Evaluation of Acetylcodeine as a Specific Marker of Illicit Heroin in Human Hair

Transcription

1 Evaluation of Acetylcodeine as a Specific Marker of Illicit Heroin in Human Hair Pascal Kintzl, *, Carole Jamey 1, Vincent Cirimele 1, Rudolf Brenneisen 2, and Bertrand Ludes 1 9 l lnstitut de M~decine L~gale, 11, rue Humann, F trasbourg, France and 2Institute of Pharmacy, Baltzerstrasse 5, CH-3012 Bern, Switzerland Abstract I In addition to acetylmorphine (6-AM), acetylcodeine (AC) has been suggested as a marker for the use of illicit heroin. Because no procedure was available for AC testing in hair, a new method was developed for the simultaneous identification and quantitation of morphine (MOR), codeine (COD), 6-AM, and AC. After decontamination, each hair specimen was cut into l-ram pieces. A 50-mg aliquot was incubated overnight at 50~ in 1 ml Soerensen buffer (ph 7.6) in presence of 200 ng of MOR-d 3, COD-d3, 6-A/VI-d3, and AC-d 3. After ph adjustment to 8.4, the analytes were extracted in 5 ml of chloroform/isopropanol/nheptane (25:10:65, v/v/v). The organic phase was removed and evaporated to dryness, and the residue was derivatized by silylation (BSTFA + 1% TMCS). Drugs were analyzed by gas chromatography-mass spectrometry in electron impact mode. Limits of quantitation were set to 0.1 rig/rag. Fifty hair specimens obtained from subjects who died from fatal opiate overdose were analyzed. AC was detected in 22 samples in concentrations ranging from 0.17 to 5.60 ng/mg with a mean value of 1.04 ng/mg. 6-AM was also present in these samples at concentrations ranging from 1.35 to ng/mg with a mean value of 7.79 ns/mg. Of the 28 specimens negative for AC, 21 were positive for 6-AM at concentrations ranging from 0.18 to 7.13 ng/mg. When detected, the AC concentrations were an average of 15.5% (2.8 to 32.6%) of the 6-AM concentrations. There was a positive relationship between AC concentrations and 6-AM concentrations (r = 0.915, p = 0.001). Neither AC nor COD was identified in hair specimens collected from 20 subjects taking part in a heroin-maintenance program in Switzerland and receiving pure pharmaceutical heroin hydrochloride daily. Although it is indicative of illicit heroin use, AC would not make a suitable biomarker in place of 6-AM because of its low concentration in hair compared with that of 6-AM and its absence in about 50% of the specimens that tested positive for 6-AM. *Author to whom corres~x~ence should be addressed. Introduction The detection and quantitative measurement of opiates in the hair of human subjects, first reported in 1979 (1), has proven to be a powerful clinical and forensic tool for the investigation of the use or abuse of licit codeine (COD) or morphine (MOR) and illicit heroin (HER). Hair analysis provides long-term information (from weeks to months, depending on the length of the hair shaft) on an individual's drug use, ifi contrast to urinalysis' shortterm information of a few days (2). In fact, the two tests complement each other. It is generally accepted that HER exposure can be documented by detecting 6-acetylmorphine (6-AM) in both urine (3) and hair (4). For several years, it has been demonstrated that 6-AM is the major analyte found in hair of HER users, which is why analytical procedures to prevent hydrolysis of 6-AM to MOR during hair extraction were established. Most procedures used to determine opiates in hair are based on organic solvent incubation (5), enzymatic hydrolysis (6), or acid hydrolysis (7) as demonstrated by the report of the participants to various international interlaboratory comparisons (8). Today, there is no consensus among the active investigators on a cutoff value for 6-AM in hair, and the interpretation of the results can be complicated by the simultaneous use of COD, MOR, and/or HER and a partial conversion by hydrolysis of 6-AM to MOR during the hair preparation. Acetylcodeine (AC) is an impurity of manufacture found in HER. Typical amounts of AC in illicit heroin in eastern France are in the range of 2 to 12%. AC has been recently proposed as a marker complementary to 6-AM in the urine of HER abusers (9,10). However, the authors concluded that although it is indicative of illicit HER use, AC would not make a suitable biomarker in place of 6-AM because of the low concentration in urine compared with that of 6-AM. In order to determine if AC would be a suitable indicator of illicit long-term heroin use, we developed a sensitive gas chromatography-mass spectrometry (GC--MS) procedure for the simultaneous identification of 6-AM, AC, COD, and MOR in hair specimens. Reproduction (~otocopying) of editorial content of this journal is prohibited widlout publisher's permission. 425

2 Materials and Methods Chemicals Chloroform, isopropanol, n-heptane, and dichloromethane were high-performance liquid chromatography grade (Merck, Darmstadt, Germany). All other chemicals were of analytical Table I. Selected Ions (m/z) and Retention Times (R t) for Each Analyte Analytes R t (rain) m/z Codeine-TMS Codeine-d3-TMS !4 Morphine-diTMS Morphine-drdiTMS 9,98 43,_22 Acetylcodeine _,[ Acetylcodeine-d Acetylmorphine-TMS _.9_ Acetylmorphine-d3-TMS _22 * The underlined ions were used for quantitation. Table II. Correlation Coefficients for tinearity, Within-Run Precision (N= 6), Extraction Recoveries (N=3), and timits of Detection for Each Analyte* Wlthin-run limit of precision detection Recovery Analytes r (C.V.%) (pg/mg) (%) Codeine Morphine Acetylcodeine Acetylmorphine * The analytical validation was done using a 50-rag hair sample in all cases. grade and provided by Merck. 6-AM, 6-AM-d3, MOR, MOR-d3, COD, and COD-d3 were purchased from Promochem (Molsheim, France). AC and AC-d3 were purchased from Euromedex (Souffelweyersheim, France). N,O,-Bis(trimethylsilyl)trifluoroacetamide (BSTFA) + 1% trimethylchlorosilane (TMCS) was purchased from Interchim (Montiu{;on, France). Soerensen buffer was prepared by adding 38.8 ml KH2PO4 buffer (9.07 g/l) to 61.2 ml NazHPO4 (11.87 g/l). The ph was adjusted to the value 7.6 with sodium hydroxide. Phosphate buffer (ph 8.4) was prepared with a saturated aqueous solution of di-ammonium hydrogen phosphate. Specimen collection Hair samples were obtained from 50 subjects who had died from fatal HER overdosage, as demonstrated by high blood MOR concentration and the simultaneous identification of 6-AM in urine. Samples were also obtained from 20 subjects taking part in the Swiss heroin-maintenance program in Bern. The latter subjects self-administered HER hydrochloride intravenously under controlled conditions two or three times a day. The total dose was mg/day (11). Strands of hair in the vertex posterior region were cut as closely as possible to the skin, dried, and stored in tubes at room temperature. Before analysis, samples (approximately 100 rag) were decontaminated twice in 5 ml of methylene chloride, for 2 rain, at room temperature. The decontamination washes were not analyzed. Sample extraction Fifty milligrams of hair was cut into small pieces of about 1 mm and incubated overnight at 50~ in 1 ml of Soerensen buffer (ph 7.6) in presence of 200 ng of MOR-d3, COD-d3, 6-AM-d3, and AC-d3, used as internal standards. After the addition of i ml of saturated phosphate buffer at ph 8.4, the homogenate 3, Scan 464 ( min): D (-) m r i ~ ~11~2~6~ 18~ , I 252 2i h'l...,',,... ~',.,i',~,,,2,ll'l'i",,,,'',hh""h'h'i"'h'~',,"j'itl,,, ',',hh'"2 'I,"I" "',' I '%'0'"id6"k~6'"l,~d"i~d 'l~6"'266"'2~d"2~6"'2~6"2~6""366'"3~d '"3~6''" Figure 1. Mass spectrum of acetylcodeine. m/z 426

3 then was extracted with 5 ml of chloroforrrdisopropanol/ n-heptane (25:10:65, v/v/v). After horizontal agitation (20 rain at 100 cycles/rain) and centrifugation (15 rain at 2200 xg), the organic phase was removed and evaporated to dryness. The residue was derivatized using 35 IJL BSTFA-TMCS for 20 rain at 70~ GC-MS A 1.5-pL portion of the derivatized extract was injected into the column of a Hewlett Packard (Palo Alto, CA) GC (5890, series II) via a Hewlett Packard (7673) autosampler. The flow of cartier gas (helium, purity grade N55) through the column (HP5-2ooooo MS capillary column, 5% phenyl-95% methyl- analyte; less than 4% of these specimens tested positive with higher concentrations of COD, clearly indicating COD use. It was sometimes difficult to evaluate the pattern of drug use, particularly when the measured concentrations of 6-AM, MOR, or COD were below 1 ng/mg because of COD contamination of HER and the metabolic conversion of COD to MOR. It has been reported that AC is present in HER in varying amounts, depending on the source and the extent to which MOR is purified from opium. Soine (12) reported that the AC concentration Ion to ): D i0 12 siloxane, 30 m x 0.25-ram i.d., ~m film! thickness) was 1.0 mumin, Injector temperature was 260~ and split- zi< less injection was employed with a split valve off-time of 1.0 rain. The column oven temperature was programmed to rise from an initial temperature of 60~ kept for 1 rain, to 295~ at 30~ and maintained at 295~ for the final 6 rain. The detector was a Hewlett Packard 5972 mass selective detector operated in electron impact (EI) mode. The electron multiplier voltage was set at +400 V above the EI-tune voltage. Analytes were identified and quantifled on the basis of comparison of retention times and the relative abundance of three ions with the deuterated internal standards. Analytical validation For the different analytes, standard calibration curves were obtained by adding 5 (0.1 ng/mg), 25 (0.5 ng/mg), 50 (1.0 ng/mg), 250 (5.0 ng/mg), 500 (10.0 ng/mg), and 2500 ng (50 ng/mg) of each compound prepared in acetonitrile (1 and 10 mg/l) to 50 mg of pulverized blank control hair (obtained from laboratory personnel and previously tested to be free of opiates). Recovery and within-run precision for the analytes were determined by adding 100 ng of each opiate to 50 mg of powdered blank control hair, corresponding to a final concentration of 2 ng/mg hair. The detection limits were evaluated with decreasing concentrations of the drugs, until a response equivalent to three times the background noise was observed. Results and Discussion ij. I zooooo i soooo l 'so''' kolod ' ' ' ' ~ 6 o ' Ion to ): D I0 12.'so' ' ' ko'.oo' ''' ' i0'.20''' i0140''' i016d Time (min) Ion ( to ): D 9.80''' io.oo''' i0.20''' i0140'' i01go' ' ' Ion ( to ): D '8o''' io'.oo''' '10120' ' ' io'.4o''' io'.6o' Over the years, more than 3000 hair specimens were tested for opiates in this laboratory. In most cases, 6-AM was the major Figure 2. Selected ion chromatograms for acetylcodeine (m/z 341,282, and 229) and acetyicodeine-cl 3 (m/z 344). Concentration in hair was 0.31 ng/mg. 427

4 is usually 2 to 20% relative to HER and may be as high as 45%. At the initial stages of this study, we tried to apply our classic procedure for opiates (7) to target AC. However, AC was unstable using a 0.1M HCI hair hydrolysis with rapid conversion of AC to COD. Six replicative 50-rag aliquots of blank hair were only Table lu. Analytes Concentrations in 50 Hair Specimens Codeine Morphine Acetylcodeine Acetylmorphine Subject (ng/mg) (ng/mg) (ng/mg) (n~/n~) , ND* 0.13 ND ND ND ND ND ND , ND ND ND ND 10 ND 0.34 ND , NO ND 0.75 t6 ND 0.35 ND t ND ND ND ND ND 0.29 ND NO ND ND NO 0.54 O, , NO O ND , ND ND ND ND ND 33 ND 0.65 ND , , ND 0, ND ND 0.76 ND ND 38 NO 1.42 NO ,86 ND ND 1, ,92 1,84 0, ND 0,42 ND ND ND 46 ND ND ND ND ND ND ND ND ND 50 I.I 9 0,55 ND 0.53 " ND, not detected {under 0.1 rig/rag). spiked with 200 ng AC. After overnight incubation in 1 ml 0.1M HCl and a three-step liquid-liquid extraction and addition of 200 ng of AC-d3 and COD-d3 just before silylation, we observed a > 80% loss of AC and subsequent formation of COD. In a parallel experiment of AC-spiked hair, incubation in methanol for 4 h, a procedure that has been described for HER analysis in hair (5), induced a > 50% loss of AC. After overnight incubation in Soerensen buffer, AC was found to be stable. The conversion of AC to COD was < 8% after the entire procedure. As reported by O'NeaI and Poklis (9), it is not necessary to derivatize AC, but it is highly recommended to simultaneously analyze MOR. AC does not hydrolyze to COD during derivatization, and the silyl derivatives were found to be stable at least for 24 h. Table I shows the ions monitored for each analyte and for the deuterated internal standards and the retention times (Rt). The mass spectrum obtained for AC is shown in Figure 1. Under the chromatographic conditions used, there was no interference with the analytes or the internal standards by any extractable endogenous materials present in hair. Figure 2 is a typical chromatngram obtained after extraction of the ions m/z 341,282, and 229. The AC concentration was 0.31 ng/mg of hair. Correlation coefficients for the linearity, within-run precision (n = 6), extraction recoveries (n = 3), and limits of detection are presented in Table II. The limits of detection (calculated for a signal-to-noise ratio of 3), using a 50-rag hair sample, were in the ng/mg range. Limit of quantitation was set to 0.1 ng/mg for all the analytes. The within-run precision of the method was assessed by testing six replicate samples through the entire procedure in one analysis day. The values ranged from 7.4 to I3.2% and were acceptable for a screening procedure of several opiates. Individual results are presented in Table III. Fifty hair specimens obtained from subjects who died from fatal opiate overdose were analyzed. AC was detected in 22 samples in concentrations ranging from 0.17 to 5.60 ng/mg with a mean value of 1.04 ng/mg. 6-AM was also present in these samples at concentrations ranging from 1.35 to ng/mg with a mean value of 7.79 ng/mg. In 11 samples, the COD concentration was higher than the corresponding AC concentration. This is surprising because the parent drug is generally the predominent analyte, but it can be explained by either concomitant COD intake or degradation of AC. However, the former hypothesis is the more probable. The AC concentrations observed were low compared with those of (i-an. AC did not test positive when the 6-AN concentration was below 1 ng/mg of hair. Of the 28 specimens negative for AC, 21 were positive for 6-AN at concentrations ranging from 0.18 to 7.13 ng/mg, clearly indicating that 6-AN is a better marker of illicit HER abuse. When detected, the AC concentrations were an average of 15.5% (2.8 to 32.6%) of the 6-AN concentrations, which is in accordance with the ratio between AC and HER in the street drug. There was a positive relationship between AC concentrations and 6-AN concentrations (r = 0.915, p = 0.001), with the equation AC (ng/mg) = x 6-AM (ng/mg) Based on this formula, the average concentration of AC is about 8 times lower than the corresponding 6-AN concentration. It would therefore be necessary to have an assay 10 times more sensitive forac than 6-AM, which 428

5 is obviously not the case for GC--MS, particularly for concentrations of 6-AM < 1 ng/mg. O'Neal and Poklis (9) also found a positive relationship between AC concentrations and 6-AM concentrations in urine, with r = (10). AC and other opiates were also tested in hair collected from addicts taking part in the Swiss heroin-maintenance program in Bern (11). Subjects self-administered HER hydrochloride intravenously under controlled conditions. Neither AC nor COD was ever identified in these samples, clearly indicating that the "pharmaceutical" HER is pure. In this population, the idea of using AC as a marker of concomitant street-drug abuse should be avoided, because, as demonstrated in this work, AC is either missing or below the detection limit in approximately 50% of the cases that tested positive for illicit HER abuse. Conclusion Although it is found in illicit HER, AC would not make a suitable biomarker in place of 6-AM because of its low concentration in hair compared with that of 6-AM and its lack of detection in about 50% of the specimens that tested positive for 6-AM. References 1. A.M. Baumgartner, RE Jones, W.A. Baumgartner, and C.T. Black. Radioimmunoassay of hair for determining opiate abuse histories. ]. NucL Med. 20: (1979). 2. R Kintz. Drug testing in addicts: a comparison between urine, sweat and hair. Ther. Drug Monit. 18: (1996). 3. E.J. Cone, P. Welch, J.M. Mitchell, and B.D. Paul. Forensic drug testing for opiates: I. Detection of 6-acetylmorphine in urine as an indicator of recent heroin exposure; drug and assay considerations and detection times. J. Anal ToxicoL 15:1-7 (1991). 4. B.A. Goldberger, Y.H. Caplan, T. Maguire, and E.J. Cone. Testing human hair for drugs of abuse: III. Identification of heroin and 6-acetylmorphine as indicators of heroin use. J. Anal Toxicol. 15: (1991). 5. G. Kauert and J. R6hrich. Concentrations of A%tetrahydrocannabinol, cocaine and 6-monoacetylmorphine in hair of drug abusers. Int. J. Leg. Med. 108: (1996). 6. M.R. Moeller, P. Fey, and R. Wennig. Simultaneous determination of drugs of abuse (opiates, cocaine and amphetamines) in human hair by GC/MS and its application to a methadone treatment program. Forensic Sci. Int. 63: (1993). 7. P. Kintz and P. Mangin. Simultaneous determination of opiates, cocaine and major metabolites of cocaine in human hair by gas chromatography/mass spectrometry. Forensic Sci. Int. 73: (1995). 8. P. Kintz. Interlaboratory comparison of quantitative determinations of drug in hair samples. Forensic Sci. Int. 70: (1995) 9. C.L. O'Neal and A. Poklis. Simultaneous determination of acetylcodeine, monoacetylmorphine, and other opiates in urine by GC-MS. J. AnaL ToxicoL 21 : (1997). 10. C.L. O'Neal and A. Poklis. The detection of acetylcodeine and 6-monoacetylmorphine in opiate positive urines. Abstract 28. Proceedings of the Society of Forensic Toxicologists, Salt Lake City, October 5-9, R Kintz, R Bundeli, R. Brenneisen, and B. Ludes. Dose-concentration in hair relationships in a controlled heroin-maintenance program. J. AnaL Toxicol. 22: (1998) 12. W.H. Soine. Clandestine drug synthesis. Med. Res. Rev. 6:41-74 (1986). Manuscript received March 3, 1998; revision received May 12,