Detection of Benzodiazepine Intake in Therapeutic Doses by Immunoanalysis of Urine: Two Techniques Evaluated and Modified for Improved Performance

Size: px

Start display at page:

Download "Detection of Benzodiazepine Intake in Therapeutic Doses by Immunoanalysis of Urine: Two Techniques Evaluated and Modified for Improved Performance"

Beryl Owens
6 years ago
Views:

1 LI.HM.38/2, (1992) Detection of Benzodiazepine Intake in Therapeutic Doses by Immunoanalys of Urine: Two Techniques valuated and Modified for Improved Performance Olof Beck, Pierre Lafolie, Paul Hjemdahl, Stefan Borg, Gudrun Odelius, and Peter Wlrbmg We evaluated the MIT (enzyme-multipled immuno technique) and FPIA (fluorescence polarization immunoassay) urine screening systems for detection of benzodiazepine intake. Healthy male volunteers were given single oral therapeutic doses of alprazolam (2 mg), chlordiazepoxide (25 mg), flunitrazepam (1 mg), lorazepam (3.75 mg), nitrazepam (5 mg), and triazolam (.25 mg), afterwhich urine was collected for the next 32 h. The MIT method failed to detect the intake of flunitrazepam, lorazepam, and nitrazepam. FPIAdid not detect the intake of chlordiazepoxide, flunitrazepam, lorazepam, nitrazepam, and triazolam. Modification of the MIT method to include enzymatic hydrolys did not significantly alter the results obtained with th method. A modification of the FPIA method to include enzymatic hydrolys and a lower cutoff value improved the results considerably, so that we reliably detected all studied substances but flunitrazepam. We conclude that (a) both MIT and FPIA techniques, when used as intended by the manufacturers, are unreliable for the detection of intake of therapeutic doses of these benzodiazepines, and (b) the described modification of the FPIA should provide a much improved tool for detection of benzodiazepineintake. AddItIonal Keyphra.es: enzyme immunoassay polarization immunoassay. Screening Analys of urine widespread for the control of substance abuse. Immunological systems to screen for illicit drugs in urine are now commercially available. They generally provide adequate quality when combined with confirmation analyses to eliminate falsepositive results (1,2). Benzodiazepines have the potential of eliciting dependence and of being abused, both in high doses and in low, therapeutic, doses (3-6). Available screening assays, however, lack sufficient sensitivity to detect some commonly prescribed benzodiazepines, e.g., nitrazepam, flunitrazepam, and lorazepam (7-9). Th will result in unacceptably high rates of false-negative samples in a screening program for benzodiazepines, and the substances may not be detected at all, even when consumed in high doses. Th will seriously undermine the drug Department of linical Pharmacology, Karolinska Hospital, S-141 Stockholm, Sweden, and Department of Psychiatry, St. Gorans Hospital, S Stockholm, Sweden. Received March 5, 1991; accepted December 2, analys program. Several factors complicate the analys for benzodiazepines in urine. One their complex metabolm, which usually results in the recovery of very little of the parent drugs in urine. For example, <1% of the given dose of mtrazepam excreted in urine as the free parent compound (1). Many benzodiazepine metabolites are excreted in urine as glucuronide conjugates (11), e.g., oxazepani (12). Another complicating factor the wide range of therapeutic doses used for the different benzodiazepines. The older generation of benzodiazepines, e.g., chlordiazepoxide, diazepam, and oxazepam are therapeutically admintrated in doses >1 mg, whereas newer benzodiazepines, e.g., aiprazolam, flunitrazepam, lorazepam, nitrazepam, and triazolam, are admintered in lower therapeutic doses (<5 mg). We recently reported on the specific detection of oxazepam in urine by immunoassays (13). By modifying the assay to include hydrolys of oxazepam glucuronide, we could reliably detect oxazepam by use of either the MITTh (enzyme-multipled immuno technique) or the FPIATh (fluorescence polarization immunoassay) systems. Here, we further evaluate these two systems for the detection of several other benzodiazepines, especially low-dose benzodiazepines. We also suggest a modification of the FP!A assay for improving its performance. Materials and Methods Biological Samples Urine samples were collected from nine healthy male volunteers (ages years, mean 37) with no htory of drug or alcohol abuse, after oral intake of the following benzodiazepines: alprazolam (2 mg, n = 6), chiordiazepoxide (25 mg, n = 5), flurntrazepam (1 mg, n = 6), lorazepam (3.75 mg, n = 6), nitrazepam (5 mg, n = 6), and triazolam (.25 mg, n = 6). The subjects ingested benzodiazepine at 22 h after collecting a control sample of urine. At each time of collection, which occurred at random time points, urine was saved both in a separate 1-mL volume and pooled in three intervals (-12 h, h, h) during the first 24 h. The interval between 24 and 32 h was covered by collecting the morning urine at about 32 h after the intake. All urine was collected in plastic containers. After the arrival to the laboratory (within 48 h of drug intake), the samples were stored at -8 #{176} and analyzed within one month. The wash-out period between experiments LIIALHMISTRY, Vol. 38, o. 2,

2 was at least one week for each individual. The experiment was approved by the local thics ommittee of the Karolinska Hospital. Table 1. ross-reactivity of Various Benzodlazeplnes and Metaboiltes In the MIT and FPIA Systems % Immunological Assays Kits for the MIT benzodiazepine assay were obtained from Syva o. (Palo Alto, A) and were used with a Monarch 2 hemtry System (Instrumentation Laboratory, Lexington, MA). The detection limit (cutoff value) used corresponds to.3 mg of oxazepam per liter (low calibrator). Th limit also corresponds to the sensitivity specified in Syva s package insert. o modification of the MIT cutoff value was possible. Kits for the benzodiazepine FPIAassay were obtained from Abbott Labs. (Abbott Park, IL) and used with an ADxTh instrument (Abbott Labs.). Results from the FPIA assay are quantified as ng/ml of nordiazepam equivalents (ig/l). The detection limit (cutoff value) used in th system, recommended by the manufacturer, corresponds to 2 pg of nordiazepam per liter (calibrator), but th may easily be changed by the operator. The sensitivity corresponds to 4 1.tgfLfor nordiazepam (according to Abbott s package insert). Because the sensitivity of the assay reported to be considerably better than the recommended cutoff value, we also evaluated the data with a lower cutoff value (1 pg/l). To do so, the daily calibration control procedure recommended by Abbott must be modified slightly to include proper control for the blank reading of the instrument. Th reading must be <4 gfl. The cross-reactivities of various benzodiazepines and metabolites of interest in the MiT and FPIA systems are given in Table 1. nzymatic hydrolys of urine was performed by usmg -glucuronidase (fi-d-glucuronide glucuronosohydrolase, ) from scherichia coli obtained from Boehringer Mannheim (Mannheim, F.R.G.), which was available as a solution in glycerol/water (1/1 by vol) and used without further treatment. The following hydrolytic procedure was used: we mixed.2 ml of urine and 2 1iL of the enzyme preparation (4 U at 37#{176}, representing a 1.1-fold dilution of the sample) in a test tube and incubated the mixture at ambient temperature (2 #{176} for 3 mm before loading the samples into the immunological analyzer systems. These conditions completely hydrolyze the conjugate forms of oxazepam (13). Using a cutoff value of 1 p.g/l and enzymatic hydrolys in the ma system improved the assay results. Th procedure referred to as the modified method in th report. reatinine concentrations in urine were determined by the Jaffe reaction, with use of the Monarch 2 hemtry System and reagents obtained from Instrumentation Laboratory. Results MIT 272 LIIAL HMISTRY, Vol. 38, o. 2, 1992 The unmodified MiT screening assay was capable of detecting alprazolam intake at all time points (up to 32 h) and in all individuals studied (Table 2). For chlorcompound T Alprazolam 3 a-hydroxyalprazolam - hlordiazepoxide Demoxepam 15 3 Flunitrazepam 6 7-Aminoflunitrazepam - orflunftrazepam - 3-Hydroxyfiunitrazepam Lorazepam 15 ltrazepam 6 7-Aminonitrazepam 3-Hydroxynitrazepam ordlazepam (calibrator FPIA) 15 Oxazepam (calibratormrt) 1 Trlazolam 6 a-hydroxytriazolam - a Data from manufacturers manuals or b our own woik. FPSA b 35b diazepoxide, the detectability appeared to be slow in onset but all individuals gave positive results during the interval h. Flunitrazepam intake produced a positive response only in one individual during the interval h. Lorazepain intake was not detected at all, and the delta absorbance (M) values obtained were blank values. itrazepam showed a pattern similar to fiunitrazepam, but two individuals produced positive samples at different time points. Finally, triazolam was detectable in all individuals tested during the first 12 h after intake; after 18 h, no positive results were obtained. nzymatic hydrolys resulted in decreased ia values and reduced sensitivity for all substances except lorazepam with the MIT assay. FPIA The original FPIA screening method readily detected alprazolam (Table 3). However, the modified assay procedure improved the detectability, completely identifying th compound at all time points. hlordiazepoxide intake was detected in all five subjects at 32 h with the modified FP!A (Table 3). Flunitrazepam, lorazepani, nitrazepain, and triazolam were not detected at all with the original ma screening method, but the modified FPIA detected flunitrazepam in three of six subjects between 12 and 24 h, lorazepam in all six subjects between 12 and 32 h, mtrazepam in five of six subjects between 18 and 32 h, and triazolam in all subjects at 12 h after intake (Table 3). The effect of the enzymatic hydrolys on the response 15b

3 Table 2. DetectIon of Single-Dose Intake of Various Benzodlazeplnes In Urine with the MIT System Tim. after Aiprazolam, hlordlazepozide, Flunltrazepam, Lorazepain, itrazepem, Trlezolam, Intice, h 2 mg 25 mg 1 mg 375 mg 5 mg.25 mg O/6 /5 /6 /6 /6 /6-12 6/ 2/3 /6 /6 /6 6/ / 4/1 1/5 /6 1/5 2/ / 5/ 1/5 /6 1/5 /6 6/ 5/ 1/5 /6 1/5 /6 a o.of positive/no,of negativeresults. Table 3. DetectIon of Single-Dose Intake of Various Benzodlazeplnes In Urine by the FPIA System (and by the Modified FPIA System) Tim. after Intake, h Aiprazolam hiordiazspoxld. Flunltrazsp.m Lor.zspam ftrazep.m Triazolem /6 (/6) /5(/5) /6(/6) /6 (/6) /6(/6) /6 (/6) -12 5/1 (6/) /5(2/3) /6 (2/4) /6 (5/1) /6 (2/4) /6 (6/) /1 (6/) /5(4/1) /6(3/3) /6 (6/) /6 (3/3) /6 (1/5) /2 (6/) /5(4/1) /6(3/3) /6 (6/) /6 (5/1) /6 (/6) 32 6/ (6/) 3/2 (5/) /6(2/4) /6 (6/) /6 (5/1) /6 (/6) Doses as In Table 2. Table 4. ffect of Hydrolys of Urine on FPIA Response after Single-Dose Intake of Various Benzodiazeplnes Relative Incass, mean (S), S Alprazolam 27 (3) hiordlazepoxide 1 (2) Flunftrazepam 24 (7) Lorazepam 488(13) ftrazepam 35 (7) Trlazolam 19 (5) an 5; all others, n = 6. in the FPIA sununarizedin Table 4, which based on data generated by using the sample from each individual with the greatest FPIA response after hydrolys. The hydrolys procedure improves the response for all compounds. When the A responses for the different compounds were related to the creatinine concentrations in urine, it was evident that triazolam was predominantly excreted during the interval -12 h. Alprazolam, flunitrazepam, and nitrazepam were excreted at greatest rates in the interval h, lorazepam at h, and chiordiazepoxide at (or beyond) 32 h after intake (Table 5). Graphic presentations of the individual urinary concentrations of iinmunoresponsive material after intake of the various benzodiazepines are given in Figure 1. Dcussion In a recent review, Sch#{252}tz (9) expressed hopes for the future of immunologically based systems for benzodiazepine screening in urine, given the possibilities of high speed, automation, and economy. However, the currently available systems, MIT and ma, have limitations in the reliable detection of the intake of various specific benzodiazepines in therapeutic doses (7-9). We recently focused our attention on the detection of oxazepam intake and showed that both the MIT and FPIA techniques lack the capability of reliably producing positive responses in urine from patients undergoing oxazepam therapy, because of the presence of oxazepam as the glucuronide conjugate in urine (13). One way of overcoming the problem was enzymatic liberation of oxazepam from the conjugated form. In the present investigation we have further documented the performance of the MIT and i PL techniques for benzodiazepine assays in urine. learly, the MIT technique does not work reliably for lorazepam, fiunitrazepam, and nitrazepam (Table 2). nzymatic hydrolys was expected to substantially improve the detect- Table 5. UrInary xcretion Rate over Time for Various Benzodlazeplnes as Related to reatinine xcretion Tim. sinc. Intake, h Alprazolam 3 ±.8k 29 ± 6 33 ± 6 25 ± 4 15 ± 2 hlordlazepoxide.5 ± ± ± ± ± 1.1 Flunltrazepam 2.6 ± ± ± ± ±.9 Lorazepam.8 ± ± ± ± ±.8 ftrazepam 2.1 ± ± ± ± ±.5 Trlazolam.9 ± ± ±.9 2. ± ±.7 Mean ±S, noidlazepam equivalents, iag/mol1creatinlne. LIIALHMISTRY,Vol. 38, o. 2,

4 I, s , o.! = z = 5) 2 a. 1 o #{149} = z ! 2 a- o. 1 a. u = = z FIg. 1. Urinaryexcretionofimmunoresponslve material (nordiazepam equivalents, in tg/l) measuredwith themodified IPIA system(see text) onan ADx aftertheingestionof 2 mgofaiprazolam(a),25 mgofchlordiazepoxide(b), 1 mg of fiunitrazepam(ci, 3.75 mgoflorazepam (d), 5 mg of nltrazepam (), and.25 mgoftriazolam(f) Data shown are taken from samples collected ateachtimeofurination(random time points) and are not corrected for excreted creatinine ability of lorazepam because, like oxazepam (13), it excreted mainly as the glucuronide conjugate (14). However, although an increase in sensitivity was indeed obtained (not shown), the response was not above the cutoff value for the MIT assay. Th may be explained by the poor cross-reactivity of lorazepam itself in the MIT system (Table 1). For other substances, the hydrolys procedure led to decreased responses in the MiT assay. Th may be due to the minor dilution introduced with the hydrolys. We therefore conclude that, with the exception of oxazepam (13), enzymatic hydrolys should not be routinely included when the MIT system used. In its original form, the ma did not reliably detect the intake of chlordiazepoxide, flunitrazepam, lorazepam, mtrazepam, and triazolam. However, the hydrolys procedure effectively increased the sensitivity of the FPIA system and, in combination with a lower cutoff value (1 g/l, i.e., our modified method), the FPIA assay offered a powerful system for general benzodiazepine detection (Table 3). The results for pooled urine (Table 3) are complemented by Figure 1, which shows for each individual the results at different time points. Figure 1 shows an increase in response over background in all suljects and for all compounds. Th further supports the usefulness of the modified A. The routine application of a 274 LIIAL HMISTRY,Vol.38, o.2, 1992

lower cutoff value in the FPIA has also been suggested by others (8). However, a lower cutoff value alone may produce a greater number of false-positive results.

5 lower cutoff value in the FPIA has also been suggested by others (8). However, a lower cutoff value alone may produce a greater number of false-positive results. Th must be considered and thoroughly controlled for. Data obtained should be validated by verification analyses (e.g., gas chromatography-mass spectrometry) regardless of screening methodology, if they are to be conclusive. hiordiazepoxide undergoes metabolm before its excretion in urine (15,16). Both the parent drug and its main metabolite demoxepam show low (-25%) inununological cross-reactivity in the original FPIA (Table 1), suggesting that the detectability may be low. A small fraction converted to oxazepam glucuromde. The small effect of enzymatic hydrolys on immunoreactive material was therefore expected. Although nordiazepam a minor metabolite of chlordiazepoxide, its higher cross-reactivity makes it more important for the contribution to immunoresponsive material. Th may explain why the detectability of chlordiazepoxide in our study slow in onset and seems to culminpte later than 32 h after intake. Lorazepam eliminated from the body predominantly after conjugation with glucuronic acid (14). The cross-reactivity of free lorazepam low (Table 1). The intake of a therapeutic dose of lorazepam would therefore be expected to produce a low response only after hydrolys of the glucuronide conjugate. Indeed, our results indicate that detectable material with the modified ma present for 32 h. Alprazolam and triazolam are both excreted in urine itlainly as their a-hydroxylated metabolites but also, to some extent, as the parent compounds (17). The metabolites presumably occur in urine as glucuronide conjugates (17). The hydroxylated metabolites of both compounds show lower cross-reactivities than the respective parent drug. In analogy with oxazepam and lorazepam, we therefore expected that our modified method would have improved the responses for alprazolam and triazolam more than it actually did. One explanation for th finding could be significant immunological cross-reactivity of the glucuronide conjugates in the assay. Another explanation could be incomplete conjugation of the metabolites. Flunitrazepam and nitrazepam both undergo extensive metabolm before excretion in urine (1, 18). Predominant metabolites are the 7-amino analogs. In addition, aromatic hydroxylation occurs, followed by glucuronidation. With fiunitrazepam, demethylation also an important metabolic step. The relative crossreactivities for the 7-amino metabolites and desmethylfiunitrazepam are about three- to fivefold less than those of the calibrators used. The cross-reactivities for the hydroxylated forms are unknown. On theoretical grounds, one might therefore predict that enzymatic hydrolys would influence assay results to only a minor degree. onsidering the low (therapeutic) doses used, the low detectability of flunitrazepam and nitrazepam intake with the immunoassays used not unexpected. In conclusion, we have found that detectability of benzodiazepine intake by urine analyses with MiT and FPIA poor, especially for lorazepam, fiunitrazepain, and nitrazepam. The simple modification of the ma technique we suggested offers a much improved screening system for benzodiazepines in general. In the future we will evaluate the performance of th modified method in samples from patients in abuse treatment programs. References 1. Vereby K, Jukofsky D, Mule S. onfirmation of MiT benzodiazepine assay with GL/PD. J Anal Toxicol 1982;6: Hawks R. Analytical methodology. In: Hawks RL, hang, edo. Urine testing for drugs of abuse. ational Institute on Drug Abuse Res Mono 73. Rockville, MD: IDA, 1986: Murphy S, Owen R, Tyrer P. Withdrawal symptoms after six weeks treatment with diazepam. Lancet 1984;ii: Montgomery S, Tyrer P. Benzodiazepines: time to withdraw. J R oil Gen Pract 1988;38: Maletaky B, Klotter J. Addiction to diazepam. mt j Addict : Tyrer P, Owen R, Dawling S. Gradual withdrawal of diazepani after long-term therapy. Lancet 1983;i: Minder, Schaubhut R, Simmier F. Toxicological screening for benzodiazepines in urine: MiT versus high-performance liquid chromatography with photo-diode array detection. Toxicol Lett 1989;45: BAumler J. Der analytche achwe von akuten Vergiftungen. MTA J 1985;7: SchlItz H. Modern screening strategies in analytical toxicology with special regard to new benzodiazepines [Review]. Z Rechstmed 1988;1: Kangas L. itrazepam and its main metabolites. Acta Pharmacol Toxicol 1979;45: Kaplan S, Jack M. Metabolm of the benzodiazepines: pharmacokinetic and pharmacodynamic considerations. In: osta, ed. The benzodiazepines: from molecular biology to clinical practice. ew York: Raven Press, 1983: Alv#{225}n G, Sievers B, Veasman J. Pharmacokinetics of in healthy volunteers. Acta Pharmacol Toxicol 1977;4: Beck, Lafolie P, Odelius G, Bor#{233}us L. Immunological screening of benzodiazepines in urine: improved detection of oxazepam intake. Toxicol Lett 199;52: Greenblatt, Joyce TH, omer WH, et al. linical pharmacokinetics of lorazepam: U. Intramuscular injection. lin Pharmacol Ther 1977;21: Koechuin B, Schwartz M, Krol G, OberhansliW. The metabolic fate of 4-iabeled chlordiazepoxide in man, in the dog and in the rat. J Pharm xp Ther 1965;148: Schwartz M, Postma. Metabolites of demoxazepam, a chiordiazepoxide metabolite in man. J Pharm Sci 1972;61: Fraser A. Urinary screening for alprazolam, triazolam, and their metaboliteswith the M1T d.a.u. benzodiazepine metabolite assay. J Anal Toxicol 1987;11: Amrein R. Zur Pharinakokinetik und zum Metabolmus von Flunitrazepaxn. In: Ahnefeld FW, Bergman H, Burn, et al., eds. Rohypnol (Flunitrazepain) pharmakologche grundlagen klinche Anwendung. Berlin: Springer-Verlag, 1978:8-24. LIIALHMISTRY, Vol. 38, o. 2,

BENZODIAZEPINE FINDINGS IN BLOOD AND URINE BY GAS CHROMATOGRAPHY AND IMMUNOASSAY

BENZODIAZEPINE FINDINGS IN BLOOD AND URINE BY GAS CHROMATOGRAPHY AND IMMUNOASSAY Ilpo RASANEN, Mikko NEUVONEN, Ilkka OJANPERÄ, Erkki VUORI Department of Forensic Medicine, University of Helsinki, Helsinki,