1.011x – 2.602 y = 1.061x – 11.280 y = 1.042x + 10.399 y = 0.968x + 9.431 y = 1.029x – 3.245 y = 1.053x – 14.844 y = 1.146x – 31.166 y = 0.993x + 16.124 y = 1.125x – 25.450

0.9992 0.9995 0.9995 0.9995 0.9988 0.9989 0.9994 0.9970 0.9998 0.9995 0.9989 0.9981 0.9977

Analyte Nordiazepam Oxazepam Temazepam Diazepam Lorazepam Alprazolam α-OH-Alprazolam α-OH-Triazolam α-OH-Ethylflurazepam Desalkylflurazepam Midazolam Clonazepam 7-Aminoclonazepam

Plasma R2

Meconium

Equation

R2

Equation

R2

y = 0.983x + 10.780 y = 0.968x + 11.330 y = 0.945x + 16.795 y = 0.969x + 0.060 y = 0.962x – 4.070 y = 0.980x + 10.018 y = 1.018x + 3.552 y = 0.988x + 10.713 y = 0.963x + 16.052 y = 1.007x + 5.828 y = 1.001x + 10.230 y = 0.979x + 11.322 y = 1.097x – 15.383

0.9982 0.9980 0.9970 0.9919 0.9913 0.9985 0.9992 0.9995 0.9962 0.9998 0.9996 0.9990 0.9986

y = 0.957x – 18.868 y = 0.966x – 27.317 y = 0.845x + 5.167 y = 0.869x + 2.263 y = 1.006x – 4.251 y = 0.958x – 2.588 y = 0.992x – 27.292 y = 0.947x – 6.016 y = 0.982x – 27.107 y = 0.929x – 2.911 y = 0.978x – 26.091 y = 0.950x + 8.937 y = 1.102x – 41.644

0.9967 0.9943 0.9990 0.9981 0.9966 0.9992 0.9959 0.9988 0.9959 0.9990 0.9951 0.9987 0.9971

* All linear correlation data had slopes between 0.85 and 1.15, y-intercepts less than 20 ng/mL or 20 ng/g (LLOQ), and a correlation coefficient greater than 0.99.

Table III. Average Recovery (%)* Urine

Serum/Plasma

Meconium

Analyte

20 ng/mL

200 ng/mL

5000 ng/mL

20 ng/mL

200 ng/mL

2500 ng/mL

20 ng/g

200 ng/g

5000 ng/g

Nordiazepam Oxazepam Temazepam Diazepam Lorazepam Alprazolam α-OH-Alprazolam α-OH-Triazolam α-OH-Ethylflurazepam Desalkylflurazepam Midazolam Clonazepam 7-Aminoclonazepam

99.3 98.7 105.1 105.1 97.1 101.6 109.7 97.7 106.0 100.4 109.4 111.0 104.4

107.8 99.6 105.2 106.7 98.7 102.1 110.5 97.2 103.1 105.2 110.6 107.5 104.1

90.0 86.9 85.5 87.3 94.9 89.7 98.3 86.5 89.5 91.9 101.0 87.9 109.4

110.1 96.1 107.8 106.1 95.4 104.2 109.3 100.3 100.3 102.8 110.0 112.7 103.2

102.9 95.8 98.4 96.2 94.9 98.8 107.7 97.6 96.6 100.6 110.5 105.2 102.3

92.9% 89.5 88.4 87.6 88.9 93.7 103.4 91.5 91.1 95.1 101.6 89.0 107.0

102.5 97.1 101.8 102.3 97.4 101.6 98.7 99.2 100.5 101.1 104.0 113.0 103.9

102.6 104.2 101.5 102.4 103.9 103.3 106.6 100.7 103.7 101.2 107.7 103.5 104.2

92.1 91.6 82.1 83.4 96.1 92.5 97.2 87.3 96.4 89.0 95.4 87.2 99.4

* Average recovery was the mean of all samples at a specific target concentration for each specimen type (nine results per concentration for each specimen type).

494

Journal of Analytical Toxicology, Vol. 32, September 2008

ratios (IMR) were within ± 20% or better for all analytes and all specimen types.

patient specimens and fortified samples were quantitated using calibrators prepared in urine. Data from the fortified samples and patient specimen correlation data demonstrate that the quantitation of serum/plasma specimens using urine calibrators was accurate and reliable. Urine was chosen as the matrix for the calibrators because the majority of specimens received in our laboratory are urine specimens. A plasma positive control fortified at 50 ng/mL was included with all batches. Urine and plasma positive controls were within ± 10% of the target concentration with %CV within 10% for all analytes, and all values were within two standard deviations of the mean during the time period data were collected (60 days). Accurate serum/plasma concentrations are important in therapeutic drug monitoring and pharmacokinetics. Serum or plasma specimens containing benzodiazepines at concentrations greater than 500 ng/mL are rarely seen in our laboratory, so the lower ULOQ for serum/plasma is well above the expected clin-

Results and Discussion The LOD for all analytes and all specimen types was 10 ng/mL or ng/g or less. The LLOQ for all analytes and specimen types was selected to be equal to two times the LOD (20 ng/mL or ng/g). The LOD of many of the analytes were below 10 ng/mL or ng/g. A blanket LOD of 10 ng/mL or ng/g and an LLOQ of 20 ng/mL or ng/g was validated for all analytes and all specimen types for continuity and ease of reporting. The ULOQ was 5000 ng/mL for urine, 2500 ng/mL for serum/plasma, and 5000 ng/g for meconium. Validated dilutions were up to 25-fold for serum/plasma and up to 50-fold for urine. Serum/plasma Table IV. Total Imprecision (%CV) Urine

Serum/Plasma

Meconium

Analyte

20 ng/mL

200 ng/mL

5000 ng/mL

20 ng/mL

200 ng/mL

2500 ng/mL

20 ng/g

200 ng/g

5000 ng/g

Nordiazepam Oxazepam Temazepam Diazepam Lorazepam Alprazolam α-OH-Alprazolam α-OH-Triazolam α-OH-Ethylflurazepam Desalkylflurazepam Midazolam Clonazepam 7-Aminoclonazepam

11.2 8.5 9.7 9.7 8.0 7.0 5.8 7.5 9.1 7.8 3.7 6.3 9.5

6.9 8.4 10.8 7.9 7.9 6.7 3.8 9.9 6.2 8.3 4.4 3.1 7.8

4.7 3.4 0.6 2.5 4.9 3.3 3.1 5.0 5.2 5.6 8.3 4.4 6.4

4.5 14.7 8.2 8.5 14.0 9.1 7.5 14.0 7.5 9.9 2.5 2.0 9.1

9.6 9.8 12.4 7.3 8.3 6.8 5.3 12.1 9.3 11.6 4.2 5.5 8.9

7.1 3.3 3.3 4.8 3.4 3.1 4.5 4.3 3.1 4.2 8.6 3.6 9.0

10.0 12.1 8.6 9.5 15.1 11.4 9.5 11.2 13.9 11.2 9.7 9.9 16.8

5.1 5.4 5.3 4.7 5.7 7.1 6.1 9.1 5.8 7.1 5.9 6.0 8.2

5.3 7.5 4.0 4.0 16.4 6.5 8.9 6.0 9.0 5.5 6.7 4.4 13.6

Table V. Average Within-Run Imprecision (%CV) Urine Analyte Nordiazepam Oxazepam Temazepam Diazepam Lorazepam Alprazolam α-OH-Alprazolam α-OH-Triazolam α-OH-Ethylflurazepam Desalkylflurazepam Midazolam Clonazepam 7-Aminoclonazepam

Serum/Plasma

Meconium

20 ng/mL

200 ng/mL

5000 ng/mL

20 ng/mL

200 ng/mL

2500 ng/mL

20 ng/g

200 ng/g

5000 ng/g

2.9 2.0 2.4 1.8 3.0 2.4 1.4 3.5 6.2 3.7 2.1 5.9 5.8

3.2 3.3 2.1 2.0 4.9 3.6 3.5 2.6 2.7 3.7 2.4 2.9 3.8

2.0 0.9 0.4 1.8 4.5 1.4 2.0 3.5 4.4 3.6 1.9 3.4 3.5

2.6 2.0 3.3 2.2 2.6 3.9 1.5 3.9 4.6 4.4 2.4 1.3 3.7

4.3 2.9 3.5 4.2 3.6 4.2 5.2 3.1 5.1 6.5 3.8 4.3 2.8

3.9 3.2 2.6 3.8 3.2 1.9 3.2 4.2 2.2 3.3 2.9 2.7 3.1

4.0 5.7 6.2 6.5 7.2 4.9 4.7 5.2 9.6 8.7 3.5 8.3 8.5

3.8 3.6 4.3 4.2 4.1 4.3 3.5 4.0 4.5 5.4 5.2 3.5 7.1

2.8 6.8 2.4 3.8 14.5 2.6 4.9 5.6 4.0 2.5 4.8 3.2 3.4

495

Journal of Analytical Toxicology, Vol. 32, September 2008

ical range. Urine concentrations are often seen in excess of 5000 ng/mL; therefore, a dilution must be performed if accurate quantitation is required. A 50-fold dilution provides a clinically reportable range of 10–250,000 ng/mL for urine specimens. Meconium positive controls were also within ± 10% of the target concentration with %CV within 10% for all analytes. All values were within two standard deviations of the mean. Accurate concentrations in meconium have little clinical significance because the correlation between concentrations and details of drug use and exposure are not well defined. It is, however, clinically important to provide evidence of drug exposure to health care providers as soon as possible, so it is also important to have a fast, accurate, reliable, and robust method. Linear correlation data for the fortified samples at 15 concentrations from the LOD to the ULOQ are listed in Table II. Results for average recovery of samples fortified at the LLOQ (20 ng/mL or ng/g), high calibrator (200 ng/mL or ng/g), and ULOQ (2500 or 5000 ng/mL and 5000 ng/g) are listed in Table III. Total imprecision is reported in Table IV. Within-run and between-run imprecision are reported in Tables V and VI, respectively. Linear correlation results for positive patient specimens are plotted in Figures 1 and 2. Figure 1 shows the results of 167 urine specimens compared to the previous in-house validated GC–MS method. The slope of the line of the best least-squares fit was 0.927 with a y-intercept of –3.534. The correlation coefficient (R2) was 0.971. Total positive results were n = 256 because some specimens were positive for more than one benzodiazepine. The benzodiazepines quantitated in urine specimens were α-OH-alprazolam (n = 110), lorazepam (n = 38), oxazepam (n = 49), nordiazepam (n = 24), and temazepam (n = 35). Additional benzodiazepines (alprazolam, desalkylflurazepam, diazepam, midazolam, clonazepam, and 7aminoclonazepam) not included in the original GC–MS method were identified and quantitated in 24 urine patient specimens using the new LC–MS–MS method.

Figure 2 shows results of 21 urine and 17 serum or plasma specimens comparing the new method to a validated LC–MS–MS method from a certified outside laboratory. The benzodiazepines quantitated in the 17 serum/plasma specimens by LC–MS–MS were alprazolam (n = 5), diazepam (n = 4), nordiazepam (n = 4), lorazepam (n = 2), temazepam (n = 1), clonazepam (n = 4), and midazolam (n = 1). The benzodiazepines quantitated in the 21 urine specimens by LC–MS–MS were alprazolam (n = 5), α-OH-alprazolam (n = 2), clonazepam (n = 3), 7-aminoclonazepam (n = 9), lorazepam (n = 6), nordiazepam (n = 8), oxazepam (n = 11), and temazepam (n = 9). The slope of the line of the best least-squares fit was 1.002, the y-intercept was 6.077, and the correlation coefficient was 0.992. Carryover was defined as being present if the analyte was observed at a concentration above the LOD in a blank preceded by a sample fortified at 100,000 ng/mL or ng/g and all qualitative criteria were met. No carryover was observed for any of the analytes. Ion suppression was evaluated for each matrix by extracting a blank urine, plasma, and meconium sample and injecting the sample while infusing a sample containing all 13 analytes and internal standards. Figure 3 shows the total ion chromatograms (TIC) for an unextracted control (A) and the blank urine (B), plasma (C), and meconium (D) injected during infusion. All three specimen types showed very minimal ion suppression. None of the observed ion suppression caused a loss in sensitivity significant enough to interfere with accurate quantitation. The use of a deuterated internal standard for all 13 analytes compensates for the minimal ion suppression that occurs. Previously accepted laboratory values for stability were accepted for this validation. Patient specimens were collected and frozen for analysis over the course of one year. Specimens stored at ambient temperature appeared to be stable for one week. Refrigerated urine specimens were stable for one month, serum/plasma for two weeks, and meconium specimens for

Table VI. Average Between-Run Imprecision (%CV)* Urine

Serum/Plasma

Meconium

Analyte

20 ng/mL

200 ng/mL

5000 ng/mL

20 ng/mL

200 ng/mL

2500 ng/mL

20 ng/g

200 ng/g

5000 ng/g

Nordiazepam Oxazepam Temazepam Diazepam Lorazepam Alprazolam α-OH-Alprazolam α-OH-Triazolam α-OH-Ethylflurazepam Desalkylflurazepam Midazolam Clonazepam 7-Aminoclonazepam

10.9 8.3 9.4 9.6 7.4 6.6 5.7 6.6 6.7 6.9 3.1 2.1 7.6

6.1 7.7 10.6 7.6 6.3 5.7 1.3 9.5 5.6 7.4 3.7 1.1 6.8

4.3 3.3 0.4 1.8 2.1 3.0 2.4 3.5 2.8 4.3 8.1 2.8 5.4

3.7 14.5 7.6 8.2 13.7 8.2 7.4 13.4 5.9 8.9 0.7 1.6 8.4

8.6 9.4 11.9 6.0 7.5 5.4 1.1 11.7 7.8 9.6 1.8 3.4 8.5

6.0 0.9 2.0 2.9 1.2 2.4 3.1 0.5 2.2 2.7 8.1 2.3 8.4

9.1 10.7 6.0 7.0 13.3 10.3 8.3 9.9 10.1 7.1 9.0 5.3 14.4

3.4 4.0 3.0 2.0 4.0 5.6 5.0 8.1 3.7 4.6 2.8 4.9 4.2

4.5 3.2 3.2 1.4 7.6 5.9 7.4 2.2 8.1 4.8 4.7 3.0 13.2

* Average between-run imprecision was the standard deviation of the average concentrations for each day at each target concentration. Within-run imprecision also required CVs within ± 15% (20% for meconium) of the target concentration.

496

Journal of Analytical Toxicology, Vol. 32, September 2008

Figure 1. Patient specimen correlation data for 167 urine specimens. Results from the new LC–MS–MS method were compared to results from the previous in-house GC–MS method.

three months. Frozen serum/plasma and meconium specimens were stable for one year, and frozen urine specimens were stable for three years. Stability of extracted samples in autosampler vials stored at 4°C was at least 72 h for all specimen types. The hydrolysis control, oxazepam glucuronide, was not available when the GC–MS method was developed. A hydrolysis control fortified to yield 50 ng/mL or ng/g after enzyme hydrolysis was included with each batch of samples. Average recovery was 46.9 ± 5.2 (93.9%) for urine (n = 50) and 50.2 ± 2.7 (100.4%) for meconium (n = 25). Two urine values were slightly high (115.6% and 116% recovery) and outside the range of two standard deviations, and one value was lower than two standard deviations with only 60.6% recovery. One meconium value was outside the range of two standard deviations at 111.8% recovery. Oxazepam glucuronide was determined to be an effective control to monitor the enzyme hydrolysis of the urine and meconium specimens.

Conclusions

Figure 2. Patient specimen correlation data for 17 serum/plasma and 21 urine specimens. Results from the new LC–MS–MS method were compared to results from an LC–MS–MS method from a nationally certified reference laboratory.

A single method for the confirmation of 13 benzodiazepines in 3 specimen types was developed and validated. Calibrators prepared in urine were successfully used to quantitate the serum/plasma specimens. Accuracy and imprecision of fortified samples were ± 15% for urine and serum/plasma, and ± 20% for meconium. Results for 337 positive results (205 positive patient specimens) had good agreement with the previous inhouse GC–MS method or analysis by an outside laboratory using LC–MS–MS. A hydrolysis control demonstrated that enzyme hydrolysis averaged 94% complete for urine and 100% complete for meconium.

Acknowledgments Funding, instrumentation, and physical facilities to conduct this research were provided by the ARUP Institute for Clinical and Experimental Pathology™ and ARUP Laboratories, Inc. The authors are grateful to Heidi Carlisle, Lori Brophy, and Bryan Lawlor for performing some of the urine and serum/ plasma extractions. They also thank Natalie Rasmussen and Jared Allred for assistance with the meconium extractions and Mario Gonzales and Amberly Johnson for collecting the LC–MS–MS data. Anna Christensen prepared the calibrators and controls. Erin Campbell compiled the urine, plasma, and meconium positive control data.

Figure 3. Ion suppression study showing an unextracted control fortified at 100 ng/mL or ng/g of each analyte and 500 ng/mL or ng/g of each internal standard (A) and a blank urine (B), plasma (C), and meconium (D) sample injected while infusing a sample containing all of the analytes and internal standards.

References 1. Top 200 generic drugs by units in 2006. Drug Topics, March 5, 2007. 2. J. Gareri, J. Klein, and G. Koren. Drugs of abuse testing in meconium. Clin. Chim. Acta 366: 101–111 (2006).

497

Journal of Analytical Toxicology, Vol. 32, September 2008

3. R.C. Baselt. Disposition of Toxic Drugs and Chemicals in Man, 7th ed. Biomedical Publications, Foster City, CA, 2004. 4. O.H. Drummer. Methods for the measurement of benzodiazepines in biological samples. J. Chromatogr. B Biomed. Sci. Appl. 713: 201–225 (1998). 5. H.M. Rivera, G.S. Walker, D.N. Sims, and P.C. Stockholm. Application of liquid chromatography–tandem mass spectrometry to the analysis of benzodiazepines in blood. Eur. J. Mass Spectrom. 9: 599–607 (2003). 6. B.E. Smink, J.E. Brandsma, A. Dijkhuizen, K.J. Lusthof, J.J. Gier, A.C.G. Egberts, and D.R.A. Uges. Quantitative analysis of 22 benzodiazepines, metabolites, and benzodiazepine-like substances in whole blood by liquid chromatography–tandem mass spectrometry. J. Chromatogr. B 811: 13–20 (2004). 7. H.H. Maurer. Multi-analyte procedures for screening for and quantification of drugs in blood, plasma, or serum by liquid chromatography–single stage or tandem mass spectrometry (LC–MS or LC–MS–MS) relevant to clinical and forensic toxicology. Clin. Biochem. 38: 310–318 (2005). 8. M. Laloup, M. Fernandez, G. De Boeck, M. Wood, and V. Maes. Validation of a liquid chromatography–tandem mass spectrometry method for the simultaneous determination of 26 benzodiazepines and metabolites, zolpidem and zopiclone in blood, urine, and hair. J. Anal. Toxicol. 29: 616–626 (2005). 9. S. Hegstad, E.L. Oiestad, U. Johansen, and A.S. Christophersen. Determination of benzodiazepines in human urine using solidphase extraction and high-performance liquid chromatog-

498

10.

11.

12.

13.

14.

raphy–electrospray ionization tandem mass spectrometry. J. Anal Toxicol. 30: 31–37 (2006). O. Quintela, F. Sauvage, F. Charvier, J. Gaulier, G. Lachatre, and P. Marquet. Liquid chromatography–tandem mass spectrometry for detection of low concentrations of 21 benzodiazepines, metabolites, and analogs in urine: method with forensic applications. Clin. Chem. 52: 1346–1355 (2006). B.E. Smink, M.P.M. Mathijssen, K.J. Lusthof, J.J. Gier, A.C.G. Egberts, and D.R.A. Uges. Comparison of urine and oral fluid as matrices for screening of 33 benzodiazepines and benzodiazepine-like substances using immunoassay and LC–MS–MS. J. Anal. Toxicol. 30: 478–485 (2006). H.H. Maurer. Current role of liquid chromatography–mass spectrometry in clinical and forensic toxicology. Anal. Bioanal. Chem. 388: 1315–1325 (2007). M. Laloup, M. Fernandez, M. Wood, V. Maes, G. De Boeck, Y. Vanbeckevoort, and N. Samyn. Detection of diazepam in urine, hair, and preserved oral fluid samples with LC–MS–MS after single and repeated administration of Myolastan and Valium. Anal. Bioanal. Chem. 388: 1545–1556 (2007). J. Feng, L. Wang, I. Dai, T. Harmon, and J.T. Bernert. Simultaneous determination of multiple drugs of abuse and relevant metabolites in urine by LC–MS–MS. J. Anal. Toxicol. 31: 359–368 (2007). Manuscript received January 2, 2008; revision received April 8, 2008.