Detection and quantification of codeine-6-glucuronide, hydromorphone-3-glucuronide, oxymorphone-3-glucuronide, morphine 3-glucuronide and morphine-6-glucuronide in human hair from opioid users by LC–MS-MS

Detection and quantification of codeine-6-glucuronide, hydromorphone-3-glucuronide,... Abstract Current hair testing methods that rely solely on quantification of parent drug compounds are unable to definitively distinguish between drug use and external contamination. One possible solution to this problem is to confirm the presence of unique drug metabolites that cannot be present through contamination, such as phase II glucuronide conjugates. This work demonstrates for the first time that codeine-6-glucuronide, hydromorphone-3-glucuronide, oxymorphone-3-glucuronide, morphine-3-glucuronide and morphine-6-glucuronide are present at sufficient concentrations to be quantifiable in hair of opioid users and that their concentrations generally increase as the concentrations of the corresponding parent compounds increase. Here, we present a validated liquid chromatography tandem mass spectrometry method to quantify codeine-6-glucuronide, dihydrocodeine-6-glucuronide, hydromorphone-3-glucuronide, morphine-3-glucuronide, morphine-6-glucuronide, oxymorphone-3-glucuronide, codeine, dihydrocodeine, dihydromorphine, hydrocodone, hydromorphone, morphine, oxycodone, oxymorphone and 6-acetylmorphine in human hair. The method was used to analyze 46 human hair samples from known drug users that were confirmed positive for opioids by an independent laboratory. Glucuronide concentrations in samples positive for parent analytes ranged from ~1 to 25 pg/mg, and most samples had glucuronide concentrations in the range of ~1 to 5 pg/mg. Relative to the parent concentrations, the average concentrations of the four detected glucuronides were as follows: codeine-6-glucuronide, 2.33%; hydromorphone-3-glucuronide, 0.94%; oxymorphone-3-glucuronide, 0.77%; morphine 3-glucuronide, 0.59%; and morphine-6-glucuronide, 0.93%. Introduction In 2004, the Substance Abuse and Mental Health Services Administration (SAMHSA) proposed a revision to the mandatory guidelines for federal workplace drug testing programs to expand the types of specimens that may be tested to include head hair, sweat and oral fluid, in addition to urine (1). Hair is seen as an advantageous matrix for drug testing for a number of reasons including, but not limited to, its long detection window, difficulty to adulterate, and ease of sample collection, transportation and storage. The proposed guidelines allow for testing of up to 1.5 inches of head hair closest to the scalp, which represents a period of ~90 days (2–4). For opiates, the currently proposed confirmatory compounds are codeine, morphine and 6-acetylmorphine, with cutoffs at 200 pg/mg for each (1). These analytes and cutoffs parallel guidelines set forth by the Society of Hair Testing (SOHT) (5, 6). Although hair testing is currently used for a variety of applications, one issue that remains under significant debate regarding its use in criminal investigations and federal workplace drug testing is the ability to definitively distinguish drug use from external contamination (7–17). Drugs may become incorporated into hair through multiple physiological processes (4, 18–21) or through passive environmental exposure (13, 18, 22–24). One approach to address the issue of environmental contamination, pioneered by Baumgartner and Hill, is to incorporate an extensive wash into the sample preparation procedure to remove surface contamination, followed by analysis of the final wash solution and application of a wash criterion (15, 25, 26). This procedure, when combined with determination of suitable metabolite ratios, has been used to identify samples that are contaminated rather than positive because of drug use (8, 15, 27). Another approach to distinguish use from contamination is to develop methods to detect unique metabolites that are definitive indicators of consumption instead of, or in addition to, the parent drugs. Schafer et al. (28) recently explored this approach and developed an assay to detect hydroxycocaine metabolites as evidence of cocaine ingestion rather than relying on the presence of cocaethylene and norcocaine. In March 2017, the Federal Bureau of Investigation (FBI) published a new set of criteria for reporting cocaine use that depends on analysis of wash solutions and the presence of hydroxycocaine metabolites (29). Despite these new reporting criteria for cocaine, the FBI Laboratory “continues to encourage research into the identification of drug metabolites or other markers that are uniquely associated with drug consumption” (29). In this study we focus on detection of metabolites in place of or in addition to parent compounds of opioid drugs. The choice of metabolite is critical (30) because some metabolites can be present as process impurities, meaning they have a distinct probability of being present as a result of contamination by the parent compounds. Other metabolites may be degradation products formed because of exposure to hair care products subsequent to parent drug contamination. Also, some metabolites are commercially available drugs themselves. Opioids in particular are problematic in this regard as many opioid metabolic products are active and marketed as standalone drugs. For example, hydromorphone is a metabolite of both morphine and hydrocodone, which are themselves metabolites of codeine (31). All of these compounds are separately marketed prescription drugs. Phase II conjugated metabolites are ideal markers of use because they are not products of common degradation pathways, as is the case for many phase I metabolites, and are not commercially available drugs. The most common phase II metabolic transformation is glucuronidation (conjugation with glucuronic acid) to form a glucuronide conjugate (32). Aside from ethyl glucuronide, very few publications have addressed the detection of glucuronides in human hair. Wang et al. (33) developed a method for the detection of oxazepam glucuronide and temazepam glucuronide in human hair but did not detect either of these metabolites in hair from study volunteers who consumed a single 10-mg dose of diazepam. Pichini et al. (34) used liquid chromatography tandem mass spectrometry (LC–MS-MS) to detect 11-nor-delta-9-tetrahydrocannabinol-9-carboxylic acid glucuronide (THC-COOH-glu) in 20 user hair samples at concentrations ranging from 0.5 to 8.6 pg/mg hair. Beasley et al. (35) attempted to detect THC-COOH-glu in human hair using matrix-assisted laser desorption/ionization (MALDI)-MS, but during analysis of hair spiked with THC-COOH-glu, the glucuronide fragmented to form THC-COOH, resulting in low sensitivity, and the method was not applied to user hair. Kim et al. (36) used LC–MS-MS to detect propofol glucuronide in hair from five users. Wang et al.(37) used LC–MS-MS to quantify gamma-hydroxybutyric acid (GHB) glucuronide in human hair and determined that GHB glucuronide did not accumulate appreciably in the hair of two GHB abusers. Several glucuronide conjugates have been detected in rat hair, including propofol (38, 39), morphine (40–42) and codeine (41). Here, we present a method for the extraction and quantification of glucuronide conjugates of opioids in human head hair and apply the method to 46 opioid-positive user hair samples. Structures of the opioid analytes and glucuronide conjugates included in the method are presented in Figures 1 and 2, respectively. Figure 1. View largeDownload slide Structures of the opioid analytes included in the validated method. Figure 1. View largeDownload slide Structures of the opioid analytes included in the validated method. Figure 2. View largeDownload slide Structures of the glucuronide conjugates included in the validated method. Figure 2. View largeDownload slide Structures of the glucuronide conjugates included in the validated method. Methods Materials High-performance LC (HPLC)-grade methanol and water were purchased from Fisher Scientific (Fair Lawn, NJ). Hydrochloric acid (HCl), isopropanol, potassium phosphate monobasic, potassium phosphate dibasic and bovine serum albumin (BSA) were purchased from Sigma Aldrich (St. Louis, MO), and ammonium hydroxide was acquired from Fisher Scientific. Ammonium formate and formic acid were purchased from Alfa Aesar (Ward Hill, MA) and Electron Microscopy Sciences (Hatfield, PA), respectively. M3 Reagent was purchased from Comedical (Trento, Italy). Deionized water was obtained from an in-house system. Solid-phase extraction columns (Oasis MCX) were purchased from Waters (Milford, MA). Reference standards and internal standards listed in Table I, as well as heroin, norcodeine, norhydrocodone, normorphine, noroxycodone and noroxymorphone were purchased from Cerilliant (Round Rock, TX). Drug-free human hair was collected from volunteers under Institutional Review Board (IRB)-approved protocols. Drug user human hair was obtained from a local substance abuse treatment center under IRB-approved protocols. Table I. MS–MS method parameters Analytes  Retention time (min)  Precursor ion (m/z)  Product ion 1 (m/z)  Product ion 2 (m/z)  CE 1 (V)  CE 2 (V)  Codeine  2.97  300.2  165.0  152.1  56  52  Codeine-6B-D-glucuronide  2.87  476.2  300.2  215.1  32  44  Dihydrocodeine  2.95  302.2  199.0  128.1  36  70  Dihydrocodeine-6B-D-glucuronide  2.94  478.2  302.3  199.2  32  56  Dihydromorphine  1.29  288.2  185.1  157.0  36  52  Dihydromorphine-3B-D-glucuronide  0.80  464.2  288.2  185.1  36  64  Hydrocodone  3.51  300.2  199.2  128.0  32  68  Hydromorphone  1.81  286.2  185.1  157.1  36  52  Hydromorphone-3B-D-glucuronide  1.04  462.2  286.2  185.1  32  56  Morphine  1.29  286.2  152.1  165.2  44  70  Morphine-3B-D-glucuronide  0.81  462.2  286.2  –  36  –  Morphine-6B-D-glucuronide  1.25  462.2  286.3  –  36  –  Oxycodone  3.27  316.2  298.1  241.1  20  36  Oxymorphone  1.54  302.1  284.2  227.0  24  32  Oxymorphone-3B-D-glucuronide  0.81  478.2  284.2  227.0  36  56  6-Acetylmorphine  3.72  328.2  211.1  164.9  28  40  Codeine-d6  2.93  306.2  165.0  –  44  –  Codeine-6B-D-glucuronide-d3  2.85  479.2  61.1  –  40  –  Dihydrocodeine-d6  2.90  308.2  171.2  –  44  –  Hydrocodone-d6  3.47  306.2  202.3  –  32  –  Hydromorphone-d3  1.80  289.2  185.0  –  32  –  Morphine-d6  1.27  292.2  151.9  –  68  –  Morphine-3B-D-glucuronide-d3  0.81  465.2  289.2  –  36  –  Morphine-6B-D-glucuronide-d3  1.24  465.2  289.0  –  36  –  Oxycodone-d6  3.23  322.2  304.1  –  20  –  Oxymorphone-d3  1.53  305.2  287.1  –  24  –  Oxymorphone-3B-D-glucuronide-d3  0.79  481.2  287.1  –  32  –  6-Acetylmorphine-d6  3.70  334.2  164.9  –  44  –  Analytes  Retention time (min)  Precursor ion (m/z)  Product ion 1 (m/z)  Product ion 2 (m/z)  CE 1 (V)  CE 2 (V)  Codeine  2.97  300.2  165.0  152.1  56  52  Codeine-6B-D-glucuronide  2.87  476.2  300.2  215.1  32  44  Dihydrocodeine  2.95  302.2  199.0  128.1  36  70  Dihydrocodeine-6B-D-glucuronide  2.94  478.2  302.3  199.2  32  56  Dihydromorphine  1.29  288.2  185.1  157.0  36  52  Dihydromorphine-3B-D-glucuronide  0.80  464.2  288.2  185.1  36  64  Hydrocodone  3.51  300.2  199.2  128.0  32  68  Hydromorphone  1.81  286.2  185.1  157.1  36  52  Hydromorphone-3B-D-glucuronide  1.04  462.2  286.2  185.1  32  56  Morphine  1.29  286.2  152.1  165.2  44  70  Morphine-3B-D-glucuronide  0.81  462.2  286.2  –  36  –  Morphine-6B-D-glucuronide  1.25  462.2  286.3  –  36  –  Oxycodone  3.27  316.2  298.1  241.1  20  36  Oxymorphone  1.54  302.1  284.2  227.0  24  32  Oxymorphone-3B-D-glucuronide  0.81  478.2  284.2  227.0  36  56  6-Acetylmorphine  3.72  328.2  211.1  164.9  28  40  Codeine-d6  2.93  306.2  165.0  –  44  –  Codeine-6B-D-glucuronide-d3  2.85  479.2  61.1  –  40  –  Dihydrocodeine-d6  2.90  308.2  171.2  –  44  –  Hydrocodone-d6  3.47  306.2  202.3  –  32  –  Hydromorphone-d3  1.80  289.2  185.0  –  32  –  Morphine-d6  1.27  292.2  151.9  –  68  –  Morphine-3B-D-glucuronide-d3  0.81  465.2  289.2  –  36  –  Morphine-6B-D-glucuronide-d3  1.24  465.2  289.0  –  36  –  Oxycodone-d6  3.23  322.2  304.1  –  20  –  Oxymorphone-d3  1.53  305.2  287.1  –  24  –  Oxymorphone-3B-D-glucuronide-d3  0.79  481.2  287.1  –  32  –  6-Acetylmorphine-d6  3.70  334.2  164.9  –  44  –  Calibrator, quality control and internal standard preparation Working solutions were prepared in water by diluting methanolic stock solutions. Negative hair specimens were fortified with 50 μL of the appropriate calibrator working solutions to create seven-point curves for all analytes. The calibration curves of all glucuronide conjugates were constructed with concentrations of 2, 4, 10, 20, 40, 80 and 120 pg/mg, whereas those for all non-conjugated compounds were constructed with concentrations of 40, 80, 120, 300, 500, 800 and 1,200 pg/mg. Quality control (QC) working solutions were independently prepared in water by diluting methanolic stock solutions. Negative hair specimens were fortified with 50 μL of the appropriate QC working solution at low, medium, and high concentrations. The QC concentrations were 6, 30 and 100 pg/mg for all glucuronide conjugates and 120, 450 and 960 pg/mg for all non-conjugated compounds. Internal standard (ISTD) working solutions were prepared in water using methanolic stock solutions. For all calibrators, QCs, and user hair samples, 50 μL of ISTD working solution was added to achieve final concentrations of 30 pg/mg for all glucuronide conjugates and 200 pg/mg for all non-conjugated compounds. Hair wash procedure The wash procedure used was adapted from a method published by Cairns et al. (26). Hair samples were placed in 500-mL amber glass bottles with Teflon coated screw caps. A sufficient volume of dry isopropanol to cover the hair was added to each, and the bottles were shaken at 39°C for 15 min. The isopropanol was decanted, and sufficient phosphate buffer (0.01 M phosphate buffer with 0.01% BSA, pH 6) was added to cover the hair. The bottles were shaken at 39°C for 30 min, after which the phosphate buffer was decanted. The 30-min phosphate buffer wash was repeated two more times, followed by two 60-min phosphate buffer washes. Sample preparation Hair samples (25 ± 0.2 mg) were accurately weighed and cut into pieces <1 cm using scissors. The hair samples were then placed into conical glass tubes with screw caps, and then, 50 μL of ISTD working solution and 500 μL of M3 Reagent were added. The solutions were vortexed then centrifuged at 4,000 rpm for 5 min. The vials were placed in a 100°C heating block for 30 min, removed, gently mixed, then returned to the heating block for an additional 30 min. Subsequently, the samples were cooled to room temperature on the bench and centrifuged, and the supernatant was removed and placed in a 12 × 75 glass culture tube. Then, 500 μL of 0.1 M HCl was added to the supernatant, vortexed, and poured onto Waters Oasis MCX extraction cartridges (30 cc/60 mg) that had been pretreated with 2 mL of methanol followed by 2 mL of deionized water. The cartridges were then washed with 4 mL of 0.1 M HCl and dried for 5 min under nitrogen. Samples were eluted into silanized glass culture tubes using 2 mL of 5% ammonium hydroxide in methanol, evaporated at 40°C under nitrogen, and then reconstituted with 100 μL of the starting mobile phase composition (5 mM ammonium formate:methanol [95:5/v:v] with 0.1% formic acid). LC–MS-MS parameters Samples were analyzed on an Agilent 1290 LC coupled to an Agilent 6490 triple-quadrupole MS with an electrospray source (Santa Clara, CA). All analyses were conducted in multiple reaction monitoring mode using the MS/MS ion transitions and optimized collision energies listed in Table I. Two ion transitions were monitored for each target analyte, except the morphine glucuronides. For these two compounds, abundances of the qualifier ion transitions employed in assays of urine (43) and serum (44) or determined experimentally during the analytical optimization of the standards were unacceptably low, and reliable ion ratios could not be established. To increase the sensitivity for these analytes, the low-intensity qualifier transitions were not included in the final method. All analytes were acquired using the same dwell time (25 ms), fragmentation voltage (380 V), cell acceleration voltage (4 V) and positive ionization mode. Data were acquired using the following source parameters: gas temperature, 250°C; gas flow, 15 L/min; and capillary voltage, 3,000 V. Samples were injected (4 μL) onto an Agilent poroshell 120 SB C-18 column (2.7 μm, 2.1 × 100 mm) held at 50°C. A gradient elution was used, which consisted of 5-mM ammonium formate with 0.1% formic acid (mobile phase A) and methanol with 0.1% formic acid (mobile phase B) with a flow rate of 0.50 mL/min. The mobile phase composition was held at 5% B for 0.50 min, increased linearly to 75% B from 0.50 to 5.00 min, increased further to 90% from 5.10 to 6.10 min, and then decreased to 5% by 6.20 min, followed by a 2.00-min post run, for a total runtime of 8.20 min. An alternate LC method that achieves baseline separation between morphine and morphine-6-glucuronide was used for a portion of the interference study. That method was isocratic with a mobile phase consisting of acetonitrile:methanol:10-mM ammonium formate with 0.1% formic acid (2.5:2.5:95, v:v:v) at a flow rate of 0.300 mL/min. Data analysis was performed using Agilent MassHunter software. Chromatograms of the LC–MS-MS transitions for all analytes and internal standards are presented in Figures S1 through S3 of the Supplemental information. Validation Calibration model The calibration model was established by analyzing seven non-zero calibrators spanning approximately two orders of magnitude. For codeine, dihydrocodeine, dihydromorphine, hydrocodone, hydromorphone, morphine, oxycodone, oxymorphone and 6-acetylmorphine, the calibrator concentrations were 40, 80, 120, 300, 500, 800 and 1,200 pg/mg. For all glucuronide conjugates, the calibrator concentrations were 2, 4, 10, 20, 40, 80 and 120 pg/mg. The calibration curve was considered acceptable if all calibrators were within 20% of their target concentrations for all analytes at each level over five individual runs (n = 5). Additionally, the retention times and qualifier/quantifier ion ratios (where applicable) for individual calibrators were required to be within 2 and 20%, respectively, of the average values of those parameters for all calibrators and QCs over the five runs. Limit of detection, lower limit of quantification and carryover The limit of detection (LOD) was defined as the lowest concentration that produced an average instrument response greater than or equal to the average peak area of 10 blank matrix samples plus 3.3 times the standard deviation of the blank samples (45) and had a qualifier/quantifier ratio within 20% of the average ratio of the calibrators and controls. The LOD was determined by analyzing each drug in matrix at decreasing concentrations in duplicate over three runs in three different lots (n = 18) until the results no longer met the criteria described above. The lower limit of quantification (LLOQ) was administratively set as the concentration of the lowest calibrator, which was determined during the establishment of the calibration curve. Carryover was evaluated by analyzing blank matrix samples immediately after the highest calibrators in the calibration runs. Carryover was considered present if the peak area of the blank sample exceeded 20% of the peak area of the lowest calibrator (LLOQ). Accuracy and precision Accuracy was determined by comparing the calculated concentration to the target concentration at low, medium, and high analyte levels in triplicate over five individual runs (n = 45). For non-conjugated compounds, the target concentrations were 120, 450 and 960 pg/mg, whereas those for glucuronide conjugates were 6, 30 and 100 pg/mg. Dihydrocodeine-6B-D-glucuronide was the exception. Because the upper LOQ (ULOQ) was determined to be 80 pg/mg, accuracy and precision were calculated using the low and medium QC concentrations (6 and 30 pg/mg, respectively; n = 30). Within-run precision was calculated as the average of the percent covariance (% CV) in the accuracy for each concentration within a run, averaged across all five runs. Between-run precision was calculated as the % CV in the accuracy of the samples at a single concentration in all five runs (n = 15), averaged across the three concentrations. Acceptable accuracy and precision limits were considered to be ±20%. For samples with target analyte concentrations above 1,200 pg/mg, an alternate preparation technique utilizing 10 mg of hair (instead of 25 mg), comparable to a dilution integrity test for liquid samples, was included in the validation. Dilution samples (n = 15) were prepared at a target analyte concentration of 2,500 pg/mg. For these samples, 10 mg of hair was extracted and quantified against a calibration curve prepared from 25-mg extractions to verify that the accuracy and precision were within acceptable limits. Stability Processed sample stability was determined by extracting and analyzing low, medium, and high QC samples in triplicate to establish a baseline (t0) and then re-analyzing the same extracted samples after 24, 48 and 72 h of storage at room temperature. Processed samples were considered stable if the calculated concentrations were within 20% of the t0 concentrations. Interferences and specificity Blank matrix samples containing all target analytes, all ISTDs, and a single potentially interfering compound were prepared. These samples contained the single potentially interfering compound at 500 pg/mg, non-conjugated analytes at 80 pg/mg, glucuronide conjugates at 3 pg/mg, non-conjugated ISTDs at 200 pg/mg and glucuronide conjugate ISTDs at 30 pg/mg. The potentially interfering compounds evaluated were heroin, norcodeine, norhydrocodone, normorphine, noroxycodone and noroxymorphone (Figure 3). An additional set of samples, each of which contained only a single potentially interfering compound and all ISTDs at the same concentrations listed above, was also evaluated. Interference was considered present for any compound that produced a peak area >20% of the peak area at the LLOQ for any analyte or that caused any calculated analyte concentration to deviate by more than 20% from its target concentration when interferences were mixed with analytes. Figure 3. View largeDownload slide Structures of the potentially interfering compounds that were evaluated during method validation. Figure 3. View largeDownload slide Structures of the potentially interfering compounds that were evaluated during method validation. Five blank matrix samples fortified with ISTDs were analyzed to demonstrate the absence of interference originating from the ISTDs. Ten additional matrix samples without the addition of analyte or ISTD were also analyzed to demonstrate the absence of interference from the matrix. In both cases, interference with an analyte was identified if the average peak area of the blank sample was >20% of the established LLOQ peak area for that analyte. To investigate the possibility of glucuronide conjugate interference arising during the sample preparation and extraction processes, blank hair samples were fortified with codeine, dihydrocodeine, dihydromorphine, hydrocodone, hydromorphone, morphine, oxycodone, oxymorphone and 6-acetylmorphine at 500, 800, 1,200 and 3,000 pg/mg (n = 3 at each concentration) and all ISTDS. These samples were extracted and analyzed for the presence of the glucuronide conjugates listed in Table I. Matrix effects Matrix effects were evaluated according to the method described by Matuszewski et al. (46) using 10 lots of blank hair. Three sets of samples were fortified with each target analyte at 3 pg/mg for glucuronide conjugates and 80 pg/mg for all other compounds. Type C samples (i.e., pre-extraction spiked samples) were made by fortifying blank hair with target analytes and ISTDs prior to solid-phase extraction. Type B samples (i.e., post-extraction spiked samples) were made by reconstituting the dried eluent from the solid-phase extraction of blank hair with mobile phase fortified with target analytes and ISTDs at equivalent concentrations to type C samples. Type A samples contained the target analytes and ISTDs in mobile phase A, at equivalent concentrations to type B and C samples. Matrix effect, extraction recovery, and process efficiency were calculated as follows:   Matrixeffect(%)=B/A×100  Extractionrecovery(%)=C/B×100  Processefficiency(%)=C/A×100where A, B and C are the mean peak areas for the quantitative transitions of type A, B and C samples, respectively. The matrix effect defined above is referred to as the absolute matrix effect because it is a comparison of the response in matrix to the response in neat mobile phase. The relative matrix effect was also assessed by comparing the % CV in the response across the 10 matrix lots for type B samples to the % CV in the response for type A samples. Results and Discussion Validation The method presented here produced calibration data with reproducible linear correlations with weighting factors of 1/x for each analyte. Throughout the use of the method for validation and sample analysis, consistent peak shapes, ion ratios, ISTD responses and retention times were achieved. The validation results are summarized in Table II. The method is linear from 40 to 1,200 pg/mg for all non-conjugated compounds and from 2 to 120 pg/mg for all glucuronide conjugates, except dihydrocodeine-6B-D-glucuronide, which has an ULOQ of 80 pg/mg. LODs range from 0.2 pg/mg to 1 pg/mg for all analytes. Overall accuracy was within acceptable limits and ranged from 94 to 108%. Within-run precision was consistent across all analytes, with a narrow range of % CVs (3.4–6.7); morphine-6B-D-glucuronide showed the most variability. As expected, between-run precision was slightly more variable and ranged from 5.1 to 11.2%; again, morphine 6B-D-glucuronide exhibited the most variability. Reducing the sample amount from 25 to 10 mg for high-concentration samples did not affect the accuracy or precision by more than ±15%. No carryover was observed in blank samples analyzed immediately following high calibrators, and processed samples were stable in the autosampler, with <10% deviation from t0 observed in calculated concentrations up to 72 h post-extraction. Table II. Validation summary Analytes     Precision (n=45)  LOD (pg/mg)  LLOQ (pg/mg)  ULOQ (pg/mg)  r2  Overall average accuracy % (n = 45)  Average within run precision %CV  Average between run precision %CV  Codeine  1  40  1,200  0.9978  94.2  4.1  6.6  Codeine-6B-d-glucuronide  0.5  2  120  0.9961  99.2  4.7  6.7  Dihydrocodeine  0.5  40  1,200  0.9950  94.9  4.6  7.9  Dihydrocodeine-6B-d-glucuronide  1  2  80  0.9960  96.8  4.2  7.5  Dihydromorphine  0.5  40  1,200  0.9977  101.1  5.1  7.4  Hydrocodone  1  40  1,200  0.9980  93.2  3.4  5.9  Hydromorphone  0.5  40  1,200  0.9979  107.8  3.6  5.5  Hydromorphone-3B-d-glucuronide  0.2  2  120  0.9928  99.5  6.0  10.0  Morphine  0.2  40  1,200  0.9960  97.3  4.1  6.2  Morphine-3B-d-glucuronide  0.5  2  120  0.9969  98.8  3.7  5.5  Morphine-6B-d-glucuronide  0.5  2  120  0.9964  102.0  6.7  11.2  Oxycodone  0.5  40  1,200  0.9977  92.6  3.5  5.1  Oxymorphone  0.5  40  1,200  0.9985  95.4  4.4  6.2  Oxymorphone-3B-d-glucuronide  0.5  2  120  0.9968  98.3  5.4  8.1  6-Acetylmorphine  1  40  1,200  0.9978  103.0  4.4  5.2  Analytes     Precision (n=45)  LOD (pg/mg)  LLOQ (pg/mg)  ULOQ (pg/mg)  r2  Overall average accuracy % (n = 45)  Average within run precision %CV  Average between run precision %CV  Codeine  1  40  1,200  0.9978  94.2  4.1  6.6  Codeine-6B-d-glucuronide  0.5  2  120  0.9961  99.2  4.7  6.7  Dihydrocodeine  0.5  40  1,200  0.9950  94.9  4.6  7.9  Dihydrocodeine-6B-d-glucuronide  1  2  80  0.9960  96.8  4.2  7.5  Dihydromorphine  0.5  40  1,200  0.9977  101.1  5.1  7.4  Hydrocodone  1  40  1,200  0.9980  93.2  3.4  5.9  Hydromorphone  0.5  40  1,200  0.9979  107.8  3.6  5.5  Hydromorphone-3B-d-glucuronide  0.2  2  120  0.9928  99.5  6.0  10.0  Morphine  0.2  40  1,200  0.9960  97.3  4.1  6.2  Morphine-3B-d-glucuronide  0.5  2  120  0.9969  98.8  3.7  5.5  Morphine-6B-d-glucuronide  0.5  2  120  0.9964  102.0  6.7  11.2  Oxycodone  0.5  40  1,200  0.9977  92.6  3.5  5.1  Oxymorphone  0.5  40  1,200  0.9985  95.4  4.4  6.2  Oxymorphone-3B-d-glucuronide  0.5  2  120  0.9968  98.3  5.4  8.1  6-Acetylmorphine  1  40  1,200  0.9978  103.0  4.4  5.2  Interferences and specificity At 500 pg/mg, heroin was the only interferent that produced a signal >20% of the LLOQ response for any analyte or caused the quantification of any analyte to deviate by more than 20% from the target concentration when it was mixed with analytes. In both cases, 6-acetylmorphine was the only impacted analyte. An interferent concentration of 500 pg/mg was used because it is 2.5 times the suggested cutoff concentration for opioids in hair: 200 pg/mg. This is analogous to the National Laboratory Certification Program (NLCP) guidelines for opiates in urine, which require interference testing at 5,000 ng/mL (i.e., 2.5 times the cutoff concentration of 2,000 ng/mL). No glucuronide conjugate interference was detected in extracted blank hair samples from 10 non-drug users or from blank hair fortified with codeine, dihydrocodeine, dihydromorphine, hydrocodone, hydromorphone, oxycodone, oxymorphone or 6-acetylmorphine at concentrations up to 3,000 pg/mg. Additionally, no glucuronide conjugate interference was detected in blank hair fortified with morphine at 500, 800 or 1,200 pg/mg. At 3,000 pg/mg, a peak was observed that interfered with the morphine-6-glucuronide transition both in the presence and absence of ISTDs; its peak area was approximately equal to that of the LLOQ. No interference was observed when standard analytes were spiked into mobile phase and extracted, indicating that this interference is attributable to an interaction between morphine and hair during the extraction process. Morphine samples (3,000 pg/mg) were prepared in triplicate in two additional lots of blank hair. In both of these lots, the average peak area for the interference was approximately one-third of the peak area of the LLOQ. Care should be taken in interpreting morphine-6-glucuronide results for samples with morphine concentrations above 3,000 pg/mg using this assay. Only one out of the 46 user hair samples tested had a morphine concentration above 3,000 pg/mg, (estimated to be 3,113 pg/mg), and its morphine-6-glucuronide concentration was >7 × LLOQ. Therefore, the impact of this interference on the results presented here is expected to be minimal. Using an alternate LC method, it was possible to separate this interference from morphine-6-glucuronide, indicating it is a true interference and not attributable to the formation of morphine-6-glucuroinde in the samples during the extraction process. Although the impact of this interference is minimal, it can be overcome via additional optimization of the LC method. Additional studies are underway to determine the cause of this interference. Matrix effects The matrix effects ranged from 43.9 to 111.5% and are summarized in Table III. For most analytes, there was ion suppression, indicated by matrix effect values of <100%. The only exceptions were codeine-6B-D-glucuronide and dihydrocodeine-6B-D-glucuronide, which exhibited slight ion enhancement, as indicated by a matrix effect >100%. Although the results show significant ion suppression and enhancement, this is not unexpected in an LC–MS-MS assay. The method presented here also appears to have a relative matrix effect, as evidenced by the increased variability in the peak areas of analytes spiked into 10 different lots of hair extraction eluent (type B samples) compared with that in the responses of analytes spiked into mobile phase only (type A samples). However, the % CVs of the peak areas for type B samples are all <13%. The relative matrix effect and the ion suppression and enhancement are effectively controlled for using matrix-matched calibrators and controls and appropriate ISTDs, as evidenced by the accuracy of the results obtained for type C samples (i.e., pre-extraction spiked samples). Extraction recovery and process efficiency ranged from 56.6 to 68.1% and from 28.4 to 73.0%, respectively, for glucuronide conjugates, whereas for non-conjugated compounds, these values ranged from 64.9 to 75.1% and from 28.5 to 58.6%, respectively. The overlapping ranges indicate no significant differences in the extraction recovery or process efficiency for the two groups of analytes. Table III. Matrix effect results Analytes  Matrix effect (%)  Extraction recovery (%)  Process efficiency (%)  Accuracy C samples  Accuracy %CV C samples  Peak area % CV  A samples  B samples  C samples  Codeine  75.8  69.9  53.0  100.4  4.5  2.3  8.3  7.9  Codeine-6B-d-glucuronide  107.9  67.7  73.0  84.1  7.0  2.8  12.1  8.0  Dihydrocodeine  75.0  71.9  53.9  106.6  5.4  1.5  9.7  8.9  Dihydrocodeine-6B-d-glucuronide  111.5  56.6  63.1  103.2  9.7  1.5  3.7  21.8  Dihydromorphine  79.9  71.0  56.7  107.8  2.7  2.2  10.6  2.1  Hydrocodone  53.7  66.9  35.9  106.3  3.2  1.6  10.0  14.2  Hydromorphone  75.6  71.4  54.0  102.4  4.1  0.8  10.6  4.8  Hydromorphone-3B-d-glucuronide  46.3  61.5  28.4  124.8  6.7  1.9  9.3  4.9  Morphine  73.0  75.1  54.8  96.7  2.7  1.8  9.2  3.9  Morphine-3B-d-glucuronide  45.5  68.1  30.9  85.4  4.8  2.4  11.1  3.4  Morphine-6B-d-glucuronide  77.8  65.6  51.0  88.7  4.2  2.1  11.9  5.1  Oxycodone  79.3  72.1  57.1  96.4  3.4  1.4  8.8  4.8  Oxymorphone  80.6  72.7  58.6  105.8  0.2  1.3  9.8  2.7  Oxymorphone-3B-d-glucuronide  52.5  67.6  35.5  83.9  11.3  2.2  12.0  12.0  6-Acetylmorphine  43.9  64.9  28.5  115.9  5.0  1.1  9.6  16.4  Analytes  Matrix effect (%)  Extraction recovery (%)  Process efficiency (%)  Accuracy C samples  Accuracy %CV C samples  Peak area % CV  A samples  B samples  C samples  Codeine  75.8  69.9  53.0  100.4  4.5  2.3  8.3  7.9  Codeine-6B-d-glucuronide  107.9  67.7  73.0  84.1  7.0  2.8  12.1  8.0  Dihydrocodeine  75.0  71.9  53.9  106.6  5.4  1.5  9.7  8.9  Dihydrocodeine-6B-d-glucuronide  111.5  56.6  63.1  103.2  9.7  1.5  3.7  21.8  Dihydromorphine  79.9  71.0  56.7  107.8  2.7  2.2  10.6  2.1  Hydrocodone  53.7  66.9  35.9  106.3  3.2  1.6  10.0  14.2  Hydromorphone  75.6  71.4  54.0  102.4  4.1  0.8  10.6  4.8  Hydromorphone-3B-d-glucuronide  46.3  61.5  28.4  124.8  6.7  1.9  9.3  4.9  Morphine  73.0  75.1  54.8  96.7  2.7  1.8  9.2  3.9  Morphine-3B-d-glucuronide  45.5  68.1  30.9  85.4  4.8  2.4  11.1  3.4  Morphine-6B-d-glucuronide  77.8  65.6  51.0  88.7  4.2  2.1  11.9  5.1  Oxycodone  79.3  72.1  57.1  96.4  3.4  1.4  8.8  4.8  Oxymorphone  80.6  72.7  58.6  105.8  0.2  1.3  9.8  2.7  Oxymorphone-3B-d-glucuronide  52.5  67.6  35.5  83.9  11.3  2.2  12.0  12.0  6-Acetylmorphine  43.9  64.9  28.5  115.9  5.0  1.1  9.6  16.4  Analysis of drug user hair samples User hair samples were obtained from a local substance abuse treatment facility from October 2015 through September 2016. All samples received were washed at RTI International using the wash procedure described in the methods section. Aliquots of each were sent to Psychemedics Corporation (Acton, MA) for initial screening and confirmation of codeine, hydrocodone, hydromorphone, morphine, oxycodone, oxymorphone and 6-acetylmorphine. Forty-six samples were confirmed positive for at least one of these analytes and were subsequently extracted and analyzed at RTI using the validated method for opioids and their glucuronide conjugates described here. The results of the analyses performed at RTI are presented in Table IV. Table IV. RTI quantitative results for user hair samples Sample ID  6-AM  HYC  OXYC  COD  COD-6G  DHYC  DHYC-6G  HYM  HYM-3G  OXYM  OXYM-3G  DHYM  DHYM-3G  MOR  MOR-3G  MOR-6G  1  798  692  154  105  <2 (1.87)  –  –  52  –  –  –  –  –  752  2.78  4.96  2  >3000 (5199)  60  –  427  4.13  –  –  257  2.37  –  –  –  –  >3000 (3113)  6.89  14.96  3  –  407  –  –  –  –  –  –  –  –  –  –  –  –  –  –  4  1067  –  –  –  –  –  –  –  –  –  –  –  –  131  –  –  5  2496  49  299  44  –  –  –  –  –  89  <2 (1.08)  –  –  562  –  <2 (1.24)  6  –  77  1585  272  5.65  –  –  –  –  –  –  –  –  –  –  –  7  212  –  –  –  –  –  –  –  –  –  –  –  –  72  –  –  8  –  85  661  65  –  –  –  –  –  120  –  –  –  –  –  –  9  64  568  259  –  –  41  –  –  –  –  –  –  –  –  –  –  10  –  50  499  –  –  –  –  –  –  –  –  –  –  –  –  –  11  –  40  436  –  –  –  –  –  –  –  –  –  –  –  –  –  12  >3000 (4248)  45  –  285  4.22  –  –  137  <2 (1.79)  –  –  –  –  2832  9.41  13.86  13  –  –  975  –  –  –  –  843  4.97  2183  3.42  –  –  125  –  –  14  508  –  –  65  –  –  –  75  –  –  –  –  –  473  2.00  2.70  15  1370  398  299  118  7.22  –  –  64  –  –  –  –  –  457  2.26  7.01  16  53  –  –  –  –  –  –  –  –  –  –  –  –  118  <2 (1.22)  <2 (1.08)  17  –  106  –  –  –  –  –  –  –  –  –  –  –  –  –  –  18  –  364  966  –  –  –  –  –  –  78  –  –  –  –  –  –  19  194  –  –  –  –  –  –  –  –  –  –  –  –  89  –  <2 (1.20)  20  –  447  –  –  –  43  –  –  –  –  –  –  –  –  –  21  238  –  –  –  –  –  –  –  –  –  –  –  –  –  –  –  22  134  –  1490  –  –  –  –  72  –  1712  3.60  –  –  70  –  –  23  –  –  105  –  –  –  –  –  –  –  –  –  –  –  –  –  24  440  41  –  46  <2 (1.87)  –  –  –  –  –  –  –  –  402  3.90  5.57  25  419  –  –  –  –  –  –  –  –  –  –  –  –  223  <2 (1.05)  <2 (1.56)  26  314  78  480  –  –  –  –  –  –  –  –  –  –  206  –  –  27  448  –  2536  46  <2 (1.05)  –  –  –  –  1190  8.93  –  –  489  2.30  4.03  28  162  –  2349  –  –  –  –  –  –  –  –  –  –  70  –  –  29  2401  –  –  78  2.52  –  –  82  –  665  4.80  –  –  1083  3.62  8.90  30  –  982  2988  –  –  –  –  –  –  436  <2 (0.98)  –  –  200  –  –  31  –  187  237  –  –  –  –  –  –  –  –  –  –  –  –  32  2934  74  916  75  –  –  –  –  –  –  –  –  –  469  <2 (1.01)  <2 (1.20)  33  >3000 (6846)  –  –  280  5.37  –  –  209  –  –  –  –  –  2373  9.35  24.97  34  –  –  484  –  –  –  –  –  –  –  –  –  –  –  –  35  –  150  483  –  –  –  –  –  –  –  –  –  –  –  –  –  36  –  81  2932  –  –  –  –  –  –  134  2.86  –  –  –  –  –  37  429  –  69  76  <2 (1.89)  –  –  46  –  –  –  –  –  606  4.63  7.95  38  2158  92  –  269  2.86  –  –  83  –  425  3.25  –  –  1422  4.32  8.39  39  643  –  43  –  –  –  –  –  –  –  –  –  –  80  –  –  40  –  619  68  –  –  –  –  –  <2 (1.05)  –  –  –  –  –  –  –  41  191  –  160  –  –  –  –  –  –  –  –  –  –  44  <2 (0.92)  <2 (1.20)  42  206  –  -  –  –  –  –  –  –  –  –  –  –  –  –  –  43  –  –  1334  –  –  –  –  –  –  –  –  –  –  –  –  –  44  303  102  –  –  –  –  –  –  –  –  –  –  –  254  <2 (1.68)  2.29  45  >3000 (5132)  60  52  51  –  –  –  –  –  –  –  –  –  640  <2 (0.98)  <2 (1.31)  46  –  296  283  –  –  –  –  –  –  –  –  –  –  –  –  –  Sample ID  6-AM  HYC  OXYC  COD  COD-6G  DHYC  DHYC-6G  HYM  HYM-3G  OXYM  OXYM-3G  DHYM  DHYM-3G  MOR  MOR-3G  MOR-6G  1  798  692  154  105  <2 (1.87)  –  –  52  –  –  –  –  –  752  2.78  4.96  2  >3000 (5199)  60  –  427  4.13  –  –  257  2.37  –  –  –  –  >3000 (3113)  6.89  14.96  3  –  407  –  –  –  –  –  –  –  –  –  –  –  –  –  –  4  1067  –  –  –  –  –  –  –  –  –  –  –  –  131  –  –  5  2496  49  299  44  –  –  –  –  –  89  <2 (1.08)  –  –  562  –  <2 (1.24)  6  –  77  1585  272  5.65  –  –  –  –  –  –  –  –  –  –  –  7  212  –  –  –  –  –  –  –  –  –  –  –  –  72  –  –  8  –  85  661  65  –  –  –  –  –  120  –  –  –  –  –  –  9  64  568  259  –  –  41  –  –  –  –  –  –  –  –  –  –  10  –  50  499  –  –  –  –  –  –  –  –  –  –  –  –  –  11  –  40  436  –  –  –  –  –  –  –  –  –  –  –  –  –  12  >3000 (4248)  45  –  285  4.22  –  –  137  <2 (1.79)  –  –  –  –  2832  9.41  13.86  13  –  –  975  –  –  –  –  843  4.97  2183  3.42  –  –  125  –  –  14  508  –  –  65  –  –  –  75  –  –  –  –  –  473  2.00  2.70  15  1370  398  299  118  7.22  –  –  64  –  –  –  –  –  457  2.26  7.01  16  53  –  –  –  –  –  –  –  –  –  –  –  –  118  <2 (1.22)  <2 (1.08)  17  –  106  –  –  –  –  –  –  –  –  –  –  –  –  –  –  18  –  364  966  –  –  –  –  –  –  78  –  –  –  –  –  –  19  194  –  –  –  –  –  –  –  –  –  –  –  –  89  –  <2 (1.20)  20  –  447  –  –  –  43  –  –  –  –  –  –  –  –  –  21  238  –  –  –  –  –  –  –  –  –  –  –  –  –  –  –  22  134  –  1490  –  –  –  –  72  –  1712  3.60  –  –  70  –  –  23  –  –  105  –  –  –  –  –  –  –  –  –  –  –  –  –  24  440  41  –  46  <2 (1.87)  –  –  –  –  –  –  –  –  402  3.90  5.57  25  419  –  –  –  –  –  –  –  –  –  –  –  –  223  <2 (1.05)  <2 (1.56)  26  314  78  480  –  –  –  –  –  –  –  –  –  –  206  –  –  27  448  –  2536  46  <2 (1.05)  –  –  –  –  1190  8.93  –  –  489  2.30  4.03  28  162  –  2349  –  –  –  –  –  –  –  –  –  –  70  –  –  29  2401  –  –  78  2.52  –  –  82  –  665  4.80  –  –  1083  3.62  8.90  30  –  982  2988  –  –  –  –  –  –  436  <2 (0.98)  –  –  200  –  –  31  –  187  237  –  –  –  –  –  –  –  –  –  –  –  –  32  2934  74  916  75  –  –  –  –  –  –  –  –  –  469  <2 (1.01)  <2 (1.20)  33  >3000 (6846)  –  –  280  5.37  –  –  209  –  –  –  –  –  2373  9.35  24.97  34  –  –  484  –  –  –  –  –  –  –  –  –  –  –  –  35  –  150  483  –  –  –  –  –  –  –  –  –  –  –  –  –  36  –  81  2932  –  –  –  –  –  –  134  2.86  –  –  –  –  –  37  429  –  69  76  <2 (1.89)  –  –  46  –  –  –  –  –  606  4.63  7.95  38  2158  92  –  269  2.86  –  –  83  –  425  3.25  –  –  1422  4.32  8.39  39  643  –  43  –  –  –  –  –  –  –  –  –  –  80  –  –  40  –  619  68  –  –  –  –  –  <2 (1.05)  –  –  –  –  –  –  –  41  191  –  160  –  –  –  –  –  –  –  –  –  –  44  <2 (0.92)  <2 (1.20)  42  206  –  -  –  –  –  –  –  –  –  –  –  –  –  –  –  43  –  –  1334  –  –  –  –  –  –  –  –  –  –  –  –  –  44  303  102  –  –  –  –  –  –  –  –  –  –  –  254  <2 (1.68)  2.29  45  >3000 (5132)  60  52  51  –  –  –  –  –  –  –  –  –  640  <2 (0.98)  <2 (1.31)  46  –  296  283  –  –  –  –  –  –  –  –  –  –  –  –  –  Concentrations are in pg/mg. All calculated concentrations >0.9 pg/mg are included. Values >3000 or <2 pg/mg are out of the validated calibration range of the method. 6-Acetylmorphine (6-AM), hydrocodone (HYC), ixycodone (OXYC), codeine (COD), codeine-6-glucuronide (COD-6G), dihydrocodeine (DHYC), dihydrocodeine-6-glucuronide (DHYC-6G), hydromorphone (HYM), hydromorphone-3-glucuronide (HYM-3G), oxymorphone (OXYM), oxymorphone-3-glucuronide (OXYM-3G), dihydromorphine (DHYM), dihydromorphine-3-glucuronide (DHYM-3G), morphine (MOR), morphine-3-glucuronide (MOR-3G), morphine-6-glucuronide (MOR-6G). Codeine-6-glucuronide, hydromorphone-3-glucuronide, oxymorphone-3-glucuronide, morphine-3-glucuronide and morphine-6-glucuronide were reliably detected in samples containing the corresponding parent compound, and the concentrations of the glucuronide conjugates generally increased as the parent concentrations increased (Figure 4). Morphine-6-glucuronide was present at a higher concentration than morphine-3-glucuronide. Every sample that contained morphine 3-glucuronide also contained morphine-6-glucuronide, but two samples with morphine-6-glucuronide did not contain detectable levels of morphine-3-glucuronide. Two samples had relatively high concentrations of 6-acetylmorphine (1,067 and 643 pg/mg) but much lower concentrations of morphine (131 and 80 pg/mg) and no detectable morphine-3-glucuronide or morphine-6-glucuronide. Figure 4. View largeDownload slide Glucuronide concentrations as a function of the corresponding parent concentrations in user hair samples. Figure 4. View largeDownload slide Glucuronide concentrations as a function of the corresponding parent concentrations in user hair samples. Five out of five samples that had codeine concentrations above 200 pg/mg also had codeine-6-glucuronide concentrations above 2 pg/mg. Two additional samples had codeine-6-glucuronide concentrations above 2 pg/mg with codeine concentrations of only 118 and 78 pg/mg. Two out of three samples that had hydromorphone concentrations above 200 pg/mg also had hydromorphone-3-glucuronide concentrations above 2 pg/mg. Five out of six samples that had oxymorphone concentrations >200 pg/mg also had oxymorphone-3-glucuronide concentrations >2 pg/mg. One additional sample had an oxymorphone-3-glucuronide concentration above 2 pg/mg with an oxymorphone concentration of only 134 pg/mg. Twelve out of 16 samples that had morphine concentrations above 200 pg/mg also had morphine-6-glucuronide concentrations above 2 pg/mg, and 11 out of 16 had morphine-3-glucuroinde concentrations above 2 pg/mg. Hydromorphone-3-glucuronide present in the samples could have originated from hydromorphone or hydrocodone. Although the dataset is limited, it indicates that hydromorphone-3-glucuronide most likely originated from hydromorphone. The two samples with hydromorphone-3-glucuronide concentrations above the LLOQ had hydromorphone concentrations of 257 and 843 pg/mg. One of these samples contained no detectable hydrocodone, whereas the other had only 60 pg/mg hydrocodone. There are also several samples with high hydrocodone concentrations (up to 982 pg/mg) but no detectable hydromorphone-3-glucuronide. Similarly, oxymorphone-3-glucuronide could have originated from oxymorphone or oxycodone, but the dataset suggests that its origin was oxymorphone. All samples containing oxymorphone-3-glucuronide also contained oxymorphone, but two of these samples (samples 38 and 29, which had oxymorphone concentrations of 425 and 665 pg/mg, respectively) had no detectable oxycodone. There are numerous samples with high oxycodone concentrations (several over 1,000 pg/mg) but no detectable oxymorphone-3-glucuronide. The glucuronide conjugate concentrations relative to parent concentrations (in samples with both present) were as follows: Codeine-6-glucuronide ranged from 0.97 to 6.12%, with average and % CV values of 2.33 and 67.65%, respectively. Hydromorphone-3-glucuronide ranged from 0.59 to 1.30%, with average and % CV values of 0.94 and 31.03%, respectively. Oxymorphone-3-glucuronide ranged from 0.16 to 2.13%, with average and % CV values of 0.77 and 79.94%, respectively. Morphine-3-glucuronide ranged from 0.15 to 2.00%, with average and % CV values of 0.59 and 76.84%, respectively. Morphine-6-glucuronide ranged from 0.20 to 2.73%, with average and % CV values of 0.93 and 61.88%, respectively. For samples with calculated glucuronide conjugate concentrations between 0.9 and 2 pg/mg, the estimated concentrations denoted in parenthesis in Table IV were used in these calculations. Only two samples had quantifiable levels of dihydrocodeine: one at 41 pg/mg and one at 43 pg/mg. Dihydrocodeine-6-glucuronide was not detected in any of the samples. Dihydromorphine-3-glucuronide was not included in the validated assay because it interfered with morphine-3-glucuronide-d3 transitions. However, instrument parameters were optimized for this analyte, and its presence was monitored in all samples. None of the user hair samples had a dihydromorphine-3-glucuronide peak area that was greater than the average area plus 3.3 times the standard deviation of the peak area for this transition in blank samples. This result is not unexpected because none of the samples had dihydromorphine concentrations above the LLOQ (40 pg/mg). The results from the analyses performed by Psychemedics are presented in Table V. The results for analytes detected at both laboratories are compared in Table VI. Samples were sent to Psychemedics for analysis in multiple batches between October 2015 and December 2016, whereas the 46 opioid-positive samples were extracted and analyzed at RTI over 2 days in February 2017. The two data sets agree very well, despite the differences in the timing and location of the analyses. Thus, the methods used to detect parent compounds at both laboratories are robust, and the analytes are stable in hair stored under ambient conditions. The two discrepancies in the number of samples >200 pg/mg both correspond to samples very near this suggested cutoff. One sample was determined to contain 186 pg/mg hydrocodone by RTI and 238 pg/mg hydrocodone by Psychemedics. The other sample was determined to contain 206 pg/mg morphine at RTI and 176 pg/mg morphine by Psychemedics. The average and median concentration differences for all analytes with concentrations exceeding 40 pg/mg at both laboratories are less than 30 and 20%, respectively. Table V. Psychemedics quantitative results for user hair samples Sample ID  6-AM  HYC  OXYC  COD  HYM  OXYM  MOR  1  769  716  182  117  59  –  621  2  6010  59  –  566  273  –  3025  3  –  594  –  –  –  –  –  4  1565  –  –  –  –  –  177  5  2610  –  334  54  –  91  545  6  –  84  1740  446  –  –  –  7  242  –  –  –  –  –  102  8  –  72  497  48  –  117  –  9  102  838  412  –  –  –  64  10  –  –  462  –  –  –  –  11  –  45  379  –  –  –  –  12  4975  60  –  365  146  –  3370  13  26  –  1035  –  847  1765  113  14  701  –  68  66  74  –  515  15  1390  376  276  93  56  –  403  16  68  –  –  33  –  –  168  17  –  111  –  –  –  –  –  18  –  508  1390  –  –  109  –  19  183  –  –  –  –  –  98  20  –  496  –  –  –  –  –  21  271  –  –  –  –  –  51  22  160  –  1405  –  71  1670  87  23  –  –  121  –  –  –  –  24  420  47  –  52  36  –  432  25  405  –  –  –  –  –  214  26  340  69  329  –  –  –  176  27  392  –  2435  –  –  1180  437  28  185  –  2225  –  –  40  114  29  1415  –  –  66  69  475  864  30  –  1004  2655  –  –  382  152  31  –  238  206  –  27  –  –  32  2150  124  1036  71  –  –  493  33  6950  –  –  246  202  –  2380  34  –  –  375  –  –  –  –  35  –  162  482  –  –  –  –  36  –  83  3240  –  –  119  –  37  741  –  126  158  86  –  1145  38  2425  74  –  220  71  226  1032  39  513  –  31  –  –  –  66  40  –  536  62  –  –  –  –  41  180  –  136  –  –  –  56  42  303  –  –  –  –  –  –  43  –  –  942  –  –  –  –  44  258  –  –  –  –  –  398  45  3699  –  –  –  –  –  857  46  –  209  293  –  –  –  –  Sample ID  6-AM  HYC  OXYC  COD  HYM  OXYM  MOR  1  769  716  182  117  59  –  621  2  6010  59  –  566  273  –  3025  3  –  594  –  –  –  –  –  4  1565  –  –  –  –  –  177  5  2610  –  334  54  –  91  545  6  –  84  1740  446  –  –  –  7  242  –  –  –  –  –  102  8  –  72  497  48  –  117  –  9  102  838  412  –  –  –  64  10  –  –  462  –  –  –  –  11  –  45  379  –  –  –  –  12  4975  60  –  365  146  –  3370  13  26  –  1035  –  847  1765  113  14  701  –  68  66  74  –  515  15  1390  376  276  93  56  –  403  16  68  –  –  33  –  –  168  17  –  111  –  –  –  –  –  18  –  508  1390  –  –  109  –  19  183  –  –  –  –  –  98  20  –  496  –  –  –  –  –  21  271  –  –  –  –  –  51  22  160  –  1405  –  71  1670  87  23  –  –  121  –  –  –  –  24  420  47  –  52  36  –  432  25  405  –  –  –  –  –  214  26  340  69  329  –  –  –  176  27  392  –  2435  –  –  1180  437  28  185  –  2225  –  –  40  114  29  1415  –  –  66  69  475  864  30  –  1004  2655  –  –  382  152  31  –  238  206  –  27  –  –  32  2150  124  1036  71  –  –  493  33  6950  –  –  246  202  –  2380  34  –  –  375  –  –  –  –  35  –  162  482  –  –  –  –  36  –  83  3240  –  –  119  –  37  741  –  126  158  86  –  1145  38  2425  74  –  220  71  226  1032  39  513  –  31  –  –  –  66  40  –  536  62  –  –  –  –  41  180  –  136  –  –  –  56  42  303  –  –  –  –  –  –  43  –  –  942  –  –  –  –  44  258  –  –  –  –  –  398  45  3699  –  –  –  –  –  857  46  –  209  293  –  –  –  –  Concentrations are in pg/mg. LOD/LLOQ for all is 25 pg/mg, ULOL is 10,000 pg/mg for morphine, 15,000 pg/mg for all other analytes. Table VI. Comparison of quantitative results from RTI and Psychemedics   Number of samples, >40 pg/mg  Number of samples, >200 pg/mg  Average absolute concentration difference (%)  Median absolute concentration difference (%)  RTI  Psychemedics  Codeine  14  5  5  27  20  Hydrocodone  22  9  10  19  13  Hydromorphone  11  3  3  15  6  Morphine  27  17  16  23  18  Oxycodone  26  21  21  18  12  Oxymorphone  10  6  6  17  12  6-Acetylmorphine  28  22  22  21  15    Number of samples, >40 pg/mg  Number of samples, >200 pg/mg  Average absolute concentration difference (%)  Median absolute concentration difference (%)  RTI  Psychemedics  Codeine  14  5  5  27  20  Hydrocodone  22  9  10  19  13  Hydromorphone  11  3  3  15  6  Morphine  27  17  16  23  18  Oxycodone  26  21  21  18  12  Oxymorphone  10  6  6  17  12  6-Acetylmorphine  28  22  22  21  15  Conclusion This work demonstrates for the first time that codeine-6-glucuronide, hydromorphone-3-glucuronide, oxymorphone-3-glucuronide, morphine 3-glucuronide and morphine 6-glucuronide are present at sufficient concentrations to be quantifiable in hair from opioid users. Of the 46 samples analyzed in this study, only two contained dihydrocodeine at concentrations >40 pg/mg, and none had quantifiable concentrations of dihydromorphine. Further studies on hair positive for these analytes are needed to assess the detectability of dihydrocodeine glucuronide and dihydromorphine glucuronide in human hair. The LLOQ of this method for glucuronide conjugates is 2 pg/mg, but based on the determined concentrations of the analytes in user hair samples, a lower LLOQ (i.e., 1 pg/mg or lower) would be desirable. Using cutoffs of 200 pg/mg for parent and 1 pg/mg for glucuronide metabolites sixteen out of sixteen morphine positive samples were positive for morphine-6-glucuronide, five out of five codeine positive samples were positive for codeine-6-glucuronide, two out of three hydromorphone positive samples were positive for hydromorphone-3-glucuronide and six out of six oxymorphone positive samples were positive for oxymorphone-3-glucuronide. The overall process efficiency and extraction recovery of this method were less than ideal and could likely be improved with further refinement of the extraction method to achieve suitable sensitivity and LLOQs for glucuronide conjugates in human hair. Supplementary Data Supplementary data is available at Journal of Analytical Toxicology online. Acknowledgments The authors thank Dr Edward J. Cone for helpful discussions throughout the project and for reviewing the article. We also thank Amy Evans, Susan Crumpton, Frank Esposito and the United States Substance Abuse and Mental Health Services Administration (SAMHSA). Without the coordinated effort of all these people, this study would not have been possible. References 1 SAMHSA. ( 2004) Proposed revisions to mandatory guidelines for federal workplace drug testing programs. Federal Register , 69, 19673– 19732. 2 Henderson, G.L., Harkey, M.R., Zhou, C., Jones, R.T., Jacob, P., III ( 1996) Incorporation of isotopically labeled cocaine and metabolites into human hair: 1. Dose-response relationships. Journal of Analytical Toxicology , 20, 1– 12. Google Scholar CrossRef Search ADS PubMed  3 LeBeau, M.A., Montgomery, M.A., Brewer, J.D. ( 2011) The role of variations in growth rate and sample collection on interpreting results of segmental analyses of hair. Forensic Science International , 210, 110– 116. Google Scholar CrossRef Search ADS PubMed  4 Nakahara, Y. ( 1999) Hair analysis for abused and therapeutic drugs. Journal of Chromatography B , 733, 161– 180. Google Scholar CrossRef Search ADS   5 SOHT. ( 2004) Recommendations for hair testing in forensic cases. Forensic Science International , 145, 83– 84. CrossRef Search ADS PubMed  6 Cooper, G.A.A., Kronstrand, R., Kintz, P. ( 2012) Society of Hair Testing guidelines for drug testing in hair. Forensic Science International , 218, 20– 24. Google Scholar CrossRef Search ADS PubMed  7 Baumgartner, W.A., Hill, V.A., Blahd, W.H. ( 1989) Hair analysis for drugs of abuse. Journal of Forensic Sciences , 34, 1433– 1453. Google Scholar CrossRef Search ADS   8 Cairns, T., Hill, V., Schaffer, M., Thistle, W. ( 2004) Levels of cocaine and its metabolites in washed hair of demonstrated cocaine users and workplace subjects. Forensic Science International , 145, 175– 181. Google Scholar CrossRef Search ADS PubMed  9 Cairns, T., Hill, V., Schaffer, M., Thistle, W. ( 2004) Amphetamines in washed hair of demonstrated users and workplace subjects. Forensic Science International , 145, 137– 142. Google Scholar CrossRef Search ADS PubMed  10 Kidwell, D., Smith, F. Passive exposure, decontamination procedures, cutoffs, and bias. In: Kintz, P. (ed). Analytical and Practical Aspects of Drug Testing in Hair . Boca Raton, FL: CRC Press, 2006; pp. 25– 72. Chapter 2. Google Scholar CrossRef Search ADS   11 Kidwell, D.A., Smith, F.P., Shepherd, A.R. ( 2015) Ethnic hair care products may increase false positives in hair drug testing. Forensic Science International , 257, 160– 164. Google Scholar CrossRef Search ADS PubMed  12 Koren, G., Klein, J., Forman, R., Graham, K. ( 1992) Hair analysis of cocaine: differentiation between systemic exposure and external contamination. The Journal of Clinical Pharmacology , 32, 671– 675. Google Scholar CrossRef Search ADS PubMed  13 Morris-Kukoski, C.L., Montgomery, M.A., Hammer, R.L. ( 2014) Analysis of extensively washed hair from cocaine users and drug chemists to establish new reporting criteria. Journal of Analytical Toxicology , 38, 628– 636. Google Scholar CrossRef Search ADS PubMed  14 Ropero-Miller, J.D., Huestis, M.A., Stout, P.R. ( 2012) Cocaine analytes in human hair: evaluation of concentration ratios in different cocaine sources, drug-user populations and surface-contaminated specimens. Journal of Analytical Toxicology , 36, 390– 398. Google Scholar CrossRef Search ADS PubMed  15 Schaffer, M., Hill, V., Cairns, T. ( 2005) Hair analysis for cocaine: the requirement for effective wash procedures and effects of drug concentration and hair porosity in contamination and decontamination. Journal of Analytical Toxicology , 29, 319– 326. Google Scholar CrossRef Search ADS PubMed  16 Stout, P.R., Horn, C.K., Klette, K.L., Given, J. ( 2006) Occupational exposure to methamphetamine in workers preparing training aids for drug detection dogs. Journal of Analytical Toxicology , 30, 551– 553. Google Scholar CrossRef Search ADS PubMed  17 Welch, M.J., Sniegoski, L.T., Allgood, C.C., Habram, M. ( 1993) Hair analysis for drugs of abuse: evaluation of analytical methods, environmental issues, and development of reference materials. Journal of Analytical Toxicology , 17, 389– 398. Google Scholar CrossRef Search ADS PubMed  18 Cone, E.J. ( 1996) Mechanisms of drug incorporation into hair. Therapeutic Drug Monitoring , 18, 438– 443. Google Scholar CrossRef Search ADS PubMed  19 Henderson, G.L. ( 1993) Mechanisms of drug incorporation into hair. Forensic Science International , 63, 19– 29. Google Scholar CrossRef Search ADS PubMed  20 Porta, T., Grivet, C., Kraemer, T., Varesio, E., Hopfgartner, G. ( 2011) Single hair cocaine consumption monitoring by mass spectrometric imaging. Analytical Chemistry , 83, 4266– 4272. Google Scholar CrossRef Search ADS PubMed  21 Potsch, L., Skopp, G., Moeller, M.R. ( 1997) Biochemical approach on the conservation of drug molecules during hair fiber formation. Forensic Science International , 84, 25– 35. Google Scholar CrossRef Search ADS PubMed  22 Bassindale, T. ( 2012) Quantitative analysis of methamphetamine in hair of children removed from clandestine laboratories—evidence of passive exposure? Forensic Science International , 219, 179– 182. Google Scholar CrossRef Search ADS PubMed  23 Farst, K., Reading Meyer, J.A., Mac Bird, T., James, L., Robbins, J.M. ( 2011) Hair drug testing of children suspected of exposure to the manufacture of methamphetamine. Journal of Forensic and Legal Medicine , 18, 110– 114. Google Scholar CrossRef Search ADS PubMed  24 Martyny, J.W., Arbuckle, S.L., McCammon, C.S., Jr, Erb, N., Van Dyke, M. ( 2008) Methamphetamine contamination on environmental surfaces caused by simulated smoking of methamphetamine. Journal of Chemical Health and Safety , 15, 25– 31. Google Scholar CrossRef Search ADS   25 Baumgartner, W., Hill, V. Hair analysis for drugs of abuse: decontamination issues. In: Sunshine, I. (ed). Recent Developments in Therapeutic Drug Monitoring and Clinical Toxicology . Marcel Dekker: New York, 1992; pp. 577– 597. 26 Cairns, T., Hill, V., Schaffer, M., Thistle, W. ( 2004) Removing and identifying drug contamination in the analysis of human hair. Forensic Science International , 145, 97– 108. Google Scholar CrossRef Search ADS PubMed  27 Hill, V., Loni, E., Cairns, T., Sommer, J., Schaffer, M. ( 2014) Identification and analysis of damaged or porous hair. Drug Testing and Analysis , 6, 42– 54. Google Scholar CrossRef Search ADS PubMed  28 Schaffer, M., Cheng, C.-C., Chao, O., Hill, V., Matsui, P. ( 2016) Analysis of cocaine and metabolites in hair: validation and application of measurement of hydroxycocaine metabolites as evidence of cocaine ingestion. Analytical and Bioanalytical Chemistry , 408, 2043– 2054. Google Scholar CrossRef Search ADS PubMed  29 Montgomery, M., LeBeau, M., Morris-Kukoski, C. ( 2016) New hair testing conclusions. Journal of Analytical Toxicology , 41, 161– 162. 30 White, R.M. ( 2017) Drugs in hair. Part I. Metabolisms of major drug classes. Forensic Science Review , 29, 23– 55. Google Scholar PubMed  31 Smith, H.S. ( 2009) Opioid metabolism. Mayo Clinic Proceedings , 84, 613– 624. Google Scholar CrossRef Search ADS PubMed  32 Burchell, B., Brierley, C.H., Rance, D. ( 1995) Specificity of human UDP-glucuronosyltransferases and xenobiotic glucuronidation. Life sciences , 57, 1819– 1831. Google Scholar CrossRef Search ADS PubMed  33 Wang, X., Johansen, S.S., Zhang, Y., Jia, J., Rao, Y., Jiang, F., et al.  . ( 2017) Deposition of diazepam and its metabolites in hair following a single dose of diazepam. International Journal of Legal Medecine , 131, 131– 141. Google Scholar CrossRef Search ADS   34 Pichini, S., Marchei, E., Martello, S., Gottardi, M., Pellegrini, M., Svaizer, F., et al.  . ( 2015) Identification and quantification of 11-nor-delta9-tetrahydrocannabinol-9-carboxylic acid glucuronide (THC-COOH-glu) in hair by ultra-performance liquid chromatography tandem mass spectrometry as a potential hair biomarker of cannabis use. Forensic Science International , 249, 47– 51. Google Scholar CrossRef Search ADS PubMed  35 Beasley, E., Francese, S., Bassindale, T. ( 2016) Detection and mapping of cannabinoids in single hair samples through rapid derivatization and matrix-assisted laser desorption ionization mass spectrometry. Analytical Chemistry , 88, 10328– 10334. Google Scholar CrossRef Search ADS PubMed  36 Kim, H.S., Cheong, J.C., Lee, J.I., In, M.K. ( 2013) Rapid and sensitive determination of propofol glucuronide in hair by liquid chromatography and tandem mass spectrometry. Journal of Pharmaceutical and Biomedical Analysis , 85, 33– 39. Google Scholar CrossRef Search ADS PubMed  37 Wang, X., Linnet, K., Johansen, S.S. ( 2016) Development of a UPLC-MS/MS method for determining gamma-hydroxybutyric acid (GHB) and GHB glucuronide concentrations in hair and application to forensic cases. Forensic Toxicology , 34, 51– 60. Google Scholar CrossRef Search ADS   38 Kim, J., In, S., Park, Y., Park, M., Kim, E., Lee, S. ( 2013) Quantitative analysis of propofol-glucuronide in hair as a marker for propofol abuse. Analytical and Bioanalytical Chemistry , 405, 6807– 6814. Google Scholar CrossRef Search ADS PubMed  39 Kwak, J.H., Kim, H.K., Choe, S., In, S., Pyo, J.S. ( 2016) Determination of propofol glucuronide from hair sample by using mixed mode anion exchange cartridge and liquid chromatography tandem mass spectrometry. Journal of Chromatography B , 1015–1016, 209– 213. Google Scholar CrossRef Search ADS   40 Gygi, S.P., Joseph, R.E., Jr, Cone, E.J., Wilkins, D.G., Rollins, D.E. ( 1996) Incorporation of codeine and metabolites into hair. Role of pigmentation. Drug Metabolism and Disposition , 24, 495– 501. Google Scholar PubMed  41 Gygi, S.P., Colón, F., Raftogianis, R.B., Galinsky, R.E., Wilkins, D.G., Rollins, D.E. ( 1996) Dose-related distribution of codeine and its metabolites into rat hair. Drug Metabolism and Disposition , 24, 282– 287. Google Scholar PubMed  42 Toyooka, T., Yano, M., Kato, M., Nakahara, Y. ( 2001) Simultaneous determination of morphine and its glucuronides in rat hair and rat plasma by reversed-phase liquid chromatography with electrospray ionization mass spectrometry. Analyst , 126, 1339– 1345. Google Scholar CrossRef Search ADS PubMed  43 Waters Corporation. ( 2015) A simplified, mixed-mode sample preparation strategy for urinary forensic toxicology screening by UPLC-MS/MS [updated 2015] http://www.waters.com/webassets/cms/library/docs/720005290en.pdf. 44 Lott, S., Musshoff, F., Madea, B. ( 2013) LC/MS/MS method of 6-MAM, morphine, morphine-3-glucuro-nide (M3G) and morphine-6-glucuronide (M6G) for quantitative analysis in serum. Toxichem Krimtech , 80, 363– 366. 45 SWGTOX. ( 2013) Scientific Working Group for Forensic Toxicology (SWGTOX) standard practices for method validation in forensic toxicology [updated 2013]. http://www.swgtox.org/documents/Validation3.pdf. 46 Matuszewski, B.K., Constanzer, M.L., Chavez-Eng, C.M. ( 2003) Strategies for the assessment of matrix effect in quantitative bioanalytical methods based on HPLC-MS/MS. Analytical Chemistry , 75, 3019– 3030. Google Scholar CrossRef Search ADS PubMed  © The Author 2017. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Analytical Toxicology Oxford University Press

Detection and quantification of codeine-6-glucuronide, hydromorphone-3-glucuronide, oxymorphone-3-glucuronide, morphine 3-glucuronide and morphine-6-glucuronide in human hair from opioid users by LC–MS-MS

Loading next page...
 
/lp/ou_press/detection-and-quantification-of-codeine-6-glucuronide-hydromorphone-3-0bJJjZT0Ky
Publisher
Oxford University Press
Copyright
© The Author 2017. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com
ISSN
0146-4760
eISSN
1945-2403
D.O.I.
10.1093/jat/bkx086
Publisher site
See Article on Publisher Site

Abstract

Abstract Current hair testing methods that rely solely on quantification of parent drug compounds are unable to definitively distinguish between drug use and external contamination. One possible solution to this problem is to confirm the presence of unique drug metabolites that cannot be present through contamination, such as phase II glucuronide conjugates. This work demonstrates for the first time that codeine-6-glucuronide, hydromorphone-3-glucuronide, oxymorphone-3-glucuronide, morphine-3-glucuronide and morphine-6-glucuronide are present at sufficient concentrations to be quantifiable in hair of opioid users and that their concentrations generally increase as the concentrations of the corresponding parent compounds increase. Here, we present a validated liquid chromatography tandem mass spectrometry method to quantify codeine-6-glucuronide, dihydrocodeine-6-glucuronide, hydromorphone-3-glucuronide, morphine-3-glucuronide, morphine-6-glucuronide, oxymorphone-3-glucuronide, codeine, dihydrocodeine, dihydromorphine, hydrocodone, hydromorphone, morphine, oxycodone, oxymorphone and 6-acetylmorphine in human hair. The method was used to analyze 46 human hair samples from known drug users that were confirmed positive for opioids by an independent laboratory. Glucuronide concentrations in samples positive for parent analytes ranged from ~1 to 25 pg/mg, and most samples had glucuronide concentrations in the range of ~1 to 5 pg/mg. Relative to the parent concentrations, the average concentrations of the four detected glucuronides were as follows: codeine-6-glucuronide, 2.33%; hydromorphone-3-glucuronide, 0.94%; oxymorphone-3-glucuronide, 0.77%; morphine 3-glucuronide, 0.59%; and morphine-6-glucuronide, 0.93%. Introduction In 2004, the Substance Abuse and Mental Health Services Administration (SAMHSA) proposed a revision to the mandatory guidelines for federal workplace drug testing programs to expand the types of specimens that may be tested to include head hair, sweat and oral fluid, in addition to urine (1). Hair is seen as an advantageous matrix for drug testing for a number of reasons including, but not limited to, its long detection window, difficulty to adulterate, and ease of sample collection, transportation and storage. The proposed guidelines allow for testing of up to 1.5 inches of head hair closest to the scalp, which represents a period of ~90 days (2–4). For opiates, the currently proposed confirmatory compounds are codeine, morphine and 6-acetylmorphine, with cutoffs at 200 pg/mg for each (1). These analytes and cutoffs parallel guidelines set forth by the Society of Hair Testing (SOHT) (5, 6). Although hair testing is currently used for a variety of applications, one issue that remains under significant debate regarding its use in criminal investigations and federal workplace drug testing is the ability to definitively distinguish drug use from external contamination (7–17). Drugs may become incorporated into hair through multiple physiological processes (4, 18–21) or through passive environmental exposure (13, 18, 22–24). One approach to address the issue of environmental contamination, pioneered by Baumgartner and Hill, is to incorporate an extensive wash into the sample preparation procedure to remove surface contamination, followed by analysis of the final wash solution and application of a wash criterion (15, 25, 26). This procedure, when combined with determination of suitable metabolite ratios, has been used to identify samples that are contaminated rather than positive because of drug use (8, 15, 27). Another approach to distinguish use from contamination is to develop methods to detect unique metabolites that are definitive indicators of consumption instead of, or in addition to, the parent drugs. Schafer et al. (28) recently explored this approach and developed an assay to detect hydroxycocaine metabolites as evidence of cocaine ingestion rather than relying on the presence of cocaethylene and norcocaine. In March 2017, the Federal Bureau of Investigation (FBI) published a new set of criteria for reporting cocaine use that depends on analysis of wash solutions and the presence of hydroxycocaine metabolites (29). Despite these new reporting criteria for cocaine, the FBI Laboratory “continues to encourage research into the identification of drug metabolites or other markers that are uniquely associated with drug consumption” (29). In this study we focus on detection of metabolites in place of or in addition to parent compounds of opioid drugs. The choice of metabolite is critical (30) because some metabolites can be present as process impurities, meaning they have a distinct probability of being present as a result of contamination by the parent compounds. Other metabolites may be degradation products formed because of exposure to hair care products subsequent to parent drug contamination. Also, some metabolites are commercially available drugs themselves. Opioids in particular are problematic in this regard as many opioid metabolic products are active and marketed as standalone drugs. For example, hydromorphone is a metabolite of both morphine and hydrocodone, which are themselves metabolites of codeine (31). All of these compounds are separately marketed prescription drugs. Phase II conjugated metabolites are ideal markers of use because they are not products of common degradation pathways, as is the case for many phase I metabolites, and are not commercially available drugs. The most common phase II metabolic transformation is glucuronidation (conjugation with glucuronic acid) to form a glucuronide conjugate (32). Aside from ethyl glucuronide, very few publications have addressed the detection of glucuronides in human hair. Wang et al. (33) developed a method for the detection of oxazepam glucuronide and temazepam glucuronide in human hair but did not detect either of these metabolites in hair from study volunteers who consumed a single 10-mg dose of diazepam. Pichini et al. (34) used liquid chromatography tandem mass spectrometry (LC–MS-MS) to detect 11-nor-delta-9-tetrahydrocannabinol-9-carboxylic acid glucuronide (THC-COOH-glu) in 20 user hair samples at concentrations ranging from 0.5 to 8.6 pg/mg hair. Beasley et al. (35) attempted to detect THC-COOH-glu in human hair using matrix-assisted laser desorption/ionization (MALDI)-MS, but during analysis of hair spiked with THC-COOH-glu, the glucuronide fragmented to form THC-COOH, resulting in low sensitivity, and the method was not applied to user hair. Kim et al. (36) used LC–MS-MS to detect propofol glucuronide in hair from five users. Wang et al.(37) used LC–MS-MS to quantify gamma-hydroxybutyric acid (GHB) glucuronide in human hair and determined that GHB glucuronide did not accumulate appreciably in the hair of two GHB abusers. Several glucuronide conjugates have been detected in rat hair, including propofol (38, 39), morphine (40–42) and codeine (41). Here, we present a method for the extraction and quantification of glucuronide conjugates of opioids in human head hair and apply the method to 46 opioid-positive user hair samples. Structures of the opioid analytes and glucuronide conjugates included in the method are presented in Figures 1 and 2, respectively. Figure 1. View largeDownload slide Structures of the opioid analytes included in the validated method. Figure 1. View largeDownload slide Structures of the opioid analytes included in the validated method. Figure 2. View largeDownload slide Structures of the glucuronide conjugates included in the validated method. Figure 2. View largeDownload slide Structures of the glucuronide conjugates included in the validated method. Methods Materials High-performance LC (HPLC)-grade methanol and water were purchased from Fisher Scientific (Fair Lawn, NJ). Hydrochloric acid (HCl), isopropanol, potassium phosphate monobasic, potassium phosphate dibasic and bovine serum albumin (BSA) were purchased from Sigma Aldrich (St. Louis, MO), and ammonium hydroxide was acquired from Fisher Scientific. Ammonium formate and formic acid were purchased from Alfa Aesar (Ward Hill, MA) and Electron Microscopy Sciences (Hatfield, PA), respectively. M3 Reagent was purchased from Comedical (Trento, Italy). Deionized water was obtained from an in-house system. Solid-phase extraction columns (Oasis MCX) were purchased from Waters (Milford, MA). Reference standards and internal standards listed in Table I, as well as heroin, norcodeine, norhydrocodone, normorphine, noroxycodone and noroxymorphone were purchased from Cerilliant (Round Rock, TX). Drug-free human hair was collected from volunteers under Institutional Review Board (IRB)-approved protocols. Drug user human hair was obtained from a local substance abuse treatment center under IRB-approved protocols. Table I. MS–MS method parameters Analytes  Retention time (min)  Precursor ion (m/z)  Product ion 1 (m/z)  Product ion 2 (m/z)  CE 1 (V)  CE 2 (V)  Codeine  2.97  300.2  165.0  152.1  56  52  Codeine-6B-D-glucuronide  2.87  476.2  300.2  215.1  32  44  Dihydrocodeine  2.95  302.2  199.0  128.1  36  70  Dihydrocodeine-6B-D-glucuronide  2.94  478.2  302.3  199.2  32  56  Dihydromorphine  1.29  288.2  185.1  157.0  36  52  Dihydromorphine-3B-D-glucuronide  0.80  464.2  288.2  185.1  36  64  Hydrocodone  3.51  300.2  199.2  128.0  32  68  Hydromorphone  1.81  286.2  185.1  157.1  36  52  Hydromorphone-3B-D-glucuronide  1.04  462.2  286.2  185.1  32  56  Morphine  1.29  286.2  152.1  165.2  44  70  Morphine-3B-D-glucuronide  0.81  462.2  286.2  –  36  –  Morphine-6B-D-glucuronide  1.25  462.2  286.3  –  36  –  Oxycodone  3.27  316.2  298.1  241.1  20  36  Oxymorphone  1.54  302.1  284.2  227.0  24  32  Oxymorphone-3B-D-glucuronide  0.81  478.2  284.2  227.0  36  56  6-Acetylmorphine  3.72  328.2  211.1  164.9  28  40  Codeine-d6  2.93  306.2  165.0  –  44  –  Codeine-6B-D-glucuronide-d3  2.85  479.2  61.1  –  40  –  Dihydrocodeine-d6  2.90  308.2  171.2  –  44  –  Hydrocodone-d6  3.47  306.2  202.3  –  32  –  Hydromorphone-d3  1.80  289.2  185.0  –  32  –  Morphine-d6  1.27  292.2  151.9  –  68  –  Morphine-3B-D-glucuronide-d3  0.81  465.2  289.2  –  36  –  Morphine-6B-D-glucuronide-d3  1.24  465.2  289.0  –  36  –  Oxycodone-d6  3.23  322.2  304.1  –  20  –  Oxymorphone-d3  1.53  305.2  287.1  –  24  –  Oxymorphone-3B-D-glucuronide-d3  0.79  481.2  287.1  –  32  –  6-Acetylmorphine-d6  3.70  334.2  164.9  –  44  –  Analytes  Retention time (min)  Precursor ion (m/z)  Product ion 1 (m/z)  Product ion 2 (m/z)  CE 1 (V)  CE 2 (V)  Codeine  2.97  300.2  165.0  152.1  56  52  Codeine-6B-D-glucuronide  2.87  476.2  300.2  215.1  32  44  Dihydrocodeine  2.95  302.2  199.0  128.1  36  70  Dihydrocodeine-6B-D-glucuronide  2.94  478.2  302.3  199.2  32  56  Dihydromorphine  1.29  288.2  185.1  157.0  36  52  Dihydromorphine-3B-D-glucuronide  0.80  464.2  288.2  185.1  36  64  Hydrocodone  3.51  300.2  199.2  128.0  32  68  Hydromorphone  1.81  286.2  185.1  157.1  36  52  Hydromorphone-3B-D-glucuronide  1.04  462.2  286.2  185.1  32  56  Morphine  1.29  286.2  152.1  165.2  44  70  Morphine-3B-D-glucuronide  0.81  462.2  286.2  –  36  –  Morphine-6B-D-glucuronide  1.25  462.2  286.3  –  36  –  Oxycodone  3.27  316.2  298.1  241.1  20  36  Oxymorphone  1.54  302.1  284.2  227.0  24  32  Oxymorphone-3B-D-glucuronide  0.81  478.2  284.2  227.0  36  56  6-Acetylmorphine  3.72  328.2  211.1  164.9  28  40  Codeine-d6  2.93  306.2  165.0  –  44  –  Codeine-6B-D-glucuronide-d3  2.85  479.2  61.1  –  40  –  Dihydrocodeine-d6  2.90  308.2  171.2  –  44  –  Hydrocodone-d6  3.47  306.2  202.3  –  32  –  Hydromorphone-d3  1.80  289.2  185.0  –  32  –  Morphine-d6  1.27  292.2  151.9  –  68  –  Morphine-3B-D-glucuronide-d3  0.81  465.2  289.2  –  36  –  Morphine-6B-D-glucuronide-d3  1.24  465.2  289.0  –  36  –  Oxycodone-d6  3.23  322.2  304.1  –  20  –  Oxymorphone-d3  1.53  305.2  287.1  –  24  –  Oxymorphone-3B-D-glucuronide-d3  0.79  481.2  287.1  –  32  –  6-Acetylmorphine-d6  3.70  334.2  164.9  –  44  –  Calibrator, quality control and internal standard preparation Working solutions were prepared in water by diluting methanolic stock solutions. Negative hair specimens were fortified with 50 μL of the appropriate calibrator working solutions to create seven-point curves for all analytes. The calibration curves of all glucuronide conjugates were constructed with concentrations of 2, 4, 10, 20, 40, 80 and 120 pg/mg, whereas those for all non-conjugated compounds were constructed with concentrations of 40, 80, 120, 300, 500, 800 and 1,200 pg/mg. Quality control (QC) working solutions were independently prepared in water by diluting methanolic stock solutions. Negative hair specimens were fortified with 50 μL of the appropriate QC working solution at low, medium, and high concentrations. The QC concentrations were 6, 30 and 100 pg/mg for all glucuronide conjugates and 120, 450 and 960 pg/mg for all non-conjugated compounds. Internal standard (ISTD) working solutions were prepared in water using methanolic stock solutions. For all calibrators, QCs, and user hair samples, 50 μL of ISTD working solution was added to achieve final concentrations of 30 pg/mg for all glucuronide conjugates and 200 pg/mg for all non-conjugated compounds. Hair wash procedure The wash procedure used was adapted from a method published by Cairns et al. (26). Hair samples were placed in 500-mL amber glass bottles with Teflon coated screw caps. A sufficient volume of dry isopropanol to cover the hair was added to each, and the bottles were shaken at 39°C for 15 min. The isopropanol was decanted, and sufficient phosphate buffer (0.01 M phosphate buffer with 0.01% BSA, pH 6) was added to cover the hair. The bottles were shaken at 39°C for 30 min, after which the phosphate buffer was decanted. The 30-min phosphate buffer wash was repeated two more times, followed by two 60-min phosphate buffer washes. Sample preparation Hair samples (25 ± 0.2 mg) were accurately weighed and cut into pieces <1 cm using scissors. The hair samples were then placed into conical glass tubes with screw caps, and then, 50 μL of ISTD working solution and 500 μL of M3 Reagent were added. The solutions were vortexed then centrifuged at 4,000 rpm for 5 min. The vials were placed in a 100°C heating block for 30 min, removed, gently mixed, then returned to the heating block for an additional 30 min. Subsequently, the samples were cooled to room temperature on the bench and centrifuged, and the supernatant was removed and placed in a 12 × 75 glass culture tube. Then, 500 μL of 0.1 M HCl was added to the supernatant, vortexed, and poured onto Waters Oasis MCX extraction cartridges (30 cc/60 mg) that had been pretreated with 2 mL of methanol followed by 2 mL of deionized water. The cartridges were then washed with 4 mL of 0.1 M HCl and dried for 5 min under nitrogen. Samples were eluted into silanized glass culture tubes using 2 mL of 5% ammonium hydroxide in methanol, evaporated at 40°C under nitrogen, and then reconstituted with 100 μL of the starting mobile phase composition (5 mM ammonium formate:methanol [95:5/v:v] with 0.1% formic acid). LC–MS-MS parameters Samples were analyzed on an Agilent 1290 LC coupled to an Agilent 6490 triple-quadrupole MS with an electrospray source (Santa Clara, CA). All analyses were conducted in multiple reaction monitoring mode using the MS/MS ion transitions and optimized collision energies listed in Table I. Two ion transitions were monitored for each target analyte, except the morphine glucuronides. For these two compounds, abundances of the qualifier ion transitions employed in assays of urine (43) and serum (44) or determined experimentally during the analytical optimization of the standards were unacceptably low, and reliable ion ratios could not be established. To increase the sensitivity for these analytes, the low-intensity qualifier transitions were not included in the final method. All analytes were acquired using the same dwell time (25 ms), fragmentation voltage (380 V), cell acceleration voltage (4 V) and positive ionization mode. Data were acquired using the following source parameters: gas temperature, 250°C; gas flow, 15 L/min; and capillary voltage, 3,000 V. Samples were injected (4 μL) onto an Agilent poroshell 120 SB C-18 column (2.7 μm, 2.1 × 100 mm) held at 50°C. A gradient elution was used, which consisted of 5-mM ammonium formate with 0.1% formic acid (mobile phase A) and methanol with 0.1% formic acid (mobile phase B) with a flow rate of 0.50 mL/min. The mobile phase composition was held at 5% B for 0.50 min, increased linearly to 75% B from 0.50 to 5.00 min, increased further to 90% from 5.10 to 6.10 min, and then decreased to 5% by 6.20 min, followed by a 2.00-min post run, for a total runtime of 8.20 min. An alternate LC method that achieves baseline separation between morphine and morphine-6-glucuronide was used for a portion of the interference study. That method was isocratic with a mobile phase consisting of acetonitrile:methanol:10-mM ammonium formate with 0.1% formic acid (2.5:2.5:95, v:v:v) at a flow rate of 0.300 mL/min. Data analysis was performed using Agilent MassHunter software. Chromatograms of the LC–MS-MS transitions for all analytes and internal standards are presented in Figures S1 through S3 of the Supplemental information. Validation Calibration model The calibration model was established by analyzing seven non-zero calibrators spanning approximately two orders of magnitude. For codeine, dihydrocodeine, dihydromorphine, hydrocodone, hydromorphone, morphine, oxycodone, oxymorphone and 6-acetylmorphine, the calibrator concentrations were 40, 80, 120, 300, 500, 800 and 1,200 pg/mg. For all glucuronide conjugates, the calibrator concentrations were 2, 4, 10, 20, 40, 80 and 120 pg/mg. The calibration curve was considered acceptable if all calibrators were within 20% of their target concentrations for all analytes at each level over five individual runs (n = 5). Additionally, the retention times and qualifier/quantifier ion ratios (where applicable) for individual calibrators were required to be within 2 and 20%, respectively, of the average values of those parameters for all calibrators and QCs over the five runs. Limit of detection, lower limit of quantification and carryover The limit of detection (LOD) was defined as the lowest concentration that produced an average instrument response greater than or equal to the average peak area of 10 blank matrix samples plus 3.3 times the standard deviation of the blank samples (45) and had a qualifier/quantifier ratio within 20% of the average ratio of the calibrators and controls. The LOD was determined by analyzing each drug in matrix at decreasing concentrations in duplicate over three runs in three different lots (n = 18) until the results no longer met the criteria described above. The lower limit of quantification (LLOQ) was administratively set as the concentration of the lowest calibrator, which was determined during the establishment of the calibration curve. Carryover was evaluated by analyzing blank matrix samples immediately after the highest calibrators in the calibration runs. Carryover was considered present if the peak area of the blank sample exceeded 20% of the peak area of the lowest calibrator (LLOQ). Accuracy and precision Accuracy was determined by comparing the calculated concentration to the target concentration at low, medium, and high analyte levels in triplicate over five individual runs (n = 45). For non-conjugated compounds, the target concentrations were 120, 450 and 960 pg/mg, whereas those for glucuronide conjugates were 6, 30 and 100 pg/mg. Dihydrocodeine-6B-D-glucuronide was the exception. Because the upper LOQ (ULOQ) was determined to be 80 pg/mg, accuracy and precision were calculated using the low and medium QC concentrations (6 and 30 pg/mg, respectively; n = 30). Within-run precision was calculated as the average of the percent covariance (% CV) in the accuracy for each concentration within a run, averaged across all five runs. Between-run precision was calculated as the % CV in the accuracy of the samples at a single concentration in all five runs (n = 15), averaged across the three concentrations. Acceptable accuracy and precision limits were considered to be ±20%. For samples with target analyte concentrations above 1,200 pg/mg, an alternate preparation technique utilizing 10 mg of hair (instead of 25 mg), comparable to a dilution integrity test for liquid samples, was included in the validation. Dilution samples (n = 15) were prepared at a target analyte concentration of 2,500 pg/mg. For these samples, 10 mg of hair was extracted and quantified against a calibration curve prepared from 25-mg extractions to verify that the accuracy and precision were within acceptable limits. Stability Processed sample stability was determined by extracting and analyzing low, medium, and high QC samples in triplicate to establish a baseline (t0) and then re-analyzing the same extracted samples after 24, 48 and 72 h of storage at room temperature. Processed samples were considered stable if the calculated concentrations were within 20% of the t0 concentrations. Interferences and specificity Blank matrix samples containing all target analytes, all ISTDs, and a single potentially interfering compound were prepared. These samples contained the single potentially interfering compound at 500 pg/mg, non-conjugated analytes at 80 pg/mg, glucuronide conjugates at 3 pg/mg, non-conjugated ISTDs at 200 pg/mg and glucuronide conjugate ISTDs at 30 pg/mg. The potentially interfering compounds evaluated were heroin, norcodeine, norhydrocodone, normorphine, noroxycodone and noroxymorphone (Figure 3). An additional set of samples, each of which contained only a single potentially interfering compound and all ISTDs at the same concentrations listed above, was also evaluated. Interference was considered present for any compound that produced a peak area >20% of the peak area at the LLOQ for any analyte or that caused any calculated analyte concentration to deviate by more than 20% from its target concentration when interferences were mixed with analytes. Figure 3. View largeDownload slide Structures of the potentially interfering compounds that were evaluated during method validation. Figure 3. View largeDownload slide Structures of the potentially interfering compounds that were evaluated during method validation. Five blank matrix samples fortified with ISTDs were analyzed to demonstrate the absence of interference originating from the ISTDs. Ten additional matrix samples without the addition of analyte or ISTD were also analyzed to demonstrate the absence of interference from the matrix. In both cases, interference with an analyte was identified if the average peak area of the blank sample was >20% of the established LLOQ peak area for that analyte. To investigate the possibility of glucuronide conjugate interference arising during the sample preparation and extraction processes, blank hair samples were fortified with codeine, dihydrocodeine, dihydromorphine, hydrocodone, hydromorphone, morphine, oxycodone, oxymorphone and 6-acetylmorphine at 500, 800, 1,200 and 3,000 pg/mg (n = 3 at each concentration) and all ISTDS. These samples were extracted and analyzed for the presence of the glucuronide conjugates listed in Table I. Matrix effects Matrix effects were evaluated according to the method described by Matuszewski et al. (46) using 10 lots of blank hair. Three sets of samples were fortified with each target analyte at 3 pg/mg for glucuronide conjugates and 80 pg/mg for all other compounds. Type C samples (i.e., pre-extraction spiked samples) were made by fortifying blank hair with target analytes and ISTDs prior to solid-phase extraction. Type B samples (i.e., post-extraction spiked samples) were made by reconstituting the dried eluent from the solid-phase extraction of blank hair with mobile phase fortified with target analytes and ISTDs at equivalent concentrations to type C samples. Type A samples contained the target analytes and ISTDs in mobile phase A, at equivalent concentrations to type B and C samples. Matrix effect, extraction recovery, and process efficiency were calculated as follows:   Matrixeffect(%)=B/A×100  Extractionrecovery(%)=C/B×100  Processefficiency(%)=C/A×100where A, B and C are the mean peak areas for the quantitative transitions of type A, B and C samples, respectively. The matrix effect defined above is referred to as the absolute matrix effect because it is a comparison of the response in matrix to the response in neat mobile phase. The relative matrix effect was also assessed by comparing the % CV in the response across the 10 matrix lots for type B samples to the % CV in the response for type A samples. Results and Discussion Validation The method presented here produced calibration data with reproducible linear correlations with weighting factors of 1/x for each analyte. Throughout the use of the method for validation and sample analysis, consistent peak shapes, ion ratios, ISTD responses and retention times were achieved. The validation results are summarized in Table II. The method is linear from 40 to 1,200 pg/mg for all non-conjugated compounds and from 2 to 120 pg/mg for all glucuronide conjugates, except dihydrocodeine-6B-D-glucuronide, which has an ULOQ of 80 pg/mg. LODs range from 0.2 pg/mg to 1 pg/mg for all analytes. Overall accuracy was within acceptable limits and ranged from 94 to 108%. Within-run precision was consistent across all analytes, with a narrow range of % CVs (3.4–6.7); morphine-6B-D-glucuronide showed the most variability. As expected, between-run precision was slightly more variable and ranged from 5.1 to 11.2%; again, morphine 6B-D-glucuronide exhibited the most variability. Reducing the sample amount from 25 to 10 mg for high-concentration samples did not affect the accuracy or precision by more than ±15%. No carryover was observed in blank samples analyzed immediately following high calibrators, and processed samples were stable in the autosampler, with <10% deviation from t0 observed in calculated concentrations up to 72 h post-extraction. Table II. Validation summary Analytes     Precision (n=45)  LOD (pg/mg)  LLOQ (pg/mg)  ULOQ (pg/mg)  r2  Overall average accuracy % (n = 45)  Average within run precision %CV  Average between run precision %CV  Codeine  1  40  1,200  0.9978  94.2  4.1  6.6  Codeine-6B-d-glucuronide  0.5  2  120  0.9961  99.2  4.7  6.7  Dihydrocodeine  0.5  40  1,200  0.9950  94.9  4.6  7.9  Dihydrocodeine-6B-d-glucuronide  1  2  80  0.9960  96.8  4.2  7.5  Dihydromorphine  0.5  40  1,200  0.9977  101.1  5.1  7.4  Hydrocodone  1  40  1,200  0.9980  93.2  3.4  5.9  Hydromorphone  0.5  40  1,200  0.9979  107.8  3.6  5.5  Hydromorphone-3B-d-glucuronide  0.2  2  120  0.9928  99.5  6.0  10.0  Morphine  0.2  40  1,200  0.9960  97.3  4.1  6.2  Morphine-3B-d-glucuronide  0.5  2  120  0.9969  98.8  3.7  5.5  Morphine-6B-d-glucuronide  0.5  2  120  0.9964  102.0  6.7  11.2  Oxycodone  0.5  40  1,200  0.9977  92.6  3.5  5.1  Oxymorphone  0.5  40  1,200  0.9985  95.4  4.4  6.2  Oxymorphone-3B-d-glucuronide  0.5  2  120  0.9968  98.3  5.4  8.1  6-Acetylmorphine  1  40  1,200  0.9978  103.0  4.4  5.2  Analytes     Precision (n=45)  LOD (pg/mg)  LLOQ (pg/mg)  ULOQ (pg/mg)  r2  Overall average accuracy % (n = 45)  Average within run precision %CV  Average between run precision %CV  Codeine  1  40  1,200  0.9978  94.2  4.1  6.6  Codeine-6B-d-glucuronide  0.5  2  120  0.9961  99.2  4.7  6.7  Dihydrocodeine  0.5  40  1,200  0.9950  94.9  4.6  7.9  Dihydrocodeine-6B-d-glucuronide  1  2  80  0.9960  96.8  4.2  7.5  Dihydromorphine  0.5  40  1,200  0.9977  101.1  5.1  7.4  Hydrocodone  1  40  1,200  0.9980  93.2  3.4  5.9  Hydromorphone  0.5  40  1,200  0.9979  107.8  3.6  5.5  Hydromorphone-3B-d-glucuronide  0.2  2  120  0.9928  99.5  6.0  10.0  Morphine  0.2  40  1,200  0.9960  97.3  4.1  6.2  Morphine-3B-d-glucuronide  0.5  2  120  0.9969  98.8  3.7  5.5  Morphine-6B-d-glucuronide  0.5  2  120  0.9964  102.0  6.7  11.2  Oxycodone  0.5  40  1,200  0.9977  92.6  3.5  5.1  Oxymorphone  0.5  40  1,200  0.9985  95.4  4.4  6.2  Oxymorphone-3B-d-glucuronide  0.5  2  120  0.9968  98.3  5.4  8.1  6-Acetylmorphine  1  40  1,200  0.9978  103.0  4.4  5.2  Interferences and specificity At 500 pg/mg, heroin was the only interferent that produced a signal >20% of the LLOQ response for any analyte or caused the quantification of any analyte to deviate by more than 20% from the target concentration when it was mixed with analytes. In both cases, 6-acetylmorphine was the only impacted analyte. An interferent concentration of 500 pg/mg was used because it is 2.5 times the suggested cutoff concentration for opioids in hair: 200 pg/mg. This is analogous to the National Laboratory Certification Program (NLCP) guidelines for opiates in urine, which require interference testing at 5,000 ng/mL (i.e., 2.5 times the cutoff concentration of 2,000 ng/mL). No glucuronide conjugate interference was detected in extracted blank hair samples from 10 non-drug users or from blank hair fortified with codeine, dihydrocodeine, dihydromorphine, hydrocodone, hydromorphone, oxycodone, oxymorphone or 6-acetylmorphine at concentrations up to 3,000 pg/mg. Additionally, no glucuronide conjugate interference was detected in blank hair fortified with morphine at 500, 800 or 1,200 pg/mg. At 3,000 pg/mg, a peak was observed that interfered with the morphine-6-glucuronide transition both in the presence and absence of ISTDs; its peak area was approximately equal to that of the LLOQ. No interference was observed when standard analytes were spiked into mobile phase and extracted, indicating that this interference is attributable to an interaction between morphine and hair during the extraction process. Morphine samples (3,000 pg/mg) were prepared in triplicate in two additional lots of blank hair. In both of these lots, the average peak area for the interference was approximately one-third of the peak area of the LLOQ. Care should be taken in interpreting morphine-6-glucuronide results for samples with morphine concentrations above 3,000 pg/mg using this assay. Only one out of the 46 user hair samples tested had a morphine concentration above 3,000 pg/mg, (estimated to be 3,113 pg/mg), and its morphine-6-glucuronide concentration was >7 × LLOQ. Therefore, the impact of this interference on the results presented here is expected to be minimal. Using an alternate LC method, it was possible to separate this interference from morphine-6-glucuronide, indicating it is a true interference and not attributable to the formation of morphine-6-glucuroinde in the samples during the extraction process. Although the impact of this interference is minimal, it can be overcome via additional optimization of the LC method. Additional studies are underway to determine the cause of this interference. Matrix effects The matrix effects ranged from 43.9 to 111.5% and are summarized in Table III. For most analytes, there was ion suppression, indicated by matrix effect values of <100%. The only exceptions were codeine-6B-D-glucuronide and dihydrocodeine-6B-D-glucuronide, which exhibited slight ion enhancement, as indicated by a matrix effect >100%. Although the results show significant ion suppression and enhancement, this is not unexpected in an LC–MS-MS assay. The method presented here also appears to have a relative matrix effect, as evidenced by the increased variability in the peak areas of analytes spiked into 10 different lots of hair extraction eluent (type B samples) compared with that in the responses of analytes spiked into mobile phase only (type A samples). However, the % CVs of the peak areas for type B samples are all <13%. The relative matrix effect and the ion suppression and enhancement are effectively controlled for using matrix-matched calibrators and controls and appropriate ISTDs, as evidenced by the accuracy of the results obtained for type C samples (i.e., pre-extraction spiked samples). Extraction recovery and process efficiency ranged from 56.6 to 68.1% and from 28.4 to 73.0%, respectively, for glucuronide conjugates, whereas for non-conjugated compounds, these values ranged from 64.9 to 75.1% and from 28.5 to 58.6%, respectively. The overlapping ranges indicate no significant differences in the extraction recovery or process efficiency for the two groups of analytes. Table III. Matrix effect results Analytes  Matrix effect (%)  Extraction recovery (%)  Process efficiency (%)  Accuracy C samples  Accuracy %CV C samples  Peak area % CV  A samples  B samples  C samples  Codeine  75.8  69.9  53.0  100.4  4.5  2.3  8.3  7.9  Codeine-6B-d-glucuronide  107.9  67.7  73.0  84.1  7.0  2.8  12.1  8.0  Dihydrocodeine  75.0  71.9  53.9  106.6  5.4  1.5  9.7  8.9  Dihydrocodeine-6B-d-glucuronide  111.5  56.6  63.1  103.2  9.7  1.5  3.7  21.8  Dihydromorphine  79.9  71.0  56.7  107.8  2.7  2.2  10.6  2.1  Hydrocodone  53.7  66.9  35.9  106.3  3.2  1.6  10.0  14.2  Hydromorphone  75.6  71.4  54.0  102.4  4.1  0.8  10.6  4.8  Hydromorphone-3B-d-glucuronide  46.3  61.5  28.4  124.8  6.7  1.9  9.3  4.9  Morphine  73.0  75.1  54.8  96.7  2.7  1.8  9.2  3.9  Morphine-3B-d-glucuronide  45.5  68.1  30.9  85.4  4.8  2.4  11.1  3.4  Morphine-6B-d-glucuronide  77.8  65.6  51.0  88.7  4.2  2.1  11.9  5.1  Oxycodone  79.3  72.1  57.1  96.4  3.4  1.4  8.8  4.8  Oxymorphone  80.6  72.7  58.6  105.8  0.2  1.3  9.8  2.7  Oxymorphone-3B-d-glucuronide  52.5  67.6  35.5  83.9  11.3  2.2  12.0  12.0  6-Acetylmorphine  43.9  64.9  28.5  115.9  5.0  1.1  9.6  16.4  Analytes  Matrix effect (%)  Extraction recovery (%)  Process efficiency (%)  Accuracy C samples  Accuracy %CV C samples  Peak area % CV  A samples  B samples  C samples  Codeine  75.8  69.9  53.0  100.4  4.5  2.3  8.3  7.9  Codeine-6B-d-glucuronide  107.9  67.7  73.0  84.1  7.0  2.8  12.1  8.0  Dihydrocodeine  75.0  71.9  53.9  106.6  5.4  1.5  9.7  8.9  Dihydrocodeine-6B-d-glucuronide  111.5  56.6  63.1  103.2  9.7  1.5  3.7  21.8  Dihydromorphine  79.9  71.0  56.7  107.8  2.7  2.2  10.6  2.1  Hydrocodone  53.7  66.9  35.9  106.3  3.2  1.6  10.0  14.2  Hydromorphone  75.6  71.4  54.0  102.4  4.1  0.8  10.6  4.8  Hydromorphone-3B-d-glucuronide  46.3  61.5  28.4  124.8  6.7  1.9  9.3  4.9  Morphine  73.0  75.1  54.8  96.7  2.7  1.8  9.2  3.9  Morphine-3B-d-glucuronide  45.5  68.1  30.9  85.4  4.8  2.4  11.1  3.4  Morphine-6B-d-glucuronide  77.8  65.6  51.0  88.7  4.2  2.1  11.9  5.1  Oxycodone  79.3  72.1  57.1  96.4  3.4  1.4  8.8  4.8  Oxymorphone  80.6  72.7  58.6  105.8  0.2  1.3  9.8  2.7  Oxymorphone-3B-d-glucuronide  52.5  67.6  35.5  83.9  11.3  2.2  12.0  12.0  6-Acetylmorphine  43.9  64.9  28.5  115.9  5.0  1.1  9.6  16.4  Analysis of drug user hair samples User hair samples were obtained from a local substance abuse treatment facility from October 2015 through September 2016. All samples received were washed at RTI International using the wash procedure described in the methods section. Aliquots of each were sent to Psychemedics Corporation (Acton, MA) for initial screening and confirmation of codeine, hydrocodone, hydromorphone, morphine, oxycodone, oxymorphone and 6-acetylmorphine. Forty-six samples were confirmed positive for at least one of these analytes and were subsequently extracted and analyzed at RTI using the validated method for opioids and their glucuronide conjugates described here. The results of the analyses performed at RTI are presented in Table IV. Table IV. RTI quantitative results for user hair samples Sample ID  6-AM  HYC  OXYC  COD  COD-6G  DHYC  DHYC-6G  HYM  HYM-3G  OXYM  OXYM-3G  DHYM  DHYM-3G  MOR  MOR-3G  MOR-6G  1  798  692  154  105  <2 (1.87)  –  –  52  –  –  –  –  –  752  2.78  4.96  2  >3000 (5199)  60  –  427  4.13  –  –  257  2.37  –  –  –  –  >3000 (3113)  6.89  14.96  3  –  407  –  –  –  –  –  –  –  –  –  –  –  –  –  –  4  1067  –  –  –  –  –  –  –  –  –  –  –  –  131  –  –  5  2496  49  299  44  –  –  –  –  –  89  <2 (1.08)  –  –  562  –  <2 (1.24)  6  –  77  1585  272  5.65  –  –  –  –  –  –  –  –  –  –  –  7  212  –  –  –  –  –  –  –  –  –  –  –  –  72  –  –  8  –  85  661  65  –  –  –  –  –  120  –  –  –  –  –  –  9  64  568  259  –  –  41  –  –  –  –  –  –  –  –  –  –  10  –  50  499  –  –  –  –  –  –  –  –  –  –  –  –  –  11  –  40  436  –  –  –  –  –  –  –  –  –  –  –  –  –  12  >3000 (4248)  45  –  285  4.22  –  –  137  <2 (1.79)  –  –  –  –  2832  9.41  13.86  13  –  –  975  –  –  –  –  843  4.97  2183  3.42  –  –  125  –  –  14  508  –  –  65  –  –  –  75  –  –  –  –  –  473  2.00  2.70  15  1370  398  299  118  7.22  –  –  64  –  –  –  –  –  457  2.26  7.01  16  53  –  –  –  –  –  –  –  –  –  –  –  –  118  <2 (1.22)  <2 (1.08)  17  –  106  –  –  –  –  –  –  –  –  –  –  –  –  –  –  18  –  364  966  –  –  –  –  –  –  78  –  –  –  –  –  –  19  194  –  –  –  –  –  –  –  –  –  –  –  –  89  –  <2 (1.20)  20  –  447  –  –  –  43  –  –  –  –  –  –  –  –  –  21  238  –  –  –  –  –  –  –  –  –  –  –  –  –  –  –  22  134  –  1490  –  –  –  –  72  –  1712  3.60  –  –  70  –  –  23  –  –  105  –  –  –  –  –  –  –  –  –  –  –  –  –  24  440  41  –  46  <2 (1.87)  –  –  –  –  –  –  –  –  402  3.90  5.57  25  419  –  –  –  –  –  –  –  –  –  –  –  –  223  <2 (1.05)  <2 (1.56)  26  314  78  480  –  –  –  –  –  –  –  –  –  –  206  –  –  27  448  –  2536  46  <2 (1.05)  –  –  –  –  1190  8.93  –  –  489  2.30  4.03  28  162  –  2349  –  –  –  –  –  –  –  –  –  –  70  –  –  29  2401  –  –  78  2.52  –  –  82  –  665  4.80  –  –  1083  3.62  8.90  30  –  982  2988  –  –  –  –  –  –  436  <2 (0.98)  –  –  200  –  –  31  –  187  237  –  –  –  –  –  –  –  –  –  –  –  –  32  2934  74  916  75  –  –  –  –  –  –  –  –  –  469  <2 (1.01)  <2 (1.20)  33  >3000 (6846)  –  –  280  5.37  –  –  209  –  –  –  –  –  2373  9.35  24.97  34  –  –  484  –  –  –  –  –  –  –  –  –  –  –  –  35  –  150  483  –  –  –  –  –  –  –  –  –  –  –  –  –  36  –  81  2932  –  –  –  –  –  –  134  2.86  –  –  –  –  –  37  429  –  69  76  <2 (1.89)  –  –  46  –  –  –  –  –  606  4.63  7.95  38  2158  92  –  269  2.86  –  –  83  –  425  3.25  –  –  1422  4.32  8.39  39  643  –  43  –  –  –  –  –  –  –  –  –  –  80  –  –  40  –  619  68  –  –  –  –  –  <2 (1.05)  –  –  –  –  –  –  –  41  191  –  160  –  –  –  –  –  –  –  –  –  –  44  <2 (0.92)  <2 (1.20)  42  206  –  -  –  –  –  –  –  –  –  –  –  –  –  –  –  43  –  –  1334  –  –  –  –  –  –  –  –  –  –  –  –  –  44  303  102  –  –  –  –  –  –  –  –  –  –  –  254  <2 (1.68)  2.29  45  >3000 (5132)  60  52  51  –  –  –  –  –  –  –  –  –  640  <2 (0.98)  <2 (1.31)  46  –  296  283  –  –  –  –  –  –  –  –  –  –  –  –  –  Sample ID  6-AM  HYC  OXYC  COD  COD-6G  DHYC  DHYC-6G  HYM  HYM-3G  OXYM  OXYM-3G  DHYM  DHYM-3G  MOR  MOR-3G  MOR-6G  1  798  692  154  105  <2 (1.87)  –  –  52  –  –  –  –  –  752  2.78  4.96  2  >3000 (5199)  60  –  427  4.13  –  –  257  2.37  –  –  –  –  >3000 (3113)  6.89  14.96  3  –  407  –  –  –  –  –  –  –  –  –  –  –  –  –  –  4  1067  –  –  –  –  –  –  –  –  –  –  –  –  131  –  –  5  2496  49  299  44  –  –  –  –  –  89  <2 (1.08)  –  –  562  –  <2 (1.24)  6  –  77  1585  272  5.65  –  –  –  –  –  –  –  –  –  –  –  7  212  –  –  –  –  –  –  –  –  –  –  –  –  72  –  –  8  –  85  661  65  –  –  –  –  –  120  –  –  –  –  –  –  9  64  568  259  –  –  41  –  –  –  –  –  –  –  –  –  –  10  –  50  499  –  –  –  –  –  –  –  –  –  –  –  –  –  11  –  40  436  –  –  –  –  –  –  –  –  –  –  –  –  –  12  >3000 (4248)  45  –  285  4.22  –  –  137  <2 (1.79)  –  –  –  –  2832  9.41  13.86  13  –  –  975  –  –  –  –  843  4.97  2183  3.42  –  –  125  –  –  14  508  –  –  65  –  –  –  75  –  –  –  –  –  473  2.00  2.70  15  1370  398  299  118  7.22  –  –  64  –  –  –  –  –  457  2.26  7.01  16  53  –  –  –  –  –  –  –  –  –  –  –  –  118  <2 (1.22)  <2 (1.08)  17  –  106  –  –  –  –  –  –  –  –  –  –  –  –  –  –  18  –  364  966  –  –  –  –  –  –  78  –  –  –  –  –  –  19  194  –  –  –  –  –  –  –  –  –  –  –  –  89  –  <2 (1.20)  20  –  447  –  –  –  43  –  –  –  –  –  –  –  –  –  21  238  –  –  –  –  –  –  –  –  –  –  –  –  –  –  –  22  134  –  1490  –  –  –  –  72  –  1712  3.60  –  –  70  –  –  23  –  –  105  –  –  –  –  –  –  –  –  –  –  –  –  –  24  440  41  –  46  <2 (1.87)  –  –  –  –  –  –  –  –  402  3.90  5.57  25  419  –  –  –  –  –  –  –  –  –  –  –  –  223  <2 (1.05)  <2 (1.56)  26  314  78  480  –  –  –  –  –  –  –  –  –  –  206  –  –  27  448  –  2536  46  <2 (1.05)  –  –  –  –  1190  8.93  –  –  489  2.30  4.03  28  162  –  2349  –  –  –  –  –  –  –  –  –  –  70  –  –  29  2401  –  –  78  2.52  –  –  82  –  665  4.80  –  –  1083  3.62  8.90  30  –  982  2988  –  –  –  –  –  –  436  <2 (0.98)  –  –  200  –  –  31  –  187  237  –  –  –  –  –  –  –  –  –  –  –  –  32  2934  74  916  75  –  –  –  –  –  –  –  –  –  469  <2 (1.01)  <2 (1.20)  33  >3000 (6846)  –  –  280  5.37  –  –  209  –  –  –  –  –  2373  9.35  24.97  34  –  –  484  –  –  –  –  –  –  –  –  –  –  –  –  35  –  150  483  –  –  –  –  –  –  –  –  –  –  –  –  –  36  –  81  2932  –  –  –  –  –  –  134  2.86  –  –  –  –  –  37  429  –  69  76  <2 (1.89)  –  –  46  –  –  –  –  –  606  4.63  7.95  38  2158  92  –  269  2.86  –  –  83  –  425  3.25  –  –  1422  4.32  8.39  39  643  –  43  –  –  –  –  –  –  –  –  –  –  80  –  –  40  –  619  68  –  –  –  –  –  <2 (1.05)  –  –  –  –  –  –  –  41  191  –  160  –  –  –  –  –  –  –  –  –  –  44  <2 (0.92)  <2 (1.20)  42  206  –  -  –  –  –  –  –  –  –  –  –  –  –  –  –  43  –  –  1334  –  –  –  –  –  –  –  –  –  –  –  –  –  44  303  102  –  –  –  –  –  –  –  –  –  –  –  254  <2 (1.68)  2.29  45  >3000 (5132)  60  52  51  –  –  –  –  –  –  –  –  –  640  <2 (0.98)  <2 (1.31)  46  –  296  283  –  –  –  –  –  –  –  –  –  –  –  –  –  Concentrations are in pg/mg. All calculated concentrations >0.9 pg/mg are included. Values >3000 or <2 pg/mg are out of the validated calibration range of the method. 6-Acetylmorphine (6-AM), hydrocodone (HYC), ixycodone (OXYC), codeine (COD), codeine-6-glucuronide (COD-6G), dihydrocodeine (DHYC), dihydrocodeine-6-glucuronide (DHYC-6G), hydromorphone (HYM), hydromorphone-3-glucuronide (HYM-3G), oxymorphone (OXYM), oxymorphone-3-glucuronide (OXYM-3G), dihydromorphine (DHYM), dihydromorphine-3-glucuronide (DHYM-3G), morphine (MOR), morphine-3-glucuronide (MOR-3G), morphine-6-glucuronide (MOR-6G). Codeine-6-glucuronide, hydromorphone-3-glucuronide, oxymorphone-3-glucuronide, morphine-3-glucuronide and morphine-6-glucuronide were reliably detected in samples containing the corresponding parent compound, and the concentrations of the glucuronide conjugates generally increased as the parent concentrations increased (Figure 4). Morphine-6-glucuronide was present at a higher concentration than morphine-3-glucuronide. Every sample that contained morphine 3-glucuronide also contained morphine-6-glucuronide, but two samples with morphine-6-glucuronide did not contain detectable levels of morphine-3-glucuronide. Two samples had relatively high concentrations of 6-acetylmorphine (1,067 and 643 pg/mg) but much lower concentrations of morphine (131 and 80 pg/mg) and no detectable morphine-3-glucuronide or morphine-6-glucuronide. Figure 4. View largeDownload slide Glucuronide concentrations as a function of the corresponding parent concentrations in user hair samples. Figure 4. View largeDownload slide Glucuronide concentrations as a function of the corresponding parent concentrations in user hair samples. Five out of five samples that had codeine concentrations above 200 pg/mg also had codeine-6-glucuronide concentrations above 2 pg/mg. Two additional samples had codeine-6-glucuronide concentrations above 2 pg/mg with codeine concentrations of only 118 and 78 pg/mg. Two out of three samples that had hydromorphone concentrations above 200 pg/mg also had hydromorphone-3-glucuronide concentrations above 2 pg/mg. Five out of six samples that had oxymorphone concentrations >200 pg/mg also had oxymorphone-3-glucuronide concentrations >2 pg/mg. One additional sample had an oxymorphone-3-glucuronide concentration above 2 pg/mg with an oxymorphone concentration of only 134 pg/mg. Twelve out of 16 samples that had morphine concentrations above 200 pg/mg also had morphine-6-glucuronide concentrations above 2 pg/mg, and 11 out of 16 had morphine-3-glucuroinde concentrations above 2 pg/mg. Hydromorphone-3-glucuronide present in the samples could have originated from hydromorphone or hydrocodone. Although the dataset is limited, it indicates that hydromorphone-3-glucuronide most likely originated from hydromorphone. The two samples with hydromorphone-3-glucuronide concentrations above the LLOQ had hydromorphone concentrations of 257 and 843 pg/mg. One of these samples contained no detectable hydrocodone, whereas the other had only 60 pg/mg hydrocodone. There are also several samples with high hydrocodone concentrations (up to 982 pg/mg) but no detectable hydromorphone-3-glucuronide. Similarly, oxymorphone-3-glucuronide could have originated from oxymorphone or oxycodone, but the dataset suggests that its origin was oxymorphone. All samples containing oxymorphone-3-glucuronide also contained oxymorphone, but two of these samples (samples 38 and 29, which had oxymorphone concentrations of 425 and 665 pg/mg, respectively) had no detectable oxycodone. There are numerous samples with high oxycodone concentrations (several over 1,000 pg/mg) but no detectable oxymorphone-3-glucuronide. The glucuronide conjugate concentrations relative to parent concentrations (in samples with both present) were as follows: Codeine-6-glucuronide ranged from 0.97 to 6.12%, with average and % CV values of 2.33 and 67.65%, respectively. Hydromorphone-3-glucuronide ranged from 0.59 to 1.30%, with average and % CV values of 0.94 and 31.03%, respectively. Oxymorphone-3-glucuronide ranged from 0.16 to 2.13%, with average and % CV values of 0.77 and 79.94%, respectively. Morphine-3-glucuronide ranged from 0.15 to 2.00%, with average and % CV values of 0.59 and 76.84%, respectively. Morphine-6-glucuronide ranged from 0.20 to 2.73%, with average and % CV values of 0.93 and 61.88%, respectively. For samples with calculated glucuronide conjugate concentrations between 0.9 and 2 pg/mg, the estimated concentrations denoted in parenthesis in Table IV were used in these calculations. Only two samples had quantifiable levels of dihydrocodeine: one at 41 pg/mg and one at 43 pg/mg. Dihydrocodeine-6-glucuronide was not detected in any of the samples. Dihydromorphine-3-glucuronide was not included in the validated assay because it interfered with morphine-3-glucuronide-d3 transitions. However, instrument parameters were optimized for this analyte, and its presence was monitored in all samples. None of the user hair samples had a dihydromorphine-3-glucuronide peak area that was greater than the average area plus 3.3 times the standard deviation of the peak area for this transition in blank samples. This result is not unexpected because none of the samples had dihydromorphine concentrations above the LLOQ (40 pg/mg). The results from the analyses performed by Psychemedics are presented in Table V. The results for analytes detected at both laboratories are compared in Table VI. Samples were sent to Psychemedics for analysis in multiple batches between October 2015 and December 2016, whereas the 46 opioid-positive samples were extracted and analyzed at RTI over 2 days in February 2017. The two data sets agree very well, despite the differences in the timing and location of the analyses. Thus, the methods used to detect parent compounds at both laboratories are robust, and the analytes are stable in hair stored under ambient conditions. The two discrepancies in the number of samples >200 pg/mg both correspond to samples very near this suggested cutoff. One sample was determined to contain 186 pg/mg hydrocodone by RTI and 238 pg/mg hydrocodone by Psychemedics. The other sample was determined to contain 206 pg/mg morphine at RTI and 176 pg/mg morphine by Psychemedics. The average and median concentration differences for all analytes with concentrations exceeding 40 pg/mg at both laboratories are less than 30 and 20%, respectively. Table V. Psychemedics quantitative results for user hair samples Sample ID  6-AM  HYC  OXYC  COD  HYM  OXYM  MOR  1  769  716  182  117  59  –  621  2  6010  59  –  566  273  –  3025  3  –  594  –  –  –  –  –  4  1565  –  –  –  –  –  177  5  2610  –  334  54  –  91  545  6  –  84  1740  446  –  –  –  7  242  –  –  –  –  –  102  8  –  72  497  48  –  117  –  9  102  838  412  –  –  –  64  10  –  –  462  –  –  –  –  11  –  45  379  –  –  –  –  12  4975  60  –  365  146  –  3370  13  26  –  1035  –  847  1765  113  14  701  –  68  66  74  –  515  15  1390  376  276  93  56  –  403  16  68  –  –  33  –  –  168  17  –  111  –  –  –  –  –  18  –  508  1390  –  –  109  –  19  183  –  –  –  –  –  98  20  –  496  –  –  –  –  –  21  271  –  –  –  –  –  51  22  160  –  1405  –  71  1670  87  23  –  –  121  –  –  –  –  24  420  47  –  52  36  –  432  25  405  –  –  –  –  –  214  26  340  69  329  –  –  –  176  27  392  –  2435  –  –  1180  437  28  185  –  2225  –  –  40  114  29  1415  –  –  66  69  475  864  30  –  1004  2655  –  –  382  152  31  –  238  206  –  27  –  –  32  2150  124  1036  71  –  –  493  33  6950  –  –  246  202  –  2380  34  –  –  375  –  –  –  –  35  –  162  482  –  –  –  –  36  –  83  3240  –  –  119  –  37  741  –  126  158  86  –  1145  38  2425  74  –  220  71  226  1032  39  513  –  31  –  –  –  66  40  –  536  62  –  –  –  –  41  180  –  136  –  –  –  56  42  303  –  –  –  –  –  –  43  –  –  942  –  –  –  –  44  258  –  –  –  –  –  398  45  3699  –  –  –  –  –  857  46  –  209  293  –  –  –  –  Sample ID  6-AM  HYC  OXYC  COD  HYM  OXYM  MOR  1  769  716  182  117  59  –  621  2  6010  59  –  566  273  –  3025  3  –  594  –  –  –  –  –  4  1565  –  –  –  –  –  177  5  2610  –  334  54  –  91  545  6  –  84  1740  446  –  –  –  7  242  –  –  –  –  –  102  8  –  72  497  48  –  117  –  9  102  838  412  –  –  –  64  10  –  –  462  –  –  –  –  11  –  45  379  –  –  –  –  12  4975  60  –  365  146  –  3370  13  26  –  1035  –  847  1765  113  14  701  –  68  66  74  –  515  15  1390  376  276  93  56  –  403  16  68  –  –  33  –  –  168  17  –  111  –  –  –  –  –  18  –  508  1390  –  –  109  –  19  183  –  –  –  –  –  98  20  –  496  –  –  –  –  –  21  271  –  –  –  –  –  51  22  160  –  1405  –  71  1670  87  23  –  –  121  –  –  –  –  24  420  47  –  52  36  –  432  25  405  –  –  –  –  –  214  26  340  69  329  –  –  –  176  27  392  –  2435  –  –  1180  437  28  185  –  2225  –  –  40  114  29  1415  –  –  66  69  475  864  30  –  1004  2655  –  –  382  152  31  –  238  206  –  27  –  –  32  2150  124  1036  71  –  –  493  33  6950  –  –  246  202  –  2380  34  –  –  375  –  –  –  –  35  –  162  482  –  –  –  –  36  –  83  3240  –  –  119  –  37  741  –  126  158  86  –  1145  38  2425  74  –  220  71  226  1032  39  513  –  31  –  –  –  66  40  –  536  62  –  –  –  –  41  180  –  136  –  –  –  56  42  303  –  –  –  –  –  –  43  –  –  942  –  –  –  –  44  258  –  –  –  –  –  398  45  3699  –  –  –  –  –  857  46  –  209  293  –  –  –  –  Concentrations are in pg/mg. LOD/LLOQ for all is 25 pg/mg, ULOL is 10,000 pg/mg for morphine, 15,000 pg/mg for all other analytes. Table VI. Comparison of quantitative results from RTI and Psychemedics   Number of samples, >40 pg/mg  Number of samples, >200 pg/mg  Average absolute concentration difference (%)  Median absolute concentration difference (%)  RTI  Psychemedics  Codeine  14  5  5  27  20  Hydrocodone  22  9  10  19  13  Hydromorphone  11  3  3  15  6  Morphine  27  17  16  23  18  Oxycodone  26  21  21  18  12  Oxymorphone  10  6  6  17  12  6-Acetylmorphine  28  22  22  21  15    Number of samples, >40 pg/mg  Number of samples, >200 pg/mg  Average absolute concentration difference (%)  Median absolute concentration difference (%)  RTI  Psychemedics  Codeine  14  5  5  27  20  Hydrocodone  22  9  10  19  13  Hydromorphone  11  3  3  15  6  Morphine  27  17  16  23  18  Oxycodone  26  21  21  18  12  Oxymorphone  10  6  6  17  12  6-Acetylmorphine  28  22  22  21  15  Conclusion This work demonstrates for the first time that codeine-6-glucuronide, hydromorphone-3-glucuronide, oxymorphone-3-glucuronide, morphine 3-glucuronide and morphine 6-glucuronide are present at sufficient concentrations to be quantifiable in hair from opioid users. Of the 46 samples analyzed in this study, only two contained dihydrocodeine at concentrations >40 pg/mg, and none had quantifiable concentrations of dihydromorphine. Further studies on hair positive for these analytes are needed to assess the detectability of dihydrocodeine glucuronide and dihydromorphine glucuronide in human hair. The LLOQ of this method for glucuronide conjugates is 2 pg/mg, but based on the determined concentrations of the analytes in user hair samples, a lower LLOQ (i.e., 1 pg/mg or lower) would be desirable. Using cutoffs of 200 pg/mg for parent and 1 pg/mg for glucuronide metabolites sixteen out of sixteen morphine positive samples were positive for morphine-6-glucuronide, five out of five codeine positive samples were positive for codeine-6-glucuronide, two out of three hydromorphone positive samples were positive for hydromorphone-3-glucuronide and six out of six oxymorphone positive samples were positive for oxymorphone-3-glucuronide. The overall process efficiency and extraction recovery of this method were less than ideal and could likely be improved with further refinement of the extraction method to achieve suitable sensitivity and LLOQs for glucuronide conjugates in human hair. Supplementary Data Supplementary data is available at Journal of Analytical Toxicology online. Acknowledgments The authors thank Dr Edward J. Cone for helpful discussions throughout the project and for reviewing the article. We also thank Amy Evans, Susan Crumpton, Frank Esposito and the United States Substance Abuse and Mental Health Services Administration (SAMHSA). Without the coordinated effort of all these people, this study would not have been possible. References 1 SAMHSA. ( 2004) Proposed revisions to mandatory guidelines for federal workplace drug testing programs. Federal Register , 69, 19673– 19732. 2 Henderson, G.L., Harkey, M.R., Zhou, C., Jones, R.T., Jacob, P., III ( 1996) Incorporation of isotopically labeled cocaine and metabolites into human hair: 1. Dose-response relationships. Journal of Analytical Toxicology , 20, 1– 12. Google Scholar CrossRef Search ADS PubMed  3 LeBeau, M.A., Montgomery, M.A., Brewer, J.D. ( 2011) The role of variations in growth rate and sample collection on interpreting results of segmental analyses of hair. Forensic Science International , 210, 110– 116. Google Scholar CrossRef Search ADS PubMed  4 Nakahara, Y. ( 1999) Hair analysis for abused and therapeutic drugs. Journal of Chromatography B , 733, 161– 180. Google Scholar CrossRef Search ADS   5 SOHT. ( 2004) Recommendations for hair testing in forensic cases. Forensic Science International , 145, 83– 84. CrossRef Search ADS PubMed  6 Cooper, G.A.A., Kronstrand, R., Kintz, P. ( 2012) Society of Hair Testing guidelines for drug testing in hair. Forensic Science International , 218, 20– 24. Google Scholar CrossRef Search ADS PubMed  7 Baumgartner, W.A., Hill, V.A., Blahd, W.H. ( 1989) Hair analysis for drugs of abuse. Journal of Forensic Sciences , 34, 1433– 1453. Google Scholar CrossRef Search ADS   8 Cairns, T., Hill, V., Schaffer, M., Thistle, W. ( 2004) Levels of cocaine and its metabolites in washed hair of demonstrated cocaine users and workplace subjects. Forensic Science International , 145, 175– 181. Google Scholar CrossRef Search ADS PubMed  9 Cairns, T., Hill, V., Schaffer, M., Thistle, W. ( 2004) Amphetamines in washed hair of demonstrated users and workplace subjects. Forensic Science International , 145, 137– 142. Google Scholar CrossRef Search ADS PubMed  10 Kidwell, D., Smith, F. Passive exposure, decontamination procedures, cutoffs, and bias. In: Kintz, P. (ed). Analytical and Practical Aspects of Drug Testing in Hair . Boca Raton, FL: CRC Press, 2006; pp. 25– 72. Chapter 2. Google Scholar CrossRef Search ADS   11 Kidwell, D.A., Smith, F.P., Shepherd, A.R. ( 2015) Ethnic hair care products may increase false positives in hair drug testing. Forensic Science International , 257, 160– 164. Google Scholar CrossRef Search ADS PubMed  12 Koren, G., Klein, J., Forman, R., Graham, K. ( 1992) Hair analysis of cocaine: differentiation between systemic exposure and external contamination. The Journal of Clinical Pharmacology , 32, 671– 675. Google Scholar CrossRef Search ADS PubMed  13 Morris-Kukoski, C.L., Montgomery, M.A., Hammer, R.L. ( 2014) Analysis of extensively washed hair from cocaine users and drug chemists to establish new reporting criteria. Journal of Analytical Toxicology , 38, 628– 636. Google Scholar CrossRef Search ADS PubMed  14 Ropero-Miller, J.D., Huestis, M.A., Stout, P.R. ( 2012) Cocaine analytes in human hair: evaluation of concentration ratios in different cocaine sources, drug-user populations and surface-contaminated specimens. Journal of Analytical Toxicology , 36, 390– 398. Google Scholar CrossRef Search ADS PubMed  15 Schaffer, M., Hill, V., Cairns, T. ( 2005) Hair analysis for cocaine: the requirement for effective wash procedures and effects of drug concentration and hair porosity in contamination and decontamination. Journal of Analytical Toxicology , 29, 319– 326. Google Scholar CrossRef Search ADS PubMed  16 Stout, P.R., Horn, C.K., Klette, K.L., Given, J. ( 2006) Occupational exposure to methamphetamine in workers preparing training aids for drug detection dogs. Journal of Analytical Toxicology , 30, 551– 553. Google Scholar CrossRef Search ADS PubMed  17 Welch, M.J., Sniegoski, L.T., Allgood, C.C., Habram, M. ( 1993) Hair analysis for drugs of abuse: evaluation of analytical methods, environmental issues, and development of reference materials. Journal of Analytical Toxicology , 17, 389– 398. Google Scholar CrossRef Search ADS PubMed  18 Cone, E.J. ( 1996) Mechanisms of drug incorporation into hair. Therapeutic Drug Monitoring , 18, 438– 443. Google Scholar CrossRef Search ADS PubMed  19 Henderson, G.L. ( 1993) Mechanisms of drug incorporation into hair. Forensic Science International , 63, 19– 29. Google Scholar CrossRef Search ADS PubMed  20 Porta, T., Grivet, C., Kraemer, T., Varesio, E., Hopfgartner, G. ( 2011) Single hair cocaine consumption monitoring by mass spectrometric imaging. Analytical Chemistry , 83, 4266– 4272. Google Scholar CrossRef Search ADS PubMed  21 Potsch, L., Skopp, G., Moeller, M.R. ( 1997) Biochemical approach on the conservation of drug molecules during hair fiber formation. Forensic Science International , 84, 25– 35. Google Scholar CrossRef Search ADS PubMed  22 Bassindale, T. ( 2012) Quantitative analysis of methamphetamine in hair of children removed from clandestine laboratories—evidence of passive exposure? Forensic Science International , 219, 179– 182. Google Scholar CrossRef Search ADS PubMed  23 Farst, K., Reading Meyer, J.A., Mac Bird, T., James, L., Robbins, J.M. ( 2011) Hair drug testing of children suspected of exposure to the manufacture of methamphetamine. Journal of Forensic and Legal Medicine , 18, 110– 114. Google Scholar CrossRef Search ADS PubMed  24 Martyny, J.W., Arbuckle, S.L., McCammon, C.S., Jr, Erb, N., Van Dyke, M. ( 2008) Methamphetamine contamination on environmental surfaces caused by simulated smoking of methamphetamine. Journal of Chemical Health and Safety , 15, 25– 31. Google Scholar CrossRef Search ADS   25 Baumgartner, W., Hill, V. Hair analysis for drugs of abuse: decontamination issues. In: Sunshine, I. (ed). Recent Developments in Therapeutic Drug Monitoring and Clinical Toxicology . Marcel Dekker: New York, 1992; pp. 577– 597. 26 Cairns, T., Hill, V., Schaffer, M., Thistle, W. ( 2004) Removing and identifying drug contamination in the analysis of human hair. Forensic Science International , 145, 97– 108. Google Scholar CrossRef Search ADS PubMed  27 Hill, V., Loni, E., Cairns, T., Sommer, J., Schaffer, M. ( 2014) Identification and analysis of damaged or porous hair. Drug Testing and Analysis , 6, 42– 54. Google Scholar CrossRef Search ADS PubMed  28 Schaffer, M., Cheng, C.-C., Chao, O., Hill, V., Matsui, P. ( 2016) Analysis of cocaine and metabolites in hair: validation and application of measurement of hydroxycocaine metabolites as evidence of cocaine ingestion. Analytical and Bioanalytical Chemistry , 408, 2043– 2054. Google Scholar CrossRef Search ADS PubMed  29 Montgomery, M., LeBeau, M., Morris-Kukoski, C. ( 2016) New hair testing conclusions. Journal of Analytical Toxicology , 41, 161– 162. 30 White, R.M. ( 2017) Drugs in hair. Part I. Metabolisms of major drug classes. Forensic Science Review , 29, 23– 55. Google Scholar PubMed  31 Smith, H.S. ( 2009) Opioid metabolism. Mayo Clinic Proceedings , 84, 613– 624. Google Scholar CrossRef Search ADS PubMed  32 Burchell, B., Brierley, C.H., Rance, D. ( 1995) Specificity of human UDP-glucuronosyltransferases and xenobiotic glucuronidation. Life sciences , 57, 1819– 1831. Google Scholar CrossRef Search ADS PubMed  33 Wang, X., Johansen, S.S., Zhang, Y., Jia, J., Rao, Y., Jiang, F., et al.  . ( 2017) Deposition of diazepam and its metabolites in hair following a single dose of diazepam. International Journal of Legal Medecine , 131, 131– 141. Google Scholar CrossRef Search ADS   34 Pichini, S., Marchei, E., Martello, S., Gottardi, M., Pellegrini, M., Svaizer, F., et al.  . ( 2015) Identification and quantification of 11-nor-delta9-tetrahydrocannabinol-9-carboxylic acid glucuronide (THC-COOH-glu) in hair by ultra-performance liquid chromatography tandem mass spectrometry as a potential hair biomarker of cannabis use. Forensic Science International , 249, 47– 51. Google Scholar CrossRef Search ADS PubMed  35 Beasley, E., Francese, S., Bassindale, T. ( 2016) Detection and mapping of cannabinoids in single hair samples through rapid derivatization and matrix-assisted laser desorption ionization mass spectrometry. Analytical Chemistry , 88, 10328– 10334. Google Scholar CrossRef Search ADS PubMed  36 Kim, H.S., Cheong, J.C., Lee, J.I., In, M.K. ( 2013) Rapid and sensitive determination of propofol glucuronide in hair by liquid chromatography and tandem mass spectrometry. Journal of Pharmaceutical and Biomedical Analysis , 85, 33– 39. Google Scholar CrossRef Search ADS PubMed  37 Wang, X., Linnet, K., Johansen, S.S. ( 2016) Development of a UPLC-MS/MS method for determining gamma-hydroxybutyric acid (GHB) and GHB glucuronide concentrations in hair and application to forensic cases. Forensic Toxicology , 34, 51– 60. Google Scholar CrossRef Search ADS   38 Kim, J., In, S., Park, Y., Park, M., Kim, E., Lee, S. ( 2013) Quantitative analysis of propofol-glucuronide in hair as a marker for propofol abuse. Analytical and Bioanalytical Chemistry , 405, 6807– 6814. Google Scholar CrossRef Search ADS PubMed  39 Kwak, J.H., Kim, H.K., Choe, S., In, S., Pyo, J.S. ( 2016) Determination of propofol glucuronide from hair sample by using mixed mode anion exchange cartridge and liquid chromatography tandem mass spectrometry. Journal of Chromatography B , 1015–1016, 209– 213. Google Scholar CrossRef Search ADS   40 Gygi, S.P., Joseph, R.E., Jr, Cone, E.J., Wilkins, D.G., Rollins, D.E. ( 1996) Incorporation of codeine and metabolites into hair. Role of pigmentation. Drug Metabolism and Disposition , 24, 495– 501. Google Scholar PubMed  41 Gygi, S.P., Colón, F., Raftogianis, R.B., Galinsky, R.E., Wilkins, D.G., Rollins, D.E. ( 1996) Dose-related distribution of codeine and its metabolites into rat hair. Drug Metabolism and Disposition , 24, 282– 287. Google Scholar PubMed  42 Toyooka, T., Yano, M., Kato, M., Nakahara, Y. ( 2001) Simultaneous determination of morphine and its glucuronides in rat hair and rat plasma by reversed-phase liquid chromatography with electrospray ionization mass spectrometry. Analyst , 126, 1339– 1345. Google Scholar CrossRef Search ADS PubMed  43 Waters Corporation. ( 2015) A simplified, mixed-mode sample preparation strategy for urinary forensic toxicology screening by UPLC-MS/MS [updated 2015] http://www.waters.com/webassets/cms/library/docs/720005290en.pdf. 44 Lott, S., Musshoff, F., Madea, B. ( 2013) LC/MS/MS method of 6-MAM, morphine, morphine-3-glucuro-nide (M3G) and morphine-6-glucuronide (M6G) for quantitative analysis in serum. Toxichem Krimtech , 80, 363– 366. 45 SWGTOX. ( 2013) Scientific Working Group for Forensic Toxicology (SWGTOX) standard practices for method validation in forensic toxicology [updated 2013]. http://www.swgtox.org/documents/Validation3.pdf. 46 Matuszewski, B.K., Constanzer, M.L., Chavez-Eng, C.M. ( 2003) Strategies for the assessment of matrix effect in quantitative bioanalytical methods based on HPLC-MS/MS. Analytical Chemistry , 75, 3019– 3030. Google Scholar CrossRef Search ADS PubMed  © The Author 2017. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com

Journal

Journal of Analytical ToxicologyOxford University Press

Published: Mar 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off