sign in sign up
<<
Home , Cyclobenzaprine, Desipramine, Doxepin, Norclomipramine, Nordoxepin, Nortriptyline

Available online at www.annclinlabsci.org Annals of Clinical & Laboratory Science, vol. 44, no. 4, 2014 431 Method Validation of a Drug Panel in by UPLC-MS/MS

Chia-Ni Lin1,2, JoEtta M. Juenke3, and Kamisha L. Johnson-Davis2,3

1Department of Laboratory , Chang-Gung Memorial Hospital, Taoyuan, Taiwan, 2Department of Pathology, University of Utah Health Sciences Center, Salt Lake City, UT, and 3ARUP Institute for Clinical and Experimental Pathology, Salt Lake City, UT,

Abstract. (TCA) drugs are also used as adjunctive therapy to treat . To monitor patient compliance to therapy, urine specimens may be preferred since collection is non-invasive, and the specimen can provide a longer detection window. TCA drugs are frequently monitored by immuno- assay; however, poor antibody specificity may compromise results. The purpose for this study was to develop a confirmation method for determining TCA in urine specimens by ultra-pressure liquid chromatography- tandem mass spectrometry (UPLC-MS/MS). Our method can quantify 9 TCA drugs in less than 5 min. Method validation experiments were performed, and the coefficient variation for inter- and intra-day preci- sion was within 12% for each analyte at five different concentrations. Accuracy studies had good agreement with another laboratory that performs testing by GC/MS.

Highlights: 1. Tricyclic are commonly used to treat , , and neuropath- ic pain. 2. A confirmation method by ultra-pressure liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) to determine TCA in urine specimens was developed. 3. Urine testing for TCA is pre- ferred for monitoring adherence to therapy due to ease of collection.

Key words: Tricyclic antidepressants (TCA), Urine drug testing, UPLC-MS/MS

Introduction TCA, which can also be used as adjunctive therapy for chronic and , are primarily me- Tricyclic Antidepressants (TCA), which have a tabolized by the . Seventy percent of the total characteristic three-ringed nucleus structure, are dose is excreted via the kidney as inactive metabo- one of the oldest classes of antidepressants and are lites. Drug concentrations and metabolites tend to still used extensively. In addition to the treatment be higher in urine than in serum, allowing longer of various forms of depression, TCA also have effi- detection time. Urine TCA testing may be a useful cacy in the treatment of anxiety, obsessive-compul- approach to identify compliance to therapy. sive disorders, post-traumatic stress disorders, eat- Although immunoassay is widely used for measure- ing disorders, attention deficit disorders, and a ment of many drugs, the assay has poor specificity number of different types of chronic pain. The rela- and can generate false positives due to cross-reactiv- tively narrow therapeutic range of TCA necessi- ity with three-ring structured compounds from dif- tates clinical monitoring for effective treatment. ferent drug classes. For example, immunoassay test- Clinical presentation of overdose from TCA inges- ing cannot differentiate between parent and tion includes convulsions, coma, life-threatening metabolite drugs, such as and nor- and cardiac conduction disturbances. triptyline. Due to poor specificity, false positives for An overdose can be lethal [1,2]. TCA can occur from cross-reactivity with quetiap- ine ( drug) and drugs Address correspondence to Kamisha L. Johnson-Davis, PhD; [3-6]. Several analytical methods have been devel- University of Utah, Department of Pathology, ARUP Laboratories: 500 Chipeta Way, Salt Lake City, UT 84108, USA; phone: +1 11 oped to quantify TCA in biofluids using high per- 801 583 2787, ext 2982; fax: +1 11 801 584 5207; e mail: kamisha. formance liquid chromatography (HPLC)[3,4], [email protected] gas-chromatography (GC)[5], gas-chromatography

0091-7370/14/0400-431. © 2014 by the Association of Clinical Scientists, Inc. 432 Annals of Clinical & Laboratory Science, vol. 44, no. 4, 2014

Table 1. Mass Spectrometry parameters for Tricyclic antidepressant drugs and metabolites.

Analyte Precursor cone Product 1 Collision Product 2 Collision (m/z) (m/z) energy (m/z) energy

Amitryptyline 278.20 35 91.0 35 105.5 35 -d3 283.15 30 107.0 25 235.10 15 Doxepin 280.10 30 107.05 25 235.10 15 266.10 30 107.05 20 235.10 15 276.10 35 215.10 40 58.00 20 295.10 25 100.00 15 58.00 15 -d3 267.15 30 91.00 20 233.10 15 -d3 284.20 30 89.10 15 61.10 30 Imipramine 281.15 30 86.05 15 58.10 30 -d3 270.15 25 75.05 15 47.10 30 Desipramine 267.15 30 72.05 20 44.10 30 -d3 267.20 35 155.10 20 191.10 25 Protriptyline 264.20 35 155.05 20 191.10 25 -d3 318.15 30 89.10 20 61.10 30 Clomipramine 315.10 30 86.05 20 58.05 30 301.10 30 72.05 15 44.05 35

mass spectrometry (GC-MS)[6], HPLC-MS[7], methanol, and deuterated solutions of imipramine-D3, HPLC-MS-MS[8,9] and capillary electrophoresis nortriptypine-D3, desipramine-D3, doxepin-D3, pro- (CE)[10], and turbulent flow liquid chromatogra- triptyline-D3 and clomipramine-D3 were purchased as phy [11]. Most of these methods have an analysis 0.1 µg/µL from Cerilliant (Round Rock, TX). Residual urine samples sent for routine testing were de-identified time greater than 10 min per sample [5-7,9,10]. according to a protocol approved by the University of Utah Institutional Review Board. Drug-free urine was The purpose of this study was to develop a confir- pooled and used as a matrix to create different concen- mation method to determine TCA in urine speci- tration of samples for validation. mens, supporting TCA adherence testing for in laboratories that need confirmation Sample preparation. The calibrators, controls, and pa- testing for positive TCA urine samples by immuno- tient samples were processed by protein precipitation, assay or urine drug screen cup. The method was using a reagent that contains methanol and acetonitrile developed by ultra-pressure liquid chromatography (50:50). After the addition of precipitation reagent forti- tandem mass spectrometry (UPLC-MS/MS) to re- fied with internal standard (IS), the samples were capped, vortexed and centrifuged. The supernatant was then duce run-time and support high-volume testing to poured into an autosampler vial. Separation and detec- quantify amtriptyline, nortriptyline, doxepin, des- tion of drugs were performed using UPLC-MS/MS; methyldoxepin (nordoxepin), imipramine, desim- Waters Acquity UPLC TQD with Masslynx software pramine, clomipramine, desmethylclomipramine (Waters Corp., Milford, MA). (norclomipramine), and protriptyline in less than 5 minutes. The method can also detect the presence Liquid chromatography-electrospray tandem mass of trimipramine and cyclobenzaprine, which is a spectrometry. The specimen was injected into an Acquity that is known to cause false posi- UPLC HSS T3 column with dimensions of 2.1x50 mm tives for TCA by immunoassay. and a 8 µm particle size (Waters Corp., Milford, MA) at room temperature. The mobile phase flow rate was 0.6 mL/min, and the injection volume was 2 µL. Elution Experimental occurred using a linear gradient of acetonitrile and water, each having 0.1% formic acid added. The analytes were Chemicals, reagents and samples. Stock solutions of detected with a Waters Acquity UPLC TQD operating amitriptyline, imipramine, nortriptypine, desipramine, in positive electrospray ionization and multiple reaction doxepin, nordoxepin, protriptyline, clomipramine, and monitoring (MRM) mode. The capillary voltage was 3 norclomipramine were purchased as 1 mg/mL in kV, and the desolvation gas flow was set at 25 L/hr, the Tricyclic antidepressant panel by UPLC-MS/MS 433

Table 2. Linearity was evaluated by analyzing five batches of samples at 5 concentrations in duplicate over the stated ana- lytical range of the assay.

Analytes Analytical range Linearity correlation data

Amitriptyline 100-2000 ng/mL Slope =0.99 r: 0.9958 y-intercept =4.98 n= 10 Nortriptyline 100-2000 ng/mL Slope =1.00 r: 0.9966 y-intercept =2.82 n= 10 Imipramine 100-2000 ng/mL Slope =0.99 r: 0.9988 y-intercept =6.33 n= 10 Desipramine 100-2000 ng/mL Slope =0.99 r: 0.9983 y-intercept =4.33 n= 10 Doxepin 100-2000 ng/mL Slope =0.99 r: 0.9970 y-intercept =3.80 n= 10 Nordoxepin 100-2000 ng/mL Slope =1.00 r: 0.9969 y-intercept =-3.64 n= 10 Clomipramine 200-4000 ng/mL Slope =0.99 r: 0.9989 y-intercept =10.78 n= 10 Norclomipramine 200-4000 ng/mL Slope =1.01 r: 0.9979 y-intercept =-17.97 n= 10 Protriptyline 100-2000 ng/mL Slope =1.00 r: 0.9974 y-intercept =-2.91 n= 10

source temperature was 120°C. The precursor ions, the total 33 positive samples, 7 samples were positive for product ions, cone voltage, and collision energies for amitriptyline and nortriptyline, 6 samples for nortripty- each analyte and internal standard are listed in Table 1. line, 1 sample for amitriptyline, 5 samples for doxepin and nordoxepin, 3 samples for imipramine and desipra- Ion suppression. Ion suppression was evaluated by ana- mine, 2 samples for protriptyline, and 3 samples for clo- lyzing extracted samples while infusing the compounds mipramine and norclomipramine. Others included 1 of interest. The effect of signal interference was evaluated sample for clomipramine, norclomipramine, and pro- in specimens containing the following drugs: , triptylin; 1 sample for amitriptyline and nortriptyline, norsertraline, , norclozapine, , di- clomipramine, and norclomipramine; 2 samples for phenhydramine, , , , amitriptyline and nortriptyline, doxepin and nordoxe- , , , , pin; 1 sample for imipramine desipramine and protrip- , dihydrocodone, , carbam- tyline; and 1 sample for imipramine, desipramine, clo- azepine, and epoxide. mipramine, and norclomipramine.

Validation of the method. The method was validated for Results linearity, imprecision, carryover, and accuracy. Linearity was determined with fortified urine calibrators prepared Linearity. Linearity plots had R2 values >0.995 in five replicates at 100, 200, 400, 1000, 2000 ng/mL from 100-2000 ng/mL for amitriptyline, imipra- for amitriptyline, imipramine, nortriptypine, desipra- mine, doxepin, nordoxepin, and protriptyline, and at mine, nortriptypine, desipramine, doxepin, nor- 200, 400, 800, 2000, 4000 ng/mL for clomipramine doxepin, protriptyline, and at 200-4000 ng/mL, and norclomipramine. Five-day intra- and inter-day im- for clomipramine and norclomipramine. Results precision was determined by analyzing urine fortified are shown in Table 2. with 100, 300, 1000, 2500, and 5000 ng/mL and quan- tified using calibrators, except with clomipramine and Imprecision. Imprecision was determined through norclomipramine, where four concentrations were eval- analysis of five concentrations from urine samples uated. Carryover was evaluated by analyzing a set of high fortified with TCA prepared in quintuplicate for concentration (5000 ng/mL) and a set of low concentra- five days. Intra- and inter- assay imprecision was tion (100 ng/mL) in urine with a certain sequence as less than 12% for each concentration (Table 3). described at the CLSI protocol EP10-A3 [12]. To evalu- ate accuracy, drug-free urine was pooled and used as a Percent recovery of the assay was calculated by matrix to spike 22 samples with different TCA at differ- comparing the target value of the control samples ent concentrations. An additional 53 authentic patient and to the measured results (n=25). Percent recov- samples were collected, de-identified, and sent to an al- ery was >80% recovery for most of the drugs, ex- ternate laboratory (NMS Labs, Willow Grove, PA) for cept clomipramine and norclomipramine at 100 split sample comparison by gas chromatography-mass ng/mL. The welo r limit of quantification was set at spectrometry (GC/MS). The 53 samples contained 42 200 ng/mL for clomipramine and negative samples and 11 positive samples. Adding the 22 norclomipramine. spike samples, 33 positive samples were evaluated. Of 434 Annals of Clinical & Laboratory Science, vol. 44, no. 4, 2014

Accuracy. Fifty-three patient samples were collect- ed, de-identified, and sent to an alternate laboratory (NMS Labs, Willow Grove, PA) for split sample comparison by gas chromatography-mass spec- trometry. Of the 53 patient results, there were two samples that demonstrated a significant negative bias in the UPLC-MS/MS method. One sample was reported to contain 550 ng/mL of amitriptyline and 1300ng/mL of nortriptyline by GC/MS but was reported to contain 394 ng/mL of amitriptyline and 81.2 ng/mL by the UPLC-MS/MS method. The other was reported to contain 167.5 ng/mL of nortriptyline by GC/MS and only 61ng/mL by the UPLC-MS/MS method. The low bias of the former Figure 1. Comparing the UPLC-MS/MS results with GC samples could be due to the different sample extrac- MS showed excellent agreement between these two method- tion preparations between the two laboratories. The ologies for different tricyclic antidepressant drugs in patient correlation was shown in Figure 1. There were no urine. Of the 53 urine samples run in parallel between GC/MS and UPLC-MS/MS method, 3 samples were ex- false positive results observed. An additional 22 cluded, due to poor correlation to GC/MS and the concen- drug-free urine samples were pooled and fortified tration range of the drugs were between 100 to 5000 ng/ with TCA at different concentrations. There were mL. (y=1.0088x-2.1377, R²=0.9854). no false positive results or negative bias in results for these fortified samples.

Ion suppression and interference. TCA have simi- Discussion lar structures, and some analytes have similar mass- es. The TCA drugs that have similar nominal mass- Because TCA are used as adjunctive therapy to treat es can exhibit the same multiple reaction monitoring chronic and neuropathic pain, adherence to therapy transitions, leading to isobaric interferences within must be monitored. Urine testing for TCA is pre- the assay. The most prominent isobaric interference ferred due to ease of collection. Clinical monitoring was nortriptyline/protriptyline and nortriptyline- of therapeutic drugs is most easily achieved using D3/protriptyline-D3. These two analytes were dif- immunoassay methods, either automated instru- ficult to separate by LC, due to complete overlap of ments or point of care devices. However, assay per- transitions; therefore, baseline peak separation was formance is limited by false positive results and critical. All of the other analytes were separated ei- poor specificity in identification of specific TCA ther by LC or by distinct MRM transitions. Ion drugs, as a result of their similar molecular struc- suppression was evaluated by extracting five differ- tures [13-16]. Consequently, confirmatory meth- ent drug-free human urine samples and injecting ods are needed to identify specific TCA and them into the mass spectrometer, while infusing a metabolites. sample containing all of the analytes of interest. There was no significant ion suppression observed There are published, quantitative methods for urine at the retention times of the analytes. Structurally TCA by gas chromatography-mass spectrometry similar compounds and other drugs that may be [17] and liquid chromatography tandem mass spec- co-administered were analyzed for interference us- trometry [18]. Both methods tend to have longer ing drug-free urine fortified with various concen- analysis time. Our current UPLC-MS/MS method, trations of the analytes as described in materials however, has a run-time of less than 5 min. This and methods. The compounds that were fortified method employs a simple sample preparation and into urine did not cause interference in this assay. produces reliable and reproducible results for the Tricyclic antidepressant panel by UPLC-MS/MS 435

Table 3. Imprecision was determined by analyzing three levels of control material prepared in quintuplicate at five con- centrations for five days (n=25 for each concentration).

Expected Mean Within Run Between Day Total Drug Sample N (ng/mL) (ng/mL) SD CV (%) SD CV (%) SD CV (%)

Amitriptyline 1 25 100 80.88 4.89 6.05 2.12 2.62 5.33 6.59 2 25 300 270.5 15.68 5.69 10.54 3.83 18.90 6.86 3 25 1000 1029 50.75 4.93 40.28 3.91 64.79 6.30 4 25 2500 2567 108.65 4.23 116.02 4.52 158.95 6.19 5 25 5000 5160.7 200.01 3.88 579.48 11.23 613.02 11.88 Nortriptyline 1 25 100 76.15 4.12 5.41 2.91 3.82 5.04 6.62 2 25 300 268.25 15.84 5.90 11.83 4.41 19.77 7.37 3 25 1000 985.09 57.14 5.80 45.34 4.60 72.94 7.40 4 25 2500 2513.7 100.37 3.99 114.35 4.55 152.15 6.05 5 25 5000 5083.3 193.11 3.80 401.53 7.90 445.56 8.77 Imipramine 1 25 100 79.24 3.50 4.42 5.00 6.31 6.10 7.70 2 25 300 273.72 10.75 3.93 9.16 3.35 14.13 5.16 3 25 1000 975.27 34.48 3.54 39.06 4.01 52.10 5.34 4 25 2500 2571.2 83.81 3.26 105.27 4.09 134.56 5.23 5 25 5000 5157.5 211.69 4.10 339.50 6.58 400.09 7.76 Desipramine 1 25 100 78.44 5.45 6.95 6.60 8.41 8.56 10.91 2 25 300 266.98 11.26 4.22 8.76 3.28 14.27 5.34 3 25 1000 959.12 39.40 4.11 12.84 1.34 41.44 4.32 4 25 2500 2456.3 76.37 3.11 87.75 3.57 116.33 4.74 5 25 5000 4971.6 157.51 3.17 241.00 4.85 287.91 5.79 Doxepin 1 25 100 82.4 4.54 5.51 2.71 3.29 5.29 6.42 2 25 300 281.44 17.97 6.38 2.79 0.99 18.18 6.46 3 25 1000 982.36 37.14 3.78 17.61 1.79 41.10 4.18 4 25 2500 2509.7 87.15 3.47 67.26 2.68 110.09 4.39 5 25 5000 5051.7 216.90 4.29 319.41 6.32 386.09 7.64 Nordoxepin 1 25 100 83.34 5.06 6.08 3.84 4.61 6.35 7.62 2 25 300 283.22 12.75 4.50 9.04 3.19 15.63 5.52 3 25 1000 996.76 41.55 4.17 21.24 2.13 46.66 4.68 4 25 2500 2589.8 108.54 4.19 72.76 2.81 130.67 5.05 5 25 5000 5301.3 160.10 3.02 251.54 4.75 298.17 5.62 Clomipramine 1 25 100 57.7 5.03 8.71 9.12 15.81 10.42 18.05 2 25 300 252.19 12.87 5.10 4.58 1.82 13.67 5.42 3 25 1000 968.44 28.78 2.97 40.96 4.23 50.06 5.17 4 25 2500 2577.7 75.67 2.94 171.79 6.66 187.72 7.28 5 25 5000 5347.1 157.93 2.95 254.31 4.76 299.36 5.60 Norclomipramine 1 25 100 61.67 4.08 6.61 12.29 19.94 12.95 21.00 2 25 300 235.9 11.33 4.80 7.13 3.02 13.39 5.67 3 25 1000 898.62 28.68 3.19 32.34 3.60 43.22 4.81 4 25 2500 2480.3 73.08 2.95 190.11 7.67 203.67 8.21 5 25 5000 5224.8 149.67 2.86 361.69 6.92 391.44 7.49 Protriptyline 1 25 100 84.34 4.52 5.36 7.30 8.66 8.59 10.18 2 25 300 284.5 18.25 6.42 5.80 2.04 19.15 6.73 3 25 1000 1018.2 32.78 3.22 39.18 3.85 51.08 5.02 4 25 2500 2602.9 83.16 3.19 112.78 4.33 140.13 5.38 5 25 5000 5406.6 144.69 2.68 229.41 4.24 271.23 5.02 436 Annals of Clinical & Laboratory Science, vol. 44, no. 4, 2014 detection and quantification of nine TCA in urine multi-level, single-sample approach. Journal of chromatogra- phy. B, Analytical technologies in the biomedical and life sci- specimens. In addition, this method has the ability ences 843, 100-113 to detect cyclobenzaprine, which can cause false 9. Kollroser, M., and Schober, C. (2002) Simultaneous determi- positive for TCA by immunoassay. nation of seven tricyclic antidepressant drugs in human plasma by direct-injection HPLC-APCI-MS-MS with an ion trap de- tector. Therapeutic drug monitoring 24, 537-544 References 10. Jinno, K., Kawazoe, M., Saito, Y., Takeichi, T., and Hayashida, M. (2001) Sample preparation with fiber-in-tube solid-phase 1. Preskorn, S. H., Othmer, S. C., Lai, C. W., and Othmer, E. microextraction for capillary electrophoretic separation of tri- (1984) Tricyclic-induced electroencephalogram abnormalities cyclic antidepressant drugs in human urine. Electrophoresis and plasma drug concentrations. Journal of clinical psycho- 22, 3785-3790 pharmacology 4, 262-264 11. Sauvage, F. L., Gaulier, J. M., Lachatre, G., and Marquet, P. 2. Varley, C. K., and McClellan, J. (1997) Case study: two addi- (2006) A fully automated turbulent-flow liquid chromatogra- tional sudden deaths with tricyclic antidepressants. Journal of phy-tandem mass spectrometry technique for monitoring anti- the American Academy of Child and Adolescent 36, in human serum. Therapeutic drug monitoring 28, 390-394 123-130 3. Theurillat, R., and Thormann, W. (1998) Monitoring of tricy- 12. (2006) Clinical and Laboratory Standards Institute. clic antidepressants in human serum and plasma by HPLC: Preliminary Evaluation of Quantitative Clinical Laboratory characterization of a simple, laboratory developed method via Measurement Procedures; Approved Guideline-Third Edition. external quality assessment. Journal of pharmaceutical and CLSI document EP10-A3 (ISBN 1-56238-622-0). . Clinical biomedical analysis 18, 751-760 and Laboratory Standards Institute,940 West Valley Road, 4. Tanaka, E., Terada, M., Nakamura, T., Misawa, S., and Suite 1400, Wayne, Pennsylvania 19087-1898 USA,2006 Wakasugi, C. (1997) Forensic analysis of eleven cyclic antide- 13. Sorisky, A., and Watson, D. C. (1986) Positive diphenhydr- pressants in human biological samples using a new reversed- interference in the EMIT-st assay for tricyclic antide- phase chromatographic column of 2 microm porous micro- pressants in serum. Clinical chemistry 32, 715 spherical silica gel. Journal of chromatography. B, Biomedical 14. Dasgupta, A., Wells, A., and Datta, P. (2007) False-positive sciences and applications 692, 405-412 serum tricyclic antidepressant concentrations using fluores- 5. Martinez, M. A., Sanchez de la Torre, C., and Almarza, E. cence polarization immunoassay due to the presence of hy- (2003) A comparative solid-phase extraction study for the si- droxyzine and . Therapeutic drug monitoring 29, multaneous determination of , amitriptyline, nor- 134-139 triptyline, trimipramine, , clomipramine, and tra- 15. Hendrickson, R. G., and Morocco, A. P. (2003) zodone in whole blood by capillary gas-liquid chromatography Cross-Reactivity Among Three Tricyclic Antidepressant with nitrogen-phosphorus detection. Journal of analytical toxi- Immunoassays. Clinical toxicology 41, 105-108 cology 27, 353-358 16. Schussler, J. M., Juenke, J. M., and Schussler, I. (2003) 6. Paterson, S., Cordero, R., and Burlinson, S. (2004) Screening Quetiapine and falsely elevated nortriptyline level. The and semi-quantitative analysis of post mortem blood for basic American journal of psychiatry 160, 589 drugs using gas chromatography/ion trap mass spectrometry. 17. Rana, S., Uralets, V. P., and Ross, W. (2008) A new method for Journal of chromatography. B, Analytical technologies in the simultaneous determination of cyclic antidepressants and their biomedical and life sciences 813, 323-330 metabolites in urine using enzymatic hydrolysis and fast GC- 7. Tolonen, A., Turpeinen, M., Uusitalo, J., and Pelkonen, O. MS.J Anal Toxicol 32, 355-363 (2005) A simple method for differentiation of monoisotopic 18. Ito, R., Ushiro, M., Takahashi, Y., Saito, K., Ookubo, T., drug metabolites with hydrogen-deuterium exchange liquid Iwasaki, Y., and Nakazawa, H. (2011) Improvement and vali- chromatography/electrospray mass spectrometry. European dation the method using dispersive liquid-liquid microextrac- journal of pharmaceutical sciences : official journal of the tion with in situ derivatization followed by gas chromatogra- European Federation for Pharmaceutical Sciences 25, 155-162 phy-mass spectrometry for determination of tricyclic 8. Kirchherr, H., and Kuhn-Velten, W. N. (2006) Quantitative antidepressants in human urine samples.J Chromatogr B determination of forty-eight antidepressants and antipsychot- Analyt Technol Biomed Life Sci 879, 3714-3720 ics in human serum by HPLC tandem mass spectrometry: a