Simultaneous Detection and Quantification of Amphetamines ...

Journal of Analytical Toxicology, Vol. 32, September 2008

Simultaneous Detection and Quantification of Amphetamines,Diazepam and its Metabolites, Cocaine and its Metabolites, and Opiates in Hair by LC–ESI-MS–MS Using a Single Extraction Method Eleanor I. Miller*, Fiona M. Wylie, and John S. Oliver Forensic Medicine and Science, University of Glasgow, University Place, Glasgow, G12 8QQ, Scotland

Abstract A liquid chromatography–tandem mass spectrometry method was developed and validated for the simultaneous identification and quantification of amphetamines, diazepam and its metabolites, cocaine and its metabolites, and opiates from hair using a single extraction method. As part of the method development, Gemini C18, Synergi Hydro RP, and Zorbax Stablebond-Phenyl LC columns were tested with three different mobile phases. Analyte recovery and limit of detection were evaluated for two different solid-phase extraction methods that used Bond Elut Certify™ and Clean Screen® cartridges. Phosphate buffer (pH 5.0) was chosen as the optimum hair incubation medium because of the high stability of cocaine and 6-monoacetylmorphine using this method and faster sample preparation. The optimized method was fully validated. Linearity was established over the concentration range 0.2–10 ng/mg hair, and the correlation coefficients were all greater than 0.99. Total extraction recoveries were greater than 76%, detection limits were between 0.02 and 0.09 ng/mg, and the intra- and interday imprecisions were generally less than 20% in spiked hair. The intraand interbatch imprecision of the method for a pooled authentic hair sample ranged from 1.4 to 23.4% relative standard deviation (RSD) and 8.3 to 25.4% RSD, respectively, for representative analytes from the different drug groups. The percent matrix effect ranged from 63.5 to 135.6%, with most analytes demonstrating ion suppression. Sixteen postmortem samples collected from suspected drug-related deaths were analyzed for the 17 drugs of abuse and metabolites included in the method. The method was sufficiently sensitive and specific for the analysis of drugs and metabolites in postmortem hair samples. There is scope for the inclusion of other target drugs and metabolites in the method.

Introduction Hair was first analyzed for drugs almost 30 years ago by Baumgartner et al. (1) in 1979 for the purpose of determining * Author to whom correspondence should be addressed.

opiate abuse histories. It has since been investigated as an alternative biological matrix to more conventional matrices such as blood and urine, and advantages to using hair have been identified. These include the ability of hair testing to reveal chronic drug use, which is desirable for drug monitoring; noninvasive sample collection that can be easily supervised, reducing the risk of sample adulteration; and the application of segmental analysis to determine the approximate time of drug exposure. A restriction of hair testing is the limited sample weight provided for testing and the potentially low drug concentrations. A single extraction method that can be used for the analysis of multiple drug groups is particularly useful for samples of low weight. There are, however, still some controversies with respect to the exact mechanisms of drug incorporation, passive contamination and the effectiveness of wash procedures, and the potential for bias of cosmetically treated hair and of hair from individuals of different cultures. Until these issues are resolved, hair testing should not be a single source of evidence in a toxicological investigation, but where possible should be used in combination with blood and urine analysis to provide information on both chronic and acute drug use or to rule out the use of drugs. The consensus opinion of the Society of Forensic Toxicologists (SOFT), which currently proposes that hair evidence is a useful specimen in forensic investigations only when supported by additional evidence of drug use, has not changed since 1990 (2). Reports on hair testing in the scientific literature generally restrict analysis to one or two drug groups (3–9). Routine hair samples are initially screened by immunoassay (particularly in workplace testing where there are large numbers of negatives) and then positives are confirmed with a more sensitive and specific technique such as gas chromatography–mass spectrometry (GC–MS) or liquid chromatography–tandem mass spectrometry (LC–MS–MS). The Society of Hair Testing (SoHT) proposed guidelines in 2004 for immunoassay screening and confirmation cut-offs for amphetamines, cocaine and its

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

457


metabolites, and opiates in hair (10). GC–MS and LC–MS–MS methods have been reported for the simultaneous qualitative analysis of common drugs of abuse in hair (11–13) and the simultaneous quantification of two or three target drug groups (3–9,14–19). Recently, methods using capillary zone electrophoresis–MS have been reported for the quantitative analysis of three drug groups in hair (16,17). The simultaneous quantification of amphetamines, benzodiazepines, cocaine and its metabolites, and opiates in hair by GC–MS has been reported by one group (20). The method involved an overnight acid extraction at 50°C followed by a mixed mode solid-phase extraction (SPE) clean-up and a single GC– MS injection, operated in full scan mode. An earlier study by the same group developed a qualitative screening method for the same drugs reported in their more recent quantitative method. The qualitative method used an overnight incubation in methanol for 18 h at 45°C with no clean-up step. Significant baseline noise was observed using this method and the authors noted that this could potentially cause difficulty in the identification of low drug concentrations. A qualitative LC–MS–MS screening method has been reported for the analysis of amphetamines, benzodiazepines, cocaine and its metabolites, and opiates (12). The drugs were extracted from the hair by incubation in acidic mobile phase. An aliquot of extract was injected into the LC–MS–MS system. The authors proposed that this method could be used as an alternative to immunoassay screening. This method was not, however, developed for quantitative purposes. This present study reports the development and validation of a single extraction procedure for the quantification of amphetamines, diazepam, and its metabolites, cocaine and its metabolites, and opiates in hair collected from suspected drugrelated deaths. The drugs included in the method were those that are commonly encountered in our laboratory in blood samples as identified by the in-house database. The method was validated for 17 drugs of abuse and metabolites and applied to postmortem samples generally in the range of 10–30 mg for segments excluding roots, but also to root samples, some of which weighed less than 1 mg. A single extraction method that could be applied in the analysis of a wide range of drugs would prove to be very useful, particularly for the analysis of weight-limited hair samples taken from polydrug users, including drug-related death cases where several drug classes could be present in each sample.

Experimental Samples

All samples were postmortem head hair samples submitted to the toxicology laboratory of the Forensic Medicine and Science Section at the University of Glasgow for routine testing. The samples were tested for diagnostic purposes, and the results were reported to Procurators Fiscal. Chemicals

Methanol, acetone, acetonitrile, acetic acid, ammonium

458

hydroxide, ethyl acetate, formic acid, dichloromethane, propan-2-ol, and potassium dihydrogen phosphate were purchased from BDH (Poole, U.K.) and were of analytical grade. Ammonium formate and sodium dodecyl sulfate were purchased from Sigma-Aldrich (Dorset, U.K.). d,l-Amphetamine, d,l-methamphetamine, d,l-MDA, d,l-MDMA, d,l-MDEA, diazepam, nordiazepam, oxazepam, temazepam, cocaine, benzoylecgonine, ecgonine methyl ester, cocaethylene, morphine, 6-monoacetylmorphine (6-MAM), codeine, and dihydrocodeine were obtained from Promochem (Teddington, U.K.). Deuterated internal standards, including d,l-amphetamine-d5, d,l-methamphetamine-d5, d,l-MDA-d5, d,lMDMA-d5, d,l-MDEA-d5, diazepam-d5, nordiazepam-d5, oxazepam-d5, temazepam-d5, cocaine-d3, benzoylecgonine-d3, ecgonine methyl ester-d3, cocaethylene-d3, morphine-d3, 6MAM-d3, codeine-d3, and dihydrocodeine-d6, were obtained from LGC Promochem. World Wide Monitoring Clean Screen® SPE columns were purchased from Kinesis (Cambridgeshire, U.K.). Bond Elut Certify™ SPE columns were purchased from Crawford Scientific (Strathaven, U.K.). Standard solutions

Individual drug stock standard solutions and deuterated drug standards were obtained as 100 µg/mL prepared in methanol. A combined working drug solution of d,l-amphetamine, d,l-methamphetamine, d,l-MDA, d,l-MDMA, d,lMDEA, diazepam, nordiazepam, oxazepam, temazepam, cocaine, benzoylecgonine, ecgonine methyl ester, cocaethylene, morphine, 6-MAM, codeine, and dihydrocodeine was prepared at 1 µg/mL by 100-fold dilution with methanol. This was achieved by adding 25 µL of each 100 µg/mL drug solution into a 25-mL volumetric flask and making up the volume to the 25-mL mark with methanol. A combined working deuterated internal standard solution of d,l-amphetamine-d5, d,lmethamphetamine-d5, d,l-MDA-d5, d,l-MDMA-d5, d,l-MDEAd5, diazepam-d5, nordiazepam-d5, oxazepam-d5, temazepamd5, cocaine-d3, benzoylecgonine-d3, ecgonine methyl ester-d3, cocaethylene-d3, morphine-d3, 6-MAM-d3, codeine-d3, and dihydrocodeine-d6 was also prepared at 1 µg/mL in a similar way to the working drug solution. Instrumentation

LC–MS–MS analysis was carried out using a Surveyor HPLC system with an LCQ Deca XP Plus™ ion trap MS (Thermo Finnigan, San José, CA). During method development, three different LC columns were tested. The Gemini and Synergi Hydro RP columns were purchased from Phenomenex (Torrance, CA). Both were C18 columns but the Synergi Hydro RP column also had polar end-capping. The Zorbax Stablebond (SB) Phenyl column was purchased from ChromTech (Cheshire, U.K.). This was a non-endcapped phenyl modified silica column. The columns (and their dimensions) used were the Gemini C18 (150 mm × 2.0 mm, 3-µm particle size), the Synergi Hydro RP (150 mm × 2.0 mm, 4-µm particle size), and the Zorbax Stablebond (SB) (50 mm × 2.1 mm, 3.5-µm particle size). These columns were fitted with guard columns with identical packing material which was purchased from the same companies as the LC columns. Guard column dimensions were


4 mm × 2.0 mm, 5-µm particle size for the Gemini column; 4 mm × 2.0 mm, 4-µm particle size for the Synergi Hydro RP column; and 12.5 × 2.1 mm, 5-µm particle size for the Zorbax SB Phenyl column.

Method Development Initial tuning and ion identification

The optimum tuning parameters, precursor, and product ions were identified for each analyte. This was achieved through a combination of automatic and manual tuning. Optimum LC and mobile phase combination

Initially, three mobile phase systems and three LC columns were investigated to determine which mobile phase system and column produced the greatest peak area response for the target analytes (5,12,21). The Gemini C18, Synergi Hydro RP, and Zorbax SB columns were each tested with 3 mM ammonium formate + 0.001% formic acid and acetonitrile, 10 mM ammonium acetate + 0.001% formic acid and acetonitrile, and 10:10:80 methanol/acetonitrile/20 mM formate buffer and 35:35:65 methanol/acetonitrile/20 mM formate buffer gradient systems. Three 20-µL injections of 50 ng drug/200 µL mobile phase were run for each column and mobile phase system combination. The retention times of the analytes were also compared for each system. Stability of cocaine and 6-MAM

The stability of cocaine and 6-MAM was tested in a variety of incubation media including methanol, phosphate buffer pH 5.0, 0.1 M, 0.05 M, and 0.1 M hydrochloric acid and methanol containing 1, 2, 5, 10, and 25% ammonium hydroxide (v/v). Three samples were prepared by spiking 1.5 mL of the particular incubation medium with 50 ng of cocaine and 6-MAM. The ammoniated methanol incubations were subsequently left to incubate for 18 h at room temperature. The aqueous acidic media, methanol, and phosphate buffer incubations were subsequently left for 18 h at 45°C. After incubation, 50 ng of cocaine-d3 and 6-MAM-d3 were added to the vials and with the exception of the phosphate buffer extract, the vial contents were evaporated to dryness at room temperature under nitrogen. The phosphate buffer extracts were extracted by SPE , and the results from the extracts were compared to three unextracted standards to determine stability. Three unextracted standards were prepared at the same concentration and kept in the fridge at 4°C during the incubation period. Fifty nanograms of deuterated internal standard was added to the unextracted standards at the same time as the incubated samples, and the vial contents were evaporated to dryness under nitrogen, with the exception of the phosphate buffer extract. The samples, once dry, were reconstituted in 200 µL of mobile phase initial conditions, and 20 µL was injected. SPE methods

Bond Elut Certify and Clean Screen DAU SPE columns were

investigated for the simultaneous extraction of the drugs of abuse from hair. The Bond Elut Certify method was previously published, having been applied in the simultaneous screening of acidic, neutral, and basic drugs from oral fluid (21). The cartridges were conditioned with 2 mL methanol and 2 mL phosphate buffer (pH 6.0). After the samples were applied to the cartridges, the cartridges were washed with 1 mL deionized water followed by 0.5 mL 0.01 M acetic acid. The cartridges were then air dried under full vacuum for 10 min, and 50 µL methanol was added; the cartridges were dried for an additional 2 min. The elution step for the basic analytes was modified slightly from this published method and involved 1.5 mL ethyl acetate with 2% aqueous ammonium hydroxide and 1.5 mL dichloromethane/isopropanol/aqueous ammonium hydroxide (78:20:2, v/v/v) with a 2-min drying step between the two solutions. The Clean Screen SPE method intended for the extraction of drugs of abuse in urine and provided by the SPE manufacturers was adapted for hair extracts. Clean Screen (ZSDAU020) extraction cartridges were conditioned sequentially with 3 mL methanol, 3 mL deionized water, and 1 mL phosphate buffer (0.1 M, pH 5.0). The vortex mixed samples were loaded in 2 mL phosphate buffer (0.1 M, pH 5.0) and allowed to drip through without vacuum. The columns were then washed with 3 mL phosphate buffer (0.1 M, pH 5.0) and 1 mL acetic acid (1.0 M) and dried for 5 min under full vacuum. The drugs were eluted using 2 mL methanol/2% aqueous ammonium hydroxide. Comparison of SPE methods

SPE clean-up. Hair was collected from five members of laboratory personnel. Ten milligrams ± 0.1 mg hair was washed and subsequently extracted using the Bond Elut Certify and Clean Screen methods. The chromatograms for each blank hair extract using both SPE methods were compared visually to determine which method produced the cleanest extracts overall. Total extraction recovery from spiked hair samples. In the spiked hair experiment, 10 mg ± 0.1 mg blank hair was weighed out into three vials and was spiked with 50 ng analyte mix. Two unextracted samples were prepared at the same concentration and kept refrigerated at 4°C during the extraction. The samples were then incubated in 1.5 mL phosphate buffer (pH 5.0) for 18 h at 45°C and extracted by the Bond Elut Certify and Clean Screen methods. Fifty nanograms of deuterated internal standard was added to the eluant after the extraction and also to the unextracted samples. The samples were evaporated to dryness at room temperature under nitrogen and subsequently reconstituted in 200 µL of mobile phase initial conditions. Twenty microliters was injected for each sample. Incubation recovery from authentic hair samples. Three postmortem hair samples which had drug-positive blood results were washed once with 0.1% sodium dodecyl sulfate wash, twice with deionized water washes, and twice with dichloromethane washes. Each wash involved a 10-min ultrasonication. The root–0.5 cm segment was removed, and the 0.5–3.5 cm segment was cut up into smaller segments of 2–3 mm in length. This cut-up segment was subsequently split to compare the methanol and phosphate buffer incubation methods. After the incubation, the samples were extracted

459


using the Bond Elut Certify method described in the following section.

LOD and LOQ

These data had already been established as part of the comparison of SPE methods.

Limits of detection (LOD) and quantification (LOQ)

The LOD and LOQ were determined for each drug using spiked hair. Ten milligrams of blank washed hair was spiked with 0.1, 0.2, 0.3, 0.4, 0.5, 0.75, 1, 2, and 5 ng of each drug and 50 ng of deuterated internal standard. The samples were then incubated, extracted by both SPE methods, and analyzed by LC–MS–MS. The LOD and LOQ were calculated at a signal-tonoise ratio of 3 and 10, respectively.

Method Validation The optimized method was fully validated. The extracts were reconstituted in 150 µL of mobile phase for the validation work to allow a greater quantity of analyte to be injected onto the LC column, ultimately improving the assay sensitivity. Linearity

Linearity was determined over the concentration range 0.2– 10 ng/mg spiked hair. This sample weight was chosen for the validation because the hair case samples were analyzed in segments and the sample weights were generally low.

Imprecision of the procedure using authentic hair samples

A pooled hair sample was tested to determine the imprecision of the procedure using authentic hair samples which contained incorporated drugs. Imprecision experiments using spiked hair do not truly represent the imprecision of the procedure for authentic hair samples. An intraday imprecision between extracts was calculated for n = 3. The interday imprecision was calculated by analyzing the pooled hair sample in duplicate on two days (n = 4). Case samples

Sixteen postmortem case samples were tested for diagnostic purposes. The root–0.5 cm segment was removed from each sample. The remaining hair was cut into 3-cm segments. The segments were subsequently washed according to the procedure described previously. After removing the second dichloromethane wash, the hair samples were left to dry overnight at room temperature. The samples were then incubated and extracted using the validated method.

Method Development Results and Discussion

Total extraction recovery

The percent total extraction recovery was determined for each analyte at 0.5, 1, and 5 ng/mg. This recovery takes into account a combination of the incubation step, SPE step, and matrix effect because the reference was unextracted standards. Matrix effect

Matrix effect was assessed by comparing the mean peak-area ratio of product ion/internal standard of extracted blank hair samples spiked with standard solution following SPE (A) (n = 6) versus the mean peak-area ratio of product ion/internal standard of unextracted standards prepared in mobile phase at equivalent concentrations (B) (n = 6). Matrix effect (%) was calculated for a low standard concentration of 0.5 ng/mg, according to the calculation proposed in a previously published article (22). Accuracy and imprecision

Intrabatch accuracy and imprecision were assessed over the linear range for five extracted 10 mg blank hair samples spiked at low, medium, and high concentrations (0.5, 1, and 5 ng/mg). The intrabatch imprecision was expressed as a percent relative standard deviation (RSD) calculated using the five individual values obtained on the same day. Accuracy was calculated by dividing the mean measured concentration of five extracts by the theoretical spike concentration and was expressed as a percentage of the theoretical spike concentration. Interbatch imprecision was evaluated for five replicate hair extracts spiked at these concentrations on five separate days (ntotal = 25). It was expressed as a % RSD value, which was calculated using all 25 individual values.

460

Initial tuning and ion identification

Optimum tuning parameters, precursor and product quantitation ions are shown in Table I for the drugs tested in the method. There was only one product ion produced during fragmentation for the amphetamines because of their low molecular weights. Optimum LC column and mobile phase combination

The mobile phase system producing the greatest average peak area response for each drug was recorded as 100% and the average peak area responses for the other two mobile phase systems were calculated as a percent of the optimum mobile phase system response as shown in Table II. The Zorbax SB-Phenyl column and mobile phase system B produced the greatest average peak area response for the greatest number of analytes compared to the other two columns and mobile phase systems. Although the Zorbax SBPhenyl column was initially selected for further work, a pressure problem was encountered during further method development. The column pressure would not stabilize and exceeded the LC–MS–MS maximum pressure setting of 5800 psi. The manufacturers recommend using such mobile phases as acetonitrile/water mixtures because the column is moderately non-polar in nature, so the mobile phase selection was theoretically suitable. However, it was not feasible to use this column in practice. The Synergi Hydro RP column and mobile phase system A were therefore selected for further work because this column and mobile phase combination was stable and produced a higher relative average percent peak area response for more analytes compared with the Gemini column


and mobile phase system C. As expected, the retention times and elution order of the analytes were different depending on the LC column and mobile phase system used. However, the number of scan events required for each retention window was comparable for each system. Furthermore, it was not possible to analyze all 17 analytes in 1 injection because of the nature of the ion trap MS. Therefore, the analytes were analyzed in three separate injections. This was done to minimize the number of scan events per time window to achieve maximum sensitivity and to maximize the number of points across the peak to obtain more accurate and reproducible analyte quantifications. The first injection was for cocaine and its metabolites and opiates. The second injection was for amphetamines, and the third was for diazepam and its metabolites. Figures 1–4 show the results of some positive hair samples which were extracted using the method used in the validation work (i.e., Clean Screen extraction cartridge and the Synergi Hydro RP column and mobile phase system A).

The results obtained for the aqueous alkaline media showed an increased cocaine and 6-MAM hydrolysis with increasing concentration of aqueous ammonium hydroxide. Both the acidic and alkaline incubation media produced significant hydrolysis of 6-MAM. Cocaine showed significant hydrolysis using 25% aqueous ammonium hydroxide in methanol. It was therefore decided to compare the recovery of the methanol and phosphate buffer incubation media for future work in authentic hair samples because both cocaine and 6-MAM were relatively stable using these methods. The methanol and phosphate buffer incubation media produced 6-MAM recoveries of 88.6% and 102.9%, respectively, and cocaine recoveries were 99.4% and 102.3%, respectively.

Comparison of SPE methods SPE clean-up

The Clean Screen method produced the cleanest extracts overall for all drug classes and for hair from all five individuals.

Stability of cocaine and 6-MAM

A stability experiment for cocaine and 6-MAM in various incubation media was carried out with a view to selecting the incubation method that would yield the highest cocaine and 6MAM recoveries. The extent of cocaine and 6-MAM degradation was negligible when using the methanol and phosphate buffer incubations. This supports the findings of another paper that evaluated cocaine and 6-MAM hydrolysis following a methanol and phosphate buffer (pH 5.0) incubation (23). In contrast, all concentrations of aqueous acidic media resulted in significant 6-MAM hydrolysis, whereas cocaine degradation was negligible. This result was in agreement with previously published results (24).

Total extraction recovery

The Clean Screen method produced higher recoveries for a greater number of analytes compared to the Bond Elut Certify method for spiked hair. All of these were 76% or greater. Two analytes of significant difference were morphine and MDMA, where there was approximately 20% difference in recovery with Clean Screen, producing higher recovery. In contrast, oxazepam had a higher recovery using the Bond Elut Certify extraction by approximately 13%. In the U.K., oxazepam is likely to be present in hair as a result of metabolism rather than oxazepam use; therefore, the Clean Screen method would de-

Table I. Optimum Tuning Parameters and Precursor and Product ions for Each Analyte

Analyte EME Cocaine Cocaethylene Benzoylecgonine Morphine 6-MAM Codeine Dihydrocodeine Oxazepam Temazepam Nordiazepam Diazepam Amphetamine Methamphetamine MDA MDMA MDEA

Sheath Gas (AU)

Auxillary Gas (AU)

Capillary Temperature (°C)

Collision Energy (%)

Precursor Ion (m/z)

Product Ion (m/z)

Retention Time (tR )

10 10 10 10 20 20 10 10 20 20 20 20 15 15 15 15 15

10 10 10 10 10 10 5 5 20 20 15 20 5 5 15 15 15

210 300 210 300 200 210 240 240 300 300 300 300 210 210 220 280 220

29 30 30 28 33 34 34 33 29 29 41 42 23 28 21 25 20

200 304 318 290 286 328 300 302 287 301 271 285 136 150 180 194 208

182*, 168 182*, 150 196*, 150 168*, 150 201*, 229 211*, 268 215*, 243 245*, 201 269*, 241 283*, 255 243*, 140 257*, 222 119* 119* 163* 163* 163*

2.8 19.4 20.8 10.5 10.3 15.2 14.7 13.8 19.2 20.4 20.6 21.9 10.8 12.1 11.7 13.2 13.6

* Quantitation ion.

461


tect diazepam and other breakdown products that had a higher recovery than oxazepam. Incubation recovery from authentic hair samples using the methanol and phosphate buffer incubation

Although extraction recoveries using spiked hair do not re-

flect real cases, the previous recovery experiment was a means of testing the total extraction recovery of the analytes using a particular SPE method in the presence of hair matrix. The experiment was used as a means of determining the optimum extraction cartridge. It was difficult to obtain sufficient real samples that contained all 17 analytes. However, the extraction

Table II. Relative Percent Peak Area Response for the Three Optimum LC Column and Mobile Phase Combinations Gemini C18 + Mobile Phase System C Analyte EME Cocaine Cocaethylene Benzoylecgonine Morphine 6-MAM Codeine Dihydrocodeine Oxazepam Temazepam Nordiazepam Diazepam Amphetamine Methamphetamine MDA MDMA MDEA

Relative %PAR (n = 3)

tR (min)

100.0 46.6 48.1 42.4 100.0 100.0 100.0 100.0 45.1 37.8 4.0 4.3 85.2 31.3 89.2 46.2 54.0

1.8 11.5 15.9 5.7 2.3 4.8 4.1 3.4 23.6 25.1 25.8 27.2 4.2 4.9 4.5 5.1 6.2

Figure 1. Hair sample positive for amphetamine.

462

Synergi Hydro RP + Mobile Phase System A Relative %PAR (n = 3) 59.5 67.8 100.0 100.0 89.7 28.2 34.2 43.2 100.0 59.0 100.0 100.0 68.7 71.4 33.8 57.5 53.6

Zorbax Phenyl + Mobile Phase System B

tR (min)

Relative %PAR (n = 3)

tR (min)

4.9 19.4 20.8 10.5 10.3 15.2 14.7 13.8 19.2 20.4 20.6 21.9 10.8 12.1 11.7 13.2 13.6

62.9 100.0 71.0 72.3 47.3 33.5 31.2 43.3 79.9 100.0 83.4 49.5 100.0 100.0 100.0 100.0 100.0

3.2 21.7 23.8 11.7 9.5 16.1 14.3 13.3 19.4 20.5 20.5 21.7 9.6 12.3 12.1 14.5 16.7


recoveries of cocaine and its metabolites, amphetamine, and opiates were tested for three real cases. With the exception of the cocaine concentration detected in sample 1, the phosphate buffer extraction produced higher

quantitative results overall, extracting greater concentrations of EME, benzoylecgonine, 6-MAM, codeine, and amphetamine. Morphine was not detected in sample 3 by either incubation method. A possible explanation for this is the low levels of 6-

Figure 2. Hair sample positive for diazepam (A) and nordiazepam (B).

Figure 3. Hair sample positive for morphine (A), codeine (B), dihydrocodeine (C), and 6-MAM (D).

463


Figure 4. Hair sample positive for EME (A), cocaine (B), cocaethylene (C), and benzoylecgonine (D).

Table III. Analyte Recovery from Spiked Hair (n = 3)

Table IV. % Matrix Effect for 0.5 ng/mg Spiked Hair Extracts (n = 6)

% Recovery (n = 3) Analyte EME Cocaine Cocaethylene Benzoylecgonine Morphine 6-MAM Codeine Dihydrocodeine Oxazepam Temazepam Nordiazepam Diazepam Amphetamine Methamphetamine MDA MDMA MDEA

0.5 ng/mg

1 ng/mg

5 ng/mg

99.2 (6.6) 87.2 (8.9) 79.4 (13.5) 95.3 (8.5) 91.2 (2.2) 109.6 (7.7) 99.6 (2.3) 86.5 (7.5) 108.0 (8.8) 76.3 (7.2) 97.9 (5.6) 96.9 (6.0) 108.1 (17.0) 93.6 (2.2) 82.8 (7.9) 98.3 (12.6) 102.5 (13.2)

96.4 (3.7) 91.8 (4.9) 97.8 (6.6) 100.7 (4.5) 80.1 (6.7) 104.5 (4.6) 94.4 (5.4) 80.9 (3.2) 96.1 (3.8) 96.5 (9.4) 94.7 (3.7) 98.3 (2.1) 99.6 (6.0) 106.0 (0.8) 106.2 (12.5) 100.0 (9.7) 104.3 (10.3)

97.7 (1.0) 97.3 (3.1) 92.5 (2.8) 99.4 (3.2) 92.3 (1.1) 94.3 (2.5) 99.3 (0.2) 100.2 (0.7) 100.6 (3.3) 95.7 (6.4) 93.2 (2.7) 94.7 (0.6) 91.0 (0.6) 94.2 (2.3) 97.6 (6.1) 98.2 (0.6) 90.8 (1.6)

MAM detected in the samples. The 6-MAM/morphine ratios observed in the case samples that were tested for this paper ranged from 0.14 to 28.00 (5.47 mean). If the mean 6MAM/morphine ratio was applied to calculate a predicted morphine concentration from the 6-MAM concentration found, it would be 0.15 ng/mg, which is lower than the LOQ of the method. Overall, the phosphate buffer incubation produced better

464

Analyte EME Cocaine Cocaethylene Benzoylecgonine Morphine 6-MAM Codeine Dihydrocodeine Oxazepam Temazepam Nordiazepam Diazepam Amphetamine Methamphetamine MDA MDMA MDEA

% Matrix Effect (% RSD) 119.2 (5.9) 96.4 (13.7) 82.9 (66.1) 85.1 (21.6) 85.4 (11.1) 110.6 (18.9) 92.6 (9.5) 102.8 (5.1) 67.6 (43.8) 63.5 (65.1) 130.3 (22.6) 95.5 (13.9) 97.9 (3.4) 135.6 (10.8) 76.4 (17.2) 90.4 (8.7) 91.1 (11.9)

qualitative and quantitative results compared to the methanol incubation. Benzodiazepines were not detected in any of the samples, so it was not possible to compare the two incubations for this drug group. Analyte recoveries for the methanol incubation were generally lower for most analytes, as reported in various studies (23–25). It has been proposed that the generally higher recoveries obtained using a phosphate buffer incubation is a result


Table V. Intraday Accuracy and Precision Intrabatch Precision (% RSD, n = 5) Analyte EME Cocaine Cocaethylene Benzoylecgonine Morphine 6-MAM Codeine Dihydrocodeine Oxazepam Temazepam Nordiazepam Diazepam Amphetamine Methamphetamine MDA MDMA MDEA

0.5 ng/mg 5.5 7.2 10.0 7.1 8.4 11.4 8.9 7.0 11.6 18.7 14.4 6.6 9.2 6.0 9.1 10.6 6.7

1 ng/mg 4.9 5.8 5.0 6.3 7.5 5.2 6.5 5.3 8.9 13.9 9.6 5.6 6.7 6.1 5.0 6.6 6.6

Interbatch Precision (% RSD, n = 25)

Accuracy (% of Target, n = 5)

5 ng/mg

0.5 ng/mg

1 ng/mg

5 ng/mg

0.5 ng/mg

1 ng/mg

5 ng/mg

3.8 3.0 3.7 3.6 6.9 4.8 2.0 1.4 6.0 9.4 6.6 2.9 7.0 4.4 7.1 5.9 3.5

18.3 10.7 19.0* 22.5 16.4 20.5 18.7 18.2 16.8 20.2* 24.3* 13.3 16.9 20.7 20.8 15.9 15.1

9.9 7.7 15.8 12.0 15.5* 13.6 17.5 12.9 14.2* 18.9 22.9* 12.2 15.7 8.9 19.4 10.8 11.0

5.6 7.1 10.2 8.8 11.0 8.2 9.3 8.5 10.8* 10.6 10.6 9.2 12.2 8.6 19.6 10.0 9.8

106.0 82.0 96.0 92.0 94.0 90.0 94.0 98.0 100.0 88.0 114.0 96.0 112.0 116.0 88.0 104.0 114.0

99.0 96.0 103.0 92.0 91.0 87.0 107.0 116.0 106.0 94.0 111.0 98.0 80.0 92.0 95.0 89.0 88.0

109.6 93.6 96.2 99.6 100.6 101.6 108.4 100.4 101.6 105.8 97.6 97.2 104.2 88.6 99.2 101.6 110.0

* n = 24.

of the water molecules penetrating the keratinized hair to a greater extent than an organic solvent (26). Furthermore, the non-keratinous hair regions could potentially provide diffusion channels, both into and out of the hair, for small drug molecules in the presence of water (26).

Table VI. LOD and LOQ Values for Bond Elut Certify and Clean Screen Methods Bond Elut Certify Analyte

Clean Screen

LOD (ng/mg)

LOQ (ng/mg)

LOD (ng/mg)

LOQ (ng/mg)

0.03 0.10 0.04 0.03 0.10 0.11 0.06 0.03 0.09 0.04 0.08 0.03 0.08 0.04 0.10 0.02 0.03

0.11 0.34 0.12 0.11 0.32 0.38 0.23 0.11 0.31 0.12 0.27 0.11 0.26 0.13 0.32 0.08 0.09

0.03 0.04 0.04 0.02 0.07 0.09 0.06 0.05 0.02 0.08 0.07 0.06 0.03 0.04 0.02 0.04 0.03

0.08 0.13 0.14 0.05 0.26 0.31 0.22 0.16 0.06 0.26 0.23 0.16 0.10 0.13 0.07 0.13 0.09

LOD and LOQ

The Clean Screen SPE method produced lower detection and quantitation limits for the majority of analytes (Table III). The detection limits ranged from 0.02 to 0.09 ng/mg, and the quantitation limits ranged was 0.05 to 0.31 ng/mg. This may be a consequence of the higher recoveries and reduced matrix interference observed using the Clean Screen method compared to the Bond Elut Certify method. The Clean Screen method was therefore validated in further work.

Method Validation Results Linearity

All regression lines had R2 values > 0.99 over the concentration range 0.2–10 ng/mg. This was the acceptance criterion used for the determination of linearity. Total extraction recovery

The percent total extraction recovery was calculated as the mean peak-area ratio of product ion/internal standard for the samples in which the standard solution was added before SPE (n = 3), divided by the mean peak-area ratio of product ion/internal standard for the samples in which the standard solution was added after SPE (n = 3). The recovery values for all the analytes in spiked hair are

EME Cocaine Cocaethylene Benzoylecgonine Morphine 6-MAM Codeine Dihydrocodeine Oxazepam Temazepam Nordiazepam Diazepam Amphetamine Methamphetamine MDA MDMA MDEA

given in Table III. All gave high recoveries (> 76%) at all three concentrations and % RSD values for these were ≤ 17%. Matrix effect

The percent matrix effect ranged from 63.5 to 135.6%. Most analytes (12 out of 17) demonstrated ion suppression, some being affected to a greater extent than others. Oxazepam and temazepam product ions were suppressed the most, 36.5%

465


and 32.4%, respectively. With the exception of nordiazepam, methamphetamine, and MDA, all of the other analytes were suppressed or enhanced by ± 20%, which is the value which is currently regarded as acceptable for accuracy and imprecision experiments in forensic toxicology (2). At present, no criteria have been published regarding acceptance criteria for matrix effects. ESI is reported to be particularly prone to matrix effects compared to other ionization techniques such as atmospheric pressure chemical ionization (27). This may account for the high variation in some of the ion suppression data. In addition, hair from six individuals was used in the matrix effect evaluation where in reality a larger sample group should have been used, consisting of a range of hair colors, hair from people of different races, and a range of dyed and chemically treated hair. A larger sample group could reduce the average matrix effect or increase it from the one quoted in this paper. The results of the % matrix effect evaluation are given in Table IV.

Table VII. LC–MS–MS Results for Postmortem Cases

Hair Sample Number

Weight (mg)

(1) Root–0.5 cm

2.85

0.91 MOR* 0.97 6-MAM 1.38 COD 2.93 DHC

(1) 0.5–3.5 cm

23.80

2.02 MOR 1.31 6-MAM 0.37 COD 1.22 DHC 0.10 DZ 0.37 NDZ

(2) Root–0.5 cm

2.05

1.67 MOR 1.96 6-MAM 0.34 DZ 2.26 NDZ

(2) 0.5–3.5 cm

25.45

1.23 MOR 0.12 6-MAM 0.45 COD 0.13 DHC 0.06 COC 0.42 BZE 1.03 MDMA 0.22 OXAZ 0.10 TMZ 0.52 DZ 0.67 NDZ

(3) Root–0.5 cm

2.74

NEG for all drug groups

Accuracy and imprecision

The intrabatch accuracy and imprecision values were acceptable by SOFT guidelines of ± 20% for the three concentrations tested (2). The results are given in Table V. Intraday accuracy ranged from 80 to 116%. Imprecision (% RSD) ranged from 1.4 to 18.7% for analytes. As expected, lower precision was observed at the lower concentrations. Similarly, the interbatch imprecision was generally acceptable within SOFT guidelines of ± 20% for the three concentrations tested (2). Interbatch precision for nordiazepam was slightly higher at the lower concentrations; however, the guidelines also state that ± 25–30% may be acceptable for some analytes. Imprecision (% RSD) ranged from 5.6 to 24.3% for all analytes. As found in the intrabatch imprecision experiment, the % RSD was higher at lower concentrations. The % RSD range was higher for interbatch imprecision compared to intrabatch imprecision, due to instrumental and extraction interbatch variation and also a greater number of data values. Some of the analyte data shown in Table IV for interbatch precision was calculated for n = 24. This was due to the removal of

466

LC–MS–MS Results (ng/mg)

(3) 0.5–3.5 cm

19.09

0.09 MOR 0.52 6-MAM 0.21 COD 0.29 COC 0.16 BZE 1.26 MDMA 0.12 DZ 0.36 NDZ

(3) 3.5–6.5 cm

18.60

0.53 6-MAM 0.96 COC 0.45 BZE 1.60 MDMA

(4) Root–0.5 cm

(4) 0.5–3.5 cm

3.21

21.12

0.58 MOR 0.77 6-MAM 0.32 DHC 1.40 NDZ

Cause of Death

Bronchopneumonia due to heroin intoxication

No information available

Heroin and alcohol intoxication

Heroin and alcohol intoxication

Bronchopneumonia due to aspiration of gastric contents due to morphine intoxication

0.16 MOR 0.58 6-MAM Table continues on next page

* Abbreviations: MOR, morphine; 6-MAM, 6-monoacetylmorphine; COD, codeine; DHC, dihydrocodeine; COC, cocaine; BZE, benzoylecgonine; COCAETH, cocaethylene; EME, ecgonine methyl ester; AMP, amphetamine; DZ, diazepam; NDZ, nordiazepam; OXAZ, oxazepam; and TMZ, temazepam.


Table VII. LC–MS–MS Results for Postmortem Cases (Continued)

Hair Sample Number

Weight (mg)

LC–MS–MS Results (ng/mg)

LOD and LOQ Cause of Death

0.20 COD* 0.09 DHC 0.86 COC 0.11 BZE 0.19 AMP 1.42 MDMA 0.04 DZ 0.29 NDZ (5) Root–0.5 cm

1.71

(5) 0.5–3.5 cm

12.62

(6) Root–0.5 cm

1.03


(6) 0.5–3.5 cm

18.49

2.32 COD

(7) Root–0.5 cm

1.49

0.13 COD

(7) 0.5–3.5 cm

8.38

1.10 MOR 4.55 6-MAM 0.11 COD

(8) Root–0.5 cm

1.05


(8) 0.5–3.5 cm

9.24

0.58 6-MAM 0.08 COD 0.21 DHC

(9) Root–0.5 cm

(9) 0.5–3.5 cm

3.16

28.55

1.39 6-MAM

The LOD and LOQ values for the Clean Screen method are provided in Table VI. The LOD values ranged from 0.02 to 0.09 ng/mg hair and the LOQ ranged from 0.05 to 0.31 ng/mg hair. Imprecision of the procedure using authentic hair samples

Heroin intoxication

0.22 MOR 0.49 6-MAM 0.32 COD 10.18 COC 3.11 BZE 0.43 COCAETH 0.24 EME 0.11 MDMA 0.17 DZ 0.52 NDZ

0.71 6-MAM 2.05 COC 0.31 BZE 0.25 COCAETH

an outlier in the data sets.

Gastrointestinal hemorrhage due to esophageal ulcers due to chronic alcohol abuse

Heroin intoxication

The intrabatch imprecision for n = 3 aliquots of the pooled authentic hair sample extracted on the same day reported here as ng/mg ± SD (% RSD) were morphine 0.29 ± 0.01 (4.2%); 6-MAM 4.2 ± 0.57 (13.7%); cocaine 4.6 ± 0.26 (5.6%); benzoylecgonine 0.93 ± 0.01 (1.4%); amphetamine 3.3 ± 0.47 (14.2%); and diazepam 0.62 ± 0.15 (23.4%). The interbatch imprecision for n = 2 aliquots of the pooled authentic hair sample extracted on two different days (i.e., n = 4), reported here as ng/mg ± SD (% RSD) were morphine 0.30 ± 0.03 (9.0%); 6-MAM 3.3 ± 0.84 (25.4%); cocaine 4.7 ± 0.40 (8.4%); benzoylecgonine 0.90 ± 0.13 (14.6%); amphetamine 3.6 ± 0.30 (8.3%); and diazepam 0.58 ± 0.06 (11.1%). The interbatch imprecision for diazepam was for n = 3 as a result of a bad injection by the instrument.

Heroin intoxication

Case Sample Results

Morphine (heroin) intoxication

0.04 MOR 1.12 6-MAM 13.99 COC 4.73 BZE 0.21 EME 0.54 COCAETH 0.21 AMP 0.18 DZ 0.17 NDZ Table continues on next page


The case sample results are provided in Table VII. The acceptance criteria used for qualifier ratios was ± 20% as recommended by SOFT (2). Because the amphetamines have relatively low molecular weights and only fragment to produce one major product ion, the total ion count to major product ion ratio was used for identification of amphetamines. At least one drug class was detected in every sample. Some of the cocaine, dihydrocodeine, and amphetamine results were greater than the highest point on the calibration graph. These concentrations were calculated by extrapolation because there was an insufficient amount of sample to repeat the analysis. The hair samples were reported as positive for a particular drug group according to the SoHT guidelines for opiates, cocaine, and amphetamines (10). There are currently

467


no published guidelines for diazepam and its metabolites in hair. Amphetamine, benzoylecgonine, cocaethylene, cocaine, codeine, diazepam, dihydrocodeine, morphine, 6-MAM, and nordiazepam were detected in some of the root–0.5 cm segments. Ten of the 16 root samples tested positive for at least

one drug group. Furthermore, four of these root samples tested positive for two drug groups. These root samples were of very low weight (≤ 3.27 mg) and the high sensitivity of the LC–MS– MS instrument has been highlighted by the positive results. A greater number of analytes were detected in the segments excluding roots compared to the root samples. Also, the ranges found were much broader than the ranges obtained for the root segments. This may have been due to greater segment Table VII. LC–MS–MS Results for Postmortem Cases (Continued) weights, or alternatively, the individual may have used drugs in the past which Weight LC–MS–MS Results (mg) (ng/mg) Cause of Death Hair Sample Number they had stopped using. Oxazepam, temazepam, ecgonine methyl ester, and (10) Root–0.5 cm 3.27 0.60 COC* MDMA were detected in addition to the 0.18 BZE compounds found in the roots. The drug 0.24 COCAETH concentrations found in this present 0.31 DZ study were within the ranges reported in 0.34 NDZ Cocaine, heroin, and another study which used the same phosdiazepam intoxication phate buffer incubation (23). (10) 0.5–3.5 cm 29.47 0.08 6-MAM The cause of death is also reported for 2.50 COC each of the cases in Table VII. The hair 1.88 BZE analysis provided the pathologist with 0.05 COCAETH extra information on prior drug use his0.15 EME tory such as chronic or naïve use. The 0.26 OXAZ root–0.5 cm segment was cut from the 0.14 TMZ bulk of the hair because this segment 0.24 DZ 0.64 NDZ could be contaminated with drug-positive blood circulating in the body at the (11) 0.5–3.5 cm 8.55 0.49 MDMA Heroin and alcohol intoxication time of death. This segment was analyzed (11) 3.5–6.5 cm 8.69 0.02 COD as an indicator of recent drug use. The 0.5–3.5 cm segment therefore should not (12) Root–0.5 cm 0.35 14.48 DHC Sertraline, methadone, and be contaminated with blood from this morphine overdose source. The 0.5–3.5 cm segment was (12) 0.5–3.5 cm 10.40 0.87 MOR tested to provide information on drug use 0.12 6-MAM in the 3.5 months prior to death, as16.18 DHC suming the generally accepted growth 0.46 MDMA rate of 1 cm/month. 3.51 OXAZ 0.68 DZ 0.92 NDZ (13) Root–0.5 cm

0.21

NEG for all drugs groups

(13) 0.5–3.5 cm

3.79

0.71 6-MAM

(14) 0.5–1.5 cm

1.25


Cerebro-vascular accident (“stroke”)

(15) Root–0.5 cm

1.75

3.54 MOR 1.31 6-MAM

Heroin intoxication

(15) 0.5–3.5 cm

18.66

(16) Root–0.5 cm

2.49

5.59 AMP

(16) 0.5–3.5 cm

18.21

7.29 AMP

Heroin intoxication

1.03 MOR 1.32 6-MAM 0.31 COD 10.72 DHC Amphetamine intoxication


468

Conclusions The developed and validated LC–MS– MS method is capable of simultaneously identifying and quantifying amphetamines, diazepam and its metabolites, cocaine and its metabolites, and opiates from one hair sample of 8–30 mg for segments excluding roots. It is also capable of detecting and quantifying these drug groups in root segments of low weight (in one case, < 1 mg). The method proved to be sufficiently sensitive and specific for the analysis of 17 drugs and metabolites in postmortem hair samples. There is scope for the inclusion of other target drugs and metabolites into the method. Maximum infor-


mation is obtained from one hair sample which is extremely useful when the sample weight is limited. It is currently not possible to relate the quantity of drug detected in hair with the amount of drug ingested or the frequency of drug use; a qualitative screening method may provide all the required information for as accurate an interpretation as possible at this time. However, the quantitative method can be used to determine if an individual is a chronic drug user. It can be used to compare the segments within an individual’s hair to give an indication of drug use over a period of time. However, care must be taken with this for longer segments where some leaching of drug from the hair might have occurred through hygienic procedures and environmental exposure. The quantitative method is also useful where further analysis is required on drug users’ hair, and where possible, to analyze known users’ hair to establish an in-house database of results for this particular method and hair types.

References 1. A.M. Baumgartner, P.F. Jones, W.A. Baumgartner, and C.T. Black. Radioimmunoassay of hair for determining opiate-abuse histories. J. Nucl. Med. 20: 748–752 (1979). 2. SOFT/AAFS Forensic Laboratory Guidelines 2006. http://www.soft tox.org/docs/Guidelines%202006%20Final.pdf. Accessed July 2007. 3. K. Lachenmeier, F. Musshoff, and B. Madea. Determination of opiates and cocaine in hair using automated enzyme immunoassay screening methodologies followed by gas chromatographic–mass spectrometric (GC–MS) confirmation. Forensic Sci. Int. 159: 189– 199 (2006). 4. W.E. Brewer, R.C. Galipo, K.W. Sellers, and S.L. Morgan. Analysis of cocaine, benzoylecgonine, codeine, and morphine in hair by supercritical fluid extraction with carbon dioxide modified with methanol. Anal. Chem. 73: 2371–2376 (2001). 5. K.B. Scheidweiler and M.A. Huestis. Simultaneous quantification of opiates, cocaine and metabolites in hair by LC–APCIMS–MS. Anal. Chem. 76: 4358–4363 (2004). 6. B.A. Goldberger, Y.H. Caplan, T. Maguire, and E.J. Cone. Testing human hair for drugs of abuse, III. Identification of heroin and 6acetylmorphine as indicators of heroin use. J. Anal. Toxicol. 15: 226–231 (1991). 7. E. Cognard, S. Rudaz, S. Bouchonnet, and C. Staub. Analysis of cocaine and three of its metabolites in hair by gas chromatography–mass spectrometry using ion-trap detection for CI–MS– MS. J. Chromatogr. B 826: 17–25 (2005). 8. E. Han, W. Yang, J. Lee, Y. Park, E. Kim, M. Lim, and H. Chung. Correlation of methamphetamine results and concentrations between head, axillary, and pubic hair. Forensic Sci. Int. 147: 21– 24 (2005). 9. E.I. Miller, F.M. Wylie, and J.S. Oliver. Detection of benzodiazepines in hair by LC–ESI-MS–MS. J. Anal. Toxicol. 30(7): 441– 448 (2006). 10. Society of Hair Testing. Recommendations for hair testing in forensic cases. Forensic Sci. Int. 145: 83–84 (2004). 11. S. Paterson, N. McLachlan-Troup, R. Cordero, M. Dohnal, and

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

S. Carman. Qualitative screening for drugs of abuse in hair using GC–MS. J. Anal. Toxicol. 25: 203–208 (2001). R. Kronstrand, I. Nyström, J. Strandberg, and H. Druid. Screening for drugs of abuse in hair with ion spray LC–MS–MS. Forensic Sci. Int. 145: 183–190 (2004). S. Gentili, M. Cornetta, and T. Macchia. Rapid screening procedure based on headspace solid-phase microextraction and gas chromatography–mass spectrometry for the detection of many recreational drugs in hair. J. Chromatogr. B 801: 289–296 (2004). L. Skender, V. Karacˇić, I. Brcˇić, and A. Bagarić. Quantitative determination of amphetamines, cocaine, and opiates in human hair by gas chromatography–mass spectrometry. Forensic Sci. Int. 125: 120–126 (2002). C. Girod and C. Staub. Analysis of drugs of abuse in hair by automated solid-phase extraction, GC–EI-MS, and GC ion trap– CI-MS. Forensic Sci. Int. 107: 261–271 (2000). R. Gottardo, A. Fanigliulo, F. Bortolotti, G. De Paoli, J.P. Pascali, and F. Tagliaro. Broad spectrum toxicological analysis of hair based on capillary zone electrophoresis time-of-flight mass spectrometry. J. Chromatogr. A 1159: 190–197 (2007). R. Gottardo, F. Bortolotti, G. De Paoli, J.P. Pascali, I. Mikšík, and F. Tagliaro. Hair analysis for illicit drugs by using capillary zone electrophoresis–electrospray ionization-ion trap mass spectrometry. J. Chromatogr. A 1159: 185–189 (2007). M.R. Moeller, P. Fey, and R. Wennig. Simultaneous determination of drugs of abuse (opiates, cocaine, and amphetamine) in human hair by GC–MS and its application to a methadone treatment program. Forensic Sci. Int. 63: 185–206 (1993). M. Klys, S. Rojek, J. Kulikowska, E. Bo´zek, and M. ´Scislowski. Usefulness of multiparameter opiates-amphetamines-cocainics analysis in hair of drug users for the evaluation of an abuse profile by means of LC–APCI-MS–MS. J. Chromatogr. B 854: 299–307 (2007). R. Cordero and S. Paterson. Simultaneous quantification of opiates, amphetamines, cocaine, and metabolites and diazepam and metabolite in a single hair sample using GC–MS. J. Chromatogr. B 850: 423–431 (2007). F.M. Wylie, H. Torrance, R.A. Anderson, and J.S. Oliver. Drugs in oral fluid. Part I. Validation of an analytical procedure for licit and illicit drugs in oral fluid. Forensic Sci. Int. 150: 191–198 (2005). B.K. Matuszewski, M.L. Constanzer, and C.M. Chavez-Eng. Strategies for the assessment of matrix-effect in quantitative bioanalytical methods based on HPLC–MS–MS. Anal. Chem. 75: 3019– 3030 (2003). F.S. Romolo, M.C. Rotolo, I. Palmi, R. Pacifici, and A. Lopez . Optimized conditions for simultaneous determination of opiates, cocaine and benzoylecgonine in hair samples by GC–MS. Forensic Sci. Int. 138: 17–26 (2003). A. Polettini, C. Stramesi, C. Vignali, and M. Montagna. Determination of opiates in hair: effects of extraction methods on recovery and on stability of analytes. Forensic Sci. Int. 84: 271–279 (1997). V. Cirimele, P. Kintz, and P. Mangin. Comparison of different extraction procedures for drugs in hair of drug addicts. Biomed. Chromatogr. 10: 179–182 (1996). G. Skopp, L. Potsch, and R. Aderjan. Experimental investigations on hair fibres as diffusion bridges and opiates as solutes in solution. J. Forensic Sci. 41: 117–120 (1996). Applications of LC–MS in Toxicology, A. Polettini, Ed. Pharmaceutical Press, London, U.K., 2006. Manuscript received December 17, 2007; revision received March 20, 2008.

469