Forensic Drug Testing For Opiates. V. Urine Testing for Heroin, Morphine, and Codeine With Commercial Opiate Immunoassays Edward J. Cone* and Sandra Dickerson P.O. Box 5180, Addiction Research Center, National Institute on Drug Abuse, Ba/timore, MD 21224 Buddha D. Paul Navy Drug Screening Laboratory, S-33, Nava/ Air Station, Norfo/k, VA 23511-6295 John M. Mitchell Navy Drug Screening Laboratory, H2033, Naval Air Station, Jacksonville, FL 32212 Abstract Urine specimens collected after heroin, morphine, and codeine administration were tested by four commercial opiate immunoaasays (TDx, CAC, ABUS, and EMIT) and by GC/MS. Quantitative immunoassay results (morphine equivalents) were compared with results by GC/MS for total morphine, free morphine, or total codeine. Mean detection times for the broadly cross-reacting immunoassays (TDx, ABUS, and EMIT, 300 ng/ml cutoff) ranged from 15-44 hours following heroin and morphine administration and 33-54 hours following codeine administration. Detection times obtained with CAC (25 ng/ml cutoff) tended to be somewhat shorter as a result of the high selectivity of the antibody for free morphine. High correlations over a wide concentration range were obtained for TDx, CAC, and ABUS versus GC/MS, with specimens collected after heroin and morphine administration. EMIT showed a high correlation over a narrow concentration range (0-1000 ng/ml) with heroin and morphine specimens, but responses plateaued at higher concentrations. There was substantial variability in immunoassay responses with specimens collected after codeine administration. Generally, this study demonstrated that immunoassay responses for opiate urine testing can be used as a semi-quantitative guide for GC/MS confirmation; however, the presence of codeine increased variability and diminished the accuracy of the immunoassay response. Introduction Urine testing for drugs of abuse generally consists of an initial screening test followed by confirmation by a more specific technology like gas chromatography/mass spectrometry (GC/MS). The Mandatory Guidelines for Federal Workplace Testing requires the use of an immunoassay as the initial test method (1), the purpose of which is to eliminate negative specimens prior to assay by GC/MS. According to the Food and Drug Administration, the immunoassay result is designed to provide only preliminary analytical test results (2). Specimens that assay negative "Author to whom correspondence should be addressed. receive no further testing, whereas positive specimens may be rescreened by the same or different procedure, and confirmed by GC/MS. In addition, many laboratories rely upon the magnitude of the initial screening result to indicate appropriate dilutions for confirmation testing. Accordingly, the accuracy of the immunoassay at the cutoff and at higher analyte concentrations impacts upon the final test result. Immunoassay testing for opiates requires the accurate determination of codeine and/or morphine in urine at amounts equal to or greater than 300 ng/ml for total drug concentration and equal to or greater than 25 ng/ml for free morphine concentration. Since heroin is metabolized by hydrolysis to morphine, and morphine and codeine are metabolized by oxidative and coupling mechanisms (3), several forms of morphine and codeine may be present in clinical specimens. In addition, metabolite ratios vary with time elapsed since drug administration. Consequently, validity assessment studies should be performed on sample sets that contain both spiked specimens and true clinical specimens. Preferably, the latter should be collected at known time intervals following controlled dosing. Immunoassay of specimens containing known concentrations of analytes provides information on sensitivity, specificity, and accuracy, whereas pharmacological parameters such as individual differences in absorption, metabolism, and excretion rates are revealed only in the analysis of clinical specimens. Recently, we reported an assessment of the analytical sensitivity, specificity and accuracy of four commercial urine opiate immunoassays (4). The study was performed with prepared specimens containing known drug concentrations. The present report details the performance of these assays with clinical specimens collected from subjects who received known doses of heroin, morphine, and codeine. The immunoassay results are compared to specific GC/MS assays for 6-acetylmorphine and free and total morphine and codeine. Materials and Methods Chemicals. Heroin hydrochloride was obtained from the Research Technology Branch, National Institute on Drug Abuse. 1 56 Reproduction (photocopying) of editorial content of this journal is prohibited without publisher's permission.
Table I. Immunoassay and GCIMS Analyses of Human Urine Collected After Administration of Heroin, Morphine, and Codeine* GC/MS Total Total 6-Acetyl- Dose Time TDx CAC ABUS EMIT Morphine Morphine Codeine Codeine morphine Drug (mg) Subject (h) (ng/ml) (ng/ml) (ng/ml) (ng/ml) (ng/ml) (ng/ml) (ng/ml) (ng/ml) (ng/ml) Heroin 6 H 0 <25 0.3 <10 <10 0-0 - 0 1.2" 6117 1193 6875 1826 QNS QNS QNS QNS QNS 4.1 6869 725 6904 1826 7999 769 0 0 50 7.3 4875 247 4865 1502 QNS QNS QNS QNS QNS 9.5 3441 161 3720 1384 QNS QNS QNS QNS QNS 24 911 56 825 761 1290 65 0 0 0 36 164 8.8 213 150 203 0 0 0 0 48 89 6.1 97 53 106t 0 0 0 0 60 57 3.5 24 18 64 0 0 0 0 72 38 2.8 12 <10 0-0 - 0 84 <25 0.3 <10 <10 0-0 - 0 96 <25 0.7 <10 <10 0 0 0 0 0 108 <25 0.6 <10 <10 0 0 0 0 0 120 <25 0.4 <10 <10 0-0 - 0 Heroin 6 K 0 <25 <0.3 <10 <10 0-0 - 0 1.1 1831 338 1802 1266 1621 255 0 0 98 4.5 3456 267 3627 1561 3316 296 0 0 15 7.2 1992 83 1831 1177 2153 124 0 0 0 11.8 818 25 631 596 785 38 0 0 0 24 438 18 374 434 536 27t 0 0 0 36 146 8.1 189 131 175 0 0 0 0 48 72 2.6 82 47 114 0 0 0 0 60 26 0.6 <10 <10 0 0 0 0 0 Heroin 6 L 0 <25 0.9 <10 <10 0-0 - 0 1.2 6502 1312 7700 1856 QNS QNS QNS QNS QNS 4.3 9881 1146 8525 1915 12022 1280 0 0 188 7.4 4545 258 4573 1531 5196 254 0 0 3.4 9.0 523 27 497 477 617 27 0 0 0 10.9 1197 39 818 783 1260 54 0 0 0 24 280 14 310 277 324 0 0 0 0 36 94 5.9 81 53 111 0 0 0 0 48 48 4.0 27 18 55t 0 0 0 0 60 35 1.6 14 <10 0 0 0 0 0 72 41 1.7 <10 <10 0-0 - 0 84 <25 0.7 <10 <10 0-0 - 0 Heroin 6 M 0 <25 <0.3 <10 <10 0-0 - 0 1.1 1423 249 984 1177 1342 244 0 0 51 3.3 7377 908 7607 1797 8721 841 0 0 30 7.1 1833 140 1876 1236 2314 126 0 0 0 7.8 93 5.5 111 65 131 0 0 0 0 8.7 111 8.7 149 89 168 0 0 0 0 10.8 330 16 304 366 448 19 0 0 0 24 655 28 557 552 705 41t 0 0 0 36 244 16 274 235 311 0 0 0 0 48 86 5.0 86 53 113 0 0 0 0 60 28 1.1 <10 <10 0-0 - 0 72 <25 0.6 <10 <10 0 0 0 0 0 Heroin 6 Q 0 <25 <0.3 <10 <10 0-0 - 0 2.8 6171 947 6279 1885 7322 1201 0 0 206 3.8 934 150 669 941 1433 149 0 0 11 6.6 1069 95 631 913 1234 107 0 0 1.5 7.4 947 67 794 783 1062 84 0 0 0 9.4 216 20 172 270 311 23 0 0 0 24 375 19 366 366 429 28 0 0 0 36 94 5.2 79 41 110 0 0 0 0 48 53 3.4 21 18 66t 15t 0 0 0 60 <25 0.3 <10 <10 0-0 - 0 *GC/MS data is included for comparison to immunoassay and has been published (4,5,6), Specimens were identified by GC/MS on the basis of retention time and ion ratios. 1" indicates those specimens with one ion ratio out of range ( 157
Table I. Immunoassay and GC/MS Analyses of Human Urine Collected After Administration of Heroin, Morphine, and Codeine* (Continued) Total Total 6-Acetyl- Dose Time TDx CAC ABUS EMIT Morphine Morphine Codeine Codeine morphine Drug (mg) Subject (h) (ng/ml) (ng/ml) (ng/ml) (ng/ml) (ng/ml) (ng/ml) (ng/ml) (ng/ml) (ng/ml) Heroin 6 W 0 11 <0.3 <10 <10 0-0 - 0 3.8 5122 1061 5523 1797 5828 1111 0 0 236 5.3 3056 322 2762 1413 3166 316 0 0 22 6.4 1634 96 1370 1089 1551 122 0 0 1.4 8 332 23 336 308 359 31t 0 0 0 10 958 37 760 761 1002 54 0 0 0 24 692 36 607 698 899 47 0 0 0 36 175 11 231 I62 236 16t 0 0 0 48 47 1.9 12 <10 47t 0 0 0 0 60 32 0.9 <10 <10 0 0 0 0 0 72 <25 0.5 <10 <10 0 0 0 0 0 Morphine 20 C 0 98 20 123 59 127 32 0 0 0 1.2 8050 2048 8878 1974 8006 2123 0 0 0 4.2 44112 5977 41868 2122 44843 5254 0 0 0 8.8 19171 1694 20535 1885 23010 1598 0 0 0 11.6 3819 224 3938 1531 4335 228 0 0 0 24 2756 144 2548 1384 2579 157 0 0 0 36 1527 120 1433 971 1640 100 0 0 0 48 574 40 554 501 627 40t 0 0 0 60 89 8.0 111 53 112 0 0 0 0 72 53 6.0 28 12 54t 0 0 0 0 84 41 2,8 <10 <10 0 0 0 0 0 96 <25 1.5 <10 <10 0-0 - 0 Morphine 20 E 0 <25 0.3 <10 <10 0-0 0 0.5 1228 476 1227 1030 1250 420 0 0 0 1.2 2518 702 2590 1472 1939 499 0 0 0 5.1 12229 1819 11186 2004 12220 1527 0 0 0 8.8 9179 1123 10083 1944 11801 944 0 0 0 24 2593 188 2375 1325 2689 156 0 0 0 36 536 33 488 539 631 35 0 0 0 48 165 9.4 197 150 199 0 0 0 0 60 53 1.7 17 <10 57 0 0 0 0 72 <25 1,6 <10 <10 0 0 0 84 <25 0,7 <10 <10 0 0 0 0 0 96 <25 0.5 <10 <10 0 0 0 0 0 Morphine 20 F 0 <25 <0.3 <10 <10 0 - - - 0 1.1 5740 1576 6009 1826 5928 1630 0 0 0 9.2 15244 1620 16418 1944 18078 1773 0 0 0 11.2 508 29 476 455 555 24 0 0 0 24 2810 129 2327 1295 2641 158 0 0 0 36 2005 84 1807 1089 2020 93 0 0 0 48 984 36 744 679 1110 52 0 0 0 60 222 14 274 221 278 16 0 0 0 72 131 8.8 158 95 161 0 0 0 0 84 59 3.5 35 18 64t 0 0 0 0 96 27 1.0 <10 <10 0 0 0 0 0 Morphine 20 0 0 <25 0.8 <10 <10 0-0 - 0 1.0 7495 2011 8347 1974 6615 1715 0 0 0 3.8 9699 1168 10129 1944 11481 1208 0 0 0 7.2 9228 715 9442 1885 13446 713 0 0 0 24 3237 165 3169 1413 3976 188 0 0 0 36 570 36 531 539 704 32 0 0 0 48 451 29 434 444 636 40 0 0 0 60 236 13 227 221 267 16t 0 0 0 72 125 12 152 125 136 0 0 0 0 84 69 4.8 21 24 74 0 0 0 0 96 58 4.3 26 12 61 0 0 0-108 29 1.8 <10 <10 0-0 - 0 GC/MS *GC/MS data is included for comparison to immunoassay and has been published (4,5,6). Specimens were identified by GC/MS on the basis of retention time and ion ratios. t Indicates those specimens with one ion ratio out of range (+20%). :~ Indicates those specimens with two ion ratios out of range. 158
Table I. Immunoassay and GC/MS Analyses of Human Urine Collected After Administration of Heroin, Morphine, and Codeine* (Continued) GC]MS Total Total 6-Acetyl- Dose Time TDx CAC ABUS EMIT Morphine Morphine Codeine Codeine morphine Drug (mg) Subject (h) (ng/ml) (ng/ml) (ng/ml) (ng/ml) (ng/ml) (ng/ml) (ng/ml) (ng/ml) (ng/ml) Codeine 120 N 0 <25 <0.3 <10 <10 0 0 0 0 0 1.6 36646 308 78259 2505 3341 149 34833 9827 0 2.8 12880 137 24829 2535 2653 83 20464 3099 0 3.2 7615 102 17561 2505 2255 80 16776 2274 0 4.3 8560 117 19454 2535 2786 80 18132 1976 0 8.1 18261 209 31181 2358 QNS QNS QNS QNS QNS 10.3 11462 95 16518 2417 4140 87 21882 1473 0 24 3397 81 5242 2240 2069 71 3571 290 0 36 702 30 855 1620 711 0 259 37 0 48 642 17 697 1325 505 0 166 34 0 60 231 9.6 346 513 209 0 89 12 0 72 137 4.6 156 181 90 0 671 13 0 84 63 2.0 22 41 0 0 0 0 0 96 49 1.0 15 24 0 0 0 131 0 108 28 0.5 <10 <10 0 0 0 0 0 Codeine 120 T 0 <25 <0.3 <10 <10 0 0 0 0 0 1.1 22078 57 49037 2446 446 54 27177 6747 0 3.8 41829 158 63155 2476 2129 79 95435 9202 0 6.2 19453 104 29264 2417 1753 55 61569 3313 0 9.8 12621 67 24446 2446 1601 39 49586 2011 0 24 3451 46 5543 2240 757 0 7815 579 0 36 977 30 1648 1944 471 0 1011 80 0 48 319 9.1 468 1325 178 0 233 23 0 60 91 2.0 44 169 581 0 64 0 0 72 61 2.6 25 71 0 0 0 0 0 84 47 0.5 12 18 0 0 401 0 0 96 28 <0.3 <10 <10 0 0 0 0 0 Codeine 120 EE 0 <25 <0.3 <10 <10 0 0 0 0-2.1 98539 145 214285 2381 2920 110 156783 47157-4.1 57484 120 38824 2387 3286 75 142778 17857-7.5 20189 69 42918 2215 1785 35 82788 5248-10.3 9128 31 15589 2353 1024 0 42816 2547-24 2307 21 3814 2240 333 16 5361 445-45.2 331 7.6 588 1746 88 0 587 52-48 371 2.5 669 1690 0 0 507 70-60 131 0.3 35 590 0 0 236 211-72 84 <0.3 17 106 0 0 102 0-84 64 0.6 14 93 0 0 91 0-96 52 1.0 11 49 0 0 71 0-108 58 <0.3 10 49 0 0 58 0-120 38 <0.3 <10 12 0 0 0 0-132 49 <0.3 12 18 0 0 0 0 - Codeine 120 FF 0 <25 0.3 <10 0 0 0 0 0 2.7 24481 130 54910 2417 1456 136 30852 10708 6.3 19318 137 33994 2381 2404 103 39781 5376 24 10005 90 22205 2269 2718 0 15061 1997 36 754 26 1013 1828 484 0 800 41 48 765 22 1284 1856 569 17 461 81 60 179 6.2 280 25 134 0 117 0 72 137 4.1 134 214 72 0 63 0 84 70 2.4 30 77 0 0 40 0 96 64 1.8 18 25 0 0 0 0 108 46 0.5 11 <10 0 0 0 0 120 41 1.1 13 <10 0 0 0 0 132 42 0.4 10 <10 0 0 0 0 *GC/MS data is included for comparison to immunoassay and has been published (4,5,6). Specimens were identified by GC/MS on the basis el retention time and ion rat{os. t Indicates these specimens with one ion ratio out of range (+20%). 159
Morphine sulfate was purchased from Mallinckrodt, and codeine phosphate was purchased from Merck and Co. All solvents and reagents were of reagent grade quality. Subject, dosing and urine collection. Details of subjects' characteristics and dosing and urine collection procedures for heroin, morphine, and codeine have been published (5-7). Immunoassays. All assays were performed on freshly thawed specimen sets. The specimens were assayed in random order under blind conditions. Results were decoded only after completion of the assay. Specimen sets were tested by immunoassay with the following commercial opiate assays: TDx Opiates (TDx; Abbott Laboratories), Coat-A-Count Morphine in Urine (CAC; Diagnostic Products Corporation), Abuscreen Radioimmunoassay for Morphine (ABUS; Roche Diagnostic Systems, Inc.), and Emit d.a.u. TM Opiate Assay (EMIT; Syva Company). Manufacturer's procedures were followed for all assays. EMIT data was obtained by analysis on a Hitachi 717 System and the delta absorbance units were converted to semi-quantitative data by means of a spline curve-fitting program (courtesy of Syva Company). All quantitative results were expressed as morphine equivalents. GC/MS assay. GC/MS assay for 6-acetylmorphine (6-AM) (8), and free and total morphine and codeine (9) were performed according to published procedures at the Navy Drug Screening Laboratory (Norfolk, VA). The limits of detection for 6-AM, free morphine, total morphine, free codeine, and total codeine were 0.81, 15, 40, 10, and 40 ng/ml, respectively. concentrations considerably below that observed for total morphine (5). Concentrations of other analytes declined rapidly during the first 12 h after heroin, morphine, and codeine administration and approached the level of sensitivity of the immunoassay within 72-120 h. There was general agreement between immunoassay results and GC/MS analyses. Concordance analysis (Table II) indicated that no false positive results (as defined by >300 ng/ml of morphine equivalents by immunoassay and <300 ng/ml of total morphine by GC/MS) were generated with heroin and morphine specimens by TDx, ABUS, or EMIT assay. Two false positives (as defined by >25 ng/ml of free morphine equivalents by CAC and <25 ng/ml of free morphine by GC/MS) were generated by CAC with heroin and morphine specimens. One sample, obtained from Subject H after administration of 3 mg of heroin (data not shown in Table I), contained 26 ng/ml of free morphine by CAC and 532 ng/ml of total morphine by GC/MS, but was negative for free morphine by GC/MS. The second sample was obtained from Subject F (Table I) and contained 29 ng/ml of free morphine by CAC and 24 ng/ml of free morphine and 555 ng/ml of total morphine by GC/MS. Generally, less than 5% of the heroin and morphine specimens tested as false negatives by each of the immunoassays. Table II. Concordance of Positive and Negative Results of Commercial Opiate Assays with Quantitative GC/MS Assay Results Assay (Cutoff, ng/ml) Result, Pos/Neg Heroin/Morphine Specimens (n = 207) Total Morphine (TM) by GC/MS (ng/ml) <150 TM 150-299 TM _>300 TM Acute opiate studies were performed with healthy male volunteer subjects with a history of heroin abuse. A primary goal of these studies was immunoassay testing of urine and other body fluids that were collected under controlled conditions after heroin, morphine, and codeine administration. The studies were conducted under the guidelines for the protection of human subjects, and each volunteer gave informed consent. During the studies, all subjects resided in a closed ward at the Addiction Research Center under medical surveillance. All subjects were drug-free prior to study participation. Heroin (3.0 and 6.0 mg), morphine (10 and 20 mg), and codeine (60 and 120 mg) were administered as single doses by the intramuscular route at weekly intervals. Doses were chosen to produce minimal to moderate euphorigenic effects. Urine specimens were collected prior to and following drug administration. Specimens were collected as 12-h pooled urines with the exception that individual specimens were collected during the first 12-h interval after dosing. Immunoassay testing was performed on all specimens according to manufacturers' specifications. GC/MS analysis for total and free morphine and codeine and for 6-acetylmorphine was performed for quantitative comparison. The results of the analyses are presented in Table I. Only data for the higher dose of each opiate are included. Prior to each dose, subjects' urine samples generally tested below the level of sensitivity of each assay, with the exception of Subject C who had residual carryover from a previous dose of morphine. Immediately after dosing, specimens tested highly positive for opiates by all assays. Generally, the first or second specimen contained the highest concentration of codeine or morphine. 6-Acetylmorphine was detected only during the first 8 h after heroin administration and at TDx (300) + 0 0 75-117 11 4 ABUS (300) + 0 0 74-117 11 5 EMIT (300) + 0 0 74-117 11 5 Assay (Cutoff, ng/ml) CAC (25) + Assay (Cutoff, ng/ml) Result, Pos/Neg Result, Pos/Neg Heroin/Morphine Specimens (n= 146) Free Morphine (FM) by GC/MS (ng/ml) <15 FM 15-24 FM _>25 FM 1 1 63 65 10 6 Codeine Specimens (n= 120) Total Codeine (TC) and Total Morphine (TM) by GC/MS (ng/ml) 150-299 TC, 150-299 TC, <150 TC <300 TM _>300TM _>300 TC TDx (300) + 0 3 3 50-61 2 0 1 ABUS (300) + 1 3 3 50-60 2 0 1 EMIT (300) + 4 5 3 51-57 0 0 0 Codeine Specimens (n= 120) Assay (Cutoff, Result, Free Morphine (FM) by GC/MS (ng/ml) ng/ml) Pos/Neg <15 FM 15-24 FM >_25 FM CAC (25) + 12 2 27-74 4 1 160
Concordance analysis of results from codeine specimens was more complex because of the simultaneous occurrence of codeine and morphine in some specimens. Categories were selected in Table II that contained different combinations of total codeine and total morphine by GC/MS. Positive specimens were identified by GC/MS which contained either >300 ng/ml of total codeine or >300 ng/ml of total morphine. Negative specimens were identified by 150-299 ng/ml of total codeine and <300 ng/ml of total morphine. Some specimens in this category contained a total opiate content >300 ng/ml (total codeine plus total morphine). A final category of negative specimens by GC/MS was identified which contained <150 ng/ml of total codeine. None of the specimens in this latter category contained total opiate content _>300 ng/ml. There were 0, 1, and 4 false positives (as defined by _>300 ng/ml of morphine equivalents by immunoassay and <150 ng/ml of total codeine by GC/MS) for TDx, ABUS, and EMIT, respectively. There were additional false positives characterized as specimens which tested positive by immunoassay and contained total codeine in the range of 150-299 ng/ml in combination with <300 ng/ml of total morphine. There were 3, 3, and 5 false positives for TDx, ABUS, and EMIT, respectively, in this category. Three of these specimens contained a combined total of total codeine and total morphine exceeding 300 ng/ml. There were 14 false positives (as defined by _>25 ng/ml of free morphine equivalents by CAC and <25 ng/ml of free morphine by GC/MS). Generally, less than 1% of Table III. Detection Times (time to detection of last positive at specified cutoff) by GCIMS and Immunoassay for Opiates Following Administration of Heroin Hydrochloride, Morphine Sulfate, and Codeine Phosphate to Male Volunteer Subjects GC/MS Total Free Total Free Dose Morphine Morphine Codeine Codeine TDx CAC ABUS EMIT Drug (mg) Subject (300ng/mL) (25 ng/ml) (300 ng/ml) (25ng/mL) (300ng/mL) (25 ng/ml) (300ng/mL) (300ng/mL) Heroin 3 H 24.0 11.5 - - 24.0 24.0 24.0 24.0 K 9.3 9.3 - - 9.3 5.9 9.3 9.3 L 11.8 9.8 - - 11.8 9.8 5.5 11.8 M 24.0 6.2 - - 24.0 7.9 24.0 7.9 Q 10.2 10.2 - - 10.2 10.2 10.2 10.2 W 24.0 24.0 - - 24.0 10.0 24.0 24.0 Mean. SE 17.2+3.1 11.8 2.5 - - 17.2 3.1 11.3 16.2 14.5+3.0 6 H 24.0 24.0 - - 24.0 24.0 24.0 24.0 K 24.0 24.0 - - 24.0 7.2 24.0 24.0 L 24.0 10.9 - - 10.9 10.9 24.0 10.9 M 36.0 24.0 - - 24.0 24.0 24.0 24.0 Q 24.0 24.0 - - 24.0 7.4 24.0 24.0 W 24.0 24.0 - - 24.0 24.0 24.0 24.0 Mean +SE 26.0 21.8+2.2 - - 21.8+2.2 16.3 240+0.0 21.8+2.2 Morphine 10 A ND* ND - - 36.0 24.0 36.0 36.0 B ND ND - - 36.0 24.0 36.0 36.0 C ND ND - 36.0 36.0 36,0 36.0 E ND ND - - 12.0 12.0 12.0 12,0 F ND ND - - 48.0 24.0 48.0 48,0 0 ND ND - - 36.0 36.0 36.0 36.0 Mean ND ND - - 34.0 26.0 34.0_+4.8 34.0 +4.8 20 A ND ND - - 48.0 36.0 48.0 48.0 B ND ND - - 36.0 36.0 36.0 36.0 C 48.0 48.0 - - 48.0 48.0 48.0 48.0 E 36.0 36.0 - - 36.0 36.0 36.0 36.0 F 48.0 48.0 - - 48.0 48.0 48.0 48.0 0 48.0 48.0-48.0 48.0 48.0 48.0 Mean+SE 45.0_+ 3.0 45.0-44.0 42.0 44.0+2.5 44.0+2.5 Codeine 60 N 36.0 24.0 24.0 24.0 36.0 24.0 36.0 48,0 T 24.0 1.2 24.0 24.0 36.0 6.6 360 48.0 EE 8.0 3.1 36.0 48.0 24.0 8.0 24.0 48.0 FF 24.0 24.0 24.0 36.0 36.0 24.0 36.0 48.0 Mean+SE 23.0 13.1 +6.3 27.0+3.0 33.0_+5.7 33.0+3.0 15.7+4.8 33.0+3.0 48.0_+0.0 *ND = Not determined, 120 N 48.0 24.0 24.0 48.0 48.0 36.0 60.0 60.0 T 36.0 9.8 36.0 36.0 48.0 36.0 48.0 48.0 EE 24.0 7.5 48.0 48.0 48.0 10.3 48.0 60.0 FF 48.0 6.3 48.0 48.0 48.0 36.0 48.0 48.0 Mean +SE 39.0 11.9_+4.1 39.0_+5.7 45.0 48.0 29.6_+6.4 51.0_+3.0 54.0 +3.5 161
the codeine specimens tested as false negatives by each of the immunoassays. Detection times (time to detection of last positive at specified.-i I- Z uj,,j < > 5 o w LU Z n 45000" 30000 15000 45000 84 30000 15000 0-0 r : 0. ~ f 9,. 0 15000 30000 45000 TOTAL MORPHINE r = O. ~ ~ f9,, 15000 30000 45000 TOTAL MORPHINE cutoffs) were calculated for each assay and are listed in Table IlL Because of differences in immunoassay selectivity and in relative drug potency, detection times were longest for codeine, followed by morphine and then by heroin. Amongst ff the immunoassays, CAC demonstrated the shortest detection times. This was due to the s000 B. CAC (N = 149) high selectivity of the CAC antibody for free 4000 r=0.904/y = 1.077x - 8.3 morphine. The detection time of CAC (25 = / ~ ng/ml cutoff) generally matched those obtained by GC/MS for free morphine (25 2000 ng/ml cutoff). The detection times of TDx, ABUS, and EMIT (300 ng/ml cutoffs) were similar and matched those obtained by ~ 2000 4000 0ooo GC/MS for total morphine (300 ng/ml FREE MORPHINE cutoff). Quantitative comparisons of immunoassay response versus GC/MS assay of target ana- 2ooo D. EMIT (N = 194) lytes were made by least squares linear regression. Responses were linear across a wide 15oo y : 0.602x 6.2.. concentration range for TDx, CAC, and r = 0.973 ~ ABUS following heroin and morphine ad- 1000 9 fi'/" 9 ministration (Figure 1A-C). In contrast, ~, EMIT was linear only over a range of 0-1000 9 9 ng/ml (Figure ID). At concentrations above so0, 1000 ng/ml, EMIT plateaued and produced 0"~".... no further increase in response. Following 0 500 1000 1500 2000 codeine administration, TDx, ABUS, and TOTAL MORPHINE Figure 1. Correlations of immunoassay responses (morphine equivalents) with GC/MS assay of total (or free) morphine for clinical specimens collected following administration of heroin (3 mg and 6 rag, intramuscular) and morphine (20 mg, intramuscular)9 For EMIT, only specimens that quantitated <1000 ng/ml (morphine equivalents) by immunoassay were included in the correlation. 200000 ~: 100000 z I-- Z _J o A. TDx (N = 119) y = 0.474x + 225.3 r = 0.039 400 300 200 100 0 B. CAC (N = 119) :,./ y : 1.344x + 6.4 r = 0.913 100000 200000 100 200 300 TOTAL CODEINE FREE MORPHINE 400 EMIT showed considerable variability when plotted vs total codeine by GC/MS (Figure 2A, 2C, and 2D); similar variability was seen in CAC versus free morphine by GC/MS (Figure 2B). Based on linear regression analyses, estimates were made of the amount of GC/MS analyte needed to test positive by immunoassay at least 50% of the time (i.e., the cutoff concentration). These are listed in Table IV. For the heroin and morphine specimens, slightly higher amounts of free or total morphine were needed by GC/MS to match the immunoassay cutoffs. For codeine specimens, less opiate was required to match CAC and EMIT cutoffs. Calculations were not performed for TDx and ABUS as a result of high intercept values obtained with codeine specimens. o w l,u z -To,. 300OO0 200000 100000 1000 " C. ABUS (N = 119) D. EMIT (N = 119) :4J ~ y = 0.820x + 86 600 9 II rty :j ~, 9 9. 2oo.,~ 400 ' y = 2.7~x 0.8 r = 0.822 9 100()00 200000 300000 0 200" 4()0 6()0 8()0 I000 TOTAL CODEINE TOTAL CODEINE Figure 2. Correlations of immunoassay responses (morphine equivalents) with GC/MS assay of total codeine or free morphine for clinical specimens collected following administration of codeine (60 mg and 120 mg, intramuscular)9 For EMIT, only specimens that quantitated <1000 ng/ml (morphine equivalents) by immunoassay were included in the correlation. Discussion The validity of a test method resides in its ability to detect parent drug and/or metabolite in biological fluids after human drug administration (10). This definition of validity encompasses numerous factors which impact upon an assay: pharmacologic variables such as dose, route of administration and intersubject variability; and chemical factors such as sensitivity, specificity, and accuracy. In light of current forensic standards, it has been proposed to expand the definition of assay 162
validity to encompass a requirement for confirmation by an alternate analytical method (e.g., GC/MS) (I I). Consequently, immunoassays should not only detect drugs in a sensitive and selective fashion, but also quantitate analytes in a concentration range acceptable for GC/MS confirmation. In the present study, four commercial opiate screening immunoassays were evaluated for their ability to accurately detect drug exposure after administration of known doses of heroin, morphine, and codeine. Those assays with broadly cross-reacting antibodies (TDx, ABUS, and EMIT) were found to be exceptionally accurate for the detection of opiate use following heroin, morphine, and codeine administration. Their false negative rate was consistently low (<5%) and there were no false positives identified after heroin or morphine administration. Some codeine specimens were identified as false positives by TDX, ABUS, and EMIT as a result of the presence of multiple codeine metabolites. In addition, it has been shown that EMIT demonstrates even greater cross-reactivity with codeine than with morphine (4); hence, the EMIT response was even further exaggerated in the presence of codeine. This exaggerated response can be seen in Figure 2D, in which there is an almost three-fold increase in the immunoassay response versus the GC/MS response (slope = 2.758). The CAC immunoassay is based on detection of free morphine and demonstrates very low cross-reactivity to morphine metabolites and codeine (4). This assay was generally accurate for detection of heroin or morphine exposure due to the presence of free morphine excreted in urine after morphine or heroin administration (5,6). Only two false positives (as defined by >25 ng/ml of free morphine equivalents by CAC and <25 ng/ml of free morphine by GC/MS) were obtained with heroin and morphine specimens. Both of these samples contained >300 ng/ml of total morphine by GC/MS. However, CAC testing of codeine specimens was less accurate. Fourteen false positives (as defined by >25 ng/ml of free morphine equivalents by CAC and <25 ng/ml of free morphine by GC/MS) were identified by this assay. It should be noted that these specimens contained either total morphine or total codeine >300 ng/ml; therefore, they would not be designated as false positives if comparisons were made with total opiates. The detection time (time from administration of drug to detection of last positive specimen) of an assay is an important component in validity studies. Early studies on screening method- Table IV. Equivalent GC/MS Opiate Concentrations at the Immunoassay Cutoff* Heroin/Morphine Codeine Immuno- Specimens Specimens assay Equiv. GC/MS Equiv. GC/MS Assay Cutoff Opiate Conc.t Deviationt Opiate Conc.w Deviations TDx 300 350.8 16.9% NA n NA CAC 25 31.0 24.0% 13.8-44.8% ABUS 300 360.6 20.2% NA NA EMIT 300 424.6 41.5% 108.5-63.8% 9 Estimates are based on the linear correlations of Figures 1 and 2. 1" For TDx, ABUS, and EMIT, concentrations are expressed as ng/ml of total morphine, and for CAC as ng/ml of free morphine. :t Percent deviation of the GC/MS concentration from the immunoassay cutoff concentration. w For EMIT, concentration is expressed as ng/ml of total codeine, and for CAC as ng/ml of free morphine. # NA = not applicable. Estimates were not made due to the high intercept values obtained in the linear correlation calculations of Figures 1 and 2. ology for opiates involving both immunologic and nonimmunologic tests (10,12) produced a variety of detection times ranging from 8-80 hours following administration of morphine or heroin. The present comparison of the detection times of TDx, ABUS, and EMIT were similar, with detection times ranging from 15-44 hours following single doses of heroin or morphine and 33-54 hours following single doses of codeine. Both pharmacologic factors and antibody characteristics are likely to account for the differences between heroin and morphine detection times. Codeine has a longer excretion half-life (T1/2 = ca. 12 h) than morphine (TI/2 = ca. 8 h) (7), as well as lower potency. Furthermore, all three immunoassays have been shown to be more cross-reactive with codeine than morphine (4). CAC's detection times tended to be shorter than those of TDx, ABUS, and EMIT, likely as a result of antibody characteristics. This assay was designed to cross-react only with free morphine; and free morphine has a slightly shorter half-life (T1/2 = ca. 4-6 h) than conjugated morphine (TL/2 = ca. 8 h) (5-7). As a resuit, CAC's detection time is slightly diminished relative to that observed for the broadly reacting antibody assays. Finally, quantitative comparisons of immunoassays versus GC/MS are needed because the immunoassay result is often used as a semiquantitative guide for confirmation. The immunoassay responses following heroin and morphine administrations were demonstrated to be generally accurate for TDx, CAC, and ABUS versus GC/MS over a wide concentration range (Figure 1A, 1B, and 1C). The EMIT response demonstrated a limited useful range, only up to 1000 ng/ml; thereafter, a plateau occurred, and further increases in concentration were not detected by EMIT. A second complication in correlating immunoassay responses to GC/MS was demonstrated with codeine specimens. The variability in responses with these specimens was considerably greater as indicated by a comparison of Figures 1 and 2. The variability observed with the codeine specimens is undoubtedly due to the presence of multiple metabolites with different cross-reactivities for each immunoassay antibody. Thus, although immunoassay responses with specimens collected after heroin and morphine administration correlate highly with GC/MS responses, the presence of codeine may substantially alter the reliability of the immunoassay response. Acknowledgments The authors wish to thank the following companies for generous supply of reagents and equipment: Abbott Laboratories, Diagnostic Products Corp., and Roche Diagnostics Systems. References 1. Mandatory guidelines for federal workplace drug testing programs; Final guidelines; Notice. Fed. Regist. 53(69): 11970-89 (April 11,1988). 2. K. Aziz. Drugs-of-abuse testing; Screening and confirmation. Clin. Lab. Med. 10: 493-502 (1990). 3. L. Lemberger and A. Rubin. In Physiologic Disposition of Drugs of Abuse, Vol. I., E.S. Vesell and S. Garattini, Eds., Spectrum Publications, Inc., New York, 1976, pp. 125-61. 4. E.J. Cone, S. Dickerson, B.D. Paul, and J.M. Mitchell. Forensic 163
drug testing for opiates. IV. Analytical sensitivity, specificity, and accuracy of commercial urine opiate immunoassays. J. Anal ToxicoL 16:72-78 (1992). 5. E.J. Cone, P. Welch, J.M. Mitchell, and B.D. Paul. Forensic drug testing for opiates. I. Detection of 6-acetylmorphine in urine as an indicator of recent heroin exposure; Drug and assay considerations and detection times. J. Anal Toxicol. 15:1-7 (1991). 6. J.M. Mitchell, B.D. Paul, P. Welch, and E.J. Cone. Forensic drug testing for opiates. I1. Metabolism and excretion rate of morphine in humans after morphine administration. J. Anal ToxicoL 15:49-53 (1991 ). 7. E.J. Cone, P. Welch, J.M. Mitchell, and B.D. Paul. Forensic drug testing for opiates. III. Urinary excretion rates of morphine and codeine following codeine administration. J. Anal. ToxicoL 15: 161-66 (1991). 8. B.D. Paul, J.M. Mitchell, L.D. Mell, Jr., and J. Irving. Gas chromatography/electron impact mass fragmentometric determina- tion of urinary 6-acetylmorphine, a metabolite of heroin. J. Anal Toxicol. 13:2-7 (1989). 9. B.D. Paul, L.D. Mell, JM. Mitchell, J. Irving, and A.J. Novak. Simultaneous identification and quantitation of codeine and morphine in urine by capillary gas chromatography and mass spectroscopy. J. Anal ToxicoL 9:222-26 (1985). 10. C.W. Gorodetzky. Detection of drugs of abuse in biological fluids. In Handbook of Experimental Pharmacology, G.V.R. Born, O. Eichler, A. Farah, H. Herken, and A.D. Welch, Eds., Springer-Verlag, Berlin, 1977, Vol. 45, pp. 319-409. 11. E.J. Cone, S.L. Menchen, and J. Mitchell. Validity testing of the TDX cocaine metabolite assay with human specimens obtained after intravenous cocaine administration. Forensic ScL Int. 37:265-75 (1988). 12. D.H. Catlin. Urine testing: A comparison of five current methods for detecting morphine. Amer. J. Clin. Path. 60:719-28 (1973). Manuscript received June 29, 1992. 164