Comparison of Cozart microplate ELISA and GC-MS detection of methadone and metabolites in human hair.J Anal Toxicol. 2005 Oct; 29(7):678-81.JA
The purpose of this study was to determine the performance characteristics of the Cozart Methadone Microplate ELISA assay for the detection of methadone and methadone metabolites in hair specimens. One hundred and ten hair specimens were collected from volunteers (n=46) with a history of drug use and from drug-related deaths (n=64). The hair samples (approximately 20 mg) were extracted by sonication in methanol followed by overnight extraction in methanol at 60 degrees C. The methanol extract was evaporated to dryness, reconstituted in ELISA negative calibrator, and then analyzed. For gas chromatography-mass spectrometry (GC-MS) analysis, deuterated internal standard mixture [methadone-d9 and 2-ethylidene-1,5-dimethyl-3,3-diphenylpyrrolidine (EDDP)-d3] and 0.1M HCI were added to approximately 20 mg of specimen or spiked blank hair and sonicated for 1 h. The pH was adjusted to neutral, and methadone and its primary metabolite, EDDP, were analyzed by GC-MS following solid-phase extraction using Bond Elute Certify columns and pH 7.4 phosphate buffer (0.1M). Forty hair specimens were confirmed positive for methadone by GC-MS. Concentrations ranged from 0.10 to 8.3 ng/mg for methadone and 0.1 to 1.2 ng/mg for EDDP. The true positives, true negatives, false positives, and false negatives for different cutoffs with the ELISA were determined by comparison of the ELISA response (normalized to weight of hair extracted) to the GC-MS results with a cutoff of 0.1 ng/mg for both methadone and EDDP as the reference method. The optimum cutoff for the Cozart Methadone Microplate ELISA was determined to be between 200 and 300 pg methadone equivalents/mg hair using a 20 mg hair sample. The Cozart Methadone Microplate EIA for methadone and metabolites in hair using a cut-off of 200 pg/mg hair with a 20 mg hair sample had a sensitivity of 95 +/- 2% and a specificity of 100 +/- 3.5% (vs GC-MS) and an accuracy of 98.2 +/- 1.3%.