Monitoring Tricyclic Antidepressant Concentrations in Serum by ...

3 downloads 0 Views 1MB Size Report
veloped by Abbott to diagnose intoxication with tncyclic antidepressants was adapted for therapeutic drug moni- toring and validated with chromatographic ...
CLIN.CHEM.40/6, 929-933

(1994)

#{149} Drug

Monitoring

and Toxicology

Monitoring Tricyclic Antidepressant Concentrations in Serum by Fluorescence Polarization Immunoassay Compared with Gas Chromatography and HPLC Marie Luise Rao,1’4 Uta Staberock,1 Pierre Baumann,2 Christoph Marlyse Amey,2 Sebastian H#{228}rtter,3 and Michaela Kraemer’

The fluorescence polarization immunoassay (FPIA) developed by Abbott to diagnose intoxication with tncyclic antidepressants was adapted for therapeutic drug monitoring and validated with chromatographic methods to investigate its potential for this use. We compared serum concentrations of tricyclic antidepressants in vivo and in vitro obtained by FPIA with those by gas chromatography and HPLC. For amitriptyline, imipramine, clomipramine, and doxepin, the detection limit of the FPIA was 72, 71, 64, and 72 nmol/L (-20 gIL), respectively; that by gas chromatography was 18, 18, and 16 nmol/L (-5 g/L) for amitnptyline, imipramine and clomipramine, respectively; with HPLC the lower limit of detection for doxepin was 36 nmol/L (10 g/L). The intra- and interassay CVs ranged from 3% to 6%. In patients being treated with amitriptyline, imipramine, clomipramine, and doxepin, at steady-state the correlation coefficients between FPIA and GCIHPLC results for split samples were 0.95, 0.92, 0.90 and 0.70, respectively. However, the slopes were close to unityonly for amitriptylineand doxepin, being 0.6 for imipramine and 1.9 for clomipramine. Terms: toxicology/kit adaptation/compliance/intermethod comparison/neuroactive drugs/monitoring therapy Indexing

In 1985 the American Psychiatric Association Task Force recommended therapeutic monitoring of tricycic antidepressant drugs as a guideline for treating patients (1). To achieve this goal, monitoring by gas chromatography and liquid chromatography were introduced (2-5). However, these methods were time consuming and required highly trained personnel. For data on drug monitoring to be integrated into patient management, the results need to be available within a few days. Immunological methods to achieve this goal have been developed, and the enzyme immunoassay technique (Emit; Syva Co., Palo Alto, CA), a nonradiometric quantitative immunoassay, is widely used for drug monitoring (6-9). Here we discuss the fluorescence polarization immunoassay (FPIA), which was originally designed to detect 1Psychiatrische Klinik und Poliklinikder Rheinischen FriedrichWilhelms-UniversitSt, Sigmund-Freud-Str. 25, 53105 Bonn, Germany. 2Unit#{233} de Biochimie et Psychopharmacologie Clinique, Departement Universitaire de Psychiatrie Adulte, CH-1008 Prilly-Lausanne, Switzerland. 3Psychiatrische Kiinik der Johannes Gutenberg-Universitat Mains, Untere Zahlbacher Str. 8, 55101 Mains, Germany. 4Author for correspondence. Fax 49-228-280-3097. Received June 7, 1993; accepted February 25, 1994.

Hiemke,3

Arno

Deister,’

Christelle

Cuendet,2

serum concentrations of tricycic antidepressants (10-13). There being, to our knowledge, no previous reports comparing the tricycic antidepressant serum concentrations measured by gas-liquid chromatography (GLC) or HPLC with those by FPIA, we investigated the potential of the FPIA with respect to therapeutic drug monitoring. We also tested the cross-reactivity of antidepressants and neuroleptics in vitro with a fluorescence-labeled antibody to the tricycic drug structure in the presence of buffer containing a protein stabilizer and with the tracer used in the TDxiTDxFLxTh ToxicologyTricycic Assay FPIA (Abbott Labs., Abbott Park, IL) and carried out a comparative study on steady-state serum concentrations of antidepressants and their major metabolites with the FPIA, GLC, and HPLC. toxic

Materials and Methods Subjects and Specimen Collection The healthy subjects were blood donors from the Institute of Hematology, whose health records were checked routinely. Inpatients’ blood was obtained during routine therapeutic drug monitoring, which is an integral part of our patient care at the Psychiatric Hospital of the University of Bonn. All procedures were in accordance with the ethical standards of the Helsinki Declaration of 1975 as revised in 1983. Patients’ blood samples were included in this study, when the following conditions were fulfilled: (a) monotherapy; (b) steadystate conditions, i.e., blood taken after treatment with a stable dose for at least 1 week (although most of the patients had been treated for 3 weeks or longer); and (c) blood drawn after an overnight fast, 12 h after the last administration of the drug. To minimize the sources of statistical imprecision, we included in each analysis one sample per patient treated with amitriptyline, imipramine, clomipramine, or doxepin. Fasting blood was collected in the morning from healthy subjects and from antidepressant-treated patients by vempuncture into brown-stoppered evacuated tubes without additives (Venoject; Terumo, Leuven, Belgium). The blood was allowed to clot for 30 mm at room temperature and then centrifuged (3000g, 10 miii). The serum was divided into aliquots and stored at -25#{176}C until further use. For the intermethod investigation by GLC or HPLC, the samples were shipped on solid CO2 to the laboratory in Lausanne, Switzerland (P.B.), or Mainz, Germany (C.H.). Control sera containing 0, 50, or 200 ng of exogenous antidepressants were also included. The data were exchanged when all samples were analyzed. CUNICAL CHEMISTRY, Vol. 40, No. 6, 1994

929

Reagents Most of the drugs were gift8: amitriptyhne (Lundbeck, Copenhagen, Denmark; Merck, Sharp and Dome, M#{252}nchen,Germany); clomipramine, desmethylclomipramine, and opipramol (Ciba Geigy, Wehr, Germany); chlorpromazine (Bayer, Leverkusen, Germany); dibenzepine and thioridazine (Sandoz, Basel, Switzerland); dosulepine (The Boots Co., Nottingham, UK); doxepin and metabolites (Galenus, Mannheim, Germany; Pfizer, Karlsruhe, Germany); imipramine and derivatives and maprotiline (Ciba Geigy, Basel, Switzerland); meprobamate, nortriptyline, and promethazine (Tropon, Kfiln, Germany); mianserin (Organon, Oss, The Netherlands); perazine (Promonta, Hamburg, Germany); trazodone (Thomae, Bieberach, Germany); and trimipramine (Rh#{244}ne-Poulenc Rorer, K#{246}ln,Germany). Desipramine was purchased from Sigma (MOnchen, Germany). Procedures FPIA quantification of antidepressants. The reagents for the FPLA were purchased from Abbott (Wiesbaden, Germany). The assay (TDx/TDxFLx Tricycic Antidepressant Assay) of 50-L samples includes tricycic antidepressant antibodies raised in rabbits and fluorescein-labeled tricycic antidepressant as tracer. We calibrated the assay with imipramine calibrators of 751000 4ug/L (268-3571 nmoIJL) and measured their apparent concentrations as proposed by the manufacturer. Besides the internal standards provided by the manufacturer, we also analyzed our own internal quality-control samples, which contained concentrations as low as 10 .i.gfL. To test the cross-reactivity of various drugs, we prepared stock solutions (1.0 g/L) of them in aqueous 1 g/L ascorbic acid in water. We diluted the stock solutions with serum from healthy subjects to obtain concentrations ranging from 10 to 1000 g/L. The amount of ascorbic acid in the final volume was 0.96, slope >0.97, n = 45). Assaying the concentrations of the tricycic antidethe tracer from the Abbott TDx procedure were close to pressants in sera kept for 7 days at 20-23#{176}C or at 4#{176}C 100%. The in vitro experiments indicated no interferyielded CVs of 2% and 3%, respectively. The slope of the ence between the parent compound and metabolite (data not shown); their concentrations appeared to be addidata for day 1 vs day 7 was close to unity, indicating that the serum concentrations of the drugs were stable at tive, and the antibody did not distinguish between the tertiary amino parent compound and the secondary these temperatures for 1 week. Intermethod comparison. The drug concentrations of amino metabolites. all serum samples were above the detection limit of the The low intra- and interassay CVs of the FPIA (3-6%) FPIA, GLC, or HPLC. permit longitudinal investigations. The low detection The correlation between the sum of the concentralimit of the assay allows measurement of apparent concentrations (i.e., immunoreactivity) 20 tg/L (-70 tions of amnitriptyline plus the active metabolite nortriptyline obtained by GLC vs the total antidepressant imnmolJL). This is necessary for the clinical setting, given munoreactivity determined by the FPIA was r = 0.95 that low concentrations are likely to occur in nonre(Fig. 1A, Table 2). The concentrations obtained by the spenders (Rao et al., unpublished observations), e.g., FPIA were 19% higher than by GLC. The proportion of noncompliant patients or those who metabolize the the active metabolite nortriptyline in relation to the drugs rapidly (because of either genetic disposition or CLINICAL CHEMISTRY, Vol. 40, No. 6, 1994

931

2.000

#{149} C

A

B C

E

1.500

1.500

0

#{149}

U. V0

ai.ooo 0 =

a

#{149} -, #{149} #{149} .

C

a

1.000

E

soo

0

E

500

a

4

S

E

S..

C’ 0

#{149}

500

1.000

lmmunor.cctlve

1.500

omltriptylino

m.t#{248}bolites,nmol/L

C 0

2.000

500

plus

1.000

Immunoreoctive m.tabolites,

(FPIA)

1.500

2.000

imiprOmine nmol/L

2.500

plus

(FPIA)

‘.uu’.J 5..

C

-

.3.000

a

.

C

a

1.500

0 V

2,500 i.E

-

‘C

a

D

C

.

2.000

Ca.

1.500

V

5l

#{149} 1.000

#{149}---

#{149}‘ S

1.000

C

a

‘4’.

C 5

500

0

500

0

lmmunoreoctlve m.tabolltss,

..

1.000

1.500

clomipramine nmol/L

0

plus

200

400

lmmunoreactive metobolites.

(FPIA)

600

800

1.000

doxepln plus nmol/L (IPIA)

Fig. 1. Cross-validation of the concentrations of serum tricyclic antidepressants plus metabolites as determined by FPIA (immunoreactivity) and the sum of the antidepressant and its major metabolite analyzed with gas chromatography (A, B, and C) or HPLC (0): (A) amitriptytine, (B) imipramine, (C) clomipramine, and (0) doxepin.

substance abuse) and thus cannot build up sufficiently high concentrations of the drugs for a therapeutic response. The results of the FPLA correlated well with those by GLC and HPLC for amitriptyline, imipramine, clomipramine, and doxepin. All intercepts of the correlation curves passed close to the origin. However, the slopes of the correlation analysis between methods were close to unity only for amitriptyline and doxepin. Because the cross-reactivity of clomipramine ranged from 40% to 60% and because the proportions of parent compound and metabolite were similar throughout the concentration range tested, the concentrations measured by GLC were twice those by FPIA (i.e., the correlation slope was 1.92). Over the entire concentration range, the serum

of imipramine obtained by FPLA were 40% lower than those by GLC. This discrepancy was not due to the presence of the N-desmethylated metabolite, however, which showed a constant proportion of 62% to 72% (Table 2); the discrepancy might be due, we assume, to the putative interference of hydroxylated metabolites. The tetracycic compounds maprotiline and mianserin cross-react only slightly with the anti-tricycic antibody. In the daily setting of the clinic, patients are comedicated with other drugs; the question then is, Does this interfere with the FPLA of the tricycic antidepressants” The commonly used drugs lithium, benzodiazepines, and thioridazine showed little or no cross-reactivity with the anti-tricycic antibody and the tracer. Howconcentrations

Table 2. Cross-valIdation of steady-state tricyclic antIdepressant concentrations obtained wIth FPIA, GLC, and HPLC, and percentage of the major metabollte. Correlation n

Comparlslon method

Slope

N-Demethylated metabolite

Intercept r Metabolite 0 0.95 Nortriptyline 36 GLC 0.81 Imipramine 29 GLC 0.64 31 Desipramine 0.92 17 0.90 Desmethylclomipramine Clomipramine 39 GLC 1.92 Doxepin 48 HLPC 0.98 15 0.70 Desmethyldoxepin #{149}The sum of the concentration of the parent compound plus the N-desmethylated metabolite was taken as 100%.

Arnitriptyline

932 CLINICAL CHEMISTRY, Vol. 40, No. 6, 1994

%OhtOt&a 43 62 71 48

r

0.76 0.96 0.97 0.97

ever, perazine gave substantial interference: Therapeutic serum concentrations of perazine range from 380 to 2700 moWL, concentrations that exhibit considerable interference with determinations of tricyclic antidepressanta by FP!A. On the other hand, chlorpromazine, thioridazine, and promethazine in the therapeutic range did not interfere. Treatment with perazine, a neuroleptic drug prescribed in Germany, yields false-positive serum results, misleadingly indicating that tricycic antidepressants were involved. Therefore, when intoxication due to slow metabolism or attempted suicide with drugs is suspected, it is necessary to know the putative drug taken to interpret the results. Although the FPIA from Abbott was originally designed for only semiquantification of toxic tricycic antidepressant concentrations in serum, its sensitivity and precision render it suitable to answer questions concerning compliance and completion of a washout phase. However, there are considerable limitations with respect to the accuracy of FPIA for clomipramine and imipramine in this use. The low cross-reactivity of 50% with respect to clomipramine meant that the serum concentration reported for clomipramine-plus-desmethylclomipramine was only 50% of that given by the chromatographic methods. Similarly, the apparent concentration given by the FPIA with respect to imipramine plus desipramine was 36% and that of amitriptyline plus nortripyline 19% higher than the concentrations obtained

tricyclic

by

GLC.

Therefore,

antidepressants

quantitative

obtained

results

for

by the FPIA should

be interpreted

with caution. On the other hand, unless methods are available to the routine clinical laboratory that provide within a minimum of 2 days the serum concentrations for the broad spectrum of tricycic antidepressants currently administered in the clinical setting, FPIA offers the opportunity to answer questions concerning inappropriately low serum concentrations,

patient noncompliance, and drug excess. We are currently evaluating serum tricycic antidepressant concentrations of patients with respect to nonresponse, compliance, therapeutic window, influence of age and sex, substance abuse, and toxicity. The skillful technical assistance of C. Chapuis (Lausanne) is gratefully acknowledged. We thank P. Hanfland and R.E. Scharf, Institute for Hematologr, for providing serum from healthy donors, and we appreciate the drug industry (see text: Reagents) for providing tricycle antidepressants, neuroleptics, and their metabolites. The data reported are part of an Inaugural Dissertation (U.S.) at The Medical Faculty of the University of Bonn.

References 1. APA Task Force. Tricycic antidepressants-blood level measurements and clinical outhome: an APA Task Force report. Am J Psychiatry 1985;142:155-62.

2. Jatlow P. Therapeutic monitoring of psychotherapeutic drugs: role of the clinical laboratory. Cliii Biochem 1985;18:143-8. 3. Orsulak PJ. Therapeutic monitoring of antidepressant drugs: current methodolor and applications. J Clin Psychiatry 1986;47: 39-50. 4. Preskorn SH, Cameron Dorey R, Jerkovich GS. Therapeutic drug monitoring of tricycic antidepressants. Cliii Chem 1988;34: 822-8. 5. Baumann F, Jonzier-Perey M, Koeb L, Le PK, Tunguely D, Sch#{246}pfJ. Amitriptyline pharmacokinetics and clinical response: I. Free and total plasma amitriptyline and nortriptyline. mt Clin Psychopharmacol 1986;1:89.-1O1. 6. Pankey S, Collins C, Jaklitsch A, Izutsu A, HuM, Pirio M, Sing P. Quantitative homogeneous enzyme nnmunoassays for amitriptyhne, nortriptyline, imipramine, and desipramine. Cliii Chem 1986;32:768-72. 7. Cameron Dorey R, Preskorn

511, Widener PK. Results compared for tricycle antidepressants as assayed by liquid chromatography and enzyme inimunoassay. Cliii Chem 1988;34:2348-51. 8. Fraser AD, Bryan W, Isner AF. Evaluation of the EMIT amitriptyline and nortriptyline assays for the determination of serum clomipramine and desmethylclomipramine. Ther Drug Monit 1989;11:349-53. 9. Ernst R, Williams L, Dalbey M, Collins C, Pankey S. Homogeneous enzyme imniunoassay (EMIT) protocol for monitoring tricydie antidepressant8 on the COBAS-BIO centrifugal analyzer. Ther Drug Monit 1987;9:85-90. 10 Benitez J, Dahlqvist R, Gustafsson LL, Magnusson A, Sj#{246}qvist F. Clinical pharmacologic evaluation of an assay kit for intoxications with tricydlic antidepressants. Ther Drug Momt 1986;8: 102-5. 11. Nebinger P, Keel M. Specificity data of the tricycic antidepressants assay by fluorescent polarization immunoassay. J Anal Toxicol 1990;14:219-21. 12. Keel M, Roth A, Nebinger P. Zur Spezifitat des Nachweises von trizyklischen Antidepressiva durch Fluoreszenzpolarisation (TDx-System der Fa. Abbott). Arztl Lab 1989;35:57-9. 13. Provost Y, Farinotti R. Dosages immunologiques par polarisation de fluorescence: application aux m#{233}dicaments. Fluorescence polarization unununoassays in drug monitoring. J Pharm Cm 1984;3:197-213. 14. SoucheA, Baumann P, Koeb L, Thermoz P, Azorin JM, Dufour H. Traitement d#{233} Ia depression par l’association de la clomipramine et de la triiodothyronine (LT3). Encephale 1991;17:37-42. 15. HSrtter S, Hiemke C. Column switching and high-performance liquid chromatography in the analysis of aniitriptyline, nortriptyhne and hydroxylated metabolites in human plasma or serum. J Chromatogr 1992;578:273-82. 16. Baumann P. Therapeutisches Drug Monitoring. In: Riederer G, Laux G, Poldinger W, eds. Neuropsychopharmaka. 1. Ailgemeune Grundlagen der Pharmakopsychiatrie. New York/Wien: Springer Verlag, 1992:291-310. 17. Lana G, Riederer P. Plasrnaspiegelbestimmung von Psychopharmaka: Therapeutisches Drug-Monitoring. Stuttgart: Wissenschaftliche Verlagsgesellschaft, 1992. 18. Preskorn SH. Therapeutic drug monitoring of tricycle antidepressants: a means of avoiding toxicity. In: Dahi SG, Gram LF, eds. Clinical pharmacology in psychiatry (Psychopharmacol Ser 7). Berlin/Heidelberg: Springer Verlag, 1989:237-43. 19. Orsulak PJ, Haven MC, Burton ME, Akers LC. Issues in methodology and applications for therapeutic monitoring of antidepressant drugs. Clin Chem 1989;35:1318-25. 20. Orsulak PJ. Therapeutic monitoring of antidepressant drugs: guidelines updated. Ther Drug Moult 1989;11:497-507. 21. Preskorn SH. Tricycle antidepressants: the whys and hows of therapeutic drug monitoring. J Cliii Psychiatry 1989;50:34-42.

CUNICAL CHEMISTRY, Vol. 40, No. 6,

1994

933