Journal of Analytical Toxicology, Vol.

34, July/August 2010

Vitreous Humor as an Alternative Matrix for Comprehensive Drug Screening in Postmortem Toxicology by Liquid Chromatography Time-of-Flight Mass Spectrometry
Anna Pelander1, Johanna Ristimaa1,2, and Ilkka Ojanper1
1Department 2Laboratory

of Forensic Medicine, P.O. Box 40, FI-00014 University of Helsinki, Helsinki, Finland and of Analytical Chemistry, Department of Chemistry, P.O. Box 55, FI-00014 University of Helsinki, Helsinki, Finland

Abstract
Liquid chromatographytime-of-flight mass spectrometry (LCTOFMS) is an emerging technique in toxicological drug screening. The applicability of vitreous humor as an alternative matrix for drug screening by LCTOFMS in postmortem investigations was studied by comparison of findings in urine and vitreous humor in 50 autopsy cases. In addition, a cutoff value was determined for 70 drugs of toxicological relevance. The comparison study showed that vitreous humor is well-suited for qualitative screening analysis, although representativeness was not as good as urine because of less frequent metabolite detection. A total of 45 parent compounds and 24 metabolites were identified in vitreous samples; for urine, the figures were 55 and 39, respectively. The mean and median cutoff values were 0.072 and 0.023 mg/L, respectively, which are sufficient for routine casework according to the quantitative data available in the literature.

Introduction
Vitreous humor has been traditionally utilized in alcohol and glucose/lactate analyses in postmortem toxicology. There has been less interest in using it as an alternative matrix for drug analysis as the interpretation of quantitative results is complicated because of the blood-retinal barrier (1,2). Papers discussing vitreous drug analysis during the last decade have used target analysis approaches for specific drug groups, including amphetamines (3,4), benzodiazepines (5,6), cocaine and its metabolites (710), opiates (10), LSD and metabolites (11), cannabinoids (12), and single substance analyses (12 14). The applicability of vitreous use for immunological drugsof-abuse testing has also been evaluated (1517). Qualitative screening analysis is a key step in postmortem forensic toxicological investigation because the ability to identify unknown agents determines the scope of the examination. Urine is the primary sample material because the time window

for detection of exogenous compounds is wider in urine than in most other matrices. However, urine is not always available; in the authors laboratory, for instance, urine samples are unavailable in approximately 15% of cases. Liver has been the traditional secondary choice, yet the range of analytes detectable in liver is somewhat limited to lipophilic drugs (18), and tissue samples require complicated sample pretreatment, which results in increased time and cost. Vitreous humor, on the other hand, has a liquid character and resists putrefaction, which make it an attractive material for qualitative screening analysis in cases where urine is not available. In addition, because the analytes are carried from the blood stream by active and passive transportation via the blood-retinal barrier, vitreous humor is closer to blood than urine as a sample material, which makes it a suitable parallel sample material for blood. Liquid chromatographytime-of-flight mass spectrometry (LCTOFMS) has been used in comprehensive drug-screening procedures in the authors laboratory on a routine basis since 2004. The applications include urine drug screening (19,20), urine doping agent screening (21), and hair drug screening (22). In this study, the suitability of vitreous humor as an alternative matrix for drug-screening purposes was studied. The representativeness of vitreous was evaluated by comparing findings in urine and vitreous, and a cutoff value was determined for 70 compounds with toxicological relevance.

Experimental
Materials

All solvents and reagents were of analytical reagent grade and from Merck (Darmstadt, Germany), except for sodium hydroxide (J.T. Baker, Deventer, the Netherlands) and high-performance liquid chromatography-grade acetonitrile (Rathburn, Walkerburn, U.K.). Water was DirectQ-3 purified (Millipore, Bedford, MA). Internal standard dibenzepin was of pharma-

312

Reproduction (photocopying) of editorial content of this journal is prohibited without publishers permission.

Journal of Analytical Toxicology, Vol. 34, July/August 2010

ceutical purity. Isolute HCX-5 (130 mg) mixed-mode solidphase extraction (SPE) cartridges were from International Sorbent Technologies (Hengoed, U.K.).
Sample collection

by infusing 10 mM sodium hydroxide in isopropanol/0.2% formic acid (1:1, v/v). The post-run calibration of the individual data sets was enabled by injecting the calibrant at the beginning of each run via the six-port valve. The automated calibration cycle was included in the software (see Data analysis).
Data analysis

Vitreous samples were collected in medicolegal autopsies with a syringe and a needle. The sample was stored in a plastic tube with 100 mg sodium fluoride added. A pooled blank vitreous matrix was used for cutoff value determination.
Sample preparation

An SPE procedure was performed as described earlier for urine samples but excluding the hydrolysis step (19). Dibenzepin internal standard solution (10 L, 10 g/mL) and 2 mL pH 6 phosphate buffer were added to 1 mL vitreous sample. The SPE cartridges were conditioned with 2 mL methanol, followed by 2 mL water and 3 mL pH 6 phosphate buffer. The sample was added, followed by washing with 1 mL pH 6 phosphate buffer and drying for 5 min. The cartridge was further washed with 1 mL 1 M acetic acid and again dried for 5 min. The acidic/neutral fraction was eluted with 3 mL of ethyl acetate/hexane (25:75, v/v), and the cartridge was dried for 2 min. The cartridge was washed with 1 mL methanol and dried for 2 min. The basic fraction was eluted with 3 mL freshly made ethyl acetate/ammonia (98:2, v/v), 25% ammonia solution in water. The eluates (acidic-neutral fraction and basic fraction) were combined and evaporated to dryness at 40C, reconstituted with 150 L acetonitrile/0.1% formic acid (1:9, v/v), and taken to LCTOFMS analysis.
LCTOFMS

The LCTOFMS analysis was performed as previously described (20) with minor modifications. The LC was an Agilent 1100 series instrument including a vacuum degasser, autosampler, binary pump, and column oven (Waldbronn, Germany). Separation was performed in gradient mode with a Phenomenex Luna C18 (2) (100 2 mm, 3 m) column and a 4 2-mm pre-column at 40C (Torrance, CA). Mobile phase components were 5 mM ammonium acetate in 0.1% formic acid and acetonitrile. The flow rate was 0.3 mL/min. The proportion of acetonitrile was increased from 10% to 40% in 10 min, to 75% in 13.5 min, to 80% in 16 min, and held at 80% for 6 min. The post-time was 6 min and injection volume 10 L. The mass analyzer was a Bruker Daltonics MicrOTOF MS with an electrospray ionization (ESI) source and a six-port divert valve (Bremen, Germany). The instrument controls were performed with HyStar 3.1 and micrOTOF Control 1.1 (Bruker Daltonics) software. The nominal resolution of the instrument was 10,000. The instrument was operated in positive ion mode at m/z 50800. The capillary voltage was 4500 V, and the capillary exit was 85 V. The nebulizer gas pressure was 1.6 bar, and the drying gas flow was 8 L/min. The drying temperature was 200C. The spectra average was set to 2, and the summation was 12,000, corresponding to 0.6 s sample time. The transfer time was 40 s, and the hexapole RF was 45 Vpp. A six-point external instrument mass-scale calibration was performed prior to each sequence with sodium formate clusters

For the data analysis, TargetAnalysis 1.1 and DataAnalysis 3.4 (Bruker Daltonics) software were used. The in-house mass database included 815 monoisotopic exact masses comprising a wide chemical variety of drugs of abuse, therapeutic drugs and their metabolites, and designer drugs. Retention time information was available for approximately half of these compounds. TargetAnalysis software performed the mass-scale calibration of the data, created extracted ion chromatograms (EIC) in a m/z 0.003 window for the protonated molecule of each molecular formula included in the in-house mass database, applied peak detection and identification criteria according to mass accuracy, isotopic pattern, area, and retention time if available, and finally created an MS Excel-based result report. Isotopic pattern was expressed as SigmaFit, which is a numerical match value calculated for the deviation of the average mass spectrum of the created EIC peak compared to the theoretical isotopic pattern of the particular molecular formula. The three identification criteria had two levels: 5 and 8 ppm for mass accuracy, 0.03 and 0.05 for SigmaFit, and 0.1 and 0.3 min for retention time. Based on these criteria, a score value for the reliability of the identification was calculated for compounds with a retention time available. The highest score, 3 (identified), required identification with better values than 5 ppm mass accuracy, 0.03 SigmaFit, and 0.1 min retention time deviation. If any of the values were in the secondary level (58 ppm, 0.03 0.05 SigmaFit, 0.10.3 min retention time) the score was 2 (possibly identified). Score 1 (tentatively identified) was given to compounds without a reference retention time. The wider identification criteria windows were applied at score 1 level.
Evaluation of the applicability of vitreous as a sample matrix

The representativeness of vitreous for drug-screening purposes was evaluated by analyzing autopsy urine and vitreous samples in parallel and comparing the number of findings in the two matrices. A total of 50 successive cases representing the routine workflow of the laboratory were selected for the study based on the availability of both urine and vitreous humor. The urine samples were analyzed with a previously described screening method based on SPE and LCTOFMS (20), applying identification criteria described in the Data analysis section.
Determination of cutoff values

Cutoff values were determined for 70 drugs with a pooled blank vitreous sample spiked with selected drugs. The previously determined cutoffs for urine were used as starting concentrations (19). Methanolic solutions of the reference substances (10 drugs in each sample) were evaporated to dryness prior to blank vitreous addition. The concentration was either increased or decreased until identification at the reporting criteria area (50,000) was achieved. Three replicates were analyzed at the cutoff level.

313

Journal of Analytical Toxicology, Vol. 34, July/August 2010

Table I. Combined Data for Occurrence of Drugs and Metabolites in Urine and Vitreous Humor of 50 Autopsy Cases
Number of Findings in Urine 1 1 1 4 1 1 3 2 1 7 37 2 2 2 0 1 6 28 1 5 1 4 4 2 7 1 2 1 1 2 2 1 1 1 1 1 1 2 3 3 4 0 22 2 4 2 1 0 1 Number of Findings in Vitreous 1 0 0 3 0 1 4 2 0 7 34 2 2 2 1 1 6 29 1 3 0 1 3 2 7 1 2 1 1 0 1 0 0 0 1 0 0 0 1 3 2 1 12 0 0 0 1 1 1 Number of Findings in Urine 2 8 1 1 1 1 2 5 5 1 1 6 3 22 1 1 4 1 7 1 6 1 4 2 10 3 11 8 1 1 1 1 4 3 2 1 2 1 1 13 1 9 6 6 2 1 10 8 Number of Findings in Vitreous 1 8 0 1 1 1 1 5 5 1 0 6 0 14 0 1 0 1 3 1 4 0 4 2 3 1 0 0 1 0 1 2 3 1 1 1 0 0 1 0 0 1 6 6 3 1 8 4

Compound Alprazolam 7-Aminoclonazepam 7-Aminonitrazepam Amitriptyline Amlodipine Amphetamine Atenolol Atropine Atropine-N-oxide Bisoprolol Caffeine Carbamazepine Carbamazepine 10-11-epoxide* Chlordiazepoxide Chloroquine Citalopram Codeine Cotinine Cyclizine Demoxepam O-Desmethylgalantamine* O-Desmethylnortramadol O-Desmethyltramadol O-Desmethylvenlafaxine Diazepam Diltiazem 10-11-Dihydroxycarbamazepine* Dinorcitalopram* Dinortramadol Dinortriptyline Dinorvenlafaxine* Dipyridamole EDDP (methadone metabolite) Enalapril Ephedrine/pseudoephedrine Gabapentin Galantamine (E)-10-Hydroxyamitriptyline (Z)-10-Hydroxyamitriptyline (E)-10-Hydroxynortriptyline (Z)-10-Hydroxynortriptyline Hydroxychloroquine Hydroxycotinine 1-Hydroxymidazolam 4-Hydroxypropranolol* Iminostilbene* Ketamine Lamotrigine Levetiracetam

Compound Levomepromazine Lidocaine Lorazepam Meprobamate Methadone 4-Methylaminophenazone Metoclopramide Metoprolol Metoprolol metabolite III* Metronidazol Midazolam Mirtazapine Morphine Nicotine Noratropine* Norcitalopram Norcodeine Norcyclizine* Nordiazepam Norlevomepromazine Normirtazapine* Norolanzapine* Nortramadol Nortriptyline Norzopiclone* Olanzapine Oxazepam Paracetamol Paroxetine Paroxetine metabolite I/II* Pentoxiverine Phenytoin Propranolol Quetiapine Quinine Risperidone Salbutamol Sertraline Sulpiride Temazepam Theobromine Theophylline Tramadol Warfarin Venlafaxine Zolpidem Zopiclone Zopiclone-N-oxide*

* Tentative identification based on accurate mass, isotopic pattern, and occurrence in combination with the parent drug.

314

Journal of Analytical Toxicology, Vol. 34, July/August 2010

Figure 1. LCTOFMS total ion chromatogram of a urine extract (upper) and vitreous humor extract (lower) from the same individual. Internal standard (dibenzepin) elutes at 10.13 min.

Matrix effect

Table II. Cut-Off Values of 70 Drugs in Vitreous Humor by LCTOFMS

Compound Acebutolol Amitriptyline Amlodipine Amphetamine Atenolol Benzoylecgonine Betaxolol Bisoprolol Buprenorphine Carbamazepine Carvedilol Celiprolol Chloroquine Citalopram Clomipramine Clonidine Clozapine Cocaine Codeine Dextropropoxyphene Diazepam Diltiazem Dixyrazine Doxepin Ethylmorphine Fentanyl Flunitrazepam Fluoxetine Fluvoxamine Glipizide Indomethacin Ketoprofen LSD MDMA (Ecstasy) Melperone Cut-off (mg/L) 0.02 0.02 0.5 0.05 0.04 0.25 0.01 0.015 0.1 0.009 0.09 0.025 0.1 0.013 0.05 0.015 0.035 0.01 0.015 0.02 0.01 0.03 0.2 0.015 0.02 0.02 0.02 0.05 0.05 0.5 0.8 0.04 0.025 0.03 0.01 Compound Methadone Methamphetamine Metoprolol Mianserine Midazolam Mirtazapine Moclobemide 6-Monoacetylmorphine Morphine Nortriptyline Olanzapine Orphenadrine Oxprenolol Oxycodone Paroxetine Phenazone Phencyclidine Phenytoin Propranolol Quinine Ranitidine Risperidone Selegiline Sildenafil Sotalol Sulpiride Timolol Tizanidine Tramadol Trimipramine Venlafaxine Verapamil Warfarin Zolpidem Zopiclone Cut-off (mg/L) 0.01 0.02 0.01 0.015 0.025 0.01 0.015 0.04 0.025 0.03 0.05 0.01 0.01 0.025 0.07 0.0075 0.015 0.25 0.008 0.03 0.03 0.05 0.4 0.35 0.06 0.015 0.015 0.01 0.008 0.02 0.007 0.08 0.02 0.01 0.07

Matrix effect was studied by postcolumn infusion as described earlier by Dams et al. (23). Seven compounds presenting varying retention behavior and polarities were selected for the study: metformine, nicotine, morphine, oxycodone, moclobemide, propranolol, and clomipramine. Early eluting compounds were emphasized based on earlier experience with urine matrix effects. The analyte in a concentration of 10 g/mL in methanol/acetonitrile (1:1, v/v) was infused post-column via a T-piece with a flow rate of 5 L/min. Single-analyte solutions were used. The analysis was performed by comparing the EICs of the analyte in question in two injections: the first was performed against pure mobile phase and the second against extracted pooled blank vitreous humor. The matrix effect was expressed as the ion suppression or enhancement presented as a signal intensity difference in percentage at the elution slot of the analyte.

Results and Discussion

Results from the LCTOFMS screening analysis of urine and vitreous samples of 50 authentic cases are combined in Table I. The total number of findings in urine and vitreous were 376 and 245, respectively. The total number of compounds identified was 97, of which 39 were

315

Journal of Analytical Toxicology, Vol. 34, July/August 2010

metabolites. In urine, 55 parent compounds and 39 metabolites were identified. In vitreous, the figures were 45 and 24, respectively. Temazepam, oxazepam, nortriptyline, theophylline, and theobromine were considered parent compounds in these figures. The number of identified compounds was lower in vitreous than in urine. In addition, the proportion of identified parent compounds was higher in vitreous than in urine. Both results are to be expected: drug concentrations are in general higher in urine than in vitreous, resulting in a higher number of findings in urine, and metabolites are present predominantly in urine. Duer et al. (8) and Favretto et al. (11) have shown this for cocaine and LSD together with their metabolites, although systematic multimatrix studies are rare. The results show that vitreous does not outperform urine as a sample material for qualitative screening analysis, but it is a very good secondary choice in cases where urine is not available. Compared to the liver, where basic lipophilic compounds are mostly detectable (18), the current results show a much wider range of chemistries and polarities. However, some of the more polar analytes appear to be poorly detectable, most likely because of the blood-retinal barrier, as paracetamol, oxazepam, and temazepam were not detected in vitreous. Chloroquine, hydroxychloroquine, and lamotrigine were detected once in vitreous but not in the corresponding urine sample. In these cases, the blood sample was positive for the particular compound (data not shown). A noteworthy feature of vitreous was its low background noise compared to urine extracts. This reduces both instrument contamination and severe matrix effects. An example of urine-vitreous background difference is presented in Figure 1. Table II shows a cutoff value for 70 drugs in vitreous humor. The cutoff value was not the limit of detection (LOD) by definition but a reporting criterion for the area of a detected peak. A cutoff value was used instead of LOD because fullrange data acquisition from a complex matrix often results in an excessive number of false positives if a traditional LOD definition is applied. The cutoff values varied widely from 0.007 to 0.8 mg/L with the mean and median

Table III. Results From the Matrix Effect Study by Post-Column Infusion Against Mobile Phase and Blank Vitreous Extract
Retention Time (min) 0.82 1.02 1.37 3.15 5.97 10.33 13.90 Observed Time Window (min) 0.70.9 0.91.1 1.31.5 3.03.2 5.96.1 10.210.4 13.814.0 Ion Suppression/Enhancement (%) 15 +8 18 Not observable +12 3 5

Compound Metformin Nicotine Morphine Oxycodone Moclobemide Propranolol Clomipramine

Figure 2. Matrix effect studied by post-column infusion against mobile phase and blank vitreous extract for oxycodone (A), morphine (B), and nicotine (C). Extracted ion chromatograms for both injections of the analyte in question are superimposed. Oxycodone eluted at 3.15 min, morphine at 1.37 min, and nicotine at 1.02 min.

316

Journal of Analytical Toxicology, Vol. 34, July/August 2010

being 0.072 and 0.023 mg/L, respectively. However, only 10 compounds had a cutoff value higher than 0.05 mg/L, and the median value was closer to the lower end of the range. The cutoff values defined in the present study for vitreous humor were similar to the LODs determined for urine in an earlier study (19). Holmgren et al. (24) have studied the stability of drugs in vitreous and reported mean concentrations for 46 drugs in 469 cases. The concentrations varied from undetected (five compounds positive in blood but not in vitreous) to 10.9 g/g; the overall mean and median were 1.03 and 0.15 g/g, respectively. The cases reflected a normal casework panorama, thus giving a realistic idea of concentrations in authentic samples. In this context, the cutoff values achieved in the present study appear adequate in general. Mackey-Bojack et al. (7) studied cocaine and its metabolites in vitreous and reported median concentrations of 0.61 and 0.99 mg/L for cocaine and benzoylecgonine, respectively, in 62 authentic cases. In another study on 26 cocaine poisoning cases, Fernandez et al. (9) reported corresponding mean values of 1.54 and 0.94 mg/L, and the cutoff values achieved in the present study were 0.010 and 0.25 mg/L. The benzoylecgonine cutoff of 0.25 mg/L was one of the highest in our study; however, even that seems adequate when compared to the values reported in authentic casework. Jenkins et al. (12) reported mean and median phencyclidine concentrations of 0.14 and 0.1 mg/L in 27 samples, which is very well in line with the cutoff of 0.015 mg/L of the present study. Cutoffs for 6-monoacetylmorphine (6-MAM) and morphine were 0.04 and 0.025 mg/L in our study. According to results by Pragst et al. (25), our morphine level is adequate (mean and median of 0.15 and 0.13 in 26 cases), but the level for 6-MAM is clearly insufficient (mean and median of 0.0175 and 0.013 in 21 cases). However, in suspected heroin poisonings, a target analysis for 6-MAM in blood should always be performed anyway. Thus, the inadequate cutoff in vitreous does not cause any major drawbacks. The buprenorphine cutoff of 0.1 mg/L was high, and it is obvious that the method is unfit for buprenorphine screening. Buprenorphine would require a higher capillary exit voltage, and a complimentary screening method is recommended. Matrix effects were studied by post-column infusion of seven compounds presenting varying polarities and retention (Table III). Both suppression and enhancement was observed. Significant suppression occurred at 0.71.3 min and was 18% at its highest for morphine. This is moderate compared, for instance, to the values reported by Dams and co-workers (23) in their multimatrix study and is considered to be acceptable for a screening method. Figure 2 shows superimposed EICs for oxycodone (no suppression or enhancement), morphine (18% suppression), and nicotine (8% enhancement). The oxycodone and morphine profiles look fairly identical. However, the nicotine profile looks quite different at the beginning of the run, and indeed, some enhancement is observed. In a screening method for several hundred compounds, full validation of matrix effects is at least very complicated, if possible at all, and the presented approach gives only a general view of the phenomenon. In addition, the matrix effect variation in separate samples was evaluated by calculating the relative standard de-

viation (RSD) for internal standard (ISTD) peak area in 5 authentic sample series including 10 samples each. RSD varied from 18 to 33% with the average being 27%. This is a high but realistic deviation, and it is emphasized that the calculations are based on absolute areas rather than relative values typically used in quantitative target analysis evaluations. The instrument performance was evaluated in terms of mass accuracy and mass resolution by analyzing the data in three 15sample series collected during a three-month period with onemonth intervals. The average mass accuracy for 242 identifications of 75 compounds was 2.6 ppm or 0.64 mDa. This fulfills the 5 ppm rule for molecular formula assignment. The average mass resolution for the ISTD was 14,965, which is clearly above the instrument specification nominal resolution of 10,000 and is fairly good for a TOF instrument. However, this means that co-elution of a compound with a mass difference of 19.8 mDa or lower would potentially result in a false-negative result due to an increased mass error as a result of sum peak formation. Therefore, potential for a false-negative result due to limited mass resolution exists. A decent way to monitor and control this phenomenon is to follow a routine of two matrices analyzed by two independent analytical techniques, for instance blood by GC-based screening and urine/vitreous humor by TOFMS screening. In addition, we emphasize the role of selective sample preparation; with less selective sample preparation techniques, such as protein precipitation and liquidliquid extraction, the potential for low-resolution disturbance is increased. Detailed information about fundamentals and applications of accurate mass and TOFMS technology is available elsewhere (26).

Conclusions
Vitreous humor proved to be a feasible, easy-to-handle, and representative matrix for qualitative screening analysis. The representativeness in terms of drug and metabolite variability was lower than in urine but adequate for screening purposes. One of the advantages in postmortem work compared to clinical setting is the versatility of sample matrices available. We strongly encourage forensic toxicology laboratories to utilize this by adapting vitreous as a potential option for qualitative analysis. The cutoff values determined gave a realistic picture of the performance of the method, and although quantitative information about vitreous drug concentrations is limited in the literature, it can be concluded that the presented method is adequate for routine casework.

References
1. R.O. Bost. Analytical toxicology of vitreous humor. In Handbook of Analytical Therapeutic Drug Monitoring and Toxicology, S.H.Y. Wong and I. Sunshine, Eds. CRC Press, Boca Raton, FL, 1997, pp 281302. 2. J.G. Cunha-Vaz. The blood-retinal barriers system. Basic concepts and clinical evaluation. Exp. Eye Res. 78: 715721 (2004).

317

Journal of Analytical Toxicology, Vol. 34, July/August 2010

3. K.M. Clauwaert, J.F. Van Bocxlaer, E.A. De Letter, S. Van Calenbergh, W.E. Lambert, and A.P. De Leenheer. Determination of the designer drugs 3,4-methylenedioxymethamphetamine, 3,4-methylenedioxyethylamphetamine, and 3,4-methylenedioxyamphetamine with HPLC and fluorescence detection in whole blood, serum, vitreous humor, and urine. Clin. Chem. 46: 1968 1977 (2000). 4. E.A. De Letter, P. De Paepe, K.M. Clauwaert, F.M. Belpaire, W.E. Lambert, J.F. Van Bocxlaer, and M.H. Piette. Is vitreous humour useful for the interpretation of 3,4-methylenedioxymethamphetamine (MDMA) blood levels? Experimental approach with rabbits. Int. J. Legal Med. 114: 2935 (2000). 5. K.S. Scott and J.S. Oliver. The use of vitreous humor as an alternative to whole blood for the analysis of benzodiazepines. J. Forensic Sci. 46: 694697 (2001). 6. H.M. Teixeira, F. Reis, P. Proena, P. Ramos, O. Quintela, M. Lpez-Rivadulla, E. Marques, and D.N. Vieira. Vitreous humour as a complementary sample to blood for the detection/confirmation of diazepam: ante-mortem and post-mortem studies in an animal model. Hum. Exp. Toxicol. 23: 571577 (2004). 7. S. Mackey-Bojack, J. Kloss, and F. Apple. Cocaine, cocaine metabolite, and ethanol concentrations in postmortem blood and vitreous humor. J. Anal. Toxicol. 24: 5965 (2000). 8. W.C. Duer, D.J. Spitz, and S. McFarland. Relationships between concentrations of cocaine and its hydrolysates in peripheral blood, heart blood, vitreous humor and urine. J. Forensic Sci. 51: 421425 (2006). 9. P. Fernndez, M. Aldonza, A. Bouzas, M. Lema, A.M. Bermejo, and M.J. Tabernero. GCFID determination of cocaine and its metabolites in human bile and vitreous humor. J. Appl. Toxicol. 26: 253257 (2006). 10. H.M. Antonides, E.R. Kiely, and L.J. Marinetti. Vitreous fluid quantification of opiates, cocaine, and benzoylecgonine: comparison of calibration curves in both blood and vitreous matrices with corresponding concentrations in blood. J. Anal. Toxicol. 31: 469 476 (2007). 11. D. Favretto, G. Frison, S. Maietti, and S.D. Ferrara. LCESI-MS/MS on an ion trap for the determination of LSD, iso-LSD, nor-LSD and 2-oxo-3-hydroxy-LSD in blood, urine and vitreous humor. Int. J. Legal Med. 121: 259265 (2007). 12. A.J. Jenkins and J. Oblock. Phencyclidine and cannabinoids in vitreous humor. Leg. Med. (Tokyo) 10: 201203 (2008). 13. D. Cox, R.A. Jufer Phipps, B. Levine, A. Jacobs, and D. Fowler. Distribution of phencyclidine into vitreous humor. J. Anal. Toxicol. 31: 537539 (2007). 14. P. Kintz, M. Villain, V. Cirimele, and B. Ludes. GHB in postmortem toxicology. Discrimination between endogenous production from exposure using multiple specimens. Forensic Sci. Int. 143: 177181 (2004).

15. I. Ojanper, E. Vuori, R. Nieminen, and A. Penttil. Screening for barbiturates in vitreous humor by the EMIT-st serum enzyme immunoassay. J. Forensic Sci. 31: 707709 (1986). 16. N. Fucci, N. De Giovanni, F. De Giorgio, R. Liddi, and M. Chiarotti. An evaluation of the Cozart RapiScan system as an on-site screening tool for drugs of abuse in a non-conventional biological matrix: vitreous humor. Forensic Sci. Int. 156: 102105 (2006). 17. C.W. Chronister, A.L. Gund, and B.A. Goldberger. Rapid detection of opioids in vitreous humor by enzyme immunoassay. J. Anal. Toxicol. 32: 601604 (2008). 18. I. Ojanper, R.-L. Ojansivu, J. Nokua, and E. Vuori. Comprehensive TLC drug screening in forensic toxicology: comparison of findings in urine and liver. J. Planar Chromatogr. Modern TLC 12: 3841 (1999). 19. A. Pelander, I. Ojanper, S. Laks, I. Rasanen, and E. Vuori. Toxicological screening with formula-based metabolite identification by liquid chromatography/time-of-flight mass spectrometry. Anal. Chem. 75: 57105718 (2003). 20. S. Ojanper, A. Pelander, M. Pelzing, I. Krebs, E. Vuori, and I. Ojanper. Isotopic pattern and accurate mass determination in urine drug screening by liquid chromatography/time-of-flight mass spectrometry. Rapid Commun. Mass Spectrom. 20: 1161 1167 (2006). 21. M. Kolmonen, A. Leinonen, A. Pelander, and I. Ojanper. A general screening method for doping agents in human urine by solid phase extraction and liquid chromatography/time-of-flight mass spectrometry. Anal. Chim. Acta 585: 94102 (2007). 22. A. Pelander, J. Ristimaa, I. Rasanen, E. Vuori, and I. Ojanper. Screening for basic drugs in hair of drug addicts by liquid chromatography/time-of-flight mass spectrometry. Ther. Drug Monit. 30: 717724 (2008). 23. R. Dams, M.A. Huestis, W.E. Lambert, and C.M. Murphy. Matrix effect in bio-analysis of illicit drugs with LCMS/MS: influence of ionization type, sample preparation, and biofluid. J. Am. Soc. Mass Spectrom. 14: 12901294 (2003). 24. P. Holmgren, H. Druid, A. Holmgren, and J. Ahlner. Stability of drugs in stored postmortem femoral blood and vitreous humor. J. Forensic Sci. 49: 820825 (2004). 25. F. Pragst, K. Spiegel, U. Leuschner, and A. Hager. Detection of 6-acetylmorphine in vitreous humor and cerebrospinal fluid comparison with urinary analysis for proving heroin administration in opiate fatalities. J. Anal. Toxicol. 23: 168172 (1999). 26. Liquid Chromatography Time-of-Flight Mass Spectrometry , I. Ferrer and E.M. Thurman, Eds. Wiley, Hobogen, NJ, 2009. Manuscript received November 30, 2009; revision received January 26, 2010.

318