Determination of butorphanol in horse race urine by immunoassay and gas chromatography–mass spectrometry

An analytical procedure to screen butorphanol in horse race urine using ELISA kits and its confirmation by GC–MS is described. Urine samples (5 ml) were subjected to enzymatic hydrolysis and extracted by solid-phase extraction. The residues were then evaporated, derivatized and injected into the GC–MS system. The ELISA test (20 μl of sample) was able to detect butorphanol up to 104 h after the intramuscular administration of 8 mg of Torbugesic, and the GC–MS method detected the drug up to 24 h in FULL SCAN or 31 h in the SIM mode. Validation of the GC–MS method in the SIM mode using nalbuphine as internal standard included linearity studies (10–250 ng/ml), recovery (±100%), intra-assay (4.1–14.9%) and inter-assay (9.3–45.1%) precision, stability (10 days), limit of detection (10 ng/ml) and limit of quantitation (20 ng/ml).     

Application of solid-phase microextraction and gas chromatography–mass spectrometry for the determination of chlorophenols in urine

This study investigated the feasibility of applying solid-phase microextraction (SPME) combined with gas chromatography–mass spectrometry to analyze chlorophenols in urine. The SPME experimental procedures to extract chlorophenols in urine were optimized with a polar polyacrylate coated fiber at pH 1, extraction time for 50 min and desorption in GC injector at 290°C for 2 min. The linearity was obtained with a precision below 10% R.S.D. for the studied chlorophenols in a wide range from 0.1 to 100 μg/l. In addition, sample extraction by SPME was used to estimate the detection limits of chlorophenols in urine, with selected ion monitoring of GC–MS operated in the electron impact mode and negative chemical ionization mode. Detection limits were obtained at the low ng/l levels. The application of the methods to the determination of chlorophenols in real samples was tested by analyzing urine samples of sawmill workers. The chlorophenols were found in workers, the urinary concentration ranging from 0.02 μg/l (PCP) to 1.56 μg/l (2,4-DCP) depending on chlorophenols. The results show that trace chlorophenols have been detected with SPME–GC–MS in the workers of sawmill where chlorophenol-containing anti-stain agents had been previously used.     

Simultaneous determination of eight lipid peroxidation degradation products in urine of rats treated with carbon tetrachloride using gas chromatography with electron-capture detection

One of the major processes that occur as a result of radical-induced oxidative stress is lipid peroxidation (LPO). Degradation of lipid peroxides results in various products, including a variety of carbonyl compounds. In the present study eight different lipid degradation products, i.e., formaldehyde, acetaldehyde, acetone, propanal, butanal, pentanal, hexanal and malondialdehyde were identified and measured simultaneously and quantitatively in rat urine after derivatization with O-(2,3,4,5,6-pentafluorbenzyl)hydroxylamine hydrochloride, extraction with heptane and using gas chromatography–electron-capture detection (GC–ECD). The identity of the respective oximes in urine was confirmed by gas chromatography–negative ion chemical ionization mass spectrometry (GC–NCI-MS). Simultaneously measured standard curves were linear for all oxime-products and the detection limits were between 39.0±5.3 (n=9) and 500±23 (n=9) fmol per μl injected sample. Recoveries of all products from urine or water were 73.0±5.2% and higher. In urine of CCl₄-treated rats an increase in all eight lipid degradation products in urine was found 24 h following exposure. ACON showed the most distinct increase, followed by PROPA, BUTA and MDA. It is concluded that the rapid, selective and sensitive analytical method based on GC–ECD presented here is well suited for routine measurement of eight different lipid degradation products. These products appear to be useful as non-invasive biomarkers for in vivo oxidative stress induced in rats by CCl₄.     

Simultaneous determination of local anesthetics including ester-type anesthetics in human plasma and urine by gas chromatography–mass spectrometry with solid-phase extraction

The present study describes the simultaneous determination of seven different kinds of local anesthetics and one metabolite by GC–MS with solid-state extraction: Mepivacaine, propitocaine, lidocaine, procaine (an ester-type local anesthetics), cocaine, tetracaine (an ester-type local anesthetics), dibucaine (Dib) and monoethylglycinexylidide (a metabolite of lidocaine) were clearly separated from each other and simultaneously determined by GC–MS using a DB-1 open tubular column. Their recoveries ranged from 73–95% at the target concentrations of 1.00, 10.0 and 100 μg/ml in plasma, urine and water. Coefficients of variation of the recoveries ranged from 2.3–13.1% at these concentrations. The quantitation limits of the method were approximately 100 ng/ml for monoethylglycinexylidide, propitocaine, procaine, cocaine, tetracaine and dibucaine, and 50 ng/ml for lidocaine and mepivacaine. This method was applied to specimens of patients who had been treated with drip infusion of lidocaine, and revealed that simultaneous determination of lidocaine and monoethylglycinexylidide in the blood and urine was possible.     

Development, validation and application of assays to quantify metrifonate and 2,2-dichlorovinyl dimethylphosphate in human body fluids

Gas chromatographic procedures [GC with electron-capture detection (ECD) and GC–MS] for the quantitative analysis of metrifonate and its active metabolite 2,2-dichlorovinyl dimethylphosphate (DDVP) in human blood and urine were developed, validated, and applied to the analysis of clinical study samples. Analysis of metrifonate involved extraction of acidified blood with ethyl acetate followed by solid-phase clean-up of the organic extract. Acidified urine was extracted with dichloromethane and the residue of evaporated organic phase was reconstituted in toluene. ECD and diethyl analogue of metrifonate internal standard (I.S.) were used for quantitation of metrifonate. The metrifonate lower limit of quantitation (LOQ) was 10.0 μg/l. The DDVP metabolite was chromatographed separately after cyclohexane extraction of acidified blood and urine using d₆-DDVP I.S. and MS detection. The LOQ of DDVP was 1 μg/l. Stability studies have confirmed that the matrix should be acidified prior to storage at −20°C or −80°C to inhibit chemical and enzymatic degradation of the analytes and to avoid overestimation of DDVP concentrations. Metrifonate was found to be stable in acidified human blood after 20 months of storage at −20°C and after 23 months of storage at −80°C. Under these conditions DDVP was found to be stable after 12 months of storage. Both assay procedures were cross-validated by different world-wide laboratories and found to be accurate and robust during analyses of clinical study samples.     

High-temperature solid-phase microextraction procedure for the detection of drugs by gas chromatography–mass spectrometry

High-temperature headspace solid-phase microextraction (SPME) with simultaneous (“in situ”) derivatisation (acetylation or silylation) is a new sample preparation technique for the screening of illicit drugs in urine and for the confirmation analysis in serum by GC–MS. After extraction of urine with a small portion of an organic solvent mixture (e.g., 2 ml of hexane–ethyl acetate) at pH 9, the organic layer is separated and evaporated to dryness in a small headspace vial. A SPME-fiber (e.g., polyacrylate) doped with acetic anhydride–pyridine (for acetylation) is exposed to the vapour phase for 10 min at 200°C in a blockheater. The SPME fiber is then injected into the GC–MS for thermal desorption and analysis. After addition of perchloric acid and extraction with n-hexane to remove lipids, the serum can be analysed after adjusting to pH 9 as described for urine. Very clean extracts are obtained. The various drugs investigated could be detected and identified in urine by the total ion current technique at the following concentrations: amphetamines (200 μg/l), barbiturates (500 μg/l), benzodiazepines (100 μg/l), benzoylecgonine (150 μg/l), methadone (100 μg/l) and opiates (200 μg/l). In serum all drugs could be detected by the selected ion monitoring technique within their therapeutic range. As compared to liquid–liquid extraction only small amounts of organic solvent are needed and larger amounts of the pertinent analytes could be transferred to the GC column. In contrast to solid-phase extraction (SPE), the SPME-fiber is reusable several times (as there is no contamination by endogenous compounds). The method is time-saving and can be mechanised by the use of a dedicated autosampler.     

Determination of dichloroanilines in human urine by GC–MS, GC–MS–MS, and GC–ECD as markers of low-level pesticide exposure

Methods for the determination of 3,4-dichloroaniline (3,4-DCA) and 3,5-dichloroaniline (3,5-DCA) as common markers of eight non-persistent pesticides in human urine are presented. 3,5-DCA is a marker for the exposure to the fungicides vinclozolin, procymidone, iprodione, and chlozolinate. Furthermore the herbicides diuron, linuron, neburon, and propanil are covered using their common marker 3,4-DCA. The urine samples were treated by basic hydrolysis to degrade all pesticides, metabolites, and their conjugates containing the intact moieties completely to the corresponding dichloroanilines. After addition of the internal standard 4-chloro-2-methylaniline, simultaneous steam distillation extraction (SDE) followed by liquid–liquid extraction (LLE) was carried out to produce, concentrate and purify the dichloroaniline moieties. Gas chromatography (GC) with mass spectrometric (MS) and tandem mass spectrometric (MS–MS) detection and also detection with an electron-capture detector (ECD) after derivatisation with heptafluorobutyric anhydride (HFBA) were employed for separation, detection, and identification. Limit of detection of the GC–MS–MS and the GC–ECD methods was 0.03 and 0.05 μg/l, respectively. Absolute recoveries obtained from a urine sample spiked with the internal standard, 3,5-, and 3,4-DCA, ranged from 93 to 103% with 9–18% coefficient of variation. The three detection techniques were compared concerning their performance, expenditure and suitability for their application in human biomonitoring studies. The described procedure has been successfully applied for the determination of 3,4- and 3,5-DCA in the urine of non-occupationally exposed volunteers. The 3,4-DCA levels in these urine samples ranged between 0.13 and 0.34 μg/g creatinine or 0.11 and 0.56 μg/l, while those for 3,5-DCA were between 0.39 and 3.33 μg/g creatinine or 0.17 and 1.17 μg/l.     

Headspace solid-phase microextraction and gas chromatographic–mass spectrometric screening for volatile hydrocarbons in blood

Optimization for headspace solid-phase microextraction (SPME) was studied with a view to performing gas chromatographic–mass spectrometric (GC–MS) screening of volatile hydrocarbons (VHCs) in blood. Twenty hydrocarbons comprising aliphatic hydrocarbons ranging from n-hexane to n-tridecane, and aromatic hydrocarbons ranging from benzene to trimethylbenzenes were used in this study. This method can be used for examining a burned body to ascertain whether the victim had been alive or not when the burning incident took place. n-Hexane, n-heptane and benzene, the main indicators of gasoline components, were found as detectable peaks through the use of cryogenic oven trapping upon SPME injection into a GC–MS instrument. The optimal screening procedure was performed as follows. The analytes in the headspace of 0.2 g of blood mixed with 0.8 ml of water plus 0.2 μg of toluene-d₈ at −5°C were adsorbed to a 100-μm polydimethylsiloxane (PDMS) fiber for 30 min, and measured using the full-mass-scanning GC–MS method. The lower detection limits of all the compounds were 0.01 μg per 1 g of blood. Linearities (r²) within the range 0.01 to 4 μg per 1 g of blood were only obtained for the aromatic hydrocarbons at between 0.9638 (pseudocumene) and 0.9994 (toluene), but not for aliphatic hydrocarbons at between 0.9392 (n-tridecane) and 0.9935 (n-hexane). The coefficients of variation at 0.2 μg/g were less than 8.6% (n-undecane). In conclusion, this method is feasible for the screening of volatile hydrocarbons from blood in forensic medicine.     

Simple analysis of local anaesthetics in human blood using headspace solid-phase microextraction and gas chromatography–mass spectrometry–electron impact ionization selected ion monitoring

A simple method for analysis of five local anaesthetics in blood was developed using headspace solid-phase microextraction (HS-SPME) and gas chromatography–mass spectrometry–electron impact ionization selected ion monitoring (GC–MS–EI-SIM). Deuterated lidocaine (d₁₀-lidocaine) was synthesized and used as a desirable internal standard (I.S.). A vial containing a blood sample, 5 M sodium hydroxide and d₁₀-lidocaine (I.S.) was heated at 120°C. The extraction fiber of the SPME system was exposed for 45 min in the headspace of the vial. The compounds adsorbed on the fiber were desorbed by exposing the fiber in the injection port of a GC–MS system. The calibration curves showed linearity in the range of 0.1–20 μg/g for lidocaine and mepivacaine, 0.5–20 μg/g for bupivacaine and 1–20 μg/g for prilocaine in blood. No interfering substances were found, and the time for analysis was 65 min for one sample. In addition, this proposed method was applied to a medico–legal case where the cause of death was suspected to be acute local anaesthetics poisoning. Mepivacaine was detected in the left and right heart blood samples of the victim at concentrations of 18.6 and 15.8 μg/g, respectively.     

Measurement of azacyclonol in urine and serum of humans following terfenadine (Seldane) administration using gas chromatography—mass spectrometry

A gas chromatographic—mass spectrometric (GC—MS) method is presented for the analysis of azacyclonol (AZA), a metabolite of terfenadine in serum and urine specimens. Following an alkaline extraction, AZA and an internal standard were derivatized using heptafluorobutyric anhydride. Fourier transform infrared spectrometry suggested that two sites on the AZA molecule were derivatized. GC—MS of the extracts had a limit of quantitation (LOQ) of 1 ng/ml and linear range of 1–1000 ng/ml in urine. Four volunteers were administered a therapeutic regimen of terfenadine followed by urine and serum specimen collection(s) during the next seven days. The results indicated that following a 60-mg dose of terfenadine each 12 h for five days, (1) AZA appears in urine within 2 h, (2) urine AZA concentrations were above the LOQ 72 h following the last dose, (3) peak urine concentrations were as high as 19 000 ng/ml, and (4) mean serum concentration following the ninth dose was 59 ng/ml.     

Urinary excretion of 19-norandrosterone of endogenous origin in man: quantitative analysis by gas chromatography–mass spectrometry

A GC–MS method, using deuterium-labelled 19-noretiocholanolone as internal standard and following an extensive LC purification prior to selected ion monitoring of the bis(trimethylsilyl) ethers at ion masses m/z 405, 419, 420 and 421, allowed the quantitation of subnanogram amounts of 19-norandrosterone present in 10-ml urine samples at m/z 405. Thirty healthy men, free of anabolic androgen supply, delivered 24-h urine collections in 4 timed fractions. Accuracy was proven by the equation, relating added (0.05–1 ng/ml) to measured analyte, which had a slope not significantly different from 1. Precision (RSD) was 4% at a concentration of 0.4 ng/ml, and 14% at 0.04 ng/ml. Analytical recovery was 82%. The limit of quantitation was 0.02 ng/ml. The excretion ranges were 0.03–0.25 μg/24 h or 0.01–0.32 ng/ml in nonfractionated 24-h urine.Taking into account inter-individual variability and log-normal distribution, a threshold of 19-norandrosterone endogenous concentration of 2 ng/ml, calculated as the geometric mean plus 4 SD, was established. This value corresponds to the decision limit advised by sport authorities for declaring positive (anabolic) doping with nandrolone.     

Purge-and-trap gas chromatographic determination of styrene in urine and blood: Application to exposed workers

A simple purge-and-trap gas chromatographic method with flame ionization detection was developed for the determination of styrene in urine and blood. Styrene present in a 5 ml sample at room temperature was swept by helium at 40 ml/min for 11 min, trapped on a Tenax trap, desorbed by heating, cryofocused, and injected by flash heating into a DB-5 capillary GC column. The oven temperature program was from 80°C, held for 8 min, to 120°C at 5°C/min, and then held for 2 min. The detector temperature was 250°C. The calibration curves were linear in the range of 2.5–15 ppb styrene in urine and 25–150 ppb in blood. The detection limits calculated were 0.4 μg/l in urine and 0.6 μg/l in blood. The coefficients of variations within the day and day-to-day were 3 and 3.1%, respectively, for 2.5 ppb of styrene in urine, and 1 and 1.6% for 25 ppb of styrene in blood. The results obtained from samples taken from workers exposed to styrene were reported.     

A high-performance liquid chromatography assay with ultraviolet detection for olanzapine in human plasma and urine

Olanzapine is a commonly used atypical antipsychotic medication for which therapeutic drug monitoring has been proposed as clinically useful. A sensitive method was developed for the determination of olanzapine concentrations in plasma and urine by high-performance liquid chromatography with low-wavelength ultraviolet absorption detection (214 nm). A single-step liquid–liquid extraction procedure using heptane-iso-amyl alcohol (97.5:2.5 v/v) was employed to recover olanzapine and the internal standard (a 2-ethylated olanzapine derivative) from the biological matrices which were adjusted to pH 10 with 1 M carbonate buffer. Detector response was linear from 1–5000 ng (r²>0.98). The limit of detection of the assay (signal:noise=3:1) and the lower limit of quantitation were 0.75 ng and 1 ng/ml of olanzapine, respectively. Interday variation for olanzapine 50 ng/ml in plasma and urine was 5.2% and 7.1% (n=5), respectively, and 9.5 and 12.3% at 1 ng/ml (n=5). Intraday variation for olanzapine 50 ng/ml in plasma and urine was 8.1% and 9.6% (n=15), respectively, and 14.2 and 17.1% at 1 ng/ml (n=15). The recoveries of olanzapine (50 ng/ml) and the internal standard were 83±6 and 92±6% in plasma, respectively, and 79±7 and 89±7% in urine, respectively. Accuracy was 96% and 93% at 50 and 1 ng/ml, respectively. The applicability of the assay was demonstrated by determining plasma concentrations of olanzapine in a healthy male volunteer for 48 h following a single oral dose of 5 mg olanzapine. This method is suitable for studying olanzapine disposition in single or multiple-dose pharmacokinetic studies.     

Simultaneous determination of artemether and its major metabolite dihydroartemisinin in plasma by gas chromatography–mass spectrometry-selected ion monitoring

A sensitive, selective, and reproducible GC–MS–SIM method was developed for determination of artemether (ARM) and dihydroartemisinin (DHA) in plasma using artemisinin (ART) as internal standard. Solid phase extraction was performed using C₁₈ Bond Elut cartridges. The analysis was carried out using a HP-5MS 5% phenylmethylsiloxane capillary column. The recoveries of ARM, DHA and ART were 94.9±1.6%, 92.2±4.1% and 81.3±1.2%, respectively. The limit of quantification in plasma was 5 ng/ml (C.V.≤17.4% for ARM and 15.2% for DHA). Calibration curves were linear with R²≥0.988. Within day coefficients of variation were 3–10.4% for ARM and 7.7–14.5% for DHA. Between day coefficients of variations were 6.5–15.4% and 7.6–14.1% for ARM and DHA. The method is currently being used for pharmacokinetic studies. Preliminary data on pharmacokinetics showed C_max of 245.2 and 35.6 ng/ml reached at 2 and 3 h and AUC_0–8h of 2463.6 and 111.8 ngh/ml for ARM and DHA, respectively.     

Improved sample preparation method for selected persistent organochlorine pollutants in human serum using solid-phase disk extraction with gas chromatographic analysis

An improved solid-phase extraction (SPE) method was developed to isolate and concentrate trace levels of selected POPs (persistent organochlorine pollutants) in human serum prior to GC–MS in SIM mode or GC–ECD quantitation. The extraction involves denaturation of serum proteins with formic acid, SPE using C₁₈ Empore™ disk cartridges, followed by elimination of lipid interferences using a sulfuric acid wash of the eluate. Use of the SPE disk improved assay throughput and gave a cleaner analytical matrix compared with previously reported solid-phase and liquid–liquid extraction techniques. The extraction method provided consistent recoveries at three fortification levels using ¹³C₁₂ PCB 149 as internal standard. Recoveries ranged from 48 to 140% for organochlorine pesticides (6.25, 12.5 and 25 ng/ml) and 71 to 126% for polychlorinated biphenyls (0.625, 1.25 and 2.5 ng/ml).     

Determination of diphenylmethane antihistaminic drugs and their analogues in body fluids by gas chromatography with surface ionization detection

Eleven diphenylmethane antihistaminic drugs and their analogues were tested for their detection by capillary gas chromatography (GC) with surface ionization detection (SID). The GC—SID response was highest for doxylamine, diphenhydramine and orphenadrine and lowest for terodiline, clemastine and pipethanate. The detection limits for drugs with the highest response were 2–5 pg (ca. 6–20 fmol) on-column (100–250 pg/ml of body fluid). The detection limits with GC—SID were 10–100 times higher than those with GC with nitrogen—phosphorus detection. A detailed procedure for the isolation of the antihistaminics from human whole blood and urine by the use of Sep-Pak C₁₈ cartridges, prior to GC—SID, is also presented. The recoveries of the drugs (50 or 500 pmol), which had been added to 1 ml of body fluids, were>60%. The baselines remained steady as the column temperature was increased and the background was clean, especially for whole blood extracts.     

Sensitive determination of xylenes in whole blood by capillary gas chromatography with cryogenic trapping

A new and sensitive method for measurement of o-, m- and p-xylenes in human whole blood by capillary gas chromatography (GC) with cryogenic trapping is presented. After heating 0.5 ml of whole blood and 0.5 ml of distilled water containing the xylenes and aniline (internal standard, I.S.) in a 4.0-ml vial at 100°C for 30 min, 2 ml of the headspace vapor was drawn into a glass syringe. All vapor was introduced through the GC port into an AT-Wax middle-bore capillary column in the splitless mode at an oven temperature of 5°C to trap the entire analytes, and the oven temperature was then programmed up to 180°C. The present conditions gave sharp peaks for xylenes and aniline (I.S.), and low background noises for whole blood samples; the peaks of p- and m-xylenes showed about 90% separation with the AT-Wax column. As much as 41.0–46.3% of xylenes, which had been spiked to whole blood could be recovered. The calibration curves showed linearity in the range of 0.1–0.5 μg/0.5 ml of whole blood. The detection limit was estimated to be about 10 ng/0.5 ml. The coefficients of intra-day and inter-day variations for xylenes were not greater than 9.38%. The data for actual detection of xylenes in post-mortem blood of self-ignition suicide cases by the present method were also presented.     

Assay of 2-naphthol in human urine by high-performance liquid chromatography

This paper describes a novel liquid chromatographic method for the quantitation of 2-naphthol in human urine. Urine samples were extracted after enzymatic hydrolysis of glucuronides and sulfates; 2-naphthol was then separated using reversed-phase high-performance liquid chromatography. The corresponding detection limits were 0.04 ng/ml for the standard sample in acetonitrile and 0.13 ng/ml for urine samples. The level of urinary 2-naphthol in 100 Korean shipyard workers was analyzed using this new method. The level ranged from 0.21 ng/ml (0.26 μmol/mol creatinine) to 34.19 ng/ml (59.11 μmol/mol creatinine), and the mean±standard deviation was 5.08 ng/ml (6.60 μmol/mol creatinine)±5.75 ng/ml (9.22 μmol/mol creatinine). The mean±standard deviation of urinary 2-naphthol level of smokers, 7.03 ng/ml (8.49 μmol/mol creatinine)±6.16 ng/ml (10.23 μmol/mol creatinine), was significantly higher than that of non-smokers, 2.49 ng/ml (4.10 μmol/mol creatinine)±3.92 ng/ml (7.03 μmol/mol creatinine).     

Fluorodensitometric evaluation of gentamicin from plasma and urine by high-performance thin-layer chromatography

High-performance thin-layer chromatographic (HPTLC) analysis of gentamicin by in situ fluorodensitometric evaluation of its 4-chloro-7-nitrobenzo-2-oxa-1,3-diazole (NBD-Cl) derivative is presented. The aminoglycoside components separated on silica gel plates using chloroform–methanol–20% ammonium hydroxide (2.4:2.2:1.5, v/v/v) as the mobile phase were reacted with NB-Cl to yield highly fluorescent derivatives. The calibration curves of gentamicin in water, plasma and urine were linear in the range 40–200 ng. The mean values of intercept, slope and correlation coefficient were 16.82±0.473, 6.83±0.015 and 0.9968±0.0017 for standard curves in water, 17.35±0.375, 6.85±0.018 and 0.9941±0.0012 for standard curves in plasma and 14.35±0.286, 6.86±0.002 and 0.9933+0.0011 for standard curves in urine respectively. The analytical technique was validated for within-day and day-to-day variation. The results indicate that HPTLC, coupled with in situ fluorodensitometry, is a reliable and valuable technique for quantitative analysis of the bulk drug gentamicin and gentamicin from urine and plasma.     

Solid-phase microextraction with capillary gas-liquid chromatography and nitrogen-phosphorus selective detection for the assay of antidepressant drugs in human plasma

Solid-phase microextraction (SPME) was tested as a sample preparation for the simultaneous assay of ten antidepressant drugs and metabolites (TADs) in human plasma. Aqueous NaOH (0.5 ml, 1 M) and chloramitriptyline (50 μl, 40 μg/ml) as internal standard (I.S.) were added to a 2-ml plasma sample. This mixture was extracted with a 100-μm polydimethylsiloxane SPME fiber (Supelco) for 10 min. After washing in water and methanol (50%) and subsequent drying at room temperature, desorption of the fiber was performed in the injection port of a gas chromatograph at 260°C for 1 min (HP 5890, DB-17 30 m×0.25 mm I.D., 0.25 μm capillary; 0.7 ml/min nitrogen; nitrogen-phosphorus selective detection). The recovery was found to be very low from plasma (0.3% to 0.8%) but considerably higher from water (about 15%). Therefore, the high protein binding of antidepressants appears to be the main limiting mechanism for a better extraction. However, the analytes were well separated and the calibrations were linear between 125 ng/ml and 2000 ng/ml. The limits of quantification were about 90 ng/ml for imipramine and desipramine, 125 ng/ml for amitriptyline, trimipramine, doxepine, nortriptyline and mianserine and about 200 ng/ml for maprotiline, clomipramine and desmethylclomipramine. The recovery was improved by increasing the extraction time. The influence of the concentrations of the sum of proteins and of α-acid glycoprotein on the peak-area ratios A_TAD/A_I.S. and on absolute peak areas was studied. Peak-area ratios increased with decreasing protein concentration but were found to be independent on α-acid glycoprotein. A simple model for the explanation of the effect is presented. Measures for the improvement of sensitivity are discussed. As presented in this paper, which first describes SPME for the analysis of drugs in plasma, SPME with a short extraction time can be of only very limited value for therapeutic drug monitoring. Lower concentrations than the limit of quantification are usually found at therapeutic doses. The method can be useful for toxicological analysis after the accidental or suicidal intake of higher doses. However, an about 10-fold improvement of the sensitivity of the method seems to be possible.